id	sid	tid	token	lemma	pos
aiti-8904	1	1	2___aiti#8904___169	2___aiti#8904___169	NUM
aiti-8904	1	2	-	-	SYM
aiti-8904	1	3	180	180	NUM
aiti-8904	1	4	advances	advance	NOUN
aiti-8904	1	5	in	in	ADP
aiti-8904	1	6	technology	technology	NOUN
aiti-8904	1	7	innovation	innovation	NOUN
aiti-8904	1	8	,	,	PUNCT
aiti-8904	1	9	vol	vol	NOUN
aiti-8904	1	10	.	.	PROPN
aiti-8904	1	11	7	7	NUM
aiti-8904	1	12	,	,	PUNCT
aiti-8904	1	13	no	no	INTJ
aiti-8904	1	14	.	.	NOUN
aiti-8904	1	15	3	3	NUM
aiti-8904	1	16	,	,	PUNCT
aiti-8904	1	17	2022	2022	NUM
aiti-8904	1	18	,	,	PUNCT
aiti-8904	1	19	pp	pp	ADJ
aiti-8904	1	20	.	.	PUNCT
aiti-8904	2	1	169	169	NUM
aiti-8904	2	2	-	-	SYM
aiti-8904	2	3	180	180	NUM
aiti-8904	2	4	developing	develop	VERB
aiti-8904	2	5	and	and	CCONJ
aiti-8904	2	6	implementing	implement	VERB
aiti-8904	2	7	an	an	DET
aiti-8904	2	8	ai	ai	NOUN
aiti-8904	2	9	-	-	PUNCT
aiti-8904	2	10	based	base	VERB
aiti-8904	2	11	leak	leak	NOUN
aiti-8904	2	12	detection	detection	NOUN
aiti-8904	2	13	system	system	NOUN
aiti-8904	2	14	in	in	ADP
aiti-8904	2	15	a	a	DET
aiti-8904	2	16	long	long	ADJ
aiti-8904	2	17	-	-	PUNCT
aiti-8904	2	18	distance	distance	NOUN
aiti-8904	2	19	gas	gas	NOUN
aiti-8904	2	20	pipeline	pipeline	NOUN
aiti-8904	2	21	te	te	PROPN
aiti-8904	2	22	-	-	PUNCT
aiti-8904	2	23	kwei	kwei	PROPN
aiti-8904	2	24	wang	wang	PROPN
aiti-8904	2	25	1	1	NUM
aiti-8904	2	26	,	,	PUNCT
aiti-8904	2	27	yu	yu	PROPN
aiti-8904	2	28	-	-	PROPN
aiti-8904	2	29	hsun	hsun	PROPN
aiti-8904	2	30	lin	lin	PROPN
aiti-8904	2	31	2	2	NUM
aiti-8904	2	32	,	,	PUNCT
aiti-8904	2	33	*	*	PUNCT
aiti-8904	2	34	,	,	PUNCT
aiti-8904	2	35	jian	jian	PROPN
aiti-8904	2	36	-	-	PUNCT
aiti-8904	2	37	yuan	yuan	ADJ
aiti-8904	2	38	shen	shen	NOUN
aiti-8904	2	39	1	1	NUM
aiti-8904	2	40	1	1	NUM
aiti-8904	2	41	department	department	NOUN
aiti-8904	2	42	of	of	ADP
aiti-8904	2	43	electrical	electrical	ADJ
aiti-8904	2	44	engineering	engineering	NOUN
aiti-8904	2	45	,	,	PUNCT
aiti-8904	2	46	ming	ming	PROPN
aiti-8904	2	47	chi	chi	PROPN
aiti-8904	2	48	university	university	PROPN
aiti-8904	2	49	of	of	ADP
aiti-8904	2	50	technology	technology	NOUN
aiti-8904	2	51	,	,	PUNCT
aiti-8904	2	52	new	new	ADJ
aiti-8904	2	53	taipei	taipei	PROPN
aiti-8904	2	54	city	city	PROPN
aiti-8904	2	55	,	,	PUNCT
aiti-8904	2	56	taiwan	taiwan	PROPN
aiti-8904	2	57	2	2	NUM
aiti-8904	2	58	department	department	NOUN
aiti-8904	2	59	of	of	ADP
aiti-8904	2	60	business	business	NOUN
aiti-8904	2	61	and	and	CCONJ
aiti-8904	2	62	management	management	NOUN
aiti-8904	2	63	,	,	PUNCT
aiti-8904	2	64	ming	ming	PROPN
aiti-8904	2	65	chi	chi	PROPN
aiti-8904	2	66	university	university	PROPN
aiti-8904	2	67	of	of	ADP
aiti-8904	2	68	technology	technology	NOUN
aiti-8904	2	69	,	,	PUNCT
aiti-8904	2	70	new	new	ADJ
aiti-8904	2	71	taipei	taipei	PROPN
aiti-8904	2	72	city	city	PROPN
aiti-8904	2	73	,	,	PUNCT
aiti-8904	2	74	taiwan	taiwan	PROPN
aiti-8904	2	75	received	receive	VERB
aiti-8904	2	76	12	12	NUM
aiti-8904	2	77	november	november	PROPN
aiti-8904	2	78	2021	2021	NUM
aiti-8904	2	79	;	;	PUNCT
aiti-8904	2	80	received	receive	VERB
aiti-8904	2	81	in	in	ADP
aiti-8904	2	82	revised	revise	VERB
aiti-8904	2	83	form	form	NOUN
aiti-8904	2	84	13	13	NUM
aiti-8904	2	85	april	april	PROPN
aiti-8904	2	86	2022	2022	NUM
aiti-8904	2	87	;	;	PUNCT
aiti-8904	2	88	accepted	accept	VERB
aiti-8904	2	89	14	14	NUM
aiti-8904	2	90	april	april	PROPN
aiti-8904	2	91	2022	2022	NUM
aiti-8904	2	92	doi	doi	NOUN
aiti-8904	2	93	:	:	PUNCT
aiti-8904	2	94	https://doi.org/10.46604/aiti.2022.8904	https://doi.org/10.46604/aiti.2022.8904	PROPN
aiti-8904	2	95	abstract	abstract	ADV
aiti-8904	2	96	this	this	DET
aiti-8904	2	97	research	research	NOUN
aiti-8904	2	98	proposes	propose	VERB
aiti-8904	2	99	an	an	DET
aiti-8904	2	100	artificial	artificial	ADJ
aiti-8904	2	101	intelligence	intelligence	NOUN
aiti-8904	2	102	(	(	PUNCT
aiti-8904	2	103	ai	ai	NOUN
aiti-8904	2	104	)	)	PUNCT
aiti-8904	2	105	detection	detection	NOUN
aiti-8904	2	106	model	model	NOUN
aiti-8904	2	107	using	use	VERB
aiti-8904	2	108	convolutional	convolutional	ADJ
aiti-8904	2	109	neural	neural	ADJ
aiti-8904	2	110	networks	network	NOUN
aiti-8904	2	111	(	(	PUNCT
aiti-8904	2	112	cnn	cnn	PROPN
aiti-8904	2	113	)	)	PUNCT
aiti-8904	2	114	to	to	PART
aiti-8904	2	115	automatically	automatically	ADV
aiti-8904	2	116	detect	detect	VERB
aiti-8904	2	117	gas	gas	NOUN
aiti-8904	2	118	leaks	leak	NOUN
aiti-8904	2	119	in	in	ADP
aiti-8904	2	120	a	a	DET
aiti-8904	2	121	long	long	ADJ
aiti-8904	2	122	-	-	PUNCT
aiti-8904	2	123	distance	distance	NOUN
aiti-8904	2	124	pipeline	pipeline	NOUN
aiti-8904	2	125	.	.	PUNCT
aiti-8904	3	1	the	the	DET
aiti-8904	3	2	change	change	NOUN
aiti-8904	3	3	of	of	ADP
aiti-8904	3	4	gap	gap	NOUN
aiti-8904	3	5	pressure	pressure	NOUN
aiti-8904	3	6	is	be	AUX
aiti-8904	3	7	collected	collect	VERB
aiti-8904	3	8	when	when	SCONJ
aiti-8904	3	9	leakage	leakage	NOUN
aiti-8904	3	10	occurs	occur	VERB
aiti-8904	3	11	in	in	ADP
aiti-8904	3	12	the	the	DET
aiti-8904	3	13	pipeline	pipeline	NOUN
aiti-8904	3	14	,	,	PUNCT
aiti-8904	3	15	and	and	CCONJ
aiti-8904	3	16	thereby	thereby	ADV
aiti-8904	3	17	the	the	DET
aiti-8904	3	18	feature	feature	NOUN
aiti-8904	3	19	of	of	ADP
aiti-8904	3	20	gas	gas	NOUN
aiti-8904	3	21	leakage	leakage	NOUN
aiti-8904	3	22	is	be	AUX
aiti-8904	3	23	extracted	extract	VERB
aiti-8904	3	24	for	for	ADP
aiti-8904	3	25	building	build	VERB
aiti-8904	3	26	the	the	DET
aiti-8904	3	27	cnn	cnn	PROPN
aiti-8904	3	28	model	model	NOUN
aiti-8904	3	29	.	.	PUNCT
aiti-8904	4	1	the	the	DET
aiti-8904	4	2	gas	gas	NOUN
aiti-8904	4	3	leak	leak	NOUN
aiti-8904	4	4	patterns	pattern	NOUN
aiti-8904	4	5	in	in	ADP
aiti-8904	4	6	the	the	DET
aiti-8904	4	7	long	long	ADJ
aiti-8904	4	8	-	-	PUNCT
aiti-8904	4	9	distance	distance	NOUN
aiti-8904	4	10	pipeline	pipeline	NOUN
aiti-8904	4	11	are	be	AUX
aiti-8904	4	12	analyzed	analyze	VERB
aiti-8904	4	13	.	.	PUNCT
aiti-8904	5	1	a	a	DET
aiti-8904	5	2	pipeline	pipeline	NOUN
aiti-8904	5	3	detection	detection	NOUN
aiti-8904	5	4	model	model	NOUN
aiti-8904	5	5	based	base	VERB
aiti-8904	5	6	on	on	ADP
aiti-8904	5	7	ai	ai	PRON
aiti-8904	5	8	technology	technology	NOUN
aiti-8904	5	9	for	for	ADP
aiti-8904	5	10	automatically	automatically	ADV
aiti-8904	5	11	monitoring	monitor	VERB
aiti-8904	5	12	the	the	DET
aiti-8904	5	13	leaks	leak	NOUN
aiti-8904	5	14	is	be	AUX
aiti-8904	5	15	proposed	propose	VERB
aiti-8904	5	16	by	by	ADP
aiti-8904	5	17	extracting	extract	VERB
aiti-8904	5	18	the	the	DET
aiti-8904	5	19	feature	feature	NOUN
aiti-8904	5	20	of	of	ADP
aiti-8904	5	21	gas	gas	NOUN
aiti-8904	5	22	leakage	leakage	NOUN
aiti-8904	5	23	.	.	PUNCT
aiti-8904	6	1	this	this	DET
aiti-8904	6	2	model	model	NOUN
aiti-8904	6	3	is	be	AUX
aiti-8904	6	4	tested	test	VERB
aiti-8904	6	5	by	by	ADP
aiti-8904	6	6	collecting	collect	VERB
aiti-8904	6	7	gas	gas	NOUN
aiti-8904	6	8	pressure	pressure	NOUN
aiti-8904	6	9	data	datum	NOUN
aiti-8904	6	10	from	from	ADP
aiti-8904	6	11	an	an	DET
aiti-8904	6	12	existing	exist	VERB
aiti-8904	6	13	natural	natural	ADJ
aiti-8904	6	14	gas	gas	NOUN
aiti-8904	6	15	pipeline	pipeline	NOUN
aiti-8904	6	16	system	system	NOUN
aiti-8904	6	17	starting	start	VERB
aiti-8904	6	18	from	from	ADP
aiti-8904	6	19	mailiao	mailiao	PROPN
aiti-8904	6	20	to	to	ADP
aiti-8904	6	21	taoyuan	taoyuan	PROPN
aiti-8904	6	22	in	in	ADP
aiti-8904	6	23	taiwan	taiwan	PROPN
aiti-8904	6	24	.	.	PUNCT
aiti-8904	7	1	the	the	DET
aiti-8904	7	2	testing	testing	NOUN
aiti-8904	7	3	result	result	NOUN
aiti-8904	7	4	shows	show	VERB
aiti-8904	7	5	that	that	SCONJ
aiti-8904	7	6	the	the	DET
aiti-8904	7	7	reduced	reduced	ADJ
aiti-8904	7	8	model	model	NOUN
aiti-8904	7	9	of	of	ADP
aiti-8904	7	10	leak	leak	NOUN
aiti-8904	7	11	detection	detection	NOUN
aiti-8904	7	12	can	can	AUX
aiti-8904	7	13	be	be	AUX
aiti-8904	7	14	used	use	VERB
aiti-8904	7	15	to	to	PART
aiti-8904	7	16	detect	detect	VERB
aiti-8904	7	17	the	the	DET
aiti-8904	7	18	leaks	leak	NOUN
aiti-8904	7	19	from	from	ADP
aiti-8904	7	20	the	the	DET
aiti-8904	7	21	upstream	upstream	ADJ
aiti-8904	7	22	and	and	CCONJ
aiti-8904	7	23	downstream	downstream	ADJ
aiti-8904	7	24	pipelines	pipeline	NOUN
aiti-8904	7	25	,	,	PUNCT
aiti-8904	7	26	and	and	CCONJ
aiti-8904	7	27	the	the	DET
aiti-8904	7	28	ai	ai	NOUN
aiti-8904	7	29	-	-	PUNCT
aiti-8904	7	30	based	base	VERB
aiti-8904	7	31	pipeline	pipeline	NOUN
aiti-8904	7	32	leak	leak	NOUN
aiti-8904	7	33	detection	detection	NOUN
aiti-8904	7	34	system	system	NOUN
aiti-8904	7	35	can	can	AUX
aiti-8904	7	36	obtain	obtain	VERB
aiti-8904	7	37	a	a	DET
aiti-8904	7	38	satisfactory	satisfactory	ADJ
aiti-8904	7	39	result	result	NOUN
aiti-8904	7	40	.	.	PUNCT
aiti-8904	8	1	keywords	keyword	NOUN
aiti-8904	8	2	:	:	PUNCT
aiti-8904	8	3	artificial	artificial	ADJ
aiti-8904	8	4	intelligence	intelligence	NOUN
aiti-8904	8	5	,	,	PUNCT
aiti-8904	8	6	convolutional	convolutional	ADJ
aiti-8904	8	7	neural	neural	ADJ
aiti-8904	8	8	network	network	NOUN
aiti-8904	8	9	,	,	PUNCT
aiti-8904	8	10	pipeline	pipeline	NOUN
aiti-8904	8	11	leak	leak	NOUN
aiti-8904	8	12	,	,	PUNCT
aiti-8904	8	13	leak	leak	VERB
aiti-8904	8	14	detection	detection	NOUN
aiti-8904	8	15	1	1	NUM
aiti-8904	8	16	.	.	PUNCT
aiti-8904	8	17	introduction	introduction	NOUN
aiti-8904	8	18	pipeline	pipeline	NOUN
aiti-8904	8	19	transportation	transportation	NOUN
aiti-8904	8	20	systems	system	NOUN
aiti-8904	8	21	are	be	AUX
aiti-8904	8	22	generally	generally	ADV
aiti-8904	8	23	used	use	VERB
aiti-8904	8	24	to	to	PART
aiti-8904	8	25	transport	transport	VERB
aiti-8904	8	26	oil	oil	NOUN
aiti-8904	8	27	,	,	PUNCT
aiti-8904	8	28	natural	natural	ADJ
aiti-8904	8	29	gas	gas	NOUN
aiti-8904	8	30	,	,	PUNCT
aiti-8904	8	31	domestic	domestic	ADJ
aiti-8904	8	32	water	water	NOUN
aiti-8904	8	33	,	,	PUNCT
aiti-8904	8	34	etc	etc	X
aiti-8904	8	35	.	.	X
aiti-8904	9	1	the	the	DET
aiti-8904	9	2	long	long	ADJ
aiti-8904	9	3	-	-	PUNCT
aiti-8904	9	4	distance	distance	NOUN
aiti-8904	9	5	transportation	transportation	NOUN
aiti-8904	9	6	system	system	NOUN
aiti-8904	9	7	may	may	AUX
aiti-8904	9	8	encounter	encounter	VERB
aiti-8904	9	9	leak	leak	VERB
aiti-8904	9	10	problems	problem	NOUN
aiti-8904	9	11	due	due	ADJ
aiti-8904	9	12	to	to	ADP
aiti-8904	9	13	old	old	ADJ
aiti-8904	9	14	pipelines	pipeline	NOUN
aiti-8904	9	15	.	.	PUNCT
aiti-8904	10	1	also	also	ADV
aiti-8904	10	2	,	,	PUNCT
aiti-8904	10	3	thieves	thief	NOUN
aiti-8904	10	4	may	may	AUX
aiti-8904	10	5	destroy	destroy	VERB
aiti-8904	10	6	pipelines	pipeline	NOUN
aiti-8904	10	7	and	and	CCONJ
aiti-8904	10	8	steal	steal	VERB
aiti-8904	10	9	the	the	DET
aiti-8904	10	10	resources	resource	NOUN
aiti-8904	10	11	from	from	ADP
aiti-8904	10	12	the	the	DET
aiti-8904	10	13	pipelines	pipeline	NOUN
aiti-8904	10	14	.	.	PUNCT
aiti-8904	11	1	thus	thus	ADV
aiti-8904	11	2	,	,	PUNCT
aiti-8904	11	3	leak	leak	NOUN
aiti-8904	11	4	detection	detection	NOUN
aiti-8904	11	5	is	be	AUX
aiti-8904	11	6	a	a	DET
aiti-8904	11	7	critical	critical	ADJ
aiti-8904	11	8	issue	issue	NOUN
aiti-8904	11	9	for	for	ADP
aiti-8904	11	10	effective	effective	ADJ
aiti-8904	11	11	pipeline	pipeline	NOUN
aiti-8904	11	12	management	management	NOUN
aiti-8904	11	13	.	.	PUNCT
aiti-8904	12	1	leak	leak	NOUN
aiti-8904	12	2	detection	detection	NOUN
aiti-8904	12	3	techniques	technique	NOUN
aiti-8904	12	4	based	base	VERB
aiti-8904	12	5	on	on	ADP
aiti-8904	12	6	acoustics	acoustic	NOUN
aiti-8904	12	7	and	and	CCONJ
aiti-8904	12	8	infrared	infrared	ADJ
aiti-8904	12	9	(	(	PUNCT
aiti-8904	12	10	ir	ir	NOUN
aiti-8904	12	11	)	)	PUNCT
aiti-8904	12	12	are	be	AUX
aiti-8904	12	13	the	the	DET
aiti-8904	12	14	most	most	ADV
aiti-8904	12	15	widely	widely	ADV
aiti-8904	12	16	used	use	VERB
aiti-8904	12	17	techniques	technique	NOUN
aiti-8904	12	18	for	for	ADP
aiti-8904	12	19	the	the	DET
aiti-8904	12	20	detection	detection	NOUN
aiti-8904	12	21	of	of	ADP
aiti-8904	12	22	liquid	liquid	ADJ
aiti-8904	12	23	and	and	CCONJ
aiti-8904	12	24	gas	gas	NOUN
aiti-8904	12	25	leaks	leak	NOUN
aiti-8904	12	26	,	,	PUNCT
aiti-8904	12	27	respectively	respectively	ADV
aiti-8904	12	28	.	.	PUNCT
aiti-8904	13	1	however	however	ADV
aiti-8904	13	2	,	,	PUNCT
aiti-8904	13	3	these	these	DET
aiti-8904	13	4	two	two	NUM
aiti-8904	13	5	techniques	technique	NOUN
aiti-8904	13	6	do	do	AUX
aiti-8904	13	7	not	not	PART
aiti-8904	13	8	work	work	VERB
aiti-8904	13	9	sometimes	sometimes	ADV
aiti-8904	13	10	[	[	X
aiti-8904	13	11	1	1	NUM
aiti-8904	13	12	]	]	PUNCT
aiti-8904	13	13	.	.	PUNCT
aiti-8904	14	1	for	for	ADP
aiti-8904	14	2	instance	instance	NOUN
aiti-8904	14	3	,	,	PUNCT
aiti-8904	14	4	ir	ir	PROPN
aiti-8904	14	5	approaches	approach	NOUN
aiti-8904	14	6	can	can	AUX
aiti-8904	14	7	not	not	PART
aiti-8904	14	8	accurately	accurately	ADV
aiti-8904	14	9	detect	detect	VERB
aiti-8904	14	10	leaks	leak	NOUN
aiti-8904	14	11	in	in	ADP
aiti-8904	14	12	wet	wet	ADJ
aiti-8904	14	13	weather	weather	NOUN
aiti-8904	14	14	,	,	PUNCT
aiti-8904	14	15	while	while	SCONJ
aiti-8904	14	16	acoustic	acoustic	ADJ
aiti-8904	14	17	sensors	sensor	NOUN
aiti-8904	14	18	may	may	AUX
aiti-8904	14	19	not	not	PART
aiti-8904	14	20	accurately	accurately	ADV
aiti-8904	14	21	detect	detect	VERB
aiti-8904	14	22	gas	gas	NOUN
aiti-8904	14	23	leaks	leak	NOUN
aiti-8904	14	24	in	in	ADP
aiti-8904	14	25	the	the	DET
aiti-8904	14	26	target	target	NOUN
aiti-8904	14	27	area	area	NOUN
aiti-8904	14	28	because	because	SCONJ
aiti-8904	14	29	of	of	ADP
aiti-8904	14	30	noise	noise	NOUN
aiti-8904	14	31	[	[	X
aiti-8904	14	32	1	1	NUM
aiti-8904	14	33	]	]	PUNCT
aiti-8904	14	34	.	.	PUNCT
aiti-8904	15	1	due	due	ADP
aiti-8904	15	2	to	to	ADP
aiti-8904	15	3	the	the	DET
aiti-8904	15	4	need	need	NOUN
aiti-8904	15	5	to	to	PART
aiti-8904	15	6	purchase	purchase	VERB
aiti-8904	15	7	and	and	CCONJ
aiti-8904	15	8	install	install	VERB
aiti-8904	15	9	new	new	ADJ
aiti-8904	15	10	sensors	sensor	NOUN
aiti-8904	15	11	,	,	PUNCT
aiti-8904	15	12	the	the	DET
aiti-8904	15	13	additional	additional	ADJ
aiti-8904	15	14	cost	cost	NOUN
aiti-8904	15	15	of	of	ADP
aiti-8904	15	16	buying	buy	VERB
aiti-8904	15	17	hardware	hardware	NOUN
aiti-8904	15	18	equipment	equipment	NOUN
aiti-8904	15	19	may	may	AUX
aiti-8904	15	20	be	be	AUX
aiti-8904	15	21	incurred	incur	VERB
aiti-8904	15	22	.	.	PUNCT
aiti-8904	16	1	in	in	ADP
aiti-8904	16	2	recent	recent	ADJ
aiti-8904	16	3	years	year	NOUN
aiti-8904	16	4	,	,	PUNCT
aiti-8904	16	5	many	many	ADJ
aiti-8904	16	6	methods	method	NOUN
aiti-8904	16	7	have	have	AUX
aiti-8904	16	8	been	be	AUX
aiti-8904	16	9	proposed	propose	VERB
aiti-8904	16	10	based	base	VERB
aiti-8904	16	11	on	on	ADP
aiti-8904	16	12	artificial	artificial	ADJ
aiti-8904	16	13	intelligence	intelligence	NOUN
aiti-8904	16	14	(	(	PUNCT
aiti-8904	16	15	ai	ai	NOUN
aiti-8904	16	16	)	)	PUNCT
aiti-8904	16	17	techniques	technique	NOUN
aiti-8904	16	18	for	for	ADP
aiti-8904	16	19	developing	develop	VERB
aiti-8904	16	20	leak	leak	NOUN
aiti-8904	16	21	detection	detection	NOUN
aiti-8904	16	22	systems	system	NOUN
aiti-8904	17	1	[	[	X
aiti-8904	17	2	2	2	NUM
aiti-8904	17	3	-	-	SYM
aiti-8904	17	4	3	3	NUM
aiti-8904	17	5	]	]	PUNCT
aiti-8904	17	6	.	.	PUNCT
aiti-8904	18	1	for	for	ADP
aiti-8904	18	2	example	example	NOUN
aiti-8904	18	3	,	,	PUNCT
aiti-8904	18	4	yang	yang	PROPN
aiti-8904	18	5	and	and	CCONJ
aiti-8904	18	6	zhao	zhao	PROPN
aiti-8904	19	1	[	[	X
aiti-8904	19	2	2	2	X
aiti-8904	19	3	]	]	PUNCT
aiti-8904	19	4	used	use	VERB
aiti-8904	19	5	an	an	DET
aiti-8904	19	6	optimally	optimally	ADV
aiti-8904	19	7	pruned	prune	VERB
aiti-8904	19	8	extreme	extreme	ADJ
aiti-8904	19	9	learning	learning	NOUN
aiti-8904	19	10	machine	machine	NOUN
aiti-8904	19	11	(	(	PUNCT
aiti-8904	19	12	opelm	opelm	NOUN
aiti-8904	19	13	)	)	PUNCT
aiti-8904	19	14	to	to	PART
aiti-8904	19	15	improve	improve	VERB
aiti-8904	19	16	the	the	DET
aiti-8904	19	17	accuracy	accuracy	NOUN
aiti-8904	19	18	of	of	ADP
aiti-8904	19	19	the	the	DET
aiti-8904	19	20	pressure	pressure	NOUN
aiti-8904	19	21	point	point	NOUN
aiti-8904	19	22	analysis	analysis	NOUN
aiti-8904	19	23	(	(	PUNCT
aiti-8904	19	24	ppa	ppa	PROPN
aiti-8904	19	25	)	)	PUNCT
aiti-8904	19	26	,	,	PUNCT
aiti-8904	19	27	and	and	CCONJ
aiti-8904	19	28	used	use	VERB
aiti-8904	19	29	bidirectional	bidirectional	ADJ
aiti-8904	19	30	long	long	ADJ
aiti-8904	19	31	-	-	PUNCT
aiti-8904	19	32	short	short	ADJ
aiti-8904	19	33	term	term	NOUN
aiti-8904	19	34	memory	memory	NOUN
aiti-8904	19	35	(	(	PUNCT
aiti-8904	19	36	bilstm	bilstm	NOUN
aiti-8904	19	37	)	)	PUNCT
aiti-8904	19	38	to	to	PART
aiti-8904	19	39	construct	construct	VERB
aiti-8904	19	40	a	a	DET
aiti-8904	19	41	leak	leak	NOUN
aiti-8904	19	42	detection	detection	NOUN
aiti-8904	19	43	system	system	NOUN
aiti-8904	19	44	[	[	X
aiti-8904	19	45	2	2	NUM
aiti-8904	19	46	]	]	PUNCT
aiti-8904	19	47	.	.	PUNCT
aiti-8904	20	1	zhou	zhou	PROPN
aiti-8904	20	2	et	et	PROPN
aiti-8904	20	3	al	al	PROPN
aiti-8904	20	4	.	.	PUNCT
aiti-8904	21	1	[	[	X
aiti-8904	21	2	3	3	X
aiti-8904	21	3	]	]	PUNCT
aiti-8904	21	4	used	use	VERB
aiti-8904	21	5	improved	improve	VERB
aiti-8904	21	6	spline	spline	ADJ
aiti-8904	21	7	-	-	PUNCT
aiti-8904	21	8	local	local	ADJ
aiti-8904	21	9	mean	mean	NOUN
aiti-8904	21	10	decomposition	decomposition	NOUN
aiti-8904	21	11	(	(	PUNCT
aiti-8904	21	12	islmd	islmd	NOUN
aiti-8904	21	13	)	)	PUNCT
aiti-8904	21	14	to	to	PART
aiti-8904	21	15	analyze	analyze	VERB
aiti-8904	21	16	the	the	DET
aiti-8904	21	17	internal	internal	ADJ
aiti-8904	21	18	pressure	pressure	NOUN
aiti-8904	21	19	value	value	NOUN
aiti-8904	21	20	of	of	ADP
aiti-8904	21	21	pipelines	pipeline	NOUN
aiti-8904	21	22	.	.	PUNCT
aiti-8904	22	1	they	they	PRON
aiti-8904	22	2	converted	convert	VERB
aiti-8904	22	3	the	the	DET
aiti-8904	22	4	data	datum	NOUN
aiti-8904	22	5	into	into	ADP
aiti-8904	22	6	an	an	DET
aiti-8904	22	7	image	image	NOUN
aiti-8904	22	8	,	,	PUNCT
aiti-8904	22	9	employed	employ	VERB
aiti-8904	22	10	alexnet	alexnet	NOUN
aiti-8904	22	11	to	to	PART
aiti-8904	22	12	build	build	VERB
aiti-8904	22	13	a	a	DET
aiti-8904	22	14	model	model	NOUN
aiti-8904	22	15	to	to	PART
aiti-8904	22	16	determine	determine	VERB
aiti-8904	22	17	the	the	DET
aiti-8904	22	18	occurrence	occurrence	NOUN
aiti-8904	22	19	of	of	ADP
aiti-8904	22	20	leakage	leakage	NOUN
aiti-8904	22	21	,	,	PUNCT
aiti-8904	22	22	and	and	CCONJ
aiti-8904	22	23	used	use	VERB
aiti-8904	22	24	a	a	DET
aiti-8904	22	25	cross	cross	ADJ
aiti-8904	22	26	-	-	ADJ
aiti-8904	22	27	correlation	correlation	ADJ
aiti-8904	22	28	function	function	NOUN
aiti-8904	22	29	to	to	PART
aiti-8904	22	30	calculate	calculate	VERB
aiti-8904	22	31	the	the	DET
aiti-8904	22	32	time	time	NOUN
aiti-8904	22	33	difference	difference	NOUN
aiti-8904	22	34	between	between	ADP
aiti-8904	22	35	upstream	upstream	ADJ
aiti-8904	22	36	and	and	CCONJ
aiti-8904	22	37	downstream	downstream	ADJ
aiti-8904	22	38	sensor	sensor	NOUN
aiti-8904	22	39	values	value	NOUN
aiti-8904	22	40	to	to	PART
aiti-8904	22	41	find	find	VERB
aiti-8904	22	42	the	the	DET
aiti-8904	22	43	location	location	NOUN
aiti-8904	22	44	of	of	ADP
aiti-8904	22	45	the	the	DET
aiti-8904	22	46	leakage	leakage	NOUN
aiti-8904	22	47	point	point	NOUN
aiti-8904	22	48	[	[	X
aiti-8904	22	49	3	3	NUM
aiti-8904	22	50	]	]	PUNCT
aiti-8904	22	51	.	.	PUNCT
aiti-8904	23	1	a	a	DET
aiti-8904	23	2	growing	grow	VERB
aiti-8904	23	3	number	number	NOUN
aiti-8904	23	4	of	of	ADP
aiti-8904	23	5	ai	ai	ADJ
aiti-8904	23	6	techniques	technique	NOUN
aiti-8904	23	7	for	for	ADP
aiti-8904	23	8	developing	develop	VERB
aiti-8904	23	9	leak	leak	NOUN
aiti-8904	23	10	detection	detection	NOUN
aiti-8904	23	11	systems	system	NOUN
aiti-8904	23	12	are	be	AUX
aiti-8904	23	13	based	base	VERB
aiti-8904	23	14	on	on	ADP
aiti-8904	23	15	convolutional	convolutional	ADJ
aiti-8904	23	16	neural	neural	ADJ
aiti-8904	23	17	networks	network	NOUN
aiti-8904	23	18	(	(	PUNCT
aiti-8904	23	19	cnns	cnns	PROPN
aiti-8904	23	20	)	)	PUNCT
aiti-8904	24	1	[	[	X
aiti-8904	24	2	3	3	NUM
aiti-8904	24	3	-	-	SYM
aiti-8904	24	4	5	5	NUM
aiti-8904	24	5	]	]	PUNCT
aiti-8904	24	6	.	.	PUNCT
aiti-8904	25	1	cnns	cnns	PROPN
aiti-8904	25	2	are	be	AUX
aiti-8904	25	3	based	base	VERB
aiti-8904	25	4	on	on	ADP
aiti-8904	25	5	neurons	neuron	NOUN
aiti-8904	25	6	arranged	arrange	VERB
aiti-8904	25	7	in	in	ADP
aiti-8904	25	8	layers	layer	NOUN
aiti-8904	25	9	and	and	CCONJ
aiti-8904	25	10	can	can	AUX
aiti-8904	25	11	therefore	therefore	ADV
aiti-8904	25	12	learn	learn	VERB
aiti-8904	25	13	hierarchical	hierarchical	ADJ
aiti-8904	25	14	representations	representation	NOUN
aiti-8904	25	15	;	;	PUNCT
aiti-8904	25	16	moreover	moreover	ADV
aiti-8904	25	17	,	,	PUNCT
aiti-8904	25	18	weights	weight	NOUN
aiti-8904	25	19	and	and	CCONJ
aiti-8904	25	20	biases	bias	NOUN
aiti-8904	25	21	link	link	VERB
aiti-8904	25	22	neurons	neuron	NOUN
aiti-8904	25	23	from	from	ADP
aiti-8904	25	24	one	one	NUM
aiti-8904	25	25	layer	layer	NOUN
aiti-8904	25	26	to	to	ADP
aiti-8904	25	27	the	the	DET
aiti-8904	25	28	next	next	ADJ
aiti-8904	25	29	[	[	X
aiti-8904	25	30	6	6	NUM
aiti-8904	25	31	]	]	PUNCT
aiti-8904	25	32	.	.	PUNCT
aiti-8904	26	1	the	the	DET
aiti-8904	26	2	first	first	ADJ
aiti-8904	26	3	layer	layer	NOUN
aiti-8904	26	4	acts	act	VERB
aiti-8904	26	5	as	as	ADP
aiti-8904	26	6	the	the	DET
aiti-8904	26	7	input	input	NOUN
aiti-8904	26	8	layer	layer	NOUN
aiti-8904	26	9	for	for	ADP
aiti-8904	26	10	receiving	receive	VERB
aiti-8904	26	11	the	the	DET
aiti-8904	26	12	*	*	PUNCT
aiti-8904	26	13	corresponding	correspond	VERB
