id	sid	tid	token	lemma	pos
fcis-30111	1	1	frontiers	frontier	NOUN
fcis-30111	1	2	in	in	ADP
fcis-30111	1	3	computing	computing	NOUN
fcis-30111	1	4	and	and	CCONJ
fcis-30111	1	5	intelligent	intelligent	ADJ
fcis-30111	1	6	systems	system	NOUN
fcis-30111	1	7	issn	issn	VERB
fcis-30111	1	8	:	:	PUNCT
fcis-30111	1	9	2832	2832	NUM
fcis-30111	1	10	-	-	SYM
fcis-30111	1	11	6024	6024	NUM
fcis-30111	1	12	|	|	NOUN
fcis-30111	1	13	vol	vol	NOUN
fcis-30111	1	14	.	.	PROPN
fcis-30111	2	1	11	11	NUM
fcis-30111	2	2	,	,	PUNCT
fcis-30111	2	3	no	no	INTJ
fcis-30111	2	4	.	.	NOUN
fcis-30111	2	5	3	3	NUM
fcis-30111	2	6	,	,	PUNCT
fcis-30111	2	7	2025	2025	NUM
fcis-30111	2	8	69	69	NUM
fcis-30111	2	9	multimodal	multimodal	NOUN
fcis-30111	2	10	data	datum	NOUN
fcis-30111	2	11	fusion	fusion	NOUN
fcis-30111	2	12	for	for	ADP
fcis-30111	2	13	digital	digital	ADJ
fcis-30111	2	14	monitoring	monitoring	NOUN
fcis-30111	2	15	of	of	ADP
fcis-30111	2	16	railway	railway	NOUN
fcis-30111	2	17	vehicle	vehicle	NOUN
fcis-30111	2	18	wheelset	wheelset	PROPN
fcis-30111	2	19	health	health	PROPN
fcis-30111	2	20	status	status	PROPN
fcis-30111	2	21	jun	jun	PROPN
fcis-30111	2	22	zhang	zhang	PROPN
fcis-30111	2	23	1	1	NUM
fcis-30111	2	24	,	,	PUNCT
fcis-30111	2	25	*	*	PROPN
fcis-30111	2	26	,	,	PUNCT
fcis-30111	2	27	jiaqi	jiaqi	PROPN
fcis-30111	2	28	zhang	zhang	PROPN
fcis-30111	2	29	2	2	NUM
fcis-30111	2	30	,	,	PUNCT
fcis-30111	2	31	a	a	DET
fcis-30111	2	32	1	1	NUM
fcis-30111	2	33	jilin	jilin	PROPN
fcis-30111	2	34	city	city	PROPN
fcis-30111	2	35	zhijian	zhijian	PROPN
fcis-30111	2	36	technology	technology	PROPN
fcis-30111	2	37	co.	co.	PROPN
fcis-30111	2	38	,	,	PUNCT
fcis-30111	2	39	ltd	ltd	PROPN
fcis-30111	2	40	.	.	PROPN
fcis-30111	2	41	,	,	PUNCT
fcis-30111	2	42	jilin	jilin	PROPN
fcis-30111	2	43	,	,	PUNCT
fcis-30111	2	44	jilin	jilin	PROPN
fcis-30111	2	45	,	,	PUNCT
fcis-30111	2	46	132000	132000	NUM
fcis-30111	2	47	,	,	PUNCT
fcis-30111	2	48	china	china	PROPN
fcis-30111	2	49	2	2	NUM
fcis-30111	2	50	jilin	jilin	PROPN
fcis-30111	2	51	railway	railway	PROPN
fcis-30111	2	52	technology	technology	PROPN
fcis-30111	2	53	college	college	PROPN
fcis-30111	2	54	,	,	PUNCT
fcis-30111	2	55	jilin	jilin	PROPN
fcis-30111	2	56	,	,	PUNCT
fcis-30111	2	57	jilin	jilin	PROPN
fcis-30111	2	58	,	,	PUNCT
fcis-30111	2	59	132200	132200	NUM
fcis-30111	2	60	,	,	PUNCT
fcis-30111	2	61	china	china	PROPN
fcis-30111	2	62	*	*	PUNCT
fcis-30111	2	63	corresponding	correspond	VERB
fcis-30111	2	64	author	author	NOUN
fcis-30111	2	65	:	:	PUNCT
fcis-30111	2	66	jun	jun	PROPN
fcis-30111	2	67	zhang	zhang	PROPN
fcis-30111	2	68	(	(	PUNCT
fcis-30111	2	69	email	email	NOUN
fcis-30111	2	70	:	:	PUNCT
fcis-30111	2	71	13578516081@163.com	13578516081@163.com	NUM
fcis-30111	2	72	)	)	PUNCT
fcis-30111	2	73	,	,	PUNCT
fcis-30111	2	74	a	a	DET
fcis-30111	2	75	2025799766@qq.com	2025799766@qq.com	NUM
fcis-30111	2	76	abstract	abstract	NOUN
fcis-30111	2	77	:	:	PUNCT
fcis-30111	2	78	this	this	DET
fcis-30111	2	79	study	study	NOUN
fcis-30111	2	80	proposes	propose	VERB
fcis-30111	2	81	a	a	DET
fcis-30111	2	82	multimodal	multimodal	ADJ
fcis-30111	2	83	data	datum	NOUN
fcis-30111	2	84	fusion	fusion	NOUN
fcis-30111	2	85	-	-	PUNCT
fcis-30111	2	86	based	base	VERB
fcis-30111	2	87	method	method	NOUN
fcis-30111	2	88	for	for	ADP
fcis-30111	2	89	monitoring	monitor	VERB
fcis-30111	2	90	the	the	DET
fcis-30111	2	91	health	health	NOUN
fcis-30111	2	92	status	status	NOUN
fcis-30111	2	93	of	of	ADP
fcis-30111	2	94	railway	railway	NOUN
fcis-30111	2	95	wheelsets	wheelset	NOUN
fcis-30111	2	96	by	by	ADP
fcis-30111	2	97	integrating	integrate	VERB
fcis-30111	2	98	vibration	vibration	NOUN
fcis-30111	2	99	,	,	PUNCT
fcis-30111	2	100	acoustic	acoustic	ADJ
fcis-30111	2	101	,	,	PUNCT
fcis-30111	2	102	temperature	temperature	NOUN
fcis-30111	2	103	and	and	CCONJ
fcis-30111	2	104	visual	visual	ADJ
fcis-30111	2	105	sensing	sense	VERB
fcis-30111	2	106	information	information	NOUN
fcis-30111	2	107	,	,	PUNCT
fcis-30111	2	108	and	and	CCONJ
fcis-30111	2	109	constructing	construct	VERB
fcis-30111	2	110	an	an	DET
fcis-30111	2	111	‘	'	PUNCT
fcis-30111	2	112	edge	edge	ADJ
fcis-30111	2	113	-	-	PUNCT
fcis-30111	2	114	cloud	cloud	NOUN
fcis-30111	2	115	’	'	PUNCT
fcis-30111	2	116	collaborative	collaborative	ADJ
fcis-30111	2	117	computing	computing	NOUN
fcis-30111	2	118	architecture	architecture	NOUN
fcis-30111	2	119	.	.	PUNCT
fcis-30111	3	1	the	the	DET
fcis-30111	3	2	research	research	NOUN
fcis-30111	3	3	innovatively	innovatively	ADV
fcis-30111	3	4	adopts	adopt	VERB
fcis-30111	3	5	a	a	DET
fcis-30111	3	6	deep	deep	ADJ
fcis-30111	3	7	fusion	fusion	NOUN
fcis-30111	3	8	model	model	NOUN
fcis-30111	3	9	with	with	ADP
fcis-30111	3	10	cross	cross	ADJ
fcis-30111	3	11	-	-	ADJ
fcis-30111	3	12	modal	modal	ADJ
fcis-30111	3	13	attention	attention	NOUN
fcis-30111	3	14	mechanism	mechanism	NOUN
fcis-30111	3	15	,	,	PUNCT
fcis-30111	3	16	and	and	CCONJ
fcis-30111	3	17	realises	realise	VERB
fcis-30111	3	18	high	high	ADJ
fcis-30111	3	19	-	-	PUNCT
fcis-30111	3	20	precision	precision	NOUN
fcis-30111	3	21	identification	identification	NOUN
fcis-30111	3	22	and	and	CCONJ
fcis-30111	3	23	early	early	ADJ
fcis-30111	3	24	warning	warning	NOUN
fcis-30111	3	25	of	of	ADP
fcis-30111	3	26	early	early	ADJ
fcis-30111	3	27	failure	failure	NOUN
fcis-30111	3	28	of	of	ADP
fcis-30111	3	29	wheelsets	wheelset	NOUN
fcis-30111	3	30	.	.	PUNCT
fcis-30111	4	1	practical	practical	ADJ
fcis-30111	4	2	engineering	engineering	NOUN
fcis-30111	4	3	applications	application	NOUN
fcis-30111	4	4	verify	verify	VERB
fcis-30111	4	5	that	that	SCONJ
fcis-30111	4	6	the	the	DET
fcis-30111	4	7	method	method	NOUN
fcis-30111	4	8	significantly	significantly	ADV
fcis-30111	4	9	reduces	reduce	VERB
fcis-30111	4	10	the	the	DET
fcis-30111	4	11	number	number	NOUN
fcis-30111	4	12	of	of	ADP
fcis-30111	4	13	unscheduled	unscheduled	ADJ
fcis-30111	4	14	wheelset	wheelset	NOUN
fcis-30111	4	15	replacements	replacement	NOUN
fcis-30111	4	16	and	and	CCONJ
fcis-30111	4	17	maintenance	maintenance	NOUN
fcis-30111	4	18	costs	cost	NOUN
fcis-30111	4	19	,	,	PUNCT
fcis-30111	4	20	and	and	CCONJ
fcis-30111	4	21	shows	show	VERB
fcis-30111	4	22	strong	strong	ADJ
fcis-30111	4	23	robustness	robustness	NOUN
fcis-30111	4	24	under	under	ADP
fcis-30111	4	25	severe	severe	ADJ
fcis-30111	4	26	operating	operating	NOUN
fcis-30111	4	27	conditions	condition	NOUN
fcis-30111	4	28	.	.	PUNCT
fcis-30111	5	1	the	the	DET
fcis-30111	5	2	reliability	reliability	NOUN
fcis-30111	5	3	and	and	CCONJ
fcis-30111	5	4	early	early	ADJ
fcis-30111	5	5	warning	warning	NOUN
fcis-30111	5	6	capability	capability	NOUN
fcis-30111	5	7	of	of	ADP
fcis-30111	5	8	the	the	DET
fcis-30111	5	9	system	system	NOUN
fcis-30111	5	10	are	be	AUX
fcis-30111	5	11	greatly	greatly	ADV
fcis-30111	5	12	improved	improve	VERB
fcis-30111	5	13	,	,	PUNCT
fcis-30111	5	14	providing	provide	VERB
fcis-30111	5	15	a	a	DET
fcis-30111	5	16	new	new	ADJ
fcis-30111	5	17	paradigm	paradigm	NOUN
fcis-30111	5	18	for	for	ADP
fcis-30111	5	19	early	early	ADJ
fcis-30111	5	20	warning	warning	NOUN
fcis-30111	5	21	of	of	ADP
fcis-30111	5	22	wheelset	wheelset	ADJ
fcis-30111	5	23	failures	failure	NOUN
fcis-30111	5	24	with	with	ADP
fcis-30111	5	25	obvious	obvious	ADJ
fcis-30111	5	26	technical	technical	ADJ
fcis-30111	5	27	and	and	CCONJ
fcis-30111	5	28	economic	economic	ADJ
fcis-30111	5	29	benefits	benefit	NOUN
fcis-30111	5	30	.	.	PUNCT
fcis-30111	6	1	keywords	keyword	NOUN
fcis-30111	6	2	:	:	PUNCT
fcis-30111	6	3	multimodal	multimodal	NOUN
fcis-30111	6	4	data	datum	NOUN
fcis-30111	6	5	fusion	fusion	NOUN
fcis-30111	6	6	;	;	PUNCT
fcis-30111	6	7	wheelset	wheelset	PROPN
fcis-30111	6	8	health	health	NOUN
fcis-30111	6	9	monitoring	monitoring	NOUN
fcis-30111	6	10	;	;	PUNCT
fcis-30111	6	11	cross	cross	ADJ
fcis-30111	6	12	-	-	ADJ
fcis-30111	6	13	modal	modal	ADJ
fcis-30111	6	14	attention	attention	NOUN
fcis-30111	6	15	mechanism	mechanism	NOUN
fcis-30111	6	16	;	;	PUNCT
fcis-30111	6	17	failure	failure	NOUN
fcis-30111	6	18	early	early	ADJ
fcis-30111	6	19	warning	warning	NOUN
fcis-30111	6	20	;	;	PUNCT
fcis-30111	6	21	edge	edge	NOUN
fcis-30111	6	22	computing	computing	NOUN
fcis-30111	6	23	.	.	PUNCT
fcis-30111	7	1	1	1	X
fcis-30111	7	2	.	.	X
fcis-30111	7	3	introduction	introduction	NOUN
fcis-30111	7	4	as	as	ADP
fcis-30111	7	5	a	a	DET
fcis-30111	7	6	key	key	ADJ
fcis-30111	7	7	component	component	NOUN
fcis-30111	7	8	of	of	ADP
fcis-30111	7	9	train	train	NOUN
fcis-30111	7	10	operation	operation	NOUN
fcis-30111	7	11	,	,	PUNCT
fcis-30111	7	12	the	the	DET
fcis-30111	7	13	health	health	NOUN
fcis-30111	7	14	status	status	NOUN
fcis-30111	7	15	of	of	ADP
fcis-30111	7	16	railway	railway	NOUN
fcis-30111	7	17	vehicle	vehicle	NOUN
fcis-30111	7	18	wheelsets	wheelset	NOUN
fcis-30111	7	19	is	be	AUX
fcis-30111	7	20	directly	directly	ADV
fcis-30111	7	21	related	relate	VERB
fcis-30111	7	22	to	to	PART
fcis-30111	7	23	train	train	VERB
fcis-30111	7	24	safety	safety	NOUN
fcis-30111	7	25	and	and	CCONJ
fcis-30111	7	26	operational	operational	ADJ
fcis-30111	7	27	efficiency	efficiency	NOUN
fcis-30111	7	28	[	[	X
fcis-30111	7	29	1	1	NUM
fcis-30111	7	30	]	]	PUNCT
fcis-30111	7	31	.	.	PUNCT
fcis-30111	8	1	traditional	traditional	ADJ
fcis-30111	8	2	wheelset	wheelset	ADJ
fcis-30111	8	3	monitoring	monitoring	NOUN
fcis-30111	8	4	technology	technology	NOUN
fcis-30111	8	5	mostly	mostly	ADV
fcis-30111	8	6	relies	rely	VERB
fcis-30111	8	7	on	on	ADP
fcis-30111	8	8	single	single	ADJ
fcis-30111	8	9	-	-	PUNCT
fcis-30111	8	10	mode	mode	NOUN
fcis-30111	8	11	sensors	sensor	NOUN
fcis-30111	8	12	,	,	PUNCT
fcis-30111	8	13	which	which	PRON
fcis-30111	8	14	has	have	VERB
fcis-30111	8	15	problems	problem	NOUN
fcis-30111	8	16	such	such	ADJ
fcis-30111	8	17	as	as	ADP
fcis-30111	8	18	poor	poor	ADJ
fcis-30111	8	19	environmental	environmental	ADJ
fcis-30111	8	20	adaptability	adaptability	NOUN
fcis-30111	8	21	and	and	CCONJ
fcis-30111	8	22	high	high	ADJ
fcis-30111	8	23	leakage	leakage	NOUN
fcis-30111	8	24	rate	rate	NOUN
fcis-30111	8	25	[	[	X
fcis-30111	8	26	2	2	NUM
fcis-30111	8	27	]	]	PUNCT
fcis-30111	8	28	.	.	PUNCT
fcis-30111	9	1	in	in	ADP
fcis-30111	9	2	recent	recent	ADJ
fcis-30111	9	3	years	year	NOUN
fcis-30111	9	4	,	,	PUNCT
fcis-30111	9	5	the	the	DET
fcis-30111	9	6	rapid	rapid	ADJ
fcis-30111	9	7	development	development	NOUN
fcis-30111	9	8	of	of	ADP
fcis-30111	9	9	data	datum	NOUN
fcis-30111	9	10	fusion	fusion	NOUN
fcis-30111	9	11	technology	technology	NOUN
fcis-30111	9	12	in	in	ADP
fcis-30111	9	13	the	the	DET
fcis-30111	9	14	industrial	industrial	ADJ
fcis-30111	9	15	field	field	NOUN
fcis-30111	9	16	has	have	AUX
fcis-30111	9	17	provided	provide	VERB
fcis-30111	9	18	new	new	ADJ
fcis-30111	9	19	ideas	idea	NOUN
fcis-30111	9	20	for	for	ADP
fcis-30111	9	21	wheelset	wheelset	ADJ
fcis-30111	9	22	condition	condition	NOUN
fcis-30111	9	23	monitoring	monitoring	NOUN
fcis-30111	9	24	.	.	PUNCT
fcis-30111	10	1	scholars	scholar	NOUN
fcis-30111	10	2	at	at	ADP
fcis-30111	10	3	home	home	ADV
fcis-30111	10	4	and	and	CCONJ
fcis-30111	10	5	abroad	abroad	ADV
fcis-30111	10	6	have	have	AUX
fcis-30111	10	7	carried	carry	VERB
fcis-30111	10	8	out	out	ADP
fcis-30111	10	9	research	research	NOUN
fcis-30111	10	10	on	on	ADP
fcis-30111	10	11	acoustic	acoustic	ADJ
fcis-30111	10	12	characterisation	characterisation	NOUN
fcis-30111	10	13	and	and	CCONJ
fcis-30111	10	14	vibration	vibration	NOUN
fcis-30111	10	15	signal	signal	NOUN
fcis-30111	10	16	processing	processing	NOUN
fcis-30111	10	17	for	for	ADP
fcis-30111	10	18	different	different	ADJ
fcis-30111	10	19	failure	failure	NOUN
fcis-30111	10	20	modes	mode	NOUN
fcis-30111	10	21	of	of	ADP
fcis-30111	10	22	wheelsets	wheelset	NOUN
fcis-30111	10	23	,	,	PUNCT
fcis-30111	10	24	but	but	CCONJ
fcis-30111	10	25	the	the	DET
fcis-30111	10	26	comprehensive	comprehensive	ADJ
fcis-30111	10	27	utilisation	utilisation	NOUN
fcis-30111	10	28	of	of	ADP
fcis-30111	10	29	heterogeneous	heterogeneous	ADJ
fcis-30111	10	30	data	datum	NOUN
fcis-30111	10	31	from	from	ADP
fcis-30111	10	32	multiple	multiple	ADJ
fcis-30111	10	33	sources	source	NOUN
fcis-30111	10	34	still	still	ADV
fcis-30111	10	35	faces	face	VERB
fcis-30111	10	36	many	many	ADJ
fcis-30111	10	37	challenges	challenge	NOUN
fcis-30111	10	38	[	[	X
fcis-30111	10	39	3	3	NUM
fcis-30111	10	40	]	]	PUNCT
fcis-30111	10	41	.	.	PUNCT
fcis-30111	11	1	this	this	DET
fcis-30111	11	2	study	study	NOUN
fcis-30111	11	3	focuses	focus	VERB
fcis-30111	11	4	on	on	ADP
fcis-30111	11	5	the	the	DET
fcis-30111	11	6	fusion	fusion	NOUN
fcis-30111	11	7	of	of	ADP
fcis-30111	11	8	multimodal	multimodal	NOUN
fcis-30111	11	9	monitoring	monitoring	NOUN
fcis-30111	11	10	data	datum	NOUN
fcis-30111	11	11	of	of	ADP
fcis-30111	11	12	wheelsets	wheelset	NOUN
fcis-30111	11	13	,	,	PUNCT
fcis-30111	11	14	constructs	construct	VERB
fcis-30111	11	15	a	a	DET
fcis-30111	11	16	complete	complete	ADJ
fcis-30111	11	17	technical	technical	ADJ
fcis-30111	11	18	system	system	NOUN
fcis-30111	11	19	from	from	ADP
fcis-30111	11	20	data	datum	NOUN
fcis-30111	11	21	acquisition	acquisition	NOUN
fcis-30111	11	22	,	,	PUNCT
fcis-30111	11	23	feature	feature	NOUN
fcis-30111	11	24	extraction	extraction	NOUN
fcis-30111	11	25	to	to	ADP
fcis-30111	11	26	fusion	fusion	NOUN
fcis-30111	11	27	decision	decision	NOUN
fcis-30111	11	28	-	-	PUNCT
fcis-30111	11	29	making	making	NOUN
fcis-30111	11	30	,	,	PUNCT
fcis-30111	11	31	proposes	propose	VERB
fcis-30111	11	32	a	a	DET
fcis-30111	11	33	deep	deep	ADJ
fcis-30111	11	34	fusion	fusion	NOUN
fcis-30111	11	35	model	model	NOUN
fcis-30111	11	36	based	base	VERB
fcis-30111	11	37	on	on	ADP
fcis-30111	11	38	cross	cross	ADJ
fcis-30111	11	39	-	-	ADJ
fcis-30111	11	40	modal	modal	ADJ
fcis-30111	11	41	attention	attention	NOUN
fcis-30111	11	42	mechanism	mechanism	NOUN
fcis-30111	11	43	,	,	PUNCT
fcis-30111	11	44	and	and	CCONJ
fcis-30111	11	45	develops	develop	VERB
fcis-30111	11	46	a	a	DET
fcis-30111	11	47	realtime	realtime	NOUN
fcis-30111	11	48	monitoring	monitoring	NOUN
fcis-30111	11	49	and	and	CCONJ
fcis-30111	11	50	warning	warning	NOUN
fcis-30111	11	51	platform	platform	NOUN
fcis-30111	11	52	.	.	PUNCT
fcis-30111	12	1	the	the	DET
fcis-30111	12	2	adaptability	adaptability	NOUN
fcis-30111	12	3	and	and	CCONJ
fcis-30111	12	4	reliability	reliability	NOUN
fcis-30111	12	5	of	of	ADP
fcis-30111	12	6	this	this	DET
fcis-30111	12	7	technology	technology	NOUN
fcis-30111	12	8	in	in	ADP
fcis-30111	12	9	complex	complex	ADJ
fcis-30111	12	10	environments	environment	NOUN
fcis-30111	12	11	are	be	AUX
fcis-30111	12	12	verified	verify	VERB
fcis-30111	12	13	through	through	ADP
fcis-30111	12	14	engineering	engineering	NOUN
fcis-30111	12	15	practice	practice	NOUN
fcis-30111	12	16	,	,	PUNCT
fcis-30111	12	17	providing	provide	VERB
fcis-30111	12	18	a	a	DET
fcis-30111	12	19	systematic	systematic	ADJ
fcis-30111	12	20	solution	solution	NOUN
fcis-30111	12	21	for	for	ADP
fcis-30111	12	22	digital	digital	ADJ
fcis-30111	12	23	monitoring	monitoring	NOUN
fcis-30111	12	24	of	of	ADP
fcis-30111	12	25	wheelset	wheelset	ADJ
fcis-30111	12	26	health	health	NOUN
fcis-30111	12	27	status	status	NOUN
fcis-30111	12	28	in	in	ADP
fcis-30111	12	29	high	high	ADJ
fcis-30111	12	30	-	-	PUNCT
fcis-30111	12	31	speed	speed	NOUN
fcis-30111	12	32	railways	railway	NOUN
fcis-30111	12	33	.	.	PUNCT
fcis-30111	13	1	2	2	X
fcis-30111	13	2	.	.	X
fcis-30111	13	3	multimodal	multimodal	NOUN
fcis-30111	13	4	data	datum	NOUN
fcis-30111	13	5	acquisition	acquisition	NOUN
fcis-30111	13	6	for	for	ADP
fcis-30111	13	7	wheelset	wheelset	ADJ
fcis-30111	13	8	health	health	NOUN
fcis-30111	13	9	condition	condition	NOUN
fcis-30111	13	10	monitoring	monitor	VERB
fcis-30111	13	11	2.1	2.1	NUM
fcis-30111	13	12	.	.	PUNCT
fcis-30111	14	1	analysis	analysis	NOUN
fcis-30111	14	2	of	of	ADP
fcis-30111	14	3	wheelset	wheelset	ADJ
fcis-30111	14	4	structure	structure	NOUN
fcis-30111	14	5	and	and	CCONJ
fcis-30111	14	6	common	common	ADJ
fcis-30111	14	7	failure	failure	NOUN
fcis-30111	14	8	types	type	NOUN
fcis-30111	14	9	as	as	ADP
fcis-30111	14	10	a	a	DET
fcis-30111	14	11	key	key	ADJ
fcis-30111	14	12	component	component	NOUN
fcis-30111	14	13	of	of	ADP
fcis-30111	14	14	the	the	DET
fcis-30111	14	15	train	train	NOUN
fcis-30111	14	16	,	,	PUNCT
fcis-30111	14	17	the	the	DET
fcis-30111	14	18	wheelset	wheelset	NOUN
fcis-30111	14	19	is	be	AUX
fcis-30111	14	20	mainly	mainly	ADV
fcis-30111	14	21	composed	compose	VERB
fcis-30111	14	22	of	of	ADP
fcis-30111	14	23	wheels	wheel	NOUN
fcis-30111	14	24	,	,	PUNCT
fcis-30111	14	25	axles	axle	NOUN
fcis-30111	14	26	and	and	CCONJ
fcis-30111	14	27	bearings	bearing	NOUN
fcis-30111	14	28	,	,	PUNCT
fcis-30111	14	29	and	and	CCONJ
fcis-30111	14	30	its	its	PRON
fcis-30111	14	31	health	health	NOUN
fcis-30111	14	32	status	status	NOUN
fcis-30111	14	33	directly	directly	ADV
fcis-30111	14	34	affects	affect	VERB
fcis-30111	14	35	the	the	DET
fcis-30111	14	36	safety	safety	NOUN
fcis-30111	14	37	of	of	ADP
fcis-30111	14	38	train	train	NOUN
fcis-30111	14	39	operation	operation	NOUN
fcis-30111	14	40	.	.	PUNCT
fcis-30111	15	1	according	accord	VERB
fcis-30111	15	2	to	to	ADP
fcis-30111	15	3	the	the	DET
fcis-30111	15	4	statistics	statistic	NOUN
fcis-30111	15	5	of	of	ADP
fcis-30111	15	6	railway	railway	PROPN
fcis-30111	15	7	department	department	PROPN
fcis-30111	15	8	,	,	PUNCT
fcis-30111	15	9	the	the	DET
fcis-30111	15	10	common	common	ADJ
fcis-30111	15	11	failures	failure	NOUN
fcis-30111	15	12	of	of	ADP
fcis-30111	15	13	wheelsets	wheelset	NOUN
fcis-30111	15	14	mainly	mainly	ADV
fcis-30111	15	15	include	include	VERB
fcis-30111	15	16	wheel	wheel	NOUN
fcis-30111	15	17	abrasion	abrasion	NOUN
fcis-30111	15	18	,	,	PUNCT
fcis-30111	15	19	rim	rim	NOUN
fcis-30111	15	20	crack	crack	NOUN
fcis-30111	15	21	,	,	PUNCT
fcis-30111	15	22	axle	axle	NOUN
fcis-30111	15	23	wear	wear	VERB
fcis-30111	15	24	and	and	CCONJ
fcis-30111	15	25	bearing	bear	VERB
fcis-30111	15	26	damage	damage	NOUN
fcis-30111	15	27	,	,	PUNCT
fcis-30111	15	28	etc	etc	X
fcis-30111	15	29	.	.	X
fcis-30111	16	1	these	these	DET
fcis-30111	16	2	failures	failure	NOUN
fcis-30111	16	3	will	will	AUX
fcis-30111	16	4	lead	lead	VERB
fcis-30111	16	5	to	to	ADP
fcis-30111	16	6	serious	serious	ADJ
fcis-30111	16	7	consequences	consequence	NOUN
fcis-30111	16	8	if	if	SCONJ
fcis-30111	16	9	they	they	PRON
fcis-30111	16	10	are	be	AUX
fcis-30111	16	11	not	not	PART
fcis-30111	16	12	detected	detect	VERB
fcis-30111	16	13	in	in	ADP
fcis-30111	16	14	time	time	NOUN
fcis-30111	16	15	.	.	PUNCT
fcis-30111	17	1	the	the	DET
fcis-30111	17	2	annual	annual	ADJ
fcis-30111	17	3	data	datum	NOUN
fcis-30111	17	4	of	of	ADP
fcis-30111	17	5	many	many	ADJ
fcis-30111	17	6	domestic	domestic	ADJ
fcis-30111	17	7	high	high	ADJ
fcis-30111	17	8	-	-	PUNCT
fcis-30111	17	9	speed	speed	NOUN
fcis-30111	17	10	railway	railway	NOUN
fcis-30111	17	11	lines	line	NOUN
fcis-30111	17	12	show	show	VERB
fcis-30111	17	13	that	that	SCONJ
fcis-30111	17	14	wheel	wheel	NOUN
fcis-30111	17	15	tread	tread	NOUN
fcis-30111	17	16	abrasion	abrasion	NOUN
fcis-30111	17	17	accounts	account	VERB
fcis-30111	17	18	for	for	ADP
fcis-30111	17	19	about	about	ADV
fcis-30111	17	20	42	42	NUM
fcis-30111	17	21	%	%	NOUN
fcis-30111	17	22	of	of	ADP
fcis-30111	17	23	the	the	DET
fcis-30111	17	24	total	total	ADJ
fcis-30111	17	25	number	number	NOUN
fcis-30111	17	26	of	of	ADP
fcis-30111	17	27	wheelset	wheelset	ADJ
fcis-30111	17	28	failures	failure	NOUN
fcis-30111	17	29	,	,	PUNCT
fcis-30111	17	30	bearing	bear	VERB
fcis-30111	17	31	early	early	ADJ
fcis-30111	17	32	damage	damage	NOUN
fcis-30111	17	33	accounts	account	NOUN
fcis-30111	17	34	for	for	ADP
fcis-30111	17	35	about	about	ADV
fcis-30111	17	36	23	23	NUM
fcis-30111	17	37	%	%	NOUN
fcis-30111	17	38	,	,	PUNCT
fcis-30111	17	39	axle	axle	NOUN
fcis-30111	17	40	connecting	connect	VERB
fcis-30111	17	41	part	part	NOUN
fcis-30111	17	42	loosening	loosen	VERB
fcis-30111	17	43	accounts	account	NOUN
fcis-30111	17	44	for	for	ADP
fcis-30111	17	45	about	about	ADV
fcis-30111	17	46	18	18	NUM
fcis-30111	17	47	%	%	NOUN
fcis-30111	17	48	,	,	PUNCT
fcis-30111	17	49	and	and	CCONJ
fcis-30111	17	50	the	the	DET
fcis-30111	17	51	rest	rest	NOUN
fcis-30111	17	52	are	be	AUX
fcis-30111	17	53	compound	compound	NOUN
fcis-30111	17	54	failures[4	failures[4	PROPN
fcis-30111	17	55	]	]	PUNCT
fcis-30111	17	56	.	.	PUNCT
fcis-30111	18	1	based	base	VERB
fcis-30111	18	2	on	on	ADP
fcis-30111	18	3	the	the	DET
fcis-30111	18	4	analysis	analysis	NOUN
fcis-30111	18	5	of	of	ADP
fcis-30111	18	6	engineering	engineering	NOUN
fcis-30111	18	7	practice	practice	NOUN
fcis-30111	18	8	,	,	PUNCT
fcis-30111	18	9	most	most	ADJ
fcis-30111	18	10	of	of	ADP
fcis-30111	18	11	the	the	DET
fcis-30111	18	12	wheelset	wheelset	ADJ
fcis-30111	18	13	failures	failure	NOUN
fcis-30111	18	14	in	in	ADP
fcis-30111	18	15	the	the	DET
fcis-30111	18	16	early	early	ADJ
fcis-30111	18	17	stage	stage	NOUN
