id sid tid token lemma pos ajst-29929 1 1 academic academic ADJ ajst-29929 1 2 journal journal NOUN ajst-29929 1 3 of of ADP ajst-29929 1 4 science science NOUN ajst-29929 1 5 and and CCONJ ajst-29929 1 6 technology technology NOUN ajst-29929 1 7 issn issn NOUN ajst-29929 1 8 : : PUNCT ajst-29929 1 9 2771 2771 NUM ajst-29929 1 10 - - SYM ajst-29929 1 11 3032 3032 NUM ajst-29929 1 12 | | NOUN ajst-29929 1 13 vol vol NOUN ajst-29929 1 14 . . PUNCT ajst-29929 2 1 14 14 NUM ajst-29929 2 2 , , PUNCT ajst-29929 2 3 no no INTJ ajst-29929 2 4 . . NOUN ajst-29929 2 5 2 2 NUM ajst-29929 2 6 , , PUNCT ajst-29929 2 7 2025 2025 NUM ajst-29929 2 8 191 191 NUM ajst-29929 2 9 a a DET ajst-29929 2 10 bearing bear VERB ajst-29929 2 11 fault fault NOUN ajst-29929 2 12 diagnosis diagnosis NOUN ajst-29929 2 13 method method NOUN ajst-29929 2 14 based base VERB ajst-29929 2 15 on on ADP ajst-29929 2 16 fusion fusion NOUN ajst-29929 2 17 of of ADP ajst-29929 2 18 cnn‐bilstm‐transformer cnn‐bilstm‐transformer PROPN ajst-29929 2 19 and and CCONJ ajst-29929 2 20 cross‐attention cross‐attention NOUN ajst-29929 2 21 xun xun PROPN ajst-29929 3 1 yuan1 yuan1 PROPN ajst-29929 3 2 , , PUNCT ajst-29929 3 3 a a PRON ajst-29929 3 4 , , PUNCT ajst-29929 3 5 wei wei X ajst-29929 3 6 wu1 wu1 PROPN ajst-29929 4 1 1school 1school NUM ajst-29929 4 2 of of ADP ajst-29929 4 3 mechanical mechanical ADJ ajst-29929 4 4 engineering engineering NOUN ajst-29929 4 5 , , PUNCT ajst-29929 4 6 xi'an xi'an PROPN ajst-29929 4 7 shiyou shiyou PROPN ajst-29929 4 8 university university PROPN ajst-29929 4 9 , , PUNCT ajst-29929 4 10 china china PROPN ajst-29929 4 11 ayuanxun1221@163.com ayuanxun1221@163.com X ajst-29929 4 12 abstract abstract NOUN ajst-29929 4 13 : : PUNCT ajst-29929 4 14 to to PART ajst-29929 4 15 address address VERB ajst-29929 4 16 the the DET ajst-29929 4 17 limitations limitation NOUN ajst-29929 4 18 of of ADP ajst-29929 4 19 traditional traditional ADJ ajst-29929 4 20 diagnostic diagnostic ADJ ajst-29929 4 21 models model NOUN ajst-29929 4 22 in in ADP ajst-29929 4 23 achieving achieve VERB ajst-29929 4 24 high high ADJ ajst-29929 4 25 accuracy accuracy NOUN ajst-29929 4 26 in in ADP ajst-29929 4 27 rolling rolling ADJ ajst-29929 4 28 bearing bear VERB ajst-29929 4 29 fault fault NOUN ajst-29929 4 30 diagnosis diagnosis NOUN ajst-29929 4 31 and and CCONJ ajst-29929 4 32 their their PRON ajst-29929 4 33 inadequate inadequate ADJ ajst-29929 4 34 extraction extraction NOUN ajst-29929 4 35 of of ADP ajst-29929 4 36 vibration vibration NOUN ajst-29929 4 37 signal signal NOUN ajst-29929 4 38 features feature VERB ajst-29929 4 39 across across ADP ajst-29929 4 40 both both CCONJ ajst-29929 4 41 frequency frequency NOUN ajst-29929 4 42 and and CCONJ ajst-29929 4 43 time time NOUN ajst-29929 4 44 domains domain NOUN ajst-29929 4 45 , , PUNCT ajst-29929 4 46 this this DET ajst-29929 4 47 paper paper NOUN ajst-29929 4 48 introduces introduce VERB ajst-29929 4 49 a a DET ajst-29929 4 50 fault fault NOUN ajst-29929 4 51 diagnosis diagnosis NOUN ajst-29929 4 52 approach approach NOUN ajst-29929 4 53 that that PRON ajst-29929 4 54 integrates integrate VERB ajst-29929 4 55 multiple multiple ADJ ajst-29929 4 56 neural neural ADJ ajst-29929 4 57 networks network NOUN ajst-29929 4 58 with with ADP ajst-29929 4 59 a a DET ajst-29929 4 60 cross cross ADJ ajst-29929 4 61 - - ADJ ajst-29929 4 62 attention attention ADJ ajst-29929 4 63 mechanism mechanism NOUN ajst-29929 4 64 . . PUNCT ajst-29929 5 1 initially initially ADV ajst-29929 5 2 , , PUNCT ajst-29929 5 3 onedimensional onedimensional ADJ ajst-29929 5 4 fault fault NOUN ajst-29929 5 5 signals signal NOUN ajst-29929 5 6 undergo undergo VERB ajst-29929 5 7 processing processing NOUN ajst-29929 5 8 through through ADP ajst-29929 5 9 the the DET ajst-29929 5 10 fast fast ADJ ajst-29929 5 11 fourier fourier NOUN ajst-29929 5 12 transform transform NOUN ajst-29929 5 13 ( ( PUNCT ajst-29929 5 14 fft fft PROPN ajst-29929 5 15 ) ) PUNCT ajst-29929 5 16 and and CCONJ ajst-29929 5 17 variational variational ADJ ajst-29929 5 18 mode mode NOUN ajst-29929 5 19 decomposition decomposition NOUN ajst-29929 5 20 ( ( PUNCT ajst-29929 5 21 vmd vmd NOUN ajst-29929 5 22 ) ) PUNCT ajst-29929 5 23 techniques technique NOUN ajst-29929 5 24 to to PART ajst-29929 5 25 obtain obtain VERB ajst-29929 5 26 spectral spectral ADJ ajst-29929 5 27 information information NOUN ajst-29929 5 28 and and CCONJ ajst-29929 5 29 time time NOUN ajst-29929 5 30 - - PUNCT ajst-29929 5 31 domain domain NOUN ajst-29929 5 32 features feature NOUN ajst-29929 5 33 , , PUNCT ajst-29929 5 34 respectively respectively ADV ajst-29929 5 35 . . PUNCT ajst-29929 6 1 the the DET ajst-29929 6 2 combined combined ADJ ajst-29929 6 3 application application NOUN ajst-29929 6 4 of of ADP ajst-29929 6 5 fft fft PROPN ajst-29929 6 6 and and CCONJ ajst-29929 6 7 vmd vmd PROPN ajst-29929 6 8 facilitates facilitate VERB ajst-29929 6 9 the the DET ajst-29929 6 10 efficient efficient ADJ ajst-29929 6 11 extraction extraction NOUN ajst-29929 6 12 of of ADP ajst-29929 6 13 multi multi ADJ ajst-29929 6 14 - - ADJ ajst-29929 6 15 scale scale ADJ ajst-29929 6 16 features feature NOUN ajst-29929 6 17 from from ADP ajst-29929 6 18 the the DET ajst-29929 6 19 signals signal NOUN ajst-29929 6 20 . . PUNCT ajst-29929 7 1 subsequently subsequently ADV ajst-29929 7 2 , , PUNCT ajst-29929 7 3 a a DET ajst-29929 7 4 cnn cnn PROPN ajst-29929 7 5 - - PUNCT ajst-29929 7 6 transformer transformer NOUN ajst-29929 7 7 network network NOUN ajst-29929 7 8 is be AUX ajst-29929 7 9 utilized utilize VERB ajst-29929 7 10 to to PART ajst-29929 7 11 capture capture VERB ajst-29929 7 12 the the DET ajst-29929 7 13 spatial spatial ADJ ajst-29929 7 14 characteristics characteristic NOUN ajst-29929 7 15 of of ADP ajst-29929 7 16 the the DET ajst-29929 7 17 pre pre ADJ ajst-29929 7 18 - - ADJ ajst-29929 7 19 processed process VERB ajst-29929 7 20 multi multi ADJ ajst-29929 7 21 - - ADJ ajst-29929 7 22 scale scale ADJ ajst-29929 7 23 fault fault NOUN ajst-29929 7 24 signal signal NOUN ajst-29929 7 25 features feature NOUN ajst-29929 7 26 . . PUNCT ajst-29929 8 1 this this PRON ajst-29929 8 2 is be AUX ajst-29929 8 3 followed follow VERB ajst-29929 8 4 by by ADP ajst-29929 8 5 the the DET ajst-29929 8 6 deployment deployment NOUN ajst-29929 8 7 of of ADP ajst-29929 8 8 a a DET ajst-29929 8 9 bidirectional bidirectional ADJ ajst-29929 8 10 long long ADJ ajst-29929 8 11 short short ADJ ajst-29929 8 12 - - PUNCT ajst-29929 8 13 term term NOUN ajst-29929 8 14 memory memory NOUN ajst-29929 8 15 ( ( PUNCT ajst-29929 8 16 bilstm bilstm NOUN ajst-29929 8 17 ) ) PUNCT ajst-29929 8 18 network network NOUN ajst-29929 8 19 to to PART ajst-29929 8 20 extract extract VERB ajst-29929 8 21 sequential sequential ADJ ajst-29929 8 22 information information NOUN ajst-29929 8 23 , , PUNCT ajst-29929 8 24 with with ADP ajst-29929 8 25 a a DET ajst-29929 8 26 particular particular ADJ ajst-29929 8 27 emphasis emphasis NOUN ajst-29929 8 28 on on ADP ajst-29929 8 29 pivotal pivotal ADJ ajst-29929 8 30 temporal temporal ADJ ajst-29929 8 31 features feature NOUN ajst-29929 8 32 . . PUNCT ajst-29929 9 1 additionally additionally ADV ajst-29929 9 2 , , PUNCT ajst-29929 9 3 the the DET ajst-29929 9 4 integration integration NOUN ajst-29929 9 5 of of ADP ajst-29929 9 6 the the DET ajst-29929 9 7 transformer transformer NOUN ajst-29929 9 8 network network NOUN ajst-29929 9 9 further far ADV ajst-29929 9 10 enhances enhance VERB ajst-29929 9 11 feature feature NOUN ajst-29929 9 12 extraction extraction NOUN ajst-29929 9 13 and and CCONJ ajst-29929 9 14 fusion fusion NOUN ajst-29929 9 15 capabilities capability NOUN ajst-29929 9 16 . . PUNCT ajst-29929 10 1 ultimately ultimately ADV ajst-29929 10 2 , , PUNCT ajst-29929 10 3 a a DET ajst-29929 10 4 cross cross ADJ ajst-29929 10 5 - - ADJ ajst-29929 10 6 attention attention ADJ ajst-29929 10 7 mechanism mechanism NOUN ajst-29929 10 8 is be AUX ajst-29929 10 9 implemented implement VERB ajst-29929 10 10 to to PART ajst-29929 10 11 seamlessly seamlessly ADV ajst-29929 10 12 integrate integrate VERB ajst-29929 10 13 time time NOUN ajst-29929 10 14 - - PUNCT ajst-29929 10 15 domain domain NOUN ajst-29929 10 16 and and CCONJ ajst-29929 10 17 frequency frequency NOUN ajst-29929 10 18 - - PUNCT ajst-29929 10 19 domain domain NOUN ajst-29929 10 20 features feature NOUN ajst-29929 10 21 , , PUNCT ajst-29929 10 22 thereby thereby ADV ajst-29929 10 23 bolstering bolster VERB ajst-29929 10 24 the the DET ajst-29929 10 25 model model NOUN ajst-29929 10 26 's 's PART ajst-29929 10 27 performance performance NOUN ajst-29929 10 28 and and CCONJ ajst-29929 10 29 generalization generalization NOUN ajst-29929 10 30 capacity capacity NOUN ajst-29929 10 31 for for ADP ajst-29929 10 32 fault fault NOUN ajst-29929 10 33 classification classification NOUN ajst-29929 10 34 tasks task NOUN ajst-29929 10 35 . . PUNCT ajst-29929 11 1 experimental experimental ADJ ajst-29929 11 2 findings finding NOUN ajst-29929 11 3 validate validate VERB ajst-29929 11 4 the the DET ajst-29929 11 5 proposed propose VERB ajst-29929 11 6 model model NOUN ajst-29929 11 7 's 's PART ajst-29929 11 8 diagnostic diagnostic ADJ ajst-29929 11 9 effectiveness effectiveness NOUN ajst-29929 11 10 , , PUNCT ajst-29929 11 11 achieving achieve VERB ajst-29929 11 12 an an DET ajst-29929 11 13 accuracy accuracy NOUN ajst-29929 11 14 and and CCONJ ajst-29929 11 15 recall recall NOUN ajst-29929 11 16 rate rate NOUN ajst-29929 11 17 exceeding exceed VERB ajst-29929 11 18 99 99 NUM ajst-29929 11 19 % % NOUN ajst-29929 11 20 in in ADP ajst-29929 11 21 rolling rolling ADJ ajst-29929 11 22 bearing bear VERB ajst-29929 11 23 fault fault NOUN ajst-29929 11 24 diagnosis diagnosis NOUN ajst-29929 11 25 . . PUNCT ajst-29929 12 1 keywords keyword NOUN ajst-29929 12 2 : : PUNCT ajst-29929 12 3 