aiti-8904	26	14	author	author	NOUN
aiti-8904	26	15	.	.	PUNCT
aiti-8904	27	1	e	e	X
aiti-8904	27	2	-	-	NOUN
aiti-8904	27	3	mail	mail	NOUN
aiti-8904	27	4	address	address	NOUN
aiti-8904	27	5	:	:	PUNCT
aiti-8904	27	6	yslin@mail.mcut.edu.tw	yslin@mail.mcut.edu.tw	PROPN
aiti-8904	27	7	tel	tel	PROPN
aiti-8904	27	8	.	.	PUNCT
aiti-8904	27	9	:	:	PUNCT
aiti-8904	28	1	+886	+886	NUM
aiti-8904	28	2	-	-	SYM
aiti-8904	28	3	2	2	NUM
aiti-8904	28	4	-	-	PUNCT
aiti-8904	28	5	29089899#3168	29089899#3168	NUM
aiti-8904	28	6	;	;	PUNCT
aiti-8904	28	7	fax	fax	NOUN
aiti-8904	28	8	:	:	PUNCT
aiti-8904	28	9	+886	+886	NUM
aiti-8904	28	10	-	-	PUNCT
aiti-8904	28	11	2	2	NUM
aiti-8904	28	12	-	-	PUNCT
aiti-8904	28	13	29084533	29084533	NUM
aiti-8904	28	14	advances	advance	NOUN
aiti-8904	28	15	in	in	ADP
aiti-8904	28	16	technology	technology	NOUN
aiti-8904	28	17	innovation	innovation	NOUN
aiti-8904	28	18	,	,	PUNCT
aiti-8904	28	19	vol	vol	NOUN
aiti-8904	28	20	.	.	PROPN
aiti-8904	29	1	7	7	NUM
aiti-8904	29	2	,	,	PUNCT
aiti-8904	29	3	no	no	INTJ
aiti-8904	29	4	.	.	NOUN
aiti-8904	29	5	3	3	NUM
aiti-8904	29	6	,	,	PUNCT
aiti-8904	29	7	2022	2022	NUM
aiti-8904	29	8	,	,	PUNCT
aiti-8904	29	9	pp	pp	ADJ
aiti-8904	29	10	.	.	PUNCT
aiti-8904	30	1	169	169	NUM
aiti-8904	30	2	-	-	SYM
aiti-8904	30	3	180	180	NUM
aiti-8904	30	4	transformed	transform	VERB
aiti-8904	30	5	vector	vector	NOUN
aiti-8904	30	6	from	from	ADP
aiti-8904	30	7	image	image	NOUN
aiti-8904	30	8	-	-	PUNCT
aiti-8904	30	9	based	base	VERB
aiti-8904	30	10	data	datum	NOUN
aiti-8904	30	11	,	,	PUNCT
aiti-8904	30	12	e.g.	e.g.	ADV
aiti-8904	30	13	,	,	PUNCT
aiti-8904	30	14	remote	remote	ADJ
aiti-8904	30	15	leak	leak	NOUN
aiti-8904	30	16	data	datum	NOUN
aiti-8904	30	17	.	.	PUNCT
aiti-8904	31	1	the	the	DET
aiti-8904	31	2	last	last	ADJ
aiti-8904	31	3	layer	layer	NOUN
aiti-8904	31	4	is	be	AUX
aiti-8904	31	5	the	the	DET
aiti-8904	31	6	output	output	NOUN
aiti-8904	31	7	,	,	PUNCT
aiti-8904	31	8	e.g.	e.g.	ADV
aiti-8904	31	9	,	,	PUNCT
aiti-8904	31	10	predicting	predict	VERB
aiti-8904	31	11	whether	whether	SCONJ
aiti-8904	31	12	a	a	DET
aiti-8904	31	13	pipeline	pipeline	NOUN
aiti-8904	31	14	leaks	leak	NOUN
aiti-8904	31	15	or	or	CCONJ
aiti-8904	31	16	not	not	PART
aiti-8904	31	17	.	.	PUNCT
aiti-8904	32	1	the	the	DET
aiti-8904	32	2	feature	feature	NOUN
aiti-8904	32	3	derived	derive	VERB
aiti-8904	32	4	from	from	ADP
aiti-8904	32	5	the	the	DET
aiti-8904	32	6	previous	previous	ADJ
aiti-8904	32	7	layer	layer	NOUN
aiti-8904	32	8	serves	serve	VERB
aiti-8904	32	9	as	as	ADP
aiti-8904	32	10	an	an	DET
aiti-8904	32	11	input	input	NOUN
aiti-8904	32	12	for	for	ADP
aiti-8904	32	13	the	the	DET
aiti-8904	32	14	following	follow	VERB
aiti-8904	32	15	layer	layer	NOUN
aiti-8904	32	16	.	.	PUNCT
aiti-8904	33	1	the	the	DET
aiti-8904	33	2	final	final	ADJ
aiti-8904	33	3	layer	layer	NOUN
aiti-8904	33	4	predicts	predict	VERB
aiti-8904	33	5	an	an	DET
aiti-8904	33	6	outcome	outcome	NOUN
aiti-8904	33	7	according	accord	VERB
aiti-8904	33	8	to	to	ADP
aiti-8904	33	9	the	the	DET
aiti-8904	33	10	detected	detect	VERB
aiti-8904	33	11	features	feature	NOUN
aiti-8904	33	12	.	.	PUNCT
aiti-8904	34	1	layers	layer	NOUN
aiti-8904	34	2	between	between	ADP
aiti-8904	34	3	the	the	DET
aiti-8904	34	4	first	first	ADJ
aiti-8904	34	5	and	and	CCONJ
aiti-8904	34	6	the	the	DET
aiti-8904	34	7	last	last	ADJ
aiti-8904	34	8	layer	layer	NOUN
aiti-8904	34	9	are	be	AUX
aiti-8904	34	10	hidden	hide	VERB
aiti-8904	34	11	layers	layer	NOUN
aiti-8904	34	12	that	that	PRON
aiti-8904	34	13	transform	transform	VERB
aiti-8904	34	14	the	the	DET
aiti-8904	34	15	feature	feature	NOUN
aiti-8904	34	16	values	value	NOUN
aiti-8904	34	17	from	from	ADP
aiti-8904	34	18	the	the	DET
aiti-8904	34	19	input	input	NOUN
aiti-8904	34	20	to	to	ADP
aiti-8904	34	21	the	the	DET
aiti-8904	34	22	output	output	NOUN
aiti-8904	34	23	.	.	PUNCT
aiti-8904	35	1	traditionally	traditionally	ADV
aiti-8904	35	2	,	,	PUNCT
aiti-8904	35	3	cnns	cnn	NOUN
aiti-8904	35	4	consist	consist	VERB
aiti-8904	35	5	of	of	ADP
aiti-8904	35	6	at	at	ADV
aiti-8904	35	7	least	least	ADV
aiti-8904	35	8	one	one	NUM
aiti-8904	35	9	convolutional	convolutional	ADJ
aiti-8904	35	10	layer	layer	NOUN
aiti-8904	35	11	as	as	ADP
aiti-8904	35	12	a	a	DET
aiti-8904	35	13	hidden	hidden	ADJ
aiti-8904	35	14	layer	layer	NOUN
aiti-8904	35	15	for	for	ADP
aiti-8904	35	16	exploiting	exploit	VERB
aiti-8904	35	17	spatial	spatial	ADJ
aiti-8904	35	18	patterns	pattern	NOUN
aiti-8904	35	19	[	[	X
aiti-8904	35	20	6	6	NUM
aiti-8904	35	21	]	]	PUNCT
aiti-8904	35	22	.	.	PUNCT
aiti-8904	36	1	melo	melo	PROPN
aiti-8904	37	1	[	[	X
aiti-8904	37	2	4	4	X
aiti-8904	37	3	]	]	PUNCT
aiti-8904	37	4	used	use	VERB
aiti-8904	37	5	gradient	gradient	NOUN
aiti-8904	37	6	-	-	PUNCT
aiti-8904	37	7	weighted	weight	VERB
aiti-8904	37	8	class	class	NOUN
aiti-8904	37	9	activation	activation	NOUN
aiti-8904	37	10	mapping	mapping	NOUN
aiti-8904	37	11	algorithm	algorithm	NOUN
aiti-8904	37	12	(	(	PUNCT
aiti-8904	37	13	grad	grad	NOUN
aiti-8904	37	14	-	-	PUNCT
aiti-8904	37	15	cam	cam	NOUN
aiti-8904	37	16	)	)	PUNCT
aiti-8904	37	17	and	and	CCONJ
aiti-8904	37	18	cnn	cnn	PROPN
aiti-8904	37	19	models	model	NOUN
aiti-8904	37	20	to	to	PART
aiti-8904	37	21	judge	judge	VERB
aiti-8904	37	22	whether	whether	SCONJ
aiti-8904	37	23	the	the	DET
aiti-8904	37	24	images	image	NOUN
aiti-8904	37	25	recorded	record	VERB
aiti-8904	37	26	by	by	ADP
aiti-8904	37	27	closed	closed	ADJ
aiti-8904	37	28	-	-	PUNCT
aiti-8904	37	29	circuit	circuit	NOUN
aiti-8904	37	30	television	television	NOUN
aiti-8904	37	31	(	(	PUNCT
aiti-8904	37	32	cctv	cctv	PROPN
aiti-8904	37	33	)	)	PUNCT
aiti-8904	37	34	have	have	VERB
aiti-8904	37	35	a	a	DET
aiti-8904	37	36	leak	leak	NOUN
aiti-8904	37	37	response	response	NOUN
aiti-8904	37	38	;	;	PUNCT
aiti-8904	37	39	the	the	DET
aiti-8904	37	40	experimental	experimental	ADJ
aiti-8904	37	41	results	result	NOUN
aiti-8904	37	42	reached	reach	VERB
aiti-8904	37	43	an	an	DET
aiti-8904	37	44	accuracy	accuracy	NOUN
aiti-8904	37	45	of	of	ADP
aiti-8904	37	46	99.78	99.78	NUM
aiti-8904	37	47	%	%	NOUN
aiti-8904	37	48	.	.	PUNCT
aiti-8904	38	1	li	li	PROPN
aiti-8904	38	2	et	et	PROPN
aiti-8904	38	3	al	al	PROPN
aiti-8904	38	4	.	.	PUNCT
aiti-8904	39	1	[	[	X
aiti-8904	39	2	5	5	NUM
aiti-8904	39	3	]	]	PUNCT
aiti-8904	39	4	proposed	propose	VERB
aiti-8904	39	5	a	a	DET
aiti-8904	39	6	method	method	NOUN
aiti-8904	39	7	for	for	ADP
aiti-8904	39	8	small	small	ADJ
aiti-8904	39	9	-	-	PUNCT
aiti-8904	39	10	scale	scale	NOUN
aiti-8904	39	11	natural	natural	ADJ
aiti-8904	39	12	gas	gas	NOUN
aiti-8904	39	13	pipeline	pipeline	NOUN
aiti-8904	39	14	leak	leak	NOUN
aiti-8904	39	15	detection	detection	NOUN
aiti-8904	39	16	.	.	PUNCT
aiti-8904	40	1	the	the	DET
aiti-8904	40	2	model	model	NOUN
aiti-8904	40	3	converts	convert	VERB
aiti-8904	40	4	the	the	DET
aiti-8904	40	5	value	value	NOUN
aiti-8904	40	6	of	of	ADP
aiti-8904	40	7	the	the	DET
aiti-8904	40	8	acoustic	acoustic	ADJ
aiti-8904	40	9	sensor	sensor	NOUN
aiti-8904	40	10	into	into	ADP
aiti-8904	40	11	the	the	DET
aiti-8904	40	12	frequency	frequency	NOUN
aiti-8904	40	13	domain	domain	NOUN
aiti-8904	40	14	and	and	CCONJ
aiti-8904	40	15	then	then	ADV
aiti-8904	40	16	uses	use	VERB
aiti-8904	40	17	a	a	DET
aiti-8904	40	18	one	one	NUM
aiti-8904	40	19	-	-	PUNCT
aiti-8904	40	20	dimensional	dimensional	ADJ
aiti-8904	40	21	cnn	cnn	NOUN
aiti-8904	40	22	to	to	PART
aiti-8904	40	23	train	train	VERB
aiti-8904	40	24	the	the	DET
aiti-8904	40	25	model	model	NOUN
aiti-8904	40	26	.	.	PUNCT
aiti-8904	41	1	the	the	DET
aiti-8904	41	2	final	final	ADJ
aiti-8904	41	3	trained	train	VERB
aiti-8904	41	4	model	model	NOUN
aiti-8904	41	5	has	have	VERB
aiti-8904	41	6	higher	high	ADJ
aiti-8904	41	7	accuracy	accuracy	NOUN
aiti-8904	41	8	than	than	ADP
aiti-8904	41	9	the	the	DET
aiti-8904	41	10	leakage	leakage	NOUN
aiti-8904	41	11	detection	detection	NOUN
aiti-8904	41	12	performance	performance	NOUN
aiti-8904	41	13	of	of	ADP
aiti-8904	41	14	the	the	DET
aiti-8904	41	15	traditional	traditional	ADJ
aiti-8904	41	16	two	two	NUM
aiti-8904	41	17	-	-	PUNCT
aiti-8904	41	18	dimensional	dimensional	ADJ
aiti-8904	41	19	cnn	cnn	NOUN
aiti-8904	41	20	model	model	NOUN
aiti-8904	41	21	[	[	X
aiti-8904	41	22	5	5	NUM
aiti-8904	41	23	]	]	PUNCT
aiti-8904	41	24	.	.	PUNCT
aiti-8904	42	1	kang	kang	PROPN
aiti-8904	42	2	et	et	PROPN
aiti-8904	42	3	al	al	PROPN
aiti-8904	42	4	.	.	PUNCT
aiti-8904	43	1	[	[	X
aiti-8904	43	2	7	7	X
aiti-8904	43	3	]	]	PUNCT
aiti-8904	43	4	used	use	VERB
aiti-8904	43	5	ensemble	ensemble	ADJ
aiti-8904	43	6	cnn	cnn	PROPN
aiti-8904	43	7	-	-	PUNCT
aiti-8904	43	8	svm	svm	PROPN
aiti-8904	43	9	for	for	ADP
aiti-8904	43	10	leak	leak	NOUN
aiti-8904	43	11	detection	detection	NOUN
aiti-8904	43	12	in	in	ADP
aiti-8904	43	13	a	a	DET
aiti-8904	43	14	water	water	NOUN
aiti-8904	43	15	distribution	distribution	NOUN
aiti-8904	43	16	system	system	NOUN
aiti-8904	43	17	,	,	PUNCT
aiti-8904	43	18	and	and	CCONJ
aiti-8904	43	19	used	use	VERB
aiti-8904	43	20	a	a	DET
aiti-8904	43	21	graph	graph	NOUN
aiti-8904	43	22	-	-	PUNCT
aiti-8904	43	23	based	base	VERB
aiti-8904	43	24	algorithm	algorithm	NOUN
aiti-8904	43	25	to	to	PART
aiti-8904	43	26	determine	determine	VERB
aiti-8904	43	27	the	the	DET
aiti-8904	43	28	leak	leak	NOUN
aiti-8904	43	29	location	location	NOUN
aiti-8904	43	30	according	accord	VERB
aiti-8904	43	31	to	to	ADP
aiti-8904	43	32	the	the	DET
aiti-8904	43	33	time	time	NOUN
aiti-8904	43	34	difference	difference	NOUN
aiti-8904	43	35	of	of	ADP
aiti-8904	43	36	sensor	sensor	NOUN
aiti-8904	43	37	value	value	NOUN
aiti-8904	43	38	changes	change	NOUN
aiti-8904	43	39	.	.	PUNCT
aiti-8904	44	1	the	the	DET
aiti-8904	44	2	experimental	experimental	ADJ
aiti-8904	44	3	results	result	NOUN
aiti-8904	44	4	can	can	AUX
aiti-8904	44	5	reach	reach	VERB
aiti-8904	44	6	99.3	99.3	NUM
aiti-8904	44	7	%	%	NOUN
aiti-8904	44	8	of	of	ADP
aiti-8904	44	9	the	the	DET
aiti-8904	44	10	leak	leak	NOUN
aiti-8904	44	11	detection	detection	NOUN
aiti-8904	44	12	accuracy	accuracy	NOUN
aiti-8904	44	13	rate	rate	NOUN
aiti-8904	44	14	and	and	CCONJ
aiti-8904	44	15	control	control	VERB
aiti-8904	44	16	the	the	DET
aiti-8904	44	17	leakage	leakage	NOUN
aiti-8904	44	18	point	point	NOUN
aiti-8904	44	19	position	position	NOUN
aiti-8904	44	20	error	error	NOUN
aiti-8904	44	21	within	within	ADP
aiti-8904	44	22	3	3	NUM
aiti-8904	44	23	meters	meter	NOUN
aiti-8904	44	24	[	[	X
aiti-8904	44	25	7	7	NUM
aiti-8904	44	26	]	]	PUNCT
aiti-8904	44	27	.	.	PUNCT
aiti-8904	45	1	as	as	SCONJ
aiti-8904	45	2	shown	show	VERB
aiti-8904	45	3	by	by	ADP
aiti-8904	45	4	the	the	DET
aiti-8904	45	5	literature	literature	PROPN
aiti-8904	45	6	review	review	NOUN
aiti-8904	45	7	above	above	ADV
aiti-8904	45	8	,	,	PUNCT
aiti-8904	45	9	employing	employ	VERB
aiti-8904	45	10	ai	ai	ADP
aiti-8904	45	11	techniques	technique	NOUN
aiti-8904	45	12	to	to	PART
aiti-8904	45	13	develop	develop	VERB
aiti-8904	45	14	leak	leak	NOUN
aiti-8904	45	15	detection	detection	NOUN
aiti-8904	45	16	has	have	AUX
aiti-8904	45	17	received	receive	VERB
aiti-8904	45	18	growing	grow	VERB
aiti-8904	45	19	attention	attention	NOUN
aiti-8904	45	20	.	.	PUNCT
aiti-8904	46	1	however	however	ADV
aiti-8904	46	2	,	,	PUNCT
aiti-8904	46	3	few	few	ADJ
aiti-8904	46	4	studies	study	NOUN
aiti-8904	46	5	have	have	AUX
aiti-8904	46	6	investigated	investigate	VERB
aiti-8904	46	7	ai	ai	NOUN
aiti-8904	46	8	-	-	PUNCT
aiti-8904	46	9	based	base	VERB
aiti-8904	46	10	leak	leak	NOUN
aiti-8904	46	11	detection	detection	NOUN
aiti-8904	46	12	systems	system	NOUN
aiti-8904	46	13	based	base	VERB
aiti-8904	46	14	on	on	ADP
aiti-8904	46	15	cnn	cnn	PROPN
aiti-8904	46	16	for	for	ADP
aiti-8904	46	17	automatically	automatically	ADV
aiti-8904	46	18	monitoring	monitor	VERB
aiti-8904	46	19	the	the	DET
aiti-8904	46	20	leaks	leak	NOUN
aiti-8904	46	21	in	in	ADP
aiti-8904	46	22	a	a	DET
aiti-8904	46	23	real	real	ADJ
aiti-8904	46	24	context	context	NOUN
aiti-8904	46	25	.	.	PUNCT
aiti-8904	47	1	to	to	PART
aiti-8904	47	2	fill	fill	VERB
aiti-8904	47	3	the	the	DET
aiti-8904	47	4	gap	gap	NOUN
aiti-8904	47	5	,	,	PUNCT
aiti-8904	47	6	the	the	DET
aiti-8904	47	7	present	present	ADJ
aiti-8904	47	8	study	study	NOUN
aiti-8904	47	9	aims	aim	VERB
aiti-8904	47	10	to	to	PART
aiti-8904	47	11	develop	develop	VERB
aiti-8904	47	12	an	an	DET
aiti-8904	47	13	ai	ai	NOUN
aiti-8904	47	14	-	-	PUNCT
aiti-8904	47	15	based	base	VERB
aiti-8904	47	16	model	model	NOUN
aiti-8904	47	17	to	to	PART
aiti-8904	47	18	detect	detect	VERB
aiti-8904	47	19	the	the	DET
aiti-8904	47	20	leaks	leak	NOUN
aiti-8904	47	21	of	of	ADP
aiti-8904	47	22	a	a	DET
aiti-8904	47	23	long	long	ADJ
aiti-8904	47	24	-	-	PUNCT
aiti-8904	47	25	distance	distance	NOUN
aiti-8904	47	26	pipeline	pipeline	NOUN
aiti-8904	47	27	and	and	CCONJ
aiti-8904	47	28	test	test	VERB
aiti-8904	47	29	the	the	DET
aiti-8904	47	30	model	model	NOUN
aiti-8904	47	31	’s	’s	PART
aiti-8904	47	32	accuracy	accuracy	NOUN
aiti-8904	47	33	in	in	ADP
aiti-8904	47	34	a	a	DET
aiti-8904	47	35	real	real	ADJ
aiti-8904	47	36	context	context	NOUN
aiti-8904	47	37	.	.	PUNCT
aiti-8904	48	1	in	in	ADP
aiti-8904	48	2	this	this	DET
aiti-8904	48	3	research	research	NOUN
aiti-8904	48	4	,	,	PUNCT
aiti-8904	48	5	pressure	pressure	NOUN
aiti-8904	48	6	sensors	sensor	NOUN
aiti-8904	48	7	(	(	PUNCT
aiti-8904	48	8	the	the	DET
aiti-8904	48	9	inherent	inherent	ADJ
aiti-8904	48	10	hardware	hardware	NOUN
aiti-8904	48	11	equipment	equipment	NOUN
aiti-8904	48	12	of	of	ADP
aiti-8904	48	13	a	a	DET
aiti-8904	48	14	pipeline	pipeline	NOUN
aiti-8904	48	15	)	)	PUNCT
aiti-8904	48	16	are	be	AUX
aiti-8904	48	17	used	use	VERB
aiti-8904	48	18	to	to	PART
aiti-8904	48	19	read	read	VERB
aiti-8904	48	20	the	the	DET
aiti-8904	48	21	gas	gas	NOUN
aiti-8904	48	22	pressure	pressure	NOUN
aiti-8904	48	23	in	in	ADP
aiti-8904	48	24	the	the	DET
aiti-8904	48	25	pipeline	pipeline	NOUN
aiti-8904	48	26	,	,	PUNCT
aiti-8904	48	27	and	and	CCONJ
aiti-8904	48	28	the	the	DET
aiti-8904	48	29	collected	collect	VERB
aiti-8904	48	30	gas	gas	NOUN
aiti-8904	48	31	pressure	pressure	NOUN
aiti-8904	48	32	data	datum	NOUN
aiti-8904	48	33	are	be	AUX
aiti-8904	48	34	used	use	VERB
aiti-8904	48	35	for	for	ADP
aiti-8904	48	36	feature	feature	NOUN
aiti-8904	48	37	extraction	extraction	NOUN
aiti-8904	48	38	and	and	CCONJ
aiti-8904	48	39	model	model	NOUN
aiti-8904	48	40	training	training	NOUN
aiti-8904	48	41	.	.	PUNCT
aiti-8904	49	1	additionally	additionally	ADV
aiti-8904	49	2	,	,	PUNCT
aiti-8904	49	3	the	the	DET
aiti-8904	49	4	present	present	ADJ
aiti-8904	49	5	study	study	NOUN
aiti-8904	49	6	employs	employ	VERB
aiti-8904	49	7	the	the	DET
aiti-8904	49	8	pipeline	pipeline	NOUN
aiti-8904	49	9	detection	detection	NOUN
aiti-8904	49	10	data	datum	NOUN
aiti-8904	49	11	from	from	ADP
aiti-8904	49	12	supervisory	supervisory	ADJ
aiti-8904	49	13	control	control	NOUN
aiti-8904	49	14	and	and	CCONJ
aiti-8904	49	15	data	datum	NOUN
aiti-8904	49	16	acquisition	acquisition	NOUN
aiti-8904	49	17	(	(	PUNCT
aiti-8904	49	18	scada	scada	PROPN
aiti-8904	49	19	)	)	PUNCT
aiti-8904	49	20	as	as	ADP
aiti-8904	49	21	the	the	DET
aiti-8904	49	22	training	training	NOUN
aiti-8904	49	23	data	datum	NOUN
aiti-8904	49	24	for	for	ADP
aiti-8904	49	25	deep	deep	ADJ
aiti-8904	49	26	learning	learning	NOUN
aiti-8904	49	27	.	.	PUNCT
aiti-8904	50	1	by	by	ADP
aiti-8904	50	2	doing	do	VERB
aiti-8904	50	3	so	so	ADV
aiti-8904	50	4	,	,	PUNCT
aiti-8904	50	5	the	the	DET
aiti-8904	50	6	smart	smart	ADJ
aiti-8904	50	7	detection	detection	NOUN
aiti-8904	50	8	of	of	ADP
aiti-8904	50	9	pipeline	pipeline	NOUN
aiti-8904	50	10	leaks	leak	NOUN
aiti-8904	50	11	can	can	AUX
aiti-8904	50	12	be	be	AUX
aiti-8904	50	13	realized	realize	VERB
aiti-8904	50	14	,	,	PUNCT
aiti-8904	50	15	and	and	CCONJ
aiti-8904	50	16	the	the	DET
aiti-8904	50	17	hardware	hardware	NOUN
aiti-8904	50	18	cost	cost	NOUN
aiti-8904	50	19	of	of	ADP
aiti-8904	50	20	the	the	DET
aiti-8904	50	21	sensor	sensor	NOUN
aiti-8904	50	22	can	can	AUX
aiti-8904	50	23	be	be	AUX
aiti-8904	50	24	saved	save	VERB
aiti-8904	50	25	.	.	PUNCT
aiti-8904	51	1	specifically	specifically	ADV
aiti-8904	51	2	,	,	PUNCT
aiti-8904	51	3	this	this	DET
aiti-8904	51	4	research	research	NOUN
aiti-8904	51	5	uses	use	VERB
aiti-8904	51	6	the	the	DET
aiti-8904	51	7	pressure	pressure	NOUN
aiti-8904	51	8	sensor	sensor	NOUN
aiti-8904	51	9	data	datum	NOUN
aiti-8904	51	10	recorded	record	VERB
aiti-8904	51	11	in	in	ADP
aiti-8904	51	12	scada	scada	PROPN
aiti-8904	51	13	to	to	PART
aiti-8904	51	14	analyze	analyze	VERB
aiti-8904	51	15	and	and	CCONJ
aiti-8904	51	16	extract	extract	VERB
aiti-8904	51	17	suitable	suitable	ADJ
aiti-8904	51	18	data	datum	NOUN
aiti-8904	51	19	features	feature	NOUN
aiti-8904	51	20	,	,	PUNCT
aiti-8904	51	21	and	and	CCONJ
aiti-8904	51	22	then	then	ADV
aiti-8904	51	23	labels	label	VERB
aiti-8904	51	24	the	the	DET
aiti-8904	51	25	data	data	NOUN
aiti-8904	51	26	features	feature	VERB
aiti-8904	51	27	separately	separately	ADV
aiti-8904	51	28	and	and	CCONJ
aiti-8904	51	29	conducts	conduct	VERB
aiti-8904	51	30	model	model	NOUN
aiti-8904	51	31	training	training	NOUN
aiti-8904	51	32	through	through	ADP
aiti-8904	51	33	cnn	cnn	PROPN
aiti-8904	51	34	.	.	PUNCT
aiti-8904	52	1	the	the	DET
aiti-8904	52	2	model	model	NOUN
aiti-8904	52	3	trained	train	VERB
aiti-8904	52	4	in	in	ADP
aiti-8904	52	5	this	this	DET
aiti-8904	52	6	study	study	NOUN
aiti-8904	52	7	can	can	AUX
aiti-8904	52	8	realize	realize	VERB
aiti-8904	52	9	a	a	DET
aiti-8904	52	10	rapid	rapid	ADJ
aiti-8904	52	11	and	and	CCONJ
aiti-8904	52	12	accurate	accurate	ADJ
aiti-8904	52	13	function	function	NOUN
aiti-8904	52	14	of	of	ADP
aiti-8904	52	15	leak	leak	NOUN
aiti-8904	52	16	detection	detection	NOUN
aiti-8904	52	17	.	.	PUNCT
aiti-8904	53	1	the	the	DET
aiti-8904	53	2	developed	develop	VERB
aiti-8904	53	3	system	system	NOUN
aiti-8904	53	4	is	be	AUX
aiti-8904	53	5	beneficial	beneficial	ADJ
aiti-8904	53	6	for	for	SCONJ
aiti-8904	53	7	managers	manager	NOUN
aiti-8904	53	8	to	to	PART
aiti-8904	53	9	detect	detect	VERB
aiti-8904	53	10	leaks	leak	NOUN
aiti-8904	53	11	in	in	ADP
aiti-8904	53	12	the	the	DET
aiti-8904	53	13	long	long	ADJ
aiti-8904	53	14	-	-	PUNCT
aiti-8904	53	15	distance	distance	NOUN
aiti-8904	53	16	pipeline	pipeline	NOUN
aiti-8904	53	17	.	.	PUNCT
aiti-8904	54	1	2	2	X
aiti-8904	54	2	.	.	X
aiti-8904	54	3	system	system	NOUN
aiti-8904	54	4	structure	structure	NOUN
aiti-8904	54	5	this	this	DET
aiti-8904	54	6	research	research	NOUN
aiti-8904	54	7	obtains	obtain	VERB
aiti-8904	54	8	the	the	DET
aiti-8904	54	9	dataset	dataset	NOUN
aiti-8904	54	10	from	from	ADP
aiti-8904	54	11	the	the	DET
aiti-8904	54	12	pressure	pressure	NOUN
aiti-8904	54	13	sensors	sensor	NOUN
aiti-8904	54	14	in	in	ADP
aiti-8904	54	15	the	the	DET
aiti-8904	54	16	scada	scada	PROPN
aiti-8904	54	17	system	system	NOUN
aiti-8904	54	18	provided	provide	VERB
aiti-8904	54	19	by	by	ADP
aiti-8904	54	20	a	a	DET
aiti-8904	54	21	petrochemical	petrochemical	NOUN
aiti-8904	54	22	manufacturer	manufacturer	NOUN
aiti-8904	54	23	in	in	ADP
aiti-8904	54	24	taiwan	taiwan	PROPN
aiti-8904	54	25	.	.	PUNCT
aiti-8904	55	1	fig	fig	NOUN
aiti-8904	55	2	.	.	PUNCT
aiti-8904	56	1	1	1	NUM
aiti-8904	56	2	depicts	depict	VERB
aiti-8904	56	3	the	the	DET
aiti-8904	56	4	schematic	schematic	ADJ
aiti-8904	56	5	diagram	diagram	NOUN
aiti-8904	56	6	of	of	ADP
aiti-8904	56	7	this	this	DET
aiti-8904	56	8	study	study	NOUN
aiti-8904	56	9	by	by	ADP
aiti-8904	56	10	showing	show	VERB
aiti-8904	56	11	the	the	DET
aiti-8904	56	12	pipeline	pipeline	NOUN
aiti-8904	56	13	transportation	transportation	NOUN
aiti-8904	56	14	system	system	NOUN
aiti-8904	56	15	.	.	PUNCT
aiti-8904	57	1	pt	pt	PROPN
aiti-8904	57	2	is	be	AUX
aiti-8904	57	3	a	a	DET
aiti-8904	57	4	pressure	pressure	NOUN
aiti-8904	57	5	gauge	gauge	NOUN
aiti-8904	57	6	.	.	PUNCT
aiti-8904	58	1	the	the	DET
aiti-8904	58	2	pipeline	pipeline	NOUN
aiti-8904	58	3	transportation	transportation	NOUN
aiti-8904	58	4	starts	start	VERB
aiti-8904	58	5	from	from	ADP
aiti-8904	58	6	plant	plant	NOUN
aiti-8904	58	7	a	a	PRON
aiti-8904	58	8	to	to	PART
aiti-8904	58	9	plant	plant	VERB
aiti-8904	58	10	c	c	NOUN
aiti-8904	58	11	via	via	ADP
aiti-8904	58	12	plant	plant	PROPN
aiti-8904	58	13	b.	b.	PROPN
aiti-8904	58	14	plant	plant	PROPN
aiti-8904	58	15	c	c	PROPN
aiti-8904	58	16	is	be	AUX
aiti-8904	58	17	the	the	DET
aiti-8904	58	18	end	end	NOUN
aiti-8904	58	19	of	of	ADP
aiti-8904	58	20	the	the	DET
aiti-8904	58	21	entire	entire	ADJ
aiti-8904	58	22	pipeline	pipeline	NOUN
aiti-8904	58	23	transportation	transportation	NOUN
aiti-8904	58	24	system	system	NOUN
aiti-8904	58	25	.	.	PUNCT
aiti-8904	59	1	the	the	DET
aiti-8904	59	2	distance	distance	NOUN
aiti-8904	59	3	between	between	ADP
aiti-8904	59	4	plant	plant	NOUN
aiti-8904	59	5	a	a	PRON
aiti-8904	59	6	and	and	CCONJ
aiti-8904	59	7	plant	plant	NOUN
aiti-8904	59	8	b	b	NOUN
aiti-8904	59	9	is	be	AUX
aiti-8904	59	10	about	about	ADV
aiti-8904	59	11	63	63	NUM
aiti-8904	59	12	kilometers	kilometer	NOUN
aiti-8904	59	13	.	.	PUNCT
aiti-8904	60	1	the	the	DET
aiti-8904	60	2	distance	distance	NOUN
aiti-8904	60	3	between	between	ADP
aiti-8904	60	4	plant	plant	NOUN
aiti-8904	60	5	b	b	NOUN
aiti-8904	60	6	and	and	CCONJ
aiti-8904	60	7	plant	plant	NOUN
aiti-8904	60	8	c	c	NOUN
aiti-8904	60	9	is	be	AUX
aiti-8904	60	10	about	about	ADV
aiti-8904	60	11	136	136	NUM