fcis-30111	18	18	have	have	VERB
fcis-30111	18	19	precursors	precursor	NOUN
fcis-30111	18	20	such	such	ADJ
fcis-30111	18	21	as	as	ADP
fcis-30111	18	22	abnormal	abnormal	ADJ
fcis-30111	18	23	sound	sound	NOUN
fcis-30111	18	24	,	,	PUNCT
fcis-30111	18	25	temperature	temperature	NOUN
fcis-30111	18	26	rise	rise	NOUN
fcis-30111	18	27	or	or	CCONJ
fcis-30111	18	28	change	change	NOUN
fcis-30111	18	29	of	of	ADP
fcis-30111	18	30	vibration	vibration	NOUN
fcis-30111	18	31	characteristics	characteristic	NOUN
fcis-30111	18	32	,	,	PUNCT
fcis-30111	18	33	etc	etc	X
fcis-30111	18	34	.	.	X
fcis-30111	19	1	the	the	DET
fcis-30111	19	2	reasonably	reasonably	ADV
fcis-30111	19	3	-	-	PUNCT
fcis-30111	19	4	designed	design	VERB
fcis-30111	19	5	monitoring	monitoring	NOUN
fcis-30111	19	6	system	system	NOUN
fcis-30111	19	7	can	can	AUX
fcis-30111	19	8	effectively	effectively	ADV
fcis-30111	19	9	capture	capture	VERB
fcis-30111	19	10	these	these	DET
fcis-30111	19	11	changes	change	NOUN
fcis-30111	19	12	in	in	ADP
fcis-30111	19	13	characteristics	characteristic	NOUN
fcis-30111	19	14	,	,	PUNCT
fcis-30111	19	15	achieve	achieve	VERB
fcis-30111	19	16	early	early	ADJ
fcis-30111	19	17	warning	warning	NOUN
fcis-30111	19	18	of	of	ADP
fcis-30111	19	19	failures	failure	NOUN
fcis-30111	19	20	,	,	PUNCT
fcis-30111	19	21	and	and	CCONJ
fcis-30111	19	22	fundamentally	fundamentally	ADV
fcis-30111	19	23	reduce	reduce	VERB
fcis-30111	19	24	the	the	DET
fcis-30111	19	25	risk	risk	NOUN
fcis-30111	19	26	rate	rate	NOUN
fcis-30111	19	27	of	of	ADP
fcis-30111	19	28	wheelset	wheelset	ADJ
fcis-30111	19	29	failures	failure	NOUN
fcis-30111	19	30	up	up	ADP
fcis-30111	19	31	to	to	PART
fcis-30111	19	32	more	more	ADJ
fcis-30111	19	33	than	than	ADP
fcis-30111	19	34	80	80	NUM
fcis-30111	19	35	%	%	NOUN
fcis-30111	19	36	.	.	PUNCT
fcis-30111	20	1	2.2	2.2	NUM
fcis-30111	20	2	.	.	PUNCT
fcis-30111	21	1	multi	multi	ADJ
fcis-30111	21	2	-	-	ADJ
fcis-30111	21	3	sensor	sensor	ADJ
fcis-30111	21	4	monitoring	monitoring	NOUN
fcis-30111	21	5	system	system	NOUN
fcis-30111	21	6	architecture	architecture	NOUN
fcis-30111	21	7	the	the	DET
fcis-30111	21	8	architecture	architecture	NOUN
fcis-30111	21	9	of	of	ADP
fcis-30111	21	10	the	the	DET
fcis-30111	21	11	wheelset	wheelset	PROPN
fcis-30111	21	12	health	health	NOUN
fcis-30111	21	13	condition	condition	NOUN
fcis-30111	21	14	monitoring	monitoring	NOUN
fcis-30111	21	15	system	system	NOUN
fcis-30111	21	16	adopts	adopt	VERB
fcis-30111	21	17	a	a	DET
fcis-30111	21	18	layered	layered	ADJ
fcis-30111	21	19	structure	structure	NOUN
fcis-30111	21	20	design	design	NOUN
fcis-30111	21	21	,	,	PUNCT
fcis-30111	21	22	including	include	VERB
fcis-30111	21	23	sensing	sense	VERB
fcis-30111	21	24	layer	layer	NOUN
fcis-30111	21	25	,	,	PUNCT
fcis-30111	21	26	transmission	transmission	NOUN
fcis-30111	21	27	layer	layer	NOUN
fcis-30111	21	28	,	,	PUNCT
fcis-30111	21	29	processing	processing	NOUN
fcis-30111	21	30	layer	layer	NOUN
fcis-30111	21	31	and	and	CCONJ
fcis-30111	21	32	application	application	NOUN
fcis-30111	21	33	layer	layer	NOUN
fcis-30111	21	34	,	,	PUNCT
fcis-30111	21	35	as	as	SCONJ
fcis-30111	21	36	shown	show	VERB
fcis-30111	21	37	in	in	ADP
fcis-30111	21	38	fig	fig	NOUN
fcis-30111	21	39	.	.	PUNCT
fcis-30111	22	1	1	1	X
fcis-30111	22	2	.	.	PUNCT
fcis-30111	22	3	multiple	multiple	ADJ
fcis-30111	22	4	sensors	sensor	NOUN
fcis-30111	22	5	are	be	AUX
fcis-30111	22	6	arranged	arrange	VERB
fcis-30111	22	7	in	in	ADP
fcis-30111	22	8	the	the	DET
fcis-30111	22	9	sensing	sense	VERB
fcis-30111	22	10	layer	layer	NOUN
fcis-30111	22	11	,	,	PUNCT
fcis-30111	22	12	including	include	VERB
fcis-30111	22	13	vibration	vibration	NOUN
fcis-30111	22	14	sensors	sensor	NOUN
fcis-30111	22	15	(	(	PUNCT
fcis-30111	22	16	with	with	ADP
fcis-30111	22	17	a	a	DET
fcis-30111	22	18	sensitivity	sensitivity	NOUN
fcis-30111	22	19	of	of	ADP
fcis-30111	22	20	100mv	100mv	NOUN
fcis-30111	22	21	/	/	SYM
fcis-30111	22	22	g	g	PROPN
fcis-30111	22	23	)	)	PUNCT
fcis-30111	22	24	,	,	PUNCT
fcis-30111	22	25	temperature	temperature	NOUN
fcis-30111	22	26	sensors	sensor	NOUN
fcis-30111	22	27	(	(	PUNCT
fcis-30111	22	28	with	with	ADP
fcis-30111	22	29	a	a	DET
fcis-30111	22	30	measuring	measure	VERB
fcis-30111	22	31	range	range	NOUN
fcis-30111	22	32	of	of	ADP
fcis-30111	22	33	-40	-40	PROPN
fcis-30111	22	34	℃	℃	PROPN
fcis-30111	22	35	to	to	ADP
fcis-30111	22	36	150	150	NUM
fcis-30111	22	37	℃	℃	PROPN
fcis-30111	22	38	)	)	PUNCT
fcis-30111	22	39	,	,	PUNCT
fcis-30111	22	40	acoustic	acoustic	ADJ
fcis-30111	22	41	sensors	sensor	NOUN
fcis-30111	22	42	(	(	PUNCT
fcis-30111	22	43	with	with	ADP
fcis-30111	22	44	a	a	DET
fcis-30111	22	45	frequency	frequency	NOUN
fcis-30111	22	46	response	response	NOUN
fcis-30111	22	47	of	of	ADP
fcis-30111	22	48	20hz-20khz	20hz-20khz	NUM
fcis-30111	22	49	)	)	PUNCT
fcis-30111	22	50	and	and	CCONJ
fcis-30111	22	51	optical	optical	ADJ
fcis-30111	22	52	sensors	sensor	NOUN
fcis-30111	22	53	,	,	PUNCT
fcis-30111	22	54	etc	etc	X
fcis-30111	22	55	.	.	X
fcis-30111	22	56	,	,	PUNCT
fcis-30111	22	57	to	to	PART
fcis-30111	22	58	form	form	VERB
fcis-30111	22	59	an	an	DET
fcis-30111	22	60	omni	omni	ADJ
fcis-30111	22	61	-	-	PUNCT
fcis-30111	22	62	directional	directional	ADJ
fcis-30111	22	63	monitoring	monitoring	NOUN
fcis-30111	22	64	network	network	NOUN
fcis-30111	22	65	.	.	PUNCT
fcis-30111	23	1	monitoring	monitor	VERB
fcis-30111	23	2	network	network	NOUN
fcis-30111	23	3	[	[	X
fcis-30111	23	4	5	5	NUM
fcis-30111	23	5	]	]	PUNCT
fcis-30111	23	6	.	.	PUNCT
fcis-30111	24	1	the	the	DET
fcis-30111	24	2	transmission	transmission	NOUN
fcis-30111	24	3	layer	layer	NOUN
fcis-30111	24	4	adopts	adopt	VERB
fcis-30111	24	5	a	a	DET
fcis-30111	24	6	combination	combination	NOUN
fcis-30111	24	7	of	of	ADP
fcis-30111	24	8	industrial	industrial	ADJ
fcis-30111	24	9	ethernet	ethernet	NOUN
fcis-30111	24	10	and	and	CCONJ
fcis-30111	24	11	wireless	wireless	ADJ
fcis-30111	24	12	transmission	transmission	NOUN
fcis-30111	24	13	to	to	PART
fcis-30111	24	14	ensure	ensure	VERB
fcis-30111	24	15	that	that	SCONJ
fcis-30111	24	16	the	the	DET
fcis-30111	24	17	data	data	NOUN
fcis-30111	24	18	transmission	transmission	NOUN
fcis-30111	24	19	rate	rate	NOUN
fcis-30111	24	20	reaches	reach	VERB
fcis-30111	24	21	more	more	ADJ
fcis-30111	24	22	than	than	ADP
fcis-30111	24	23	10mbps	10mbps	NUM
fcis-30111	24	24	to	to	PART
fcis-30111	24	25	meet	meet	VERB
fcis-30111	24	26	real	real	ADJ
fcis-30111	24	27	-	-	PUNCT
fcis-30111	24	28	time	time	NOUN
fcis-30111	24	29	monitoring	monitoring	NOUN
fcis-30111	24	30	requirements	requirement	NOUN
fcis-30111	24	31	.	.	PUNCT
fcis-30111	25	1	the	the	DET
fcis-30111	25	2	processing	processing	NOUN
fcis-30111	25	3	layer	layer	NOUN
fcis-30111	25	4	is	be	AUX
fcis-30111	25	5	equipped	equip	VERB
fcis-30111	25	6	with	with	ADP
fcis-30111	25	7	edge	edge	NOUN
fcis-30111	25	8	computing	computing	NOUN
fcis-30111	25	9	devices	device	NOUN
fcis-30111	25	10	,	,	PUNCT
fcis-30111	25	11	adopting	adopt	VERB
fcis-30111	25	12	dual	dual	ADJ
fcis-30111	25	13	-	-	PUNCT
fcis-30111	25	14	core	core	NOUN
fcis-30111	25	15	processors	processor	NOUN
fcis-30111	25	16	with	with	ADP
fcis-30111	25	17	a	a	DET
fcis-30111	25	18	computing	compute	VERB
fcis-30111	25	19	power	power	NOUN
fcis-30111	25	20	of	of	ADP
fcis-30111	25	21	4ghz	4ghz	PROPN
fcis-30111	25	22	to	to	PART
fcis-30111	25	23	perform	perform	VERB
fcis-30111	25	24	signal	signal	ADJ
fcis-30111	25	25	pre	pre	ADJ
fcis-30111	25	26	-	-	ADJ
fcis-30111	25	27	processing	processing	ADJ
fcis-30111	25	28	and	and	CCONJ
fcis-30111	25	29	feature	feature	NOUN
fcis-30111	25	30	extraction	extraction	NOUN
fcis-30111	25	31	.	.	PUNCT
fcis-30111	26	1	the	the	DET
fcis-30111	26	2	application	application	NOUN
fcis-30111	26	3	layer	layer	NOUN
fcis-30111	26	4	is	be	AUX
fcis-30111	26	5	deployed	deploy	VERB
fcis-30111	26	6	in	in	ADP
fcis-30111	26	7	the	the	DET
fcis-30111	26	8	cloud	cloud	NOUN
fcis-30111	26	9	server	server	NOUN
fcis-30111	26	10	,	,	PUNCT
fcis-30111	26	11	utilising	utilise	VERB
fcis-30111	26	12	64	64	NUM
fcis-30111	26	13	gb	gb	NOUN
fcis-30111	26	14	of	of	ADP
fcis-30111	26	15	memory	memory	NOUN
fcis-30111	26	16	and	and	CCONJ
fcis-30111	26	17	8	8	NUM
fcis-30111	26	18	tb	tb	NOUN
fcis-30111	26	19	of	of	ADP
fcis-30111	26	20	storage	storage	NOUN
fcis-30111	26	21	to	to	PART
fcis-30111	26	22	run	run	VERB
fcis-30111	26	23	fault	fault	NOUN
fcis-30111	26	24	diagnosis	diagnosis	NOUN
fcis-30111	26	25	algorithms	algorithm	NOUN
fcis-30111	26	26	and	and	CCONJ
fcis-30111	26	27	health	health	NOUN
fcis-30111	26	28	assessment	assessment	NOUN
fcis-30111	26	29	models	model	NOUN
fcis-30111	26	30	.	.	PUNCT
fcis-30111	27	1	actual	actual	ADJ
fcis-30111	27	2	engineering	engineering	NOUN
fcis-30111	27	3	applications	application	NOUN
fcis-30111	27	4	show	show	VERB
fcis-30111	27	5	that	that	SCONJ
fcis-30111	27	6	the	the	DET
fcis-30111	27	7	architecture	architecture	NOUN
fcis-30111	27	8	has	have	VERB
fcis-30111	27	9	a	a	DET
fcis-30111	27	10	system	system	NOUN
fcis-30111	27	11	response	response	NOUN
fcis-30111	27	12	time	time	NOUN
fcis-30111	27	13	of	of	ADP
fcis-30111	27	14	less	less	ADJ
fcis-30111	27	15	than	than	ADP
fcis-30111	27	16	200	200	NUM
fcis-30111	27	17	ms	ms	NOUN
fcis-30111	27	18	and	and	CCONJ
fcis-30111	27	19	a	a	DET
fcis-30111	27	20	fault	fault	NOUN
fcis-30111	27	21	identification	identification	NOUN
fcis-30111	27	22	accuracy	accuracy	NOUN
fcis-30111	27	23	of	of	ADP
fcis-30111	27	24	more	more	ADJ
fcis-30111	27	25	than	than	ADP
fcis-30111	27	26	95	95	NUM
fcis-30111	27	27	%	%	NOUN
fcis-30111	27	28	in	in	ADP
fcis-30111	27	29	the	the	DET
fcis-30111	27	30	high	high	ADJ
fcis-30111	27	31	-	-	PUNCT
fcis-30111	27	32	speed	speed	NOUN
fcis-30111	27	33	railway	railway	NOUN
fcis-30111	27	34	field	field	NOUN
fcis-30111	27	35	environment	environment	NOUN
fcis-30111	27	36	.	.	PUNCT
fcis-30111	28	1	70	70	NUM
fcis-30111	28	2	figure	figure	NOUN
fcis-30111	28	3	1	1	NUM
fcis-30111	28	4	.	.	PUNCT
fcis-30111	28	5	multi	multi	ADJ
fcis-30111	28	6	-	-	ADJ
fcis-30111	28	7	sensor	sensor	ADJ
fcis-30111	28	8	monitoring	monitoring	NOUN
fcis-30111	28	9	system	system	NOUN
fcis-30111	28	10	architecture	architecture	NOUN
fcis-30111	28	11	2.3	2.3	NUM
fcis-30111	28	12	multimodal	multimodal	NOUN
fcis-30111	28	13	data	datum	NOUN
fcis-30111	28	14	acquisition	acquisition	NOUN
fcis-30111	28	15	method	method	VERB
fcis-30111	28	16	the	the	DET
fcis-30111	28	17	multimodal	multimodal	NOUN
fcis-30111	28	18	data	datum	NOUN
fcis-30111	28	19	acquisition	acquisition	NOUN
fcis-30111	28	20	method	method	NOUN
fcis-30111	28	21	combines	combine	VERB
fcis-30111	28	22	the	the	DET
fcis-30111	28	23	characteristics	characteristic	NOUN
fcis-30111	28	24	of	of	ADP
fcis-30111	28	25	the	the	DET
fcis-30111	28	26	wheel	wheel	NOUN
fcis-30111	28	27	-	-	PUNCT
fcis-30111	28	28	rail	rail	NOUN
fcis-30111	28	29	system	system	NOUN
fcis-30111	28	30	and	and	CCONJ
fcis-30111	28	31	adopts	adopt	VERB
fcis-30111	28	32	a	a	DET
fcis-30111	28	33	combination	combination	NOUN
fcis-30111	28	34	of	of	ADP
fcis-30111	28	35	fixed	fix	VERB
fcis-30111	28	36	and	and	CCONJ
fcis-30111	28	37	mobile	mobile	ADJ
fcis-30111	28	38	modes	mode	NOUN
fcis-30111	28	39	.	.	PUNCT
fcis-30111	29	1	fixed	fix	VERB
fcis-30111	29	2	acquisition	acquisition	NOUN
fcis-30111	29	3	equipment	equipment	NOUN
fcis-30111	29	4	is	be	AUX
fcis-30111	29	5	installed	instal	VERB
fcis-30111	29	6	in	in	ADP
fcis-30111	29	7	stations	station	NOUN
fcis-30111	29	8	or	or	CCONJ
fcis-30111	29	9	specific	specific	ADJ
fcis-30111	29	10	line	line	NOUN
fcis-30111	29	11	sections	section	NOUN
fcis-30111	29	12	,	,	PUNCT
fcis-30111	29	13	using	use	VERB
fcis-30111	29	14	high	high	ADJ
fcis-30111	29	15	-	-	PUNCT
fcis-30111	29	16	speed	speed	NOUN
fcis-30111	29	17	cameras	camera	NOUN
fcis-30111	29	18	(	(	PUNCT
fcis-30111	29	19	acquisition	acquisition	NOUN
fcis-30111	29	20	rate	rate	NOUN
fcis-30111	29	21	of	of	ADP
fcis-30111	29	22	500	500	NUM
fcis-30111	29	23	frames	frame	NOUN
fcis-30111	29	24	per	per	ADP
fcis-30111	29	25	second	second	NOUN
fcis-30111	29	26	)	)	PUNCT
fcis-30111	29	27	to	to	PART
fcis-30111	29	28	capture	capture	VERB
fcis-30111	29	29	images	image	NOUN
fcis-30111	29	30	of	of	ADP
fcis-30111	29	31	the	the	DET
fcis-30111	29	32	wheelset	wheelset	PROPN
fcis-30111	29	33	surface	surface	NOUN
fcis-30111	29	34	,	,	PUNCT
fcis-30111	29	35	together	together	ADV
fcis-30111	29	36	with	with	ADP
fcis-30111	29	37	infrared	infrared	ADJ
fcis-30111	29	38	thermal	thermal	ADJ
fcis-30111	29	39	cameras	camera	NOUN
fcis-30111	29	40	(	(	PUNCT
fcis-30111	29	41	temperature	temperature	NOUN
fcis-30111	29	42	resolution	resolution	NOUN
fcis-30111	29	43	of	of	ADP
fcis-30111	29	44	0.05	0.05	NUM
fcis-30111	29	45	℃	℃	PROPN
fcis-30111	29	46	)	)	PUNCT
fcis-30111	29	47	to	to	PART
fcis-30111	29	48	record	record	VERB
fcis-30111	29	49	the	the	DET
fcis-30111	29	50	temperature	temperature	NOUN
fcis-30111	29	51	distribution	distribution	NOUN
fcis-30111	29	52	.	.	PUNCT
fcis-30111	30	1	the	the	DET
fcis-30111	30	2	mobile	mobile	ADJ
fcis-30111	30	3	system	system	NOUN
fcis-30111	30	4	is	be	AUX
fcis-30111	30	5	mounted	mount	VERB
fcis-30111	30	6	on	on	ADP
fcis-30111	30	7	the	the	DET
fcis-30111	30	8	vehicle	vehicle	NOUN
fcis-30111	30	9	and	and	CCONJ
fcis-30111	30	10	consists	consist	VERB
fcis-30111	30	11	of	of	ADP
fcis-30111	30	12	a	a	DET
fcis-30111	30	13	triaxial	triaxial	ADJ
fcis-30111	30	14	acceleration	acceleration	NOUN
fcis-30111	30	15	sensor	sensor	NOUN
fcis-30111	30	16	(	(	PUNCT
fcis-30111	30	17	sampling	sample	VERB
fcis-30111	30	18	frequency	frequency	NOUN
fcis-30111	30	19	10khz	10khz	NOUN
fcis-30111	30	20	)	)	PUNCT
fcis-30111	30	21	to	to	PART
fcis-30111	30	22	continuously	continuously	ADV
fcis-30111	30	23	capture	capture	VERB
fcis-30111	30	24	vibration	vibration	NOUN
fcis-30111	30	25	signals	signal	NOUN
fcis-30111	30	26	,	,	PUNCT
fcis-30111	30	27	and	and	CCONJ
fcis-30111	30	28	an	an	DET
fcis-30111	30	29	array	array	NOUN
fcis-30111	30	30	of	of	ADP
fcis-30111	30	31	acoustic	acoustic	ADJ
fcis-30111	30	32	sensors	sensor	NOUN
fcis-30111	30	33	(	(	PUNCT
fcis-30111	30	34	consisting	consist	VERB
fcis-30111	30	35	of	of	ADP
fcis-30111	30	36	eight	eight	NUM
fcis-30111	30	37	microphones	microphone	NOUN
fcis-30111	30	38	)	)	PUNCT
fcis-30111	30	39	to	to	PART
fcis-30111	30	40	capture	capture	VERB
fcis-30111	30	41	operational	operational	ADJ
fcis-30111	30	42	noise	noise	NOUN
fcis-30111	30	43	characteristics	characteristic	NOUN
fcis-30111	30	44	.	.	PUNCT
fcis-30111	31	1	combined	combine	VERB
fcis-30111	31	2	with	with	ADP
fcis-30111	31	3	speed	speed	NOUN
fcis-30111	31	4	and	and	CCONJ
fcis-30111	31	5	load	load	NOUN
fcis-30111	31	6	data	datum	NOUN
fcis-30111	31	7	from	from	ADP
fcis-30111	31	8	the	the	DET
fcis-30111	31	9	train	train	NOUN
fcis-30111	31	10	control	control	NOUN
fcis-30111	31	11	system	system	NOUN
fcis-30111	31	12	are	be	AUX
fcis-30111	31	13	collected	collect	VERB
fcis-30111	31	14	together	together	ADV
fcis-30111	31	15	to	to	PART
fcis-30111	31	16	form	form	VERB
fcis-30111	31	17	a	a	DET
fcis-30111	31	18	complete	complete	ADJ
fcis-30111	31	19	data	datum	NOUN
fcis-30111	31	20	set	set	VERB
fcis-30111	31	21	.	.	PUNCT
fcis-30111	32	1	the	the	DET
fcis-30111	32	2	collected	collect	VERB
fcis-30111	32	3	data	data	NOUN
fcis-30111	32	4	is	be	AUX
fcis-30111	32	5	correlated	correlate	VERB
fcis-30111	32	6	with	with	ADP
fcis-30111	32	7	different	different	ADJ
fcis-30111	32	8	modal	modal	ADJ
fcis-30111	32	9	information	information	NOUN
fcis-30111	32	10	through	through	ADP
fcis-30111	32	11	time	time	NOUN
fcis-30111	32	12	synchronisation	synchronisation	NOUN
fcis-30111	32	13	technology	technology	NOUN
fcis-30111	32	14	(	(	PUNCT
fcis-30111	32	15	with	with	ADP
fcis-30111	32	16	microsecond	microsecond	ADJ
fcis-30111	32	17	accuracy	accuracy	NOUN
fcis-30111	32	18	)	)	PUNCT
fcis-30111	32	19	,	,	PUNCT
fcis-30111	32	20	and	and	CCONJ
fcis-30111	32	21	a	a	DET
fcis-30111	32	22	data	data	NOUN
fcis-30111	32	23	compression	compression	NOUN
fcis-30111	32	24	algorithm	algorithm	NOUN
fcis-30111	32	25	is	be	AUX
fcis-30111	32	26	used	use	VERB
fcis-30111	32	27	to	to	PART
fcis-30111	32	28	compress	compress	VERB
fcis-30111	32	29	the	the	DET
fcis-30111	32	30	original	original	ADJ
fcis-30111	32	31	data	datum	NOUN
fcis-30111	32	32	to	to	ADP
fcis-30111	32	33	35	35	NUM
fcis-30111	32	34	%	%	NOUN
fcis-30111	32	35	of	of	ADP
fcis-30111	32	36	its	its	PRON
fcis-30111	32	37	original	original	ADJ
fcis-30111	32	38	size	size	NOUN
fcis-30111	32	39	.	.	PUNCT
fcis-30111	33	1	the	the	DET
fcis-30111	33	2	field	field	NOUN
fcis-30111	33	3	test	test	NOUN
fcis-30111	33	4	proves	prove	VERB
fcis-30111	33	5	that	that	SCONJ
fcis-30111	33	6	the	the	DET
fcis-30111	33	7	method	method	NOUN
fcis-30111	33	8	can	can	AUX
fcis-30111	33	9	still	still	ADV
fcis-30111	33	10	collect	collect	VERB
fcis-30111	33	11	data	datum	NOUN
fcis-30111	33	12	stably	stably	ADV
fcis-30111	33	13	on	on	ADP
fcis-30111	33	14	a	a	DET
fcis-30111	33	15	high	high	ADJ
fcis-30111	33	16	-	-	PUNCT
fcis-30111	33	17	speed	speed	NOUN
fcis-30111	33	18	train	train	NOUN
fcis-30111	33	19	with	with	ADP
fcis-30111	33	20	a	a	DET
fcis-30111	33	21	speed	speed	NOUN
fcis-30111	33	22	of	of	ADP
fcis-30111	33	23	350	350	NUM
fcis-30111	33	24	kilometres	kilometre	NOUN
fcis-30111	33	25	per	per	ADP
fcis-30111	33	26	hour	hour	NOUN
fcis-30111	33	27	,	,	PUNCT
fcis-30111	33	28	and	and	CCONJ
fcis-30111	33	29	the	the	DET
fcis-30111	33	30	loss	loss	NOUN
fcis-30111	33	31	of	of	ADP
fcis-30111	33	32	signal	signal	ADJ
fcis-30111	33	33	quality	quality	NOUN
fcis-30111	33	34	is	be	AUX
fcis-30111	33	35	no	no	DET
fcis-30111	33	36	more	more	ADJ
fcis-30111	33	37	than	than	ADP
fcis-30111	33	38	3	3	NUM
fcis-30111	33	39	%	%	NOUN
fcis-30111	33	40	,	,	PUNCT
fcis-30111	33	41	providing	provide	VERB
fcis-30111	33	42	comprehensive	comprehensive	ADJ
fcis-30111	33	43	and	and	CCONJ
fcis-30111	33	44	reliable	reliable	ADJ
fcis-30111	33	45	data	data	NOUN
fcis-30111	33	46	support	support	NOUN
fcis-30111	33	47	for	for	ADP
fcis-30111	33	48	fault	fault	NOUN
fcis-30111	33	49	diagnosis	diagnosis	NOUN
fcis-30111	33	50	.	.	PUNCT
fcis-30111	34	1	3	3	X
fcis-30111	34	2	.	.	X
fcis-30111	34	3	multimodal	multimodal	ADJ
fcis-30111	34	4	data	datum	NOUN
fcis-30111	34	5	preprocessing	preprocessing	NOUN
fcis-30111	34	6	and	and	CCONJ
fcis-30111	34	7	feature	feature	NOUN
fcis-30111	34	8	extraction	extraction	NOUN
fcis-30111	34	9	for	for	ADP
fcis-30111	34	10	wheel	wheel	NOUN
fcis-30111	34	11	pair	pair	NOUN
fcis-30111	34	12	monitoring	monitor	VERB
fcis-30111	34	13	3.1	3.1	NUM
fcis-30111	34	14	.	.	PUNCT
fcis-30111	35	1	data	datum	NOUN
fcis-30111	35	2	preprocessing	preprocessing	NOUN
fcis-30111	35	3	methods	method	NOUN
fcis-30111	35	4	the	the	DET
fcis-30111	35	5	wheelset	wheelset	PROPN
fcis-30111	35	6	monitoring	monitor	VERB
fcis-30111	35	7	data	datum	NOUN
fcis-30111	35	8	preprocessing	preprocessing	NOUN
fcis-30111	35	9	process	process	NOUN
fcis-30111	35	10	is	be	AUX
fcis-30111	35	11	crucial	crucial	ADJ
fcis-30111	35	12	for	for	ADP
fcis-30111	35	13	subsequent	subsequent	ADJ
fcis-30111	35	14	fault	fault	NOUN
fcis-30111	35	15	diagnosis	diagnosis	NOUN
fcis-30111	35	16	,	,	PUNCT
fcis-30111	35	17	and	and	CCONJ
fcis-30111	35	18	its	its	PRON
fcis-30111	35	19	core	core	NOUN
fcis-30111	35	20	steps	step	NOUN
fcis-30111	35	21	include	include	VERB
fcis-30111	35	22	data	datum	NOUN
fcis-30111	35	23	cleaning	cleaning	NOUN
fcis-30111	35	24	,	,	PUNCT
fcis-30111	35	25	denoising	denoising	NOUN
fcis-30111	35	26	and	and	CCONJ
fcis-30111	35	27	standardisation	standardisation	NOUN
fcis-30111	35	28	.	.	PUNCT
fcis-30111	36	1	in	in	ADP
fcis-30111	36	2	actual	actual	ADJ
fcis-30111	36	3	engineering	engineering	NOUN
fcis-30111	36	4	,	,	PUNCT
fcis-30111	36	5	the	the	DET
fcis-30111	36	6	collected	collect	VERB
fcis-30111	36	7	vibration	vibration	NOUN
fcis-30111	36	8	signals	signal	NOUN
fcis-30111	36	9	are	be	AUX
fcis-30111	36	10	often	often	ADV
fcis-30111	36	11	interfered	interfere	VERB
fcis-30111	36	12	by	by	ADP
fcis-30111	36	13	environmental	environmental	ADJ
fcis-30111	36	14	noise	noise	NOUN
fcis-30111	36	15	,	,	PUNCT
fcis-30111	36	16	and	and	CCONJ
fcis-30111	36	17	the	the	DET
fcis-30111	36	18	signal	signal	NOUN
fcis-30111	36	19	-	-	PUNCT
fcis-30111	36	20	to	to	ADP
fcis-30111	36	21	-	-	PUNCT
fcis-30111	36	22	noise	noise	NOUN
fcis-30111	36	23	ratio	ratio	NOUN
fcis-30111	36	24	can	can	AUX
fcis-30111	36	25	be	be	AUX
fcis-30111	36	26	improved	improve	VERB
fcis-30111	36	27	by	by	ADP
fcis-30111	36	28	6.8	6.8	NUM
fcis-30111	36	29	db	db	NOUN
fcis-30111	36	30	after	after	ADP
fcis-30111	36	31	processing	process	VERB
fcis-30111	36	32	with	with	ADP
fcis-30111	36	33	the	the	DET
fcis-30111	36	34	wavelet	wavelet	NOUN
fcis-30111	36	35	transform	transform	NOUN
fcis-30111	36	36	denoising	denoise	VERB