rolling rolling ADJ ajst-29929 12 4 bearing bearing NOUN ajst-29929 12 5 ; ; PUNCT ajst-29929 12 6 convolutional convolutional ADJ ajst-29929 12 7 neural neural ADJ ajst-29929 12 8 network network NOUN ajst-29929 12 9 ( ( PUNCT ajst-29929 12 10 cnn cnn PROPN ajst-29929 12 11 ) ) PUNCT ajst-29929 12 12 ; ; PUNCT ajst-29929 12 13 bidirectional bidirectional ADJ ajst-29929 12 14 long long ADJ ajst-29929 12 15 short short ADJ ajst-29929 12 16 - - PUNCT ajst-29929 12 17 term term NOUN ajst-29929 12 18 memory memory NOUN ajst-29929 12 19 ( ( PUNCT ajst-29929 12 20 bilstm bilstm NOUN ajst-29929 12 21 ) ) PUNCT ajst-29929 12 22 ; ; PUNCT ajst-29929 12 23 cross cross NOUN ajst-29929 12 24 - - ADJ ajst-29929 12 25 attention attention ADJ ajst-29929 12 26 ; ; PUNCT ajst-29929 12 27 transformer transformer NOUN ajst-29929 12 28 . . PUNCT ajst-29929 13 1 1 1 X ajst-29929 13 2 . . X ajst-29929 13 3 introduction introduction NOUN ajst-29929 13 4 in in ADP ajst-29929 13 5 the the DET ajst-29929 13 6 operation operation NOUN ajst-29929 13 7 of of ADP ajst-29929 13 8 industrial industrial ADJ ajst-29929 13 9 equipment equipment NOUN ajst-29929 13 10 , , PUNCT ajst-29929 13 11 rolling rolling ADJ ajst-29929 13 12 bearings bearing NOUN ajst-29929 13 13 , , PUNCT ajst-29929 13 14 as as ADP ajst-29929 13 15 a a DET ajst-29929 13 16 core core NOUN ajst-29929 13 17 component component NOUN ajst-29929 13 18 , , PUNCT ajst-29929 13 19 play play VERB ajst-29929 13 20 a a DET ajst-29929 13 21 vital vital ADJ ajst-29929 13 22 role role NOUN ajst-29929 13 23 in in ADP ajst-29929 13 24 ensuring ensure VERB ajst-29929 13 25 the the DET ajst-29929 13 26 safe safe ADJ ajst-29929 13 27 and and CCONJ ajst-29929 13 28 reliable reliable ADJ ajst-29929 13 29 operation operation NOUN ajst-29929 13 30 of of ADP ajst-29929 13 31 machinery machinery NOUN ajst-29929 13 32 . . PUNCT ajst-29929 14 1 the the DET ajst-29929 14 2 detection detection NOUN ajst-29929 14 3 of of ADP ajst-29929 14 4 bearing bear VERB ajst-29929 14 5 faults fault NOUN ajst-29929 14 6 is be AUX ajst-29929 14 7 therefore therefore ADV ajst-29929 14 8 of of ADP ajst-29929 14 9 paramount paramount ADJ ajst-29929 14 10 importance importance NOUN ajst-29929 14 11 . . PUNCT ajst-29929 15 1 traditional traditional ADJ ajst-29929 15 2 fault fault NOUN ajst-29929 15 3 diagnosis diagnosis NOUN ajst-29929 15 4 methods method NOUN ajst-29929 15 5 rely rely VERB ajst-29929 15 6 heavily heavily ADV ajst-29929 15 7 on on ADP ajst-29929 15 8 expert expert NOUN ajst-29929 15 9 experience experience NOUN ajst-29929 15 10 and and CCONJ ajst-29929 15 11 basic basic ADJ ajst-29929 15 12 signal signal NOUN ajst-29929 15 13 processing processing NOUN ajst-29929 15 14 techniques technique NOUN ajst-29929 15 15 , , PUNCT ajst-29929 15 16 which which PRON ajst-29929 15 17 often often ADV ajst-29929 15 18 struggle struggle VERB ajst-29929 15 19 to to PART ajst-29929 15 20 effectively effectively ADV ajst-29929 15 21 extract extract VERB ajst-29929 15 22 and and CCONJ ajst-29929 15 23 analyze analyze VERB ajst-29929 15 24 the the DET ajst-29929 15 25 complex complex ADJ ajst-29929 15 26 characteristics characteristic NOUN ajst-29929 15 27 of of ADP ajst-29929 15 28 vibration vibration NOUN ajst-29929 15 29 signals signal NOUN ajst-29929 15 30 , , PUNCT ajst-29929 15 31 resulting result VERB ajst-29929 15 32 in in ADP ajst-29929 15 33 relatively relatively ADV ajst-29929 15 34 low low ADJ ajst-29929 15 35 diagnostic diagnostic ADJ ajst-29929 15 36 accuracy accuracy NOUN ajst-29929 15 37 . . PUNCT ajst-29929 16 1 with with ADP ajst-29929 16 2 the the DET ajst-29929 16 3 rapid rapid ADJ ajst-29929 16 4 development development NOUN ajst-29929 16 5 of of ADP ajst-29929 16 6 deep deep ADJ ajst-29929 16 7 learning learning NOUN ajst-29929 16 8 technology technology NOUN ajst-29929 16 9 , , PUNCT ajst-29929 16 10 fault fault VERB ajst-29929 16 11 diagnosis diagnosis NOUN ajst-29929 16 12 methods method NOUN ajst-29929 16 13 based base VERB ajst-29929 16 14 on on ADP ajst-29929 16 15 deep deep ADJ ajst-29929 16 16 neural neural ADJ ajst-29929 16 17 networks network NOUN ajst-29929 16 18 have have AUX ajst-29929 16 19 gradually gradually ADV ajst-29929 16 20 demonstrated demonstrate VERB ajst-29929 16 21 their their PRON ajst-29929 16 22 superiority superiority NOUN ajst-29929 16 23 . . PUNCT ajst-29929 17 1 however however ADV ajst-29929 17 2 , , PUNCT ajst-29929 17 3 how how SCONJ ajst-29929 17 4 to to PART ajst-29929 17 5 fully fully ADV ajst-29929 17 6 utilize utilize VERB ajst-29929 17 7 both both DET ajst-29929 17 8 the the DET ajst-29929 17 9 frequency frequency NOUN ajst-29929 17 10 - - PUNCT ajst-29929 17 11 domain domain NOUN ajst-29929 17 12 and and CCONJ ajst-29929 17 13 time time NOUN ajst-29929 17 14 - - PUNCT ajst-29929 17 15 domain domain NOUN ajst-29929 17 16 features feature NOUN ajst-29929 17 17 of of ADP ajst-29929 17 18 vibration vibration NOUN ajst-29929 17 19 signals signal NOUN ajst-29929 17 20 remains remain VERB ajst-29929 17 21 a a DET ajst-29929 17 22 significant significant ADJ ajst-29929 17 23 research research NOUN ajst-29929 17 24 challenge challenge NOUN ajst-29929 17 25 . . PUNCT ajst-29929 18 1 in in ADP ajst-29929 18 2 the the DET ajst-29929 18 3 field field NOUN ajst-29929 18 4 of of ADP ajst-29929 18 5 vibration vibration NOUN ajst-29929 18 6 signal signal NOUN ajst-29929 18 7 extraction extraction NOUN ajst-29929 18 8 and and CCONJ ajst-29929 18 9 analysis analysis NOUN ajst-29929 18 10 , , PUNCT ajst-29929 18 11 fast fast ADJ ajst-29929 18 12 fourier fourier NOUN ajst-29929 18 13 transform transform NOUN ajst-29929 18 14 ( ( PUNCT ajst-29929 18 15 fft fft PROPN ajst-29929 18 16 ) ) PUNCT ajst-29929 18 17 and and CCONJ ajst-29929 18 18 variational variational ADJ ajst-29929 18 19 mode mode NOUN ajst-29929 18 20 decomposition decomposition NOUN ajst-29929 18 21 ( ( PUNCT ajst-29929 18 22 vmd vmd NOUN ajst-29929 18 23 ) ) PUNCT ajst-29929 18 24 have have AUX ajst-29929 18 25 been be AUX ajst-29929 18 26 widely widely ADV ajst-29929 18 27 adopted adopt VERB ajst-29929 18 28 as as ADP ajst-29929 18 29 the the DET ajst-29929 18 30 primary primary ADJ ajst-29929 18 31 methods method NOUN ajst-29929 18 32 . . PUNCT ajst-29929 19 1 for for ADP ajst-29929 19 2 instance instance NOUN ajst-29929 19 3 , , PUNCT ajst-29929 19 4 reference reference NOUN ajst-29929 19 5 [ [ X ajst-29929 19 6 3 3 NUM ajst-29929 19 7 ] ] PUNCT ajst-29929 19 8 introduces introduce VERB ajst-29929 19 9 a a DET ajst-29929 19 10 fault fault NOUN ajst-29929 19 11 diagnosis diagnosis NOUN ajst-29929 19 12 technique technique NOUN ajst-29929 19 13 that that PRON ajst-29929 19 14 combines combine VERB ajst-29929 19 15 parameter parameter NOUN ajst-29929 19 16 - - PUNCT ajst-29929 19 17 optimized optimize VERB ajst-29929 19 18 variational variational ADJ ajst-29929 19 19 mode mode NOUN ajst-29929 19 20 decomposition decomposition NOUN ajst-29929 19 21 ( ( PUNCT ajst-29929 19 22 vmd vmd NOUN ajst-29929 19 23 ) ) PUNCT ajst-29929 19 24 with with ADP ajst-29929 19 25 convolutional convolutional ADJ ajst-29929 19 26 neural neural ADJ ajst-29929 19 27 networks network NOUN ajst-29929 19 28 ( ( PUNCT ajst-29929 19 29 cnn cnn PROPN ajst-29929 19 30 ) ) PUNCT ajst-29929 19 31 . . PUNCT ajst-29929 20 1 reference reference NOUN ajst-29929 20 2 [ [ X ajst-29929 20 3 4 4 X ajst-29929 20 4 ] ] PUNCT ajst-29929 20 5 proposes propose VERB ajst-29929 20 6 a a DET ajst-29929 20 7 method method NOUN ajst-29929 20 8 based base VERB ajst-29929 20 9 on on ADP ajst-29929 20 10 vmd vmd PROPN ajst-29929 20 11 optimized optimize VERB ajst-29929 20 12 by by ADP ajst-29929 20 13 the the DET ajst-29929 20 14 majorizing majorize VERB ajst-29929 20 15 sparrow sparrow NOUN ajst-29929 20 16 search search NOUN ajst-29929 20 17 algorithm algorithm NOUN ajst-29929 20 18 ( ( PUNCT ajst-29929 20 19 mssa mssa PROPN ajst-29929 20 20 - - PUNCT ajst-29929 20 21 vmd vmd NOUN ajst-29929 20 22 ) ) PUNCT ajst-29929 20 23 . . PUNCT ajst-29929 21 1 reference reference NOUN ajst-29929 21 2 [ [ X ajst-29929 21 3 5 5 NUM ajst-29929 21 4 ] ] PUNCT ajst-29929 21 5 presents present VERB ajst-29929 21 6 a a DET ajst-29929 21 7 rolling rolling ADJ ajst-29929 21 8 bearing bear VERB ajst-29929 21 9 fault fault NOUN ajst-29929 21 10 diagnosis diagnosis NOUN ajst-29929 21 11 approach approach NOUN ajst-29929 21 12 using use VERB ajst-29929 21 13 improved improve VERB ajst-29929 21 14 multiscale multiscale ADJ ajst-29929 21 15 sample sample NOUN ajst-29929 21 16 entropy entropy NOUN ajst-29929 21 17 ( ( PUNCT ajst-29929 21 18 imse imse NOUN ajst-29929 21 19 ) ) PUNCT ajst-29929 21 20 and and CCONJ ajst-29929 21 21 parameter parameter NOUN ajst-29929 21 22 - - PUNCT ajst-29929 21 23 optimized optimize VERB ajst-29929 21 24 vmd vmd NOUN ajst-29929 21 25 . . PUNCT ajst-29929 22 1 reference reference NOUN ajst-29929 23 1 [ [ X ajst-29929 23 2 6 6 NUM ajst-29929 23 3 ] ] PUNCT ajst-29929 23 4 employs employ VERB ajst-29929 23 5 vmd vmd PROPN ajst-29929 23 6 to to PART ajst-29929 23 7 extract extract VERB ajst-29929 23 8 features feature NOUN ajst-29929 23 9 from from ADP ajst-29929 23 10 main main ADJ ajst-29929 23 11 bearing bearing NOUN ajst-29929 23 12 vibration vibration NOUN ajst-29929 23 13 data datum NOUN ajst-29929 23 14 and and CCONJ ajst-29929 23 15 then then ADV ajst-29929 23 16 applies apply VERB ajst-29929 23 17 deep deep ADJ ajst-29929 23 18 neural neural ADJ ajst-29929 23 19 networkbased networkbase VERB ajst-29929 23 20 fault fault NOUN ajst-29929 23 21 diagnosis diagnosis NOUN ajst-29929 23 22 methods method NOUN ajst-29929 23 23 . . PUNCT ajst-29929 24 1 reference reference NOUN ajst-29929 24 2 [ [ X ajst-29929 24 3 7 7 X ajst-29929 24 4 ] ] PUNCT ajst-29929 24 5 demonstrates demonstrate VERB ajst-29929 24 6 a a DET ajst-29929 24 7 technique technique NOUN ajst-29929 24 8 for for ADP ajst-29929 