aiti-8904	60	12	kilometers	kilometer	NOUN
aiti-8904	60	13	.	.	PUNCT
aiti-8904	61	1	the	the	DET
aiti-8904	61	2	pipeline	pipeline	NOUN
aiti-8904	61	3	in	in	ADP
aiti-8904	61	4	plant	plant	NOUN
aiti-8904	61	5	b	b	NOUN
aiti-8904	61	6	will	will	AUX
aiti-8904	61	7	be	be	AUX
aiti-8904	61	8	pressurized	pressurize	VERB
aiti-8904	61	9	by	by	ADP
aiti-8904	61	10	pumps	pump	NOUN
aiti-8904	61	11	before	before	ADP
aiti-8904	61	12	sending	send	VERB
aiti-8904	61	13	resources	resource	NOUN
aiti-8904	61	14	to	to	PART
aiti-8904	61	15	plant	plant	VERB
aiti-8904	61	16	c.	c.	NOUN
aiti-8904	61	17	in	in	ADP
aiti-8904	61	18	this	this	DET
aiti-8904	61	19	study	study	NOUN
aiti-8904	61	20	,	,	PUNCT
aiti-8904	61	21	the	the	DET
aiti-8904	61	22	pressure	pressure	NOUN
aiti-8904	61	23	sensor	sensor	NOUN
aiti-8904	61	24	,	,	PUNCT
aiti-8904	61	25	pt	pt	NOUN
aiti-8904	61	26	-	-	PUNCT
aiti-8904	61	27	m621	m621	PROPN
aiti-8904	61	28	,	,	PUNCT
aiti-8904	61	29	in	in	ADP
aiti-8904	61	30	plant	plant	NOUN
aiti-8904	61	31	b	b	NOUN
aiti-8904	61	32	is	be	AUX
aiti-8904	61	33	used	use	VERB
aiti-8904	61	34	to	to	PART
aiti-8904	61	35	collect	collect	VERB
aiti-8904	61	36	data	datum	NOUN
aiti-8904	61	37	for	for	ADP
aiti-8904	61	38	subsequent	subsequent	ADJ
aiti-8904	61	39	data	datum	NOUN
aiti-8904	61	40	analysis	analysis	NOUN
aiti-8904	61	41	and	and	CCONJ
aiti-8904	61	42	model	model	NOUN
aiti-8904	61	43	training	training	NOUN
aiti-8904	61	44	,	,	PUNCT
aiti-8904	61	45	and	and	CCONJ
aiti-8904	61	46	the	the	DET
aiti-8904	61	47	pressure	pressure	NOUN
aiti-8904	61	48	sensor	sensor	NOUN
aiti-8904	61	49	,	,	PUNCT
aiti-8904	61	50	pt	pt	PROPN
aiti-8904	61	51	-	-	PUNCT
aiti-8904	61	52	m622	m622	PROPN
aiti-8904	61	53	,	,	PUNCT
aiti-8904	61	54	is	be	AUX
aiti-8904	61	55	used	use	VERB
aiti-8904	61	56	to	to	PART
aiti-8904	61	57	test	test	VERB
aiti-8904	61	58	and	and	CCONJ
aiti-8904	61	59	validate	validate	VERB
aiti-8904	61	60	the	the	DET
aiti-8904	61	61	model	model	NOUN
aiti-8904	61	62	.	.	PUNCT
aiti-8904	62	1	fig	fig	NOUN
aiti-8904	62	2	.	.	PUNCT
aiti-8904	63	1	1	1	NUM
aiti-8904	63	2	schematic	schematic	ADJ
aiti-8904	63	3	diagram	diagram	NOUN
aiti-8904	63	4	of	of	ADP
aiti-8904	63	5	this	this	DET
aiti-8904	63	6	study	study	NOUN
aiti-8904	63	7	170	170	NUM
aiti-8904	63	8	advances	advance	NOUN
aiti-8904	63	9	in	in	ADP
aiti-8904	63	10	technology	technology	NOUN
aiti-8904	63	11	innovation	innovation	NOUN
aiti-8904	63	12	,	,	PUNCT
aiti-8904	63	13	vol	vol	NOUN
aiti-8904	63	14	.	.	PROPN
aiti-8904	64	1	7	7	NUM
aiti-8904	64	2	,	,	PUNCT
aiti-8904	64	3	no	no	INTJ
aiti-8904	64	4	.	.	NOUN
aiti-8904	64	5	3	3	NUM
aiti-8904	64	6	,	,	PUNCT
aiti-8904	64	7	2022	2022	NUM
aiti-8904	64	8	,	,	PUNCT
aiti-8904	64	9	pp	pp	ADJ
aiti-8904	64	10	.	.	PUNCT
aiti-8904	65	1	169	169	NUM
aiti-8904	65	2	-	-	SYM
aiti-8904	65	3	180	180	NUM
aiti-8904	65	4	fig	fig	NOUN
aiti-8904	65	5	.	.	PUNCT
aiti-8904	66	1	2	2	NUM
aiti-8904	66	2	system	system	NOUN
aiti-8904	66	3	architecture	architecture	NOUN
aiti-8904	66	4	fig	fig	NOUN
aiti-8904	66	5	.	.	PUNCT
aiti-8904	67	1	3	3	NUM
aiti-8904	67	2	data	datum	NOUN
aiti-8904	67	3	from	from	ADP
aiti-8904	67	4	the	the	DET
aiti-8904	67	5	real	real	ADJ
aiti-8904	67	6	pipeline	pipeline	NOUN
aiti-8904	67	7	leak	leak	NOUN
aiti-8904	67	8	test	test	VERB
aiti-8904	67	9	the	the	DET
aiti-8904	67	10	research	research	NOUN
aiti-8904	67	11	system	system	NOUN
aiti-8904	67	12	architecture	architecture	NOUN
aiti-8904	67	13	is	be	AUX
aiti-8904	67	14	shown	show	VERB
aiti-8904	67	15	in	in	ADP
aiti-8904	67	16	fig	fig	NOUN
aiti-8904	67	17	.	.	PUNCT
aiti-8904	68	1	2	2	X
aiti-8904	68	2	.	.	X
aiti-8904	68	3	first	first	ADV
aiti-8904	68	4	,	,	PUNCT
aiti-8904	68	5	the	the	DET
aiti-8904	68	6	data	datum	NOUN
aiti-8904	68	7	received	receive	VERB
aiti-8904	68	8	from	from	ADP
aiti-8904	68	9	the	the	DET
aiti-8904	68	10	scada	scada	PROPN
aiti-8904	68	11	system	system	NOUN
aiti-8904	68	12	is	be	AUX
aiti-8904	68	13	classified	classify	VERB
aiti-8904	68	14	into	into	ADP
aiti-8904	68	15	the	the	DET
aiti-8904	68	16	normal	normal	ADJ
aiti-8904	68	17	data	datum	NOUN
aiti-8904	68	18	group	group	NOUN
aiti-8904	68	19	and	and	CCONJ
aiti-8904	68	20	the	the	DET
aiti-8904	68	21	leak	leak	NOUN
aiti-8904	68	22	data	datum	NOUN
aiti-8904	68	23	group	group	NOUN
aiti-8904	68	24	.	.	PUNCT
aiti-8904	69	1	next	next	ADV
aiti-8904	69	2	,	,	PUNCT
aiti-8904	69	3	the	the	DET
aiti-8904	69	4	characteristics	characteristic	NOUN
aiti-8904	69	5	of	of	ADP
aiti-8904	69	6	the	the	DET
aiti-8904	69	7	two	two	NUM
aiti-8904	69	8	data	datum	NOUN
aiti-8904	69	9	groups	group	NOUN
aiti-8904	69	10	are	be	AUX
aiti-8904	69	11	extracted	extract	VERB
aiti-8904	69	12	through	through	ADP
aiti-8904	69	13	feature	feature	NOUN
aiti-8904	69	14	engineering	engineering	NOUN
aiti-8904	69	15	.	.	PUNCT
aiti-8904	70	1	therefore	therefore	ADV
aiti-8904	70	2	,	,	PUNCT
aiti-8904	70	3	the	the	DET
aiti-8904	70	4	two	two	NUM
aiti-8904	70	5	groups	group	NOUN
aiti-8904	70	6	’	'	PUNCT
aiti-8904	70	7	data	datum	NOUN
aiti-8904	70	8	are	be	AUX
aiti-8904	70	9	merged	merge	VERB
aiti-8904	70	10	and	and	CCONJ
aiti-8904	70	11	divided	divide	VERB
aiti-8904	70	12	into	into	ADP
aiti-8904	70	13	the	the	DET
aiti-8904	70	14	training	training	NOUN
aiti-8904	70	15	data	datum	NOUN
aiti-8904	70	16	set	set	VERB
aiti-8904	70	17	and	and	CCONJ
aiti-8904	70	18	the	the	DET
aiti-8904	70	19	test	test	NOUN
aiti-8904	70	20	data	datum	NOUN
aiti-8904	70	21	set	set	VERB
aiti-8904	70	22	for	for	ADP
aiti-8904	70	23	model	model	NOUN
aiti-8904	70	24	training	training	NOUN
aiti-8904	70	25	.	.	PUNCT
aiti-8904	71	1	finally	finally	ADV
aiti-8904	71	2	,	,	PUNCT
aiti-8904	71	3	this	this	DET
aiti-8904	71	4	study	study	NOUN
aiti-8904	71	5	tests	test	VERB
aiti-8904	71	6	whether	whether	SCONJ
aiti-8904	71	7	the	the	DET
aiti-8904	71	8	model	model	NOUN
aiti-8904	71	9	can	can	AUX
aiti-8904	71	10	distinguish	distinguish	VERB
aiti-8904	71	11	the	the	DET
aiti-8904	71	12	leak	leak	NOUN
aiti-8904	71	13	data	datum	NOUN
aiti-8904	71	14	.	.	PUNCT
aiti-8904	72	1	overall	overall	ADV
aiti-8904	72	2	,	,	PUNCT
aiti-8904	72	3	this	this	DET
aiti-8904	72	4	study	study	NOUN
aiti-8904	72	5	employs	employ	VERB
aiti-8904	72	6	the	the	DET
aiti-8904	72	7	cnn	cnn	PROPN
aiti-8904	72	8	architecture	architecture	NOUN
aiti-8904	72	9	to	to	PART
aiti-8904	72	10	construct	construct	VERB
aiti-8904	72	11	the	the	DET
aiti-8904	72	12	model	model	NOUN
aiti-8904	72	13	.	.	PUNCT
aiti-8904	73	1	if	if	SCONJ
aiti-8904	73	2	the	the	DET
aiti-8904	73	3	model	model	NOUN
aiti-8904	73	4	can	can	AUX
aiti-8904	73	5	not	not	PART
aiti-8904	73	6	accurately	accurately	ADV
aiti-8904	73	7	classify	classify	VERB
aiti-8904	73	8	the	the	DET
aiti-8904	73	9	two	two	NUM
aiti-8904	73	10	types	type	NOUN
aiti-8904	73	11	of	of	ADP
aiti-8904	73	12	data	datum	NOUN
aiti-8904	73	13	,	,	PUNCT
aiti-8904	73	14	the	the	DET
aiti-8904	73	15	feature	feature	NOUN
aiti-8904	73	16	engineering	engineering	NOUN
aiti-8904	73	17	needs	need	VERB
aiti-8904	73	18	to	to	PART
aiti-8904	73	19	be	be	AUX
aiti-8904	73	20	adjusted	adjust	VERB
aiti-8904	73	21	until	until	SCONJ
aiti-8904	73	22	the	the	DET
aiti-8904	73	23	model	model	NOUN
aiti-8904	73	24	can	can	AUX
aiti-8904	73	25	effectively	effectively	ADV
aiti-8904	73	26	classify	classify	VERB
aiti-8904	73	27	the	the	DET
aiti-8904	73	28	two	two	NUM
aiti-8904	73	29	kinds	kind	NOUN
aiti-8904	73	30	of	of	ADP
aiti-8904	73	31	data	datum	NOUN
aiti-8904	73	32	.	.	PUNCT
aiti-8904	74	1	fig	fig	NOUN
aiti-8904	74	2	.	.	PUNCT
aiti-8904	75	1	3	3	NUM
aiti-8904	75	2	shows	show	VERB
aiti-8904	75	3	the	the	DET
aiti-8904	75	4	full	full	ADJ
aiti-8904	75	5	-	-	PUNCT
aiti-8904	75	6	day	day	NOUN
aiti-8904	75	7	data	datum	NOUN
aiti-8904	75	8	on	on	ADP
aiti-8904	75	9	the	the	DET
aiti-8904	75	10	day	day	NOUN
aiti-8904	75	11	of	of	ADP
aiti-8904	75	12	the	the	DET
aiti-8904	75	13	actual	actual	ADJ
aiti-8904	75	14	leak	leak	NOUN
aiti-8904	75	15	test	test	NOUN
aiti-8904	75	16	.	.	PUNCT
aiti-8904	76	1	the	the	DET
aiti-8904	76	2	time	time	NOUN
aiti-8904	76	3	points	point	NOUN
aiti-8904	76	4	of	of	ADP
aiti-8904	76	5	the	the	DET
aiti-8904	76	6	three	three	NUM
aiti-8904	76	7	leak	leak	NOUN
aiti-8904	76	8	tests	test	NOUN
aiti-8904	76	9	for	for	ADP
aiti-8904	76	10	data	datum	NOUN
aiti-8904	76	11	analysis	analysis	NOUN
aiti-8904	76	12	and	and	CCONJ
aiti-8904	76	13	feature	feature	NOUN
aiti-8904	76	14	engineering	engineering	NOUN
aiti-8904	76	15	are	be	AUX
aiti-8904	76	16	marked	mark	VERB
aiti-8904	76	17	in	in	ADP
aiti-8904	76	18	fig	fig	NOUN
aiti-8904	76	19	.	.	PUNCT
aiti-8904	77	1	3	3	X
aiti-8904	77	2	.	.	X
aiti-8904	77	3	the	the	DET
aiti-8904	77	4	x	x	ADJ
aiti-8904	77	5	-	-	ADJ
aiti-8904	77	6	axis	axis	ADJ
aiti-8904	77	7	presents	present	NOUN
aiti-8904	77	8	the	the	DET
aiti-8904	77	9	time	time	NOUN
aiti-8904	77	10	point	point	NOUN
aiti-8904	77	11	of	of	ADP
aiti-8904	77	12	measuring	measure	VERB
aiti-8904	77	13	the	the	DET
aiti-8904	77	14	pressure	pressure	NOUN
aiti-8904	77	15	.	.	PUNCT
aiti-8904	78	1	the	the	DET
aiti-8904	78	2	time	time	NOUN
aiti-8904	78	3	interval	interval	NOUN
aiti-8904	78	4	between	between	ADP
aiti-8904	78	5	pieces	piece	NOUN
aiti-8904	78	6	is	be	AUX
aiti-8904	78	7	one	one	NUM
aiti-8904	78	8	second	second	NOUN
aiti-8904	78	9	.	.	PUNCT
aiti-8904	79	1	2.1	2.1	NUM
aiti-8904	79	2	.	.	PUNCT
aiti-8904	80	1	data	datum	NOUN
aiti-8904	80	2	analysis	analysis	NOUN
aiti-8904	80	3	and	and	CCONJ
aiti-8904	80	4	feature	feature	NOUN
aiti-8904	80	5	extraction	extraction	NOUN
aiti-8904	80	6	fig	fig	NOUN
aiti-8904	80	7	.	.	PUNCT
aiti-8904	81	1	4	4	NUM
aiti-8904	81	2	illustrates	illustrate	VERB
aiti-8904	81	3	the	the	DET
aiti-8904	81	4	data	datum	NOUN
aiti-8904	81	5	sampling	sample	VERB
aiti-8904	81	6	for	for	ADP
aiti-8904	81	7	the	the	DET
aiti-8904	81	8	leak	leak	NOUN
aiti-8904	81	9	test	test	NOUN
aiti-8904	81	10	.	.	PUNCT
aiti-8904	82	1	for	for	ADP
aiti-8904	82	2	observing	observe	VERB
aiti-8904	82	3	the	the	DET
aiti-8904	82	4	data	data	NOUN
aiti-8904	82	5	variance	variance	NOUN
aiti-8904	82	6	in	in	ADP
aiti-8904	82	7	detail	detail	NOUN
aiti-8904	82	8	,	,	PUNCT
aiti-8904	82	9	the	the	DET
aiti-8904	82	10	data	datum	NOUN
aiti-8904	82	11	for	for	ADP
aiti-8904	82	12	analysis	analysis	NOUN
aiti-8904	82	13	is	be	AUX
aiti-8904	82	14	captured	capture	VERB
aiti-8904	82	15	after	after	ADP
aiti-8904	82	16	the	the	DET
aiti-8904	82	17	start	start	NOUN
aiti-8904	82	18	of	of	ADP
aiti-8904	82	19	the	the	DET
aiti-8904	82	20	leak	leak	NOUN
aiti-8904	82	21	test	test	NOUN
aiti-8904	82	22	.	.	PUNCT
aiti-8904	83	1	in	in	ADP
aiti-8904	83	2	fig	fig	NOUN
aiti-8904	83	3	.	.	PUNCT
aiti-8904	84	1	4	4	NUM
aiti-8904	84	2	,	,	PUNCT
aiti-8904	84	3	the	the	DET
aiti-8904	84	4	positions	position	NOUN
aiti-8904	84	5	pointed	point	VERB
aiti-8904	84	6	to	to	ADP
aiti-8904	84	7	by	by	ADP
aiti-8904	84	8	the	the	DET
aiti-8904	84	9	red	red	ADJ
aiti-8904	84	10	arrow	arrow	NOUN
aiti-8904	84	11	are	be	AUX
aiti-8904	84	12	the	the	DET
aiti-8904	84	13	data	data	NOUN
aiti-8904	84	14	points	point	NOUN
aiti-8904	84	15	during	during	ADP
aiti-8904	84	16	the	the	DET
aiti-8904	84	17	leak	leak	NOUN
aiti-8904	84	18	test	test	NOUN
aiti-8904	84	19	;	;	PUNCT
aiti-8904	84	20	through	through	ADP
aiti-8904	84	21	data	data	NOUN
aiti-8904	84	22	visualization	visualization	NOUN
aiti-8904	84	23	,	,	PUNCT
aiti-8904	84	24	there	there	PRON
aiti-8904	84	25	are	be	VERB
aiti-8904	84	26	obvious	obvious	ADJ
aiti-8904	84	27	pressure	pressure	NOUN
aiti-8904	84	28	fluctuations	fluctuation	NOUN
aiti-8904	84	29	when	when	SCONJ
aiti-8904	84	30	leakage	leakage	NOUN
aiti-8904	84	31	occurs	occur	VERB
aiti-8904	84	32	.	.	PUNCT
aiti-8904	85	1	this	this	DET
aiti-8904	85	2	study	study	NOUN
aiti-8904	85	3	uses	use	VERB
aiti-8904	85	4	these	these	DET
aiti-8904	85	5	three	three	NUM
aiti-8904	85	6	leakage	leakage	NOUN
aiti-8904	85	7	points	point	NOUN
aiti-8904	85	8	as	as	SCONJ
aiti-8904	85	9	reference	reference	NOUN
aiti-8904	85	10	points	point	VERB
aiti-8904	85	11	to	to	PART
aiti-8904	85	12	find	find	VERB
aiti-8904	85	13	more	more	ADV
aiti-8904	85	14	suitable	suitable	ADJ
aiti-8904	85	15	data	datum	NOUN
aiti-8904	85	16	for	for	ADP
aiti-8904	85	17	feature	feature	NOUN
aiti-8904	85	18	engineering	engineering	NOUN
aiti-8904	85	19	.	.	PUNCT
aiti-8904	86	1	although	although	SCONJ
aiti-8904	86	2	the	the	DET
aiti-8904	86	3	leakage	leakage	NOUN
aiti-8904	86	4	appears	appear	VERB
aiti-8904	86	5	at	at	ADP
aiti-8904	86	6	these	these	DET
aiti-8904	86	7	points	point	NOUN
aiti-8904	86	8	,	,	PUNCT
aiti-8904	86	9	the	the	DET
aiti-8904	86	10	data	datum	NOUN
aiti-8904	86	11	of	of	ADP
aiti-8904	86	12	the	the	DET
aiti-8904	86	13	leak	leak	NOUN
aiti-8904	86	14	pattern	pattern	NOUN
aiti-8904	86	15	for	for	ADP
aiti-8904	86	16	extracting	extract	VERB
aiti-8904	86	17	the	the	DET
aiti-8904	86	18	feature	feature	NOUN
aiti-8904	86	19	can	can	AUX
aiti-8904	86	20	not	not	PART
aiti-8904	86	21	be	be	AUX
aiti-8904	86	22	reflected	reflect	VERB
aiti-8904	86	23	by	by	ADP
aiti-8904	86	24	these	these	DET
aiti-8904	86	25	points	point	NOUN
aiti-8904	86	26	.	.	PUNCT
aiti-8904	87	1	therefore	therefore	ADV
aiti-8904	87	2	,	,	PUNCT
aiti-8904	87	3	according	accord	VERB
aiti-8904	87	4	to	to	ADP
aiti-8904	87	5	the	the	DET
aiti-8904	87	6	rule	rule	NOUN
aiti-8904	87	7	of	of	ADP
aiti-8904	87	8	thumb	thumb	NOUN
aiti-8904	87	9	for	for	ADP
aiti-8904	87	10	capturing	capture	VERB
aiti-8904	87	11	the	the	DET
aiti-8904	87	12	leak	leak	NOUN
aiti-8904	87	13	feature	feature	NOUN
aiti-8904	87	14	found	find	VERB
aiti-8904	87	15	in	in	ADP
aiti-8904	87	16	this	this	DET
aiti-8904	87	17	study	study	NOUN
aiti-8904	87	18	,	,	PUNCT
aiti-8904	87	19	1000	1000	NUM
aiti-8904	87	20	pieces	piece	NOUN
aiti-8904	87	21	of	of	ADP
aiti-8904	87	22	data	datum	NOUN
aiti-8904	87	23	are	be	AUX
aiti-8904	87	24	taken	take	VERB
aiti-8904	87	25	before	before	ADV
aiti-8904	87	26	and	and	CCONJ
aiti-8904	87	27	after	after	SCONJ
aiti-8904	87	28	the	the	DET
aiti-8904	87	29	leakage	leakage	NOUN
aiti-8904	87	30	points	point	VERB
aiti-8904	87	31	to	to	PART
aiti-8904	87	32	capture	capture	VERB
aiti-8904	87	33	the	the	DET
aiti-8904	87	34	graphical	graphical	ADJ
aiti-8904	87	35	pattern	pattern	NOUN
aiti-8904	87	36	of	of	ADP
aiti-8904	87	37	leak	leak	NOUN
aiti-8904	87	38	occurrence	occurrence	NOUN
aiti-8904	87	39	,	,	PUNCT
aiti-8904	87	40	as	as	SCONJ
aiti-8904	87	41	illustrated	illustrate	VERB
aiti-8904	87	42	in	in	ADP
aiti-8904	87	43	fig	fig	NOUN
aiti-8904	87	44	.	.	PUNCT
aiti-8904	88	1	4	4	X
aiti-8904	88	2	.	.	X
aiti-8904	88	3	the	the	DET
aiti-8904	88	4	time	time	NOUN
aiti-8904	88	5	interval	interval	NOUN
aiti-8904	88	6	for	for	ADP
aiti-8904	88	7	the	the	DET
aiti-8904	88	8	pressure	pressure	NOUN
aiti-8904	88	9	data	datum	NOUN
aiti-8904	88	10	sampling	sample	VERB
aiti-8904	88	11	between	between	ADP
aiti-8904	88	12	pieces	piece	NOUN
aiti-8904	88	13	is	be	AUX
aiti-8904	88	14	one	one	NUM
aiti-8904	88	15	second	second	NOUN
aiti-8904	88	16	.	.	PUNCT
aiti-8904	89	1	in	in	ADP
aiti-8904	89	2	doing	do	VERB
aiti-8904	89	3	so	so	ADV
aiti-8904	89	4	,	,	PUNCT
aiti-8904	89	5	this	this	DET
aiti-8904	89	6	study	study	NOUN
aiti-8904	89	7	does	do	AUX
aiti-8904	89	8	observe	observe	VERB
aiti-8904	89	9	and	and	CCONJ
aiti-8904	89	10	find	find	VERB
aiti-8904	89	11	the	the	DET
aiti-8904	89	12	data	datum	NOUN
aiti-8904	89	13	suitable	suitable	ADJ
aiti-8904	89	14	for	for	ADP
aiti-8904	89	15	presenting	present	VERB
aiti-8904	89	16	the	the	DET
aiti-8904	89	17	characteristics	characteristic	NOUN
aiti-8904	89	18	of	of	ADP
aiti-8904	89	19	the	the	DET
aiti-8904	89	20	leakage	leakage	NOUN
aiti-8904	89	21	data	datum	NOUN
aiti-8904	89	22	.	.	PUNCT
aiti-8904	90	1	in	in	ADP
aiti-8904	90	2	practice	practice	NOUN
aiti-8904	90	3	,	,	PUNCT
aiti-8904	90	4	the	the	DET
aiti-8904	90	5	operators	operator	NOUN
aiti-8904	90	6	test	test	VERB
aiti-8904	90	7	the	the	DET
aiti-8904	90	8	functionality	functionality	NOUN
aiti-8904	90	9	of	of	ADP
aiti-8904	90	10	the	the	DET
aiti-8904	90	11	pressure	pressure	NOUN
aiti-8904	90	12	sensor	sensor	NOUN
aiti-8904	90	13	by	by	ADP
aiti-8904	90	14	opening	opening	NOUN
aiti-8904	90	15	and	and	CCONJ
aiti-8904	90	16	then	then	ADV
aiti-8904	90	17	closing	close	VERB
aiti-8904	90	18	the	the	DET
aiti-8904	90	19	valve	valve	NOUN
aiti-8904	90	20	after	after	ADP
aiti-8904	90	21	30	30	NUM
aiti-8904	90	22	seconds	second	NOUN
aiti-8904	90	23	and	and	CCONJ
aiti-8904	90	24	four	four	NUM
aiti-8904	90	25	minutes	minute	NOUN
aiti-8904	90	26	to	to	PART
aiti-8904	90	27	simulate	simulate	VERB
aiti-8904	90	28	the	the	DET
aiti-8904	90	29	gas	gas	NOUN
aiti-8904	90	30	leaks	leak	NOUN
aiti-8904	90	31	.	.	PUNCT
aiti-8904	91	1	fig	fig	NOUN
aiti-8904	91	2	.	.	PUNCT
aiti-8904	92	1	5(a	5(a	NUM
aiti-8904	92	2	)	)	PUNCT
aiti-8904	92	3	and	and	CCONJ
aiti-8904	92	4	fig	fig	NOUN
aiti-8904	92	5	.	.	PUNCT
aiti-8904	93	1	6(a	6(a	X
aiti-8904	93	2	)	)	PUNCT
aiti-8904	94	1	show	show	VERB
aiti-8904	94	2	the	the	DET
aiti-8904	94	3	data	datum	NOUN
aiti-8904	94	4	collected	collect	VERB
aiti-8904	94	5	from	from	ADP
aiti-8904	94	6	the	the	DET
aiti-8904	94	7	interval	interval	NOUN
aiti-8904	94	8	between	between	ADP
aiti-8904	94	9	opening	open	VERB
aiti-8904	94	10	the	the	DET
aiti-8904	94	11	leak	leak	NOUN
aiti-8904	94	12	valve	valve	NOUN
aiti-8904	94	13	instantly	instantly	ADV
aiti-8904	94	14	and	and	CCONJ
aiti-8904	94	15	waiting	wait	VERB
aiti-8904	94	16	for	for	ADP
aiti-8904	94	17	30	30	NUM
aiti-8904	94	18	seconds	second	NOUN
aiti-8904	94	19	to	to	PART
aiti-8904	94	20	close	close	VERB
aiti-8904	94	21	the	the	DET
aiti-8904	94	22	leak	leak	NOUN
aiti-8904	94	23	valve	valve	NOUN
aiti-8904	94	24	instantly	instantly	ADV
aiti-8904	94	25	,	,	PUNCT
aiti-8904	94	26	respectively	respectively	ADV
aiti-8904	94	27	.	.	PUNCT
aiti-8904	95	1	the	the	DET
aiti-8904	95	2	curves	curve	NOUN
aiti-8904	95	3	marked	mark	VERB
aiti-8904	95	4	171	171	NUM
aiti-8904	95	5	advances	advance	NOUN
aiti-8904	95	6	in	in	ADP
aiti-8904	95	7	technology	technology	NOUN
aiti-8904	95	8	innovation	innovation	NOUN
aiti-8904	95	9	,	,	PUNCT
aiti-8904	95	10	vol	vol	NOUN
aiti-8904	95	11	.	.	PROPN
aiti-8904	96	1	7	7	NUM
aiti-8904	96	2	,	,	PUNCT
aiti-8904	96	3	no	no	INTJ
aiti-8904	96	4	.	.	NOUN
aiti-8904	96	5	3	3	NUM
aiti-8904	96	6	,	,	PUNCT
aiti-8904	96	7	2022	2022	NUM
aiti-8904	96	8	,	,	PUNCT
aiti-8904	96	9	pp	pp	ADJ
aiti-8904	96	10	.	.	PUNCT
aiti-8904	97	1	169	169	NUM
aiti-8904	97	2	-	-	SYM
aiti-8904	97	3	180	180	NUM
aiti-8904	97	4	with	with	ADP
aiti-8904	97	5	red	red	ADJ
aiti-8904	97	6	squares	square	NOUN
aiti-8904	97	7	in	in	ADP
aiti-8904	97	8	the	the	DET
aiti-8904	97	9	two	two	NUM
aiti-8904	97	10	figures	figure	NOUN
aiti-8904	97	11	are	be	AUX
aiti-8904	97	12	the	the	DET
aiti-8904	97	13	pressure	pressure	NOUN
aiti-8904	97	14	fluctuations	fluctuation	NOUN
aiti-8904	97	15	during	during	ADP
aiti-8904	97	16	the	the	DET
aiti-8904	97	17	leak	leak	NOUN
aiti-8904	97	18	test	test	NOUN
aiti-8904	97	19	.	.	PUNCT
aiti-8904	98	1	as	as	SCONJ
aiti-8904	98	2	shown	show	VERB
aiti-8904	98	3	in	in	ADP
aiti-8904	98	4	fig	fig	NOUN
aiti-8904	98	5	.	.	PUNCT
aiti-8904	99	1	5(a	5(a	NUM
aiti-8904	99	2	)	)	PUNCT
aiti-8904	99	3	and	and	CCONJ
aiti-8904	99	4	fig	fig	NOUN
aiti-8904	99	5	.	.	PUNCT
aiti-8904	100	1	6(a	6(a	NUM
aiti-8904	100	2	)	)	PUNCT
aiti-8904	100	3	,	,	PUNCT
aiti-8904	100	4	the	the	DET
aiti-8904	100	5	width	width	NOUN
aiti-8904	100	6	of	of	ADP
aiti-8904	100	7	the	the	DET
aiti-8904	100	8	red	red	ADJ
aiti-8904	100	9	region	region	NOUN
aiti-8904	100	10	is	be	AUX
aiti-8904	100	11	the	the	DET
aiti-8904	100	12	period	period	NOUN
aiti-8904	100	13	for	for	ADP
aiti-8904	100	14	the	the	DET
aiti-8904	100	15	opening	opening	NOUN
aiti-8904	100	16	and	and	CCONJ
aiti-8904	100	17	closing	closing	NOUN
aiti-8904	100	18	of	of	ADP
aiti-8904	100	19	the	the	DET
aiti-8904	100	20	valve	valve	NOUN
aiti-8904	100	21	and	and	CCONJ
aiti-8904	100	22	is	be	AUX
aiti-8904	100	23	enlarged	enlarge	VERB
aiti-8904	100	24	in	in	ADP
aiti-8904	100	25	fig	fig	NOUN
aiti-8904	100	26	.	.	PUNCT
aiti-8904	101	1	5(b	5(b	NUM
aiti-8904	101	2	)	)	PUNCT
aiti-8904	101	3	and	and	CCONJ
aiti-8904	101	4	fig	fig	NOUN
aiti-8904	101	5	.	.	PUNCT
aiti-8904	102	1	6(b	6(b	NUM
aiti-8904	102	2	)	)	PUNCT
aiti-8904	102	3	.	.	PUNCT
aiti-8904	103	1	the	the	DET
aiti-8904	103	2	pressure	pressure	NOUN
aiti-8904	103	3	variance	variance	NOUN
aiti-8904	103	4	of	of	ADP
aiti-8904	103	5	the	the	DET
aiti-8904	103	6	first	first	ADJ
aiti-8904	103	7	leak	leak	NOUN
aiti-8904	103	8	test	test	NOUN
aiti-8904	103	9	and	and	CCONJ
aiti-8904	103	10	the	the	DET
aiti-8904	103	11	second	second	ADJ
aiti-8904	103	12	leak	leak	NOUN
aiti-8904	103	13	test	test	NOUN
aiti-8904	103	14	are	be	AUX
aiti-8904	103	15	shown	show	VERB
aiti-8904	103	16	in	in	ADP
aiti-8904	103	17	fig	fig	NOUN
aiti-8904	103	18	.	.	PUNCT
aiti-8904	104	1	5(b	5(b	NUM
aiti-8904	104	2	)	)	PUNCT
aiti-8904	104	3	and	and	CCONJ
aiti-8904	104	4	fig	fig	NOUN
aiti-8904	104	5	.	.	PUNCT
aiti-8904	105	1	6(b	6(b	NUM
aiti-8904	105	2	)	)	PUNCT
aiti-8904	105	3	,	,	PUNCT
aiti-8904	105	4	respectively	respectively	ADV
aiti-8904	105	5	.	.	PUNCT
aiti-8904	106	1	as	as	SCONJ
aiti-8904	106	2	depicted	depict	VERB
aiti-8904	106	3	in	in	ADP
aiti-8904	106	4	fig	fig	NOUN
aiti-8904	106	5	.	.	PUNCT
aiti-8904	107	1	6(a	6(a	NUM
aiti-8904	107	2	)	)	PUNCT