fcis-30111	36	37	method[6].the	method[6].the	DET
fcis-30111	36	38	discrete	discrete	ADJ
fcis-30111	36	39	wavelet	wavelet	NOUN
fcis-30111	36	40	transform	transform	NOUN
fcis-30111	36	41	decomposes	decompose	VERB
fcis-30111	36	42	the	the	DET
fcis-30111	36	43	signal	signal	NOUN
fcis-30111	36	44	by	by	ADP
fcis-30111	36	45	the	the	DET
fcis-30111	36	46	formula	formula	NOUN
fcis-30111	36	47	:	:	PUNCT
fcis-30111	36	48	dwt	dwt	PROPN
fcis-30111	36	49	j	j	PROPN
fcis-30111	36	50	,	,	PUNCT
fcis-30111	36	51	k	k	PROPN
fcis-30111	36	52	x	x	X
fcis-30111	36	53	t	t	PROPN
fcis-30111	36	54	ψ	ψ	X
fcis-30111	36	55	dtz	dtz	X
fcis-30111	36	56	(	(	PUNCT
fcis-30111	36	57	1	1	NUM
fcis-30111	36	58	)	)	PUNCT
fcis-30111	36	59	where	where	SCONJ
fcis-30111	36	60	,	,	PUNCT
fcis-30111	36	61	dwt(j	dwt(j	PROPN
fcis-30111	36	62	,	,	PUNCT
fcis-30111	36	63	k	k	NOUN
fcis-30111	36	64	)	)	PUNCT
fcis-30111	36	65	is	be	AUX
fcis-30111	36	66	the	the	DET
fcis-30111	36	67	wavelet	wavelet	NOUN
fcis-30111	36	68	coefficient	coefficient	NOUN
fcis-30111	36	69	at	at	ADP
fcis-30111	36	70	scale	scale	PROPN
fcis-30111	36	71	j	j	PROPN
fcis-30111	36	72	and	and	CCONJ
fcis-30111	36	73	position	position	NOUN
fcis-30111	36	74	k.	k.	PROPN
fcis-30111	37	1	the	the	DET
fcis-30111	37	2	db4	db4	ADJ
fcis-30111	37	3	wavelet	wavelet	NOUN
fcis-30111	37	4	is	be	AUX
fcis-30111	37	5	selected	select	VERB
fcis-30111	37	6	for	for	ADP
fcis-30111	37	7	5	5	NUM
fcis-30111	37	8	-	-	PUNCT
fcis-30111	37	9	layer	layer	NOUN
fcis-30111	37	10	decomposition	decomposition	NOUN
fcis-30111	37	11	,	,	PUNCT
fcis-30111	37	12	and	and	CCONJ
fcis-30111	37	13	the	the	DET
fcis-30111	37	14	noise	noise	NOUN
fcis-30111	37	15	reduction	reduction	NOUN
fcis-30111	37	16	signal	signal	NOUN
fcis-30111	37	17	can	can	AUX
fcis-30111	37	18	be	be	AUX
fcis-30111	37	19	obtained	obtain	VERB
fcis-30111	37	20	by	by	ADP
fcis-30111	37	21	reconstruction	reconstruction	NOUN
fcis-30111	37	22	after	after	ADP
fcis-30111	37	23	thresholding	thresholde	VERB
fcis-30111	37	24	.	.	PUNCT
fcis-30111	38	1	for	for	ADP
fcis-30111	38	2	the	the	DET
fcis-30111	38	3	temperature	temperature	NOUN
fcis-30111	38	4	data	datum	NOUN
fcis-30111	38	5	,	,	PUNCT
fcis-30111	38	6	a	a	DET
fcis-30111	38	7	sliding	slide	VERB
fcis-30111	38	8	median	median	ADJ
fcis-30111	38	9	filter	filter	NOUN
fcis-30111	38	10	(	(	PUNCT
fcis-30111	38	11	window	window	NOUN
fcis-30111	38	12	width	width	NOUN
fcis-30111	38	13	of	of	ADP
fcis-30111	38	14	15	15	NUM
fcis-30111	38	15	samples	sample	NOUN
fcis-30111	38	16	)	)	PUNCT
fcis-30111	38	17	is	be	AUX
fcis-30111	38	18	used	use	VERB
fcis-30111	38	19	to	to	PART
fcis-30111	38	20	eliminate	eliminate	VERB
fcis-30111	38	21	the	the	DET
fcis-30111	38	22	mutation	mutation	NOUN
fcis-30111	38	23	points	point	NOUN
fcis-30111	38	24	,	,	PUNCT
fcis-30111	38	25	and	and	CCONJ
fcis-30111	38	26	the	the	DET
fcis-30111	38	27	smoothness	smoothness	NOUN
fcis-30111	38	28	of	of	ADP
fcis-30111	38	29	the	the	DET
fcis-30111	38	30	temperature	temperature	NOUN
fcis-30111	38	31	curve	curve	NOUN
fcis-30111	38	32	is	be	AUX
fcis-30111	38	33	improved	improve	VERB
fcis-30111	38	34	by	by	ADP
fcis-30111	38	35	about	about	ADV
fcis-30111	38	36	83	83	NUM
fcis-30111	38	37	%	%	NOUN
fcis-30111	38	38	after	after	ADP
fcis-30111	38	39	filtering	filter	VERB
fcis-30111	38	40	.	.	PUNCT
fcis-30111	39	1	for	for	ADP
fcis-30111	39	2	image	image	NOUN
fcis-30111	39	3	data	datum	NOUN
fcis-30111	39	4	,	,	PUNCT
fcis-30111	39	5	the	the	DET
fcis-30111	39	6	contrast	contrast	NOUN
fcis-30111	39	7	is	be	AUX
fcis-30111	39	8	enhanced	enhance	VERB
fcis-30111	39	9	by	by	ADP
fcis-30111	39	10	histogram	histogram	NOUN
fcis-30111	39	11	equalisation	equalisation	NOUN
fcis-30111	39	12	and	and	CCONJ
fcis-30111	39	13	a	a	DET
fcis-30111	39	14	gaussian	gaussian	ADJ
fcis-30111	39	15	filter	filter	NOUN
fcis-30111	39	16	(	(	PUNCT
fcis-30111	39	17	σ	σ	NOUN
fcis-30111	39	18	=	=	SYM
fcis-30111	39	19	1.2	1.2	NUM
fcis-30111	39	20	)	)	PUNCT
fcis-30111	39	21	is	be	AUX
fcis-30111	39	22	applied	apply	VERB
fcis-30111	39	23	to	to	PART
fcis-30111	39	24	eliminate	eliminate	VERB
fcis-30111	39	25	noise	noise	NOUN
fcis-30111	39	26	.	.	PUNCT
fcis-30111	40	1	the	the	DET
fcis-30111	40	2	multimodal	multimodal	NOUN
fcis-30111	40	3	data	datum	NOUN
fcis-30111	40	4	are	be	AUX
fcis-30111	40	5	also	also	ADV
fcis-30111	40	6	time	time	NOUN
fcis-30111	40	7	-	-	PUNCT
fcis-30111	40	8	aligned	align	VERB
fcis-30111	40	9	and	and	CCONJ
fcis-30111	40	10	sampling	sample	VERB
fcis-30111	40	11	rate	rate	NOUN
fcis-30111	40	12	-	-	PUNCT
fcis-30111	40	13	unified	unify	VERB
fcis-30111	40	14	,	,	PUNCT
fcis-30111	40	15	and	and	CCONJ
fcis-30111	40	16	the	the	DET
fcis-30111	40	17	data	datum	NOUN
fcis-30111	40	18	from	from	ADP
fcis-30111	40	19	different	different	ADJ
fcis-30111	40	20	sensors	sensor	NOUN
fcis-30111	40	21	are	be	AUX
fcis-30111	40	22	resampled	resample	VERB
fcis-30111	40	23	to	to	ADP
fcis-30111	40	24	a	a	DET
fcis-30111	40	25	uniform	uniform	ADJ
fcis-30111	40	26	frequency	frequency	NOUN
fcis-30111	40	27	of	of	ADP
fcis-30111	40	28	10	10	NUM
fcis-30111	40	29	khz	khz	NOUN
fcis-30111	40	30	by	by	ADP
fcis-30111	40	31	linear	linear	ADJ
fcis-30111	40	32	interpolation	interpolation	NOUN
fcis-30111	40	33	to	to	PART
fcis-30111	40	34	ensure	ensure	VERB
fcis-30111	40	35	that	that	SCONJ
fcis-30111	40	36	the	the	DET
fcis-30111	40	37	data	data	NOUN
fcis-30111	40	38	timestamp	timestamp	NOUN
fcis-30111	40	39	error	error	NOUN
fcis-30111	40	40	does	do	AUX
fcis-30111	40	41	not	not	PART
fcis-30111	40	42	exceed	exceed	VERB
fcis-30111	40	43	1	1	NUM
fcis-30111	40	44	ms	ms	PROPN
fcis-30111	40	45	.	.	PUNCT
fcis-30111	41	1	the	the	DET
fcis-30111	41	2	standardisation	standardisation	NOUN
fcis-30111	41	3	process	process	NOUN
fcis-30111	41	4	adopts	adopt	VERB
fcis-30111	41	5	the	the	DET
fcis-30111	41	6	z	z	NOUN
fcis-30111	41	7	-	-	PUNCT
fcis-30111	41	8	score	score	NOUN
fcis-30111	41	9	method	method	NOUN
fcis-30111	41	10	to	to	PART
fcis-30111	41	11	make	make	VERB
fcis-30111	41	12	all	all	DET
fcis-30111	41	13	the	the	DET
fcis-30111	41	14	feature	feature	NOUN
fcis-30111	41	15	magnitudes	magnitude	NOUN
fcis-30111	41	16	consistent	consistent	ADJ
fcis-30111	41	17	,	,	PUNCT
fcis-30111	41	18	which	which	PRON
fcis-30111	41	19	lays	lay	VERB
fcis-30111	41	20	the	the	DET
fcis-30111	41	21	foundation	foundation	NOUN
fcis-30111	41	22	for	for	ADP
fcis-30111	41	23	subsequent	subsequent	ADJ
fcis-30111	41	24	feature	feature	NOUN
fcis-30111	41	25	extraction	extraction	NOUN
fcis-30111	41	26	.	.	PUNCT
fcis-30111	42	1	3.2	3.2	NUM
fcis-30111	42	2	.	.	PUNCT
fcis-30111	43	1	traditional	traditional	ADJ
fcis-30111	43	2	feature	feature	NOUN
fcis-30111	43	3	extraction	extraction	NOUN
fcis-30111	43	4	techniques	technique	NOUN
fcis-30111	43	5	traditional	traditional	ADJ
fcis-30111	43	6	feature	feature	NOUN
fcis-30111	43	7	extraction	extraction	NOUN
fcis-30111	43	8	techniques	technique	NOUN
fcis-30111	43	9	are	be	AUX
fcis-30111	43	10	widely	widely	ADV
fcis-30111	43	11	used	use	VERB
fcis-30111	43	12	in	in	ADP
fcis-30111	43	13	wheelset	wheelset	ADJ
fcis-30111	43	14	health	health	NOUN
fcis-30111	43	15	monitoring	monitoring	NOUN
fcis-30111	43	16	,	,	PUNCT
fcis-30111	43	17	mainly	mainly	ADV
fcis-30111	43	18	focusing	focus	VERB
fcis-30111	43	19	on	on	ADP
fcis-30111	43	20	the	the	DET
fcis-30111	43	21	time	time	NOUN
fcis-30111	43	22	domain	domain	NOUN
fcis-30111	43	23	,	,	PUNCT
fcis-30111	43	24	frequency	frequency	NOUN
fcis-30111	43	25	domain	domain	NOUN
fcis-30111	43	26	and	and	CCONJ
fcis-30111	43	27	time	time	NOUN
fcis-30111	43	28	-	-	PUNCT
fcis-30111	43	29	frequency	frequency	NOUN
fcis-30111	43	30	domain	domain	NOUN
fcis-30111	43	31	feature	feature	NOUN
fcis-30111	43	32	extraction	extraction	NOUN
fcis-30111	43	33	of	of	ADP
fcis-30111	43	34	the	the	DET
fcis-30111	43	35	signal	signal	NOUN
fcis-30111	43	36	.	.	PUNCT
fcis-30111	44	1	in	in	ADP
fcis-30111	44	2	time	time	NOUN
fcis-30111	44	3	-	-	PUNCT
fcis-30111	44	4	domain	domain	NOUN
fcis-30111	44	5	analysis	analysis	NOUN
fcis-30111	44	6	,	,	PUNCT
fcis-30111	44	7	parameters	parameter	NOUN
fcis-30111	44	8	such	such	ADJ
fcis-30111	44	9	as	as	ADP
fcis-30111	44	10	the	the	DET
fcis-30111	44	11	root	root	NOUN
fcis-30111	44	12	mean	mean	NOUN
fcis-30111	44	13	square	square	PROPN
fcis-30111	44	14	(	(	PUNCT
fcis-30111	44	15	rms	rm	NOUN
fcis-30111	44	16	)	)	PUNCT
fcis-30111	44	17	value	value	NOUN
fcis-30111	44	18	and	and	CCONJ
fcis-30111	44	19	the	the	DET
fcis-30111	44	20	peak	peak	NOUN
fcis-30111	44	21	factor	factor	NOUN
fcis-30111	44	22	(	(	PUNCT
fcis-30111	44	23	generally	generally	ADV
fcis-30111	44	24	2.8	2.8	NUM
fcis-30111	44	25	-	-	SYM
fcis-30111	44	26	3.5	3.5	NUM
fcis-30111	44	27	under	under	ADP
fcis-30111	44	28	normal	normal	ADJ
fcis-30111	44	29	working	working	NOUN
fcis-30111	44	30	conditions	condition	NOUN
fcis-30111	44	31	,	,	PUNCT
fcis-30111	44	32	and	and	CCONJ
fcis-30111	44	33	6.5	6.5	NUM
fcis-30111	44	34	-	-	SYM
fcis-30111	44	35	8.0	8.0	NUM
fcis-30111	44	36	under	under	ADP
fcis-30111	44	37	abnormal	abnormal	ADJ
fcis-30111	44	38	conditions	condition	NOUN
fcis-30111	44	39	)	)	PUNCT
fcis-30111	44	40	are	be	AUX
fcis-30111	44	41	extracted	extract	VERB
fcis-30111	44	42	from	from	ADP
fcis-30111	44	43	the	the	DET
fcis-30111	44	44	vibration	vibration	NOUN
fcis-30111	44	45	signals	signal	NOUN
fcis-30111	44	46	to	to	PART
fcis-30111	44	47	effectively	effectively	ADV
fcis-30111	44	48	differentiate	differentiate	VERB
fcis-30111	44	49	between	between	ADP
fcis-30111	44	50	fault	fault	NOUN
fcis-30111	44	51	types	type	NOUN
fcis-30111	44	52	[	[	X
fcis-30111	44	53	7	7	NUM
fcis-30111	44	54	]	]	PUNCT
fcis-30111	44	55	.	.	PUNCT
fcis-30111	45	1	the	the	DET
fcis-30111	45	2	frequency	frequency	NOUN
fcis-30111	45	3	domain	domain	NOUN
fcis-30111	45	4	analysis	analysis	NOUN
fcis-30111	45	5	uses	use	VERB
fcis-30111	45	6	fast	fast	ADJ
fcis-30111	45	7	fourier	fourier	NOUN
fcis-30111	45	8	transform	transform	NOUN
fcis-30111	45	9	to	to	PART
fcis-30111	45	10	extract	extract	VERB
fcis-30111	45	11	spectral	spectral	ADJ
fcis-30111	45	12	features	feature	NOUN
fcis-30111	45	13	:	:	PUNCT
fcis-30111	45	14	x	x	X
fcis-30111	45	15	k	k	X
fcis-30111	45	16	∑	∑	PUNCT
fcis-30111	45	17	x	x	PROPN
fcis-30111	45	18	n	n	X
fcis-30111	45	19	e	e	X
fcis-30111	45	20	(	(	PUNCT
fcis-30111	45	21	2	2	NUM
fcis-30111	45	22	)	)	PUNCT
fcis-30111	45	23	where	where	SCONJ
fcis-30111	45	24	x(k	x(k	NOUN
fcis-30111	45	25	)	)	PUNCT
fcis-30111	45	26	is	be	AUX
fcis-30111	45	27	the	the	DET
fcis-30111	45	28	kth	kth	PROPN
fcis-30111	45	29	frequency	frequency	NOUN
fcis-30111	45	30	component	component	NOUN
fcis-30111	45	31	in	in	ADP
fcis-30111	45	32	the	the	DET
fcis-30111	45	33	frequency	frequency	NOUN
fcis-30111	45	34	domain	domain	NOUN
fcis-30111	45	35	and	and	CCONJ
fcis-30111	45	36	x(n	x(n	NOUN
fcis-30111	45	37	)	)	PUNCT
fcis-30111	45	38	is	be	AUX
fcis-30111	45	39	the	the	DET
fcis-30111	45	40	value	value	NOUN
fcis-30111	45	41	of	of	ADP
fcis-30111	45	42	the	the	DET
fcis-30111	45	43	time	time	NOUN
fcis-30111	45	44	domain	domain	NOUN
fcis-30111	45	45	signal	signal	NOUN
fcis-30111	45	46	at	at	ADP
fcis-30111	45	47	the	the	DET
fcis-30111	45	48	nth	nth	ADJ
fcis-30111	45	49	moment	moment	NOUN
fcis-30111	45	50	.	.	PUNCT
fcis-30111	46	1	for	for	ADP
fcis-30111	46	2	example	example	NOUN
fcis-30111	46	3	,	,	PUNCT
fcis-30111	46	4	an	an	DET
fcis-30111	46	5	energy	energy	NOUN
fcis-30111	46	6	increase	increase	NOUN
fcis-30111	46	7	of	of	ADP
fcis-30111	46	8	35	35	NUM
fcis-30111	46	9	%	%	NOUN
fcis-30111	46	10	or	or	CCONJ
fcis-30111	46	11	more	more	ADJ
fcis-30111	46	12	in	in	ADP
fcis-30111	46	13	the	the	DET
fcis-30111	46	14	2500hz-3000hz	2500hz-3000hz	PROPN
fcis-30111	46	15	band	band	NOUN
fcis-30111	46	16	often	often	ADV
fcis-30111	46	17	indicates	indicate	VERB
fcis-30111	46	18	early	early	ADJ
fcis-30111	46	19	damage	damage	NOUN
fcis-30111	46	20	to	to	ADP
fcis-30111	46	21	the	the	DET
fcis-30111	46	22	inner	inner	ADJ
fcis-30111	46	23	ring	ring	NOUN
fcis-30111	46	24	of	of	ADP
fcis-30111	46	25	the	the	DET
fcis-30111	46	26	bearing	bearing	NOUN
fcis-30111	46	27	.	.	PUNCT
fcis-30111	47	1	in	in	ADP
fcis-30111	47	2	practice	practice	NOUN
fcis-30111	47	3	,	,	PUNCT
fcis-30111	47	4	the	the	DET
fcis-30111	47	5	analysis	analysis	NOUN
fcis-30111	47	6	of	of	ADP
fcis-30111	47	7	30	30	NUM
fcis-30111	47	8	sets	set	NOUN
fcis-30111	47	9	of	of	ADP
fcis-30111	47	10	bearing	bear	VERB
fcis-30111	47	11	data	datum	NOUN
fcis-30111	47	12	shows	show	VERB
fcis-30111	47	13	that	that	SCONJ
fcis-30111	47	14	the	the	DET
fcis-30111	47	15	energy	energy	NOUN
fcis-30111	47	16	difference	difference	NOUN
fcis-30111	47	17	between	between	ADP
fcis-30111	47	18	healthy	healthy	ADJ
fcis-30111	47	19	bearings	bearing	NOUN
fcis-30111	47	20	and	and	CCONJ
fcis-30111	47	21	faulty	faulty	ADJ
fcis-30111	47	22	bearings	bearing	NOUN
fcis-30111	47	23	in	in	ADP
fcis-30111	47	24	this	this	DET
fcis-30111	47	25	band	band	NOUN
fcis-30111	47	26	is	be	AUX
fcis-30111	47	27	as	as	ADV
fcis-30111	47	28	significant	significant	ADJ
fcis-30111	47	29	as	as	ADP
fcis-30111	47	30	p<0.01	p<0.01	PROPN
fcis-30111	47	31	.	.	PUNCT
fcis-30111	48	1	the	the	DET
fcis-30111	48	2	time	time	NOUN
fcis-30111	48	3	-	-	PUNCT
fcis-30111	48	4	frequency	frequency	NOUN
fcis-30111	48	5	analysis	analysis	NOUN
fcis-30111	48	6	adopts	adopt	VERB
fcis-30111	48	7	short	short	ADJ
fcis-30111	48	8	-	-	PUNCT
fcis-30111	48	9	time	time	NOUN
fcis-30111	48	10	fourier	fourier	NOUN
fcis-30111	48	11	transform	transform	NOUN
fcis-30111	48	12	and	and	CCONJ
fcis-30111	48	13	wavelet	wavelet	NOUN
fcis-30111	48	14	packet	packet	NOUN
fcis-30111	48	15	decomposition	decomposition	NOUN
fcis-30111	48	16	,	,	PUNCT
fcis-30111	48	17	which	which	PRON
fcis-30111	48	18	is	be	AUX
fcis-30111	48	19	able	able	ADJ
fcis-30111	48	20	to	to	PART
fcis-30111	48	21	capture	capture	VERB
fcis-30111	48	22	the	the	DET
fcis-30111	48	23	nonstationary	nonstationary	ADJ
fcis-30111	48	24	characteristics	characteristic	NOUN
fcis-30111	48	25	of	of	ADP
fcis-30111	48	26	the	the	DET
fcis-30111	48	27	signals	signal	NOUN
fcis-30111	48	28	.	.	PUNCT
fcis-30111	49	1	for	for	ADP
fcis-30111	49	2	the	the	DET
fcis-30111	49	3	temperature	temperature	NOUN
fcis-30111	49	4	features	feature	NOUN
fcis-30111	49	5	,	,	PUNCT
fcis-30111	49	6	the	the	DET
fcis-30111	49	7	thermal	thermal	ADJ
fcis-30111	49	8	imaging	imaging	NOUN
fcis-30111	49	9	data	datum	NOUN
fcis-30111	49	10	is	be	AUX
fcis-30111	49	11	used	use	VERB
fcis-30111	49	12	to	to	PART
fcis-30111	49	13	assist	assist	VERB
fcis-30111	49	14	in	in	ADP
fcis-30111	49	15	locating	locate	VERB
fcis-30111	49	16	the	the	DET
fcis-30111	49	17	abnormal	abnormal	ADJ
fcis-30111	49	18	hot	hot	ADJ
fcis-30111	49	19	spot	spot	NOUN
fcis-30111	49	20	through	through	ADP
fcis-30111	49	21	regional	regional	ADJ
fcis-30111	49	22	statistics	statistic	NOUN
fcis-30111	49	23	(	(	PUNCT
fcis-30111	49	24	maximum	maximum	ADJ
fcis-30111	49	25	temperature	temperature	NOUN
fcis-30111	49	26	of	of	ADP
fcis-30111	49	27	175.3	175.3	NUM
fcis-30111	49	28	°	°	NOUN
fcis-30111	49	29	c	c	NOUN
fcis-30111	49	30	,	,	PUNCT
fcis-30111	49	31	temperature	temperature	NOUN
fcis-30111	49	32	gradient	gradient	NOUN
fcis-30111	49	33	of	of	ADP
fcis-30111	49	34	4.2	4.2	NUM
fcis-30111	49	35	°	°	NOUN
fcis-30111	49	36	c	c	NOUN
fcis-30111	49	37	/	/	SYM
fcis-30111	49	38	cm	cm	NOUN
fcis-30111	49	39	,	,	PUNCT
fcis-30111	49	40	etc	etc	X
fcis-30111	49	41	.	.	X
fcis-30111	49	42	)	)	PUNCT
fcis-30111	49	43	.	.	PUNCT
fcis-30111	50	1	the	the	DET
fcis-30111	50	2	feature	feature	NOUN
fcis-30111	50	3	selection	selection	NOUN
fcis-30111	50	4	stage	stage	NOUN
fcis-30111	50	5	uses	use	VERB
fcis-30111	50	6	correlation	correlation	NOUN
fcis-30111	50	7	analysis	analysis	NOUN
fcis-30111	50	8	and	and	CCONJ
fcis-30111	50	9	random	random	ADJ
fcis-30111	50	10	forest	forest	NOUN
fcis-30111	50	11	importance	importance	NOUN
fcis-30111	50	12	ranking	rank	VERB
fcis-30111	50	13	to	to	PART
fcis-30111	50	14	filter	filter	VERB
fcis-30111	50	15	out	out	ADP
fcis-30111	50	16	32	32	NUM
fcis-30111	50	17	key	key	ADJ
fcis-30111	50	18	features	feature	NOUN
fcis-30111	50	19	from	from	ADP
fcis-30111	50	20	the	the	DET
fcis-30111	50	21	initial	initial	ADJ
fcis-30111	50	22	128	128	NUM
fcis-30111	50	23	-	-	PUNCT
fcis-30111	50	24	dimensional	dimensional	ADJ
fcis-30111	50	25	features	feature	NOUN
fcis-30111	50	26	,	,	PUNCT
fcis-30111	50	27	which	which	PRON
fcis-30111	50	28	improves	improve	VERB
fcis-30111	50	29	the	the	DET
fcis-30111	50	30	computational	computational	ADJ
fcis-30111	50	31	efficiency	efficiency	NOUN
fcis-30111	50	32	by	by	ADP
fcis-30111	50	33	about	about	ADV
fcis-30111	50	34	65	65	NUM
fcis-30111	50	35	%	%	NOUN
fcis-30111	50	36	while	while	SCONJ
fcis-30111	50	37	maintaining	maintain	VERB
fcis-30111	50	38	93.8	93.8	NUM
fcis-30111	50	39	%	%	NOUN
fcis-30111	50	40	diagnostic	diagnostic	ADJ
fcis-30111	50	41	accuracy	accuracy	NOUN
fcis-30111	50	42	,	,	PUNCT
fcis-30111	50	43	as	as	SCONJ
fcis-30111	50	44	shown	show	VERB
fcis-30111	50	45	in	in	ADP
fcis-30111	50	46	figure	figure	NOUN
fcis-30111	50	47	2	2	NUM
fcis-30111	50	48	.	.	PUNCT
fcis-30111	50	49	figure	figure	NOUN
fcis-30111	50	50	2	2	NUM
fcis-30111	50	51	.	.	NOUN
fcis-30111	50	52	feature	feature	NOUN
fcis-30111	50	53	selection	selection	NOUN
fcis-30111	50	54	performance	performance	NOUN
fcis-30111	50	55	analysis	analysis	NOUN
fcis-30111	50	56	71	71	NUM
fcis-30111	50	57	3.3	3.3	NUM
fcis-30111	50	58	.	.	PUNCT
fcis-30111	51	1	deep	deep	ADJ
fcis-30111	51	2	learning	learning	NOUN
fcis-30111	51	3	based	base	VERB
fcis-30111	51	4	feature	feature	NOUN
fcis-30111	51	5	extraction	extraction	NOUN
fcis-30111	51	6	methods	method	NOUN
fcis-30111	51	7	the	the	DET
fcis-30111	51	8	feature	feature	NOUN
fcis-30111	51	9	extraction	extraction	NOUN
fcis-30111	51	10	method	method	NOUN
fcis-30111	51	11	based	base	VERB
fcis-30111	51	12	on	on	ADP
fcis-30111	51	13	deep	deep	ADJ
fcis-30111	51	14	learning	learning	NOUN
fcis-30111	51	15	breaks	break	NOUN
fcis-30111	51	16	through	through	ADP
fcis-30111	51	17	the	the	DET
fcis-30111	51	18	limitations	limitation	NOUN
fcis-30111	51	19	of	of	ADP
fcis-30111	51	20	traditional	traditional	ADJ
fcis-30111	51	21	feature	feature	NOUN
fcis-30111	51	22	engineering	engineering	NOUN
fcis-30111	51	23	and	and	CCONJ
fcis-30111	51	24	can	can	AUX
fcis-30111	51	25	automatically	automatically	ADV
fcis-30111	51	26	learn	learn	VERB
fcis-30111	51	27	the	the	DET
fcis-30111	51	28	deep	deep	ADJ
fcis-30111	51	29	feature	feature	NOUN
fcis-30111	51	30	representation	representation	NOUN
fcis-30111	51	31	of	of	ADP
fcis-30111	51	32	multimodal	multimodal	NOUN
fcis-30111	51	33	data	datum	NOUN
fcis-30111	51	34	.	.	PUNCT
fcis-30111	52	1	for	for	ADP
fcis-30111	52	2	vibration	vibration	NOUN
fcis-30111	52	3	signals	signal	NOUN
fcis-30111	52	4	,	,	PUNCT
fcis-30111	52	5	a	a	DET
fcis-30111	52	6	one	one	NUM
fcis-30111	52	7	-	-	PUNCT
fcis-30111	52	8	dimensional	dimensional	ADJ
fcis-30111	52	9	convolutional	convolutional	ADJ
fcis-30111	52	10	neural	neural	ADJ
fcis-30111	52	11	network	network	NOUN
fcis-30111	52	12	(	(	PUNCT
fcis-30111	52	13	1d	1d	PROPN
fcis-30111	52	14	-	-	PUNCT
fcis-30111	52	15	cnn	cnn	NOUN
fcis-30111	52	16	)	)	PUNCT
fcis-30111	52	17	is	be	AUX
fcis-30111	52	18	used	use	VERB
fcis-30111	52	19	to	to	PART
fcis-30111	52	20	construct	construct	VERB
fcis-30111	52	21	a	a	DET
fcis-30111	52	22	5	5	NUM
fcis-30111	52	23	-	-	PUNCT
fcis-30111	52	24	layer	layer	NOUN
fcis-30111	52	25	structure	structure	NOUN
fcis-30111	52	26	containing	contain	VERB
fcis-30111	52	27	three	three	NUM
fcis-30111	52	28	convolutional	convolutional	ADJ
fcis-30111	52	29	layers	layer	NOUN
fcis-30111	52	30	(	(	PUNCT
fcis-30111	52	31	with	with	ADP
fcis-30111	52	32	convolution	convolution	NOUN
fcis-30111	52	33	kernel	kernel	NOUN
fcis-30111	52	34	sizes	size	NOUN
fcis-30111	52	35	of	of	ADP
fcis-30111	52	36	64×3	64×3	PROPN
fcis-30111	52	37	,	,	PUNCT
fcis-30111	52	38	32×3	32×3	PROPN
fcis-30111	52	39	,	,	PUNCT
fcis-30111	52	40	and	and	CCONJ
fcis-30111	52	41	16×3	16×3	NUM
fcis-30111	52	42	,	,	PUNCT
fcis-30111	52	43	respectively	respectively	ADV
fcis-30111	52	44	)	)	PUNCT
fcis-30111	52	45	and	and	CCONJ
fcis-30111	52	46	two	two	NUM
fcis-30111	52	47	fully	fully	ADV
fcis-30111	52	48	connected	connected	ADJ
fcis-30111	52	49	layers	layer	NOUN
fcis-30111	52	50	(	(	PUNCT
fcis-30111	52	51	with	with	ADP
fcis-30111	52	52	node	node	ADJ
fcis-30111	52	53	numbers	number	NOUN
fcis-30111	52	54	of	of	ADP
fcis-30111	52	55	128	128	NUM
fcis-30111	52	56	and	and	CCONJ
fcis-30111	52	57	64	64	NUM
fcis-30111	52	58	)	)	PUNCT
fcis-30111	52	59	.	.	PUNCT
fcis-30111	53	1	the	the	DET
fcis-30111	53	2	convolution	convolution	NOUN
fcis-30111	53	3	operation	operation	NOUN
fcis-30111	53	4	extracts	extract	NOUN
fcis-30111	53	5	local	local	ADJ