24 9 gearbox gearbox NOUN ajst-29929 24 10 fault fault NOUN ajst-29929 24 11 identification identification NOUN ajst-29929 24 12 using use VERB ajst-29929 24 13 adaptive adaptive ADJ ajst-29929 24 14 variational variational ADJ ajst-29929 24 15 mode mode NOUN ajst-29929 24 16 decomposition decomposition NOUN ajst-29929 24 17 . . PUNCT ajst-29929 25 1 reference reference NOUN ajst-29929 25 2 [ [ X ajst-29929 25 3 8 8 NUM ajst-29929 25 4 ] ] PUNCT ajst-29929 25 5 introduces introduce VERB ajst-29929 25 6 a a DET ajst-29929 25 7 method method NOUN ajst-29929 25 8 for for ADP ajst-29929 25 9 early early ADJ ajst-29929 25 10 fault fault NOUN ajst-29929 25 11 detection detection NOUN ajst-29929 25 12 in in ADP ajst-29929 25 13 wind wind NOUN ajst-29929 25 14 turbine turbine NOUN ajst-29929 25 15 gearboxes gearbox NOUN ajst-29929 25 16 using use VERB ajst-29929 25 17 zoom zoom NOUN ajst-29929 25 18 - - PUNCT ajst-29929 25 19 fft fft NOUN ajst-29929 25 20 - - PUNCT ajst-29929 25 21 ceemd ceemd NOUN ajst-29929 25 22 combined combine VERB ajst-29929 25 23 with with ADP ajst-29929 25 24 wavelet wavelet NOUN ajst-29929 25 25 packet packet NOUN ajst-29929 25 26 denoising denoising NOUN ajst-29929 25 27 . . PUNCT ajst-29929 26 1 reference reference NOUN ajst-29929 26 2 [ [ X ajst-29929 26 3 9 9 NUM ajst-29929 26 4 ] ] PUNCT ajst-29929 26 5 applies apply VERB ajst-29929 26 6 hilbert hilbert NOUN ajst-29929 26 7 transform transform NOUN ajst-29929 26 8 combined combine VERB ajst-29929 26 9 with with ADP ajst-29929 26 10 cepstral cepstral ADJ ajst-29929 26 11 analysis analysis NOUN ajst-29929 26 12 to to PART ajst-29929 26 13 process process VERB ajst-29929 26 14 vibration vibration NOUN ajst-29929 26 15 data datum NOUN ajst-29929 26 16 from from ADP ajst-29929 26 17 gearboxes gearbox NOUN ajst-29929 26 18 , , PUNCT ajst-29929 26 19 aiming aim VERB ajst-29929 26 20 to to PART ajst-29929 26 21 identify identify VERB ajst-29929 26 22 and and CCONJ ajst-29929 26 23 extract extract VERB ajst-29929 26 24 fault fault NOUN ajst-29929 26 25 feature feature NOUN ajst-29929 26 26 frequencies frequency NOUN ajst-29929 26 27 for for ADP ajst-29929 26 28 effective effective ADJ ajst-29929 26 29 diagnosis diagnosis NOUN ajst-29929 26 30 of of ADP ajst-29929 26 31 wind wind NOUN ajst-29929 26 32 turbine turbine NOUN ajst-29929 26 33 gearbox gearbox NOUN ajst-29929 26 34 faults fault NOUN ajst-29929 26 35 . . PUNCT ajst-29929 27 1 in in ADP ajst-29929 27 2 contrast contrast NOUN ajst-29929 27 3 , , PUNCT ajst-29929 27 4 reference reference NOUN ajst-29929 27 5 [ [ X ajst-29929 27 6 10 10 NUM ajst-29929 27 7 ] ] PUNCT ajst-29929 27 8 proposes propose VERB ajst-29929 27 9 an an DET ajst-29929 27 10 innovative innovative ADJ ajst-29929 27 11 method method NOUN ajst-29929 27 12 that that PRON ajst-29929 27 13 combines combine VERB ajst-29929 27 14 the the DET ajst-29929 27 15 visualization visualization NOUN ajst-29929 27 16 of of ADP ajst-29929 27 17 vmd vmd PROPN ajst-29929 27 18 signals signal NOUN ajst-29929 27 19 with with ADP ajst-29929 27 20 deep deep ADJ ajst-29929 27 21 learning learn VERB ajst-29929 27 22 neural neural ADJ ajst-29929 27 23 networks network NOUN ajst-29929 27 24 for for ADP ajst-29929 27 25 precise precise ADJ ajst-29929 27 26 bearing bear VERB ajst-29929 27 27 fault fault NOUN ajst-29929 27 28 diagnosis diagnosis NOUN ajst-29929 27 29 . . PUNCT ajst-29929 28 1 with with ADP ajst-29929 28 2 the the DET ajst-29929 28 3 continuous continuous ADJ ajst-29929 28 4 development development NOUN ajst-29929 28 5 of of ADP ajst-29929 28 6 deep deep ADJ ajst-29929 28 7 learning learning NOUN ajst-29929 28 8 , , PUNCT ajst-29929 28 9 various various ADJ ajst-29929 28 10 deep deep ADJ ajst-29929 28 11 learning learning NOUN ajst-29929 28 12 models model NOUN ajst-29929 28 13 have have AUX ajst-29929 28 14 been be AUX ajst-29929 28 15 extensively extensively ADV ajst-29929 28 16 applied apply VERB ajst-29929 28 17 in in ADP ajst-29929 28 18 bearing bear VERB ajst-29929 28 19 fault fault NOUN ajst-29929 28 20 diagnosis diagnosis NOUN ajst-29929 28 21 . . PUNCT ajst-29929 29 1 specifically specifically ADV ajst-29929 29 2 , , PUNCT ajst-29929 29 3 reference reference NOUN ajst-29929 29 4 [ [ X ajst-29929 29 5 11 11 NUM ajst-29929 29 6 ] ] PUNCT ajst-29929 29 7 describes describe VERB ajst-29929 29 8 a a DET ajst-29929 29 9 rolling rolling ADJ ajst-29929 29 10 bearing bearing NOUN ajst-29929 29 11 fault fault NOUN ajst-29929 29 12 recognition recognition NOUN ajst-29929 29 13 strategy strategy NOUN ajst-29929 29 14 that that PRON ajst-29929 29 15 integrates integrate VERB ajst-29929 29 16 ensemble ensemble ADJ ajst-29929 29 17 empirical empirical ADJ ajst-29929 29 18 mode mode NOUN ajst-29929 29 19 decomposition decomposition NOUN ajst-29929 29 20 ( ( PUNCT ajst-29929 29 21 eemd eemd PROPN ajst-29929 29 22 ) ) PUNCT ajst-29929 29 23 with with ADP ajst-29929 29 24 convolutional convolutional ADJ ajst-29929 29 25 neural neural ADJ ajst-29929 29 26 network network NOUN ajst-29929 29 27 - - PUNCT ajst-29929 29 28 support support NOUN ajst-29929 29 29 vector vector NOUN ajst-29929 29 30 machine machine NOUN ajst-29929 29 31 ( ( PUNCT ajst-29929 29 32 cnn cnn NOUN ajst-29929 29 33 - - PUNCT ajst-29929 29 34 svm svm PROPN ajst-29929 29 35 ) ) PUNCT ajst-29929 29 36 . . PUNCT ajst-29929 30 1 reference reference NOUN ajst-29929 30 2 [ [ X ajst-29929 30 3 12 12 NUM ajst-29929 30 4 ] ] PUNCT ajst-29929 30 5 uses use VERB ajst-29929 30 6 backpropagation backpropagation NOUN ajst-29929 30 7 ( ( PUNCT ajst-29929 30 8 bp bp PROPN ajst-29929 30 9 ) ) PUNCT ajst-29929 30 10 neural neural ADJ ajst-29929 30 11 networks network NOUN ajst-29929 30 12 for for ADP ajst-29929 30 13 bearing bear VERB ajst-29929 30 14 fault fault NOUN ajst-29929 30 15 identification identification NOUN ajst-29929 30 16 . . PUNCT ajst-29929 31 1 reference reference NOUN ajst-29929 31 2 [ [ X ajst-29929 31 3 13 13 NUM ajst-29929 31 4 ] ] PUNCT ajst-29929 31 5 introduces introduce VERB ajst-29929 31 6 a a DET ajst-29929 31 7 fault fault NOUN ajst-29929 31 8 diagnosis diagnosis NOUN ajst-29929 31 9 method method NOUN ajst-29929 31 10 based base VERB ajst-29929 31 11 on on ADP ajst-29929 31 12 empirical empirical ADJ ajst-29929 31 13 mode mode NOUN ajst-29929 31 14 decomposition decomposition NOUN ajst-29929 31 15 ( ( PUNCT ajst-29929 31 16 emd emd PROPN ajst-29929 31 17 ) ) PUNCT ajst-29929 31 18 and and CCONJ ajst-29929 31 19 optimized optimize VERB ajst-29929 31 20 bp bp PROPN ajst-29929 31 21 neural neural PROPN ajst-29929 31 22 networks network NOUN ajst-29929 31 23 . . PUNCT ajst-29929 32 1 reference reference NOUN ajst-29929 32 2 [ [ X ajst-29929 32 3 14 14 NUM ajst-29929 32 4 ] ] PUNCT ajst-29929 32 5 proposes propose VERB ajst-29929 32 6 a a DET ajst-29929 32 7 deep deep ADJ ajst-29929 32 8 neural neural ADJ ajst-29929 32 9 network network NOUN ajst-29929 32 10 model model NOUN ajst-29929 32 11 that that PRON ajst-29929 32 12 combines combine VERB ajst-29929 32 13 convolutional convolutional ADJ ajst-29929 32 14 neural neural ADJ ajst-29929 32 15 networks network NOUN ajst-29929 32 16 ( ( PUNCT ajst-29929 32 17 cnn cnn PROPN ajst-29929 32 18 ) ) PUNCT ajst-29929 32 19 , , PUNCT ajst-29929 32 20 bidirectional bidirectional ADJ ajst-29929 32 21 long long ADJ ajst-29929 32 22 short short ADJ ajst-29929 32 23 - - PUNCT ajst-29929 32 24 term term NOUN ajst-29929 32 25 memory memory NOUN ajst-29929 32 26 networks network NOUN ajst-29929 32 27 ( ( PUNCT ajst-29929 32 28 bilstm bilstm NOUN ajst-29929 32 29 ) ) PUNCT ajst-29929 32 30 , , PUNCT ajst-29929 32 31 and and CCONJ ajst-29929 32 32 attention attention NOUN ajst-29929 32 33 mechanism mechanism NOUN ajst-29929 32 34 ( ( PUNCT ajst-29929 32 35 am be AUX ajst-29929 32 36 ) ) PUNCT ajst-29929 32 37 for for ADP ajst-29929 32 38 early early ADJ ajst-29929 32 39 fault fault NOUN ajst-29929 32 40 diagnosis diagnosis NOUN ajst-29929 32 41 . . PUNCT ajst-29929 33 1 reference reference NOUN ajst-29929 33 2 [ [ X ajst-29929 33 3 15 15 NUM ajst-29929 33 4 ] ] PUNCT ajst-29929 33 5 describes describe VERB ajst-29929 33 6 a a DET ajst-29929 33 7 fault fault NOUN ajst-29929 33 8 diagnosis diagnosis NOUN ajst-29929 33 9 technique technique NOUN ajst-29929 33 10 using use VERB ajst-29929 33 11 rvine rvine NOUN ajst-29929 33 12 copula copula NOUN ajst-29929 33 13 models model NOUN ajst-29929 33 14 combined combine VERB ajst-29929 33 15 with with ADP ajst-29929 33 16 dynamic dynamic ADJ ajst-29929 33 17 bayesian bayesian NOUN ajst-29929 33 18 networks network NOUN ajst-29929 33 19 . . PUNCT ajst-29929 34 1 reference reference NOUN ajst-29929 34 2 [ [ X ajst-29929 34 3 16 16 NUM ajst-29929 34 4 ] ] PUNCT ajst-29929 34 5 presents present VERB ajst-29929 34 6 a a DET ajst-29929 34 7 gearbox gearbox ADJ ajst-29929 34 8 fault fault NOUN ajst-29929 34 9 diagnosis diagnosis NOUN ajst-29929 34 10 scheme scheme NOUN ajst-29929 34 11 based base VERB ajst-29929 34 12 on on ADP ajst-29929 34 13 multi multi ADJ ajst-29929 34 14 - - ADJ ajst-29929 34 15 sensor sensor ADJ ajst-29929 34 16 information information NOUN ajst-29929 34 17 fusion fusion NOUN ajst-29929 34 18 and and CCONJ ajst-29929 34 19 dualstream dualstream NOUN ajst-29929 34 20 convolutional convolutional ADJ ajst-29929 34 21 neural neural ADJ ajst-29929 34 22 networks network NOUN ajst-29929 34 23 ( ( PUNCT ajst-29929 34 24 cnn cnn PROPN ajst-29929 34 25 ) ) PUNCT ajst-29929 34 26 . . PUNCT ajst-29929 35 1 to to PART ajst-29929 35 2 enhance enhance VERB ajst-29929 35 3 the the DET ajst-29929 35 4 accuracy accuracy NOUN ajst-29929 35 5 of of ADP ajst-29929 35 6 bearing bear VERB ajst-29929 35 7 fault fault NOUN ajst-29929 35 8 diagnosis diagnosis NOUN ajst-29929 35 9 , , PUNCT ajst-29929 35 10 this this DET ajst-29929 35 11 paper paper NOUN ajst-29929 35 12 proposes propose VERB ajst-29929 35 13 a a DET ajst-29929 