aiti-8904	107	3	,	,	PUNCT
aiti-8904	107	4	the	the	DET
aiti-8904	107	5	internal	internal	ADJ
aiti-8904	107	6	pressure	pressure	NOUN
aiti-8904	107	7	of	of	ADP
aiti-8904	107	8	the	the	DET
aiti-8904	107	9	pipeline	pipeline	NOUN
aiti-8904	107	10	does	do	AUX
aiti-8904	107	11	not	not	PART
aiti-8904	107	12	return	return	VERB
aiti-8904	107	13	to	to	ADP
aiti-8904	107	14	a	a	DET
aiti-8904	107	15	stable	stable	ADJ
aiti-8904	107	16	state	state	NOUN
aiti-8904	107	17	during	during	ADP
aiti-8904	107	18	the	the	DET
aiti-8904	107	19	second	second	ADJ
aiti-8904	107	20	leak	leak	NOUN
aiti-8904	107	21	test	test	NOUN
aiti-8904	107	22	.	.	PUNCT
aiti-8904	108	1	therefore	therefore	ADV
aiti-8904	108	2	,	,	PUNCT
aiti-8904	108	3	the	the	DET
aiti-8904	108	4	level	level	NOUN
aiti-8904	108	5	of	of	ADP
aiti-8904	108	6	pressure	pressure	NOUN
aiti-8904	108	7	fluctuation	fluctuation	NOUN
aiti-8904	108	8	in	in	ADP
aiti-8904	108	9	the	the	DET
aiti-8904	108	10	second	second	ADJ
aiti-8904	108	11	test	test	NOUN
aiti-8904	108	12	is	be	AUX
aiti-8904	108	13	smaller	small	ADJ
aiti-8904	108	14	than	than	ADP
aiti-8904	108	15	in	in	ADP
aiti-8904	108	16	the	the	DET
aiti-8904	108	17	first	first	ADJ
aiti-8904	108	18	test	test	NOUN
aiti-8904	108	19	.	.	PUNCT
aiti-8904	109	1	fig	fig	NOUN
aiti-8904	109	2	.	.	PUNCT
aiti-8904	110	1	7(a	7(a	NUM
aiti-8904	110	2	)	)	PUNCT
aiti-8904	111	1	shows	show	VERB
aiti-8904	111	2	the	the	DET
aiti-8904	111	3	data	datum	NOUN
aiti-8904	111	4	collected	collect	VERB
aiti-8904	111	5	from	from	ADP
aiti-8904	111	6	the	the	DET
aiti-8904	111	7	pressure	pressure	NOUN
aiti-8904	111	8	sensor	sensor	NOUN
aiti-8904	111	9	during	during	ADP
aiti-8904	111	10	the	the	DET
aiti-8904	111	11	third	third	ADJ
aiti-8904	111	12	leak	leak	NOUN
aiti-8904	111	13	test	test	NOUN
aiti-8904	111	14	.	.	PUNCT
aiti-8904	112	1	in	in	ADP
aiti-8904	112	2	the	the	DET
aiti-8904	112	3	third	third	ADJ
aiti-8904	112	4	leak	leak	NOUN
aiti-8904	112	5	test	test	NOUN
aiti-8904	112	6	,	,	PUNCT
aiti-8904	112	7	the	the	DET
aiti-8904	112	8	data	data	NOUN
aiti-8904	112	9	collection	collection	NOUN
aiti-8904	112	10	process	process	NOUN
aiti-8904	112	11	begins	begin	VERB
aiti-8904	112	12	when	when	SCONJ
aiti-8904	112	13	the	the	DET
aiti-8904	112	14	leaking	leak	VERB
aiti-8904	112	15	valve	valve	NOUN
aiti-8904	112	16	is	be	AUX
aiti-8904	112	17	opened	open	VERB
aiti-8904	112	18	at	at	ADP
aiti-8904	112	19	a	a	DET
aiti-8904	112	20	slower	slow	ADJ
aiti-8904	112	21	speed	speed	NOUN
aiti-8904	112	22	and	and	CCONJ
aiti-8904	112	23	ends	end	VERB
aiti-8904	112	24	when	when	SCONJ
aiti-8904	112	25	the	the	DET
aiti-8904	112	26	valve	valve	NOUN
aiti-8904	112	27	is	be	AUX
aiti-8904	112	28	fully	fully	ADV
aiti-8904	112	29	opened	open	VERB
aiti-8904	112	30	.	.	PUNCT
aiti-8904	113	1	the	the	DET
aiti-8904	113	2	time	time	NOUN
aiti-8904	113	3	interval	interval	NOUN
aiti-8904	113	4	for	for	ADP
aiti-8904	113	5	the	the	DET
aiti-8904	113	6	opening	opening	NOUN
aiti-8904	113	7	and	and	CCONJ
aiti-8904	113	8	closing	closing	NOUN
aiti-8904	113	9	of	of	ADP
aiti-8904	113	10	the	the	DET
aiti-8904	113	11	valve	valve	NOUN
aiti-8904	113	12	is	be	AUX
aiti-8904	113	13	four	four	NUM
aiti-8904	113	14	minutes	minute	NOUN
aiti-8904	113	15	.	.	PUNCT
aiti-8904	114	1	after	after	ADP
aiti-8904	114	2	four	four	NUM
aiti-8904	114	3	minutes	minute	NOUN
aiti-8904	114	4	,	,	PUNCT
aiti-8904	114	5	the	the	DET
aiti-8904	114	6	leakage	leakage	NOUN
aiti-8904	114	7	valve	valve	NOUN
aiti-8904	114	8	is	be	AUX
aiti-8904	114	9	closed	close	VERB
aiti-8904	114	10	at	at	ADP
aiti-8904	114	11	a	a	DET
aiti-8904	114	12	slow	slow	ADJ
aiti-8904	114	13	speed	speed	NOUN
aiti-8904	114	14	.	.	PUNCT
aiti-8904	115	1	the	the	DET
aiti-8904	115	2	width	width	NOUN
aiti-8904	115	3	of	of	ADP
aiti-8904	115	4	the	the	DET
aiti-8904	115	5	red	red	ADJ
aiti-8904	115	6	region	region	NOUN
aiti-8904	115	7	in	in	ADP
aiti-8904	115	8	fig	fig	NOUN
aiti-8904	115	9	.	.	PUNCT
aiti-8904	116	1	7(a	7(a	NUM
aiti-8904	116	2	)	)	PUNCT
aiti-8904	116	3	is	be	AUX
aiti-8904	116	4	the	the	DET
aiti-8904	116	5	period	period	NOUN
aiti-8904	116	6	for	for	ADP
aiti-8904	116	7	the	the	DET
aiti-8904	116	8	opening	opening	NOUN
aiti-8904	116	9	and	and	CCONJ
aiti-8904	116	10	closing	closing	NOUN
aiti-8904	116	11	of	of	ADP
aiti-8904	116	12	the	the	DET
aiti-8904	116	13	valve	valve	NOUN
aiti-8904	116	14	and	and	CCONJ
aiti-8904	116	15	is	be	AUX
aiti-8904	116	16	enlarged	enlarge	VERB
aiti-8904	116	17	in	in	ADP
aiti-8904	116	18	fig	fig	NOUN
aiti-8904	116	19	.	.	PUNCT
aiti-8904	117	1	7(b	7(b	X
aiti-8904	117	2	)	)	PUNCT
aiti-8904	117	3	.	.	PUNCT
aiti-8904	118	1	thus	thus	ADV
aiti-8904	118	2	,	,	PUNCT
aiti-8904	118	3	the	the	DET
aiti-8904	118	4	pressure	pressure	NOUN
aiti-8904	118	5	variance	variance	NOUN
aiti-8904	118	6	of	of	ADP
aiti-8904	118	7	the	the	DET
aiti-8904	118	8	third	third	ADJ
aiti-8904	118	9	leak	leak	NOUN
aiti-8904	118	10	test	test	NOUN
aiti-8904	118	11	can	can	AUX
aiti-8904	118	12	be	be	AUX
aiti-8904	118	13	clearly	clearly	ADV
aiti-8904	118	14	seen	see	VERB
aiti-8904	118	15	in	in	ADP
aiti-8904	118	16	fig	fig	NOUN
aiti-8904	118	17	.	.	PUNCT
aiti-8904	119	1	7(b	7(b	X
aiti-8904	119	2	)	)	PUNCT
aiti-8904	119	3	.	.	PUNCT
aiti-8904	120	1	fig	fig	NOUN
aiti-8904	120	2	.	.	PUNCT
aiti-8904	121	1	4	4	NUM
aiti-8904	121	2	illustration	illustration	NOUN
aiti-8904	121	3	of	of	ADP
aiti-8904	121	4	the	the	DET
aiti-8904	121	5	pressure	pressure	NOUN
aiti-8904	121	6	data	datum	NOUN
aiti-8904	121	7	for	for	ADP
aiti-8904	121	8	detecting	detect	VERB
aiti-8904	121	9	real	real	ADJ
aiti-8904	121	10	pipeline	pipeline	NOUN
aiti-8904	121	11	leaks	leak	NOUN
aiti-8904	121	12	(	(	PUNCT
aiti-8904	121	13	a	a	X
aiti-8904	121	14	)	)	PUNCT
aiti-8904	121	15	result	result	NOUN
aiti-8904	121	16	of	of	ADP
aiti-8904	121	17	the	the	DET
aiti-8904	121	18	first	first	ADJ
aiti-8904	121	19	leak	leak	NOUN
aiti-8904	121	20	test	test	NOUN
aiti-8904	121	21	showing	show	VERB
aiti-8904	121	22	leak	leak	NOUN
aiti-8904	121	23	point	point	NOUN
aiti-8904	121	24	1	1	NUM
aiti-8904	121	25	(	(	PUNCT
aiti-8904	121	26	b	b	NOUN
aiti-8904	121	27	)	)	PUNCT
aiti-8904	121	28	close	close	ADJ
aiti-8904	121	29	-	-	PUNCT
aiti-8904	121	30	up	up	ADP
aiti-8904	121	31	view	view	NOUN
aiti-8904	121	32	of	of	ADP
aiti-8904	121	33	the	the	DET
aiti-8904	121	34	red	red	ADJ
aiti-8904	121	35	region	region	NOUN
aiti-8904	121	36	in	in	ADP
aiti-8904	121	37	fig	fig	NOUN
aiti-8904	121	38	.	.	PUNCT
aiti-8904	122	1	5(a	5(a	NUM
aiti-8904	122	2	)	)	PUNCT
aiti-8904	122	3	(	(	PUNCT
aiti-8904	122	4	showing	show	VERB
aiti-8904	122	5	the	the	DET
aiti-8904	122	6	test	test	NOUN
aiti-8904	122	7	interval	interval	NOUN
aiti-8904	122	8	)	)	PUNCT
aiti-8904	122	9	fig	fig	NOUN
aiti-8904	122	10	.	.	PUNCT
aiti-8904	123	1	5	5	NUM
aiti-8904	123	2	the	the	DET
aiti-8904	123	3	first	first	ADJ
aiti-8904	123	4	leak	leak	NOUN
aiti-8904	123	5	test	test	NOUN
aiti-8904	123	6	(	(	PUNCT
aiti-8904	123	7	a	a	X
aiti-8904	123	8	)	)	PUNCT
aiti-8904	123	9	result	result	NOUN
aiti-8904	123	10	of	of	ADP
aiti-8904	123	11	the	the	DET
aiti-8904	123	12	second	second	ADJ
aiti-8904	123	13	leak	leak	NOUN
aiti-8904	123	14	test	test	NOUN
aiti-8904	123	15	showing	show	VERB
aiti-8904	123	16	leak	leak	NOUN
aiti-8904	123	17	point	point	NOUN
aiti-8904	123	18	2	2	NUM
aiti-8904	123	19	fig	fig	NOUN
aiti-8904	123	20	.	.	PUNCT
aiti-8904	124	1	6	6	NUM
aiti-8904	124	2	the	the	DET
aiti-8904	124	3	second	second	ADJ
aiti-8904	124	4	leak	leak	NOUN
aiti-8904	124	5	test	test	VERB
aiti-8904	124	6	172	172	NUM
aiti-8904	124	7	advances	advance	NOUN
aiti-8904	124	8	in	in	ADP
aiti-8904	124	9	technology	technology	NOUN
aiti-8904	124	10	innovation	innovation	NOUN
aiti-8904	124	11	,	,	PUNCT
aiti-8904	124	12	vol	vol	NOUN
aiti-8904	124	13	.	.	PROPN
aiti-8904	125	1	7	7	NUM
aiti-8904	125	2	,	,	PUNCT
aiti-8904	125	3	no	no	INTJ
aiti-8904	125	4	.	.	NOUN
aiti-8904	125	5	3	3	NUM
aiti-8904	125	6	,	,	PUNCT
aiti-8904	125	7	2022	2022	NUM
aiti-8904	125	8	,	,	PUNCT
aiti-8904	125	9	pp	pp	ADJ
aiti-8904	125	10	.	.	PUNCT
aiti-8904	126	1	169	169	NUM
aiti-8904	126	2	-	-	SYM
aiti-8904	126	3	180	180	NUM
aiti-8904	126	4	(	(	PUNCT
aiti-8904	126	5	b	b	NOUN
aiti-8904	126	6	)	)	PUNCT
aiti-8904	126	7	close	close	ADJ
aiti-8904	126	8	-	-	PUNCT
aiti-8904	126	9	up	up	ADP
aiti-8904	126	10	view	view	NOUN
aiti-8904	126	11	of	of	ADP
aiti-8904	126	12	the	the	DET
aiti-8904	126	13	red	red	ADJ
aiti-8904	126	14	region	region	NOUN
aiti-8904	126	15	in	in	ADP
aiti-8904	126	16	fig	fig	NOUN
aiti-8904	126	17	.	.	PUNCT
aiti-8904	127	1	6(a	6(a	NUM
aiti-8904	127	2	)	)	PUNCT
aiti-8904	127	3	(	(	PUNCT
aiti-8904	127	4	showing	show	VERB
aiti-8904	127	5	the	the	DET
aiti-8904	127	6	test	test	NOUN
aiti-8904	127	7	interval	interval	NOUN
aiti-8904	127	8	)	)	PUNCT
aiti-8904	127	9	fig	fig	NOUN
aiti-8904	127	10	.	.	PUNCT
aiti-8904	128	1	6	6	NUM
aiti-8904	128	2	the	the	DET
aiti-8904	128	3	second	second	ADJ
aiti-8904	128	4	leak	leak	NOUN
aiti-8904	128	5	test	test	NOUN
aiti-8904	128	6	(	(	PUNCT
aiti-8904	128	7	continued	continue	VERB
aiti-8904	128	8	)	)	PUNCT
aiti-8904	128	9	(	(	PUNCT
aiti-8904	128	10	a	a	X
aiti-8904	128	11	)	)	PUNCT
aiti-8904	128	12	result	result	NOUN
aiti-8904	128	13	of	of	ADP
aiti-8904	128	14	the	the	DET
aiti-8904	128	15	third	third	ADJ
aiti-8904	128	16	leak	leak	NOUN
aiti-8904	128	17	test	test	NOUN
aiti-8904	128	18	showing	show	VERB
aiti-8904	128	19	leak	leak	NOUN
aiti-8904	128	20	point	point	NOUN
aiti-8904	128	21	3	3	NUM
aiti-8904	128	22	(	(	PUNCT
aiti-8904	128	23	b	b	NOUN
aiti-8904	128	24	)	)	PUNCT
aiti-8904	128	25	close	close	ADJ
aiti-8904	128	26	-	-	PUNCT
aiti-8904	128	27	up	up	ADP
aiti-8904	128	28	view	view	NOUN
aiti-8904	128	29	of	of	ADP
aiti-8904	128	30	the	the	DET
aiti-8904	128	31	red	red	ADJ
aiti-8904	128	32	region	region	NOUN
aiti-8904	128	33	in	in	ADP
aiti-8904	128	34	fig	fig	NOUN
aiti-8904	128	35	.	.	PUNCT
aiti-8904	129	1	7(a	7(a	NUM
aiti-8904	129	2	)	)	PUNCT
aiti-8904	129	3	(	(	PUNCT
aiti-8904	129	4	showing	show	VERB
aiti-8904	129	5	the	the	DET
aiti-8904	129	6	test	test	NOUN
aiti-8904	129	7	interval	interval	NOUN
aiti-8904	129	8	)	)	PUNCT
aiti-8904	129	9	fig	fig	NOUN
aiti-8904	129	10	.	.	PUNCT
aiti-8904	130	1	7	7	NUM
aiti-8904	131	1	the	the	DET
aiti-8904	131	2	third	third	ADJ
aiti-8904	131	3	leak	leak	NOUN
aiti-8904	131	4	test	test	NOUN
aiti-8904	131	5	fig	fig	NOUN
aiti-8904	131	6	.	.	PUNCT
aiti-8904	132	1	8	8	NUM
aiti-8904	132	2	fod	fod	ADJ
aiti-8904	132	3	processing	processing	NOUN
aiti-8904	132	4	result	result	NOUN
aiti-8904	132	5	for	for	ADP
aiti-8904	132	6	the	the	DET
aiti-8904	132	7	data	datum	NOUN
aiti-8904	132	8	of	of	ADP
aiti-8904	132	9	real	real	ADJ
aiti-8904	132	10	pipeline	pipeline	NOUN
aiti-8904	132	11	leaks	leak	NOUN
aiti-8904	132	12	after	after	ADP
aiti-8904	132	13	training	train	VERB
aiti-8904	132	14	the	the	DET
aiti-8904	132	15	model	model	NOUN
aiti-8904	132	16	many	many	ADJ
aiti-8904	132	17	times	time	NOUN
aiti-8904	132	18	,	,	PUNCT
aiti-8904	132	19	even	even	ADV
aiti-8904	132	20	if	if	SCONJ
aiti-8904	132	21	the	the	DET
aiti-8904	132	22	number	number	NOUN
aiti-8904	132	23	of	of	ADP
aiti-8904	132	24	the	the	DET
aiti-8904	132	25	convolutional	convolutional	ADJ
aiti-8904	132	26	layers	layer	NOUN
aiti-8904	132	27	,	,	PUNCT
aiti-8904	132	28	the	the	DET
aiti-8904	132	29	number	number	NOUN
aiti-8904	132	30	of	of	ADP
aiti-8904	132	31	the	the	DET
aiti-8904	132	32	filters	filter	NOUN
aiti-8904	132	33	,	,	PUNCT
aiti-8904	132	34	and	and	CCONJ
aiti-8904	132	35	the	the	DET
aiti-8904	132	36	size	size	NOUN
aiti-8904	132	37	of	of	ADP
aiti-8904	132	38	the	the	DET
aiti-8904	132	39	convolution	convolution	NOUN
aiti-8904	132	40	kernel	kernel	NOUN
aiti-8904	132	41	are	be	AUX
aiti-8904	132	42	increased	increase	VERB
aiti-8904	132	43	,	,	PUNCT
aiti-8904	132	44	the	the	DET
aiti-8904	132	45	resisting	resist	VERB
aiti-8904	132	46	noise	noise	NOUN
aiti-8904	132	47	interference	interference	NOUN
aiti-8904	132	48	does	do	AUX
aiti-8904	132	49	not	not	PART
aiti-8904	132	50	reach	reach	VERB
aiti-8904	132	51	the	the	DET
aiti-8904	132	52	requirement	requirement	NOUN
aiti-8904	132	53	.	.	PUNCT
aiti-8904	133	1	the	the	DET
aiti-8904	133	2	result	result	NOUN
aiti-8904	133	3	presents	present	VERB
aiti-8904	133	4	that	that	SCONJ
aiti-8904	133	5	a	a	DET
aiti-8904	133	6	single	single	ADJ
aiti-8904	133	7	feature	feature	NOUN
aiti-8904	133	8	is	be	AUX
aiti-8904	133	9	less	less	ADV
aiti-8904	133	10	effective	effective	ADJ
aiti-8904	133	11	in	in	ADP
aiti-8904	133	12	resisting	resist	VERB
aiti-8904	133	13	noise	noise	NOUN
aiti-8904	133	14	interference	interference	NOUN
aiti-8904	133	15	.	.	PUNCT
aiti-8904	134	1	thus	thus	ADV
aiti-8904	134	2	,	,	PUNCT
aiti-8904	134	3	the	the	DET
aiti-8904	134	4	data	data	NOUN
aiti-8904	134	5	is	be	AUX
aiti-8904	134	6	processed	process	VERB
aiti-8904	134	7	by	by	ADP
aiti-8904	134	8	the	the	DET
aiti-8904	134	9	first	first	ADJ
aiti-8904	134	10	-	-	PUNCT
aiti-8904	134	11	order	order	NOUN
aiti-8904	134	12	differential	differential	NOUN
aiti-8904	134	13	(	(	PUNCT
aiti-8904	134	14	fod	fod	ADJ
aiti-8904	134	15	)	)	PUNCT
aiti-8904	134	16	processing	processing	NOUN
aiti-8904	134	17	and	and	CCONJ
aiti-8904	134	18	its	its	PRON
aiti-8904	134	19	characteristics	characteristic	NOUN
aiti-8904	134	20	are	be	AUX
aiti-8904	134	21	analyzed	analyze	VERB
aiti-8904	134	22	.	.	PUNCT
aiti-8904	135	1	as	as	SCONJ
aiti-8904	135	2	shown	show	VERB
aiti-8904	135	3	in	in	ADP
aiti-8904	135	4	fig	fig	NOUN
aiti-8904	135	5	.	.	PUNCT
aiti-8904	136	1	8	8	NUM
aiti-8904	136	2	,	,	PUNCT
aiti-8904	136	3	the	the	DET
aiti-8904	136	4	red	red	ADJ
aiti-8904	136	5	arrows	arrow	NOUN
aiti-8904	136	6	point	point	VERB
aiti-8904	136	7	to	to	ADP
aiti-8904	136	8	the	the	DET
aiti-8904	136	9	data	data	NOUN
aiti-8904	136	10	points	point	NOUN
aiti-8904	136	11	of	of	ADP
aiti-8904	136	12	the	the	DET
aiti-8904	136	13	real	real	ADJ
aiti-8904	136	14	leak	leak	NOUN
aiti-8904	136	15	test	test	NOUN
aiti-8904	136	16	.	.	PUNCT
aiti-8904	137	1	when	when	SCONJ
aiti-8904	137	2	observing	observe	VERB
aiti-8904	137	3	the	the	DET
aiti-8904	137	4	curve	curve	NOUN
aiti-8904	137	5	of	of	ADP
aiti-8904	137	6	the	the	DET
aiti-8904	137	7	data	datum	NOUN
aiti-8904	137	8	analysis	analysis	NOUN
aiti-8904	137	9	after	after	ADP
aiti-8904	137	10	the	the	DET
aiti-8904	137	11	fod	fod	PROPN
aiti-8904	137	12	processing	processing	NOUN
aiti-8904	137	13	,	,	PUNCT
aiti-8904	137	14	it	it	PRON
aiti-8904	137	15	can	can	AUX
aiti-8904	137	16	be	be	AUX
aiti-8904	137	17	understood	understand	VERB
aiti-8904	137	18	that	that	SCONJ
aiti-8904	137	19	relatively	relatively	ADV
aiti-8904	137	20	large	large	ADJ
aiti-8904	137	21	pressure	pressure	NOUN
aiti-8904	137	22	fluctuations	fluctuation	NOUN
aiti-8904	137	23	will	will	AUX
aiti-8904	137	24	be	be	AUX
aiti-8904	137	25	generated	generate	VERB
aiti-8904	137	26	in	in	ADP
aiti-8904	137	27	the	the	DET
aiti-8904	137	28	interval	interval	NOUN
aiti-8904	137	29	of	of	ADP
aiti-8904	137	30	the	the	DET
aiti-8904	137	31	leak	leak	NOUN
aiti-8904	137	32	test	test	NOUN
aiti-8904	137	33	.	.	PUNCT
aiti-8904	138	1	thus	thus	ADV
aiti-8904	138	2	,	,	PUNCT
aiti-8904	138	3	these	these	DET
aiti-8904	138	4	features	feature	NOUN
aiti-8904	138	5	are	be	AUX
aiti-8904	138	6	adopted	adopt	VERB
aiti-8904	138	7	as	as	ADP
aiti-8904	138	8	the	the	DET
aiti-8904	138	9	second	second	ADJ
aiti-8904	138	10	feature	feature	NOUN
aiti-8904	138	11	of	of	ADP
aiti-8904	138	12	model	model	NOUN
aiti-8904	138	13	training	training	NOUN
aiti-8904	138	14	.	.	PUNCT
aiti-8904	139	1	as	as	SCONJ
aiti-8904	139	2	shown	show	VERB
aiti-8904	139	3	in	in	ADP
aiti-8904	139	4	fig	fig	NOUN
aiti-8904	139	5	.	.	PUNCT
aiti-8904	140	1	9	9	NUM
aiti-8904	140	2	,	,	PUNCT
aiti-8904	140	3	fig	fig	NOUN
aiti-8904	140	4	.	.	PUNCT
aiti-8904	141	1	10	10	NUM
aiti-8904	141	2	,	,	PUNCT
aiti-8904	141	3	and	and	CCONJ
aiti-8904	141	4	fig	fig	NOUN
aiti-8904	141	5	11	11	NUM
aiti-8904	141	6	,	,	PUNCT
aiti-8904	142	1	1000	1000	NUM
aiti-8904	142	2	points	point	NOUN
aiti-8904	142	3	of	of	ADP
aiti-8904	142	4	data	datum	NOUN
aiti-8904	142	5	extracted	extract	VERB
aiti-8904	142	6	from	from	ADP
aiti-8904	142	7	the	the	DET
aiti-8904	142	8	interval	interval	NOUN
aiti-8904	142	9	of	of	ADP
aiti-8904	142	10	the	the	DET
aiti-8904	142	11	first	first	ADJ
aiti-8904	142	12	,	,	PUNCT
aiti-8904	142	13	second	second	ADJ
aiti-8904	142	14	,	,	PUNCT
aiti-8904	142	15	and	and	CCONJ
aiti-8904	142	16	third	third	ADJ
aiti-8904	142	17	leak	leak	NOUN
aiti-8904	142	18	tests	test	NOUN
aiti-8904	142	19	are	be	AUX
aiti-8904	142	20	used	use	VERB
aiti-8904	142	21	to	to	PART
aiti-8904	142	22	analyze	analyze	VERB
aiti-8904	142	23	the	the	DET
aiti-8904	142	24	features	feature	NOUN
aiti-8904	142	25	.	.	PUNCT
aiti-8904	143	1	fig	fig	NOUN
aiti-8904	143	2	.	.	PUNCT
aiti-8904	144	1	9(a	9(a	NUM
aiti-8904	144	2	)	)	PUNCT
aiti-8904	144	3	,	,	PUNCT
aiti-8904	144	4	fig	fig	NOUN
aiti-8904	144	5	.	.	PUNCT
aiti-8904	145	1	10(a	10(a	NUM
aiti-8904	145	2	)	)	PUNCT
aiti-8904	145	3	,	,	PUNCT
aiti-8904	145	4	and	and	CCONJ
aiti-8904	145	5	fig	fig	NOUN
aiti-8904	145	6	.	.	PUNCT
aiti-8904	146	1	11(a	11(a	NUM
aiti-8904	146	2	)	)	PUNCT
aiti-8904	147	1	are	be	AUX
aiti-8904	147	2	the	the	DET
aiti-8904	147	3	fod	fod	PROPN
aiti-8904	147	4	data	datum	NOUN
aiti-8904	147	5	from	from	ADP
aiti-8904	147	6	the	the	DET
aiti-8904	147	7	first	first	ADJ
aiti-8904	147	8	leak	leak	NOUN
aiti-8904	147	9	test	test	NOUN
aiti-8904	147	10	,	,	PUNCT
aiti-8904	147	11	the	the	DET
aiti-8904	147	12	second	second	ADJ
aiti-8904	147	13	leak	leak	NOUN
aiti-8904	147	14	test	test	NOUN
aiti-8904	147	15	,	,	PUNCT
aiti-8904	147	16	and	and	CCONJ
aiti-8904	147	17	the	the	DET
aiti-8904	147	18	third	third	ADJ
aiti-8904	147	19	leak	leak	NOUN
aiti-8904	147	20	test	test	NOUN
aiti-8904	147	21	,	,	PUNCT
aiti-8904	147	22	respectively	respectively	ADV
aiti-8904	147	23	.	.	PUNCT
aiti-8904	148	1	the	the	DET
aiti-8904	148	2	width	width	NOUN
aiti-8904	148	3	of	of	ADP
aiti-8904	148	4	the	the	DET
aiti-8904	148	5	red	red	ADJ
aiti-8904	148	6	region	region	NOUN
aiti-8904	148	7	in	in	ADP
aiti-8904	148	8	fig	fig	NOUN
aiti-8904	148	9	.	.	PUNCT
aiti-8904	149	1	9(a	9(a	NUM
aiti-8904	149	2	)	)	PUNCT
aiti-8904	149	3	,	,	PUNCT
aiti-8904	149	4	fig	fig	NOUN
aiti-8904	149	5	.	.	PUNCT
aiti-8904	150	1	10(a	10(a	NUM
aiti-8904	150	2	)	)	PUNCT
aiti-8904	150	3	,	,	PUNCT
aiti-8904	150	4	and	and	CCONJ
aiti-8904	150	5	fig	fig	NOUN
aiti-8904	150	6	11(a	11(a	NUM
aiti-8904	150	7	)	)	PUNCT
aiti-8904	150	8	is	be	AUX
aiti-8904	150	9	the	the	DET
aiti-8904	150	10	period	period	NOUN
aiti-8904	150	11	for	for	ADP
aiti-8904	150	12	173	173	NUM
aiti-8904	150	13	advances	advance	NOUN
aiti-8904	150	14	in	in	ADP
aiti-8904	150	15	technology	technology	NOUN
aiti-8904	150	16	innovation	innovation	NOUN
aiti-8904	150	17	,	,	PUNCT
aiti-8904	150	18	vol	vol	NOUN
aiti-8904	150	19	.	.	PROPN
aiti-8904	151	1	7	7	NUM
aiti-8904	151	2	,	,	PUNCT
aiti-8904	151	3	no	no	INTJ
aiti-8904	151	4	.	.	NOUN
aiti-8904	151	5	3	3	NUM
aiti-8904	151	6	,	,	PUNCT
aiti-8904	151	7	2022	2022	NUM
aiti-8904	151	8	,	,	PUNCT
aiti-8904	151	9	pp	pp	ADJ
aiti-8904	151	10	.	.	PUNCT
aiti-8904	152	1	169	169	NUM
aiti-8904	152	2	-	-	SYM
aiti-8904	152	3	180	180	NUM
aiti-8904	152	4	the	the	DET
aiti-8904	152	5	opening	opening	NOUN
aiti-8904	152	6	and	and	CCONJ
aiti-8904	152	7	closing	closing	NOUN
aiti-8904	152	8	of	of	ADP
aiti-8904	152	9	the	the	DET
aiti-8904	152	10	valve	valve	NOUN
aiti-8904	152	11	and	and	CCONJ
aiti-8904	152	12	is	be	AUX
aiti-8904	152	13	enlarged	enlarge	VERB
aiti-8904	152	14	in	in	ADP
aiti-8904	152	15	fig	fig	NOUN
aiti-8904	152	16	.	.	PUNCT
aiti-8904	153	1	9(b	9(b	NUM
aiti-8904	153	2	)	)	PUNCT
aiti-8904	153	3	,	,	PUNCT
aiti-8904	153	4	fig	fig	NOUN
aiti-8904	153	5	.	.	PUNCT
aiti-8904	154	1	10(b	10(b	NUM
aiti-8904	154	2	)	)	PUNCT
aiti-8904	154	3	,	,	PUNCT
aiti-8904	154	4	and	and	CCONJ
aiti-8904	154	5	fig	fig	NOUN
aiti-8904	154	6	11(b	11(b	NUM
aiti-8904	154	7	)	)	PUNCT
aiti-8904	154	8	,	,	PUNCT
aiti-8904	154	9	respectively	respectively	ADV
aiti-8904	154	10	.	.	PUNCT
aiti-8904	155	1	taken	take	VERB
aiti-8904	155	2	together	together	ADV
aiti-8904	155	3	,	,	PUNCT
aiti-8904	155	4	after	after	SCONJ
aiti-8904	155	5	the	the	DET
aiti-8904	155	6	data	data	NOUN
aiti-8904	155	7	is	be	AUX
aiti-8904	155	8	transformed	transform	VERB
aiti-8904	155	9	by	by	ADP
aiti-8904	155	10	the	the	DET
aiti-8904	155	11	fod	fod	PROPN
aiti-8904	155	12	processing	processing	NOUN
aiti-8904	155	13	,	,	PUNCT
aiti-8904	155	14	the	the	DET
aiti-8904	155	15	pressure	pressure	NOUN
aiti-8904	155	16	fluctuation	fluctuation	NOUN
aiti-8904	155	17	produced	produce	VERB
aiti-8904	155	18	by	by	ADP
aiti-8904	155	19	the	the	DET
aiti-8904	155	20	first	first	ADJ
aiti-8904	155	21	leak	leak	NOUN
aiti-8904	155	22	test	test	NOUN
aiti-8904	155	23	is	be	AUX
aiti-8904	155	24	more	more	ADV
aiti-8904	155	25	obvious	obvious	ADJ
aiti-8904	155	26	than	than	ADP
aiti-8904	155	27	the	the	DET
aiti-8904	155	28	other	other	ADJ
aiti-8904	155	29	two	two	NUM
aiti-8904	155	30	.	.	PUNCT
aiti-8904	156	1	therefore	therefore	ADV
aiti-8904	156	2	,	,	PUNCT
aiti-8904	156	3	the	the	DET
aiti-8904	156	4	fod	fod	ADJ
aiti-8904	156	5	data	datum	NOUN
aiti-8904	156	6	of	of	ADP
aiti-8904	156	7	the	the	DET
aiti-8904	156	8	first	first	ADJ
aiti-8904	156	9	leak	leak	NOUN
aiti-8904	156	10	test	test	NOUN