fcis-30111	53	6	features	feature	NOUN
fcis-30111	53	7	by	by	ADP
fcis-30111	53	8	the	the	DET
fcis-30111	53	9	following	follow	VERB
fcis-30111	53	10	formula	formula	NOUN
fcis-30111	53	11	:	:	PUNCT
fcis-30111	53	12	y	y	PROPN
fcis-30111	53	13	n	n	PROPN
fcis-30111	53	14	∑	∑	PROPN
fcis-30111	53	15	w	w	VERB
fcis-30111	53	16	i	i	NOUN
fcis-30111	53	17	x	x	PROPN
fcis-30111	53	18	n	n	VERB
fcis-30111	53	19	i	i	PRON
fcis-30111	53	20	(	(	PUNCT
fcis-30111	53	21	3	3	NUM
fcis-30111	53	22	)	)	PUNCT
fcis-30111	53	23	where	where	SCONJ
fcis-30111	53	24	y(n	y(n	PRON
fcis-30111	53	25	)	)	PUNCT
fcis-30111	53	26	is	be	AUX
fcis-30111	53	27	the	the	DET
fcis-30111	53	28	nth	nth	NOUN
fcis-30111	53	29	value	value	NOUN
fcis-30111	53	30	of	of	ADP
fcis-30111	53	31	the	the	DET
fcis-30111	53	32	output	output	NOUN
fcis-30111	53	33	signal	signal	NOUN
fcis-30111	53	34	.	.	PUNCT
fcis-30111	54	1	experiments	experiment	NOUN
fcis-30111	54	2	show	show	VERB
fcis-30111	54	3	that	that	SCONJ
fcis-30111	54	4	the	the	DET
fcis-30111	54	5	model	model	NOUN
fcis-30111	54	6	has	have	VERB
fcis-30111	54	7	an	an	DET
fcis-30111	54	8	accuracy	accuracy	NOUN
fcis-30111	54	9	of	of	ADP
fcis-30111	54	10	97.8	97.8	NUM
fcis-30111	54	11	%	%	NOUN
fcis-30111	54	12	in	in	ADP
fcis-30111	54	13	bearing	bear	VERB
fcis-30111	54	14	fault	fault	NOUN
fcis-30111	54	15	identification	identification	NOUN
fcis-30111	54	16	,	,	PUNCT
fcis-30111	54	17	which	which	PRON
fcis-30111	54	18	is	be	AUX
fcis-30111	54	19	5.6	5.6	NUM
fcis-30111	54	20	percentage	percentage	NOUN
fcis-30111	54	21	points	point	NOUN
fcis-30111	54	22	higher	high	ADJ
fcis-30111	54	23	than	than	ADP
fcis-30111	54	24	the	the	DET
fcis-30111	54	25	traditional	traditional	ADJ
fcis-30111	54	26	method	method	NOUN
fcis-30111	54	27	.	.	PUNCT
fcis-30111	55	1	the	the	DET
fcis-30111	55	2	acoustic	acoustic	ADJ
fcis-30111	55	3	signal	signal	NOUN
fcis-30111	55	4	features	feature	NOUN
fcis-30111	55	5	are	be	AUX
fcis-30111	55	6	extracted	extract	VERB
fcis-30111	55	7	using	use	VERB
fcis-30111	55	8	a	a	DET
fcis-30111	55	9	long	long	ADJ
fcis-30111	55	10	short	short	ADJ
fcis-30111	55	11	-	-	PUNCT
fcis-30111	55	12	term	term	NOUN
fcis-30111	55	13	memory	memory	NOUN
fcis-30111	55	14	network	network	NOUN
fcis-30111	55	15	(	(	PUNCT
fcis-30111	55	16	lstm	lstm	PROPN
fcis-30111	55	17	)	)	PUNCT
fcis-30111	55	18	,	,	PUNCT
fcis-30111	55	19	and	and	CCONJ
fcis-30111	55	20	the	the	DET
fcis-30111	55	21	temporal	temporal	ADJ
fcis-30111	55	22	dependence	dependence	NOUN
fcis-30111	55	23	is	be	AUX
fcis-30111	55	24	established	establish	VERB
fcis-30111	55	25	by	by	ADP
fcis-30111	55	26	the	the	DET
fcis-30111	55	27	equation	equation	NOUN
fcis-30111	55	28	f	f	PROPN
fcis-30111	55	29	σ	σ	PROPN
fcis-30111	55	30	w	w	PROPN
fcis-30111	55	31	∙	∙	PROPN
fcis-30111	55	32	h	h	NOUN
fcis-30111	55	33	,	,	PUNCT
fcis-30111	55	34	x	x	PROPN
fcis-30111	55	35	b	b	X
fcis-30111	55	36	(	(	PUNCT
fcis-30111	55	37	4	4	NUM
fcis-30111	55	38	)	)	PUNCT
fcis-30111	55	39	where	where	SCONJ
fcis-30111	55	40	f	f	PROPN
fcis-30111	55	41	t	t	PROPN
fcis-30111	55	42	is	be	AUX
fcis-30111	55	43	the	the	DET
fcis-30111	55	44	output	output	NOUN
fcis-30111	55	45	of	of	ADP
fcis-30111	55	46	the	the	DET
fcis-30111	55	47	forgetting	forget	VERB
fcis-30111	55	48	gate	gate	NOUN
fcis-30111	55	49	at	at	ADP
fcis-30111	55	50	moment	moment	NOUN
fcis-30111	55	51	t.	t.	PROPN
fcis-30111	55	52	the	the	DET
fcis-30111	55	53	network	network	NOUN
fcis-30111	55	54	contains	contain	VERB
fcis-30111	55	55	64	64	NUM
fcis-30111	55	56	lstm	lstm	ADJ
fcis-30111	55	57	units	unit	NOUN
fcis-30111	55	58	,	,	PUNCT
fcis-30111	55	59	which	which	PRON
fcis-30111	55	60	can	can	AUX
fcis-30111	55	61	effectively	effectively	ADV
fcis-30111	55	62	capture	capture	VERB
fcis-30111	55	63	the	the	DET
fcis-30111	55	64	time	time	NOUN
fcis-30111	55	65	-	-	PUNCT
fcis-30111	55	66	varying	vary	VERB
fcis-30111	55	67	characteristics	characteristic	NOUN
fcis-30111	55	68	of	of	ADP
fcis-30111	55	69	sound	sound	NOUN
fcis-30111	55	70	.	.	PUNCT
fcis-30111	56	1	the	the	DET
fcis-30111	56	2	feature	feature	NOUN
fcis-30111	56	3	extraction	extraction	NOUN
fcis-30111	56	4	of	of	ADP
fcis-30111	56	5	image	image	NOUN
fcis-30111	56	6	data	datum	NOUN
fcis-30111	56	7	adopts	adopt	VERB
fcis-30111	56	8	the	the	DET
fcis-30111	56	9	resnet-18	resnet-18	PROPN
fcis-30111	56	10	structure	structure	NOUN
fcis-30111	56	11	,	,	PUNCT
fcis-30111	56	12	which	which	PRON
fcis-30111	56	13	alleviates	alleviate	VERB
fcis-30111	56	14	the	the	DET
fcis-30111	56	15	problem	problem	NOUN
fcis-30111	56	16	of	of	ADP
fcis-30111	56	17	difficult	difficult	ADJ
fcis-30111	56	18	training	training	NOUN
fcis-30111	56	19	of	of	ADP
fcis-30111	56	20	the	the	DET
fcis-30111	56	21	deep	deep	ADJ
fcis-30111	56	22	network	network	NOUN
fcis-30111	56	23	through	through	ADP
fcis-30111	56	24	the	the	DET
fcis-30111	56	25	residual	residual	ADJ
fcis-30111	56	26	connection	connection	NOUN
fcis-30111	56	27	$	$	SYM
fcis-30111	56	28	h(x	h(x	PROPN
fcis-30111	56	29	)	)	PUNCT
fcis-30111	57	1	=	=	SYM
fcis-30111	57	2	f(x	f(x	PROPN
fcis-30111	57	3	)	)	PUNCT
fcis-30111	58	1	+	+	NUM
fcis-30111	58	2	x$.	x$.	VERB
fcis-30111	58	3	multimodal	multimodal	NOUN
fcis-30111	58	4	feature	feature	NOUN
fcis-30111	58	5	fusion	fusion	NOUN
fcis-30111	58	6	adopts	adopt	VERB
fcis-30111	58	7	the	the	DET
fcis-30111	58	8	attention	attention	NOUN
fcis-30111	58	9	mechanism	mechanism	NOUN
fcis-30111	58	10	,	,	PUNCT
fcis-30111	58	11	which	which	PRON
fcis-30111	58	12	dynamically	dynamically	ADV
fcis-30111	58	13	adjusts	adjust	VERB
fcis-30111	58	14	the	the	DET
fcis-30111	58	15	importance	importance	NOUN
fcis-30111	58	16	of	of	ADP
fcis-30111	58	17	each	each	DET
fcis-30111	58	18	modality	modality	NOUN
fcis-30111	58	19	by	by	ADP
fcis-30111	58	20	calculating	calculate	VERB
fcis-30111	58	21	the	the	DET
fcis-30111	58	22	weights	weight	NOUN
fcis-30111	58	23	.	.	PUNCT
fcis-30111	59	1	in	in	ADP
fcis-30111	59	2	the	the	DET
fcis-30111	59	3	actual	actual	ADJ
fcis-30111	59	4	deployment	deployment	NOUN
fcis-30111	59	5	,	,	PUNCT
fcis-30111	59	6	the	the	DET
fcis-30111	59	7	size	size	NOUN
fcis-30111	59	8	of	of	ADP
fcis-30111	59	9	the	the	DET
fcis-30111	59	10	model	model	NOUN
fcis-30111	59	11	is	be	AUX
fcis-30111	59	12	only	only	ADV
fcis-30111	59	13	8.4	8.4	NUM
fcis-30111	59	14	mb	mb	NOUN
fcis-30111	59	15	after	after	ADP
fcis-30111	59	16	quantisation	quantisation	NOUN
fcis-30111	59	17	,	,	PUNCT
fcis-30111	59	18	which	which	PRON
fcis-30111	59	19	can	can	AUX
fcis-30111	59	20	achieve	achieve	VERB
fcis-30111	59	21	the	the	DET
fcis-30111	59	22	inference	inference	NOUN
fcis-30111	59	23	speed	speed	NOUN
fcis-30111	59	24	within	within	ADP
fcis-30111	59	25	15ms	15ms	NOUN
fcis-30111	59	26	on	on	ADP
fcis-30111	59	27	the	the	DET
fcis-30111	59	28	edge	edge	NOUN
fcis-30111	59	29	computing	compute	VERB
fcis-30111	59	30	device	device	NOUN
fcis-30111	59	31	to	to	PART
fcis-30111	59	32	meet	meet	VERB
fcis-30111	59	33	the	the	DET
fcis-30111	59	34	real	real	ADJ
fcis-30111	59	35	-	-	PUNCT
fcis-30111	59	36	time	time	NOUN
fcis-30111	59	37	monitoring	monitoring	NOUN
fcis-30111	59	38	requirements	requirement	NOUN
fcis-30111	59	39	,	,	PUNCT
fcis-30111	59	40	as	as	SCONJ
fcis-30111	59	41	shown	show	VERB
fcis-30111	59	42	in	in	ADP
fcis-30111	59	43	figure	figure	NOUN
fcis-30111	59	44	3	3	NUM
fcis-30111	59	45	.	.	PUNCT
fcis-30111	59	46	figure	figure	VERB
fcis-30111	59	47	3	3	NUM
fcis-30111	59	48	.	.	PUNCT
fcis-30111	59	49	model	model	NOUN
fcis-30111	59	50	inference	inference	NOUN
fcis-30111	59	51	speed	speed	NOUN
fcis-30111	59	52	on	on	ADP
fcis-30111	59	53	edge	edge	NOUN
fcis-30111	59	54	devices	device	NOUN
fcis-30111	59	55	4	4	NUM
fcis-30111	59	56	.	.	PUNCT
fcis-30111	59	57	multimodal	multimodal	NOUN
fcis-30111	59	58	data	datum	NOUN
fcis-30111	59	59	fusion	fusion	NOUN
fcis-30111	59	60	models	model	NOUN
fcis-30111	59	61	and	and	CCONJ
fcis-30111	59	62	algorithms	algorithm	NOUN
fcis-30111	59	63	4.1	4.1	NUM
fcis-30111	59	64	.	.	PUNCT
fcis-30111	60	1	data	datum	NOUN
fcis-30111	60	2	and	and	CCONJ
fcis-30111	60	3	feature	feature	NOUN
fcis-30111	60	4	level	level	NOUN
fcis-30111	60	5	fusion	fusion	NOUN
fcis-30111	60	6	method	method	NOUN
fcis-30111	60	7	the	the	DET
fcis-30111	60	8	data	datum	NOUN
fcis-30111	60	9	and	and	CCONJ
fcis-30111	60	10	feature	feature	NOUN
fcis-30111	60	11	level	level	NOUN
fcis-30111	60	12	fusion	fusion	NOUN
fcis-30111	60	13	method	method	NOUN
fcis-30111	60	14	are	be	AUX
fcis-30111	60	15	the	the	DET
fcis-30111	60	16	basic	basic	ADJ
fcis-30111	60	17	link	link	NOUN
fcis-30111	60	18	of	of	ADP
fcis-30111	60	19	the	the	DET
fcis-30111	60	20	wheelset	wheelset	NOUN
fcis-30111	60	21	monitoring	monitoring	NOUN
fcis-30111	60	22	system	system	NOUN
fcis-30111	60	23	to	to	PART
fcis-30111	60	24	process	process	VERB
fcis-30111	60	25	multimodal	multimodal	ADJ
fcis-30111	60	26	data	datum	NOUN
fcis-30111	60	27	.	.	PUNCT
fcis-30111	61	1	based	base	VERB
fcis-30111	61	2	on	on	ADP
fcis-30111	61	3	the	the	DET
fcis-30111	61	4	actual	actual	ADJ
fcis-30111	61	5	project	project	NOUN
fcis-30111	61	6	implementation	implementation	NOUN
fcis-30111	61	7	experience	experience	NOUN
fcis-30111	61	8	,	,	PUNCT
fcis-30111	61	9	the	the	DET
fcis-30111	61	10	data	datum	NOUN
fcis-30111	61	11	level	level	NOUN
fcis-30111	61	12	fusion	fusion	NOUN
fcis-30111	61	13	in	in	ADP
fcis-30111	61	14	wheelset	wheelset	ADJ
fcis-30111	61	15	monitoring	monitoring	NOUN
fcis-30111	61	16	mainly	mainly	ADV
fcis-30111	61	17	adopts	adopt	VERB
fcis-30111	61	18	the	the	DET
fcis-30111	61	19	time	time	NOUN
fcis-30111	61	20	sequence	sequence	NOUN
fcis-30111	61	21	alignment	alignment	NOUN
fcis-30111	61	22	and	and	CCONJ
fcis-30111	61	23	standardisation	standardisation	NOUN
fcis-30111	61	24	method	method	NOUN
fcis-30111	61	25	.	.	PUNCT
fcis-30111	62	1	the	the	DET
fcis-30111	62	2	vibration	vibration	NOUN
fcis-30111	62	3	,	,	PUNCT
fcis-30111	62	4	acoustic	acoustic	ADJ
fcis-30111	62	5	,	,	PUNCT
fcis-30111	62	6	temperature	temperature	NOUN
fcis-30111	62	7	and	and	CCONJ
fcis-30111	62	8	visual	visual	ADJ
fcis-30111	62	9	data	datum	NOUN
fcis-30111	62	10	collected	collect	VERB
fcis-30111	62	11	through	through	ADP
fcis-30111	62	12	the	the	DET
fcis-30111	62	13	wireless	wireless	ADJ
fcis-30111	62	14	sensor	sensor	NOUN
fcis-30111	62	15	network	network	NOUN
fcis-30111	62	16	are	be	AUX
fcis-30111	62	17	synchronised	synchronise	VERB
fcis-30111	62	18	with	with	ADP
fcis-30111	62	19	timestamps	timestamp	NOUN
fcis-30111	62	20	,	,	PUNCT
fcis-30111	62	21	and	and	CCONJ
fcis-30111	62	22	tensor	tensor	NOUN
fcis-30111	62	23	fusion	fusion	NOUN
fcis-30111	62	24	is	be	AUX
fcis-30111	62	25	used	use	VERB
fcis-30111	62	26	to	to	PART
fcis-30111	62	27	construct	construct	VERB
fcis-30111	62	28	a	a	DET
fcis-30111	62	29	unified	unified	ADJ
fcis-30111	62	30	data	datum	NOUN
fcis-30111	62	31	representation[8	representation[8	NOUN
fcis-30111	62	32	]	]	PUNCT
fcis-30111	62	33	.	.	PUNCT
fcis-30111	63	1	the	the	DET
fcis-30111	63	2	feature	feature	NOUN
fcis-30111	63	3	-	-	PUNCT
fcis-30111	63	4	level	level	NOUN
fcis-30111	63	5	fusion	fusion	NOUN
fcis-30111	63	6	adopts	adopt	VERB
fcis-30111	63	7	a	a	DET
fcis-30111	63	8	weighted	weighted	ADJ
fcis-30111	63	9	fusion	fusion	NOUN
fcis-30111	63	10	strategy	strategy	NOUN
fcis-30111	63	11	,	,	PUNCT
fcis-30111	63	12	and	and	CCONJ
fcis-30111	63	13	the	the	DET
fcis-30111	63	14	weighting	weighting	NOUN
fcis-30111	63	15	coefficients	coefficient	NOUN
fcis-30111	63	16	are	be	AUX
fcis-30111	63	17	dynamically	dynamically	ADV
fcis-30111	63	18	adjusted	adjust	VERB
fcis-30111	63	19	according	accord	VERB
fcis-30111	63	20	to	to	ADP
fcis-30111	63	21	different	different	ADJ
fcis-30111	63	22	working	working	NOUN
fcis-30111	63	23	conditions	condition	NOUN
fcis-30111	63	24	.	.	PUNCT
fcis-30111	64	1	in	in	ADP
fcis-30111	64	2	the	the	DET
fcis-30111	64	3	project	project	NOUN
fcis-30111	64	4	test	test	NOUN
fcis-30111	64	5	,	,	PUNCT
fcis-30111	64	6	the	the	DET
fcis-30111	64	7	effect	effect	NOUN
fcis-30111	64	8	of	of	ADP
fcis-30111	64	9	feature	feature	NOUN
fcis-30111	64	10	fusion	fusion	NOUN
fcis-30111	64	11	for	for	ADP
fcis-30111	64	12	four	four	NUM
fcis-30111	64	13	types	type	NOUN
fcis-30111	64	14	of	of	ADP
fcis-30111	64	15	wheelset	wheelset	ADJ
fcis-30111	64	16	faults	fault	NOUN
fcis-30111	64	17	(	(	PUNCT
fcis-30111	64	18	rim	rim	NOUN
fcis-30111	64	19	cracks	crack	NOUN
fcis-30111	64	20	,	,	PUNCT
fcis-30111	64	21	tread	tread	NOUN
fcis-30111	64	22	wear	wear	NOUN
fcis-30111	64	23	,	,	PUNCT
fcis-30111	64	24	bearing	bear	VERB
fcis-30111	64	25	damage	damage	NOUN
fcis-30111	64	26	,	,	PUNCT
fcis-30111	64	27	and	and	CCONJ
fcis-30111	64	28	axle	axle	NOUN
fcis-30111	64	29	bending	bending	NOUN
fcis-30111	64	30	)	)	PUNCT
fcis-30111	64	31	is	be	AUX
fcis-30111	64	32	shown	show	VERB
fcis-30111	64	33	in	in	ADP
fcis-30111	64	34	table	table	NOUN
fcis-30111	64	35	1	1	NUM
fcis-30111	64	36	.	.	PUNCT
fcis-30111	65	1	as	as	SCONJ
fcis-30111	65	2	seen	see	VERB
fcis-30111	65	3	in	in	ADP
fcis-30111	65	4	the	the	DET
fcis-30111	65	5	table	table	NOUN
fcis-30111	65	6	,	,	PUNCT
fcis-30111	65	7	the	the	DET
fcis-30111	65	8	recognition	recognition	NOUN
fcis-30111	65	9	rate	rate	NOUN
fcis-30111	65	10	after	after	SCONJ
fcis-30111	65	11	feature	feature	NOUN
fcis-30111	65	12	fusion	fusion	NOUN
fcis-30111	65	13	is	be	AUX
fcis-30111	65	14	generally	generally	ADV
fcis-30111	65	15	higher	high	ADJ
fcis-30111	65	16	than	than	ADP
fcis-30111	65	17	that	that	PRON
fcis-30111	65	18	of	of	ADP
fcis-30111	65	19	a	a	DET
fcis-30111	65	20	single	single	ADJ
fcis-30111	65	21	feature	feature	NOUN
fcis-30111	65	22	,	,	PUNCT
fcis-30111	65	23	with	with	ADP
fcis-30111	65	24	an	an	DET
fcis-30111	65	25	average	average	ADJ
fcis-30111	65	26	improvement	improvement	NOUN
fcis-30111	65	27	of	of	ADP
fcis-30111	65	28	17.3	17.3	NUM
fcis-30111	65	29	%	%	NOUN
fcis-30111	65	30	.	.	PUNCT
fcis-30111	66	1	in	in	ADP
fcis-30111	66	2	the	the	DET
fcis-30111	66	3	field	field	NOUN
fcis-30111	66	4	test	test	NOUN
fcis-30111	66	5	of	of	ADP
fcis-30111	66	6	high	high	ADJ
fcis-30111	66	7	-	-	PUNCT
fcis-30111	66	8	speed	speed	NOUN
fcis-30111	66	9	train	train	NOUN
fcis-30111	66	10	,	,	PUNCT
fcis-30111	66	11	the	the	DET
fcis-30111	66	12	feature	feature	NOUN
fcis-30111	66	13	fusion	fusion	NOUN
fcis-30111	66	14	method	method	NOUN
fcis-30111	66	15	can	can	AUX
fcis-30111	66	16	still	still	ADV
fcis-30111	66	17	maintain	maintain	VERB
fcis-30111	66	18	91.5	91.5	NUM
fcis-30111	66	19	%	%	NOUN
fcis-30111	66	20	recognition	recognition	NOUN
fcis-30111	66	21	accuracy	accuracy	NOUN
fcis-30111	66	22	under	under	ADP
fcis-30111	66	23	the	the	DET
fcis-30111	66	24	noisy	noisy	ADJ
fcis-30111	66	25	environment	environment	NOUN
fcis-30111	66	26	(	(	PUNCT
fcis-30111	66	27	signal	signal	NOUN
fcis-30111	66	28	-	-	PUNCT
fcis-30111	66	29	tonoise	tonoise	NOUN
fcis-30111	66	30	ratio	ratio	NOUN
fcis-30111	66	31	-5db	-5db	NUM
fcis-30111	66	32	)	)	PUNCT
fcis-30111	66	33	,	,	PUNCT
fcis-30111	66	34	and	and	CCONJ
fcis-30111	66	35	the	the	DET
fcis-30111	66	36	anti	anti	ADJ
fcis-30111	66	37	-	-	ADJ
fcis-30111	66	38	interference	interference	ADJ
fcis-30111	66	39	ability	ability	NOUN
fcis-30111	66	40	is	be	AUX
fcis-30111	66	41	significantly	significantly	ADV
fcis-30111	66	42	improved	improve	VERB
fcis-30111	66	43	.	.	PUNCT
fcis-30111	67	1	the	the	DET
fcis-30111	67	2	dynamic	dynamic	ADJ
fcis-30111	67	3	weighting	weighting	NOUN
fcis-30111	67	4	mechanism	mechanism	NOUN
fcis-30111	67	5	ensures	ensure	VERB
fcis-30111	67	6	the	the	DET
fcis-30111	67	7	stability	stability	NOUN
fcis-30111	67	8	of	of	ADP
fcis-30111	67	9	the	the	DET
fcis-30111	67	10	fusion	fusion	NOUN
fcis-30111	67	11	results	result	NOUN
fcis-30111	67	12	by	by	ADP
fcis-30111	67	13	evaluating	evaluate	VERB
fcis-30111	67	14	the	the	DET
fcis-30111	67	15	data	data	NOUN
fcis-30111	67	16	quality	quality	NOUN
fcis-30111	67	17	of	of	ADP
fcis-30111	67	18	each	each	DET
fcis-30111	67	19	sensor	sensor	NOUN
fcis-30111	67	20	and	and	CCONJ
fcis-30111	67	21	automatically	automatically	ADV
fcis-30111	67	22	reducing	reduce	VERB
fcis-30111	67	23	its	its	PRON
fcis-30111	67	24	weight	weight	NOUN
fcis-30111	67	25	when	when	SCONJ
fcis-30111	67	26	the	the	DET
fcis-30111	67	27	reliability	reliability	NOUN
fcis-30111	67	28	of	of	ADP
fcis-30111	67	29	a	a	DET
fcis-30111	67	30	sensor	sensor	NOUN
fcis-30111	67	31	decreases	decrease	VERB
fcis-30111	67	32	.	.	PUNCT
fcis-30111	68	1	table	table	NOUN
fcis-30111	68	2	1	1	NUM
fcis-30111	68	3	.	.	PUNCT
fcis-30111	68	4	comparison	comparison	NOUN
fcis-30111	68	5	of	of	ADP
fcis-30111	68	6	the	the	DET
fcis-30111	68	7	recognition	recognition	NOUN
fcis-30111	68	8	rate	rate	NOUN
fcis-30111	68	9	of	of	ADP
fcis-30111	68	10	wheelset	wheelset	ADJ
fcis-30111	68	11	faults	fault	NOUN
fcis-30111	68	12	by	by	ADP
fcis-30111	68	13	different	different	ADJ
fcis-30111	68	14	feature	feature	NOUN
fcis-30111	68	15	fusion	fusion	NOUN
fcis-30111	68	16	methods	method	NOUN
fcis-30111	68	17	.	.	PUNCT
fcis-30111	69	1	fault	fault	NOUN
fcis-30111	69	2	type	type	NOUN
fcis-30111	69	3	vibration	vibration	NOUN
fcis-30111	69	4	feature	feature	VERB
fcis-30111	69	5	acoustic	acoustic	ADJ
fcis-30111	69	6	feature	feature	NOUN
fcis-30111	69	7	temperature	temperature	NOUN
fcis-30111	69	8	feature	feature	NOUN
fcis-30111	69	9	visual	visual	ADJ
fcis-30111	69	10	feature	feature	NOUN
fcis-30111	69	11	feature	feature	NOUN
fcis-30111	69	12	concatenation	concatenation	NOUN
fcis-30111	69	13	fusion	fusion	NOUN
fcis-30111	69	14	dynamic	dynamic	ADJ
fcis-30111	69	15	weighted	weight	VERB
fcis-30111	69	16	fusion	fusion	NOUN
fcis-30111	69	17	rim	rim	NOUN
fcis-30111	69	18	crack	crack	VERB
fcis-30111	69	19	78.3	78.3	NUM
fcis-30111	69	20	72.6	72.6	NUM
fcis-30111	69	21	65.4	65.4	NUM
fcis-30111	69	22	82.1	82.1	NUM
fcis-30111	69	23	89.5	89.5	NUM
fcis-30111	69	24	93.8	93.8	NUM
fcis-30111	69	25	tread	tread	NOUN
fcis-30111	69	26	wear	wear	VERB
fcis-30111	69	27	72.5	72.5	NUM
fcis-30111	69	28	68.4	68.4	NUM
fcis-30111	69	29	58.9	58.9	NUM
fcis-30111	69	30	85.7	85.7	NUM
fcis-30111	69	31	88.2	88.2	NUM
fcis-30111	69	32	92.5	92.5	NUM
fcis-30111	69	33	bearing	bearing	NOUN
fcis-30111	69	34	damage	damage	NOUN
fcis-30111	69	35	86.4	86.4	NUM
fcis-30111	69	36	82.1	82.1	NUM
fcis-30111	69	37	89.3	89.3	NUM
fcis-30111	69	38	70.6	70.6	NUM
fcis-30111	69	39	91.7	91.7	NUM
fcis-30111	69	40	95.2	95.2	NUM
fcis-30111	69	41	axle	axle	NOUN
fcis-30111	69	42	bending	bend	VERB
fcis-30111	69	43	80.2	80.2	NUM
fcis-30111	69	44	75.9	75.9	NUM
fcis-30111	69	45	64.8	64.8	NUM
fcis-30111	69	46	79.5	79.5	NUM
fcis-30111	69	47	87.6	87.6	NUM
fcis-30111	69	48	90.3	90.3	NUM
fcis-30111	69	49	4.2	4.2	NUM
fcis-30111	69	50	.	.	PUNCT
fcis-30111	70	1	decision	decision	NOUN
fcis-30111	70	2	-	-	PUNCT
fcis-30111	70	3	level	level	NOUN
fcis-30111	70	4	fusion	fusion	NOUN
fcis-30111	70	5	and	and	CCONJ
fcis-30111	70	6	deep	deep	ADJ
fcis-30111	70	7	learning	learning	NOUN
fcis-30111	70	8	fusion	fusion	NOUN
fcis-30111	70	9	architecture	architecture	NOUN
fcis-30111	70	10	the	the	DET
fcis-30111	70	11	decision	decision	NOUN
fcis-30111	70	12	-	-	PUNCT
fcis-30111	70	13	level	level	NOUN
fcis-30111	70	14	fusion	fusion	NOUN
fcis-30111	70	15	and	and	CCONJ
fcis-30111	70	16	deep	deep	ADJ
fcis-30111	70	17	learning	learning	NOUN
fcis-30111	70	18	fusion	fusion	NOUN
fcis-30111	70	19	architecture	architecture	NOUN
fcis-30111	70	20	constructs	construct	VERB
fcis-30111	70	21	a	a	DET
fcis-30111	70	22	multi	multi	ADJ
fcis-30111	70	23	-	-	ADJ
fcis-30111	70	24	level	level	ADJ
fcis-30111	70	25	fusion	fusion	NOUN
fcis-30111	70	26	framework	framework	NOUN
fcis-30111	70	27	for	for	ADP
fcis-30111	70	28	wheelset	wheelset	ADJ
fcis-30111	70	29	condition	condition	NOUN
fcis-30111	70	30	monitoring	monitoring	NOUN
fcis-30111	70	31	.	.	PUNCT
fcis-30111	71	1	the	the	DET
fcis-30111	71	2	multi	multi	ADJ
fcis-30111	71	3	-	-	ADJ
fcis-30111	71	4	stream	stream	ADJ
fcis-30111	71	5	deep	deep	ADJ
fcis-30111	71	6	learning	learning	NOUN
fcis-30111	71	7	fusion	fusion	NOUN
fcis-30111	71	8	architecture	architecture	NOUN
fcis-30111	71	9	based	base	VERB
fcis-30111	71	10	on	on	ADP
fcis-30111	71	11	the	the	DET
fcis-30111	71	12	actual	actual	ADJ
fcis-30111	71	13	monitoring	monitoring	NOUN
fcis-30111	71	14	data	datum	NOUN
fcis-30111	71	15	is	be	AUX
fcis-30111	71	16	shown	show	VERB
fcis-30111	71	17	in	in	ADP
fcis-30111	71	18	figure	figure	NOUN
fcis-30111	71	19	4	4	NUM
fcis-30111	71	20	,	,	PUNCT
fcis-30111	71	21	which	which	PRON
fcis-30111	71	22	contains	contain	VERB
fcis-30111	71	23	four	four	NUM
fcis-30111	71	24	parallel	parallel	ADJ
fcis-30111	71	25	processing	processing	NOUN
fcis-30111	71	26	branches	branch	NOUN
fcis-30111	71	27	and	and	CCONJ
fcis-30111	71	28	one	one	NUM
fcis-30111	71	29	fusion	fusion	NOUN
fcis-30111	71	30	module	module	NOUN
fcis-30111	71	31	.	.	PUNCT
fcis-30111	72	1	each	each	DET