35 14 method method NOUN ajst-29929 35 15 that that PRON ajst-29929 35 16 combines combine VERB ajst-29929 35 17 fast fast ADJ ajst-29929 35 18 fourier fourier NOUN ajst-29929 35 19 transform transform NOUN ajst-29929 35 20 ( ( PUNCT ajst-29929 35 21 fft fft PROPN ajst-29929 35 22 ) ) PUNCT ajst-29929 35 23 and and CCONJ ajst-29929 35 24 variational variational ADJ ajst-29929 35 25 mode mode NOUN ajst-29929 35 26 decomposition decomposition NOUN ajst-29929 35 27 ( ( PUNCT ajst-29929 35 28 vmd vmd NOUN ajst-29929 35 29 ) ) PUNCT ajst-29929 35 30 for for ADP ajst-29929 35 31 data data NOUN ajst-29929 35 32 preprocessing preprocessing NOUN ajst-29929 35 33 , , PUNCT ajst-29929 35 34 followed follow VERB ajst-29929 35 35 by by ADP ajst-29929 35 36 a a DET ajst-29929 35 37 deep deep ADJ ajst-29929 35 38 learning learning NOUN ajst-29929 35 39 model model NOUN ajst-29929 35 40 integrating integrate VERB ajst-29929 35 41 cnn cnn PROPN ajst-29929 35 42 , , PUNCT ajst-29929 35 43 bilstm bilstm NOUN ajst-29929 35 44 , , PUNCT ajst-29929 35 45 transformer transformer NOUN ajst-29929 35 46 , , PUNCT ajst-29929 35 47 and and CCONJ ajst-29929 35 48 crossattention crossattention NOUN ajst-29929 35 49 mechanisms mechanism NOUN ajst-29929 35 50 ( ( PUNCT ajst-29929 35 51 cnn cnn NOUN ajst-29929 35 52 - - PUNCT ajst-29929 35 53 bilstm bilstm NOUN ajst-29929 35 54 - - PUNCT ajst-29929 35 55 transformercrossattention transformercrossattention NOUN ajst-29929 35 56 ) ) PUNCT ajst-29929 35 57 . . PUNCT ajst-29929 36 1 initially initially ADV ajst-29929 36 2 , , PUNCT ajst-29929 36 3 one one NUM ajst-29929 36 4 - - PUNCT ajst-29929 36 5 dimensional dimensional ADJ ajst-29929 36 6 fault fault NOUN ajst-29929 36 7 signals signal NOUN ajst-29929 36 8 are be AUX ajst-29929 36 9 192 192 NUM ajst-29929 36 10 preprocessed preprocesse VERB ajst-29929 36 11 using use VERB ajst-29929 36 12 fft fft PROPN ajst-29929 36 13 and and CCONJ ajst-29929 36 14 vmd vmd PROPN ajst-29929 36 15 to to PART ajst-29929 36 16 obtain obtain VERB ajst-29929 36 17 spectral spectral ADJ ajst-29929 36 18 information information NOUN ajst-29929 36 19 and and CCONJ ajst-29929 36 20 multi multi ADJ ajst-29929 36 21 - - ADJ ajst-29929 36 22 scale scale ADJ ajst-29929 36 23 time time NOUN ajst-29929 36 24 - - PUNCT ajst-29929 36 25 domain domain NOUN ajst-29929 36 26 features feature NOUN ajst-29929 36 27 . . PUNCT ajst-29929 37 1 fft fft PROPN ajst-29929 37 2 effectively effectively ADV ajst-29929 37 3 captures capture VERB ajst-29929 37 4 the the DET ajst-29929 37 5 frequency frequency NOUN ajst-29929 37 6 - - PUNCT ajst-29929 37 7 domain domain NOUN ajst-29929 37 8 characteristics characteristic NOUN ajst-29929 37 9 of of ADP ajst-29929 37 10 signals signal NOUN ajst-29929 37 11 , , PUNCT ajst-29929 37 12 while while SCONJ ajst-29929 37 13 vmd vmd PROPN ajst-29929 37 14 aids aids PROPN ajst-29929 37 15 in in ADP ajst-29929 37 16 extracting extract VERB ajst-29929 37 17 multi multi ADJ ajst-29929 37 18 - - ADJ ajst-29929 37 19 scale scale ADJ ajst-29929 37 20 timedomain timedomain NOUN ajst-29929 37 21 features feature NOUN ajst-29929 37 22 . . PUNCT ajst-29929 38 1 the the DET ajst-29929 38 2 combination combination NOUN ajst-29929 38 3 of of ADP ajst-29929 38 4 fft fft PROPN ajst-29929 38 5 and and CCONJ ajst-29929 38 6 vmd vmd PROPN ajst-29929 38 7 allows allow VERB ajst-29929 38 8 for for ADP ajst-29929 38 9 the the DET ajst-29929 38 10 comprehensive comprehensive ADJ ajst-29929 38 11 extraction extraction NOUN ajst-29929 38 12 of of ADP ajst-29929 38 13 useful useful ADJ ajst-29929 38 14 information information NOUN ajst-29929 38 15 from from ADP ajst-29929 38 16 the the DET ajst-29929 38 17 signals signal NOUN ajst-29929 38 18 . . PUNCT ajst-29929 39 1 in in ADP ajst-29929 39 2 the the DET ajst-29929 39 3 feature feature NOUN ajst-29929 39 4 extraction extraction NOUN ajst-29929 39 5 stage stage NOUN ajst-29929 39 6 , , PUNCT ajst-29929 39 7 convolutional convolutional ADJ ajst-29929 39 8 neural neural ADJ ajst-29929 39 9 networks network NOUN ajst-29929 39 10 ( ( PUNCT ajst-29929 39 11 cnn cnn PROPN ajst-29929 39 12 ) ) PUNCT ajst-29929 39 13 and and CCONJ ajst-29929 39 14 transformer transformer NOUN ajst-29929 39 15 networks network NOUN ajst-29929 39 16 are be AUX ajst-29929 39 17 employed employ VERB ajst-29929 39 18 to to PART ajst-29929 39 19 perform perform VERB ajst-29929 39 20 convolution convolution NOUN ajst-29929 39 21 and and CCONJ ajst-29929 39 22 pooling pool VERB ajst-29929 39 23 operations operation NOUN ajst-29929 39 24 on on ADP ajst-29929 39 25 the the DET ajst-29929 39 26 preprocessed preprocesse VERB ajst-29929 39 27 features feature NOUN ajst-29929 39 28 to to PART ajst-29929 39 29 extract extract VERB ajst-29929 39 30 spatial spatial ADJ ajst-29929 39 31 features feature NOUN ajst-29929 39 32 . . PUNCT ajst-29929 40 1 meanwhile meanwhile ADV ajst-29929 40 2 , , PUNCT ajst-29929 40 3 bidirectional bidirectional ADJ ajst-29929 40 4 long long ADJ ajst-29929 40 5 short short ADJ ajst-29929 40 6 - - PUNCT ajst-29929 40 7 term term NOUN ajst-29929 40 8 memory memory NOUN ajst-29929 40 9 networks network NOUN ajst-29929 40 10 ( ( PUNCT ajst-29929 40 11 bilstm bilstm NOUN ajst-29929 40 12 ) ) PUNCT ajst-29929 40 13 are be AUX ajst-29929 40 14 used use VERB ajst-29929 40 15 to to PART ajst-29929 40 16 learn learn VERB ajst-29929 40 17 sequence sequence NOUN ajst-29929 40 18 information information NOUN ajst-29929 40 19 and and CCONJ ajst-29929 40 20 capture capture VERB ajst-29929 40 21 important important ADJ ajst-29929 40 22 temporal temporal ADJ ajst-29929 40 23 features feature NOUN ajst-29929 40 24 . . PUNCT ajst-29929 41 1 the the DET ajst-29929 41 2 introduction introduction NOUN ajst-29929 41 3 of of ADP ajst-29929 41 4 transformer transformer NOUN ajst-29929 41 5 networks network NOUN ajst-29929 41 6 enhances enhance VERB ajst-29929 41 7 the the DET ajst-29929 41 8 features feature NOUN ajst-29929 41 9 extracted extract VERB ajst-29929 41 10 by by ADP ajst-29929 41 11 bilstm bilstm NOUN ajst-29929 41 12 . . PUNCT ajst-29929 42 1 finally finally ADV ajst-29929 42 2 , , PUNCT ajst-29929 42 3 crossattention crossattention NOUN ajst-29929 42 4 mechanisms mechanism NOUN ajst-29929 42 5 are be AUX ajst-29929 42 6 applied apply VERB ajst-29929 42 7 to to PART ajst-29929 42 8 fuse fuse VERB ajst-29929 42 9 the the DET ajst-29929 42 10 features feature NOUN ajst-29929 42 11 , , PUNCT ajst-29929 42 12 further far ADV ajst-29929 42 13 improving improve VERB ajst-29929 42 14 the the DET ajst-29929 42 15 extraction extraction NOUN ajst-29929 42 16 and and CCONJ ajst-29929 42 17 integration integration NOUN ajst-29929 42 18 capabilities capability NOUN ajst-29929 42 19 of of ADP ajst-29929 42 20 the the DET ajst-29929 42 21 model model NOUN ajst-29929 42 22 . . PUNCT ajst-29929 43 1 this this DET ajst-29929 43 2 process process NOUN ajst-29929 43 3 enables enable VERB ajst-29929 43 4 the the DET ajst-29929 43 5 effective effective ADJ ajst-29929 43 6 fusion fusion NOUN ajst-29929 43 7 of of ADP ajst-29929 43 8 timedomain timedomain NOUN ajst-29929 43 9 and and CCONJ ajst-29929 43 10 frequency frequency NOUN ajst-29929 43 11 - - PUNCT ajst-29929 43 12 domain domain NOUN ajst-29929 43 13 features feature NOUN ajst-29929 43 14 , , PUNCT ajst-29929 43 15 thereby thereby ADV ajst-29929 43 16 enhancing enhance VERB ajst-29929 43 17 the the DET ajst-29929 43 18 diagnostic diagnostic ADJ ajst-29929 43 19 performance performance NOUN ajst-29929 43 20 and and CCONJ ajst-29929 43 21 generalization generalization NOUN ajst-29929 43 22 ability ability NOUN ajst-29929 43 23 of of ADP ajst-29929 43 24 the the DET ajst-29929 43 25 model model NOUN ajst-29929 43 26 . . PUNCT ajst-29929 44 1 2 2 X ajst-29929 44 2 . . X ajst-29929 44 3 design design NOUN ajst-29929 44 4 of of ADP ajst-29929 44 5 the the DET ajst-29929 44 6 cnn cnn PROPN ajst-29929 44 7 - - PUNCT ajst-29929 44 8 bilstmtransformer bilstmtransformer PROPN ajst-29929 44 9 model model NOUN ajst-29929 44 10 before before ADP ajst-29929 44 11 feeding feed VERB ajst-29929 44 12 one one NUM ajst-29929 44 13 - - PUNCT ajst-29929 44 14 dimensional dimensional ADJ ajst-29929 44 15 fault fault NOUN ajst-29929 44 16 signals signal NOUN ajst-29929 44 17 into into ADP ajst-29929 44 18 the the DET ajst-29929 44 19 model model NOUN ajst-29929 44 20 , , PUNCT ajst-29929 44 21 they they PRON ajst-29929 44 22 must must AUX ajst-29929 44 23 undergo undergo VERB ajst-29929 44 24 preprocessing preprocessing NOUN ajst-29929 44 25 steps step NOUN ajst-29929 44 26 . . PUNCT ajst-29929 45 1 these these DET ajst-29929 45 2 steps step NOUN ajst-29929 45 3 typically typically ADV ajst-29929 45 4 involve involve VERB ajst-29929 45 5 decomposing decompose VERB ajst-29929 45 6 the the DET ajst-29929 45 7 signals signal NOUN ajst-29929 45 8 in in ADP ajst-29929 45 9 both both CCONJ ajst-29929 45 10 the the DET ajst-29929 45 11 time time NOUN ajst-29929 45 12 and and CCONJ ajst-29929 45 13 frequency frequency NOUN ajst-29929 45 14 domains domain NOUN ajst-29929 45 15 to to PART ajst-29929 45 16 extract extract VERB ajst-29929 45 17 relevant relevant ADJ ajst-29929 45 18 feature feature NOUN ajst-29929 45 19 information information NOUN ajst-29929 45 20 . . PUNCT ajst-29929 46 1 in in ADP ajst-29929 46 2 the the DET ajst-29929 46 3 frequency frequency NOUN ajst-29929 46 4 domain domain NOUN ajst-29929 46 5 , , PUNCT ajst-29929 46 6 although although SCONJ ajst-29929 46 7 fast fast ADJ ajst-29929 46 8 fourier fourier NOUN ajst-29929 46 9 transform transform NOUN ajst-29929 46 10 ( ( PUNCT ajst-29929 46 11 fft fft PROPN ajst-29929 46 12 ) ) PUNCT ajst-29929 46 13 can can AUX ajst-29929 46 14 rapidly rapidly ADV ajst-29929 46 15 extract extract VERB ajst-29929 46 16 the the DET ajst-29929 46 17 spectral spectral ADJ ajst-29929 46 18 features feature NOUN ajst-29929 