aiti-8904	156	11	is	be	AUX
aiti-8904	156	12	selected	select	VERB
aiti-8904	156	13	as	as	ADP
aiti-8904	156	14	the	the	DET
aiti-8904	156	15	second	second	ADJ
aiti-8904	156	16	feature	feature	NOUN
aiti-8904	156	17	of	of	ADP
aiti-8904	156	18	the	the	DET
aiti-8904	156	19	training	training	NOUN
aiti-8904	156	20	model	model	NOUN
aiti-8904	156	21	of	of	ADP
aiti-8904	156	22	this	this	DET
aiti-8904	156	23	research	research	NOUN
aiti-8904	156	24	.	.	PUNCT
aiti-8904	157	1	(	(	PUNCT
aiti-8904	157	2	a	a	X
aiti-8904	157	3	)	)	PUNCT
aiti-8904	157	4	fod	fod	VERB
aiti-8904	157	5	processing	processing	NOUN
aiti-8904	157	6	result	result	NOUN
aiti-8904	157	7	of	of	ADP
aiti-8904	157	8	the	the	DET
aiti-8904	157	9	first	first	ADJ
aiti-8904	157	10	leak	leak	NOUN
aiti-8904	157	11	test	test	NOUN
aiti-8904	157	12	showing	show	VERB
aiti-8904	157	13	leak	leak	NOUN
aiti-8904	157	14	point	point	NOUN
aiti-8904	157	15	1	1	NUM
aiti-8904	157	16	(	(	PUNCT
aiti-8904	157	17	b	b	NOUN
aiti-8904	157	18	)	)	PUNCT
aiti-8904	157	19	close	close	ADJ
aiti-8904	157	20	-	-	PUNCT
aiti-8904	157	21	up	up	ADP
aiti-8904	157	22	view	view	NOUN
aiti-8904	157	23	of	of	ADP
aiti-8904	157	24	the	the	DET
aiti-8904	157	25	red	red	ADJ
aiti-8904	157	26	region	region	NOUN
aiti-8904	157	27	in	in	ADP
aiti-8904	157	28	fig	fig	NOUN
aiti-8904	157	29	.	.	PUNCT
aiti-8904	158	1	9(a	9(a	X
aiti-8904	158	2	)	)	PUNCT
aiti-8904	158	3	(	(	PUNCT
aiti-8904	158	4	showing	show	VERB
aiti-8904	158	5	the	the	DET
aiti-8904	158	6	test	test	NOUN
aiti-8904	158	7	interval	interval	NOUN
aiti-8904	158	8	)	)	PUNCT
aiti-8904	158	9	fig	fig	NOUN
aiti-8904	158	10	.	.	PUNCT
aiti-8904	159	1	9	9	NUM
aiti-8904	159	2	first	first	ADJ
aiti-8904	159	3	-	-	PUNCT
aiti-8904	159	4	order	order	NOUN
aiti-8904	159	5	data	datum	NOUN
aiti-8904	159	6	differentiation	differentiation	NOUN
aiti-8904	159	7	for	for	ADP
aiti-8904	159	8	the	the	DET
aiti-8904	159	9	first	first	ADJ
aiti-8904	159	10	leak	leak	NOUN
aiti-8904	159	11	test	test	NOUN
aiti-8904	159	12	(	(	PUNCT
aiti-8904	159	13	a	a	X
aiti-8904	159	14	)	)	PUNCT
aiti-8904	159	15	fod	fod	VERB
aiti-8904	159	16	processing	processing	NOUN
aiti-8904	159	17	result	result	NOUN
aiti-8904	159	18	of	of	ADP
aiti-8904	159	19	the	the	DET
aiti-8904	159	20	second	second	ADJ
aiti-8904	159	21	leak	leak	NOUN
aiti-8904	159	22	test	test	NOUN
aiti-8904	159	23	data	datum	NOUN
aiti-8904	159	24	showing	show	VERB
aiti-8904	159	25	leak	leak	NOUN
aiti-8904	159	26	point	point	NOUN
aiti-8904	159	27	2	2	NUM
aiti-8904	159	28	(	(	PUNCT
aiti-8904	159	29	b	b	NOUN
aiti-8904	159	30	)	)	PUNCT
aiti-8904	159	31	close	close	ADJ
aiti-8904	159	32	-	-	PUNCT
aiti-8904	159	33	up	up	ADP
aiti-8904	159	34	view	view	NOUN
aiti-8904	159	35	of	of	ADP
aiti-8904	159	36	the	the	DET
aiti-8904	159	37	red	red	ADJ
aiti-8904	159	38	region	region	NOUN
aiti-8904	159	39	in	in	ADP
aiti-8904	159	40	fig	fig	NOUN
aiti-8904	159	41	.	.	PUNCT
aiti-8904	160	1	10(a	10(a	NUM
aiti-8904	160	2	)	)	PUNCT
aiti-8904	160	3	(	(	PUNCT
aiti-8904	160	4	showing	show	VERB
aiti-8904	160	5	the	the	DET
aiti-8904	160	6	test	test	NOUN
aiti-8904	160	7	interval	interval	NOUN
aiti-8904	160	8	)	)	PUNCT
aiti-8904	160	9	fig	fig	NOUN
aiti-8904	160	10	.	.	PUNCT
aiti-8904	161	1	10	10	NUM
aiti-8904	161	2	first	first	ADJ
aiti-8904	161	3	-	-	PUNCT
aiti-8904	161	4	order	order	NOUN
aiti-8904	161	5	data	datum	NOUN
aiti-8904	161	6	differentiation	differentiation	NOUN
aiti-8904	161	7	for	for	ADP
aiti-8904	161	8	the	the	DET
aiti-8904	161	9	second	second	ADJ
aiti-8904	161	10	leak	leak	NOUN
aiti-8904	161	11	test	test	NOUN
aiti-8904	161	12	(	(	PUNCT
aiti-8904	161	13	a	a	X
aiti-8904	161	14	)	)	PUNCT
aiti-8904	161	15	fod	fod	VERB
aiti-8904	161	16	processing	processing	NOUN
aiti-8904	161	17	result	result	NOUN
aiti-8904	161	18	of	of	ADP
aiti-8904	161	19	the	the	DET
aiti-8904	161	20	third	third	ADJ
aiti-8904	161	21	leak	leak	NOUN
aiti-8904	161	22	test	test	NOUN
aiti-8904	161	23	data	datum	NOUN
aiti-8904	161	24	showing	show	VERB
aiti-8904	161	25	leak	leak	NOUN
aiti-8904	161	26	point	point	VERB
aiti-8904	161	27	3	3	NUM
aiti-8904	161	28	fig	fig	NOUN
aiti-8904	161	29	.	.	PUNCT
aiti-8904	162	1	11	11	NUM
aiti-8904	162	2	first	first	ADJ
aiti-8904	162	3	-	-	PUNCT
aiti-8904	162	4	order	order	NOUN
aiti-8904	162	5	data	datum	NOUN
aiti-8904	162	6	differentiation	differentiation	NOUN
aiti-8904	162	7	for	for	ADP
aiti-8904	162	8	the	the	DET
aiti-8904	162	9	third	third	ADJ
aiti-8904	162	10	leak	leak	NOUN
aiti-8904	162	11	test	test	VERB
aiti-8904	162	12	174	174	NUM
aiti-8904	162	13	advances	advance	NOUN
aiti-8904	162	14	in	in	ADP
aiti-8904	162	15	technology	technology	NOUN
aiti-8904	162	16	innovation	innovation	NOUN
aiti-8904	162	17	,	,	PUNCT
aiti-8904	162	18	vol	vol	NOUN
aiti-8904	162	19	.	.	PROPN
aiti-8904	163	1	7	7	NUM
aiti-8904	163	2	,	,	PUNCT
aiti-8904	163	3	no	no	INTJ
aiti-8904	163	4	.	.	NOUN
aiti-8904	163	5	3	3	NUM
aiti-8904	163	6	,	,	PUNCT
aiti-8904	163	7	2022	2022	NUM
aiti-8904	163	8	,	,	PUNCT
aiti-8904	163	9	pp	pp	ADJ
aiti-8904	163	10	.	.	PUNCT
aiti-8904	164	1	169	169	NUM
aiti-8904	164	2	-	-	SYM
aiti-8904	164	3	180	180	NUM
aiti-8904	164	4	(	(	PUNCT
aiti-8904	164	5	b	b	NOUN
aiti-8904	164	6	)	)	PUNCT
aiti-8904	164	7	close	close	ADJ
aiti-8904	164	8	-	-	PUNCT
aiti-8904	164	9	up	up	ADP
aiti-8904	164	10	view	view	NOUN
aiti-8904	164	11	of	of	ADP
aiti-8904	164	12	the	the	DET
aiti-8904	164	13	red	red	ADJ
aiti-8904	164	14	region	region	NOUN
aiti-8904	164	15	in	in	ADP
aiti-8904	164	16	fig	fig	NOUN
aiti-8904	164	17	.	.	PUNCT
aiti-8904	165	1	11(a	11(a	NUM
aiti-8904	165	2	)	)	PUNCT
aiti-8904	165	3	(	(	PUNCT
aiti-8904	165	4	showing	show	VERB
aiti-8904	165	5	the	the	DET
aiti-8904	165	6	test	test	NOUN
aiti-8904	165	7	interval	interval	NOUN
aiti-8904	165	8	)	)	PUNCT
aiti-8904	165	9	fig	fig	NOUN
aiti-8904	165	10	.	.	PUNCT
aiti-8904	166	1	11	11	NUM
aiti-8904	166	2	first	first	ADJ
aiti-8904	166	3	-	-	PUNCT
aiti-8904	166	4	order	order	NOUN
aiti-8904	166	5	data	datum	NOUN
aiti-8904	166	6	differentiation	differentiation	NOUN
aiti-8904	166	7	for	for	ADP
aiti-8904	166	8	the	the	DET
aiti-8904	166	9	third	third	ADJ
aiti-8904	166	10	leak	leak	NOUN
aiti-8904	166	11	test	test	NOUN
aiti-8904	166	12	(	(	PUNCT
aiti-8904	166	13	continued	continue	VERB
aiti-8904	166	14	)	)	PUNCT
aiti-8904	166	15	2.2	2.2	NUM
aiti-8904	166	16	.	.	PUNCT
aiti-8904	167	1	model	model	NOUN
aiti-8904	167	2	training	training	NOUN
aiti-8904	167	3	data	datum	NOUN
aiti-8904	167	4	set	set	VERB
aiti-8904	167	5	and	and	CCONJ
aiti-8904	167	6	testing	testing	NOUN
aiti-8904	167	7	data	datum	NOUN
aiti-8904	167	8	set	set	VERB
aiti-8904	167	9	according	accord	VERB
aiti-8904	167	10	to	to	ADP
aiti-8904	167	11	the	the	DET
aiti-8904	167	12	analysis	analysis	NOUN
aiti-8904	167	13	result	result	NOUN
aiti-8904	167	14	,	,	PUNCT
aiti-8904	167	15	a	a	DET
aiti-8904	167	16	suitable	suitable	ADJ
aiti-8904	167	17	data	data	NOUN
aiti-8904	167	18	interval	interval	NOUN
aiti-8904	167	19	for	for	ADP
aiti-8904	167	20	identifying	identify	VERB
aiti-8904	167	21	the	the	DET
aiti-8904	167	22	features	feature	NOUN
aiti-8904	167	23	of	of	ADP
aiti-8904	167	24	the	the	DET
aiti-8904	167	25	model	model	NOUN
aiti-8904	167	26	training	training	NOUN
aiti-8904	167	27	is	be	AUX
aiti-8904	167	28	found	find	VERB
aiti-8904	167	29	.	.	PUNCT
aiti-8904	168	1	the	the	DET
aiti-8904	168	2	normal	normal	ADJ
aiti-8904	168	3	data	datum	NOUN
aiti-8904	168	4	and	and	CCONJ
aiti-8904	168	5	the	the	DET
aiti-8904	168	6	leak	leak	NOUN
aiti-8904	168	7	data	datum	NOUN
aiti-8904	168	8	are	be	AUX
aiti-8904	168	9	given	give	VERB
aiti-8904	168	10	corresponding	corresponding	ADJ
aiti-8904	168	11	labels	label	NOUN
aiti-8904	168	12	.	.	PUNCT
aiti-8904	169	1	as	as	SCONJ
aiti-8904	169	2	presented	present	VERB
aiti-8904	169	3	in	in	ADP
aiti-8904	169	4	figs	fig	NOUN
aiti-8904	169	5	.	.	PUNCT
aiti-8904	170	1	12(a	12(a	NUM
aiti-8904	170	2	)	)	PUNCT
aiti-8904	170	3	and	and	CCONJ
aiti-8904	170	4	13(a	13(a	NUM
aiti-8904	170	5	)	)	PUNCT
aiti-8904	170	6	,	,	PUNCT
aiti-8904	170	7	the	the	DET
aiti-8904	170	8	data	datum	NOUN
aiti-8904	170	9	obtained	obtain	VERB
aiti-8904	170	10	from	from	ADP
aiti-8904	170	11	the	the	DET
aiti-8904	170	12	pressure	pressure	NOUN
aiti-8904	170	13	sensors	sensor	NOUN
aiti-8904	170	14	are	be	AUX
aiti-8904	170	15	normalized	normalize	VERB
aiti-8904	170	16	to	to	PART
aiti-8904	170	17	make	make	VERB
aiti-8904	170	18	the	the	DET
aiti-8904	170	19	value	value	NOUN
aiti-8904	170	20	with	with	ADP
aiti-8904	170	21	the	the	DET
aiti-8904	170	22	range	range	NOUN
aiti-8904	170	23	from	from	ADP
aiti-8904	170	24	0	0	NUM
aiti-8904	170	25	to	to	ADP
aiti-8904	170	26	1	1	NUM
aiti-8904	170	27	,	,	PUNCT
aiti-8904	170	28	so	so	ADV
aiti-8904	170	29	the	the	DET
aiti-8904	170	30	data	datum	NOUN
aiti-8904	170	31	in	in	ADP
aiti-8904	170	32	the	the	DET
aiti-8904	170	33	two	two	NUM
aiti-8904	170	34	figures	figure	NOUN
aiti-8904	170	35	are	be	AUX
aiti-8904	170	36	not	not	PART
aiti-8904	170	37	marked	mark	VERB
aiti-8904	170	38	with	with	ADP
aiti-8904	170	39	the	the	DET
aiti-8904	170	40	x	x	ADJ
aiti-8904	170	41	-	-	ADJ
aiti-8904	170	42	axis	axis	ADJ
aiti-8904	170	43	unit	unit	NOUN
aiti-8904	170	44	.	.	PUNCT
aiti-8904	171	1	because	because	SCONJ
aiti-8904	171	2	the	the	DET
aiti-8904	171	3	feature	feature	NOUN
aiti-8904	171	4	amount	amount	NOUN
aiti-8904	171	5	of	of	ADP
aiti-8904	171	6	the	the	DET
aiti-8904	171	7	leak	leak	NOUN
aiti-8904	171	8	data	datum	NOUN
aiti-8904	171	9	is	be	AUX
aiti-8904	171	10	less	less	ADJ
aiti-8904	171	11	than	than	ADP
aiti-8904	171	12	the	the	DET
aiti-8904	171	13	feature	feature	NOUN
aiti-8904	171	14	amount	amount	NOUN
aiti-8904	171	15	of	of	ADP
aiti-8904	171	16	the	the	DET
aiti-8904	171	17	normal	normal	ADJ
aiti-8904	171	18	data	datum	NOUN
aiti-8904	171	19	,	,	PUNCT
aiti-8904	171	20	data	datum	NOUN
aiti-8904	171	21	expansion	expansion	NOUN
aiti-8904	171	22	is	be	AUX
aiti-8904	171	23	needed	need	VERB
aiti-8904	171	24	to	to	PART
aiti-8904	171	25	achieve	achieve	VERB
aiti-8904	171	26	better	well	ADJ
aiti-8904	171	27	training	training	NOUN
aiti-8904	171	28	performance	performance	NOUN
aiti-8904	171	29	.	.	PUNCT
aiti-8904	172	1	in	in	ADP
aiti-8904	172	2	order	order	NOUN
aiti-8904	172	3	to	to	PART
aiti-8904	172	4	avoid	avoid	VERB
aiti-8904	172	5	the	the	DET
aiti-8904	172	6	model	model	NOUN
aiti-8904	172	7	from	from	ADP
aiti-8904	172	8	only	only	ADV
aiti-8904	172	9	learning	learn	VERB
aiti-8904	172	10	the	the	DET
aiti-8904	172	11	characteristics	characteristic	NOUN
aiti-8904	172	12	of	of	ADP
aiti-8904	172	13	a	a	DET
aiti-8904	172	14	single	single	ADJ
aiti-8904	172	15	data	datum	NOUN
aiti-8904	172	16	and	and	CCONJ
aiti-8904	172	17	causing	cause	VERB
aiti-8904	172	18	identification	identification	NOUN
aiti-8904	172	19	failure	failure	NOUN
aiti-8904	172	20	,	,	PUNCT
aiti-8904	172	21	the	the	DET
aiti-8904	172	22	data	datum	NOUN
aiti-8904	172	23	set	set	VERB
aiti-8904	172	24	is	be	AUX
aiti-8904	172	25	balanced	balanced	ADJ
aiti-8904	172	26	.	.	PUNCT
aiti-8904	173	1	finally	finally	ADV
aiti-8904	173	2	,	,	PUNCT
aiti-8904	173	3	70	70	NUM
aiti-8904	173	4	%	%	NOUN
aiti-8904	173	5	of	of	ADP
aiti-8904	173	6	the	the	DET
aiti-8904	173	7	normal	normal	ADJ
aiti-8904	173	8	data	datum	NOUN
aiti-8904	173	9	and	and	CCONJ
aiti-8904	173	10	the	the	DET
aiti-8904	173	11	leak	leak	NOUN
aiti-8904	173	12	data	datum	NOUN
aiti-8904	173	13	are	be	AUX
aiti-8904	173	14	used	use	VERB
aiti-8904	173	15	as	as	ADP
aiti-8904	173	16	the	the	DET
aiti-8904	173	17	training	training	NOUN
aiti-8904	173	18	data	datum	NOUN
aiti-8904	173	19	.	.	PUNCT
aiti-8904	174	1	the	the	DET
aiti-8904	174	2	rest	rest	NOUN
aiti-8904	174	3	of	of	ADP
aiti-8904	174	4	the	the	DET
aiti-8904	174	5	data	datum	NOUN
aiti-8904	174	6	is	be	AUX
aiti-8904	174	7	used	use	VERB
aiti-8904	174	8	as	as	ADP
aiti-8904	174	9	the	the	DET
aiti-8904	174	10	test	test	NOUN
aiti-8904	174	11	data	datum	NOUN
aiti-8904	174	12	set	set	VERB
aiti-8904	174	13	.	.	PUNCT
aiti-8904	175	1	the	the	DET
aiti-8904	175	2	result	result	NOUN
aiti-8904	175	3	of	of	ADP
aiti-8904	175	4	experimental	experimental	ADJ
aiti-8904	175	5	tests	test	NOUN
aiti-8904	175	6	shows	show	VERB
aiti-8904	175	7	that	that	SCONJ
aiti-8904	175	8	if	if	SCONJ
aiti-8904	175	9	the	the	DET
aiti-8904	175	10	data	datum	NOUN
aiti-8904	175	11	set	set	VERB
aiti-8904	175	12	is	be	AUX
aiti-8904	175	13	randomly	randomly	ADV
aiti-8904	175	14	arranged	arrange	VERB
aiti-8904	175	15	,	,	PUNCT
aiti-8904	175	16	it	it	PRON
aiti-8904	175	17	will	will	AUX
aiti-8904	175	18	reduce	reduce	VERB
aiti-8904	175	19	the	the	DET
aiti-8904	175	20	accuracy	accuracy	NOUN
aiti-8904	175	21	of	of	ADP
aiti-8904	175	22	the	the	DET
aiti-8904	175	23	overall	overall	ADJ
aiti-8904	175	24	model	model	NOUN
aiti-8904	175	25	classification	classification	NOUN
aiti-8904	175	26	.	.	PUNCT
aiti-8904	176	1	thus	thus	ADV
aiti-8904	176	2	,	,	PUNCT
aiti-8904	176	3	this	this	DET
aiti-8904	176	4	study	study	NOUN
aiti-8904	176	5	orders	order	VERB
aiti-8904	176	6	the	the	DET
aiti-8904	176	7	data	datum	NOUN
aiti-8904	176	8	according	accord	VERB
aiti-8904	176	9	to	to	ADP
aiti-8904	176	10	the	the	DET
aiti-8904	176	11	same	same	ADJ
aiti-8904	176	12	characteristics	characteristic	NOUN
aiti-8904	176	13	and	and	CCONJ
aiti-8904	176	14	by	by	ADP
aiti-8904	176	15	the	the	DET
aiti-8904	176	16	arrangement	arrangement	NOUN
aiti-8904	176	17	method	method	NOUN
aiti-8904	176	18	.	.	PUNCT
aiti-8904	177	1	as	as	SCONJ
aiti-8904	177	2	shown	show	VERB
aiti-8904	177	3	in	in	ADP
aiti-8904	177	4	fig	fig	NOUN
aiti-8904	177	5	.	.	PUNCT
aiti-8904	178	1	12(a	12(a	NUM
aiti-8904	178	2	)	)	PUNCT
aiti-8904	178	3	and	and	CCONJ
aiti-8904	178	4	fig	fig	NOUN
aiti-8904	178	5	.	.	PUNCT
aiti-8904	179	1	12(b	12(b	NUM
aiti-8904	179	2	)	)	PUNCT
aiti-8904	180	1	,	,	PUNCT
aiti-8904	180	2	the	the	DET
aiti-8904	180	3	first	first	ADJ
aiti-8904	180	4	half	half	NOUN
aiti-8904	180	5	of	of	ADP
aiti-8904	180	6	the	the	DET
aiti-8904	180	7	training	training	NOUN
aiti-8904	180	8	data	datum	NOUN
aiti-8904	180	9	set	set	VERB
aiti-8904	180	10	is	be	AUX
aiti-8904	180	11	normal	normal	ADJ
aiti-8904	180	12	data	datum	NOUN
aiti-8904	180	13	,	,	PUNCT
aiti-8904	180	14	and	and	CCONJ
aiti-8904	180	15	the	the	DET
aiti-8904	180	16	second	second	ADJ
aiti-8904	180	17	half	half	NOUN
aiti-8904	180	18	is	be	AUX
aiti-8904	180	19	leak	leak	NOUN
aiti-8904	180	20	data	datum	NOUN
aiti-8904	180	21	.	.	PUNCT
aiti-8904	181	1	fig	fig	NOUN
aiti-8904	181	2	.	.	PUNCT
aiti-8904	182	1	13(a	13(a	NUM
aiti-8904	182	2	)	)	PUNCT
aiti-8904	182	3	and	and	CCONJ
aiti-8904	182	4	fig	fig	NOUN
aiti-8904	182	5	.	.	PUNCT
aiti-8904	183	1	13(b	13(b	X
aiti-8904	183	2	)	)	PUNCT
aiti-8904	183	3	are	be	AUX
aiti-8904	183	4	the	the	DET
aiti-8904	183	5	testing	testing	NOUN
aiti-8904	183	6	data	data	NOUN
aiti-8904	183	7	sets	set	NOUN
aiti-8904	183	8	,	,	PUNCT
aiti-8904	183	9	which	which	PRON
aiti-8904	183	10	are	be	AUX
aiti-8904	183	11	used	use	VERB
aiti-8904	183	12	to	to	PART
aiti-8904	183	13	evaluate	evaluate	VERB
aiti-8904	183	14	the	the	DET
aiti-8904	183	15	effect	effect	NOUN
aiti-8904	183	16	of	of	ADP
aiti-8904	183	17	the	the	DET
aiti-8904	183	18	training	training	NOUN
aiti-8904	183	19	model	model	NOUN
aiti-8904	183	20	during	during	ADP
aiti-8904	183	21	model	model	NOUN
aiti-8904	183	22	training	training	NOUN
aiti-8904	183	23	.	.	PUNCT
aiti-8904	184	1	(	(	PUNCT
aiti-8904	184	2	a	a	X
aiti-8904	184	3	)	)	PUNCT
aiti-8904	184	4	original	original	ADJ
aiti-8904	184	5	training	training	NOUN
aiti-8904	184	6	data	datum	NOUN
aiti-8904	184	7	set	set	VERB
aiti-8904	184	8	(	(	PUNCT
aiti-8904	184	9	b	b	NOUN
aiti-8904	184	10	)	)	PUNCT
aiti-8904	184	11	fod	fod	VERB
aiti-8904	184	12	training	training	NOUN
aiti-8904	184	13	data	datum	NOUN
aiti-8904	184	14	set	set	VERB
aiti-8904	184	15	fig	fig	NOUN
aiti-8904	184	16	.	.	PUNCT
aiti-8904	185	1	12	12	NUM
aiti-8904	185	2	training	training	NOUN
aiti-8904	185	3	data	datum	NOUN
aiti-8904	185	4	set	set	VERB
aiti-8904	185	5	175	175	NUM
aiti-8904	185	6	advances	advance	NOUN
aiti-8904	185	7	in	in	ADP
aiti-8904	185	8	technology	technology	NOUN
aiti-8904	185	9	innovation	innovation	NOUN
aiti-8904	185	10	,	,	PUNCT
aiti-8904	185	11	vol	vol	NOUN
aiti-8904	185	12	.	.	PROPN
aiti-8904	186	1	7	7	NUM
aiti-8904	186	2	,	,	PUNCT
aiti-8904	186	3	no	no	INTJ
aiti-8904	186	4	.	.	NOUN
aiti-8904	186	5	3	3	NUM
aiti-8904	186	6	,	,	PUNCT
aiti-8904	186	7	2022	2022	NUM
aiti-8904	186	8	,	,	PUNCT
aiti-8904	186	9	pp	pp	ADJ
aiti-8904	186	10	.	.	PUNCT
aiti-8904	187	1	169	169	NUM
aiti-8904	187	2	-	-	SYM
aiti-8904	187	3	180	180	NUM
aiti-8904	187	4	(	(	PUNCT
aiti-8904	187	5	a	a	PRON
aiti-8904	187	6	)	)	PUNCT
aiti-8904	187	7	original	original	ADJ
aiti-8904	187	8	testing	testing	NOUN
aiti-8904	187	9	data	datum	NOUN
aiti-8904	187	10	set	set	NOUN
aiti-8904	187	11	(	(	PUNCT
aiti-8904	187	12	b	b	NOUN
aiti-8904	187	13	)	)	PUNCT
aiti-8904	187	14	fod	fod	VERB
aiti-8904	187	15	testing	testing	NOUN
aiti-8904	187	16	data	datum	NOUN
aiti-8904	187	17	set	set	VERB
aiti-8904	187	18	fig	fig	NOUN
aiti-8904	187	19	.	.	PUNCT
aiti-8904	188	1	13	13	NUM
aiti-8904	188	2	testing	testing	NOUN
aiti-8904	188	3	data	datum	NOUN
aiti-8904	188	4	set	set	VERB
aiti-8904	188	5	2.3	2.3	NUM
aiti-8904	188	6	.	.	PUNCT
aiti-8904	189	1	model	model	NOUN
aiti-8904	189	2	building	build	VERB
aiti-8904	189	3	this	this	DET
aiti-8904	189	4	study	study	NOUN
aiti-8904	189	5	uses	use	VERB
aiti-8904	189	6	keras	keras	PROPN
aiti-8904	189	7	to	to	PART
aiti-8904	189	8	construct	construct	VERB
aiti-8904	189	9	a	a	DET
aiti-8904	189	10	cnn	cnn	PROPN
aiti-8904	189	11	model	model	NOUN
aiti-8904	189	12	.	.	PUNCT
aiti-8904	190	1	the	the	DET
aiti-8904	190	2	values	value	NOUN
aiti-8904	190	3	collected	collect	VERB
aiti-8904	190	4	from	from	ADP
aiti-8904	190	5	the	the	DET
aiti-8904	190	6	pressure	pressure	NOUN
aiti-8904	190	7	sensor	sensor	NOUN
aiti-8904	190	8	are	be	AUX
aiti-8904	190	9	time	time	NOUN
aiti-8904	190	10	series	series	NOUN
aiti-8904	190	11	and	and	CCONJ
aiti-8904	190	12	belong	belong	VERB
aiti-8904	190	13	to	to	ADP
aiti-8904	190	14	a	a	DET
aiti-8904	190	15	one	one	NUM
aiti-8904	190	16	-	-	PUNCT
aiti-8904	190	17	dimensional	dimensional	ADJ
aiti-8904	190	18	data	data	NOUN
aiti-8904	190	19	type	type	NOUN
aiti-8904	190	20	.	.	PUNCT
aiti-8904	191	1	therefore	therefore	ADV
aiti-8904	191	2	,	,	PUNCT
aiti-8904	191	3	a	a	DET
aiti-8904	191	4	one	one	NUM
aiti-8904	191	5	-	-	PUNCT
aiti-8904	191	6	dimensional	dimensional	ADJ
aiti-8904	191	7	convolutional	convolutional	ADJ
aiti-8904	191	8	layer	layer	NOUN
aiti-8904	191	9	is	be	AUX
aiti-8904	191	10	used	use	VERB
aiti-8904	191	11	to	to	PART
aiti-8904	191	12	build	build	VERB
aiti-8904	191	13	the	the	DET
aiti-8904	191	14	model	model	NOUN
aiti-8904	191	15	,	,	PUNCT
aiti-8904	191	16	and	and	CCONJ
aiti-8904	191	17	relu	relu	NOUN
aiti-8904	191	18	is	be	AUX
aiti-8904	191	19	selected	select	VERB
aiti-8904	191	20	as	as	ADP
aiti-8904	191	21	the	the	DET
aiti-8904	191	22	activation	activation	NOUN
aiti-8904	191	23	function	function	NOUN
aiti-8904	191	24	.	.	PUNCT
aiti-8904	192	1	to	to	PART
aiti-8904	192	2	avoid	avoid	VERB
aiti-8904	192	3	the	the	DET
aiti-8904	192	4	over	over	ADV
aiti-8904	192	5	-	-	PUNCT
aiti-8904	192	6	fitting	fitting	NOUN
aiti-8904	192	7	of	of	ADP
aiti-8904	192	8	the	the	DET
aiti-8904	192	9	model	model	NOUN
aiti-8904	192	10	on	on	ADP
aiti-8904	192	11	the	the	DET
aiti-8904	192	12	training	training	NOUN
aiti-8904	192	13	data	datum	NOUN
aiti-8904	192	14	and	and	CCONJ
aiti-8904	192	15	reduce	reduce	VERB
aiti-8904	192	16	the	the	DET
aiti-8904	192	17	generalization	generalization	NOUN
aiti-8904	192	18	of	of	ADP
aiti-8904	192	19	the	the	DET
aiti-8904	192	20	model	model	NOUN
aiti-8904	192	21	,	,	PUNCT
aiti-8904	192	22	this	this	DET
aiti-8904	192	23	study	study	NOUN
aiti-8904	192	24	adds	add	VERB
aiti-8904	192	25	a	a	DET
aiti-8904	192	26	dropout	dropout	NOUN
aiti-8904	192	27	layer	layer	NOUN
aiti-8904	192	28	after	after	ADP
aiti-8904	192	29	the	the	DET
aiti-8904	192	30	convolutional	convolutional	ADJ
aiti-8904	192	31	layer	layer	NOUN
aiti-8904	192	32	.	.	PUNCT
aiti-8904	193	1	furthermore	furthermore	ADV
aiti-8904	193	2	,	,	PUNCT
aiti-8904	193	3	through	through	ADP
aiti-8904	193	4	adding	add	VERB
aiti-8904	193	5	the	the	DET
aiti-8904	193	6	pooling	pool	VERB
aiti-8904	193	7	layer	layer	NOUN
aiti-8904	193	8	,	,	PUNCT
aiti-8904	193	9	the	the	DET
aiti-8904	193	10	dimensions	dimension	NOUN
aiti-8904	193	11	of	of	ADP
aiti-8904	193	12	the	the	DET
aiti-8904	193	13	model	model	NOUN
aiti-8904	193	14	features	feature	NOUN
aiti-8904	193	15	are	be	AUX
aiti-8904	193	16	reduced	reduce	VERB
aiti-8904	193	17	and	and	CCONJ
aiti-8904	193	18	output	output	NOUN
aiti-8904	193	19	to	to	ADP
aiti-8904	193	20	the	the	DET
aiti-8904	193	21	full	full	ADJ
aiti-8904	193	22	connection	connection	NOUN
aiti-8904	193	23	layer	layer	NOUN
aiti-8904	193	24	.	.	PUNCT
aiti-8904	194	1	in	in	ADP
aiti-8904	194	2	the	the	DET
aiti-8904	194	3	full	full	ADJ
aiti-8904	194	4	connection	connection	NOUN
aiti-8904	194	5	layer	layer	NOUN
aiti-8904	194	6	,	,	PUNCT
aiti-8904	194	7	softmax	softmax	NOUN
aiti-8904	194	8	is	be	AUX
aiti-8904	194	9	used	use	VERB
aiti-8904	194	10	as	as	ADP
aiti-8904	194	11	the	the	DET
aiti-8904	194	12	activation	activation	NOUN
aiti-8904	194	13	function	function	NOUN