fcis-30111	72	2	branch	branch	NOUN
fcis-30111	72	3	processes	process	VERB
fcis-30111	72	4	one	one	NUM
fcis-30111	72	5	kind	kind	NOUN
fcis-30111	72	6	of	of	ADP
fcis-30111	72	7	modal	modal	ADJ
fcis-30111	72	8	data	datum	NOUN
fcis-30111	72	9	(	(	PUNCT
fcis-30111	72	10	vibration	vibration	NOUN
fcis-30111	72	11	,	,	PUNCT
fcis-30111	72	12	acoustic	acoustic	ADJ
fcis-30111	72	13	,	,	PUNCT
fcis-30111	72	14	temperature	temperature	NOUN
fcis-30111	72	15	,	,	PUNCT
fcis-30111	72	16	and	and	CCONJ
fcis-30111	72	17	vision	vision	NOUN
fcis-30111	72	18	)	)	PUNCT
fcis-30111	72	19	,	,	PUNCT
fcis-30111	72	20	extracts	extract	NOUN
fcis-30111	72	21	modal	modal	ADJ
fcis-30111	72	22	-	-	PUNCT
fcis-30111	72	23	specific	specific	ADJ
fcis-30111	72	24	features	feature	NOUN
fcis-30111	72	25	,	,	PUNCT
fcis-30111	72	26	and	and	CCONJ
fcis-30111	72	27	then	then	ADV
fcis-30111	72	28	fuses	fuse	VERB
fcis-30111	72	29	them	they	PRON
fcis-30111	72	30	through	through	ADP
fcis-30111	72	31	the	the	DET
fcis-30111	72	32	cross	cross	ADJ
fcis-30111	72	33	-	-	ADJ
fcis-30111	72	34	modal	modal	ADJ
fcis-30111	72	35	attention	attention	NOUN
fcis-30111	72	36	mechanism	mechanism	NOUN
fcis-30111	72	37	.	.	PUNCT
fcis-30111	73	1	in	in	ADP
fcis-30111	73	2	the	the	DET
fcis-30111	73	3	field	field	NOUN
fcis-30111	73	4	application	application	NOUN
fcis-30111	73	5	of	of	ADP
fcis-30111	73	6	beijing	beijing	PROPN
fcis-30111	73	7	-	-	PUNCT
fcis-30111	73	8	shanghai	shanghai	PROPN
fcis-30111	73	9	highspeed	highspeed	PROPN
fcis-30111	73	10	railway	railway	NOUN
fcis-30111	73	11	with	with	ADP
fcis-30111	73	12	10	10	NUM
fcis-30111	73	13	trainsets	trainset	NOUN
fcis-30111	73	14	,	,	PUNCT
fcis-30111	73	15	the	the	DET
fcis-30111	73	16	architecture	architecture	NOUN
fcis-30111	73	17	detects	detect	VERB
fcis-30111	73	18	early	early	ADJ
fcis-30111	73	19	abnormalities	abnormality	NOUN
fcis-30111	73	20	of	of	ADP
fcis-30111	73	21	wheelsets	wheelset	NOUN
fcis-30111	73	22	3.5	3.5	NUM
fcis-30111	73	23	days	day	NOUN
fcis-30111	73	24	earlier	early	ADV
fcis-30111	73	25	than	than	ADP
fcis-30111	73	26	a	a	DET
fcis-30111	73	27	single	single	ADJ
fcis-30111	73	28	model	model	NOUN
fcis-30111	73	29	,	,	PUNCT
fcis-30111	73	30	which	which	PRON
fcis-30111	73	31	significantly	significantly	ADV
fcis-30111	73	32	expands	expand	VERB
fcis-30111	73	33	the	the	DET
fcis-30111	73	34	time	time	NOUN
fcis-30111	73	35	window	window	NOUN
fcis-30111	73	36	for	for	ADP
fcis-30111	73	37	fault	fault	NOUN
fcis-30111	73	38	warning	warn	VERB
fcis-30111	73	39	.	.	PUNCT
fcis-30111	74	1	the	the	DET
fcis-30111	74	2	cross	cross	ADJ
fcis-30111	74	3	-	-	ADJ
fcis-30111	74	4	modal	modal	ADJ
fcis-30111	74	5	attention	attention	NOUN
fcis-30111	74	6	mechanism	mechanism	NOUN
fcis-30111	74	7	automatically	automatically	ADV
fcis-30111	74	8	adjusts	adjust	VERB
fcis-30111	74	9	the	the	DET
fcis-30111	74	10	weights	weight	NOUN
fcis-30111	74	11	of	of	ADP
fcis-30111	74	12	each	each	DET
fcis-30111	74	13	modality	modality	NOUN
fcis-30111	74	14	under	under	ADP
fcis-30111	74	15	different	different	ADJ
fcis-30111	74	16	working	work	VERB
fcis-30111	74	17	72	72	NUM
fcis-30111	74	18	conditions	condition	NOUN
fcis-30111	74	19	,	,	PUNCT
fcis-30111	74	20	for	for	ADP
fcis-30111	74	21	example	example	NOUN
fcis-30111	74	22	,	,	PUNCT
fcis-30111	74	23	it	it	PRON
fcis-30111	74	24	automatically	automatically	ADV
fcis-30111	74	25	reduces	reduce	VERB
fcis-30111	74	26	the	the	DET
fcis-30111	74	27	weight	weight	NOUN
fcis-30111	74	28	of	of	ADP
fcis-30111	74	29	the	the	DET
fcis-30111	74	30	acoustic	acoustic	ADJ
fcis-30111	74	31	branch	branch	NOUN
fcis-30111	74	32	(	(	PUNCT
fcis-30111	74	33	from	from	ADP
fcis-30111	74	34	0.28	0.28	NUM
fcis-30111	74	35	to	to	ADP
fcis-30111	74	36	0.08	0.08	NUM
fcis-30111	74	37	)	)	PUNCT
fcis-30111	74	38	and	and	CCONJ
fcis-30111	74	39	raises	raise	VERB
fcis-30111	74	40	the	the	DET
fcis-30111	74	41	weight	weight	NOUN
fcis-30111	74	42	of	of	ADP
fcis-30111	74	43	the	the	DET
fcis-30111	74	44	vibration	vibration	NOUN
fcis-30111	74	45	branch	branch	NOUN
fcis-30111	74	46	(	(	PUNCT
fcis-30111	74	47	from	from	ADP
fcis-30111	74	48	0.32	0.32	NUM
fcis-30111	74	49	to	to	ADP
fcis-30111	74	50	0.45	0.45	NUM
fcis-30111	74	51	)	)	PUNCT
fcis-30111	74	52	when	when	SCONJ
fcis-30111	74	53	travelling	travel	VERB
fcis-30111	74	54	in	in	ADP
fcis-30111	74	55	tunnels	tunnel	NOUN
fcis-30111	74	56	.	.	PUNCT
fcis-30111	75	1	after	after	SCONJ
fcis-30111	75	2	the	the	DET
fcis-30111	75	3	model	model	NOUN
fcis-30111	75	4	is	be	AUX
fcis-30111	75	5	deployed	deploy	VERB
fcis-30111	75	6	on	on	ADP
fcis-30111	75	7	the	the	DET
fcis-30111	75	8	edge	edge	NOUN
fcis-30111	75	9	computing	compute	VERB
fcis-30111	75	10	unit	unit	NOUN
fcis-30111	75	11	,	,	PUNCT
fcis-30111	75	12	the	the	DET
fcis-30111	75	13	average	average	ADJ
fcis-30111	75	14	inference	inference	NOUN
fcis-30111	75	15	time	time	NOUN
fcis-30111	75	16	is	be	AUX
fcis-30111	75	17	87	87	NUM
fcis-30111	75	18	ms	ms	NOUN
fcis-30111	75	19	,	,	PUNCT
fcis-30111	75	20	which	which	PRON
fcis-30111	75	21	meets	meet	VERB
fcis-30111	75	22	the	the	DET
fcis-30111	75	23	real	real	ADJ
fcis-30111	75	24	-	-	PUNCT
fcis-30111	75	25	time	time	NOUN
fcis-30111	75	26	monitoring	monitoring	NOUN
fcis-30111	75	27	requirements[9	requirements[9	PROPN
fcis-30111	75	28	]	]	PUNCT
fcis-30111	75	29	.	.	PUNCT
fcis-30111	76	1	in	in	ADP
fcis-30111	76	2	response	response	NOUN
fcis-30111	76	3	to	to	ADP
fcis-30111	76	4	sudden	sudden	ADJ
fcis-30111	76	5	changes	change	NOUN
fcis-30111	76	6	in	in	ADP
fcis-30111	76	7	working	working	NOUN
fcis-30111	76	8	conditions	condition	NOUN
fcis-30111	76	9	,	,	PUNCT
fcis-30111	76	10	the	the	DET
fcis-30111	76	11	dynamic	dynamic	ADJ
fcis-30111	76	12	decision	decision	NOUN
fcis-30111	76	13	fusion	fusion	NOUN
fcis-30111	76	14	mechanism	mechanism	NOUN
fcis-30111	76	15	is	be	AUX
fcis-30111	76	16	able	able	ADJ
fcis-30111	76	17	to	to	PART
fcis-30111	76	18	complete	complete	VERB
fcis-30111	76	19	strategy	strategy	NOUN
fcis-30111	76	20	adjustment	adjustment	NOUN
fcis-30111	76	21	within	within	ADP
fcis-30111	76	22	200	200	NUM
fcis-30111	76	23	ms	ms	NOUN
fcis-30111	76	24	,	,	PUNCT
fcis-30111	76	25	ensuring	ensure	VERB
fcis-30111	76	26	system	system	NOUN
fcis-30111	76	27	stability	stability	NOUN
fcis-30111	76	28	.	.	PUNCT
fcis-30111	77	1	the	the	DET
fcis-30111	77	2	decision	decision	NOUN
fcis-30111	77	3	-	-	PUNCT
fcis-30111	77	4	level	level	NOUN
fcis-30111	77	5	fusion	fusion	NOUN
fcis-30111	77	6	module	module	NOUN
fcis-30111	77	7	significantly	significantly	ADV
fcis-30111	77	8	reduces	reduce	VERB
fcis-30111	77	9	the	the	DET
fcis-30111	77	10	false	false	ADJ
fcis-30111	77	11	alarm	alarm	NOUN
fcis-30111	77	12	rate	rate	NOUN
fcis-30111	77	13	from	from	ADP
fcis-30111	77	14	8.3	8.3	NUM
fcis-30111	77	15	%	%	NOUN
fcis-30111	77	16	to	to	PART
fcis-30111	77	17	1.2	1.2	NUM
fcis-30111	77	18	%	%	NOUN
fcis-30111	77	19	through	through	ADP
fcis-30111	77	20	the	the	DET
fcis-30111	77	21	voting	voting	NOUN
fcis-30111	77	22	mechanism	mechanism	NOUN
fcis-30111	77	23	and	and	CCONJ
fcis-30111	77	24	confidence	confidence	NOUN
fcis-30111	77	25	analysis	analysis	NOUN
fcis-30111	77	26	.	.	PUNCT
fcis-30111	78	1	figure	figure	NOUN
fcis-30111	78	2	4	4	NUM
fcis-30111	78	3	.	.	PUNCT
fcis-30111	78	4	overall	overall	ADJ
fcis-30111	78	5	architecture	architecture	NOUN
fcis-30111	78	6	of	of	ADP
fcis-30111	78	7	the	the	DET
fcis-30111	78	8	multimodal	multimodal	NOUN
fcis-30111	78	9	data	datum	NOUN
fcis-30111	78	10	fusion	fusion	NOUN
fcis-30111	78	11	system	system	NOUN
fcis-30111	78	12	for	for	ADP
fcis-30111	78	13	wheel	wheel	NOUN
fcis-30111	78	14	-	-	PUNCT
fcis-30111	78	15	to	to	ADP
fcis-30111	78	16	-	-	PUNCT
fcis-30111	78	17	wheel	wheel	NOUN
fcis-30111	78	18	health	health	NOUN
fcis-30111	78	19	monitoring	monitor	VERB
fcis-30111	78	20	4.3	4.3	NUM
fcis-30111	78	21	.	.	PUNCT
fcis-30111	78	22	fusion	fusion	NOUN
fcis-30111	78	23	model	model	NOUN
fcis-30111	78	24	optimisation	optimisation	NOUN
fcis-30111	78	25	and	and	CCONJ
fcis-30111	78	26	evaluation	evaluation	NOUN
fcis-30111	78	27	the	the	DET
fcis-30111	78	28	optimisation	optimisation	NOUN
fcis-30111	78	29	and	and	CCONJ
fcis-30111	78	30	evaluation	evaluation	NOUN
fcis-30111	78	31	phase	phase	NOUN
fcis-30111	78	32	of	of	ADP
fcis-30111	78	33	the	the	DET
fcis-30111	78	34	fusion	fusion	NOUN
fcis-30111	78	35	model	model	NOUN
fcis-30111	78	36	employs	employ	VERB
fcis-30111	78	37	a	a	DET
fcis-30111	78	38	number	number	NOUN
fcis-30111	78	39	of	of	ADP
fcis-30111	78	40	technical	technical	ADJ
fcis-30111	78	41	means	mean	NOUN
fcis-30111	78	42	to	to	PART
fcis-30111	78	43	improve	improve	VERB
fcis-30111	78	44	the	the	DET
fcis-30111	78	45	system	system	NOUN
fcis-30111	78	46	performance	performance	NOUN
fcis-30111	78	47	.	.	PUNCT
fcis-30111	79	1	the	the	DET
fcis-30111	79	2	evaluation	evaluation	NOUN
fcis-30111	79	3	of	of	ADP
fcis-30111	79	4	the	the	DET
fcis-30111	79	5	diagnostic	diagnostic	ADJ
fcis-30111	79	6	capability	capability	NOUN
fcis-30111	79	7	of	of	ADP
fcis-30111	79	8	the	the	DET
fcis-30111	79	9	fusion	fusion	NOUN
fcis-30111	79	10	model	model	NOUN
fcis-30111	79	11	for	for	ADP
fcis-30111	79	12	different	different	ADJ
fcis-30111	79	13	wheelset	wheelset	NOUN
fcis-30111	79	14	fault	fault	NOUN
fcis-30111	79	15	types	type	NOUN
fcis-30111	79	16	is	be	AUX
fcis-30111	79	17	shown	show	VERB
fcis-30111	79	18	in	in	ADP
fcis-30111	79	19	table	table	NOUN
fcis-30111	79	20	2	2	NUM
fcis-30111	79	21	,	,	PUNCT
fcis-30111	79	22	which	which	PRON
fcis-30111	79	23	demonstrates	demonstrate	VERB
fcis-30111	79	24	the	the	DET
fcis-30111	79	25	performance	performance	NOUN
fcis-30111	79	26	of	of	ADP
fcis-30111	79	27	the	the	DET
fcis-30111	79	28	fusion	fusion	NOUN
fcis-30111	79	29	model	model	NOUN
fcis-30111	79	30	on	on	ADP
fcis-30111	79	31	632	632	NUM
fcis-30111	79	32	sets	set	NOUN
fcis-30111	79	33	of	of	ADP
fcis-30111	79	34	test	test	NOUN
fcis-30111	79	35	data	datum	NOUN
fcis-30111	79	36	collected	collect	VERB
fcis-30111	79	37	from	from	ADP
fcis-30111	79	38	a	a	DET
fcis-30111	79	39	high	high	ADJ
fcis-30111	79	40	-	-	PUNCT
fcis-30111	79	41	speed	speed	NOUN
fcis-30111	79	42	railway	railway	NOUN
fcis-30111	79	43	site	site	NOUN
fcis-30111	79	44	.	.	PUNCT
fcis-30111	80	1	the	the	DET
fcis-30111	80	2	average	average	ADJ
fcis-30111	80	3	accuracy	accuracy	NOUN
fcis-30111	80	4	of	of	ADP
fcis-30111	80	5	the	the	DET
fcis-30111	80	6	model	model	NOUN
fcis-30111	80	7	is	be	AUX
fcis-30111	80	8	improved	improve	VERB
fcis-30111	80	9	by	by	ADP
fcis-30111	80	10	4.7	4.7	NUM
fcis-30111	80	11	%	%	NOUN
fcis-30111	80	12	by	by	ADP
fcis-30111	80	13	adjusting	adjust	VERB
fcis-30111	80	14	the	the	DET
fcis-30111	80	15	model	model	NOUN
fcis-30111	80	16	hyperparameters	hyperparameter	NOUN
fcis-30111	80	17	,	,	PUNCT
fcis-30111	80	18	including	include	VERB
fcis-30111	80	19	the	the	DET
fcis-30111	80	20	learning	learning	NOUN
fcis-30111	80	21	rate	rate	NOUN
fcis-30111	80	22	(	(	PUNCT
fcis-30111	80	23	optimal	optimal	ADJ
fcis-30111	80	24	value	value	NOUN
fcis-30111	80	25	of	of	ADP
fcis-30111	80	26	0.0023	0.0023	NUM
fcis-30111	80	27	)	)	PUNCT
fcis-30111	80	28	,	,	PUNCT
fcis-30111	80	29	the	the	DET
fcis-30111	80	30	batch	batch	NOUN
fcis-30111	80	31	size	size	NOUN
fcis-30111	80	32	(	(	PUNCT
fcis-30111	80	33	optimal	optimal	ADJ
fcis-30111	80	34	value	value	NOUN
fcis-30111	80	35	of	of	ADP
fcis-30111	80	36	64	64	NUM
fcis-30111	80	37	)	)	PUNCT
fcis-30111	80	38	and	and	CCONJ
fcis-30111	80	39	the	the	DET
fcis-30111	80	40	number	number	NOUN
fcis-30111	80	41	of	of	ADP
fcis-30111	80	42	attention	attention	NOUN
fcis-30111	80	43	heads	head	NOUN
fcis-30111	80	44	(	(	PUNCT
fcis-30111	80	45	optimal	optimal	ADJ
fcis-30111	80	46	value	value	NOUN
fcis-30111	80	47	of	of	ADP
fcis-30111	80	48	8)	8)	NUM
fcis-30111	80	49	through	through	ADP
fcis-30111	80	50	the	the	DET
fcis-30111	80	51	bayesian	bayesian	NOUN
fcis-30111	80	52	optimisation	optimisation	NOUN
fcis-30111	80	53	algorithm	algorithm	NOUN
fcis-30111	80	54	.	.	PUNCT
fcis-30111	81	1	to	to	PART
fcis-30111	81	2	improve	improve	VERB
fcis-30111	81	3	system	system	NOUN
fcis-30111	81	4	robustness	robustness	NOUN
fcis-30111	81	5	,	,	PUNCT
fcis-30111	81	6	noise	noise	NOUN
fcis-30111	81	7	enhancement	enhancement	NOUN
fcis-30111	81	8	and	and	CCONJ
fcis-30111	81	9	adversarial	adversarial	ADJ
fcis-30111	81	10	training	training	NOUN
fcis-30111	81	11	techniques	technique	NOUN
fcis-30111	81	12	are	be	AUX
fcis-30111	81	13	used	use	VERB
fcis-30111	81	14	to	to	PART
fcis-30111	81	15	increase	increase	VERB
fcis-30111	81	16	the	the	DET
fcis-30111	81	17	recognition	recognition	NOUN
fcis-30111	81	18	rate	rate	NOUN
fcis-30111	81	19	of	of	ADP
fcis-30111	81	20	the	the	DET
fcis-30111	81	21	model	model	NOUN
fcis-30111	81	22	in	in	ADP
fcis-30111	81	23	harsh	harsh	ADJ
fcis-30111	81	24	environments	environment	NOUN
fcis-30111	81	25	(	(	PUNCT
fcis-30111	81	26	signalto	signalto	NOUN
fcis-30111	81	27	-	-	PUNCT
fcis-30111	81	28	noise	noise	NOUN
fcis-30111	81	29	ratio	ratio	NOUN
fcis-30111	81	30	-10db	-10db	PROPN
fcis-30111	81	31	)	)	PUNCT
fcis-30111	81	32	from	from	ADP
fcis-30111	81	33	the	the	DET
fcis-30111	81	34	original	original	ADJ
fcis-30111	81	35	65.3	65.3	NUM
fcis-30111	81	36	%	%	NOUN
fcis-30111	81	37	to	to	ADP
fcis-30111	81	38	83.7	83.7	NUM
fcis-30111	81	39	%	%	NOUN
fcis-30111	81	40	.	.	PUNCT
fcis-30111	82	1	for	for	ADP
fcis-30111	82	2	edge	edge	NOUN
fcis-30111	82	3	computing	compute	VERB
fcis-30111	82	4	deployment	deployment	NOUN
fcis-30111	82	5	requirements	requirement	NOUN
fcis-30111	82	6	,	,	PUNCT
fcis-30111	82	7	knowledge	knowledge	NOUN
fcis-30111	82	8	distillation	distillation	NOUN
fcis-30111	82	9	and	and	CCONJ
fcis-30111	82	10	quantitative	quantitative	ADJ
fcis-30111	82	11	compression	compression	NOUN
fcis-30111	82	12	techniques	technique	NOUN
fcis-30111	82	13	are	be	AUX
fcis-30111	82	14	used	use	VERB
fcis-30111	82	15	to	to	PART
fcis-30111	82	16	compress	compress	VERB
fcis-30111	82	17	the	the	DET
fcis-30111	82	18	model	model	NOUN
fcis-30111	82	19	size	size	NOUN
fcis-30111	82	20	from	from	ADP
fcis-30111	82	21	the	the	DET
fcis-30111	82	22	original	original	ADJ
fcis-30111	82	23	187	187	NUM
fcis-30111	82	24	mb	mb	NOUN
fcis-30111	82	25	to	to	ADP
fcis-30111	82	26	24.5	24.5	NUM
fcis-30111	82	27	mb	mb	NOUN
fcis-30111	82	28	,	,	PUNCT
fcis-30111	82	29	increasing	increase	VERB
fcis-30111	82	30	the	the	DET
fcis-30111	82	31	inference	inference	NOUN
fcis-30111	82	32	speed	speed	NOUN
fcis-30111	82	33	by	by	ADP
fcis-30111	82	34	2.8	2.8	NUM
fcis-30111	82	35	times	time	NOUN
fcis-30111	82	36	,	,	PUNCT
fcis-30111	82	37	while	while	SCONJ
fcis-30111	82	38	the	the	DET
fcis-30111	82	39	accuracy	accuracy	NOUN
fcis-30111	82	40	rate	rate	NOUN
fcis-30111	82	41	decreases	decrease	VERB
fcis-30111	82	42	by	by	ADP
fcis-30111	82	43	only	only	ADV
fcis-30111	82	44	1.3	1.3	NUM
fcis-30111	82	45	percentage	percentage	NOUN
fcis-30111	82	46	points	point	NOUN
fcis-30111	82	47	.	.	PUNCT
fcis-30111	83	1	the	the	DET
fcis-30111	83	2	system	system	NOUN
fcis-30111	83	3	evaluation	evaluation	NOUN
fcis-30111	83	4	adopts	adopt	VERB
fcis-30111	83	5	quantitative	quantitative	ADJ
fcis-30111	83	6	indicators	indicator	NOUN
fcis-30111	83	7	combined	combine	VERB
fcis-30111	83	8	with	with	ADP
fcis-30111	83	9	expert	expert	NOUN
fcis-30111	83	10	scoring	scoring	NOUN
fcis-30111	83	11	mechanism	mechanism	NOUN
fcis-30111	83	12	,	,	PUNCT
fcis-30111	83	13	and	and	CCONJ
fcis-30111	83	14	in	in	ADP
fcis-30111	83	15	addition	addition	NOUN
fcis-30111	83	16	to	to	ADP
fcis-30111	83	17	conventional	conventional	ADJ
fcis-30111	83	18	performance	performance	NOUN
fcis-30111	83	19	indicators	indicator	NOUN
fcis-30111	83	20	,	,	PUNCT
fcis-30111	83	21	it	it	PRON
fcis-30111	83	22	also	also	ADV
fcis-30111	83	23	comprehensively	comprehensively	ADV
fcis-30111	83	24	considers	consider	VERB
fcis-30111	83	25	engineering	engineering	NOUN
fcis-30111	83	26	practicality	practicality	NOUN
fcis-30111	83	27	indicators	indicator	NOUN
fcis-30111	83	28	such	such	ADJ
fcis-30111	83	29	as	as	ADP
fcis-30111	83	30	warning	warning	NOUN
fcis-30111	83	31	lead	lead	NOUN
fcis-30111	83	32	time	time	NOUN
fcis-30111	83	33	and	and	CCONJ
fcis-30111	83	34	decision	decision	NOUN
fcis-30111	83	35	reliability	reliability	NOUN
fcis-30111	83	36	.	.	PUNCT
fcis-30111	84	1	the	the	DET
fcis-30111	84	2	fusion	fusion	NOUN
fcis-30111	84	3	model	model	NOUN
fcis-30111	84	4	achieves	achieve	VERB
fcis-30111	84	5	a	a	DET
fcis-30111	84	6	comprehensive	comprehensive	ADJ
fcis-30111	84	7	score	score	NOUN
fcis-30111	84	8	of	of	ADP
fcis-30111	84	9	94.3	94.3	NUM
fcis-30111	84	10	points	point	NOUN
fcis-30111	84	11	(	(	PUNCT
fcis-30111	84	12	out	out	ADP
fcis-30111	84	13	of	of	ADP
fcis-30111	84	14	100	100	NUM
fcis-30111	84	15	points	point	NOUN
fcis-30111	84	16	)	)	PUNCT
fcis-30111	84	17	in	in	ADP
fcis-30111	84	18	the	the	DET
fcis-30111	84	19	comparison	comparison	NOUN
fcis-30111	84	20	test	test	NOUN
fcis-30111	84	21	of	of	ADP
fcis-30111	84	22	high	high	ADJ
fcis-30111	84	23	-	-	PUNCT
fcis-30111	84	24	speed	speed	NOUN
fcis-30111	84	25	rail	rail	NOUN
fcis-30111	84	26	wheelset	wheelset	VERB
fcis-30111	84	27	fault	fault	NOUN
fcis-30111	84	28	detection	detection	NOUN
fcis-30111	84	29	in	in	ADP
fcis-30111	84	30	2024	2024	NUM
fcis-30111	84	31	,	,	PUNCT
fcis-30111	84	32	which	which	PRON
fcis-30111	84	33	is	be	AUX
fcis-30111	84	34	15.7	15.7	NUM
fcis-30111	84	35	points	point	NOUN
fcis-30111	84	36	higher	high	ADJ
fcis-30111	84	37	than	than	ADP
fcis-30111	84	38	that	that	PRON
fcis-30111	84	39	of	of	ADP
fcis-30111	84	40	the	the	DET
fcis-30111	84	41	traditional	traditional	ADJ
fcis-30111	84	42	method	method	NOUN
fcis-30111	84	43	,	,	PUNCT
fcis-30111	84	44	verifying	verify	VERB
fcis-30111	84	45	the	the	DET
fcis-30111	84	46	practical	practical	ADJ
fcis-30111	84	47	application	application	NOUN
fcis-30111	84	48	value	value	NOUN
fcis-30111	84	49	of	of	ADP
fcis-30111	84	50	the	the	DET
fcis-30111	84	51	fusion	fusion	NOUN
fcis-30111	84	52	architecture	architecture	NOUN
fcis-30111	84	53	.	.	PUNCT
fcis-30111	85	1	table	table	NOUN
fcis-30111	85	2	2	2	NUM
fcis-30111	85	3	.	.	PUNCT
fcis-30111	86	1	diagnostic	diagnostic	ADJ
fcis-30111	86	2	performance	performance	NOUN
fcis-30111	86	3	of	of	ADP
fcis-30111	86	4	the	the	DET
fcis-30111	86	5	multimodal	multimodal	NOUN
fcis-30111	86	6	fusion	fusion	NOUN
fcis-30111	86	7	model	model	NOUN
fcis-30111	86	8	on	on	ADP
fcis-30111	86	9	different	different	ADJ
fcis-30111	86	10	types	type	NOUN
fcis-30111	86	11	of	of	ADP
fcis-30111	86	12	wheelset	wheelset	ADJ
fcis-30111	86	13	faults	fault	NOUN
fcis-30111	86	14	fault	fault	VERB
fcis-30111	86	15	type	type	NOUN
fcis-30111	86	16	sample	sample	NOUN
fcis-30111	86	17	size	size	NOUN
fcis-30111	86	18	accuracy	accuracy	NOUN
fcis-30111	86	19	(	(	PUNCT
fcis-30111	86	20	%	%	INTJ
fcis-30111	86	21	)	)	PUNCT
fcis-30111	86	22	precision	precision	NOUN
fcis-30111	86	23	(	(	PUNCT
fcis-30111	86	24	%	%	INTJ
fcis-30111	86	25	)	)	PUNCT
fcis-30111	86	26	recall	recall	NOUN
fcis-30111	86	27	(	(	PUNCT
fcis-30111	86	28	%	%	NOUN
fcis-30111	86	29	)	)	PUNCT
fcis-30111	86	30	f1	f1	NOUN
fcis-30111	86	31	score	score	NOUN
fcis-30111	86	32	lead	lead	NOUN
fcis-30111	86	33	time	time	NOUN
fcis-30111	86	34	(	(	PUNCT
fcis-30111	86	35	days	day	NOUN
fcis-30111	86	36	)	)	PUNCT
fcis-30111	86	37	computation	computation	NOUN
fcis-30111	86	38	time	time	NOUN
fcis-30111	86	39	(	(	PUNCT
fcis-30111	86	40	ms	ms	NOUN
fcis-30111	86	41	)	)	PUNCT
fcis-30111	86	42	rim	rim	NOUN
fcis-30111	86	43	crack	crack	VERB
fcis-30111	86	44	142	142	NUM
fcis-30111	86	45	96.5	96.5	NUM
fcis-30111	86	46	95.8	95.8	NUM
fcis-30111	86	47	94.7	94.7	NUM
fcis-30111	86	48	0.952	0.952	NUM
fcis-30111	86	49	3.8	3.8	NUM
fcis-30111	86	50	76	76	NUM
fcis-30111	86	51	tread	tread	NOUN
fcis-30111	86	52	wear	wear	VERB
fcis-30111	86	53	186	186	NUM
fcis-30111	86	54	97.3	97.3	NUM
fcis-30111	86	55	98.1	98.1	NUM
fcis-30111	86	56	96.5	96.5	NUM
fcis-30111	86	57	0.973	0.973	NUM
fcis-30111	86	58	4.2	4.2	NUM
fcis-30111	86	59	82	82	NUM
fcis-30111	86	60	bearing	bearing	NOUN
fcis-30111	86	61	damage	damage	NOUN
fcis-30111	86	62	157	157	NUM
fcis-30111	86	63	95.2	95.2	NUM
fcis-30111	86	64	94.3	94.3	NUM
fcis-30111	86	65	95.8	95.8	NUM
fcis-30111	86	66	0.95	0.95	NUM
fcis-30111	86	67	3.1	3.1	NUM
fcis-30111	86	68	90	90	NUM
fcis-30111	86	69	axle	axle	NOUN
fcis-30111	86	70	bending	bend	VERB
fcis-30111	86	71	85	85	NUM
fcis-30111	86	72	93.8	93.8	NUM
fcis-30111	86	73	92.5	92.5	NUM
fcis-30111	86	74	94.1	94.1	NUM
fcis-30111	86	75	0.933	0.933	NUM
fcis-30111	86	76	2.7	2.7	NUM
fcis-30111	86	77	85	85	NUM
fcis-30111	86	78	early	early	ADJ
fcis-30111	86	79	composite	composite	ADJ
fcis-30111	86	80	fault	fault	NOUN
fcis-30111	86	81	62	62	NUM
fcis-30111	86	82	91.2	91.2	NUM
fcis-30111	86	83	89.7	89.7	NUM
fcis-30111	86	84	91.5	91.5	NUM