46 19 of of ADP ajst-29929 46 20 vibration vibration NOUN ajst-29929 46 21 signals signal NOUN ajst-29929 46 22 and and CCONJ ajst-29929 46 23 identify identify VERB ajst-29929 46 24 the the DET ajst-29929 46 25 main main ADJ ajst-29929 46 26 frequency frequency NOUN ajst-29929 46 27 components component NOUN ajst-29929 46 28 , , PUNCT ajst-29929 46 29 it it PRON ajst-29929 46 30 only only ADV ajst-29929 46 31 provides provide VERB ajst-29929 46 32 global global ADJ ajst-29929 46 33 spectral spectral ADJ ajst-29929 46 34 information information NOUN ajst-29929 46 35 and and CCONJ ajst-29929 46 36 is be AUX ajst-29929 46 37 unable unable ADJ ajst-29929 46 38 to to PART ajst-29929 46 39 capture capture VERB ajst-29929 46 40 the the DET ajst-29929 46 41 local local ADJ ajst-29929 46 42 time time NOUN ajst-29929 46 43 - - PUNCT ajst-29929 46 44 varying vary VERB ajst-29929 46 45 characteristics characteristic NOUN ajst-29929 46 46 of of ADP ajst-29929 46 47 the the DET ajst-29929 46 48 signals signal NOUN ajst-29929 46 49 . . PUNCT ajst-29929 47 1 variational variational ADJ ajst-29929 47 2 mode mode NOUN ajst-29929 47 3 decomposition decomposition NOUN ajst-29929 47 4 ( ( PUNCT ajst-29929 47 5 vmd vmd NOUN ajst-29929 47 6 ) ) PUNCT ajst-29929 47 7 , , PUNCT ajst-29929 47 8 as as ADP ajst-29929 47 9 an an DET ajst-29929 47 10 adaptive adaptive ADJ ajst-29929 47 11 signal signal NOUN ajst-29929 47 12 decomposition decomposition NOUN ajst-29929 47 13 technique technique NOUN ajst-29929 47 14 , , PUNCT ajst-29929 47 15 can can AUX ajst-29929 47 16 effectively effectively ADV ajst-29929 47 17 decompose decompose VERB ajst-29929 47 18 complex complex ADJ ajst-29929 47 19 vibration vibration NOUN ajst-29929 47 20 signals signal NOUN ajst-29929 47 21 into into ADP ajst-29929 47 22 multiple multiple ADJ ajst-29929 47 23 modal modal ADJ ajst-29929 47 24 components component NOUN ajst-29929 47 25 with with ADP ajst-29929 47 26 clear clear ADJ ajst-29929 47 27 physical physical ADJ ajst-29929 47 28 meanings meaning NOUN ajst-29929 47 29 . . PUNCT ajst-29929 48 1 each each DET ajst-29929 48 2 modal modal ADJ ajst-29929 48 3 component component NOUN ajst-29929 48 4 reflects reflect VERB ajst-29929 48 5 the the DET ajst-29929 48 6 signal signal NOUN ajst-29929 48 7 's 's PART ajst-29929 48 8 characteristics characteristic NOUN ajst-29929 48 9 at at ADP ajst-29929 48 10 different different ADJ ajst-29929 48 11 time time NOUN ajst-29929 48 12 scales scale NOUN ajst-29929 48 13 . . PUNCT ajst-29929 49 1 by by ADP ajst-29929 49 2 combining combine VERB ajst-29929 49 3 fft fft PROPN ajst-29929 49 4 and and CCONJ ajst-29929 49 5 vmd vmd PROPN ajst-29929 49 6 , , PUNCT ajst-29929 49 7 it it PRON ajst-29929 49 8 is be AUX ajst-29929 49 9 possible possible ADJ ajst-29929 49 10 to to PART ajst-29929 49 11 obtain obtain VERB ajst-29929 49 12 both both PRON ajst-29929 49 13 global global ADJ ajst-29929 49 14 spectral spectral ADJ ajst-29929 49 15 information information NOUN ajst-29929 49 16 and and CCONJ ajst-29929 49 17 multi multi ADJ ajst-29929 49 18 - - ADJ ajst-29929 49 19 scale scale ADJ ajst-29929 49 20 local local ADJ ajst-29929 49 21 time time NOUN ajst-29929 49 22 - - PUNCT ajst-29929 49 23 domain domain NOUN ajst-29929 49 24 features feature NOUN ajst-29929 49 25 , , PUNCT ajst-29929 49 26 thereby thereby ADV ajst-29929 49 27 achieving achieve VERB ajst-29929 49 28 comprehensive comprehensive ADJ ajst-29929 49 29 feature feature NOUN ajst-29929 49 30 extraction extraction NOUN ajst-29929 49 31 from from ADP ajst-29929 49 32 the the DET ajst-29929 49 33 signals signal NOUN ajst-29929 49 34 . . PUNCT ajst-29929 50 1 this this DET ajst-29929 50 2 multi multi ADJ ajst-29929 50 3 - - ADJ ajst-29929 50 4 scale scale ADJ ajst-29929 50 5 feature feature NOUN ajst-29929 50 6 extraction extraction NOUN ajst-29929 50 7 method method NOUN ajst-29929 50 8 can can AUX ajst-29929 50 9 more more ADV ajst-29929 50 10 accurately accurately ADV ajst-29929 50 11 capture capture VERB ajst-29929 50 12 the the DET ajst-29929 50 13 complex complex ADJ ajst-29929 50 14 features feature NOUN ajst-29929 50 15 of of ADP ajst-29929 50 16 rolling roll VERB ajst-29929 50 17 bearing bear VERB ajst-29929 50 18 fault fault NOUN ajst-29929 50 19 signals signal NOUN ajst-29929 50 20 , , PUNCT ajst-29929 50 21 thereby thereby ADV ajst-29929 50 22 enhancing enhance VERB ajst-29929 50 23 the the DET ajst-29929 50 24 accuracy accuracy NOUN ajst-29929 50 25 and and CCONJ ajst-29929 50 26 robustness robustness NOUN ajst-29929 50 27 of of ADP ajst-29929 50 28 fault fault NOUN ajst-29929 50 29 diagnosis diagnosis NOUN ajst-29929 50 30 . . PUNCT ajst-29929 51 1 moreover moreover ADV ajst-29929 51 2 , , PUNCT ajst-29929 51 3 the the DET ajst-29929 51 4 integration integration NOUN ajst-29929 51 5 of of ADP ajst-29929 51 6 frequencydomain frequencydomain ADJ ajst-29929 51 7 and and CCONJ ajst-29929 51 8 time time NOUN ajst-29929 51 9 - - PUNCT ajst-29929 51 10 domain domain NOUN ajst-29929 51 11 features feature NOUN ajst-29929 51 12 enables enable VERB ajst-29929 51 13 better well ADJ ajst-29929 51 14 identification identification NOUN ajst-29929 51 15 of of ADP ajst-29929 51 16 different different ADJ ajst-29929 51 17 types type NOUN ajst-29929 51 18 and and CCONJ ajst-29929 51 19 severity severity NOUN ajst-29929 51 20 levels level NOUN ajst-29929 51 21 of of ADP ajst-29929 51 22 faults fault NOUN ajst-29929 51 23 , , PUNCT ajst-29929 51 24 meeting meet VERB ajst-29929 51 25 the the DET ajst-29929 51 26 diagnostic diagnostic ADJ ajst-29929 51 27 requirements requirement NOUN ajst-29929 51 28 of of ADP ajst-29929 51 29 complex complex ADJ ajst-29929 51 30 industrial industrial ADJ ajst-29929 51 31 environments environment NOUN ajst-29929 51 32 . . PUNCT ajst-29929 52 1 based base VERB ajst-29929 52 2 on on ADP ajst-29929 52 3 the the DET ajst-29929 52 4 initial initial ADJ ajst-29929 52 5 signal signal NOUN ajst-29929 52 6 processing processing NOUN ajst-29929 52 7 , , PUNCT ajst-29929 52 8 the the DET ajst-29929 52 9 decomposed decompose VERB ajst-29929 52 10 results result NOUN ajst-29929 52 11 are be AUX ajst-29929 52 12 superimposed superimpose VERB ajst-29929 52 13 and and CCONJ ajst-29929 52 14 processed process VERB ajst-29929 52 15 using use VERB ajst-29929 52 16 two two NUM ajst-29929 52 17 network network NOUN ajst-29929 52 18 structures structure NOUN ajst-29929 52 19 : : PUNCT ajst-29929 52 20 cnn cnn PROPN ajst-29929 52 21 - - PUNCT ajst-29929 52 22 transformer transformer NOUN ajst-29929 52 23 and and CCONJ ajst-29929 52 24 bilstm bilstm NOUN ajst-29929 52 25 - - PUNCT ajst-29929 52 26 transformer transformer NOUN ajst-29929 52 27 , , PUNCT ajst-29929 52 28 to to PART ajst-29929 52 29 extract extract VERB ajst-29929 52 30 features feature NOUN ajst-29929 52 31 in in ADP ajst-29929 52 32 both both CCONJ ajst-29929 52 33 the the DET ajst-29929 52 34 spatial spatial ADJ ajst-29929 52 35 and and CCONJ ajst-29929 52 36 temporal temporal ADJ ajst-29929 52 37 domains domain NOUN ajst-29929 52 38 . . PUNCT ajst-29929 53 1 subsequently subsequently ADV ajst-29929 53 2 , , PUNCT ajst-29929 53 3 cross cross ADJ ajst-29929 53 4 - - ADJ ajst-29929 53 5 attention attention ADJ ajst-29929 53 6 mechanisms mechanism NOUN ajst-29929 53 7 are be AUX ajst-29929 53 8 utilized utilize VERB ajst-29929 53 9 to to PART ajst-29929 53 10 integrate integrate VERB ajst-29929 53 11 the the DET ajst-29929 53 12 time time NOUN ajst-29929 53 13 - - PUNCT ajst-29929 53 14 domain domain NOUN ajst-29929 53 15 and and CCONJ ajst-29929 53 16 frequency frequency NOUN ajst-29929 53 17 - - PUNCT ajst-29929 53 18 domain domain NOUN ajst-29929 53 19 features feature NOUN ajst-29929 53 20 . . PUNCT ajst-29929 54 1 in in ADP ajst-29929 54 2 this this DET ajst-29929 54 3 manner manner NOUN ajst-29929 54 4 , , PUNCT ajst-29929 54 5 the the DET ajst-29929 54 6 model model NOUN ajst-29929 54 7 can can AUX ajst-29929 54 8 compute compute VERB ajst-29929 54 9 and and CCONJ ajst-29929 54 10 assign assign VERB ajst-29929 54 11 appropriate appropriate ADJ ajst-29929 54 12 attention attention NOUN ajst-29929 54 13 weights weight NOUN ajst-29929 54 14 to to ADP ajst-29929 54 15 different different ADJ ajst-29929 54 16 features feature NOUN ajst-29929 54 17 , , PUNCT ajst-29929 54 18 allowing allow VERB ajst-29929 54 19 it it PRON ajst-29929 54 20 to to PART ajst-29929 54 21 focus focus VERB ajst-29929 54 22 more more ADV ajst-29929 54 23 on on ADP ajst-29929 54 24 key key ADJ ajst-29929 54 25 features feature NOUN ajst-29929 54 26 and and CCONJ ajst-29929 54 27 thereby thereby ADV ajst-29929 54 28 enhancing enhance VERB ajst-29929 54 29 its its PRON ajst-29929 54 30 overall overall ADJ ajst-29929 54 31 performance performance NOUN ajst-29929 54 32 and and CCONJ ajst-29929 54 33 generalization generalization NOUN ajst-29929 54 34 ability ability NOUN ajst-29929 54 35 . . PUNCT ajst-29929 55 1 the the DET ajst-29929 55 2 model model NOUN ajst-29929 55 3 structure structure NOUN ajst-29929 55 4 is be AUX ajst-29929 55 5 shown show VERB ajst-29929 55 6 in in ADP ajst-29929 55 7 figure figure NOUN ajst-29929 55 8 1 1 NUM ajst-29929 55 9 . . PUNCT ajst-29929 55 10 figure figure NOUN ajst-29929 55 11 1 1 NUM ajst-29929 55 12 . . PUNCT ajst-29929 55 13 model model NOUN ajst-29929 55 14 architecture architecture NOUN ajst-29929 55 15 diagram diagram NOUN ajst-29929 55 16 2.1 2.1 NUM ajst-29929 55 17 . . PUNCT ajst-29929 56 1 convolutional convolutional ADJ ajst-29929 56 2 neural neural ADJ ajst-29929 56 3 networks network NOUN ajst-29929 56 4 ( ( PUNCT ajst-29929 56 5 cnn cnn NOUN ajst-29929 56 6 ) ) PUNCT ajst-29929 56 7 convolutional convolutional ADJ ajst-29929 56 8 neural neural ADJ ajst-29929 56 9 networks network NOUN ajst-29929 56 10 ( ( PUNCT ajst-29929 56 11 cnns cnns PROPN ajst-29929 56 12 ) ) PUNCT ajst-29929 56 13 are be AUX ajst-29929 