aiti-8904	194	14	.	.	PUNCT
aiti-8904	195	1	the	the	DET
aiti-8904	195	2	loss	loss	NOUN
aiti-8904	195	3	function	function	NOUN
aiti-8904	195	4	for	for	ADP
aiti-8904	195	5	the	the	DET
aiti-8904	195	6	evaluation	evaluation	NOUN
aiti-8904	195	7	model	model	NOUN
aiti-8904	195	8	is	be	AUX
aiti-8904	195	9	categorical	categorical	ADJ
aiti-8904	195	10	cross	cross	NOUN
aiti-8904	195	11	-	-	NOUN
aiti-8904	195	12	entropy	entropy	ADJ
aiti-8904	195	13	,	,	PUNCT
aiti-8904	195	14	and	and	CCONJ
aiti-8904	195	15	the	the	DET
aiti-8904	195	16	function	function	NOUN
aiti-8904	195	17	for	for	ADP
aiti-8904	195	18	the	the	DET
aiti-8904	195	19	optimizer	optimizer	NOUN
aiti-8904	195	20	is	be	AUX
aiti-8904	195	21	the	the	DET
aiti-8904	195	22	adam	adam	PROPN
aiti-8904	195	23	function	function	NOUN
aiti-8904	195	24	.	.	PUNCT
aiti-8904	196	1	fig	fig	NOUN
aiti-8904	196	2	.	.	PUNCT
aiti-8904	197	1	14	14	NUM
aiti-8904	197	2	depicts	depict	VERB
aiti-8904	197	3	the	the	DET
aiti-8904	197	4	flowchart	flowchart	NOUN
aiti-8904	197	5	of	of	ADP
aiti-8904	197	6	model	model	PROPN
aiti-8904	197	7	building	building	NOUN
aiti-8904	197	8	in	in	ADP
aiti-8904	197	9	this	this	DET
aiti-8904	197	10	study	study	NOUN
aiti-8904	197	11	.	.	PUNCT
aiti-8904	198	1	fig	fig	NOUN
aiti-8904	198	2	.	.	PUNCT
aiti-8904	199	1	14	14	NUM
aiti-8904	199	2	model	model	NOUN
aiti-8904	199	3	building	build	VERB
aiti-8904	199	4	3	3	NUM
aiti-8904	199	5	.	.	PUNCT
aiti-8904	199	6	system	system	NOUN
aiti-8904	199	7	implementation	implementation	NOUN
aiti-8904	199	8	after	after	ADP
aiti-8904	199	9	confirming	confirm	VERB
aiti-8904	199	10	that	that	SCONJ
aiti-8904	199	11	the	the	DET
aiti-8904	199	12	training	training	NOUN
aiti-8904	199	13	model	model	NOUN
aiti-8904	199	14	can	can	AUX
aiti-8904	199	15	identify	identify	VERB
aiti-8904	199	16	the	the	DET
aiti-8904	199	17	leak	leak	NOUN
aiti-8904	199	18	data	datum	NOUN
aiti-8904	199	19	,	,	PUNCT
aiti-8904	199	20	the	the	DET
aiti-8904	199	21	training	training	NOUN
aiti-8904	199	22	model	model	NOUN
aiti-8904	199	23	is	be	AUX
aiti-8904	199	24	used	use	VERB
aiti-8904	199	25	to	to	PART
aiti-8904	199	26	perform	perform	VERB
aiti-8904	199	27	a	a	DET
aiti-8904	199	28	leak	leak	NOUN
aiti-8904	199	29	test	test	NOUN
aiti-8904	199	30	in	in	ADP
aiti-8904	199	31	the	the	DET
aiti-8904	199	32	data	data	NOUN
aiti-8904	199	33	interval	interval	NOUN
aiti-8904	199	34	,	,	PUNCT
aiti-8904	199	35	thus	thus	ADV
aiti-8904	199	36	verifying	verify	VERB
aiti-8904	199	37	whether	whether	SCONJ
aiti-8904	199	38	the	the	DET
aiti-8904	199	39	model	model	NOUN
aiti-8904	199	40	can	can	AUX
aiti-8904	199	41	correctly	correctly	ADV
aiti-8904	199	42	determine	determine	VERB
aiti-8904	199	43	the	the	DET
aiti-8904	199	44	occurrence	occurrence	NOUN
aiti-8904	199	45	of	of	ADP
aiti-8904	199	46	leaks	leak	NOUN
aiti-8904	199	47	and	and	CCONJ
aiti-8904	199	48	effectively	effectively	ADV
aiti-8904	199	49	resist	resist	VERB
aiti-8904	199	50	noise	noise	NOUN
aiti-8904	199	51	176	176	NUM
aiti-8904	199	52	advances	advance	NOUN
aiti-8904	199	53	in	in	ADP
aiti-8904	199	54	technology	technology	NOUN
aiti-8904	199	55	innovation	innovation	NOUN
aiti-8904	199	56	,	,	PUNCT
aiti-8904	199	57	vol	vol	NOUN
aiti-8904	199	58	.	.	PROPN
aiti-8904	200	1	7	7	NUM
aiti-8904	200	2	,	,	PUNCT
aiti-8904	200	3	no	no	INTJ
aiti-8904	200	4	.	.	NOUN
aiti-8904	200	5	3	3	NUM
aiti-8904	200	6	,	,	PUNCT
aiti-8904	200	7	2022	2022	NUM
aiti-8904	200	8	,	,	PUNCT
aiti-8904	200	9	pp	pp	ADJ
aiti-8904	200	10	.	.	PUNCT
aiti-8904	201	1	169	169	NUM
aiti-8904	201	2	-	-	SYM
aiti-8904	201	3	180	180	NUM
aiti-8904	201	4	interference	interference	NOUN
aiti-8904	201	5	.	.	PUNCT
aiti-8904	202	1	as	as	SCONJ
aiti-8904	202	2	shown	show	VERB
aiti-8904	202	3	in	in	ADP
aiti-8904	202	4	fig	fig	NOUN
aiti-8904	202	5	.	.	PUNCT
aiti-8904	203	1	15	15	NUM
aiti-8904	203	2	,	,	PUNCT
aiti-8904	203	3	if	if	SCONJ
aiti-8904	203	4	the	the	DET
aiti-8904	203	5	result	result	NOUN
aiti-8904	203	6	of	of	ADP
aiti-8904	203	7	the	the	DET
aiti-8904	203	8	model	model	NOUN
aiti-8904	203	9	test	test	NOUN
aiti-8904	203	10	is	be	AUX
aiti-8904	203	11	different	different	ADJ
aiti-8904	203	12	from	from	ADP
aiti-8904	203	13	the	the	DET
aiti-8904	203	14	expected	expect	VERB
aiti-8904	203	15	result	result	NOUN
aiti-8904	203	16	,	,	PUNCT
aiti-8904	203	17	then	then	ADV
aiti-8904	203	18	the	the	DET
aiti-8904	203	19	hyperparameters	hyperparameter	NOUN
aiti-8904	203	20	of	of	ADP
aiti-8904	203	21	the	the	DET
aiti-8904	203	22	model	model	NOUN
aiti-8904	203	23	are	be	AUX
aiti-8904	203	24	re	re	VERB
aiti-8904	203	25	-	-	VERB
aiti-8904	203	26	adjusted	adjust	VERB
aiti-8904	203	27	to	to	PART
aiti-8904	203	28	meet	meet	VERB
aiti-8904	203	29	the	the	DET
aiti-8904	203	30	condition	condition	NOUN
aiti-8904	203	31	where	where	SCONJ
aiti-8904	203	32	the	the	DET
aiti-8904	203	33	time	time	NOUN
aiti-8904	203	34	interval	interval	NOUN
aiti-8904	203	35	of	of	ADP
aiti-8904	203	36	the	the	DET
aiti-8904	203	37	leak	leak	NOUN
aiti-8904	203	38	occurrence	occurrence	NOUN
aiti-8904	203	39	is	be	AUX
aiti-8904	203	40	consistent	consistent	ADJ
aiti-8904	203	41	with	with	ADP
aiti-8904	203	42	the	the	DET
aiti-8904	203	43	time	time	NOUN
aiti-8904	203	44	interval	interval	NOUN
aiti-8904	203	45	of	of	ADP
aiti-8904	203	46	the	the	DET
aiti-8904	203	47	real	real	ADJ
aiti-8904	203	48	leak	leak	NOUN
aiti-8904	203	49	test	test	NOUN
aiti-8904	203	50	.	.	PUNCT
aiti-8904	204	1	fig	fig	NOUN
aiti-8904	204	2	.	.	PUNCT
aiti-8904	205	1	15	15	NUM
aiti-8904	205	2	system	system	NOUN
aiti-8904	205	3	implementation	implementation	NOUN
aiti-8904	205	4	3.1	3.1	NUM
aiti-8904	205	5	.	.	PUNCT
aiti-8904	205	6	model	model	NOUN
aiti-8904	205	7	architecture	architecture	NOUN
aiti-8904	205	8	fig	fig	NOUN
aiti-8904	205	9	.	.	PUNCT
aiti-8904	206	1	16	16	NUM
aiti-8904	206	2	shows	show	VERB
aiti-8904	206	3	the	the	DET
aiti-8904	206	4	model	model	NOUN
aiti-8904	206	5	architecture	architecture	NOUN
aiti-8904	206	6	,	,	PUNCT
aiti-8904	206	7	in	in	ADP
aiti-8904	206	8	which	which	PRON
aiti-8904	206	9	the	the	DET
aiti-8904	206	10	three	three	NUM
aiti-8904	206	11	layers	layer	NOUN
aiti-8904	206	12	of	of	ADP
aiti-8904	206	13	the	the	DET
aiti-8904	206	14	one	one	NUM
aiti-8904	206	15	-	-	PUNCT
aiti-8904	206	16	dimensional	dimensional	ADJ
aiti-8904	206	17	convolutional	convolutional	ADJ
aiti-8904	206	18	layer	layer	NOUN
aiti-8904	206	19	are	be	AUX
aiti-8904	206	20	stacked	stack	VERB
aiti-8904	206	21	,	,	PUNCT
aiti-8904	206	22	and	and	CCONJ
aiti-8904	206	23	the	the	DET
aiti-8904	206	24	dropout	dropout	NOUN
aiti-8904	206	25	layer	layer	NOUN
aiti-8904	206	26	is	be	AUX
aiti-8904	206	27	added	add	VERB
aiti-8904	206	28	to	to	PART
aiti-8904	206	29	prevent	prevent	VERB
aiti-8904	206	30	the	the	DET
aiti-8904	206	31	model	model	NOUN
aiti-8904	206	32	from	from	ADP
aiti-8904	206	33	overly	overly	ADV
aiti-8904	206	34	relying	rely	VERB
aiti-8904	206	35	on	on	ADP
aiti-8904	206	36	the	the	DET
aiti-8904	206	37	training	training	NOUN
aiti-8904	206	38	data	datum	NOUN
aiti-8904	206	39	set	set	VERB
aiti-8904	206	40	.	.	PUNCT
aiti-8904	207	1	after	after	SCONJ
aiti-8904	207	2	repeated	repeat	VERB
aiti-8904	207	3	testing	testing	NOUN
aiti-8904	207	4	is	be	AUX
aiti-8904	207	5	conducted	conduct	VERB
aiti-8904	207	6	,	,	PUNCT
aiti-8904	207	7	the	the	DET
aiti-8904	207	8	classification	classification	NOUN
aiti-8904	207	9	effect	effect	NOUN
aiti-8904	207	10	of	of	ADP
aiti-8904	207	11	using	use	VERB
aiti-8904	207	12	the	the	DET
aiti-8904	207	13	average	average	ADJ
aiti-8904	207	14	pooling	pooling	NOUN
aiti-8904	207	15	layer	layer	NOUN
aiti-8904	207	16	is	be	AUX
aiti-8904	207	17	found	find	VERB
aiti-8904	207	18	to	to	PART
aiti-8904	207	19	be	be	AUX
aiti-8904	207	20	better	well	ADJ
aiti-8904	207	21	than	than	ADP
aiti-8904	207	22	the	the	DET
aiti-8904	207	23	effect	effect	NOUN
aiti-8904	207	24	of	of	ADP
aiti-8904	207	25	the	the	DET
aiti-8904	207	26	max	max	PROPN
aiti-8904	207	27	pooling	pool	VERB
aiti-8904	207	28	layer	layer	NOUN
aiti-8904	207	29	.	.	PUNCT
aiti-8904	208	1	thus	thus	ADV
aiti-8904	208	2	,	,	PUNCT
aiti-8904	208	3	the	the	DET
aiti-8904	208	4	average	average	ADJ
aiti-8904	208	5	pooling	pooling	NOUN
aiti-8904	208	6	layer	layer	NOUN
aiti-8904	208	7	is	be	AUX
aiti-8904	208	8	used	use	VERB
aiti-8904	208	9	here	here	ADV
aiti-8904	208	10	.	.	PUNCT
aiti-8904	209	1	fig	fig	NOUN
aiti-8904	209	2	.	.	PUNCT
aiti-8904	210	1	16	16	NUM
aiti-8904	210	2	model	model	NOUN
aiti-8904	210	3	architecture	architecture	NOUN
aiti-8904	210	4	3.2	3.2	NUM
aiti-8904	210	5	.	.	PUNCT
aiti-8904	211	1	result	result	NOUN
aiti-8904	211	2	of	of	ADP
aiti-8904	211	3	leak	leak	NOUN
aiti-8904	211	4	point	point	NOUN
aiti-8904	211	5	identification	identification	NOUN
aiti-8904	211	6	after	after	ADP
aiti-8904	211	7	completing	complete	VERB
aiti-8904	211	8	the	the	DET
aiti-8904	211	9	model	model	NOUN
aiti-8904	211	10	training	training	NOUN
aiti-8904	211	11	,	,	PUNCT
aiti-8904	211	12	the	the	DET
aiti-8904	211	13	data	datum	NOUN
aiti-8904	211	14	of	of	ADP
aiti-8904	211	15	two	two	NUM
aiti-8904	211	16	different	different	ADJ
aiti-8904	211	17	pressure	pressure	NOUN
aiti-8904	211	18	sensors	sensor	NOUN
aiti-8904	211	19	are	be	AUX
aiti-8904	211	20	employed	employ	VERB
aiti-8904	211	21	to	to	PART
aiti-8904	211	22	test	test	VERB
aiti-8904	211	23	whether	whether	SCONJ
aiti-8904	211	24	the	the	DET
aiti-8904	211	25	testing	testing	NOUN
aiti-8904	211	26	model	model	NOUN
aiti-8904	211	27	can	can	AUX
aiti-8904	211	28	correctly	correctly	ADV
aiti-8904	211	29	distinguish	distinguish	VERB
aiti-8904	211	30	and	and	CCONJ
aiti-8904	211	31	predict	predict	VERB
aiti-8904	211	32	the	the	DET
aiti-8904	211	33	point	point	NOUN
aiti-8904	211	34	of	of	ADP
aiti-8904	211	35	pipeline	pipeline	NOUN
aiti-8904	211	36	leak	leak	NOUN
aiti-8904	211	37	where	where	SCONJ
aiti-8904	211	38	the	the	DET
aiti-8904	211	39	leak	leak	NOUN
aiti-8904	211	40	actually	actually	ADV
aiti-8904	211	41	occurs	occur	VERB
aiti-8904	211	42	.	.	PUNCT
aiti-8904	212	1	the	the	DET
aiti-8904	212	2	pressure	pressure	NOUN
aiti-8904	212	3	sensor	sensor	NOUN
aiti-8904	212	4	data	datum	NOUN
aiti-8904	212	5	shown	show	VERB
aiti-8904	212	6	in	in	ADP
aiti-8904	212	7	fig	fig	NOUN
aiti-8904	212	8	.	.	PUNCT
aiti-8904	213	1	16	16	NUM
aiti-8904	213	2	is	be	AUX
aiti-8904	213	3	the	the	DET
aiti-8904	213	4	data	datum	NOUN
aiti-8904	213	5	of	of	ADP
aiti-8904	213	6	the	the	DET
aiti-8904	213	7	training	training	NOUN
aiti-8904	213	8	model	model	NOUN
aiti-8904	213	9	(	(	PUNCT
aiti-8904	213	10	pt	pt	PROPN
aiti-8904	213	11	-	-	PUNCT
aiti-8904	213	12	m621	m621	NUM
aiti-8904	213	13	)	)	PUNCT
aiti-8904	213	14	.	.	PUNCT
aiti-8904	214	1	in	in	ADP
aiti-8904	214	2	fig	fig	NOUN
aiti-8904	214	3	.	.	PUNCT
aiti-8904	215	1	16	16	NUM
aiti-8904	215	2	,	,	PUNCT
aiti-8904	215	3	if	if	SCONJ
aiti-8904	215	4	the	the	DET
aiti-8904	215	5	model	model	NOUN
aiti-8904	215	6	finds	find	VERB
aiti-8904	215	7	leakage	leakage	NOUN
aiti-8904	215	8	,	,	PUNCT
aiti-8904	215	9	the	the	DET
aiti-8904	215	10	leak	leak	NOUN
aiti-8904	215	11	point	point	NOUN
aiti-8904	215	12	is	be	AUX
aiti-8904	215	13	one	one	NUM
aiti-8904	215	14	,	,	PUNCT
aiti-8904	215	15	and	and	CCONJ
aiti-8904	215	16	the	the	DET
aiti-8904	215	17	normal	normal	ADJ
aiti-8904	215	18	data	datum	NOUN
aiti-8904	215	19	is	be	AUX
aiti-8904	215	20	zero	zero	NUM
aiti-8904	215	21	.	.	PUNCT
aiti-8904	216	1	moreover	moreover	ADV
aiti-8904	216	2	,	,	PUNCT
aiti-8904	216	3	in	in	ADP
aiti-8904	216	4	order	order	NOUN
aiti-8904	216	5	to	to	PART
aiti-8904	216	6	avoid	avoid	VERB
aiti-8904	216	7	the	the	DET
aiti-8904	216	8	excessive	excessive	ADJ
aiti-8904	216	9	dependence	dependence	NOUN
aiti-8904	216	10	of	of	ADP
aiti-8904	216	11	the	the	DET
aiti-8904	216	12	model	model	NOUN
aiti-8904	216	13	on	on	ADP
aiti-8904	216	14	the	the	DET
aiti-8904	216	15	training	training	NOUN
aiti-8904	216	16	data	datum	NOUN
aiti-8904	216	17	set	set	VERB
aiti-8904	216	18	,	,	PUNCT
aiti-8904	216	19	the	the	DET
aiti-8904	216	20	data	datum	NOUN
aiti-8904	216	21	from	from	ADP
aiti-8904	216	22	the	the	DET
aiti-8904	216	23	differential	differential	ADJ
aiti-8904	216	24	pressure	pressure	NOUN
aiti-8904	216	25	sensor	sensor	NOUN
aiti-8904	216	26	(	(	PUNCT
aiti-8904	216	27	pt	pt	PROPN
aiti-8904	216	28	-	-	PUNCT
aiti-8904	216	29	m622	m622	PROPN
aiti-8904	216	30	)	)	PUNCT
aiti-8904	216	31	is	be	AUX
aiti-8904	216	32	used	use	VERB
aiti-8904	216	33	to	to	PART
aiti-8904	216	34	verify	verify	VERB
aiti-8904	216	35	the	the	DET
aiti-8904	216	36	generalization	generalization	NOUN
aiti-8904	216	37	of	of	ADP
aiti-8904	216	38	the	the	DET
aiti-8904	216	39	model	model	NOUN
aiti-8904	216	40	.	.	PUNCT
aiti-8904	217	1	the	the	DET
aiti-8904	217	2	model	model	NOUN
aiti-8904	217	3	architecture	architecture	NOUN
aiti-8904	217	4	used	use	VERB
aiti-8904	217	5	in	in	ADP
aiti-8904	217	6	figs	fig	NOUN
aiti-8904	217	7	.	.	PUNCT
aiti-8904	218	1	17	17	NUM
aiti-8904	218	2	-	-	SYM
aiti-8904	218	3	18	18	NUM
aiti-8904	218	4	is	be	AUX
aiti-8904	218	5	the	the	DET
aiti-8904	218	6	same	same	ADJ
aiti-8904	218	7	as	as	ADP
aiti-8904	218	8	the	the	DET
aiti-8904	218	9	one	one	NOUN
aiti-8904	218	10	shown	show	VERB
aiti-8904	218	11	in	in	ADP
aiti-8904	218	12	fig	fig	NOUN
aiti-8904	218	13	.	.	PUNCT
aiti-8904	219	1	16	16	NUM
aiti-8904	219	2	.	.	PUNCT
aiti-8904	220	1	in	in	ADP
aiti-8904	220	2	these	these	DET
aiti-8904	220	3	figures	figure	NOUN
aiti-8904	220	4	,	,	PUNCT
aiti-8904	220	5	the	the	DET
aiti-8904	220	6	red	red	ADJ
aiti-8904	220	7	line	line	NOUN
aiti-8904	220	8	represents	represent	VERB
aiti-8904	220	9	the	the	DET
aiti-8904	220	10	zero	zero	NUM
aiti-8904	220	11	value	value	NOUN
aiti-8904	220	12	of	of	ADP
aiti-8904	220	13	the	the	DET
aiti-8904	220	14	pressure	pressure	NOUN
aiti-8904	220	15	.	.	PUNCT
aiti-8904	221	1	in	in	ADP
aiti-8904	221	2	the	the	DET
aiti-8904	221	3	pooling	pool	VERB
aiti-8904	221	4	layer	layer	NOUN
aiti-8904	221	5	,	,	PUNCT
aiti-8904	221	6	this	this	DET
aiti-8904	221	7	study	study	NOUN
aiti-8904	221	8	chooses	choose	VERB
aiti-8904	221	9	the	the	DET
aiti-8904	221	10	max	max	PROPN
aiti-8904	221	11	pooling	pool	VERB
aiti-8904	221	12	instead	instead	ADV
aiti-8904	221	13	of	of	ADP
aiti-8904	221	14	the	the	DET
aiti-8904	221	15	average	average	ADJ
aiti-8904	221	16	pooling	pooling	NOUN
aiti-8904	221	17	as	as	ADP
aiti-8904	221	18	the	the	DET
aiti-8904	221	19	mean	mean	NOUN
aiti-8904	221	20	to	to	PART
aiti-8904	221	21	gain	gain	VERB
aiti-8904	221	22	the	the	DET
aiti-8904	221	23	results	result	NOUN
aiti-8904	221	24	of	of	ADP
aiti-8904	221	25	the	the	DET
aiti-8904	221	26	model	model	NOUN
aiti-8904	221	27	training	training	NOUN
aiti-8904	221	28	.	.	PUNCT
aiti-8904	222	1	fig	fig	NOUN
aiti-8904	222	2	.	.	PUNCT
aiti-8904	223	1	17	17	NUM
aiti-8904	223	2	and	and	CCONJ
aiti-8904	223	3	fig	fig	NOUN
aiti-8904	223	4	.	.	PUNCT
aiti-8904	224	1	18	18	NUM
aiti-8904	224	2	use	use	VERB
aiti-8904	224	3	pt	pt	NOUN
aiti-8904	224	4	-	-	PUNCT
aiti-8904	224	5	m621	m621	PROPN
aiti-8904	224	6	and	and	CCONJ
aiti-8904	224	7	pt	pt	PROPN
aiti-8904	224	8	-	-	PUNCT
aiti-8904	224	9	m622	m622	PROPN
aiti-8904	224	10	as	as	ADP
aiti-8904	224	11	the	the	DET
aiti-8904	224	12	data	datum	NOUN
aiti-8904	224	13	sources	source	NOUN
aiti-8904	224	14	of	of	ADP
aiti-8904	224	15	the	the	DET
aiti-8904	224	16	model	model	NOUN
aiti-8904	224	17	verification	verification	NOUN
aiti-8904	224	18	,	,	PUNCT
aiti-8904	224	19	respectively	respectively	ADV
aiti-8904	224	20	.	.	PUNCT
aiti-8904	225	1	fig	fig	NOUN
aiti-8904	225	2	.	.	PUNCT
aiti-8904	226	1	18	18	NUM
aiti-8904	226	2	shows	show	VERB
aiti-8904	226	3	that	that	SCONJ
aiti-8904	226	4	model	model	NOUN
aiti-8904	226	5	1	1	NUM
aiti-8904	226	6	can	can	AUX
aiti-8904	226	7	accurately	accurately	ADV
aiti-8904	226	8	detect	detect	VERB
aiti-8904	226	9	the	the	DET
aiti-8904	226	10	occurrence	occurrence	NOUN
aiti-8904	226	11	of	of	ADP
aiti-8904	226	12	leakage	leakage	NOUN
aiti-8904	226	13	,	,	PUNCT
aiti-8904	226	14	while	while	SCONJ
aiti-8904	226	15	fig	fig	NOUN
aiti-8904	226	16	.	.	PUNCT
aiti-8904	227	1	17	17	NUM
aiti-8904	227	2	can	can	AUX
aiti-8904	227	3	only	only	ADV
aiti-8904	227	4	177	177	NUM
aiti-8904	227	5	advances	advance	NOUN
aiti-8904	227	6	in	in	ADP
aiti-8904	227	7	technology	technology	NOUN
aiti-8904	227	8	innovation	innovation	NOUN
aiti-8904	227	9	,	,	PUNCT
aiti-8904	227	10	vol	vol	NOUN
aiti-8904	227	11	.	.	PROPN
aiti-8904	228	1	7	7	NUM
aiti-8904	228	2	,	,	PUNCT
aiti-8904	228	3	no	no	INTJ
aiti-8904	228	4	.	.	NOUN
aiti-8904	228	5	3	3	NUM
aiti-8904	228	6	,	,	PUNCT
aiti-8904	228	7	2022	2022	NUM
aiti-8904	228	8	,	,	PUNCT
aiti-8904	228	9	pp	pp	ADJ
aiti-8904	228	10	.	.	PUNCT
aiti-8904	229	1	169	169	NUM
aiti-8904	229	2	-	-	SYM
aiti-8904	229	3	180	180	NUM
aiti-8904	229	4	detect	detect	VERB
aiti-8904	229	5	the	the	DET
aiti-8904	229	6	signal	signal	NOUN
aiti-8904	229	7	of	of	ADP
aiti-8904	229	8	the	the	DET
aiti-8904	229	9	leakage	leakage	NOUN
aiti-8904	229	10	generated	generate	VERB
aiti-8904	229	11	during	during	ADP
aiti-8904	229	12	the	the	DET
aiti-8904	229	13	first	first	ADJ
aiti-8904	229	14	leak	leak	NOUN
aiti-8904	229	15	test	test	NOUN
aiti-8904	229	16	.	.	PUNCT
aiti-8904	230	1	fig	fig	NOUN
aiti-8904	230	2	.	.	PUNCT
aiti-8904	231	1	19	19	NUM
aiti-8904	231	2	and	and	CCONJ
aiti-8904	231	3	fig	fig	NOUN
aiti-8904	231	4	.	.	PUNCT
aiti-8904	232	1	20	20	NUM
aiti-8904	232	2	use	use	VERB
aiti-8904	232	3	pt	pt	NOUN
aiti-8904	232	4	-	-	PUNCT
aiti-8904	232	5	m621	m621	PROPN
aiti-8904	232	6	and	and	CCONJ
aiti-8904	232	7	pt	pt	PROPN
aiti-8904	232	8	-	-	PUNCT
aiti-8904	232	9	m622	m622	PROPN
aiti-8904	232	10	as	as	ADP
aiti-8904	232	11	the	the	DET
aiti-8904	232	12	data	datum	NOUN
aiti-8904	232	13	sources	source	NOUN
aiti-8904	232	14	of	of	ADP
aiti-8904	232	15	the	the	DET
aiti-8904	232	16	test	test	NOUN
aiti-8904	232	17	model	model	NOUN
aiti-8904	232	18	,	,	PUNCT
aiti-8904	232	19	respectively	respectively	ADV
aiti-8904	232	20	.	.	PUNCT
aiti-8904	233	1	the	the	DET
aiti-8904	233	2	architecture	architecture	NOUN
aiti-8904	233	3	of	of	ADP
aiti-8904	233	4	this	this	DET
aiti-8904	233	5	test	test	NOUN
aiti-8904	233	6	model	model	NOUN
aiti-8904	233	7	uses	use	VERB
aiti-8904	233	8	the	the	DET
aiti-8904	233	9	average	average	ADJ
aiti-8904	233	10	pooling	pooling	NOUN
aiti-8904	233	11	instead	instead	ADV
aiti-8904	233	12	of	of	ADP
aiti-8904	233	13	the	the	DET
aiti-8904	233	14	max	max	PROPN
aiti-8904	233	15	pooling	pooling	NOUN
aiti-8904	233	16	in	in	ADP
aiti-8904	233	17	designing	design	VERB
aiti-8904	233	18	the	the	DET
aiti-8904	233	19	pooling	pool	VERB
aiti-8904	233	20	layer	layer	NOUN
aiti-8904	233	21	,	,	PUNCT
aiti-8904	233	22	and	and	CCONJ
aiti-8904	233	23	the	the	DET
aiti-8904	233	24	size	size	NOUN
aiti-8904	233	25	of	of	ADP
aiti-8904	233	26	the	the	DET
aiti-8904	233	27	convolution	convolution	NOUN
aiti-8904	233	28	kernel	kernel	NOUN
aiti-8904	233	29	is	be	AUX
aiti-8904	233	30	3	3	NUM
aiti-8904	233	31	.	.	PUNCT
aiti-8904	233	32	as	as	SCONJ
aiti-8904	233	33	shown	show	VERB
aiti-8904	233	34	in	in	ADP
aiti-8904	233	35	fig	fig	NOUN
aiti-8904	233	36	.	.	PUNCT
aiti-8904	234	1	19	19	NUM
aiti-8904	234	2	,	,	PUNCT
aiti-8904	234	3	model	model	NOUN
aiti-8904	234	4	2	2	NUM
aiti-8904	234	5	only	only	ADV
aiti-8904	234	6	identifies	identify	VERB
aiti-8904	234	7	the	the	DET
aiti-8904	234	8	first	first	ADJ
aiti-8904	234	9	and	and	CCONJ
aiti-8904	234	10	third	third	ADJ
aiti-8904	234	11	leak	leak	NOUN
aiti-8904	234	12	occurrences	occurrence	NOUN
aiti-8904	234	13	but	but	CCONJ
aiti-8904	234	14	does	do	AUX
aiti-8904	234	15	not	not	PART
aiti-8904	234	16	identify	identify	VERB
aiti-8904	234	17	the	the	DET
aiti-8904	234	18	second	second	ADJ
aiti-8904	234	19	leak	leak	NOUN
aiti-8904	234	20	test	test	NOUN
aiti-8904	234	21	.	.	PUNCT
aiti-8904	235	1	fig	fig	NOUN
aiti-8904	235	2	.	.	PUNCT
aiti-8904	236	1	20	20	NUM
aiti-8904	236	2	shows	show	VERB
aiti-8904	236	3	that	that	SCONJ
aiti-8904	236	4	this	this	DET
aiti-8904	236	5	model	model	NOUN
aiti-8904	236	6	can	can	AUX
aiti-8904	236	7	accurately	accurately	ADV
aiti-8904	236	8	detect	detect	VERB
aiti-8904	236	9	leakage	leakage	NOUN
aiti-8904	236	10	in	in	ADP
aiti-8904	236	11	the	the	DET
aiti-8904	236	12	interval	interval	NOUN
aiti-8904	236	13	where	where	SCONJ
aiti-8904	236	14	the	the	DET
aiti-8904	236	15	leakage	leakage	NOUN
aiti-8904	236	16	occurs	occur	VERB
aiti-8904	236	17	.	.	PUNCT
aiti-8904	237	1	fig	fig	NOUN
aiti-8904	237	2	.	.	PUNCT
aiti-8904	238	1	21	21	NUM
aiti-8904	238	2	and	and	CCONJ
aiti-8904	238	3	fig	fig	NOUN
aiti-8904	238	4	.	.	PUNCT
aiti-8904	239	1	22	22	NUM
aiti-8904	239	2	use	use	VERB
aiti-8904	239	3	pt	pt	NOUN
aiti-8904	239	4	-	-	PUNCT
aiti-8904	239	5	m621	m621	PROPN
aiti-8904	239	6	and	and	CCONJ
aiti-8904	239	7	pt	pt	PROPN
aiti-8904	239	8	-	-	PUNCT
aiti-8904	239	9	m622	m622	PROPN
aiti-8904	239	10	as	as	ADP
aiti-8904	239	11	the	the	DET
aiti-8904	239	12	data	datum	NOUN
aiti-8904	239	13	sources	source	NOUN
aiti-8904	239	14	of	of	ADP
aiti-8904	239	15	the	the	DET
aiti-8904	239	16	test	test	NOUN
aiti-8904	239	17	model	model	NOUN
aiti-8904	239	18	,	,	PUNCT
aiti-8904	239	19	respectively	respectively	ADV
aiti-8904	239	20	.	.	PUNCT
aiti-8904	240	1	the	the	DET
aiti-8904	240	2	structure	structure	NOUN
aiti-8904	240	3	of	of	ADP