fcis-30111	86	85	0.906	0.906	NUM
fcis-30111	86	86	2.5	2.5	NUM
fcis-30111	86	87	95	95	NUM
fcis-30111	86	88	5	5	NUM
fcis-30111	86	89	.	.	PUNCT
fcis-30111	86	90	multimodal	multimodal	NOUN
fcis-30111	86	91	data	datum	NOUN
fcis-30111	86	92	fusion	fusion	NOUN
fcis-30111	86	93	system	system	NOUN
fcis-30111	86	94	implementation	implementation	NOUN
fcis-30111	86	95	and	and	CCONJ
fcis-30111	86	96	experimental	experimental	ADJ
fcis-30111	86	97	verification	verification	NOUN
fcis-30111	86	98	5.1	5.1	NUM
fcis-30111	86	99	.	.	PUNCT
fcis-30111	87	1	system	system	NOUN
fcis-30111	87	2	architecture	architecture	NOUN
fcis-30111	87	3	design	design	VERB
fcis-30111	87	4	the	the	DET
fcis-30111	87	5	overall	overall	ADJ
fcis-30111	87	6	architecture	architecture	NOUN
fcis-30111	87	7	of	of	ADP
fcis-30111	87	8	the	the	DET
fcis-30111	87	9	system	system	NOUN
fcis-30111	87	10	adopts	adopt	VERB
fcis-30111	87	11	a	a	DET
fcis-30111	87	12	layered	layered	ADJ
fcis-30111	87	13	design	design	NOUN
fcis-30111	87	14	idea	idea	NOUN
fcis-30111	87	15	to	to	PART
fcis-30111	87	16	construct	construct	VERB
fcis-30111	87	17	a	a	DET
fcis-30111	87	18	complete	complete	ADJ
fcis-30111	87	19	health	health	NOUN
fcis-30111	87	20	monitoring	monitoring	NOUN
fcis-30111	87	21	platform	platform	NOUN
fcis-30111	87	22	for	for	ADP
fcis-30111	87	23	the	the	DET
fcis-30111	87	24	wheelset	wheelset	NOUN
fcis-30111	87	25	.	.	PUNCT
fcis-30111	88	1	the	the	DET
fcis-30111	88	2	architecture	architecture	NOUN
fcis-30111	88	3	contains	contain	VERB
fcis-30111	88	4	four	four	NUM
fcis-30111	88	5	parts	part	NOUN
fcis-30111	88	6	:	:	PUNCT
fcis-30111	88	7	sensing	sense	VERB
fcis-30111	88	8	layer	layer	NOUN
fcis-30111	88	9	,	,	PUNCT
fcis-30111	88	10	transmission	transmission	NOUN
fcis-30111	88	11	layer	layer	NOUN
fcis-30111	88	12	,	,	PUNCT
fcis-30111	88	13	processing	processing	NOUN
fcis-30111	88	14	layer	layer	NOUN
fcis-30111	88	15	and	and	CCONJ
fcis-30111	88	16	application	application	NOUN
fcis-30111	88	17	layer[10	layer[10	NOUN
fcis-30111	88	18	]	]	PUNCT
fcis-30111	88	19	.	.	PUNCT
fcis-30111	89	1	the	the	DET
fcis-30111	89	2	sensing	sense	VERB
fcis-30111	89	3	layer	layer	NOUN
fcis-30111	89	4	deploys	deploy	VERB
fcis-30111	89	5	128	128	NUM
fcis-30111	89	6	vibration	vibration	NOUN
fcis-30111	89	7	sensors	sensor	NOUN
fcis-30111	89	8	,	,	PUNCT
fcis-30111	89	9	64	64	NUM
fcis-30111	89	10	acoustic	acoustic	ADJ
fcis-30111	89	11	sensors	sensor	NOUN
fcis-30111	89	12	,	,	PUNCT
fcis-30111	89	13	32	32	NUM
fcis-30111	89	14	temperature	temperature	NOUN
fcis-30111	89	15	sensors	sensor	NOUN
fcis-30111	89	16	and	and	CCONJ
fcis-30111	89	17	16	16	NUM
fcis-30111	89	18	high	high	ADJ
fcis-30111	89	19	-	-	PUNCT
fcis-30111	89	20	speed	speed	NOUN
fcis-30111	89	21	cameras	camera	NOUN
fcis-30111	89	22	,	,	PUNCT
fcis-30111	89	23	covering	cover	VERB
fcis-30111	89	24	12	12	NUM
fcis-30111	89	25	key	key	ADJ
fcis-30111	89	26	monitoring	monitoring	NOUN
fcis-30111	89	27	points	point	NOUN
fcis-30111	89	28	of	of	ADP
fcis-30111	89	29	wuhan	wuhan	PROPN
fcis-30111	89	30	-	-	PUNCT
fcis-30111	89	31	guangzhou	guangzhou	PROPN
fcis-30111	89	32	high	high	ADJ
fcis-30111	89	33	-	-	PUNCT
fcis-30111	89	34	speed	speed	NOUN
fcis-30111	89	35	railway	railway	NOUN
fcis-30111	89	36	.	.	PUNCT
fcis-30111	90	1	the	the	DET
fcis-30111	90	2	transmission	transmission	NOUN
fcis-30111	90	3	layer	layer	NOUN
fcis-30111	90	4	adopts	adopt	VERB
fcis-30111	90	5	a	a	DET
fcis-30111	90	6	combination	combination	NOUN
fcis-30111	90	7	of	of	ADP
fcis-30111	90	8	5	5	NUM
fcis-30111	90	9	g	g	NOUN
fcis-30111	90	10	network	network	NOUN
fcis-30111	90	11	and	and	CCONJ
fcis-30111	90	12	industrial	industrial	ADJ
fcis-30111	90	13	ethernet	ethernet	NOUN
fcis-30111	90	14	,	,	PUNCT
fcis-30111	90	15	with	with	ADP
fcis-30111	90	16	a	a	DET
fcis-30111	90	17	data	data	NOUN
fcis-30111	90	18	transmission	transmission	NOUN
fcis-30111	90	19	rate	rate	NOUN
fcis-30111	90	20	of	of	ADP
fcis-30111	90	21	48mbps	48mbps	NUM
fcis-30111	90	22	and	and	CCONJ
fcis-30111	90	23	a	a	DET
fcis-30111	90	24	latency	latency	NOUN
fcis-30111	90	25	of	of	ADP
fcis-30111	90	26	less	less	ADJ
fcis-30111	90	27	than	than	ADP
fcis-30111	90	28	15ms	15ms	ADJ
fcis-30111	90	29	.	.	PUNCT
fcis-30111	91	1	the	the	DET
fcis-30111	91	2	processing	processing	NOUN
fcis-30111	91	3	layer	layer	NOUN
fcis-30111	91	4	is	be	AUX
fcis-30111	91	5	configured	configure	VERB
fcis-30111	91	6	with	with	ADP
fcis-30111	91	7	8	8	NUM
fcis-30111	91	8	edge	edge	NOUN
fcis-30111	91	9	servers	server	NOUN
fcis-30111	91	10	(	(	PUNCT
fcis-30111	91	11	each	each	PRON
fcis-30111	91	12	equipped	equip	VERB
fcis-30111	91	13	with	with	ADP
fcis-30111	91	14	gpu	gpu	PROPN
fcis-30111	91	15	:	:	PUNCT
fcis-30111	91	16	tesla	tesla	PROPN
fcis-30111	91	17	t4	t4	PROPN
fcis-30111	91	18	,	,	PUNCT
fcis-30111	91	19	cpu	cpu	PROPN
fcis-30111	91	20	:	:	PUNCT
fcis-30111	91	21	intel	intel	PROPN
fcis-30111	91	22	xeon	xeon	PROPN
fcis-30111	91	23	6248r	6248r	NUM
fcis-30111	91	24	,	,	PUNCT
fcis-30111	91	25	ram:128	ram:128	NOUN
fcis-30111	91	26	gb	gb	NOUN
fcis-30111	91	27	)	)	PUNCT
fcis-30111	91	28	and	and	CCONJ
fcis-30111	91	29	a	a	DET
fcis-30111	91	30	central	central	ADJ
fcis-30111	91	31	cloud	cloud	NOUN
fcis-30111	91	32	platform	platform	NOUN
fcis-30111	91	33	(	(	PUNCT
fcis-30111	91	34	48	48	NUM
fcis-30111	91	35	-	-	PUNCT
fcis-30111	91	36	core	core	NOUN
fcis-30111	91	37	processor	processor	NOUN
fcis-30111	91	38	,	,	PUNCT
fcis-30111	91	39	384	384	NUM
fcis-30111	91	40	gb	gb	ADP
fcis-30111	91	41	ram	ram	NOUN
fcis-30111	91	42	.	.	PUNCT
fcis-30111	92	1	12	12	NUM
fcis-30111	92	2	tb	tb	ADP
fcis-30111	92	3	storage	storage	NOUN
fcis-30111	92	4	)	)	PUNCT
fcis-30111	92	5	,	,	PUNCT
fcis-30111	92	6	forming	form	VERB
fcis-30111	92	7	an	an	DET
fcis-30111	92	8	‘	'	PUNCT
fcis-30111	92	9	edge	edge	ADJ
fcis-30111	92	10	-	-	PUNCT
fcis-30111	92	11	cloud	cloud	NOUN
fcis-30111	92	12	’	'	PUNCT
fcis-30111	92	13	collaborative	collaborative	ADJ
fcis-30111	92	14	computing	computing	NOUN
fcis-30111	92	15	architecture	architecture	NOUN
fcis-30111	92	16	.	.	PUNCT
fcis-30111	93	1	the	the	DET
fcis-30111	93	2	application	application	NOUN
fcis-30111	93	3	layer	layer	NOUN
fcis-30111	93	4	provides	provide	VERB
fcis-30111	93	5	wheel	wheel	NOUN
fcis-30111	93	6	-	-	PUNCT
fcis-30111	93	7	to	to	ADP
fcis-30111	93	8	-	-	PUNCT
fcis-30111	93	9	wheel	wheel	NOUN
fcis-30111	93	10	health	health	NOUN
fcis-30111	93	11	status	status	NOUN
fcis-30111	93	12	assessment	assessment	NOUN
fcis-30111	93	13	,	,	PUNCT
fcis-30111	93	14	fault	fault	VERB
fcis-30111	93	15	diagnosis	diagnosis	NOUN
fcis-30111	93	16	and	and	CCONJ
fcis-30111	93	17	early	early	ADJ
fcis-30111	93	18	warning	warning	NOUN
fcis-30111	93	19	,	,	PUNCT
fcis-30111	93	20	and	and	CCONJ
fcis-30111	93	21	remote	remote	ADJ
fcis-30111	93	22	monitoring	monitoring	NOUN
fcis-30111	93	23	and	and	CCONJ
fcis-30111	93	24	management	management	NOUN
fcis-30111	93	25	through	through	ADP
fcis-30111	93	26	web	web	NOUN
fcis-30111	93	27	interface	interface	NOUN
fcis-30111	93	28	and	and	CCONJ
fcis-30111	93	29	mobile	mobile	ADJ
fcis-30111	93	30	application	application	NOUN
fcis-30111	93	31	.	.	PUNCT
fcis-30111	94	1	with	with	ADP
fcis-30111	94	2	a	a	DET
fcis-30111	94	3	daily	daily	ADJ
fcis-30111	94	4	data	data	NOUN
fcis-30111	94	5	volume	volume	NOUN
fcis-30111	94	6	of	of	ADP
fcis-30111	94	7	3.6	3.6	NUM
fcis-30111	94	8	tb	tb	NOUN
fcis-30111	94	9	,	,	PUNCT
fcis-30111	94	10	the	the	DET
fcis-30111	94	11	system	system	NOUN
fcis-30111	94	12	supports	support	VERB
fcis-30111	94	13	real	real	ADJ
fcis-30111	94	14	-	-	PUNCT
fcis-30111	94	15	time	time	NOUN
fcis-30111	94	16	identification	identification	NOUN
fcis-30111	94	17	of	of	ADP
fcis-30111	94	18	8	8	NUM
fcis-30111	94	19	key	key	ADJ
fcis-30111	94	20	fault	fault	NOUN
fcis-30111	94	21	types	type	NOUN
fcis-30111	94	22	,	,	PUNCT
fcis-30111	94	23	with	with	ADP
fcis-30111	94	24	an	an	DET
fcis-30111	94	25	average	average	ADJ
fcis-30111	94	26	identification	identification	NOUN
fcis-30111	94	27	delay	delay	NOUN
fcis-30111	94	28	of	of	ADP
fcis-30111	94	29	less	less	ADJ
fcis-30111	94	30	than	than	ADP
fcis-30111	94	31	150ms	150ms	PROPN
fcis-30111	94	32	.	.	PUNCT
fcis-30111	95	1	the	the	DET
fcis-30111	95	2	architecture	architecture	NOUN
fcis-30111	95	3	has	have	AUX
fcis-30111	95	4	been	be	AUX
fcis-30111	95	5	in	in	ADP
fcis-30111	95	6	continuous	continuous	ADJ
fcis-30111	95	7	operation	operation	NOUN
fcis-30111	95	8	for	for	ADP
fcis-30111	95	9	15	15	NUM
fcis-30111	95	10	months	month	NOUN
fcis-30111	95	11	after	after	ADP
fcis-30111	95	12	the	the	DET
fcis-30111	95	13	completion	completion	NOUN
fcis-30111	95	14	of	of	ADP
fcis-30111	95	15	its	its	PRON
fcis-30111	95	16	deployment	deployment	NOUN
fcis-30111	95	17	in	in	ADP
fcis-30111	95	18	2023	2023	NUM
fcis-30111	95	19	,	,	PUNCT
fcis-30111	95	20	with	with	ADP
fcis-30111	95	21	a	a	DET
fcis-30111	95	22	system	system	NOUN
fcis-30111	95	23	availability	availability	NOUN
fcis-30111	95	24	of	of	ADP
fcis-30111	95	25	99.7	99.7	NUM
fcis-30111	95	26	per	per	ADP
fcis-30111	95	27	cent	cent	NOUN
fcis-30111	95	28	and	and	CCONJ
fcis-30111	95	29	a	a	DET
fcis-30111	95	30	fault	fault	NOUN
fcis-30111	95	31	detection	detection	NOUN
fcis-30111	95	32	rate	rate	NOUN
fcis-30111	95	33	that	that	PRON
fcis-30111	95	34	is	be	AUX
fcis-30111	95	35	18.3	18.3	NUM
fcis-30111	95	36	per	per	ADP
fcis-30111	95	37	cent	cent	NOUN
fcis-30111	95	38	higher	high	ADJ
fcis-30111	95	39	than	than	ADP
fcis-30111	95	40	that	that	PRON
fcis-30111	95	41	of	of	ADP
fcis-30111	95	42	the	the	DET
fcis-30111	95	43	traditional	traditional	ADJ
fcis-30111	95	44	single	single	ADJ
fcis-30111	95	45	-	-	PUNCT
fcis-30111	95	46	mode	mode	NOUN
fcis-30111	95	47	system	system	NOUN
fcis-30111	95	48	.	.	PUNCT
fcis-30111	96	1	5.2	5.2	NUM
fcis-30111	96	2	.	.	PUNCT
fcis-30111	96	3	data	datum	NOUN
fcis-30111	96	4	fusion	fusion	NOUN
fcis-30111	96	5	model	model	NOUN
fcis-30111	96	6	implementation	implementation	NOUN
fcis-30111	96	7	the	the	DET
fcis-30111	96	8	data	data	NOUN
fcis-30111	96	9	fusion	fusion	NOUN
fcis-30111	96	10	model	model	NOUN
fcis-30111	96	11	is	be	AUX
fcis-30111	96	12	implemented	implement	VERB
fcis-30111	96	13	based	base	VERB
fcis-30111	96	14	on	on	ADP
fcis-30111	96	15	the	the	DET
fcis-30111	96	16	pytorch	pytorch	NOUN
fcis-30111	96	17	deep	deep	ADJ
fcis-30111	96	18	learning	learning	NOUN
fcis-30111	96	19	framework	framework	NOUN
fcis-30111	96	20	,	,	PUNCT
fcis-30111	96	21	which	which	PRON
fcis-30111	96	22	uses	use	VERB
fcis-30111	96	23	a	a	DET
fcis-30111	96	24	multi	multi	ADJ
fcis-30111	96	25	-	-	ADJ
fcis-30111	96	26	stream	stream	ADJ
fcis-30111	96	27	73	73	NUM
fcis-30111	96	28	network	network	NOUN
fcis-30111	96	29	structure	structure	NOUN
fcis-30111	96	30	to	to	PART
fcis-30111	96	31	process	process	VERB
fcis-30111	96	32	different	different	ADJ
fcis-30111	96	33	modal	modal	ADJ
fcis-30111	96	34	data	datum	NOUN
fcis-30111	96	35	.	.	PUNCT
fcis-30111	97	1	each	each	DET
fcis-30111	97	2	subnetwork	subnetwork	NOUN
fcis-30111	97	3	is	be	AUX
fcis-30111	97	4	optimally	optimally	ADV
fcis-30111	97	5	designed	design	VERB
fcis-30111	97	6	for	for	ADP
fcis-30111	97	7	specific	specific	ADJ
fcis-30111	97	8	modal	modal	ADJ
fcis-30111	97	9	signals	signal	NOUN
fcis-30111	97	10	,	,	PUNCT
fcis-30111	97	11	and	and	CCONJ
fcis-30111	97	12	finally	finally	ADV
fcis-30111	97	13	the	the	DET
fcis-30111	97	14	deep	deep	ADJ
fcis-30111	97	15	fusion	fusion	NOUN
fcis-30111	97	16	is	be	AUX
fcis-30111	97	17	achieved	achieve	VERB
fcis-30111	97	18	through	through	ADP
fcis-30111	97	19	the	the	DET
fcis-30111	97	20	attention	attention	NOUN
fcis-30111	97	21	mechanism	mechanism	NOUN
fcis-30111	97	22	.	.	PUNCT
fcis-30111	98	1	in	in	ADP
fcis-30111	98	2	the	the	DET
fcis-30111	98	3	realisation	realisation	NOUN
fcis-30111	98	4	process	process	NOUN
fcis-30111	98	5	,	,	PUNCT
fcis-30111	98	6	the	the	DET
fcis-30111	98	7	training	training	NOUN
fcis-30111	98	8	data	datum	NOUN
fcis-30111	98	9	comes	come	VERB
fcis-30111	98	10	from	from	ADP
fcis-30111	98	11	the	the	DET
fcis-30111	98	12	accumulated	accumulate	VERB
fcis-30111	98	13	wheel	wheel	NOUN
fcis-30111	98	14	-	-	PUNCT
fcis-30111	98	15	pair	pair	NOUN
fcis-30111	98	16	monitoring	monitoring	NOUN
fcis-30111	98	17	records	record	NOUN
fcis-30111	98	18	of	of	ADP
fcis-30111	98	19	beijing	beijing	PROPN
fcis-30111	98	20	-	-	PUNCT
fcis-30111	98	21	guangzhou	guangzhou	PROPN
fcis-30111	98	22	high	high	ADJ
fcis-30111	98	23	-	-	PUNCT
fcis-30111	98	24	speed	speed	NOUN
fcis-30111	98	25	railway	railway	NOUN
fcis-30111	98	26	for	for	ADP
fcis-30111	98	27	5	5	NUM
fcis-30111	98	28	years	year	NOUN
fcis-30111	98	29	,	,	PUNCT
fcis-30111	98	30	which	which	PRON
fcis-30111	98	31	contains	contain	VERB
fcis-30111	98	32	14,875	14,875	NUM
fcis-30111	98	33	sets	set	NOUN
fcis-30111	98	34	of	of	ADP
fcis-30111	98	35	normal	normal	ADJ
fcis-30111	98	36	samples	sample	NOUN
fcis-30111	98	37	and	and	CCONJ
fcis-30111	98	38	5,632	5,632	NUM
fcis-30111	98	39	sets	set	NOUN
fcis-30111	98	40	of	of	ADP
fcis-30111	98	41	fault	fault	NOUN
fcis-30111	98	42	samples	sample	NOUN
fcis-30111	98	43	(	(	PUNCT
fcis-30111	98	44	covering	cover	VERB
fcis-30111	98	45	8	8	NUM
fcis-30111	98	46	typical	typical	ADJ
fcis-30111	98	47	fault	fault	NOUN
fcis-30111	98	48	types	type	NOUN
fcis-30111	98	49	)	)	PUNCT
fcis-30111	98	50	.	.	PUNCT
fcis-30111	99	1	the	the	DET
fcis-30111	99	2	model	model	NOUN
fcis-30111	99	3	training	training	NOUN
fcis-30111	99	4	adopts	adopt	VERB
fcis-30111	99	5	a	a	DET
fcis-30111	99	6	mixed	mixed	ADJ
fcis-30111	99	7	-	-	PUNCT
fcis-30111	99	8	accuracy	accuracy	NOUN
fcis-30111	99	9	training	training	NOUN
fcis-30111	99	10	strategy	strategy	NOUN
fcis-30111	99	11	,	,	PUNCT
fcis-30111	99	12	and	and	CCONJ
fcis-30111	99	13	a	a	DET
fcis-30111	99	14	single	single	ADJ
fcis-30111	99	15	training	training	NOUN
fcis-30111	99	16	cycle	cycle	NOUN
fcis-30111	99	17	takes	take	VERB
fcis-30111	99	18	about	about	ADV
fcis-30111	99	19	18	18	NUM
fcis-30111	99	20	hours	hour	NOUN
fcis-30111	99	21	.	.	PUNCT
fcis-30111	100	1	in	in	ADP
fcis-30111	100	2	order	order	NOUN
fcis-30111	100	3	to	to	PART
fcis-30111	100	4	improve	improve	VERB
fcis-30111	100	5	the	the	DET
fcis-30111	100	6	generalisation	generalisation	NOUN
fcis-30111	100	7	ability	ability	NOUN
fcis-30111	100	8	of	of	ADP
fcis-30111	100	9	the	the	DET
fcis-30111	100	10	model	model	NOUN
fcis-30111	100	11	,	,	PUNCT
fcis-30111	100	12	data	datum	NOUN
fcis-30111	100	13	enhancement	enhancement	NOUN
fcis-30111	100	14	techniques	technique	NOUN
fcis-30111	100	15	are	be	AUX
fcis-30111	100	16	introduced	introduce	VERB
fcis-30111	100	17	,	,	PUNCT
fcis-30111	100	18	including	include	VERB
fcis-30111	100	19	adding	add	VERB
fcis-30111	100	20	gaussian	gaussian	ADJ
fcis-30111	100	21	noise	noise	NOUN
fcis-30111	100	22	to	to	ADP
fcis-30111	100	23	the	the	DET
fcis-30111	100	24	vibration	vibration	NOUN
fcis-30111	100	25	signals	signal	NOUN
fcis-30111	100	26	(	(	PUNCT
fcis-30111	100	27	signal	signal	NOUN
fcis-30111	100	28	-	-	PUNCT
fcis-30111	100	29	to	to	ADP
fcis-30111	100	30	-	-	PUNCT
fcis-30111	100	31	noise	noise	NOUN
fcis-30111	100	32	ratio	ratio	NOUN
fcis-30111	100	33	range	range	NOUN
fcis-30111	100	34	of	of	ADP
fcis-30111	100	35	15db~10db	15db~10db	NUM
fcis-30111	100	36	)	)	PUNCT
fcis-30111	100	37	,	,	PUNCT
fcis-30111	100	38	random	random	ADJ
fcis-30111	100	39	time	time	NOUN
fcis-30111	100	40	offset	offset	NOUN
fcis-30111	100	41	(	(	PUNCT
fcis-30111	100	42	±50ms	±50ms	PROPN
fcis-30111	100	43	)	)	PUNCT
fcis-30111	100	44	,	,	PUNCT
fcis-30111	100	45	and	and	CCONJ
fcis-30111	100	46	random	random	ADJ
fcis-30111	100	47	scaling	scaling	NOUN
fcis-30111	100	48	of	of	ADP
fcis-30111	100	49	the	the	DET
fcis-30111	100	50	temperature	temperature	NOUN
fcis-30111	100	51	data	datum	NOUN
fcis-30111	100	52	(	(	PUNCT
fcis-30111	100	53	±5	±5	NOUN
fcis-30111	100	54	%	%	NOUN
fcis-30111	100	55	)	)	PUNCT
fcis-30111	100	56	,	,	PUNCT
fcis-30111	100	57	etc	etc	X
fcis-30111	100	58	.	.	X
fcis-30111	100	59	,	,	PUNCT
fcis-30111	100	60	so	so	SCONJ
fcis-30111	100	61	that	that	SCONJ
fcis-30111	100	62	the	the	DET
fcis-30111	100	63	sample	sample	NOUN
fcis-30111	100	64	size	size	NOUN
fcis-30111	100	65	of	of	ADP
fcis-30111	100	66	the	the	DET
fcis-30111	100	67	enhancement	enhancement	NOUN
fcis-30111	100	68	is	be	AUX
fcis-30111	100	69	enlarged	enlarge	VERB
fcis-30111	100	70	to	to	ADP
fcis-30111	100	71	three	three	NUM
fcis-30111	100	72	times	time	NOUN
fcis-30111	100	73	of	of	ADP
fcis-30111	100	74	the	the	DET
fcis-30111	100	75	original	original	ADJ
fcis-30111	100	76	data	datum	NOUN
fcis-30111	100	77	.	.	PUNCT
fcis-30111	101	1	the	the	DET
fcis-30111	101	2	model	model	NOUN
fcis-30111	101	3	parameters	parameter	NOUN
fcis-30111	101	4	are	be	AUX
fcis-30111	101	5	optimised	optimise	VERB
fcis-30111	101	6	using	use	VERB
fcis-30111	101	7	adam	adam	PROPN
fcis-30111	101	8	optimiser	optimiser	PROPN
fcis-30111	101	9	with	with	ADP
fcis-30111	101	10	an	an	DET
fcis-30111	101	11	initial	initial	ADJ
fcis-30111	101	12	learning	learning	NOUN
fcis-30111	101	13	rate	rate	NOUN
fcis-30111	101	14	of	of	ADP
fcis-30111	101	15	0.0018	0.0018	NUM
fcis-30111	101	16	and	and	CCONJ
fcis-30111	101	17	dynamically	dynamically	ADV
fcis-30111	101	18	adjusted	adjust	VERB
fcis-30111	101	19	using	use	VERB
fcis-30111	101	20	cosine	cosine	NOUN
fcis-30111	101	21	annealing	anneal	VERB
fcis-30111	101	22	strategy	strategy	NOUN
fcis-30111	101	23	.	.	PUNCT
fcis-30111	102	1	when	when	SCONJ
fcis-30111	102	2	deployed	deploy	VERB
fcis-30111	102	3	at	at	ADP
fcis-30111	102	4	the	the	DET
fcis-30111	102	5	edge	edge	NOUN
fcis-30111	102	6	end	end	NOUN
fcis-30111	102	7	,	,	PUNCT
fcis-30111	102	8	through	through	ADP
fcis-30111	102	9	onnx	onnx	NOUN
fcis-30111	102	10	format	format	NOUN
fcis-30111	102	11	conversion	conversion	NOUN
fcis-30111	102	12	and	and	CCONJ
fcis-30111	102	13	tensorrt	tensorrt	ADJ
fcis-30111	102	14	acceleration	acceleration	NOUN
fcis-30111	102	15	,	,	PUNCT
fcis-30111	102	16	the	the	DET
fcis-30111	102	17	inference	inference	NOUN
fcis-30111	102	18	speed	speed	NOUN
fcis-30111	102	19	is	be	AUX
fcis-30111	102	20	increased	increase	VERB
fcis-30111	102	21	by	by	ADP
fcis-30111	102	22	3.2	3.2	NUM
fcis-30111	102	23	times	time	NOUN
fcis-30111	102	24	compared	compare	VERB
fcis-30111	102	25	with	with	ADP
fcis-30111	102	26	the	the	DET
fcis-30111	102	27	original	original	ADJ
fcis-30111	102	28	pytorch	pytorch	NOUN
fcis-30111	102	29	model	model	NOUN
fcis-30111	102	30	,	,	PUNCT
fcis-30111	102	31	and	and	CCONJ
fcis-30111	102	32	the	the	DET
fcis-30111	102	33	memory	memory	NOUN
fcis-30111	102	34	occupation	occupation	NOUN
fcis-30111	102	35	of	of	ADP
fcis-30111	102	36	a	a	DET
fcis-30111	102	37	single	single	ADJ
fcis-30111	102	38	inference	inference	NOUN
fcis-30111	102	39	is	be	AUX
fcis-30111	102	40	reduced	reduce	VERB
fcis-30111	102	41	from	from	ADP
fcis-30111	102	42	2.1	2.1	NUM
fcis-30111	102	43	gb	gb	NOUN
fcis-30111	102	44	to	to	ADP
fcis-30111	102	45	645	645	NUM
fcis-30111	102	46	mb	mb	NOUN
fcis-30111	102	47	,	,	PUNCT
fcis-30111	102	48	which	which	PRON
fcis-30111	102	49	meets	meet	VERB
fcis-30111	102	50	the	the	DET
fcis-30111	102	51	limitations	limitation	NOUN
fcis-30111	102	52	of	of	ADP
fcis-30111	102	53	hardware	hardware	NOUN
fcis-30111	102	54	resources	resource	NOUN
fcis-30111	102	55	in	in	ADP
fcis-30111	102	56	the	the	DET
fcis-30111	102	57	field	field	NOUN
fcis-30111	102	58	.	.	PUNCT
fcis-30111	103	1	through	through	ADP
fcis-30111	103	2	quantisation	quantisation	NOUN
fcis-30111	103	3	-	-	PUNCT
fcis-30111	103	4	aware	aware	ADJ
fcis-30111	103	5	training	training	NOUN
fcis-30111	103	6	,	,	PUNCT
fcis-30111	103	7	the	the	DET
fcis-30111	103	8	accuracy	accuracy	NOUN
fcis-30111	103	9	of	of	ADP
fcis-30111	103	10	the	the	DET
fcis-30111	103	11	int8	int8	NOUN
fcis-30111	103	12	quantisation	quantisation	NOUN
fcis-30111	103	13	model	model	NOUN
fcis-30111	103	14	only	only	ADV
fcis-30111	103	15	decreases	decrease	VERB
fcis-30111	103	16	by	by	ADP
fcis-30111	103	17	0.8	0.8	NUM
fcis-30111	103	18	%	%	NOUN
fcis-30111	103	19	,	,	PUNCT
fcis-30111	103	20	while	while	SCONJ
fcis-30111	103	21	the	the	DET
fcis-30111	103	22	inference	inference	NOUN
fcis-30111	103	23	speed	speed	NOUN
fcis-30111	103	24	is	be	AUX
fcis-30111	103	25	improved	improve	VERB
fcis-30111	103	26	by	by	ADP
fcis-30111	103	27	2.7	2.7	NUM
fcis-30111	103	28	times	time	NOUN
fcis-30111	103	29	,	,	PUNCT
fcis-30111	103	30	as	as	SCONJ
fcis-30111	103	31	shown	show	VERB
fcis-30111	103	32	in	in	ADP
fcis-30111	103	33	figure	figure	NOUN
fcis-30111	103	34	5	5	NUM
fcis-30111	103	35	.	.	PUNCT
fcis-30111	103	36	figure	figure	NOUN
fcis-30111	103	37	5	5	NUM
fcis-30111	103	38	.	.	PUNCT
fcis-30111	103	39	comparison	comparison	NOUN
fcis-30111	103	40	of	of	ADP
fcis-30111	103	41	int8	int8	NOUN
fcis-30111	103	42	quantisation	quantisation	NOUN