56 14 a a DET ajst-29929 56 15 widely widely ADV ajst-29929 56 16 u u NOUN ajst-29929 56 17 sed se VERB ajst-29929 56 18 deep deep ADJ ajst-29929 56 19 learning learning NOUN ajst-29929 56 20 architecture architecture NOUN ajst-29929 56 21 in in ADP ajst-29929 56 22 the the DET ajst-29929 56 23 field field NOUN ajst-29929 56 24 of of ADP ajst-29929 56 25 signal signal ADJ ajst-29929 56 26 proce proce PROPN ajst-29929 56 27 ssing ssing NOUN ajst-29929 56 28 . . PUNCT ajst-29929 57 1 the the DET ajst-29929 57 2 structure structure NOUN ajst-29929 57 3 of of ADP ajst-29929 57 4 the the DET ajst-29929 57 5 model model NOUN ajst-29929 57 6 is be AUX ajst-29929 57 7 shown show VERB ajst-29929 57 8 in in ADP ajst-29929 57 9 figure figure NOUN ajst-29929 57 10 2 2 NUM ajst-29929 57 11 . . PUNCT ajst-29929 58 1 t t NOUN ajst-29929 58 2 he he PRON ajst-29929 58 3 core core NOUN ajst-29929 58 4 concept concept NOUN ajst-29929 58 5 of of ADP ajst-29929 58 6 cnns cnns PROPN ajst-29929 58 7 is be AUX ajst-29929 58 8 to to PART ajst-29929 58 9 extract extract VERB ajst-29929 58 10 local local ADJ ajst-29929 58 11 features feature NOUN ajst-29929 58 12 from from ADP ajst-29929 58 13 input input NOUN ajst-29929 58 14 data datum NOUN ajst-29929 58 15 through through ADP ajst-29929 58 16 convolution convolution NOUN ajst-29929 58 17 operations operation NOUN ajst-29929 58 18 and and CCONJ ajst-29929 58 19 to to PART ajst-29929 58 20 progres progre VERB ajst-29929 58 21 sively sively ADV ajst-29929 58 22 abstract abstract ADJ ajst-29929 58 23 and and CCONJ ajst-29929 58 24 integrate integrate VERB ajst-29929 58 25 these these DET ajst-29929 58 26 features feature NOUN ajst-29929 58 27 via via ADP ajst-29929 58 28 multiple multiple ADJ ajst-29929 58 29 co co ADJ ajst-29929 58 30 nvolutional nvolutional ADJ ajst-29929 58 31 layers layer NOUN ajst-29929 58 32 , , PUNCT ajst-29929 58 33 thereby thereby ADV ajst-29929 58 34 obtaining obtain VERB ajst-29929 58 35 high high ADJ ajst-29929 58 36 - - PUNCT ajst-29929 58 37 level level NOUN ajst-29929 58 38 representat representat NOUN ajst-29929 58 39 ions ion NOUN ajst-29929 58 40 of of ADP ajst-29929 58 41 the the DET ajst-29929 58 42 input input NOUN ajst-29929 58 43 data datum NOUN ajst-29929 58 44 . . PUNCT ajst-29929 59 1 in in ADP ajst-29929 59 2 the the DET ajst-29929 59 3 context context NOUN ajst-29929 59 4 of of ADP ajst-29929 59 5 rolling rolling ADJ ajst-29929 59 6 bearing bear VERB ajst-29929 59 7 fa fa PROPN ajst-29929 59 8 ult ult PROPN ajst-29929 59 9 diagnosis diagnosis NOUN ajst-29929 59 10 , , PUNCT ajst-29929 59 11 the the DET ajst-29929 59 12 spectral spectral ADJ ajst-29929 59 13 information information NOUN ajst-29929 59 14 and and CCONJ ajst-29929 59 15 multi multi ADJ ajst-29929 59 16 - - NOUN ajst-29929 59 17 scale scale ADJ ajst-29929 59 18 tim tim PROPN ajst-29929 59 19 e e NOUN ajst-29929 59 20 - - NOUN ajst-29929 59 21 domain domain ADJ ajst-29929 59 22 features feature NOUN ajst-29929 59 23 of of ADP ajst-29929 59 24 vibration vibration NOUN ajst-29929 59 25 signals signal NOUN ajst-29929 59 26 can can AUX ajst-29929 59 27 be be AUX ajst-29929 59 28 regarded regard VERB ajst-29929 59 29 as as ADP ajst-29929 59 30 two two NUM ajst-29929 59 31 - - PUNCT ajst-29929 59 32 dimensional dimensional ADJ ajst-29929 59 33 or or CCONJ ajst-29929 59 34 one one NUM ajst-29929 59 35 - - PUNCT ajst-29929 59 36 dimensional dimensional ADJ ajst-29929 59 37 signal signal NOUN ajst-29929 59 38 images image NOUN ajst-29929 59 39 . . PUNCT ajst-29929 60 1 by by ADP ajst-29929 60 2 sli sli NOUN ajst-29929 60 3 ding de VERB ajst-29929 60 4 convolutional convolutional ADJ ajst-29929 60 5 kernels kernel NOUN ajst-29929 60 6 over over ADP ajst-29929 60 7 the the DET ajst-29929 60 8 input input NOUN ajst-29929 60 9 data datum NOUN ajst-29929 60 10 , , PUNCT ajst-29929 60 11 cnns cnn NOUN ajst-29929 60 12 can can AUX ajst-29929 60 13 efficiently efficiently ADV ajst-29929 60 14 capture capture VERB ajst-29929 60 15 and and CCONJ ajst-29929 60 16 extract extract VERB ajst-29929 60 17 spatial spatial ADJ ajst-29929 60 18 feature feature NOUN ajst-29929 60 19 information information NOUN ajst-29929 60 20 from from ADP ajst-29929 60 21 the the DET ajst-29929 60 22 signals signal NOUN ajst-29929 60 23 . . PUNCT ajst-29929 61 1 one‐ one‐ ADP ajst-29929 61 2 dimen diman NOUN ajst-29929 61 3 sional sional ADJ ajst-29929 61 4     SPACE ajst-29929 61 5 signal signal NOUN ajst-29929 61 6     SPACE ajst-29929 61 7 input input NOUN ajst-29929 61 8 fft fft PROPN ajst-29929 61 9 vmd vmd PROPN ajst-29929 61 10 cnn cnn PROPN ajst-29929 61 11 bilstm bilstm PROPN ajst-29929 61 12 transfo transfo PROPN ajst-29929 61 13 rmer rmer PROPN ajst-29929 61 14 output output PROPN ajst-29929 61 15 cross cross PROPN ajst-29929 61 16     SPACE ajst-29929 61 17 attention attention NOUN ajst-29929 61 18 input input NOUN ajst-29929 61 19     SPACE ajst-29929 61 20 layer layer NOUN ajst-29929 61 21 con con PROPN ajst-29929 61 22 volu volu PROPN ajst-29929 61 23 tion tion PROPN ajst-29929 61 24     SPACE ajst-29929 61 25 laye laye NOUN ajst-29929 61 26 r r NOUN ajst-29929 61 27 pool pool NOUN ajst-29929 61 28 ed ed NOUN ajst-29929 61 29     SPACE ajst-29929 61 30 hori hori PROPN ajst-29929 61 31 zon zon PROPN ajst-29929 61 32 con con PROPN ajst-29929 61 33 volu volu PROPN ajst-29929 61 34 tion tion PROPN ajst-29929 61 35     SPACE ajst-29929 61 36 laye laye NOUN ajst-29929 61 37 r r NOUN ajst-29929 61 38 pool pool NOUN ajst-29929 61 39 ed ed NOUN ajst-29929 61 40     SPACE ajst-29929 61 41 hori hori PROPN ajst-29929 61 42 zon zon PROPN ajst-29929 61 43 fully fully ADV ajst-29929 61 44     SPACE ajst-29929 61 45 conne conne PROPN ajst-29929 61 46 cted cte VERB ajst-29929 61 47     SPACE ajst-29929 61 48 layer layer NOUN ajst-29929 61 49 three three NUM ajst-29929 61 50     SPACE ajst-29929 61 51 featur featur NOUN ajst-29929 61 52 e e NOUN ajst-29929 61 53     SPACE ajst-29929 61 54 maps map VERB ajst-29929 61 55 three three NUM ajst-29929 61 56     SPACE ajst-29929 61 57 featur featur ADJ ajst-29929 61 58 e e NOUN ajst-29929 61 59     SPACE ajst-29929 61 60 maps map VERB ajst-29929 61 61 five five NUM ajst-29929 61 62     SPACE ajst-29929 61 63 featur featur NOUN ajst-29929 61 64 e e NOUN ajst-29929 61 65     SPACE ajst-29929 61 66 maps map VERB ajst-29929 61 67 five five NUM ajst-29929 61 68     SPACE ajst-29929 61 69 featur featur NOUN ajst-29929 61 70 e e NOUN ajst-29929 61 71     SPACE ajst-29929 61 72 maps map VERB ajst-29929 61 73 outpu outpu NOUN ajst-29929 61 74 t t PROPN ajst-29929 61 75     SPACE ajst-29929 61 76 layer layer NOUN ajst-29929 61 77 lstm lstm PROPN ajst-29929 61 78 lstm lstm PROPN ajst-29929 61 79 lstmlstm lstmlstm PROPN ajst-29929 61 80 lstm lstm PROPN ajst-29929 61 81 lstm lstm PROPN ajst-29929 61 82 at‐1 at‐1 PROPN ajst-29929 61 83 a a DET ajst-29929 61 84 at+1 at+1 X ajst-29929 61 85 forw forw NOUN ajst-29929 61 86 ard ard PROPN ajst-29929 61 87     SPACE ajst-29929 61 88 revers rever NOUN ajst-29929 61 89 e e X ajst-29929 61 90 xt‐1 xt‐1 PROPN ajst-29929 61 91 xt+1x xt+1x PROPN ajst-29929 61 92 input input PROPN ajst-29929 61 93     SPACE ajst-29929 61 94 layer layer NOUN ajst-29929 61 95 outpu outpu NOUN ajst-29929 61 96 t t PROPN ajst-29929 61 97     SPACE ajst-29929 61 98 layer layer NOUN ajst-29929 61 99 spati spati PROPN ajst-29929 61 100 al al PROPN ajst-29929 61 101     SPACE ajst-29929 61 102 featu featu PROPN ajst-29929 61 103 re re ADJ ajst-29929 61 104 time time NOUN ajst-29929 61 105     SPACE ajst-29929 61 106 doma doma NOUN ajst-29929 61 107 in in ADP ajst-29929 61 108     SPACE ajst-29929 61 109 featu featu PROPN ajst-29929 61 110 re re PROPN ajst-29929 61 111 v v ADP ajst-29929 61 112 k k PROPN ajst-29929 61 113 q q PROPN ajst-29929 61 114 193 193 NUM ajst-29929 61 115 figure figure NOUN ajst-29929 61 116 2 2 NUM ajst-29929 61 117 . . PUNCT ajst-29929 61 118 convolutional convolutional ADJ ajst-29929 61 119 neural neural ADJ ajst-29929 61 120 network network NOUN ajst-29929 61 121 model model NOUN ajst-29929 61 122 diagram diagram NOUN ajst-29929 61 123 input input NOUN ajst-29929 61 124 layer layer NOUN ajst-29929 61 125 : : PUNCT ajst-29929 61 126 this this DET ajst-29929 61 127 layer layer NOUN ajst-29929 61 128 accepts accept VERB ajst-29929 61 129 preprocessed preprocesse VERB ajst-29929 61 130 vibration vibration NOUN ajst-29929 61 131 signal signal NOUN ajst-29929 61 132 features feature NOUN ajst-29929 61 133 , , PUNCT ajst-29929 61 134 such such ADJ ajst-29929 61 135 as as ADP ajst-29929 61 136 feature feature NOUN ajst-29929 61 137 maps map NOUN ajst-29929 61 138 obtained obtain VERB ajst-29929 61 139 through through ADP ajst-29929 61 140 fast fast ADJ ajst-29929 61 141 fourier fourier NOUN ajst-29929 61 142 transform transform NOUN ajst-29929 61 143 ( ( PUNCT ajst-29929 61 144 fft fft PROPN ajst-29929 61 145 ) ) PUNCT ajst-29929 61 146 and and CCONJ ajst-29929 61 147 variational variational ADJ ajst-29929 61 148 mode mode NOUN ajst-29929 61 149 decomposition decomposition NOUN ajst-29929 61 150 ( ( PUNCT ajst-29929 61 151 vmd vmd NOUN ajst-29929 61 152 ) ) PUNCT ajst-29929 61 153 . . PUNCT ajst-29929 62 1 convolutional convolutional ADJ ajst-29929 62 2 layer layer NOUN ajst-29929 62 3 : : PUNCT ajst-29929 62 4 multiple multiple ADJ ajst-29929 62 5 convolutional convolutional ADJ ajst-29929 62 6 kernels kernel NOUN ajst-29929 62 7 are be AUX ajst-29929 62 8 applied apply VERB ajst-29929 62 9 to to ADP ajst-29929 62 10 the the DET ajst-29929 62 11 input input NOUN ajst-29929 62 12 features feature VERB ajst-29929 62 13 to to PART ajst-29929 62 14 perform perform VERB ajst-29929 62 15 convolution convolution NOUN ajst-29929 62 16 operations operation NOUN ajst-29929 62 17 . . PUNCT ajst-29929 63 1 this this DET ajst-29929 63 2 process process NOUN ajst-29929 63 3 extracts extract VERB