aiti-8904	240	4	the	the	DET
aiti-8904	240	5	test	test	NOUN
aiti-8904	240	6	model	model	NOUN
aiti-8904	240	7	is	be	AUX
aiti-8904	240	8	shown	show	VERB
aiti-8904	240	9	in	in	ADP
aiti-8904	240	10	fig	fig	NOUN
aiti-8904	240	11	.	.	PUNCT
aiti-8904	241	1	16	16	NUM
aiti-8904	241	2	.	.	PUNCT
aiti-8904	242	1	the	the	DET
aiti-8904	242	2	size	size	NOUN
aiti-8904	242	3	of	of	ADP
aiti-8904	242	4	the	the	DET
aiti-8904	242	5	convolution	convolution	NOUN
aiti-8904	242	6	kernel	kernel	NOUN
aiti-8904	242	7	is	be	AUX
aiti-8904	242	8	adjusted	adjust	VERB
aiti-8904	242	9	to	to	ADP
aiti-8904	242	10	5	5	NUM
aiti-8904	242	11	,	,	PUNCT
aiti-8904	242	12	and	and	CCONJ
aiti-8904	242	13	the	the	DET
aiti-8904	242	14	test	test	NOUN
aiti-8904	242	15	results	result	NOUN
aiti-8904	242	16	are	be	AUX
aiti-8904	242	17	shown	show	VERB
aiti-8904	242	18	in	in	ADP
aiti-8904	242	19	fig	fig	NOUN
aiti-8904	242	20	.	.	PUNCT
aiti-8904	243	1	21	21	NUM
aiti-8904	243	2	and	and	CCONJ
aiti-8904	243	3	fig	fig	NOUN
aiti-8904	243	4	.	.	PUNCT
aiti-8904	244	1	22	22	NUM
aiti-8904	244	2	.	.	PUNCT
aiti-8904	245	1	as	as	SCONJ
aiti-8904	245	2	shown	show	VERB
aiti-8904	245	3	in	in	ADP
aiti-8904	245	4	fig	fig	NOUN
aiti-8904	245	5	.	.	PUNCT
aiti-8904	246	1	21	21	NUM
aiti-8904	246	2	and	and	CCONJ
aiti-8904	246	3	fig	fig	NOUN
aiti-8904	246	4	.	.	PUNCT
aiti-8904	247	1	22	22	NUM
aiti-8904	247	2	,	,	PUNCT
aiti-8904	247	3	this	this	DET
aiti-8904	247	4	model	model	NOUN
aiti-8904	247	5	architecture	architecture	NOUN
aiti-8904	247	6	can	can	AUX
aiti-8904	247	7	accurately	accurately	ADV
aiti-8904	247	8	detect	detect	VERB
aiti-8904	247	9	leaks	leak	NOUN
aiti-8904	247	10	in	in	ADP
aiti-8904	247	11	the	the	DET
aiti-8904	247	12	time	time	NOUN
aiti-8904	247	13	interval	interval	NOUN
aiti-8904	247	14	of	of	ADP
aiti-8904	247	15	the	the	DET
aiti-8904	247	16	leak	leak	NOUN
aiti-8904	247	17	test	test	NOUN
aiti-8904	247	18	on	on	ADP
aiti-8904	247	19	two	two	NUM
aiti-8904	247	20	different	different	ADJ
aiti-8904	247	21	data	data	NOUN
aiti-8904	247	22	sets	set	NOUN
aiti-8904	247	23	,	,	PUNCT
aiti-8904	247	24	which	which	PRON
aiti-8904	247	25	confirms	confirm	VERB
aiti-8904	247	26	that	that	SCONJ
aiti-8904	247	27	the	the	DET
aiti-8904	247	28	model	model	NOUN
aiti-8904	247	29	is	be	AUX
aiti-8904	247	30	with	with	ADP
aiti-8904	247	31	good	good	ADJ
aiti-8904	247	32	generalization	generalization	NOUN
aiti-8904	247	33	and	and	CCONJ
aiti-8904	247	34	can	can	AUX
aiti-8904	247	35	be	be	AUX
aiti-8904	247	36	applied	apply	VERB
aiti-8904	247	37	to	to	ADP
aiti-8904	247	38	different	different	ADJ
aiti-8904	247	39	data	datum	NOUN
aiti-8904	247	40	sets	set	NOUN
aiti-8904	247	41	.	.	PUNCT
aiti-8904	248	1	table	table	NOUN
aiti-8904	248	2	1	1	NUM
aiti-8904	248	3	lists	list	VERB
aiti-8904	248	4	the	the	DET
aiti-8904	248	5	result	result	NOUN
aiti-8904	248	6	of	of	ADP
aiti-8904	248	7	testing	testing	NOUN
aiti-8904	248	8	leak	leak	NOUN
aiti-8904	248	9	data	datum	NOUN
aiti-8904	248	10	after	after	ADP
aiti-8904	248	11	adjusting	adjust	VERB
aiti-8904	248	12	the	the	DET
aiti-8904	248	13	hyperparameters	hyperparameter	NOUN
aiti-8904	248	14	of	of	ADP
aiti-8904	248	15	the	the	DET
aiti-8904	248	16	above	above	ADJ
aiti-8904	248	17	models	model	NOUN
aiti-8904	248	18	.	.	PUNCT
aiti-8904	249	1	finally	finally	ADV
aiti-8904	249	2	,	,	PUNCT
aiti-8904	249	3	the	the	DET
aiti-8904	249	4	model	model	NOUN
aiti-8904	249	5	developed	develop	VERB
aiti-8904	249	6	by	by	ADP
aiti-8904	249	7	this	this	DET
aiti-8904	249	8	study	study	NOUN
aiti-8904	249	9	is	be	AUX
aiti-8904	249	10	employed	employ	VERB
aiti-8904	249	11	to	to	PART
aiti-8904	249	12	test	test	VERB
aiti-8904	249	13	the	the	DET
aiti-8904	249	14	data	datum	NOUN
aiti-8904	249	15	set	set	VERB
aiti-8904	249	16	from	from	ADP
aiti-8904	249	17	different	different	ADJ
aiti-8904	249	18	sensors	sensor	NOUN
aiti-8904	249	19	.	.	PUNCT
aiti-8904	250	1	the	the	DET
aiti-8904	250	2	result	result	NOUN
aiti-8904	250	3	indicates	indicate	VERB
aiti-8904	250	4	that	that	SCONJ
aiti-8904	250	5	this	this	DET
aiti-8904	250	6	model	model	NOUN
aiti-8904	250	7	can	can	AUX
aiti-8904	250	8	still	still	ADV
aiti-8904	250	9	play	play	VERB
aiti-8904	250	10	a	a	DET
aiti-8904	250	11	role	role	NOUN
aiti-8904	250	12	in	in	ADP
aiti-8904	250	13	leak	leak	NOUN
aiti-8904	250	14	detection	detection	NOUN
aiti-8904	250	15	when	when	SCONJ
aiti-8904	250	16	a	a	DET
aiti-8904	250	17	real	real	ADJ
aiti-8904	250	18	leak	leak	NOUN
aiti-8904	250	19	occurs	occur	VERB
aiti-8904	250	20	,	,	PUNCT
aiti-8904	250	21	thus	thus	ADV
aiti-8904	250	22	confirming	confirm	VERB
aiti-8904	250	23	that	that	SCONJ
aiti-8904	250	24	this	this	DET
aiti-8904	250	25	model	model	NOUN
aiti-8904	250	26	can	can	AUX
aiti-8904	250	27	well	well	ADV
aiti-8904	250	28	function	function	VERB
aiti-8904	250	29	,	,	PUNCT
aiti-8904	250	30	receive	receive	VERB
aiti-8904	250	31	the	the	DET
aiti-8904	250	32	accurate	accurate	ADJ
aiti-8904	250	33	leak	leak	NOUN
aiti-8904	250	34	detection	detection	NOUN
aiti-8904	250	35	function	function	NOUN
aiti-8904	250	36	,	,	PUNCT
aiti-8904	250	37	and	and	CCONJ
aiti-8904	250	38	achieve	achieve	VERB
aiti-8904	250	39	the	the	DET
aiti-8904	250	40	level	level	NOUN
aiti-8904	250	41	of	of	ADP
aiti-8904	250	42	generalization	generalization	NOUN
aiti-8904	250	43	.	.	PUNCT
aiti-8904	251	1	the	the	DET
aiti-8904	251	2	model	model	NOUN
aiti-8904	251	3	developed	develop	VERB
aiti-8904	251	4	by	by	ADP
aiti-8904	251	5	this	this	DET
aiti-8904	251	6	study	study	NOUN
aiti-8904	251	7	is	be	AUX
aiti-8904	251	8	implemented	implement	VERB
aiti-8904	251	9	with	with	ADP
aiti-8904	251	10	simple	simple	ADJ
aiti-8904	251	11	architecture	architecture	NOUN
aiti-8904	251	12	.	.	PUNCT
aiti-8904	252	1	the	the	DET
aiti-8904	252	2	time	time	NOUN
aiti-8904	252	3	interval	interval	NOUN
aiti-8904	252	4	for	for	ADP
aiti-8904	252	5	the	the	DET
aiti-8904	252	6	pressure	pressure	NOUN
aiti-8904	252	7	sensor	sensor	NOUN
aiti-8904	252	8	to	to	PART
aiti-8904	252	9	record	record	NOUN
aiti-8904	252	10	data	datum	NOUN
aiti-8904	252	11	is	be	AUX
aiti-8904	252	12	one	one	NUM
aiti-8904	252	13	second	second	NOUN
aiti-8904	252	14	.	.	PUNCT
aiti-8904	253	1	thus	thus	ADV
aiti-8904	253	2	,	,	PUNCT
aiti-8904	253	3	the	the	DET
aiti-8904	253	4	data	data	NOUN
aiti-8904	253	5	volume	volume	NOUN
aiti-8904	253	6	is	be	AUX
aiti-8904	253	7	also	also	ADV
aiti-8904	253	8	equal	equal	ADJ
aiti-8904	253	9	to	to	ADP
aiti-8904	253	10	the	the	DET
aiti-8904	253	11	sampling	sampling	NOUN
aiti-8904	253	12	time	time	NOUN
aiti-8904	253	13	of	of	ADP
aiti-8904	253	14	the	the	DET
aiti-8904	253	15	data	datum	NOUN
aiti-8904	253	16	volume	volume	NOUN
aiti-8904	253	17	.	.	PUNCT
aiti-8904	254	1	if	if	SCONJ
aiti-8904	254	2	the	the	DET
aiti-8904	254	3	real	real	ADJ
aiti-8904	254	4	-	-	PUNCT
aiti-8904	254	5	time	time	NOUN
aiti-8904	254	6	leak	leak	NOUN
aiti-8904	254	7	detection	detection	NOUN
aiti-8904	254	8	function	function	NOUN
aiti-8904	254	9	is	be	AUX
aiti-8904	254	10	to	to	PART
aiti-8904	254	11	be	be	AUX
aiti-8904	254	12	achieved	achieve	VERB
aiti-8904	254	13	,	,	PUNCT
aiti-8904	254	14	the	the	DET
aiti-8904	254	15	time	time	NOUN
aiti-8904	254	16	for	for	ADP
aiti-8904	254	17	calculating	calculate	VERB
aiti-8904	254	18	the	the	DET
aiti-8904	254	19	data	datum	NOUN
aiti-8904	254	20	of	of	ADP
aiti-8904	254	21	the	the	DET
aiti-8904	254	22	model	model	NOUN
aiti-8904	254	23	must	must	AUX
aiti-8904	254	24	be	be	AUX
aiti-8904	254	25	controlled	control	VERB
aiti-8904	254	26	within	within	ADP
aiti-8904	254	27	one	one	NUM
aiti-8904	254	28	second	second	NOUN
aiti-8904	254	29	.	.	PUNCT
aiti-8904	255	1	the	the	DET
aiti-8904	255	2	calculating	calculate	VERB
aiti-8904	255	3	time	time	NOUN
aiti-8904	255	4	on	on	ADP
aiti-8904	255	5	different	different	ADJ
aiti-8904	255	6	data	datum	NOUN
aiti-8904	255	7	volumes	volume	NOUN
aiti-8904	255	8	in	in	ADP
aiti-8904	255	9	the	the	DET
aiti-8904	255	10	model	model	NOUN
aiti-8904	255	11	is	be	AUX
aiti-8904	255	12	shown	show	VERB
aiti-8904	255	13	in	in	ADP
aiti-8904	255	14	table	table	NOUN
aiti-8904	255	15	2	2	NUM
aiti-8904	255	16	.	.	PUNCT
aiti-8904	256	1	the	the	DET
aiti-8904	256	2	experimental	experimental	ADJ
aiti-8904	256	3	results	result	NOUN
aiti-8904	256	4	show	show	VERB
aiti-8904	256	5	that	that	SCONJ
aiti-8904	256	6	if	if	SCONJ
aiti-8904	256	7	the	the	DET
aiti-8904	256	8	data	datum	NOUN
aiti-8904	256	9	recorded	record	VERB
aiti-8904	256	10	every	every	DET
aiti-8904	256	11	hour	hour	NOUN
aiti-8904	256	12	,	,	PUNCT
aiti-8904	256	13	3600	3600	NUM
aiti-8904	256	14	cases	case	NOUN
aiti-8904	256	15	,	,	PUNCT
aiti-8904	256	16	is	be	AUX
aiti-8904	256	17	used	use	VERB
aiti-8904	256	18	for	for	ADP
aiti-8904	256	19	leak	leak	NOUN
aiti-8904	256	20	detection	detection	NOUN
aiti-8904	256	21	,	,	PUNCT
aiti-8904	256	22	the	the	DET
aiti-8904	256	23	calculating	calculate	VERB
aiti-8904	256	24	time	time	NOUN
aiti-8904	256	25	for	for	SCONJ
aiti-8904	256	26	the	the	DET
aiti-8904	256	27	model	model	NOUN
aiti-8904	256	28	to	to	PART
aiti-8904	256	29	predict	predict	VERB
aiti-8904	256	30	the	the	DET
aiti-8904	256	31	leakage	leakage	NOUN
aiti-8904	256	32	can	can	AUX
aiti-8904	256	33	be	be	AUX
aiti-8904	256	34	below	below	ADP
aiti-8904	256	35	one	one	NUM
aiti-8904	256	36	second	second	NOUN
aiti-8904	256	37	,	,	PUNCT
aiti-8904	256	38	thus	thus	ADV
aiti-8904	256	39	being	be	AUX
aiti-8904	256	40	able	able	ADJ
aiti-8904	256	41	to	to	PART
aiti-8904	256	42	satisfy	satisfy	VERB
aiti-8904	256	43	the	the	DET
aiti-8904	256	44	requirement	requirement	NOUN
aiti-8904	256	45	of	of	ADP
aiti-8904	256	46	real	real	ADJ
aiti-8904	256	47	-	-	PUNCT
aiti-8904	256	48	time	time	NOUN
aiti-8904	256	49	leak	leak	NOUN
aiti-8904	256	50	detection	detection	NOUN
aiti-8904	256	51	.	.	PUNCT
aiti-8904	257	1	table	table	NOUN
aiti-8904	257	2	1	1	NUM
aiti-8904	257	3	results	result	NOUN
aiti-8904	257	4	of	of	ADP
aiti-8904	257	5	testing	test	VERB
aiti-8904	257	6	the	the	DET
aiti-8904	257	7	pipeline	pipeline	NOUN
aiti-8904	257	8	leakage	leakage	NOUN
aiti-8904	257	9	with	with	ADP
aiti-8904	257	10	different	different	ADJ
aiti-8904	257	11	hyperparameters	hyperparameter	NOUN
aiti-8904	257	12	model	model	NOUN
aiti-8904	257	13	parameters	parameter	NOUN
aiti-8904	257	14	pt	pt	PROPN
aiti-8904	257	15	-	-	PUNCT
aiti-8904	257	16	m621	m621	PROPN
aiti-8904	257	17	pt	pt	PROPN
aiti-8904	257	18	-	-	PUNCT
aiti-8904	257	19	m622	m622	PROPN
aiti-8904	257	20	model	model	NOUN
aiti-8904	257	21	1	1	NUM
aiti-8904	257	22	max	max	PROPN
aiti-8904	257	23	pooling	pool	VERB
aiti-8904	257	24	kernel	kernel	PROPN
aiti-8904	257	25	size	size	NOUN
aiti-8904	257	26	=	=	SYM
aiti-8904	257	27	3	3	NUM
aiti-8904	257	28	fail	fail	VERB
aiti-8904	257	29	success	success	NOUN
aiti-8904	257	30	model	model	NOUN
aiti-8904	257	31	2	2	NUM
aiti-8904	257	32	average	average	ADJ
aiti-8904	257	33	pooling	pool	VERB
aiti-8904	257	34	kernel	kernel	NOUN
aiti-8904	257	35	size	size	NOUN
aiti-8904	257	36	=	=	SYM
aiti-8904	257	37	3	3	NUM
aiti-8904	257	38	fail	fail	VERB
aiti-8904	257	39	success	success	NOUN
aiti-8904	257	40	model	model	NOUN
aiti-8904	257	41	3	3	NUM
aiti-8904	257	42	average	average	ADJ
aiti-8904	257	43	pooling	pool	VERB
aiti-8904	257	44	kernel	kernel	NOUN
aiti-8904	257	45	size	size	NOUN
aiti-8904	257	46	=	=	SYM
aiti-8904	257	47	5	5	NUM
aiti-8904	257	48	success	success	NOUN
aiti-8904	257	49	success	success	NOUN
aiti-8904	257	50	table	table	NOUN
aiti-8904	257	51	2	2	NUM
aiti-8904	257	52	calculating	calculate	VERB
aiti-8904	257	53	time	time	NOUN
aiti-8904	257	54	of	of	ADP
aiti-8904	257	55	the	the	DET
aiti-8904	257	56	model	model	NOUN
aiti-8904	257	57	for	for	ADP
aiti-8904	257	58	different	different	ADJ
aiti-8904	257	59	data	datum	NOUN
aiti-8904	257	60	volume	volume	NOUN
aiti-8904	257	61	data	datum	NOUN
aiti-8904	257	62	volume	volume	NOUN
aiti-8904	257	63	(	(	PUNCT
aiti-8904	257	64	cases	case	NOUN
aiti-8904	257	65	)	)	PUNCT
aiti-8904	257	66	calculating	calculate	VERB
aiti-8904	257	67	time	time	NOUN
aiti-8904	257	68	(	(	PUNCT
aiti-8904	257	69	second	second	ADJ
aiti-8904	257	70	)	)	PUNCT
aiti-8904	257	71	86400	86400	NUM
aiti-8904	257	72	3.339572191238403	3.339572191238403	NUM
aiti-8904	257	73	43940	43940	NUM
aiti-8904	257	74	1.904508652279663	1.904508652279663	NUM
aiti-8904	257	75	3600	3600	NUM
aiti-8904	257	76	0.636376142501831	0.636376142501831	NUM
aiti-8904	257	77	600	600	NUM
aiti-8904	257	78	0.476074934005737	0.476074934005737	NUM
aiti-8904	257	79	fig	fig	NOUN
aiti-8904	257	80	.	.	PUNCT
aiti-8904	258	1	17	17	NUM
aiti-8904	258	2	results	result	NOUN
aiti-8904	258	3	of	of	ADP
aiti-8904	258	4	leak	leak	NOUN
aiti-8904	258	5	identification	identification	NOUN
aiti-8904	258	6	for	for	ADP
aiti-8904	258	7	model	model	NOUN
aiti-8904	258	8	1	1	NUM
aiti-8904	258	9	(	(	PUNCT
aiti-8904	258	10	using	use	VERB
aiti-8904	258	11	the	the	DET
aiti-8904	258	12	pt	pt	PROPN
aiti-8904	258	13	-	-	PUNCT
aiti-8904	258	14	m621	m621	PROPN
aiti-8904	258	15	pressure	pressure	NOUN
aiti-8904	258	16	sensor	sensor	NOUN
aiti-8904	258	17	)	)	PUNCT
aiti-8904	258	18	fig	fig	NOUN
aiti-8904	258	19	.	.	PUNCT
aiti-8904	259	1	18	18	NUM
aiti-8904	259	2	results	result	NOUN
aiti-8904	259	3	of	of	ADP
aiti-8904	259	4	leak	leak	NOUN
aiti-8904	259	5	identification	identification	NOUN
aiti-8904	259	6	for	for	ADP
aiti-8904	259	7	model	model	NOUN
aiti-8904	259	8	1	1	NUM
aiti-8904	259	9	(	(	PUNCT
aiti-8904	259	10	using	use	VERB
aiti-8904	259	11	the	the	DET
aiti-8904	259	12	pt	pt	PROPN
aiti-8904	259	13	-	-	PUNCT
aiti-8904	259	14	m622	m622	PROPN
aiti-8904	259	15	pressure	pressure	NOUN
aiti-8904	259	16	sensor	sensor	NOUN
aiti-8904	259	17	)	)	PUNCT
aiti-8904	259	18	178	178	NUM
aiti-8904	259	19	advances	advance	NOUN
aiti-8904	259	20	in	in	ADP
aiti-8904	259	21	technology	technology	NOUN
aiti-8904	259	22	innovation	innovation	NOUN
aiti-8904	259	23	,	,	PUNCT
aiti-8904	259	24	vol	vol	NOUN
aiti-8904	259	25	.	.	PROPN
aiti-8904	259	26	7	7	NUM
aiti-8904	259	27	,	,	PUNCT
aiti-8904	259	28	no	no	INTJ
aiti-8904	259	29	.	.	NOUN
aiti-8904	259	30	3	3	NUM
aiti-8904	259	31	,	,	PUNCT
aiti-8904	259	32	2022	2022	NUM
aiti-8904	259	33	,	,	PUNCT
aiti-8904	259	34	pp	pp	ADJ
aiti-8904	259	35	.	.	PUNCT
aiti-8904	259	36	169	169	NUM
aiti-8904	259	37	-	-	SYM
aiti-8904	259	38	180	180	NUM
aiti-8904	259	39	fig	fig	NOUN
aiti-8904	259	40	.	.	PUNCT
aiti-8904	260	1	19	19	NUM
aiti-8904	260	2	results	result	NOUN
aiti-8904	260	3	of	of	ADP
aiti-8904	260	4	leak	leak	NOUN
aiti-8904	260	5	identification	identification	NOUN
aiti-8904	260	6	for	for	ADP
aiti-8904	260	7	model	model	NOUN
aiti-8904	260	8	2	2	NUM
aiti-8904	260	9	(	(	PUNCT
aiti-8904	260	10	using	use	VERB
aiti-8904	260	11	the	the	DET
aiti-8904	260	12	pt	pt	PROPN
aiti-8904	260	13	-	-	PUNCT
aiti-8904	260	14	m621	m621	PROPN
aiti-8904	260	15	pressure	pressure	NOUN
aiti-8904	260	16	sensor	sensor	NOUN
aiti-8904	260	17	)	)	PUNCT
aiti-8904	260	18	fig	fig	NOUN
aiti-8904	260	19	.	.	PUNCT
aiti-8904	261	1	20	20	NUM
aiti-8904	261	2	results	result	NOUN
aiti-8904	261	3	of	of	ADP
aiti-8904	261	4	leak	leak	NOUN
aiti-8904	261	5	identification	identification	NOUN
aiti-8904	261	6	for	for	ADP
aiti-8904	261	7	model	model	NOUN
aiti-8904	261	8	2	2	NUM
aiti-8904	261	9	(	(	PUNCT
aiti-8904	261	10	using	use	VERB
aiti-8904	261	11	the	the	DET
aiti-8904	261	12	pt	pt	PROPN
aiti-8904	261	13	-	-	PUNCT
aiti-8904	261	14	m622	m622	PROPN
aiti-8904	261	15	pressure	pressure	NOUN
aiti-8904	261	16	sensor	sensor	NOUN
aiti-8904	261	17	)	)	PUNCT
aiti-8904	261	18	fig	fig	NOUN
aiti-8904	261	19	.	.	PUNCT
aiti-8904	262	1	21	21	NUM
aiti-8904	262	2	results	result	NOUN
aiti-8904	262	3	of	of	ADP
aiti-8904	262	4	leak	leak	NOUN
aiti-8904	262	5	identification	identification	NOUN
aiti-8904	262	6	for	for	ADP
aiti-8904	262	7	model	model	NOUN
aiti-8904	262	8	3	3	NUM
aiti-8904	262	9	(	(	PUNCT
aiti-8904	262	10	using	use	VERB
aiti-8904	262	11	the	the	DET
aiti-8904	262	12	pt	pt	PROPN
aiti-8904	262	13	-	-	PUNCT
aiti-8904	262	14	m621	m621	PROPN
aiti-8904	262	15	pressure	pressure	NOUN
aiti-8904	262	16	sensor	sensor	NOUN
aiti-8904	262	17	)	)	PUNCT
aiti-8904	262	18	fig	fig	NOUN
aiti-8904	262	19	.	.	PUNCT
aiti-8904	263	1	22	22	NUM
aiti-8904	263	2	results	result	NOUN
aiti-8904	263	3	of	of	ADP
aiti-8904	263	4	leak	leak	NOUN
aiti-8904	263	5	identification	identification	NOUN
aiti-8904	263	6	for	for	ADP
aiti-8904	263	7	model	model	NOUN
aiti-8904	263	8	3	3	NUM
aiti-8904	263	9	(	(	PUNCT
aiti-8904	263	10	using	use	VERB
aiti-8904	263	11	the	the	DET
aiti-8904	263	12	pt	pt	PROPN
aiti-8904	263	13	-	-	PUNCT
aiti-8904	263	14	m622	m622	PROPN
aiti-8904	263	15	pressure	pressure	NOUN
aiti-8904	263	16	sensor	sensor	NOUN
aiti-8904	263	17	)	)	PUNCT
aiti-8904	263	18	the	the	DET
aiti-8904	263	19	result	result	NOUN
aiti-8904	263	20	can	can	AUX
aiti-8904	263	21	be	be	AUX
aiti-8904	263	22	summarized	summarize	VERB
aiti-8904	263	23	as	as	SCONJ
aiti-8904	263	24	follows	follow	VERB
aiti-8904	263	25	.	.	PUNCT
aiti-8904	264	1	first	first	ADV
aiti-8904	264	2	,	,	PUNCT
aiti-8904	264	3	based	base	VERB
aiti-8904	264	4	on	on	ADP
aiti-8904	264	5	the	the	DET
aiti-8904	264	6	physical	physical	ADJ
aiti-8904	264	7	features	feature	NOUN
aiti-8904	264	8	of	of	ADP
aiti-8904	264	9	the	the	DET
aiti-8904	264	10	pipeline	pipeline	NOUN
aiti-8904	264	11	,	,	PUNCT
aiti-8904	264	12	the	the	DET
aiti-8904	264	13	ai	ai	PROPN
aiti-8904	264	14	model	model	NOUN
aiti-8904	264	15	of	of	ADP
aiti-8904	264	16	pipeline	pipeline	NOUN
aiti-8904	264	17	transportation	transportation	NOUN
aiti-8904	264	18	is	be	AUX
aiti-8904	264	19	derived	derive	VERB
aiti-8904	264	20	.	.	PUNCT
aiti-8904	265	1	second	second	ADJ
aiti-8904	265	2	,	,	PUNCT
aiti-8904	265	3	based	base	VERB
aiti-8904	265	4	on	on	ADP
aiti-8904	265	5	the	the	DET
aiti-8904	265	6	ai	ai	PROPN
aiti-8904	265	7	model	model	NOUN
aiti-8904	265	8	,	,	PUNCT
aiti-8904	265	9	the	the	DET
aiti-8904	265	10	pipeline	pipeline	NOUN
aiti-8904	265	11	leak	leak	NOUN
aiti-8904	265	12	patterns	pattern	NOUN
aiti-8904	265	13	are	be	AUX
aiti-8904	265	14	analyzed	analyze	VERB
aiti-8904	265	15	.	.	PUNCT
aiti-8904	266	1	it	it	PRON
aiti-8904	266	2	is	be	AUX
aiti-8904	266	3	found	find	VERB
aiti-8904	266	4	that	that	SCONJ
aiti-8904	266	5	the	the	DET
aiti-8904	266	6	reduced	reduced	ADJ
aiti-8904	266	7	model	model	NOUN
aiti-8904	266	8	of	of	ADP
aiti-8904	266	9	detecting	detect	VERB
aiti-8904	266	10	the	the	DET
aiti-8904	266	11	pipeline	pipeline	NOUN
aiti-8904	266	12	leakage	leakage	NOUN
aiti-8904	266	13	can	can	AUX
aiti-8904	266	14	be	be	AUX
aiti-8904	266	15	used	use	VERB
aiti-8904	266	16	to	to	PART
aiti-8904	266	17	detect	detect	VERB
aiti-8904	266	18	the	the	DET
aiti-8904	266	19	leakage	leakage	NOUN
aiti-8904	266	20	from	from	ADP
aiti-8904	266	21	the	the	DET
aiti-8904	266	22	upstream	upstream	ADJ
aiti-8904	266	23	and	and	CCONJ
aiti-8904	266	24	downstream	downstream	ADJ
aiti-8904	266	25	pipelines	pipeline	NOUN
aiti-8904	266	26	.	.	PUNCT
aiti-8904	267	1	finally	finally	ADV
aiti-8904	267	2	,	,	PUNCT
aiti-8904	267	3	the	the	DET
aiti-8904	267	4	model	model	NOUN
aiti-8904	267	5	is	be	AUX
aiti-8904	267	6	tested	test	VERB
aiti-8904	267	7	in	in	ADP
aiti-8904	267	8	an	an	DET
aiti-8904	267	9	existing	exist	VERB
aiti-8904	267	10	natural	natural	ADJ
aiti-8904	267	11	gas	gas	NOUN
aiti-8904	267	12	pipeline	pipeline	NOUN
aiti-8904	267	13	system	system	NOUN
aiti-8904	267	14	starting	start	VERB
aiti-8904	267	15	from	from	ADP
aiti-8904	267	16	mailiao	mailiao	PROPN
aiti-8904	267	17	to	to	ADP
aiti-8904	267	18	taoyuan	taoyuan	PROPN
aiti-8904	267	19	in	in	ADP
aiti-8904	267	20	taiwan	taiwan	PROPN
aiti-8904	267	21	.	.	PUNCT
aiti-8904	268	1	the	the	DET
aiti-8904	268	2	results	result	NOUN
aiti-8904	268	3	show	show	VERB
aiti-8904	268	4	that	that	SCONJ
aiti-8904	268	5	in	in	ADP
aiti-8904	268	6	the	the	DET
aiti-8904	268	7	proposed	propose	VERB
aiti-8904	268	8	ai	ai	NOUN
aiti-8904	268	9	-	-	PUNCT
aiti-8904	268	10	based	base	VERB
aiti-8904	268	11	leak	leak	NOUN
aiti-8904	268	12	detection	detection	NOUN
aiti-8904	268	13	system	system	NOUN
aiti-8904	268	14	,	,	PUNCT
aiti-8904	268	15	the	the	DET
aiti-8904	268	16	correctness	correctness	NOUN
aiti-8904	268	17	concerning	concern	VERB
aiti-8904	268	18	the	the	DET
aiti-8904	268	19	detection	detection	NOUN
aiti-8904	268	20	function	function	NOUN
aiti-8904	268	21	is	be	AUX
aiti-8904	268	22	not	not	PART
aiti-8904	268	23	affected	affect	VERB
aiti-8904	268	24	by	by	ADP
aiti-8904	268	25	various	various	ADJ
aiti-8904	268	26	operational	operational	ADJ
aiti-8904	268	27	conditions	condition	NOUN
aiti-8904	268	28	of	of	ADP
aiti-8904	268	29	the	the	DET
aiti-8904	268	30	pipelines	pipeline	NOUN
aiti-8904	268	31	.	.	PUNCT
aiti-8904	269	1	4	4	X
aiti-8904	269	2	.	.	X
aiti-8904	269	3	conclusions	conclusion	NOUN
aiti-8904	269	4	and	and	CCONJ
aiti-8904	269	5	directions	direction	NOUN
aiti-8904	269	6	for	for	ADP
aiti-8904	269	7	future	future	ADJ
aiti-8904	269	8	research	research	NOUN
aiti-8904	269	9	the	the	DET
aiti-8904	269	10	present	present	ADJ
aiti-8904	269	11	study	study	NOUN
aiti-8904	269	12	contributes	contribute	VERB
aiti-8904	269	13	to	to	ADP
aiti-8904	269	14	the	the	DET
aiti-8904	269	15	research	research	NOUN
aiti-8904	269	16	and	and	CCONJ
aiti-8904	269	17	practice	practice	NOUN
aiti-8904	269	18	of	of	ADP
aiti-8904	269	19	pipeline	pipeline	NOUN