fcis-30111	103	43	performance	performance	NOUN
fcis-30111	103	44	5.3	5.3	NUM
fcis-30111	103	45	.	.	PUNCT
fcis-30111	104	1	experimental	experimental	ADJ
fcis-30111	104	2	verification	verification	NOUN
fcis-30111	104	3	and	and	CCONJ
fcis-30111	104	4	performance	performance	NOUN
fcis-30111	104	5	evaluation	evaluation	NOUN
fcis-30111	104	6	the	the	DET
fcis-30111	104	7	experimental	experimental	ADJ
fcis-30111	104	8	validation	validation	NOUN
fcis-30111	104	9	uses	use	VERB
fcis-30111	104	10	a	a	DET
fcis-30111	104	11	combination	combination	NOUN
fcis-30111	104	12	of	of	ADP
fcis-30111	104	13	actual	actual	ADJ
fcis-30111	104	14	line	line	NOUN
fcis-30111	104	15	tests	test	NOUN
fcis-30111	104	16	and	and	CCONJ
fcis-30111	104	17	simulation	simulation	NOUN
fcis-30111	104	18	tests	test	NOUN
fcis-30111	104	19	to	to	PART
fcis-30111	104	20	evaluate	evaluate	VERB
fcis-30111	104	21	the	the	DET
fcis-30111	104	22	system	system	NOUN
fcis-30111	104	23	performance	performance	NOUN
fcis-30111	104	24	.	.	PUNCT
fcis-30111	105	1	figure	figure	NOUN
fcis-30111	105	2	6	6	NUM
fcis-30111	105	3	demonstrates	demonstrate	VERB
fcis-30111	105	4	a	a	DET
fcis-30111	105	5	comparison	comparison	NOUN
fcis-30111	105	6	of	of	ADP
fcis-30111	105	7	the	the	DET
fcis-30111	105	8	recognition	recognition	NOUN
fcis-30111	105	9	performance	performance	NOUN
fcis-30111	105	10	of	of	ADP
fcis-30111	105	11	the	the	DET
fcis-30111	105	12	fusion	fusion	NOUN
fcis-30111	105	13	model	model	NOUN
fcis-30111	105	14	for	for	ADP
fcis-30111	105	15	four	four	NUM
fcis-30111	105	16	common	common	ADJ
fcis-30111	105	17	wheelset	wheelset	ADJ
fcis-30111	105	18	faults	fault	NOUN
fcis-30111	105	19	under	under	ADP
fcis-30111	105	20	different	different	ADJ
fcis-30111	105	21	scenarios	scenario	NOUN
fcis-30111	105	22	.	.	PUNCT
fcis-30111	106	1	the	the	DET
fcis-30111	106	2	experimental	experimental	ADJ
fcis-30111	106	3	data	datum	NOUN
fcis-30111	106	4	comes	come	VERB
fcis-30111	106	5	from	from	ADP
fcis-30111	106	6	the	the	DET
fcis-30111	106	7	online	online	ADJ
fcis-30111	106	8	monitoring	monitoring	NOUN
fcis-30111	106	9	system	system	NOUN
fcis-30111	106	10	of	of	ADP
fcis-30111	106	11	wheels	wheel	NOUN
fcis-30111	106	12	in	in	ADP
fcis-30111	106	13	a	a	DET
fcis-30111	106	14	southern	southern	ADJ
fcis-30111	106	15	section	section	NOUN
fcis-30111	106	16	and	and	CCONJ
fcis-30111	106	17	contains	contain	VERB
fcis-30111	106	18	the	the	DET
fcis-30111	106	19	operation	operation	NOUN
fcis-30111	106	20	data	datum	NOUN
fcis-30111	106	21	of	of	ADP
fcis-30111	106	22	three	three	NUM
fcis-30111	106	23	designated	designate	VERB
fcis-30111	106	24	rolling	rolling	ADJ
fcis-30111	106	25	stock	stock	NOUN
fcis-30111	106	26	units	unit	NOUN
fcis-30111	106	27	throughout	throughout	ADP
fcis-30111	106	28	the	the	DET
fcis-30111	106	29	year	year	NOUN
fcis-30111	106	30	2024	2024	NUM
fcis-30111	106	31	.	.	PUNCT
fcis-30111	107	1	the	the	DET
fcis-30111	107	2	test	test	NOUN
fcis-30111	107	3	results	result	NOUN
fcis-30111	107	4	show	show	VERB
fcis-30111	107	5	that	that	SCONJ
fcis-30111	107	6	the	the	DET
fcis-30111	107	7	multimodal	multimodal	ADJ
fcis-30111	107	8	fusion	fusion	NOUN
fcis-30111	107	9	method	method	NOUN
fcis-30111	107	10	outperforms	outperform	VERB
fcis-30111	107	11	the	the	DET
fcis-30111	107	12	unimodal	unimodal	ADJ
fcis-30111	107	13	method	method	NOUN
fcis-30111	107	14	in	in	ADP
fcis-30111	107	15	all	all	DET
fcis-30111	107	16	types	type	NOUN
fcis-30111	107	17	of	of	ADP
fcis-30111	107	18	fault	fault	NOUN
fcis-30111	107	19	identification	identification	NOUN
fcis-30111	107	20	,	,	PUNCT
fcis-30111	107	21	especially	especially	ADV
fcis-30111	107	22	in	in	ADP
fcis-30111	107	23	the	the	DET
fcis-30111	107	24	harsh	harsh	ADJ
fcis-30111	107	25	environment	environment	NOUN
fcis-30111	107	26	(	(	PUNCT
fcis-30111	107	27	rain	rain	NOUN
fcis-30111	107	28	,	,	PUNCT
fcis-30111	107	29	snow	snow	NOUN
fcis-30111	107	30	,	,	PUNCT
fcis-30111	107	31	tunnel	tunnel	NOUN
fcis-30111	107	32	passage	passage	NOUN
fcis-30111	107	33	)	)	PUNCT
fcis-30111	107	34	conditions	condition	NOUN
fcis-30111	107	35	,	,	PUNCT
fcis-30111	107	36	where	where	SCONJ
fcis-30111	107	37	it	it	PRON
fcis-30111	107	38	exhibits	exhibit	VERB
fcis-30111	107	39	stronger	strong	ADJ
fcis-30111	107	40	robustness	robustness	NOUN
fcis-30111	107	41	.	.	PUNCT
fcis-30111	108	1	during	during	ADP
fcis-30111	108	2	the	the	DET
fcis-30111	108	3	test	test	NOUN
fcis-30111	108	4	,	,	PUNCT
fcis-30111	108	5	the	the	DET
fcis-30111	108	6	fusion	fusion	NOUN
fcis-30111	108	7	model	model	NOUN
fcis-30111	108	8	was	be	AUX
fcis-30111	108	9	found	find	VERB
fcis-30111	108	10	to	to	PART
fcis-30111	108	11	have	have	VERB
fcis-30111	108	12	outstanding	outstanding	ADJ
fcis-30111	108	13	performance	performance	NOUN
fcis-30111	108	14	in	in	ADP
fcis-30111	108	15	early	early	ADJ
fcis-30111	108	16	bearing	bear	VERB
fcis-30111	108	17	fault	fault	NOUN
fcis-30111	108	18	detection	detection	NOUN
fcis-30111	108	19	,	,	PUNCT
fcis-30111	108	20	sending	send	VERB
fcis-30111	108	21	out	out	ADP
fcis-30111	108	22	warning	warning	NOUN
fcis-30111	108	23	signals	signal	NOUN
fcis-30111	108	24	2.8	2.8	NUM
fcis-30111	108	25	days	day	NOUN
fcis-30111	108	26	earlier	early	ADV
fcis-30111	108	27	than	than	ADP
fcis-30111	108	28	the	the	DET
fcis-30111	108	29	temperature	temperature	NOUN
fcis-30111	108	30	single	single	ADJ
fcis-30111	108	31	-	-	PUNCT
fcis-30111	108	32	modal	modal	NOUN
fcis-30111	108	33	method	method	NOUN
fcis-30111	108	34	,	,	PUNCT
fcis-30111	108	35	which	which	PRON
fcis-30111	108	36	gained	gain	VERB
fcis-30111	108	37	sufficient	sufficient	ADJ
fcis-30111	108	38	preparation	preparation	NOUN
fcis-30111	108	39	time	time	NOUN
fcis-30111	108	40	for	for	ADP
fcis-30111	108	41	the	the	DET
fcis-30111	108	42	maintenance	maintenance	NOUN
fcis-30111	108	43	department	department	NOUN
fcis-30111	108	44	.	.	PUNCT
fcis-30111	109	1	the	the	DET
fcis-30111	109	2	performance	performance	NOUN
fcis-30111	109	3	test	test	NOUN
fcis-30111	109	4	shows	show	VERB
fcis-30111	109	5	that	that	SCONJ
fcis-30111	109	6	the	the	DET
fcis-30111	109	7	average	average	ADJ
fcis-30111	109	8	response	response	NOUN
fcis-30111	109	9	time	time	NOUN
fcis-30111	109	10	of	of	ADP
fcis-30111	109	11	the	the	DET
fcis-30111	109	12	system	system	NOUN
fcis-30111	109	13	is	be	AUX
fcis-30111	109	14	142ms	142ms	ADJ
fcis-30111	109	15	,	,	PUNCT
fcis-30111	109	16	and	and	CCONJ
fcis-30111	109	17	the	the	DET
fcis-30111	109	18	peak	peak	NOUN
fcis-30111	109	19	processing	processing	NOUN
fcis-30111	109	20	capacity	capacity	NOUN
fcis-30111	109	21	reaches	reach	VERB
fcis-30111	109	22	95	95	NUM
fcis-30111	109	23	requests	request	NOUN
fcis-30111	109	24	per	per	ADP
fcis-30111	109	25	second	second	NOUN
fcis-30111	109	26	,	,	PUNCT
fcis-30111	109	27	which	which	PRON
fcis-30111	109	28	meets	meet	VERB
fcis-30111	109	29	the	the	DET
fcis-30111	109	30	real	real	ADJ
fcis-30111	109	31	-	-	PUNCT
fcis-30111	109	32	time	time	NOUN
fcis-30111	109	33	monitoring	monitoring	NOUN
fcis-30111	109	34	requirements	requirement	NOUN
fcis-30111	109	35	.	.	PUNCT
fcis-30111	110	1	the	the	DET
fcis-30111	110	2	trade	trade	NOUN
fcis-30111	110	3	-	-	PUNCT
fcis-30111	110	4	off	off	ADP
fcis-30111	110	5	analysis	analysis	NOUN
fcis-30111	110	6	of	of	ADP
fcis-30111	110	7	the	the	DET
fcis-30111	110	8	leakage	leakage	NOUN
fcis-30111	110	9	rate	rate	NOUN
fcis-30111	110	10	and	and	CCONJ
fcis-30111	110	11	false	false	ADJ
fcis-30111	110	12	alarm	alarm	NOUN
fcis-30111	110	13	rate	rate	NOUN
fcis-30111	110	14	shows	show	VERB
fcis-30111	110	15	that	that	SCONJ
fcis-30111	110	16	the	the	DET
fcis-30111	110	17	decision	decision	NOUN
fcis-30111	110	18	threshold	threshold	NOUN
fcis-30111	110	19	set	set	VERB
fcis-30111	110	20	at	at	ADP
fcis-30111	110	21	0.78	0.78	NUM
fcis-30111	110	22	after	after	ADP
fcis-30111	110	23	optimisation	optimisation	NOUN
fcis-30111	110	24	of	of	ADP
fcis-30111	110	25	the	the	DET
fcis-30111	110	26	roc	roc	PROPN
fcis-30111	110	27	curve	curve	NOUN
fcis-30111	110	28	can	can	AUX
fcis-30111	110	29	keep	keep	VERB
fcis-30111	110	30	the	the	DET
fcis-30111	110	31	false	false	ADJ
fcis-30111	110	32	alarm	alarm	NOUN
fcis-30111	110	33	rate	rate	NOUN
fcis-30111	110	34	within	within	ADP
fcis-30111	110	35	2.3	2.3	NUM
fcis-30111	110	36	%	%	NOUN
fcis-30111	110	37	while	while	SCONJ
fcis-30111	110	38	maintaining	maintain	VERB
fcis-30111	110	39	a	a	DET
fcis-30111	110	40	detection	detection	NOUN
fcis-30111	110	41	rate	rate	NOUN
fcis-30111	110	42	of	of	ADP
fcis-30111	110	43	96.5	96.5	NUM
fcis-30111	110	44	%	%	NOUN
fcis-30111	110	45	.	.	PUNCT
fcis-30111	111	1	in	in	ADP
fcis-30111	111	2	the	the	DET
fcis-30111	111	3	system	system	NOUN
fcis-30111	111	4	stability	stability	NOUN
fcis-30111	111	5	test	test	NOUN
fcis-30111	111	6	,	,	PUNCT
fcis-30111	111	7	the	the	DET
fcis-30111	111	8	detection	detection	NOUN
fcis-30111	111	9	performance	performance	NOUN
fcis-30111	111	10	is	be	AUX
fcis-30111	111	11	not	not	PART
fcis-30111	111	12	significantly	significantly	ADV
fcis-30111	111	13	degraded	degrade	VERB
fcis-30111	111	14	under	under	ADP
fcis-30111	111	15	30	30	NUM
fcis-30111	111	16	days	day	NOUN
fcis-30111	111	17	of	of	ADP
fcis-30111	111	18	continuous	continuous	ADJ
fcis-30111	111	19	operation	operation	NOUN
fcis-30111	111	20	without	without	ADP
fcis-30111	111	21	reboot	reboot	NOUN
fcis-30111	111	22	,	,	PUNCT
fcis-30111	111	23	and	and	CCONJ
fcis-30111	111	24	the	the	DET
fcis-30111	111	25	processing	processing	NOUN
fcis-30111	111	26	delay	delay	NOUN
fcis-30111	111	27	increases	increase	VERB
fcis-30111	111	28	by	by	ADP
fcis-30111	111	29	no	no	DET
fcis-30111	111	30	more	more	ADJ
fcis-30111	111	31	than	than	ADP
fcis-30111	111	32	8	8	NUM
fcis-30111	111	33	%	%	NOUN
fcis-30111	111	34	.	.	PUNCT
fcis-30111	112	1	5.4	5.4	NUM
fcis-30111	112	2	.	.	PUNCT
fcis-30111	112	3	engineering	engineering	NOUN
fcis-30111	112	4	application	application	NOUN
fcis-30111	112	5	case	case	NOUN
fcis-30111	112	6	analysis	analysis	NOUN
fcis-30111	112	7	the	the	DET
fcis-30111	112	8	system	system	NOUN
fcis-30111	112	9	has	have	AUX
fcis-30111	112	10	been	be	AUX
fcis-30111	112	11	deployed	deploy	VERB
fcis-30111	112	12	and	and	CCONJ
fcis-30111	112	13	applied	apply	VERB
fcis-30111	112	14	in	in	ADP
fcis-30111	112	15	a	a	DET
fcis-30111	112	16	certain	certain	ADJ
fcis-30111	112	17	high	high	ADJ
fcis-30111	112	18	-	-	PUNCT
fcis-30111	112	19	speed	speed	NOUN
fcis-30111	112	20	railway	railway	NOUN
fcis-30111	112	21	from	from	ADP
fcis-30111	112	22	a	a	DET
fcis-30111	112	23	certain	certain	ADJ
fcis-30111	112	24	south	south	NOUN
fcis-30111	112	25	to	to	ADP
fcis-30111	112	26	a	a	DET
fcis-30111	112	27	certain	certain	ADJ
fcis-30111	112	28	north	north	NOUN
fcis-30111	112	29	section	section	NOUN
fcis-30111	112	30	for	for	ADP
fcis-30111	112	31	one	one	NUM
fcis-30111	112	32	year	year	NOUN
fcis-30111	112	33	,	,	PUNCT
fcis-30111	112	34	and	and	CCONJ
fcis-30111	112	35	has	have	AUX
fcis-30111	112	36	achieved	achieve	VERB
fcis-30111	112	37	significant	significant	ADJ
fcis-30111	112	38	engineering	engineering	NOUN
fcis-30111	112	39	benefits	benefit	NOUN
fcis-30111	112	40	.	.	PUNCT
fcis-30111	113	1	figure	figure	NOUN
fcis-30111	113	2	7	7	NUM
fcis-30111	113	3	demonstrates	demonstrate	VERB
fcis-30111	113	4	the	the	DET
fcis-30111	113	5	comparison	comparison	NOUN
fcis-30111	113	6	of	of	ADP
fcis-30111	113	7	wheelset	wheelset	PROPN
fcis-30111	113	8	maintenance	maintenance	NOUN
fcis-30111	113	9	data	datum	NOUN
fcis-30111	113	10	before	before	ADV
fcis-30111	113	11	and	and	CCONJ
fcis-30111	113	12	after	after	ADP
fcis-30111	113	13	the	the	DET
fcis-30111	113	14	deployment	deployment	NOUN
fcis-30111	113	15	of	of	ADP
fcis-30111	113	16	the	the	DET
fcis-30111	113	17	system	system	NOUN
fcis-30111	113	18	.	.	PUNCT
fcis-30111	114	1	through	through	ADP
fcis-30111	114	2	the	the	DET
fcis-30111	114	3	multimodal	multimodal	ADJ
fcis-30111	114	4	fusion	fusion	NOUN
fcis-30111	114	5	technology	technology	NOUN
fcis-30111	114	6	,	,	PUNCT
fcis-30111	114	7	the	the	DET
fcis-30111	114	8	number	number	NOUN
fcis-30111	114	9	of	of	ADP
fcis-30111	114	10	unscheduled	unscheduled	ADJ
fcis-30111	114	11	wheelset	wheelset	NOUN
fcis-30111	114	12	replacements	replacement	NOUN
fcis-30111	114	13	has	have	AUX
fcis-30111	114	14	been	be	AUX
fcis-30111	114	15	reduced	reduce	VERB
fcis-30111	114	16	from	from	ADP
fcis-30111	114	17	an	an	DET
fcis-30111	114	18	annual	annual	ADJ
fcis-30111	114	19	average	average	NOUN
fcis-30111	114	20	of	of	ADP
fcis-30111	114	21	42	42	NUM
fcis-30111	114	22	to	to	PART
fcis-30111	114	23	17	17	NUM
fcis-30111	114	24	,	,	PUNCT
fcis-30111	114	25	a	a	DET
fcis-30111	114	26	reduction	reduction	NOUN
fcis-30111	114	27	of	of	ADP
fcis-30111	114	28	59.5	59.5	NUM
fcis-30111	114	29	%	%	NOUN
fcis-30111	114	30	;	;	PUNCT
fcis-30111	114	31	the	the	DET
fcis-30111	114	32	number	number	NOUN
fcis-30111	114	33	of	of	ADP
fcis-30111	114	34	unplanned	unplanned	ADJ
fcis-30111	114	35	stopping	stopping	NOUN
fcis-30111	114	36	events	event	NOUN
fcis-30111	114	37	has	have	AUX
fcis-30111	114	38	been	be	AUX
fcis-30111	114	39	reduced	reduce	VERB
fcis-30111	114	40	from	from	ADP
fcis-30111	114	41	23	23	NUM
fcis-30111	114	42	to	to	PART
fcis-30111	114	43	8	8	NUM
fcis-30111	114	44	,	,	PUNCT
fcis-30111	114	45	a	a	DET
fcis-30111	114	46	reduction	reduction	NOUN
fcis-30111	114	47	of	of	ADP
fcis-30111	114	48	65.2	65.2	NUM
fcis-30111	114	49	%	%	NOUN
fcis-30111	114	50	;	;	PUNCT
fcis-30111	114	51	the	the	DET
fcis-30111	114	52	average	average	ADJ
fcis-30111	114	53	overhaul	overhaul	NOUN
fcis-30111	114	54	cycle	cycle	NOUN
fcis-30111	114	55	has	have	AUX
fcis-30111	114	56	been	be	AUX
fcis-30111	114	57	extended	extend	VERB
fcis-30111	114	58	from	from	ADP
fcis-30111	114	59	45	45	NUM
fcis-30111	114	60	to	to	PART
fcis-30111	114	61	68	68	NUM
fcis-30111	114	62	days	day	NOUN
fcis-30111	114	63	,	,	PUNCT
fcis-30111	114	64	an	an	DET
fcis-30111	114	65	increase	increase	NOUN
fcis-30111	114	66	of	of	ADP
fcis-30111	114	67	51.1	51.1	NUM
fcis-30111	114	68	%	%	NOUN
fcis-30111	114	69	;	;	PUNCT
fcis-30111	114	70	and	and	CCONJ
fcis-30111	114	71	the	the	DET
fcis-30111	114	72	annual	annual	ADJ
fcis-30111	114	73	maintenance	maintenance	NOUN
fcis-30111	114	74	cost	cost	NOUN
fcis-30111	114	75	has	have	AUX
fcis-30111	114	76	been	be	AUX
fcis-30111	114	77	reduced	reduce	VERB
fcis-30111	114	78	from	from	ADP
fcis-30111	114	79	rmb	rmb	PROPN
fcis-30111	114	80	8.67	8.67	NUM
fcis-30111	114	81	million	million	NUM
fcis-30111	114	82	to	to	PART
fcis-30111	114	83	rmb	rmb	VERB
fcis-30111	114	84	4.12	4.12	NUM
fcis-30111	114	85	million	million	NUM
fcis-30111	114	86	,	,	PUNCT
fcis-30111	114	87	a	a	DET
fcis-30111	114	88	saving	saving	NOUN
fcis-30111	114	89	of	of	ADP
fcis-30111	114	90	52.5	52.5	NUM
fcis-30111	114	91	%	%	NOUN
fcis-30111	114	92	.	.	PUNCT
fcis-30111	115	1	typical	typical	ADJ
fcis-30111	115	2	case	case	NOUN
fcis-30111	115	3	study	study	NOUN
fcis-30111	115	4	shows	show	VERB
fcis-30111	115	5	that	that	SCONJ
fcis-30111	115	6	the	the	DET
fcis-30111	115	7	system	system	NOUN
fcis-30111	115	8	successfully	successfully	ADV
fcis-30111	115	9	captured	capture	VERB
fcis-30111	115	10	the	the	DET
fcis-30111	115	11	early	early	ADJ
fcis-30111	115	12	abnormal	abnormal	ADJ
fcis-30111	115	13	vibration	vibration	NOUN
fcis-30111	115	14	signal	signal	NOUN
fcis-30111	115	15	of	of	ADP
fcis-30111	115	16	the	the	DET
fcis-30111	115	17	first	first	ADJ
fcis-30111	115	18	wheelset	wheelset	NOUN
fcis-30111	115	19	of	of	ADP
fcis-30111	115	20	the	the	DET
fcis-30111	115	21	front	front	ADJ
fcis-30111	115	22	bogie	bogie	NOUN
fcis-30111	115	23	of	of	ADP
fcis-30111	115	24	a	a	DET
fcis-30111	115	25	certain	certain	ADJ
fcis-30111	115	26	type	type	NOUN
fcis-30111	115	27	of	of	ADP
fcis-30111	115	28	train	train	NOUN
fcis-30111	115	29	set	set	VERB
fcis-30111	115	30	on	on	ADP
fcis-30111	115	31	a	a	DET
fcis-30111	115	32	certain	certain	ADJ
fcis-30111	115	33	day	day	NOUN
fcis-30111	115	34	of	of	ADP
fcis-30111	115	35	a	a	DET
fcis-30111	115	36	certain	certain	ADJ
fcis-30111	115	37	month	month	NOUN
fcis-30111	115	38	in	in	ADP
fcis-30111	115	39	2023	2023	NUM
fcis-30111	115	40	,	,	PUNCT
fcis-30111	115	41	and	and	CCONJ
fcis-30111	115	42	confirmed	confirm	VERB
fcis-30111	115	43	it	it	PRON
fcis-30111	115	44	to	to	PART
fcis-30111	115	45	be	be	AUX
fcis-30111	115	46	a	a	DET
fcis-30111	115	47	rim	rim	NOUN
fcis-30111	115	48	micro	micro	NOUN
fcis-30111	115	49	-	-	NOUN
fcis-30111	115	50	crack	crack	VERB
fcis-30111	115	51	through	through	ADP
fcis-30111	115	52	cross	cross	NOUN
fcis-30111	115	53	validation	validation	NOUN
fcis-30111	115	54	of	of	ADP
fcis-30111	115	55	multimodal	multimodal	NOUN
fcis-30111	115	56	data	datum	NOUN
fcis-30111	115	57	,	,	PUNCT
fcis-30111	115	58	providing	provide	VERB
fcis-30111	115	59	an	an	DET
fcis-30111	115	60	early	early	ADJ
fcis-30111	115	61	warning	warning	NOUN
fcis-30111	115	62	5	5	NUM
fcis-30111	115	63	days	day	NOUN
fcis-30111	115	64	in	in	ADP
fcis-30111	115	65	advance	advance	NOUN
fcis-30111	115	66	and	and	CCONJ
fcis-30111	115	67	avoiding	avoid	VERB
fcis-30111	115	68	a	a	DET
fcis-30111	115	69	certain	certain	ADJ
fcis-30111	115	70	possible	possible	ADJ
fcis-30111	115	71	emergency	emergency	NOUN
fcis-30111	115	72	stopping	stop	VERB
fcis-30111	115	73	event	event	NOUN
fcis-30111	115	74	.	.	PUNCT
fcis-30111	116	1	the	the	DET
fcis-30111	116	2	diagnostic	diagnostic	ADJ
fcis-30111	116	3	accuracy	accuracy	NOUN
fcis-30111	116	4	of	of	ADP
fcis-30111	116	5	the	the	DET
fcis-30111	116	6	system	system	NOUN
fcis-30111	116	7	reached	reach	VERB
fcis-30111	116	8	95.3	95.3	NUM
fcis-30111	116	9	%	%	NOUN
fcis-30111	116	10	,	,	PUNCT
fcis-30111	116	11	much	much	ADV
fcis-30111	116	12	higher	high	ADJ
fcis-30111	116	13	than	than	ADP
fcis-30111	116	14	the	the	DET
fcis-30111	116	15	78.4	78.4	NUM
fcis-30111	116	16	%	%	NOUN
fcis-30111	116	17	of	of	ADP
fcis-30111	116	18	the	the	DET
fcis-30111	116	19	traditional	traditional	ADJ
fcis-30111	116	20	single	single	ADJ
fcis-30111	116	21	acoustic	acoustic	ADJ
fcis-30111	116	22	method	method	NOUN
fcis-30111	116	23	and	and	CCONJ
fcis-30111	116	24	82.1	82.1	NUM
fcis-30111	116	25	%	%	NOUN
fcis-30111	116	26	of	of	ADP
fcis-30111	116	27	the	the	DET
fcis-30111	116	28	single	single	ADJ
fcis-30111	116	29	vibration	vibration	NOUN
fcis-30111	116	30	method	method	NOUN
fcis-30111	116	31	.	.	PUNCT
fcis-30111	117	1	after	after	SCONJ
fcis-30111	117	2	the	the	DET
fcis-30111	117	3	system	system	NOUN
fcis-30111	117	4	was	be	AUX
fcis-30111	117	5	put	put	VERB
fcis-30111	117	6	into	into	ADP
fcis-30111	117	7	use	use	NOUN
fcis-30111	117	8	,	,	PUNCT
fcis-30111	117	9	the	the	DET
fcis-30111	117	10	on	on	ADP
fcis-30111	117	11	-	-	PUNCT
fcis-30111	117	12	time	time	NOUN
fcis-30111	117	13	rate	rate	NOUN
fcis-30111	117	14	of	of	ADP
fcis-30111	117	15	trains	train	NOUN
fcis-30111	117	16	in	in	ADP
fcis-30111	117	17	this	this	DET
fcis-30111	117	18	section	section	NOUN
fcis-30111	117	19	increased	increase	VERB
fcis-30111	117	20	by	by	ADP
fcis-30111	117	21	2.8	2.8	NUM
fcis-30111	117	22	percentage	percentage	NOUN
fcis-30111	117	23	points	point	NOUN
fcis-30111	117	24	,	,	PUNCT
fcis-30111	117	25	the	the	DET
fcis-30111	117	26	efficiency	efficiency	NOUN
fcis-30111	117	27	of	of	ADP
fcis-30111	117	28	operation	operation	NOUN
fcis-30111	117	29	and	and	CCONJ
fcis-30111	117	30	maintenance	maintenance	NOUN
fcis-30111	117	31	personnel	personnel	NOUN
fcis-30111	117	32	increased	increase	VERB
fcis-30111	117	33	by	by	ADP
fcis-30111	117	34	about	about	ADV
fcis-30111	117	35	35	35	NUM
fcis-30111	117	36	%	%	NOUN
fcis-30111	117	37	,	,	PUNCT
fcis-30111	117	38	and	and	CCONJ
fcis-30111	117	39	the	the	DET
fcis-30111	117	40	workload	workload	NOUN
fcis-30111	117	41	of	of	ADP
fcis-30111	117	42	daily	daily	ADJ
fcis-30111	117	43	inspections	inspection	NOUN
fcis-30111	117	44	was	be	AUX
fcis-30111	117	45	reduced	reduce	VERB
fcis-30111	117	46	by	by	ADP
fcis-30111	117	47	about	about	ADV
fcis-30111	117	48	40	40	NUM
fcis-30111	117	49	%	%	NOUN
fcis-30111	117	50	,	,	PUNCT
fcis-30111	117	51	bringing	bring	VERB
fcis-30111	117	52	significant	significant	ADJ
fcis-30111	117	53	safety	safety	NOUN
fcis-30111	117	54	and	and	CCONJ
fcis-30111	117	55	economic	economic	ADJ
fcis-30111	117	56	benefits	benefit	NOUN
fcis-30111	117	57	to	to	ADP
fcis-30111	117	58	the	the	DET
fcis-30111	117	59	railway	railway	NOUN
fcis-30111	117	60	department	department	NOUN
fcis-30111	117	61	.	.	PUNCT
fcis-30111	118	1	6	6	X
fcis-30111	118	2	.	.	X
fcis-30111	118	3	conclusion	conclusion	NOUN
fcis-30111	118	4	in	in	ADP
fcis-30111	118	5	this	this	DET
fcis-30111	118	6	study	study	NOUN
fcis-30111	118	7	,	,	PUNCT
fcis-30111	118	8	a	a	DET
fcis-30111	118	9	multimodal	multimodal	ADJ
fcis-30111	118	10	data	data	NOUN
fcis-30111	118	11	fusion	fusion	NOUN