ajst-29929 63 4 different different ADJ ajst-29929 63 5 local local ADJ ajst-29929 63 6 features feature NOUN ajst-29929 63 7 from from ADP ajst-29929 63 8 the the DET ajst-29929 63 9 input input NOUN ajst-29929 63 10 data datum NOUN ajst-29929 63 11 , , PUNCT ajst-29929 63 12 capturing capture VERB ajst-29929 63 13 the the DET ajst-29929 63 14 spatial spatial ADJ ajst-29929 63 15 characteristics characteristic NOUN ajst-29929 63 16 of of ADP ajst-29929 63 17 the the DET ajst-29929 63 18 signals signal NOUN ajst-29929 63 19 . . PUNCT ajst-29929 64 1 pooling pool VERB ajst-29929 64 2 layer layer NOUN ajst-29929 64 3 : : PUNCT ajst-29929 64 4 this this DET ajst-29929 64 5 layer layer NOUN ajst-29929 64 6 reduces reduce VERB ajst-29929 64 7 the the DET ajst-29929 64 8 dimensionality dimensionality NOUN ajst-29929 64 9 and and CCONJ ajst-29929 64 10 computational computational ADJ ajst-29929 64 11 complexity complexity NOUN ajst-29929 64 12 of of ADP ajst-29929 64 13 the the DET ajst-29929 64 14 data datum NOUN ajst-29929 64 15 by by ADP ajst-29929 64 16 employing employ VERB ajst-29929 64 17 techniques technique NOUN ajst-29929 64 18 such such ADJ ajst-29929 64 19 as as ADP ajst-29929 64 20 max max PROPN ajst-29929 64 21 pooling pooling NOUN ajst-29929 64 22 or or CCONJ ajst-29929 64 23 average average ADJ ajst-29929 64 24 pooling pooling NOUN ajst-29929 64 25 while while SCONJ ajst-29929 64 26 retaining retain VERB ajst-29929 64 27 key key ADJ ajst-29929 64 28 features feature NOUN ajst-29929 64 29 . . PUNCT ajst-29929 65 1 the the DET ajst-29929 65 2 pooling pool VERB ajst-29929 65 3 operation operation NOUN ajst-29929 65 4 helps help VERB ajst-29929 65 5 to to PART ajst-29929 65 6 maintain maintain VERB ajst-29929 65 7 the the DET ajst-29929 65 8 essential essential ADJ ajst-29929 65 9 information information NOUN ajst-29929 65 10 while while SCONJ ajst-29929 65 11 reducing reduce VERB ajst-29929 65 12 the the DET ajst-29929 65 13 spatial spatial ADJ ajst-29929 65 14 size size NOUN ajst-29929 65 15 of of ADP ajst-29929 65 16 the the DET ajst-29929 65 17 feature feature NOUN ajst-29929 65 18 maps map NOUN ajst-29929 65 19 . . PUNCT ajst-29929 66 1 fully fully ADV ajst-29929 66 2 connected connected ADJ ajst-29929 66 3 layer layer NOUN ajst-29929 66 4 : : PUNCT ajst-29929 66 5 this this DET ajst-29929 66 6 layer layer NOUN ajst-29929 66 7 transforms transform VERB ajst-29929 66 8 the the DET ajst-29929 66 9 highlevel highlevel ADJ ajst-29929 66 10 features feature NOUN ajst-29929 66 11 extracted extract VERB ajst-29929 66 12 by by ADP ajst-29929 66 13 the the DET ajst-29929 66 14 convolutional convolutional ADJ ajst-29929 66 15 and and CCONJ ajst-29929 66 16 pooling pool VERB ajst-29929 66 17 layers layer NOUN ajst-29929 66 18 into into ADP ajst-29929 66 19 a a DET ajst-29929 66 20 classification classification NOUN ajst-29929 66 21 space space NOUN ajst-29929 66 22 . . PUNCT ajst-29929 67 1 it it PRON ajst-29929 67 2 plays play VERB ajst-29929 67 3 a a DET ajst-29929 67 4 crucial crucial ADJ ajst-29929 67 5 role role NOUN ajst-29929 67 6 in in ADP ajst-29929 67 7 mapping map VERB ajst-29929 67 8 the the DET ajst-29929 67 9 extracted extract VERB ajst-29929 67 10 features feature NOUN ajst-29929 67 11 to to ADP ajst-29929 67 12 the the DET ajst-29929 67 13 output output NOUN ajst-29929 67 14 classes class NOUN ajst-29929 67 15 , , PUNCT ajst-29929 67 16 enabling enable VERB ajst-29929 67 17 the the DET ajst-29929 67 18 final final ADJ ajst-29929 67 19 fault fault NOUN ajst-29929 67 20 classification classification NOUN ajst-29929 67 21 task task NOUN ajst-29929 67 22 . . PUNCT ajst-29929 68 1 output output NOUN ajst-29929 68 2 layer layer NOUN ajst-29929 68 3 : : PUNCT ajst-29929 68 4 the the DET ajst-29929 68 5 output output NOUN ajst-29929 68 6 layer layer NOUN ajst-29929 68 7 utilizes utilize VERB ajst-29929 68 8 the the DET ajst-29929 68 9 softmax softmax NOUN ajst-29929 68 10 function function NOUN ajst-29929 68 11 to to PART ajst-29929 68 12 output output VERB ajst-29929 68 13 the the DET ajst-29929 68 14 probabilities probability NOUN ajst-29929 68 15 of of ADP ajst-29929 68 16 different different ADJ ajst-29929 68 17 fault fault NOUN ajst-29929 68 18 categories category NOUN ajst-29929 68 19 , , PUNCT ajst-29929 68 20 thereby thereby ADV ajst-29929 68 21 realizing realize VERB ajst-29929 68 22 fault fault NOUN ajst-29929 68 23 classification classification NOUN ajst-29929 68 24 . . PUNCT ajst-29929 69 1 the the DET ajst-29929 69 2 softmax softmax NOUN ajst-29929 69 3 function function NOUN ajst-29929 69 4 normalizes normalize VERB ajst-29929 69 5 the the DET ajst-29929 69 6 output output NOUN ajst-29929 69 7 values value NOUN ajst-29929 69 8 into into ADP ajst-29929 69 9 a a DET ajst-29929 69 10 probability probability NOUN ajst-29929 69 11 distribution distribution NOUN ajst-29929 69 12 , , PUNCT ajst-29929 69 13 making make VERB ajst-29929 69 14 it it PRON ajst-29929 69 15 suitable suitable ADJ ajst-29929 69 16 for for ADP ajst-29929 69 17 multi multi ADJ ajst-29929 69 18 - - ADJ ajst-29929 69 19 class class ADJ ajst-29929 69 20 classification classification NOUN ajst-29929 69 21 tasks task NOUN ajst-29929 69 22 . . PUNCT ajst-29929 70 1 this this DET ajst-29929 70 2 architecture architecture NOUN ajst-29929 70 3 effectively effectively ADV ajst-29929 70 4 extracts extract NOUN ajst-29929 70 5 and and CCONJ ajst-29929 70 6 integrates integrate NOUN ajst-29929 70 7 the the DET ajst-29929 70 8 spatial spatial ADJ ajst-29929 70 9 features feature NOUN ajst-29929 70 10 of of ADP ajst-29929 70 11 vibration vibration NOUN ajst-29929 70 12 signals signal NOUN ajst-29929 70 13 , , PUNCT ajst-29929 70 14 thereby thereby ADV ajst-29929 70 15 enhancing enhance VERB ajst-29929 70 16 the the DET ajst-29929 70 17 accuracy accuracy NOUN ajst-29929 70 18 of of ADP ajst-29929 70 19 fault fault NOUN ajst-29929 70 20 diagnosis diagnosis NOUN ajst-29929 70 21 . . PUNCT ajst-29929 71 1 2.2 2.2 NUM ajst-29929 71 2 . . PUNCT ajst-29929 72 1 bidirectional bidirectional ADJ ajst-29929 72 2 long long ADJ ajst-29929 72 3 short short ADJ ajst-29929 72 4 - - PUNCT ajst-29929 72 5 term term NOUN ajst-29929 72 6 memory memory NOUN ajst-29929 72 7 network network NOUN ajst-29929 72 8 ( ( PUNCT ajst-29929 72 9 bilstm bilstm NOUN ajst-29929 72 10 ) ) PUNCT ajst-29929 72 11 the the DET ajst-29929 72 12 bidirectional bidirectional ADJ ajst-29929 72 13 long long ADJ ajst-29929 72 14 short short ADJ ajst-29929 72 15 - - PUNCT ajst-29929 72 16 term term NOUN ajst-29929 72 17 memory memory NOUN ajst-29929 72 18 network network NOUN ajst-29929 72 19 ( ( PUNCT ajst-29929 72 20 bilstm bilstm NOUN ajst-29929 72 21 ) ) PUNCT ajst-29929 72 22 is be AUX ajst-29929 72 23 a a DET ajst-29929 72 24 deep deep ADJ ajst-29929 72 25 learning learning NOUN ajst-29929 72 26 model model NOUN ajst-29929 72 27 capable capable ADJ ajst-29929 72 28 of of ADP ajst-29929 72 29 considering consider VERB ajst-29929 72 30 both both CCONJ ajst-29929 72 31 forward forward ADJ ajst-29929 72 32 and and CCONJ ajst-29929 72 33 backward backward ADJ ajst-29929 72 34 contextual contextual ADJ ajst-29929 72 35 information information NOUN ajst-29929 72 36 in in ADP ajst-29929 72 37 sequential sequential ADJ ajst-29929 72 38 data datum NOUN ajst-29929 72 39 . . PUNCT ajst-29929 73 1 by by ADP ajst-29929 73 2 employing employ VERB ajst-29929 73 3 gating gate VERB ajst-29929 73 4 mechanisms mechanism NOUN ajst-29929 73 5 such such ADJ ajst-29929 73 6 as as ADP ajst-29929 73 7 input input NOUN ajst-29929 73 8 gates gate NOUN ajst-29929 73 9 , , PUNCT ajst-29929 73 10 forget forget VERB ajst-29929 73 11 gates gate NOUN ajst-29929 73 12 , , PUNCT ajst-29929 73 13 and and CCONJ ajst-29929 73 14 output output NOUN ajst-29929 73 15 gates gate NOUN ajst-29929 73 16 , , PUNCT ajst-29929 73 17 bilstm bilstm NOUN ajst-29929 73 18 successfully successfully ADV ajst-29929 73 19 overcomes overcome VERB ajst-29929 73 20 the the DET ajst-29929 73 21 challenges challenge NOUN ajst-29929 73 22 of of ADP ajst-29929 73 23 gradient gradient ADJ ajst-29929 73 24 vanishing vanishing NOUN ajst-29929 73 25 and and CCONJ ajst-29929 73 26 gradient gradient NOUN ajst-29929 73 27 explosion explosion NOUN ajst-29929 73 28 often often ADV ajst-29929 73 29 encountered encounter VERB ajst-29929 73 30 by by ADP ajst-29929 73 31 traditional traditional ADJ ajst-29929 73 32 recurrent recurrent ADJ ajst-29929 73 33 neural neural ADJ ajst-29929 73 34 networks network NOUN ajst-29929 73 35 ( ( PUNCT ajst-29929 73 36 rnns rnns PROPN ajst-29929 73 37 ) ) PUNCT ajst-29929 73 38 when when SCONJ ajst-29929 73 39 processing process VERB ajst-29929 73 40 long long ADJ ajst-29929 73 41 sequences sequence NOUN ajst-29929 73 42 . . PUNCT ajst-29929 74 1 in in ADP ajst-29929 74 2 the the DET ajst-29929 74 3 field field NOUN ajst-29929 74 4 of of ADP ajst-29929 74 5 rolling roll VERB ajst-29929 74 6 bearing bear VERB ajst-29929 74 7 fault fault NOUN ajst-29929 74 8 diagnosis diagnosis NOUN ajst-29929 74 9 , , PUNCT ajst-29929 74 10 vibration vibration NOUN ajst-29929 74 11 signals signal NOUN ajst-29929 74 12 possess possess VERB ajst-29929 74 13 distinct distinct ADJ ajst-29929 74 14 temporal temporal ADJ ajst-29929 74 15 characteristics characteristic NOUN ajst-29929 74 16 . . PUNCT ajst-29929 75 1 bilstm bilstm NOUN ajst-29929 75 2 , , PUNCT ajst-29929 75 3 through through ADP ajst-29929 75 4 its its PRON ajst-29929 75 5 forward forward ADJ ajst-29929 75 6 and and CCONJ ajst-29929 75 7 backward backward ADJ ajst-29929 75 8 lstm lstm ADJ ajst-29929 75 9 layers layer NOUN ajst-29929 75 10 , , PUNCT ajst-29929 75 11 can can AUX ajst-29929 75 12 simultaneously simultaneously ADV ajst-29929 75 13 capture capture VERB ajst-29929 75 14 both both PRON ajst-29929 75 15 past past ADJ ajst-29929 75 16 and and CCONJ ajst-29929 75 17 future future ADJ ajst-29929 