aiti-8904	269	20	leak	leak	NOUN
aiti-8904	269	21	detection	detection	NOUN
aiti-8904	269	22	in	in	ADP
aiti-8904	269	23	the	the	DET
aiti-8904	269	24	following	following	ADJ
aiti-8904	269	25	ways	way	NOUN
aiti-8904	269	26	:	:	PUNCT
aiti-8904	269	27	(	(	PUNCT
aiti-8904	269	28	1	1	X
aiti-8904	269	29	)	)	PUNCT
aiti-8904	269	30	the	the	DET
aiti-8904	269	31	model	model	NOUN
aiti-8904	269	32	developed	develop	VERB
aiti-8904	269	33	by	by	ADP
aiti-8904	269	34	this	this	DET
aiti-8904	269	35	study	study	NOUN
aiti-8904	269	36	can	can	AUX
aiti-8904	269	37	effectively	effectively	ADV
aiti-8904	269	38	detect	detect	VERB
aiti-8904	269	39	the	the	DET
aiti-8904	269	40	leakage	leakage	NOUN
aiti-8904	269	41	in	in	ADP
aiti-8904	269	42	the	the	DET
aiti-8904	269	43	real	real	ADJ
aiti-8904	269	44	long	long	ADJ
aiti-8904	269	45	-	-	PUNCT
aiti-8904	269	46	distance	distance	NOUN
aiti-8904	269	47	transmission	transmission	NOUN
aiti-8904	269	48	pipeline	pipeline	NOUN
aiti-8904	269	49	system	system	NOUN
aiti-8904	269	50	.	.	PUNCT
aiti-8904	270	1	it	it	PRON
aiti-8904	270	2	can	can	AUX
aiti-8904	270	3	achieve	achieve	VERB
aiti-8904	270	4	the	the	DET
aiti-8904	270	5	same	same	ADJ
aiti-8904	270	6	performance	performance	NOUN
aiti-8904	270	7	as	as	ADP
aiti-8904	270	8	the	the	DET
aiti-8904	270	9	traditional	traditional	ADJ
aiti-8904	270	10	leak	leak	NOUN
aiti-8904	270	11	detection	detection	NOUN
aiti-8904	270	12	system	system	NOUN
aiti-8904	270	13	.	.	PUNCT
aiti-8904	271	1	(	(	PUNCT
aiti-8904	271	2	2	2	X
aiti-8904	271	3	)	)	PUNCT
aiti-8904	271	4	the	the	DET
aiti-8904	271	5	training	training	NOUN
aiti-8904	271	6	data	datum	NOUN
aiti-8904	271	7	used	use	VERB
aiti-8904	271	8	in	in	ADP
aiti-8904	271	9	this	this	DET
aiti-8904	271	10	study	study	NOUN
aiti-8904	271	11	is	be	AUX
aiti-8904	271	12	collected	collect	VERB
aiti-8904	271	13	from	from	ADP
aiti-8904	271	14	the	the	DET
aiti-8904	271	15	sensor	sensor	NOUN
aiti-8904	271	16	of	of	ADP
aiti-8904	271	17	the	the	DET
aiti-8904	271	18	original	original	ADJ
aiti-8904	271	19	pipeline	pipeline	NOUN
aiti-8904	271	20	system	system	NOUN
aiti-8904	271	21	.	.	PUNCT
aiti-8904	272	1	thus	thus	ADV
aiti-8904	272	2	,	,	PUNCT
aiti-8904	272	3	no	no	DET
aiti-8904	272	4	additional	additional	ADJ
aiti-8904	272	5	hardware	hardware	NOUN
aiti-8904	272	6	equipment	equipment	NOUN
aiti-8904	272	7	is	be	AUX
aiti-8904	272	8	required	require	VERB
aiti-8904	272	9	,	,	PUNCT
aiti-8904	272	10	which	which	PRON
aiti-8904	272	11	can	can	AUX
aiti-8904	272	12	save	save	VERB
aiti-8904	272	13	the	the	DET
aiti-8904	272	14	cost	cost	NOUN
aiti-8904	272	15	of	of	ADP
aiti-8904	272	16	hardware	hardware	NOUN
aiti-8904	272	17	equipment	equipment	NOUN
aiti-8904	272	18	.	.	PUNCT
aiti-8904	273	1	(	(	PUNCT
aiti-8904	273	2	3	3	X
aiti-8904	273	3	)	)	PUNCT
aiti-8904	273	4	it	it	PRON
aiti-8904	273	5	can	can	AUX
aiti-8904	273	6	run	run	VERB
aiti-8904	273	7	smoothly	smoothly	ADV
aiti-8904	273	8	without	without	ADP
aiti-8904	273	9	adding	add	VERB
aiti-8904	273	10	the	the	DET
aiti-8904	273	11	extra	extra	ADJ
aiti-8904	273	12	graphics	graphic	NOUN
aiti-8904	273	13	card	card	NOUN
aiti-8904	273	14	.	.	PUNCT
aiti-8904	274	1	the	the	DET
aiti-8904	274	2	operator	operator	NOUN
aiti-8904	274	3	can	can	AUX
aiti-8904	274	4	easily	easily	ADV
aiti-8904	274	5	load	load	VERB
aiti-8904	274	6	the	the	DET
aiti-8904	274	7	trained	train	VERB
aiti-8904	274	8	model	model	NOUN
aiti-8904	274	9	when	when	SCONJ
aiti-8904	274	10	performing	perform	VERB
aiti-8904	274	11	leak	leak	NOUN
aiti-8904	274	12	detection	detection	NOUN
aiti-8904	274	13	.	.	PUNCT
aiti-8904	275	1	nevertheless	nevertheless	ADV
aiti-8904	275	2	,	,	PUNCT
aiti-8904	275	3	some	some	DET
aiti-8904	275	4	limitations	limitation	NOUN
aiti-8904	275	5	remain	remain	VERB
aiti-8904	275	6	and	and	CCONJ
aiti-8904	275	7	need	need	VERB
aiti-8904	275	8	to	to	PART
aiti-8904	275	9	be	be	AUX
aiti-8904	275	10	resolved	resolve	VERB
aiti-8904	275	11	in	in	ADP
aiti-8904	275	12	further	further	ADJ
aiti-8904	275	13	research	research	NOUN
aiti-8904	275	14	:	:	PUNCT
aiti-8904	275	15	(	(	PUNCT
aiti-8904	275	16	1	1	X
aiti-8904	275	17	)	)	PUNCT
aiti-8904	275	18	this	this	DET
aiti-8904	275	19	study	study	NOUN
aiti-8904	275	20	uses	use	VERB
aiti-8904	275	21	the	the	DET
aiti-8904	275	22	pressure	pressure	NOUN
aiti-8904	275	23	sensor	sensor	NOUN
aiti-8904	275	24	data	datum	NOUN
aiti-8904	275	25	collected	collect	VERB
aiti-8904	275	26	from	from	ADP
aiti-8904	275	27	a	a	DET
aiti-8904	275	28	single	single	ADJ
aiti-8904	275	29	site	site	NOUN
aiti-8904	275	30	to	to	PART
aiti-8904	275	31	train	train	VERB
aiti-8904	275	32	the	the	DET
aiti-8904	275	33	model	model	NOUN
aiti-8904	275	34	.	.	PUNCT
aiti-8904	276	1	however	however	ADV
aiti-8904	276	2	,	,	PUNCT
aiti-8904	276	3	the	the	DET
aiti-8904	276	4	data	datum	NOUN
aiti-8904	276	5	recorded	record	VERB
aiti-8904	276	6	by	by	ADP
aiti-8904	276	7	the	the	DET
aiti-8904	276	8	pressure	pressure	NOUN
aiti-8904	276	9	sensor	sensor	NOUN
aiti-8904	276	10	at	at	ADP
aiti-8904	276	11	different	different	ADJ
aiti-8904	276	12	sites	site	NOUN
aiti-8904	276	13	may	may	AUX
aiti-8904	276	14	be	be	AUX
aiti-8904	276	15	different	different	ADJ
aiti-8904	276	16	.	.	PUNCT
aiti-8904	277	1	it	it	PRON
aiti-8904	277	2	will	will	AUX
aiti-8904	277	3	be	be	AUX
aiti-8904	277	4	better	well	ADJ
aiti-8904	277	5	to	to	PART
aiti-8904	277	6	collect	collect	VERB
aiti-8904	277	7	the	the	DET
aiti-8904	277	8	training	training	NOUN
aiti-8904	277	9	data	datum	NOUN
aiti-8904	277	10	from	from	ADP
aiti-8904	277	11	different	different	ADJ
aiti-8904	277	12	sites	site	NOUN
aiti-8904	277	13	in	in	ADP
aiti-8904	277	14	the	the	DET
aiti-8904	277	15	pipeline	pipeline	NOUN
aiti-8904	277	16	system	system	NOUN
aiti-8904	277	17	in	in	ADP
aiti-8904	277	18	future	future	ADJ
aiti-8904	277	19	research	research	NOUN
aiti-8904	277	20	.	.	PUNCT
aiti-8904	278	1	179	179	NUM
aiti-8904	278	2	advances	advance	NOUN
aiti-8904	278	3	in	in	ADP
aiti-8904	278	4	technology	technology	NOUN
aiti-8904	278	5	innovation	innovation	NOUN
aiti-8904	278	6	,	,	PUNCT
aiti-8904	278	7	vol	vol	NOUN
aiti-8904	278	8	.	.	PROPN
aiti-8904	278	9	7	7	NUM
aiti-8904	278	10	,	,	PUNCT
aiti-8904	278	11	no	no	INTJ
aiti-8904	278	12	.	.	NOUN
aiti-8904	278	13	3	3	NUM
aiti-8904	278	14	,	,	PUNCT
aiti-8904	278	15	2022	2022	NUM
aiti-8904	278	16	,	,	PUNCT
aiti-8904	278	17	pp	pp	ADJ
aiti-8904	278	18	.	.	PUNCT
aiti-8904	279	1	169	169	NUM
aiti-8904	279	2	-	-	SYM
aiti-8904	279	3	180	180	NUM
aiti-8904	279	4	(	(	PUNCT
aiti-8904	279	5	2	2	NUM
aiti-8904	279	6	)	)	PUNCT
aiti-8904	279	7	the	the	DET
aiti-8904	279	8	present	present	ADJ
aiti-8904	279	9	study	study	NOUN
aiti-8904	279	10	only	only	ADV
aiti-8904	279	11	uses	use	VERB
aiti-8904	279	12	the	the	DET
aiti-8904	279	13	basic	basic	ADJ
aiti-8904	279	14	cnn	cnn	NOUN
aiti-8904	279	15	architecture	architecture	NOUN
aiti-8904	279	16	for	for	ADP
aiti-8904	279	17	model	model	NOUN
aiti-8904	279	18	training	training	NOUN
aiti-8904	279	19	and	and	CCONJ
aiti-8904	279	20	collects	collect	VERB
aiti-8904	279	21	the	the	DET
aiti-8904	279	22	data	datum	NOUN
aiti-8904	279	23	from	from	ADP
aiti-8904	279	24	one	one	NUM
aiti-8904	279	25	site	site	NOUN
aiti-8904	279	26	in	in	ADP
aiti-8904	279	27	the	the	DET
aiti-8904	279	28	pipeline	pipeline	NOUN
aiti-8904	279	29	.	.	PUNCT
aiti-8904	280	1	future	future	ADJ
aiti-8904	280	2	research	research	NOUN
aiti-8904	280	3	can	can	AUX
aiti-8904	280	4	collect	collect	VERB
aiti-8904	280	5	sensor	sensor	NOUN
aiti-8904	280	6	values	value	NOUN
aiti-8904	280	7	from	from	ADP
aiti-8904	280	8	other	other	ADJ
aiti-8904	280	9	sites	site	NOUN
aiti-8904	280	10	and	and	CCONJ
aiti-8904	280	11	employ	employ	VERB
aiti-8904	280	12	rnn	rnn	NOUN
aiti-8904	280	13	or	or	CCONJ
aiti-8904	280	14	lstm	lstm	NOUN
aiti-8904	280	15	algorithms	algorithm	NOUN
aiti-8904	280	16	for	for	ADP
aiti-8904	280	17	processing	process	VERB
aiti-8904	280	18	the	the	DET
aiti-8904	280	19	time	time	NOUN
aiti-8904	280	20	-	-	PUNCT
aiti-8904	280	21	series	series	NOUN
aiti-8904	280	22	data	datum	NOUN
aiti-8904	280	23	.	.	PUNCT
aiti-8904	281	1	(	(	PUNCT
aiti-8904	281	2	3	3	NUM
aiti-8904	281	3	)	)	PUNCT
aiti-8904	281	4	only	only	ADV
aiti-8904	281	5	a	a	DET
aiti-8904	281	6	few	few	ADJ
aiti-8904	281	7	pressure	pressure	NOUN
aiti-8904	281	8	data	datum	NOUN
aiti-8904	281	9	for	for	ADP
aiti-8904	281	10	pipeline	pipeline	NOUN
aiti-8904	281	11	leakage	leakage	NOUN
aiti-8904	281	12	can	can	AUX
aiti-8904	281	13	be	be	AUX
aiti-8904	281	14	collected	collect	VERB
aiti-8904	281	15	due	due	ADP
aiti-8904	281	16	to	to	ADP
aiti-8904	281	17	the	the	DET
aiti-8904	281	18	expensive	expensive	ADJ
aiti-8904	281	19	testing	testing	NOUN
aiti-8904	281	20	cost	cost	NOUN
aiti-8904	281	21	of	of	ADP
aiti-8904	281	22	the	the	DET
aiti-8904	281	23	existing	exist	VERB
aiti-8904	281	24	detection	detection	NOUN
aiti-8904	281	25	system	system	NOUN
aiti-8904	281	26	.	.	PUNCT
aiti-8904	282	1	thus	thus	ADV
aiti-8904	282	2	,	,	PUNCT
aiti-8904	282	3	the	the	DET
aiti-8904	282	4	issue	issue	NOUN
aiti-8904	282	5	of	of	ADP
aiti-8904	282	6	overfitting	overfitte	VERB
aiti-8904	282	7	and	and	CCONJ
aiti-8904	282	8	false	false	ADJ
aiti-8904	282	9	negatives	negative	NOUN
aiti-8904	282	10	may	may	AUX
aiti-8904	282	11	occur	occur	VERB
aiti-8904	282	12	in	in	ADP
aiti-8904	282	13	the	the	DET
aiti-8904	282	14	proposed	propose	VERB
aiti-8904	282	15	model	model	NOUN
aiti-8904	282	16	.	.	PUNCT
aiti-8904	283	1	more	more	ADJ
aiti-8904	283	2	pressure	pressure	NOUN
aiti-8904	283	3	data	datum	NOUN
aiti-8904	283	4	should	should	AUX
aiti-8904	283	5	be	be	AUX
aiti-8904	283	6	collected	collect	VERB
aiti-8904	283	7	in	in	ADP
aiti-8904	283	8	further	further	ADJ
aiti-8904	283	9	research	research	NOUN
aiti-8904	283	10	.	.	PUNCT
aiti-8904	284	1	conflicts	conflict	NOUN
aiti-8904	284	2	of	of	ADP
aiti-8904	284	3	interest	interest	NOUN
aiti-8904	284	4	the	the	DET
aiti-8904	284	5	authors	author	NOUN
aiti-8904	284	6	declare	declare	VERB
aiti-8904	284	7	no	no	DET
aiti-8904	284	8	conflicts	conflict	NOUN
aiti-8904	284	9	of	of	ADP
aiti-8904	284	10	interest	interest	NOUN
aiti-8904	284	11	.	.	PUNCT
aiti-8904	285	1	references	reference	NOUN
aiti-8904	285	2	[	[	X
aiti-8904	285	3	1	1	NUM
aiti-8904	285	4	]	]	PUNCT
aiti-8904	285	5	m.	m.	NOUN
aiti-8904	285	6	meribout	meribout	NOUN
aiti-8904	285	7	,	,	PUNCT
aiti-8904	285	8	et	et	PROPN
aiti-8904	285	9	al	al	PROPN
aiti-8904	285	10	.	.	PROPN
aiti-8904	285	11	,	,	PUNCT
aiti-8904	285	12	“	"	PUNCT
aiti-8904	285	13	leak	leak	VERB
aiti-8904	285	14	detection	detection	NOUN
aiti-8904	285	15	systems	system	NOUN
aiti-8904	285	16	in	in	ADP
aiti-8904	285	17	oil	oil	NOUN
aiti-8904	285	18	and	and	CCONJ
aiti-8904	285	19	gas	gas	NOUN
aiti-8904	285	20	fields	field	NOUN
aiti-8904	285	21	:	:	PUNCT
aiti-8904	285	22	present	present	ADJ
aiti-8904	285	23	trends	trend	NOUN
aiti-8904	285	24	and	and	CCONJ
aiti-8904	285	25	future	future	ADJ
aiti-8904	285	26	prospects	prospect	NOUN
aiti-8904	285	27	,	,	PUNCT
aiti-8904	285	28	”	"	PUNCT
aiti-8904	285	29	flow	flow	NOUN
aiti-8904	285	30	measurement	measurement	NOUN
aiti-8904	285	31	and	and	CCONJ
aiti-8904	285	32	instrumentation	instrumentation	NOUN
aiti-8904	285	33	,	,	PUNCT
aiti-8904	285	34	vol	vol	NOUN
aiti-8904	285	35	.	.	PROPN
aiti-8904	285	36	75	75	NUM
aiti-8904	285	37	,	,	PUNCT
aiti-8904	285	38	article	article	NOUN
aiti-8904	285	39	no	no	NOUN
aiti-8904	285	40	.	.	PROPN
aiti-8904	285	41	101772	101772	NUM
aiti-8904	285	42	,	,	PUNCT
aiti-8904	285	43	october	october	PROPN
aiti-8904	285	44	2020	2020	NUM
aiti-8904	285	45	.	.	PUNCT
aiti-8904	286	1	[	[	X
aiti-8904	286	2	2	2	NUM
aiti-8904	286	3	]	]	PUNCT
aiti-8904	286	4	l.	l.	PROPN
aiti-8904	286	5	yang	yang	PROPN
aiti-8904	286	6	,	,	PUNCT
aiti-8904	286	7	et	et	PROPN
aiti-8904	286	8	al	al	PROPN
aiti-8904	286	9	.	.	PROPN
aiti-8904	286	10	,	,	PUNCT
aiti-8904	286	11	“	"	PUNCT
aiti-8904	286	12	a	a	DET
aiti-8904	286	13	novel	novel	NOUN
aiti-8904	286	14	ppa	ppa	NOUN
aiti-8904	286	15	method	method	NOUN
aiti-8904	286	16	for	for	ADP
aiti-8904	286	17	fluid	fluid	ADJ
aiti-8904	286	18	pipeline	pipeline	NOUN
aiti-8904	286	19	leak	leak	NOUN
aiti-8904	286	20	detection	detection	NOUN
aiti-8904	286	21	based	base	VERB
aiti-8904	286	22	on	on	ADP
aiti-8904	286	23	opelm	opelm	NOUN
aiti-8904	286	24	and	and	CCONJ
aiti-8904	286	25	bidirectional	bidirectional	ADJ
aiti-8904	286	26	lstm	lstm	NOUN
aiti-8904	286	27	,	,	PUNCT
aiti-8904	286	28	”	"	PUNCT
aiti-8904	286	29	ieee	ieee	NOUN
aiti-8904	286	30	access	access	NOUN
aiti-8904	286	31	,	,	PUNCT
aiti-8904	286	32	vol	vol	NOUN
aiti-8904	286	33	.	.	PROPN
aiti-8904	286	34	8	8	NUM
aiti-8904	286	35	,	,	PUNCT
aiti-8904	286	36	pp	pp	ADJ
aiti-8904	286	37	.	.	PUNCT
aiti-8904	287	1	107185	107185	NUM
aiti-8904	287	2	-	-	SYM
aiti-8904	287	3	107199	107199	NUM
aiti-8904	287	4	,	,	PUNCT
aiti-8904	287	5	2020	2020	NUM
aiti-8904	287	6	.	.	PUNCT
aiti-8904	288	1	[	[	X
aiti-8904	288	2	3	3	X
aiti-8904	288	3	]	]	PUNCT
aiti-8904	288	4	m.	m.	NOUN
aiti-8904	288	5	zhou	zhou	PROPN
aiti-8904	288	6	,	,	PUNCT
aiti-8904	288	7	et	et	PROPN
aiti-8904	288	8	al	al	PROPN
aiti-8904	288	9	.	.	PROPN
aiti-8904	288	10	,	,	PUNCT
aiti-8904	288	11	“	"	PUNCT
aiti-8904	288	12	leak	leak	VERB
aiti-8904	288	13	detection	detection	NOUN
aiti-8904	288	14	and	and	CCONJ
aiti-8904	288	15	location	location	NOUN
aiti-8904	288	16	based	base	VERB
aiti-8904	288	17	on	on	ADP
aiti-8904	288	18	islmd	islmd	NOUN
aiti-8904	288	19	and	and	CCONJ
aiti-8904	288	20	cnn	cnn	PROPN
aiti-8904	288	21	in	in	ADP
aiti-8904	288	22	a	a	DET
aiti-8904	288	23	pipeline	pipeline	NOUN
aiti-8904	288	24	,	,	PUNCT
aiti-8904	288	25	”	"	PUNCT
aiti-8904	288	26	ieee	ieee	NOUN
aiti-8904	288	27	access	access	NOUN
aiti-8904	288	28	,	,	PUNCT
aiti-8904	288	29	vol	vol	NOUN
aiti-8904	288	30	.	.	PROPN
aiti-8904	288	31	7	7	NUM
aiti-8904	288	32	,	,	PUNCT
aiti-8904	288	33	pp	pp	ADJ
aiti-8904	288	34	.	.	PUNCT
aiti-8904	289	1	30457	30457	NUM
aiti-8904	289	2	-	-	SYM
aiti-8904	289	3	30464	30464	NUM
aiti-8904	289	4	,	,	PUNCT
aiti-8904	289	5	2019	2019	NUM
aiti-8904	289	6	.	.	PUNCT
aiti-8904	290	1	[	[	X
aiti-8904	290	2	4	4	NUM
aiti-8904	290	3	]	]	PUNCT
aiti-8904	290	4	r.	r.	PROPN
aiti-8904	290	5	o.	o.	PROPN
aiti-8904	290	6	melo	melo	PROPN
aiti-8904	290	7	,	,	PUNCT
aiti-8904	290	8	et	et	PROPN
aiti-8904	290	9	al	al	PROPN
aiti-8904	290	10	.	.	PROPN
aiti-8904	290	11	,	,	PUNCT
aiti-8904	290	12	“	"	PUNCT
aiti-8904	290	13	applying	apply	VERB
aiti-8904	290	14	convolutional	convolutional	ADJ
aiti-8904	290	15	neural	neural	ADJ
aiti-8904	290	16	networks	network	NOUN
aiti-8904	290	17	to	to	PART
aiti-8904	290	18	detect	detect	VERB
aiti-8904	290	19	natural	natural	ADJ
aiti-8904	290	20	gas	gas	NOUN
aiti-8904	290	21	leaks	leak	NOUN
aiti-8904	290	22	in	in	ADP
aiti-8904	290	23	wellhead	wellhead	NOUN
aiti-8904	290	24	images	image	NOUN
aiti-8904	290	25	,	,	PUNCT
aiti-8904	290	26	”	"	PUNCT
aiti-8904	290	27	ieee	ieee	NOUN
aiti-8904	290	28	access	access	NOUN
aiti-8904	290	29	,	,	PUNCT
aiti-8904	290	30	vol	vol	NOUN
aiti-8904	290	31	.	.	PROPN
aiti-8904	290	32	8	8	NUM
aiti-8904	290	33	,	,	PUNCT
aiti-8904	290	34	pp	pp	ADJ
aiti-8904	290	35	.	.	PUNCT
aiti-8904	291	1	191775	191775	NUM
aiti-8904	291	2	-	-	SYM
aiti-8904	291	3	191784	191784	NUM
aiti-8904	291	4	,	,	PUNCT
aiti-8904	291	5	2020	2020	NUM
aiti-8904	291	6	.	.	PUNCT
aiti-8904	292	1	[	[	X
aiti-8904	292	2	5	5	X
aiti-8904	292	3	]	]	PUNCT
aiti-8904	292	4	j.	j.	PROPN
aiti-8904	292	5	li	li	PROPN
aiti-8904	292	6	,	,	PUNCT
aiti-8904	292	7	et	et	PROPN
aiti-8904	292	8	al	al	PROPN
aiti-8904	292	9	.	.	PROPN
aiti-8904	292	10	,	,	PUNCT
aiti-8904	292	11	“	"	PUNCT
aiti-8904	292	12	a	a	DET
aiti-8904	292	13	small	small	ADJ
aiti-8904	292	14	leakage	leakage	NOUN
aiti-8904	292	15	detection	detection	NOUN
aiti-8904	292	16	approach	approach	NOUN
aiti-8904	292	17	for	for	ADP
aiti-8904	292	18	gas	gas	NOUN
aiti-8904	292	19	pipelines	pipeline	NOUN
aiti-8904	292	20	based	base	VERB
aiti-8904	292	21	on	on	ADP
aiti-8904	292	22	cnn	cnn	PROPN
aiti-8904	292	23	,	,	PUNCT
aiti-8904	292	24	”	"	PUNCT
aiti-8904	292	25	2019	2019	NUM
aiti-8904	292	26	caa	caa	NOUN
aiti-8904	292	27	symposium	symposium	NOUN
aiti-8904	292	28	on	on	ADP
aiti-8904	292	29	fault	fault	NOUN
aiti-8904	292	30	detection	detection	NOUN
aiti-8904	292	31	,	,	PUNCT
aiti-8904	292	32	supervision	supervision	NOUN
aiti-8904	292	33	,	,	PUNCT
aiti-8904	292	34	and	and	CCONJ
aiti-8904	292	35	safety	safety	NOUN
aiti-8904	292	36	for	for	ADP
aiti-8904	292	37	technical	technical	ADJ
aiti-8904	292	38	processes	process	NOUN
aiti-8904	292	39	,	,	PUNCT
aiti-8904	292	40	pp	pp	ADP
aiti-8904	292	41	.	.	PUNCT
aiti-8904	293	1	390	390	NUM
aiti-8904	293	2	-	-	SYM
aiti-8904	293	3	394	394	NUM
aiti-8904	293	4	,	,	PUNCT
aiti-8904	293	5	july	july	PROPN
aiti-8904	293	6	2019	2019	NUM
aiti-8904	293	7	.	.	PUNCT
aiti-8904	294	1	[	[	X
aiti-8904	294	2	6	6	NUM
aiti-8904	294	3	]	]	PUNCT
aiti-8904	294	4	t.	t.	NOUN
aiti-8904	294	5	kattenborn	kattenborn	PROPN
aiti-8904	294	6	,	,	PUNCT
aiti-8904	294	7	et	et	PROPN
aiti-8904	294	8	al	al	PROPN
aiti-8904	294	9	.	.	PROPN
aiti-8904	294	10	,	,	PUNCT
aiti-8904	294	11	“	"	PUNCT
aiti-8904	294	12	review	review	NOUN
aiti-8904	294	13	on	on	ADP
aiti-8904	294	14	convolutional	convolutional	ADJ
aiti-8904	294	15	neural	neural	ADJ
aiti-8904	294	16	networks	network	NOUN
aiti-8904	294	17	(	(	PUNCT
aiti-8904	294	18	cnn	cnn	PROPN
aiti-8904	294	19	)	)	PUNCT
aiti-8904	294	20	in	in	ADP
aiti-8904	294	21	vegetation	vegetation	NOUN
aiti-8904	294	22	remote	remote	ADJ
aiti-8904	294	23	sensing	sensing	NOUN
aiti-8904	294	24	,	,	PUNCT
aiti-8904	294	25	”	"	PUNCT
aiti-8904	294	26	isprs	isprs	NOUN
aiti-8904	294	27	journal	journal	NOUN
aiti-8904	294	28	of	of	ADP
aiti-8904	294	29	photogrammetry	photogrammetry	NOUN
aiti-8904	294	30	and	and	CCONJ
aiti-8904	294	31	remote	remote	ADJ
aiti-8904	294	32	sensing	sensing	NOUN
aiti-8904	294	33	,	,	PUNCT
aiti-8904	294	34	vol	vol	NOUN
aiti-8904	294	35	.	.	PROPN
aiti-8904	294	36	173	173	NUM
aiti-8904	294	37	,	,	PUNCT
aiti-8904	294	38	pp	pp	ADJ
aiti-8904	294	39	.	.	PUNCT
aiti-8904	295	1	24	24	NUM
aiti-8904	295	2	-	-	SYM
aiti-8904	295	3	49	49	NUM
aiti-8904	295	4	,	,	PUNCT
aiti-8904	295	5	march	march	PROPN
aiti-8904	295	6	2021	2021	NUM
aiti-8904	295	7	.	.	PUNCT
aiti-8904	296	1	[	[	X
aiti-8904	296	2	7	7	X
aiti-8904	296	3	]	]	X
aiti-8904	296	4	j.	j.	PROPN
aiti-8904	296	5	kang	kang	PROPN
aiti-8904	296	6	,	,	PUNCT
aiti-8904	296	7	et	et	PROPN
aiti-8904	296	8	al	al	PROPN
aiti-8904	296	9	.	.	PROPN
aiti-8904	296	10	,	,	PUNCT
aiti-8904	296	11	“	"	PUNCT
aiti-8904	296	12	novel	novel	ADJ
aiti-8904	296	13	leakage	leakage	NOUN
aiti-8904	296	14	detection	detection	NOUN
aiti-8904	296	15	by	by	ADP
aiti-8904	296	16	ensemble	ensemble	ADJ
aiti-8904	296	17	cnn	cnn	PROPN
aiti-8904	296	18	-	-	PUNCT
aiti-8904	296	19	svm	svm	PROPN
aiti-8904	296	20	and	and	CCONJ
aiti-8904	296	21	graph	graph	NOUN
aiti-8904	296	22	-	-	PUNCT
aiti-8904	296	23	based	base	VERB
aiti-8904	296	24	localization	localization	NOUN
aiti-8904	296	25	in	in	ADP
aiti-8904	296	26	water	water	NOUN
aiti-8904	296	27	distribution	distribution	NOUN
aiti-8904	296	28	systems	system	NOUN
aiti-8904	296	29	,	,	PUNCT
aiti-8904	296	30	”	"	PUNCT
aiti-8904	296	31	ieee	ieee	NOUN
aiti-8904	296	32	transactions	transaction	NOUN
aiti-8904	296	33	on	on	ADP
aiti-8904	296	34	industrial	industrial	ADJ
aiti-8904	296	35	electronics	electronic	NOUN
aiti-8904	296	36	,	,	PUNCT
aiti-8904	296	37	vol	vol	NOUN
aiti-8904	296	38	.	.	PROPN
aiti-8904	296	39	65	65	NUM
aiti-8904	296	40	,	,	PUNCT
aiti-8904	296	41	no	no	INTJ
aiti-8904	296	42	.	.	NOUN
aiti-8904	296	43	5	5	NUM
aiti-8904	296	44	,	,	PUNCT
aiti-8904	296	45	pp	pp	ADJ
aiti-8904	296	46	.	.	PUNCT
aiti-8904	296	47	4279	4279	NUM
aiti-8904	296	48	-	-	SYM
aiti-8904	296	49	4289	4289	NUM
aiti-8904	296	50	,	,	PUNCT
aiti-8904	296	51	may	may	PROPN
aiti-8904	296	52	2018	2018	NUM
aiti-8904	296	53	.	.	PUNCT
aiti-8904	297	1	copyright	copyright	NOUN
aiti-8904	297	2	©	©	PROPN
aiti-8904	297	3	by	by	ADP
aiti-8904	297	4	the	the	DET
aiti-8904	297	5	authors	author	NOUN
aiti-8904	297	6	.	.	PUNCT
aiti-8904	298	1	licensee	licensee	PROPN
aiti-8904	298	2	taeti	taeti	PROPN
aiti-8904	298	3	,	,	PUNCT
aiti-8904	298	4	taiwan	taiwan	PROPN
aiti-8904	298	5	.	.	PUNCT
aiti-8904	299	1	this	this	DET
aiti-8904	299	2	article	article	NOUN
aiti-8904	299	3	is	be	AUX
aiti-8904	299	4	an	an	DET
aiti-8904	299	5	open	open	ADJ
aiti-8904	299	6	access	access	NOUN
aiti-8904	299	7	article	article	NOUN
aiti-8904	299	8	distributed	distribute	VERB
aiti-8904	299	9	under	under	ADP
aiti-8904	299	10	the	the	DET
aiti-8904	299	11	terms	term	NOUN
aiti-8904	299	12	and	and	CCONJ
aiti-8904	299	13	conditions	condition	NOUN
aiti-8904	299	14	of	of	ADP
aiti-8904	299	15	the	the	DET
aiti-8904	299	16	creative	creative	ADJ
aiti-8904	299	17	commons	common	NOUN
aiti-8904	299	18	attribution	attribution	NOUN
aiti-8904	299	19	(	(	PUNCT
aiti-8904	299	20	cc	cc	NOUN
aiti-8904	299	21	by	by	ADP
aiti-8904	299	22	-	-	PUNCT
aiti-8904	299	23	nc	nc	NOUN
aiti-8904	299	24	)	)	PUNCT
aiti-8904	299	25	license	license	NOUN
aiti-8904	299	26	(	(	PUNCT
aiti-8904	299	27	https://creativecommons.org/licenses/by-nc/4.0/	https://creativecommons.org/licenses/by-nc/4.0/	NOUN
aiti-8904	299	28	)	)	PUNCT
aiti-8904	299	29	.	.	PUNCT
aiti-8904	300	1	180	180	NUM