fcis-30111	118	12	system	system	NOUN
fcis-30111	118	13	for	for	ADP
fcis-30111	118	14	monitoring	monitor	VERB
fcis-30111	118	15	the	the	DET
fcis-30111	118	16	health	health	NOUN
fcis-30111	118	17	status	status	NOUN
fcis-30111	118	18	of	of	ADP
fcis-30111	118	19	railway	railway	NOUN
fcis-30111	118	20	vehicle	vehicle	NOUN
fcis-30111	118	21	wheelsets	wheelset	NOUN
fcis-30111	118	22	was	be	AUX
fcis-30111	118	23	constructed	construct	VERB
fcis-30111	118	24	,	,	PUNCT
fcis-30111	118	25	which	which	PRON
fcis-30111	118	26	realises	realise	VERB
fcis-30111	118	27	the	the	DET
fcis-30111	118	28	collaborative	collaborative	ADJ
fcis-30111	118	29	processing	processing	NOUN
fcis-30111	118	30	of	of	ADP
fcis-30111	118	31	multi	multi	ADJ
fcis-30111	118	32	-	-	ADJ
fcis-30111	118	33	source	source	NOUN
fcis-30111	118	34	information	information	NOUN
fcis-30111	118	35	such	such	ADJ
fcis-30111	118	36	as	as	ADP
fcis-30111	118	37	vibration	vibration	NOUN
fcis-30111	118	38	,	,	PUNCT
fcis-30111	118	39	acoustic	acoustic	ADJ
fcis-30111	118	40	,	,	PUNCT
fcis-30111	118	41	temperature	temperature	NOUN
fcis-30111	118	42	and	and	CCONJ
fcis-30111	118	43	vision	vision	NOUN
fcis-30111	118	44	.	.	PUNCT
fcis-30111	119	1	the	the	DET
fcis-30111	119	2	research	research	NOUN
fcis-30111	119	3	results	result	NOUN
fcis-30111	119	4	show	show	VERB
fcis-30111	119	5	that	that	SCONJ
fcis-30111	119	6	compared	compare	VERB
fcis-30111	119	7	with	with	ADP
fcis-30111	119	8	the	the	DET
fcis-30111	119	9	traditional	traditional	ADJ
fcis-30111	119	10	single	single	ADJ
fcis-30111	119	11	-	-	PUNCT
fcis-30111	119	12	modal	modal	NOUN
fcis-30111	119	13	approach	approach	NOUN
fcis-30111	119	14	,	,	PUNCT
fcis-30111	119	15	the	the	DET
fcis-30111	119	16	multimodal	multimodal	NOUN
fcis-30111	119	17	fusion	fusion	NOUN
fcis-30111	119	18	technique	technique	NOUN
fcis-30111	119	19	significantly	significantly	ADV
fcis-30111	119	20	improves	improve	VERB
fcis-30111	119	21	the	the	DET
fcis-30111	119	22	fault	fault	NOUN
fcis-30111	119	23	diagnosis	diagnosis	NOUN
fcis-30111	119	24	accuracy	accuracy	NOUN
fcis-30111	119	25	(	(	PUNCT
fcis-30111	119	26	by	by	ADP
fcis-30111	119	27	14.1	14.1	NUM
fcis-30111	119	28	%	%	NOUN
fcis-30111	119	29	)	)	PUNCT
fcis-30111	119	30	and	and	CCONJ
fcis-30111	119	31	warning	warn	VERB
fcis-30111	119	32	lead	lead	NOUN
fcis-30111	119	33	time	time	NOUN
fcis-30111	119	34	(	(	PUNCT
fcis-30111	119	35	by	by	ADP
fcis-30111	119	36	184.6	184.6	NUM
fcis-30111	119	37	%	%	NOUN
fcis-30111	119	38	)	)	PUNCT
fcis-30111	119	39	,	,	PUNCT
fcis-30111	119	40	and	and	CCONJ
fcis-30111	119	41	reduces	reduce	VERB
fcis-30111	119	42	the	the	DET
fcis-30111	119	43	number	number	NOUN
fcis-30111	119	44	of	of	ADP
fcis-30111	119	45	unscheduled	unscheduled	ADJ
fcis-30111	119	46	wheelset	wheelset	NOUN
fcis-30111	119	47	replacements	replacement	NOUN
fcis-30111	119	48	and	and	CCONJ
fcis-30111	119	49	the	the	DET
fcis-30111	119	50	maintenance	maintenance	NOUN
fcis-30111	119	51	cost	cost	NOUN
fcis-30111	119	52	(	(	PUNCT
fcis-30111	119	53	by	by	ADP
fcis-30111	119	54	59.5	59.5	NUM
fcis-30111	119	55	%	%	NOUN
fcis-30111	119	56	and	and	CCONJ
fcis-30111	119	57	52.5	52.5	NUM
fcis-30111	119	58	%	%	NOUN
fcis-30111	119	59	,	,	PUNCT
fcis-30111	119	60	respectively	respectively	ADV
fcis-30111	119	61	)	)	PUNCT
fcis-30111	119	62	in	in	ADP
fcis-30111	119	63	practical	practical	ADJ
fcis-30111	119	64	engineering	engineering	NOUN
fcis-30111	119	65	applications	application	NOUN
fcis-30111	119	66	.	.	PUNCT
fcis-30111	120	1	the	the	DET
fcis-30111	120	2	system	system	NOUN
fcis-30111	120	3	shows	show	VERB
fcis-30111	120	4	strong	strong	ADJ
fcis-30111	120	5	robustness	robustness	NOUN
fcis-30111	120	6	under	under	ADP
fcis-30111	120	7	harsh	harsh	ADJ
fcis-30111	120	8	environmental	environmental	ADJ
fcis-30111	120	9	conditions	condition	NOUN
fcis-30111	120	10	,	,	PUNCT
fcis-30111	120	11	providing	provide	VERB
fcis-30111	120	12	a	a	DET
fcis-30111	120	13	new	new	ADJ
fcis-30111	120	14	idea	idea	NOUN
fcis-30111	120	15	for	for	ADP
fcis-30111	120	16	digital	digital	ADJ
fcis-30111	120	17	monitoring	monitoring	NOUN
fcis-30111	120	18	of	of	ADP
fcis-30111	120	19	wheelset	wheelset	PROPN
fcis-30111	120	20	health	health	NOUN
fcis-30111	120	21	status	status	NOUN
fcis-30111	120	22	.	.	PUNCT
fcis-30111	121	1	future	future	ADJ
fcis-30111	121	2	research	research	NOUN
fcis-30111	121	3	will	will	AUX
fcis-30111	121	4	further	far	ADV
fcis-30111	121	5	explore	explore	VERB
fcis-30111	121	6	edge	edge	NOUN
fcis-30111	121	7	intelligence	intelligence	NOUN
fcis-30111	121	8	and	and	CCONJ
fcis-30111	121	9	adaptive	adaptive	ADJ
fcis-30111	121	10	learning	learning	NOUN
fcis-30111	121	11	techniques	technique	NOUN
fcis-30111	121	12	to	to	PART
fcis-30111	121	13	improve	improve	VERB
fcis-30111	121	14	the	the	DET
fcis-30111	121	15	system	system	NOUN
fcis-30111	121	16	's	's	PART
fcis-30111	121	17	adaptability	adaptability	NOUN
fcis-30111	121	18	in	in	ADP
fcis-30111	121	19	complex	complex	ADJ
fcis-30111	121	20	dynamic	dynamic	ADJ
fcis-30111	121	21	environments	environment	NOUN
fcis-30111	121	22	and	and	CCONJ
fcis-30111	121	23	extend	extend	VERB
fcis-30111	121	24	it	it	PRON
fcis-30111	121	25	to	to	ADP
fcis-30111	121	26	monitoring	monitor	VERB
fcis-30111	121	27	applications	application	NOUN
fcis-30111	121	28	for	for	ADP
fcis-30111	121	29	more	more	ADJ
fcis-30111	121	30	types	type	NOUN
fcis-30111	121	31	of	of	ADP
fcis-30111	121	32	railway	railway	NOUN
fcis-30111	121	33	vehicle	vehicle	NOUN
fcis-30111	121	34	components	component	NOUN
fcis-30111	121	35	.	.	PUNCT
fcis-30111	122	1	74	74	NUM
fcis-30111	122	2	(	(	PUNCT
fcis-30111	122	3	a)normal	a)normal	NOUN
fcis-30111	122	4	weather	weather	NOUN
fcis-30111	122	5	(	(	PUNCT
fcis-30111	122	6	b)rain	b)rain	NOUN
fcis-30111	122	7	/	/	SYM
fcis-30111	122	8	snow	snow	NOUN
fcis-30111	122	9	(	(	PUNCT
fcis-30111	122	10	c)tunnel	c)tunnel	NOUN
fcis-30111	122	11	passage	passage	NOUN
fcis-30111	122	12	(	(	PUNCT
fcis-30111	122	13	d)mixed	d)mixe	VERB
fcis-30111	122	14	environment	environment	NOUN
fcis-30111	122	15	figure	figure	NOUN
fcis-30111	122	16	6	6	NUM
fcis-30111	122	17	.	.	PUNCT
fcis-30111	122	18	comparison	comparison	NOUN
fcis-30111	122	19	of	of	ADP
fcis-30111	122	20	fault	fault	NOUN
fcis-30111	122	21	identification	identification	NOUN
fcis-30111	122	22	accuracy	accuracy	NOUN
fcis-30111	122	23	between	between	ADP
fcis-30111	122	24	multimodal	multimodal	NOUN
fcis-30111	122	25	fusion	fusion	NOUN
fcis-30111	122	26	model	model	NOUN
fcis-30111	122	27	and	and	CCONJ
fcis-30111	122	28	unimodal	unimodal	ADJ
fcis-30111	122	29	model	model	NOUN
fcis-30111	122	30	in	in	ADP
fcis-30111	122	31	different	different	ADJ
fcis-30111	122	32	cases	case	NOUN
fcis-30111	122	33	75	75	NUM
fcis-30111	122	34	figure	figure	NOUN
fcis-30111	122	35	7	7	NUM
fcis-30111	122	36	.	.	PUNCT
fcis-30111	122	37	comparison	comparison	NOUN
fcis-30111	122	38	of	of	ADP
fcis-30111	122	39	wheelset	wheelset	PROPN
fcis-30111	122	40	maintenance	maintenance	NOUN
fcis-30111	122	41	data	datum	NOUN
fcis-30111	122	42	before	before	ADV
fcis-30111	122	43	and	and	CCONJ
fcis-30111	122	44	after	after	ADP
fcis-30111	122	45	deployment	deployment	NOUN
fcis-30111	122	46	of	of	ADP
fcis-30111	122	47	the	the	DET
fcis-30111	122	48	multimodal	multimodal	NOUN
fcis-30111	122	49	fusion	fusion	NOUN
fcis-30111	122	50	system	system	NOUN
fcis-30111	122	51	references	reference	NOUN
fcis-30111	122	52	[	[	X
fcis-30111	122	53	1	1	NUM
fcis-30111	122	54	]	]	PUNCT
fcis-30111	122	55	lourenço	lourenço	VERB
fcis-30111	122	56	a	a	PRON
fcis-30111	122	57	,	,	PUNCT
fcis-30111	122	58	ribeiro	ribeiro	PROPN
fcis-30111	122	59	d	d	PROPN
fcis-30111	122	60	,	,	PUNCT
fcis-30111	122	61	fernandes	fernandes	PROPN
fcis-30111	122	62	m	m	PROPN
fcis-30111	122	63	,	,	PUNCT
fcis-30111	122	64	et	et	PROPN
fcis-30111	122	65	al	al	PROPN
fcis-30111	122	66	.	.	PROPN
fcis-30111	122	67	time	time	PROPN
fcis-30111	122	68	series	series	PROPN
fcis-30111	122	69	data	data	PROPN
fcis-30111	122	70	mining	mining	NOUN
fcis-30111	122	71	for	for	ADP
fcis-30111	122	72	railway	railway	NOUN
fcis-30111	122	73	wheel	wheel	NOUN
fcis-30111	122	74	and	and	CCONJ
fcis-30111	122	75	track	track	NOUN
fcis-30111	122	76	monitoring	monitoring	NOUN
fcis-30111	122	77	:	:	PUNCT
fcis-30111	122	78	a	a	DET
fcis-30111	122	79	survey[j	survey[j	PROPN
fcis-30111	122	80	]	]	PUNCT
fcis-30111	122	81	.	.	PUNCT
fcis-30111	122	82	neural	neural	ADJ
fcis-30111	122	83	computing	computing	NOUN
fcis-30111	122	84	and	and	CCONJ
fcis-30111	122	85	applications	application	NOUN
fcis-30111	122	86	,	,	PUNCT
fcis-30111	122	87	2024	2024	NUM
fcis-30111	122	88	,	,	PUNCT
fcis-30111	122	89	36(27	36(27	NUM
fcis-30111	122	90	):	):	PUNCT
fcis-30111	122	91	1670716725	1670716725	NUM
fcis-30111	122	92	.	.	PUNCT
fcis-30111	123	1	[	[	X
fcis-30111	123	2	2	2	NUM
fcis-30111	123	3	]	]	X
fcis-30111	123	4	duan	duan	PROPN
fcis-30111	123	5	l	l	PROPN
fcis-30111	123	6	,	,	PUNCT
fcis-30111	123	7	liu	liu	PROPN
fcis-30111	123	8	j.	j.	PROPN
fcis-30111	123	9	smart	smart	ADJ
fcis-30111	123	10	composite	composite	ADJ
fcis-30111	123	11	materials	material	NOUN
fcis-30111	123	12	and	and	CCONJ
fcis-30111	123	13	iot	iot	NOUN
fcis-30111	123	14	:	:	PUNCT
fcis-30111	123	15	revolutionizing	revolutionize	VERB
fcis-30111	123	16	real	real	ADJ
fcis-30111	123	17	-	-	PUNCT
fcis-30111	123	18	time	time	NOUN
fcis-30111	123	19	railway	railway	NOUN
fcis-30111	123	20	health	health	NOUN
fcis-30111	123	21	monitoring[j	monitoring[j	PROPN
fcis-30111	123	22	]	]	PUNCT
fcis-30111	123	23	.	.	PUNCT
fcis-30111	124	1	mrs	mrs	PROPN
fcis-30111	124	2	communications	communications	PROPN
fcis-30111	124	3	,	,	PUNCT
fcis-30111	124	4	2024	2024	NUM
fcis-30111	124	5	:	:	PUNCT
fcis-30111	124	6	1	1	NUM
fcis-30111	124	7	-	-	SYM
fcis-30111	124	8	17	17	NUM
fcis-30111	124	9	.	.	PUNCT
fcis-30111	125	1	[	[	X
fcis-30111	125	2	3	3	X
fcis-30111	125	3	]	]	X
fcis-30111	125	4	zhang	zhang	PROPN
fcis-30111	125	5	d	d	PROPN
fcis-30111	125	6	,	,	PUNCT
fcis-30111	125	7	xie	xie	PROPN
fcis-30111	125	8	m	m	PROPN
fcis-30111	125	9	,	,	PUNCT
fcis-30111	125	10	yang	yang	PROPN
fcis-30111	125	11	j	j	PROPN
fcis-30111	125	12	,	,	PUNCT
fcis-30111	125	13	et	et	PROPN
fcis-30111	125	14	al	al	PROPN
fcis-30111	125	15	.	.	PUNCT
fcis-30111	125	16	multi	multi	ADJ
fcis-30111	125	17	-	-	ADJ
fcis-30111	125	18	sensor	sensor	ADJ
fcis-30111	125	19	graph	graph	NOUN
fcis-30111	125	20	transfer	transfer	NOUN
fcis-30111	125	21	network	network	NOUN
fcis-30111	125	22	for	for	ADP
fcis-30111	125	23	health	health	NOUN
fcis-30111	125	24	assessment	assessment	NOUN
fcis-30111	125	25	of	of	ADP
fcis-30111	125	26	high	high	ADJ
fcis-30111	125	27	-	-	PUNCT
fcis-30111	125	28	speed	speed	NOUN
fcis-30111	125	29	rail	rail	NOUN
fcis-30111	125	30	suspension	suspension	NOUN
fcis-30111	125	31	systems[j	systems[j	NOUN
fcis-30111	125	32	]	]	PUNCT
fcis-30111	125	33	.	.	PUNCT
fcis-30111	126	1	ieee	ieee	NOUN
fcis-30111	126	2	transactions	transaction	NOUN
fcis-30111	126	3	on	on	ADP
fcis-30111	126	4	intelligent	intelligent	ADJ
fcis-30111	126	5	transportation	transportation	NOUN
fcis-30111	126	6	systems	system	NOUN
fcis-30111	126	7	,	,	PUNCT
fcis-30111	126	8	2023	2023	NUM
fcis-30111	126	9	,	,	PUNCT
fcis-30111	126	10	24(9	24(9	NUM
fcis-30111	126	11	):	):	PUNCT
fcis-30111	126	12	9425	9425	NUM
fcis-30111	126	13	-	-	SYM
fcis-30111	126	14	9434	9434	NUM
fcis-30111	126	15	.	.	PUNCT
fcis-30111	127	1	[	[	X
fcis-30111	127	2	4	4	NUM
fcis-30111	127	3	]	]	X
fcis-30111	127	4	li	li	PROPN
fcis-30111	127	5	h	h	PROPN
fcis-30111	127	6	,	,	PUNCT
fcis-30111	127	7	huang	huang	PROPN
fcis-30111	127	8	j	j	PROPN
fcis-30111	127	9	,	,	PUNCT
fcis-30111	127	10	huang	huang	PROPN
fcis-30111	127	11	j	j	PROPN
fcis-30111	127	12	,	,	PUNCT
fcis-30111	127	13	et	et	PROPN
fcis-30111	127	14	al	al	PROPN
fcis-30111	127	15	.	.	PUNCT
fcis-30111	128	1	deep	deep	ADJ
fcis-30111	128	2	multimodal	multimodal	NOUN
fcis-30111	128	3	learning	learning	NOUN
fcis-30111	128	4	and	and	CCONJ
fcis-30111	128	5	fusion	fusion	NOUN
fcis-30111	128	6	based	base	VERB
fcis-30111	128	7	intelligent	intelligent	ADJ
fcis-30111	128	8	fault	fault	NOUN
fcis-30111	128	9	diagnosis	diagnosis	NOUN
fcis-30111	128	10	approach[j	approach[j	PROPN
fcis-30111	128	11	]	]	PUNCT
fcis-30111	128	12	.	.	PUNCT
fcis-30111	129	1	journal	journal	PROPN
fcis-30111	129	2	of	of	ADP
fcis-30111	129	3	beijing	beijing	PROPN
fcis-30111	129	4	institute	institute	PROPN
fcis-30111	129	5	of	of	ADP
fcis-30111	129	6	technology	technology	NOUN
fcis-30111	129	7	,	,	PUNCT
fcis-30111	129	8	2021	2021	NUM
fcis-30111	129	9	,	,	PUNCT
fcis-30111	129	10	30(2	30(2	NUM
fcis-30111	129	11	):	):	PUNCT
fcis-30111	129	12	172	172	NUM
fcis-30111	129	13	-	-	SYM
fcis-30111	129	14	185	185	NUM
fcis-30111	129	15	.	.	PUNCT
fcis-30111	130	1	[	[	X
fcis-30111	130	2	5	5	NUM
fcis-30111	130	3	]	]	PUNCT
fcis-30111	130	4	ghaboura	ghaboura	NOUN
fcis-30111	130	5	s	s	PROPN
fcis-30111	130	6	,	,	PUNCT
fcis-30111	130	7	ferdousi	ferdousi	ADJ
fcis-30111	130	8	r	r	NOUN
fcis-30111	130	9	,	,	PUNCT
fcis-30111	130	10	laamarti	laamarti	PROPN
fcis-30111	130	11	f	f	PROPN
fcis-30111	130	12	,	,	PUNCT
fcis-30111	130	13	et	et	PROPN
fcis-30111	130	14	al	al	PROPN
fcis-30111	130	15	.	.	PUNCT
fcis-30111	131	1	digital	digital	PROPN
fcis-30111	131	2	twin	twin	NOUN
fcis-30111	131	3	for	for	ADP
fcis-30111	131	4	railway	railway	NOUN
fcis-30111	131	5	:	:	PUNCT
fcis-30111	131	6	a	a	DET
fcis-30111	131	7	comprehensive	comprehensive	ADJ
fcis-30111	131	8	survey[j	survey[j	NOUN
fcis-30111	131	9	]	]	PUNCT
fcis-30111	131	10	.	.	PUNCT
fcis-30111	132	1	ieee	ieee	NOUN
fcis-30111	132	2	access	access	NOUN
fcis-30111	132	3	,	,	PUNCT
fcis-30111	132	4	2023	2023	NUM
fcis-30111	132	5	,	,	PUNCT
fcis-30111	132	6	11	11	NUM
fcis-30111	132	7	:	:	SYM
fcis-30111	132	8	120237	120237	NUM
fcis-30111	132	9	-	-	SYM
fcis-30111	132	10	120257	120257	NUM
fcis-30111	132	11	.	.	PUNCT
fcis-30111	133	1	[	[	X
fcis-30111	133	2	6	6	NUM
fcis-30111	133	3	]	]	SYM
fcis-30111	133	4	li	li	PROPN
fcis-30111	133	5	p	p	PROPN
fcis-30111	133	6	,	,	PUNCT
fcis-30111	133	7	xue	xue	PROPN
fcis-30111	133	8	r	r	PROPN
fcis-30111	133	9	,	,	PUNCT
fcis-30111	133	10	shao	shao	PROPN
fcis-30111	133	11	s	s	PROPN
fcis-30111	133	12	,	,	PUNCT
fcis-30111	133	13	et	et	PROPN
fcis-30111	133	14	al	al	PROPN
fcis-30111	133	15	.	.	PUNCT
fcis-30111	134	1	current	current	ADJ
fcis-30111	134	2	state	state	NOUN
fcis-30111	134	3	and	and	CCONJ
fcis-30111	134	4	predicted	predict	VERB
fcis-30111	134	5	technological	technological	ADJ
fcis-30111	134	6	trends	trend	NOUN
fcis-30111	134	7	in	in	ADP
fcis-30111	134	8	global	global	ADJ
fcis-30111	134	9	railway	railway	NOUN
fcis-30111	134	10	intelligent	intelligent	ADJ
fcis-30111	134	11	digital	digital	ADJ
fcis-30111	134	12	transformation	transformation	NOUN
fcis-30111	135	1	[	[	X
fcis-30111	135	2	j	j	X
fcis-30111	135	3	]	]	X
fcis-30111	135	4	.	.	PUNCT
fcis-30111	136	1	railway	railway	NOUN
fcis-30111	136	2	sciences	sciences	PROPN
fcis-30111	136	3	,	,	PUNCT
fcis-30111	136	4	2023	2023	NUM
fcis-30111	136	5	,	,	PUNCT
fcis-30111	136	6	2(4	2(4	NUM
fcis-30111	136	7	):	):	PUNCT
fcis-30111	136	8	397	397	NUM
fcis-30111	136	9	-	-	SYM
fcis-30111	136	10	412	412	NUM
fcis-30111	136	11	.	.	PUNCT
fcis-30111	137	1	[	[	X
fcis-30111	137	2	7	7	NUM
fcis-30111	137	3	]	]	X
fcis-30111	137	4	spiryagin	spiryagin	NOUN
fcis-30111	137	5	m	m	PROPN
fcis-30111	137	6	,	,	PUNCT
fcis-30111	137	7	edelmann	edelmann	PROPN
fcis-30111	137	8	j	j	PROPN
fcis-30111	137	9	,	,	PUNCT
fcis-30111	137	10	klinger	klinger	PROPN
fcis-30111	137	11	f	f	PROPN
fcis-30111	137	12	,	,	PUNCT
fcis-30111	137	13	et	et	PROPN
fcis-30111	137	14	al	al	PROPN
fcis-30111	137	15	.	.	PROPN
fcis-30111	137	16	vehicle	vehicle	NOUN
fcis-30111	137	17	system	system	NOUN
fcis-30111	137	18	dynamics	dynamic	NOUN
fcis-30111	137	19	in	in	ADP
fcis-30111	137	20	digital	digital	ADJ
fcis-30111	137	21	twin	twin	ADJ
fcis-30111	137	22	studies	study	NOUN
fcis-30111	137	23	in	in	ADP
fcis-30111	137	24	rail	rail	NOUN
fcis-30111	137	25	and	and	CCONJ
fcis-30111	137	26	road	road	NOUN
fcis-30111	137	27	domains[j	domains[j	PROPN
fcis-30111	137	28	]	]	PUNCT
fcis-30111	137	29	.	.	PUNCT
fcis-30111	138	1	vehicle	vehicle	NOUN
fcis-30111	138	2	system	system	NOUN
fcis-30111	138	3	dynamics	dynamic	NOUN
fcis-30111	138	4	,	,	PUNCT
fcis-30111	138	5	2023	2023	NUM
fcis-30111	138	6	,	,	PUNCT
fcis-30111	138	7	61(7	61(7	NUM
fcis-30111	138	8	):	):	PUNCT
fcis-30111	138	9	1737	1737	NUM
fcis-30111	138	10	-	-	SYM
fcis-30111	138	11	1786	1786	NUM
fcis-30111	138	12	.	.	PUNCT
fcis-30111	139	1	[	[	X
fcis-30111	139	2	8	8	NUM
fcis-30111	139	3	]	]	SYM
fcis-30111	139	4	hu	hu	PROPN
fcis-30111	139	5	x	x	PROPN
fcis-30111	139	6	,	,	PUNCT
fcis-30111	139	7	wu	wu	PROPN
fcis-30111	139	8	j	j	PROPN
fcis-30111	139	9	,	,	PUNCT
fcis-30111	139	10	gao	gao	PROPN
fcis-30111	139	11	y.	y.	PROPN
fcis-30111	139	12	a	a	DET
fcis-30111	139	13	review	review	NOUN
fcis-30111	139	14	of	of	ADP
fcis-30111	139	15	structural	structural	ADJ
fcis-30111	139	16	health	health	NOUN
fcis-30111	139	17	monitoring	monitoring	NOUN
fcis-30111	139	18	for	for	ADP
fcis-30111	139	19	flexible	flexible	ADJ
fcis-30111	139	20	composite	composite	NOUN
fcis-30111	139	21	materials[j	materials[j	PROPN
fcis-30111	139	22	]	]	PUNCT
fcis-30111	139	23	.	.	PUNCT
fcis-30111	140	1	applied	apply	VERB
fcis-30111	140	2	composite	composite	ADJ
fcis-30111	140	3	materials	material	NOUN
fcis-30111	140	4	,	,	PUNCT
fcis-30111	140	5	2024	2024	NUM
fcis-30111	140	6	:	:	PUNCT
fcis-30111	140	7	1	1	NUM
fcis-30111	140	8	-	-	SYM
fcis-30111	140	9	41	41	NUM
fcis-30111	140	10	.	.	PUNCT
fcis-30111	141	1	[	[	X
fcis-30111	141	2	9	9	NUM
fcis-30111	141	3	]	]	PUNCT
fcis-30111	141	4	jamora	jamora	NOUN
fcis-30111	141	5	j	j	PROPN
fcis-30111	141	6	r	r	PROPN
fcis-30111	141	7	,	,	PUNCT
fcis-30111	141	8	sotirelis	sotirelis	NOUN
fcis-30111	141	9	p	p	X
fcis-30111	141	10	,	,	PUNCT
fcis-30111	141	11	nolan	nolan	PROPN
fcis-30111	141	12	a	a	PROPN
fcis-30111	141	13	,	,	PUNCT
fcis-30111	141	14	et	et	PROPN
fcis-30111	141	15	al	al	PROPN
fcis-30111	141	16	.	.	PUNCT
fcis-30111	141	17	multiple	multiple	ADJ
fcis-30111	141	18	modality	modality	NOUN
fcis-30111	141	19	sensor	sensor	NOUN
fcis-30111	141	20	fusion	fusion	NOUN
fcis-30111	141	21	from	from	ADP
fcis-30111	141	22	synthetic	synthetic	ADJ
fcis-30111	141	23	aperture	aperture	NOUN
fcis-30111	141	24	radar	radar	NOUN
fcis-30111	141	25	,	,	PUNCT
fcis-30111	141	26	lidar	lidar	NOUN
fcis-30111	141	27	,	,	PUNCT
fcis-30111	141	28	and	and	CCONJ
fcis-30111	141	29	electrooptical	electrooptical	ADJ
fcis-30111	141	30	systems	system	NOUN
fcis-30111	141	31	using	use	VERB
fcis-30111	141	32	three	three	NUM
fcis-30111	141	33	-	-	PUNCT
fcis-30111	141	34	dimensional	dimensional	ADJ
fcis-30111	141	35	data	datum	NOUN
fcis-30111	141	36	representations	representation	NOUN
fcis-30111	141	37	[	[	PUNCT
fcis-30111	141	38	c]//	c]//	ADJ
fcis-30111	141	39	algorithms	algorithm	NOUN
fcis-30111	141	40	for	for	ADP
fcis-30111	141	41	synthetic	synthetic	ADJ
fcis-30111	141	42	aperture	aperture	NOUN
fcis-30111	141	43	radar	radar	NOUN
fcis-30111	141	44	imagery	imagery	NOUN
fcis-30111	141	45	xxix	xxix	PROPN
fcis-30111	141	46	.	.	PUNCT
fcis-30111	142	1	spie	spie	PROPN
fcis-30111	142	2	,	,	PUNCT
fcis-30111	142	3	2022	2022	NUM
fcis-30111	142	4	,	,	PUNCT
fcis-30111	142	5	12095	12095	NUM
fcis-30111	142	6	:	:	PUNCT
fcis-30111	142	7	32	32	NUM
fcis-30111	142	8	-	-	SYM
fcis-30111	142	9	43	43	NUM
fcis-30111	142	10	.	.	PUNCT
fcis-30111	143	1	[	[	X
fcis-30111	143	2	10	10	NUM
fcis-30111	143	3	]	]	X
fcis-30111	143	4	saeed	saeed	PROPN
fcis-30111	143	5	m	m	PROPN
fcis-30111	143	6	,	,	PUNCT
fcis-30111	143	7	briz	briz	PROPN
fcis-30111	143	8	f	f	PROPN
fcis-30111	143	9	,	,	PUNCT
fcis-30111	143	10	guerrero	guerrero	PROPN
fcis-30111	143	11	j	j	PROPN
fcis-30111	143	12	m	m	PROPN
fcis-30111	143	13	,	,	PUNCT
fcis-30111	143	14	et	et	PROPN
fcis-30111	143	15	al	al	PROPN
fcis-30111	143	16	.	.	PROPN
fcis-30111	143	17	onboard	onboard	ADP
fcis-30111	143	18	energy	energy	NOUN
fcis-30111	143	19	storage	storage	NOUN
fcis-30111	143	20	systems	system	NOUN
fcis-30111	143	21	for	for	ADP
fcis-30111	143	22	railway	railway	NOUN
fcis-30111	143	23	:	:	PUNCT
fcis-30111	143	24	present	present	ADJ
fcis-30111	143	25	and	and	CCONJ
fcis-30111	143	26	trends[j	trends[j	NOUN
fcis-30111	143	27	]	]	PUNCT
fcis-30111	143	28	.	.	PUNCT
fcis-30111	144	1	ieee	ieee	PROPN
fcis-30111	144	2	open	open	PROPN
fcis-30111	144	3	journal	journal	PROPN
fcis-30111	144	4	of	of	ADP
fcis-30111	144	5	industry	industry	NOUN
fcis-30111	144	6	applications	application	NOUN
fcis-30111	144	7	,	,	PUNCT
fcis-30111	144	8	2023	2023	NUM
fcis-30111	144	9	,	,	PUNCT
fcis-30111	144	10	4	4	NUM
fcis-30111	144	11	:	:	SYM
fcis-30111	144	12	238	238	NUM
fcis-30111	144	13	-	-	SYM
fcis-30111	144	14	259	259	NUM
fcis-30111	144	15	.	.	PUNCT