75 18 information information NOUN ajst-29929 75 19 of of ADP ajst-29929 75 20 the the DET ajst-29929 75 21 signal signal NOUN ajst-29929 75 22 , , PUNCT ajst-29929 75 23 thereby thereby ADV ajst-29929 75 24 providing provide VERB ajst-29929 75 25 a a DET ajst-29929 75 26 more more ADV ajst-29929 75 27 comprehensive comprehensive ADJ ajst-29929 75 28 learning learning NOUN ajst-29929 75 29 of of ADP ajst-29929 75 30 sequential sequential ADJ ajst-29929 75 31 features feature NOUN ajst-29929 75 32 . . PUNCT ajst-29929 76 1 the the DET ajst-29929 76 2 model model NOUN ajst-29929 76 3 structure structure NOUN ajst-29929 76 4 is be AUX ajst-29929 76 5 shown show VERB ajst-29929 76 6 in in ADP ajst-29929 76 7 figure figure NOUN ajst-29929 76 8 3 3 NUM ajst-29929 76 9 . . PUNCT ajst-29929 76 10 figure figure VERB ajst-29929 76 11 3 3 NUM ajst-29929 76 12 . . PUNCT ajst-29929 76 13 bilstm bilstm NOUN ajst-29929 76 14 model model PROPN ajst-29929 76 15 diagram diagram PROPN ajst-29929 76 16 input input NOUN ajst-29929 76 17     SPACE ajst-29929 76 18 layer layer NOUN ajst-29929 76 19 convolution convolution NOUN ajst-29929 76 20     SPACE ajst-29929 76 21 layer layer NOUN ajst-29929 76 22 pooled pool VERB ajst-29929 76 23     SPACE ajst-29929 76 24 horizon horizon NOUN ajst-29929 76 25 convolution convolution NOUN ajst-29929 76 26     SPACE ajst-29929 76 27 layer layer NOUN ajst-29929 76 28 pooled pool VERB ajst-29929 76 29     SPACE ajst-29929 76 30 horizon horizon NOUN ajst-29929 76 31 fully fully ADV ajst-29929 76 32     SPACE ajst-29929 76 33 connected connect VERB ajst-29929 76 34     SPACE ajst-29929 76 35 layer layer NOUN ajst-29929 76 36 three three NUM ajst-29929 76 37     SPACE ajst-29929 76 38 feature feature NOUN ajst-29929 76 39     SPACE ajst-29929 76 40 maps map VERB ajst-29929 76 41 three three NUM ajst-29929 76 42     SPACE ajst-29929 76 43 feature feature NOUN ajst-29929 76 44     SPACE ajst-29929 76 45 maps map VERB ajst-29929 76 46 five five NUM ajst-29929 76 47     SPACE ajst-29929 76 48 feature feature NOUN ajst-29929 76 49     SPACE ajst-29929 76 50 maps map VERB ajst-29929 76 51 five five NUM ajst-29929 76 52     SPACE ajst-29929 76 53 feature feature NOUN ajst-29929 76 54     SPACE ajst-29929 76 55 maps map NOUN ajst-29929 76 56 output output NOUN ajst-29929 76 57     SPACE ajst-29929 76 58 layer layer NOUN ajst-29929 76 59 lstm lstm PROPN ajst-29929 77 1 lstm lstm PROPN ajst-29929 77 2 lstmlstm lstmlstm PROPN ajst-29929 77 3 lstm lstm PROPN ajst-29929 77 4 lstm lstm PROPN ajst-29929 77 5 at‐1 at‐1 PROPN ajst-29929 77 6 a a DET ajst-29929 77 7 at+1 at+1 INTJ ajst-29929 77 8 forward forward ADV ajst-29929 77 9     SPACE ajst-29929 77 10 reverse reverse NOUN ajst-29929 77 11 xt‐1 xt‐1 PUNCT ajst-29929 78 1 xt+1xinput xt+1xinput ADJ ajst-29929 78 2     SPACE ajst-29929 78 3 layer layer NOUN ajst-29929 78 4 output output NOUN ajst-29929 78 5     SPACE ajst-29929 78 6 layer layer NOUN ajst-29929 78 7 194 194 NUM ajst-29929 78 8 input input NOUN ajst-29929 78 9 layer layer NOUN ajst-29929 78 10 : : PUNCT ajst-29929 78 11 this this DET ajst-29929 78 12 layer layer NOUN ajst-29929 78 13 accepts accept VERB ajst-29929 78 14 the the DET ajst-29929 78 15 spatial spatial ADJ ajst-29929 78 16 features feature NOUN ajst-29929 78 17 extracted extract VERB ajst-29929 78 18 by by ADP ajst-29929 78 19 the the DET ajst-29929 78 20 convolutional convolutional ADJ ajst-29929 78 21 neural neural ADJ ajst-29929 78 22 network network NOUN ajst-29929 78 23 ( ( PUNCT ajst-29929 78 24 cnn cnn PROPN ajst-29929 78 25 ) ) PUNCT ajst-29929 78 26 as as ADP ajst-29929 78 27 the the DET ajst-29929 78 28 input input NOUN ajst-29929 78 29 sequence sequence NOUN ajst-29929 78 30 for for ADP ajst-29929 78 31 the the DET ajst-29929 78 32 bilstm bilstm NOUN ajst-29929 78 33 . . PUNCT ajst-29929 79 1 forward forward ADJ ajst-29929 79 2 lstm lstm ADJ ajst-29929 79 3 layer layer NOUN ajst-29929 79 4 : : PUNCT ajst-29929 79 5 this this DET ajst-29929 79 6 layer layer NOUN ajst-29929 79 7 processes process VERB ajst-29929 79 8 the the DET ajst-29929 79 9 input input NOUN ajst-29929 79 10 sequence sequence NOUN ajst-29929 79 11 in in ADP ajst-29929 79 12 chronological chronological ADJ ajst-29929 79 13 order order NOUN ajst-29929 79 14 to to PART ajst-29929 79 15 capture capture VERB ajst-29929 79 16 the the DET ajst-29929 79 17 forward forward ADJ ajst-29929 79 18 temporal temporal ADJ ajst-29929 79 19 features feature NOUN ajst-29929 79 20 of of ADP ajst-29929 79 21 the the DET ajst-29929 79 22 signal signal NOUN ajst-29929 79 23 . . PUNCT ajst-29929 80 1 it it PRON ajst-29929 80 2 learns learn VERB ajst-29929 80 3 the the DET ajst-29929 80 4 dependencies dependency NOUN ajst-29929 80 5 and and CCONJ ajst-29929 80 6 patterns pattern NOUN ajst-29929 80 7 from from ADP ajst-29929 80 8 past past NOUN ajst-29929 80 9 to to ADP ajst-29929 80 10 future future NOUN ajst-29929 80 11 within within ADP ajst-29929 80 12 the the DET ajst-29929 80 13 sequence sequence NOUN ajst-29929 80 14 . . PUNCT ajst-29929 81 1 backward backward ADJ ajst-29929 81 2 lstm lstm ADJ ajst-29929 81 3 layer layer NOUN ajst-29929 81 4 : : PUNCT ajst-29929 81 5 this this DET ajst-29929 81 6 layer layer NOUN ajst-29929 81 7 processes process VERB ajst-29929 81 8 the the DET ajst-29929 81 9 input input NOUN ajst-29929 81 10 sequence sequence NOUN ajst-29929 81 11 in in ADP ajst-29929 81 12 reverse reverse ADJ ajst-29929 81 13 chronological chronological ADJ ajst-29929 81 14 order order NOUN ajst-29929 81 15 to to PART ajst-29929 81 16 capture capture VERB ajst-29929 81 17 the the DET ajst-29929 81 18 backward backward ADJ ajst-29929 81 19 temporal temporal ADJ ajst-29929 81 20 features feature NOUN ajst-29929 81 21 of of ADP ajst-29929 81 22 the the DET ajst-29929 81 23 signal signal NOUN ajst-29929 81 24 . . PUNCT ajst-29929 82 1 it it PRON ajst-29929 82 2 learns learn VERB ajst-29929 82 3 the the DET ajst-29929 82 4 dependencies dependency NOUN ajst-29929 82 5 and and CCONJ ajst-29929 82 6 patterns pattern NOUN ajst-29929 82 7 from from ADP ajst-29929 82 8 future future NOUN ajst-29929 82 9 to to ADP ajst-29929 82 10 past past NOUN ajst-29929 82 11 within within ADP ajst-29929 82 12 the the DET ajst-29929 82 13 sequence sequence NOUN ajst-29929 82 14 . . PUNCT ajst-29929 83 1 concatenation concatenation NOUN ajst-29929 83 2 layer layer NOUN ajst-29929 83 3 : : PUNCT ajst-29929 83 4 the the DET ajst-29929 83 5 outputs output NOUN ajst-29929 83 6 from from ADP ajst-29929 83 7 the the DET ajst-29929 83 8 forward forward ADJ ajst-29929 83 9 and and CCONJ ajst-29929 83 10 backward backward ADJ ajst-29929 83 11 lstm lstm ADJ ajst-29929 83 12 layers layer NOUN ajst-29929 83 13 are be AUX ajst-29929 83 14 concatenated concatenate VERB ajst-29929 83 15 to to PART ajst-29929 83 16 form form VERB ajst-29929 83 17 a a DET ajst-29929 83 18 comprehensive comprehensive ADJ ajst-29929 83 19 representation representation NOUN ajst-29929 83 20 of of ADP ajst-29929 83 21 the the DET ajst-29929 83 22 temporal temporal ADJ ajst-29929 83 23 features feature NOUN ajst-29929 83 24 . . PUNCT ajst-29929 84 1 this this DET ajst-29929 84 2 integration integration NOUN ajst-29929 84 3 allows allow VERB ajst-29929 84 4 the the DET ajst-29929 84 5 model model NOUN ajst-29929 84 6 to to PART ajst-29929 84 7 leverage leverage VERB ajst-29929 84 8 both both PRON ajst-29929 84 9 past past NOUN ajst-29929 84 10 and and CCONJ ajst-29929 84 11 future future ADJ ajst-29929 84 12 context context NOUN ajst-29929 84 13 simultaneously simultaneously ADV ajst-29929 84 14 . . PUNCT ajst-29929 85 1 output output NOUN ajst-29929 85 2 layer layer NOUN ajst-29929 85 3 : : PUNCT ajst-29929 85 4 the the DET ajst-29929 85 5 output output NOUN ajst-29929 85 6 layer layer NOUN ajst-29929 85 7 utilizes utilize VERB ajst-29929 85 8 the the DET ajst-29929 85 9 softmax softmax NOUN ajst-29929 85 10 function function NOUN ajst-29929 85 11 to to PART ajst-29929 85 12 output output VERB ajst-29929 85 13 the the DET ajst-29929 85 14 probabilities probability NOUN ajst-29929 85 15 of of ADP ajst-29929 85 16 different different ADJ ajst-29929 85 17 fault fault NOUN ajst-29929 85 18 categories category NOUN ajst-29929 85 19 , , PUNCT ajst-29929 85 20 thereby thereby ADV ajst-29929 85 21 completing complete VERB ajst-29929 85 22 the the DET ajst-29929 85 23 fault fault NOUN ajst-29929 85 24 classification classification NOUN ajst-29929 85 25 task task NOUN ajst-29929 85 26 . . PUNCT ajst-29929 86 1 the the DET ajst-29929 86 2 application application NOUN ajst-29929 86 3 of of ADP ajst-29929 86 4 bilstm bilstm NOUN ajst-29929 86 5 enables enable VERB ajst-29929 86 6 more more ADV ajst-29929 86 7 accurate accurate ADJ ajst-29929 86 8 extraction extraction NOUN ajst-29929 86 9 of of ADP ajst-29929 86 10 temporal temporal ADJ ajst-29929 86 11 features feature NOUN ajst-29929 86 12 from from ADP ajst-29929 86 13 vibration vibration NOUN ajst-29929 86 14 signals signal NOUN ajst-29929 86 15 , , PUNCT ajst-29929 86 16 thereby thereby ADV ajst-29929 86 17 enhancing enhance VERB ajst-29929 86 18 the the DET ajst-29929 86 19 performance performance NOUN ajst-29929 86 20 of of ADP ajst-29929 86 21 fault fault NOUN ajst-29929 86 22 diagnosis diagnosis NOUN ajst-29929 86 23 . . PUNCT ajst-29929 87 1 2.3 2.3 NUM ajst-29929 87 2 . . PUNCT ajst-29929 87 3 transformer transformer NOUN ajst-29929 87 4 encoder encoder NOUN ajst-29929 87 5 layer layer NOUN ajst-29929 87 6 the the DET ajst-29929 87 7 transformer transformer NOUN ajst-29929 87 8 is be AUX ajst-29929 87 9 a a DET ajst-29929 87 10 deep deep ADJ ajst-29929 87 11 learning learning NOUN ajst-29929 87 12 architecture architecture NOUN ajst-29929 87 13 that that PRON ajst-29929 87 14 originated originate VERB ajst-29929 87 15 in in ADP ajst-29929 87 16 the the DET ajst-29929 87 17 field field NOUN ajst-29929 87 18 of of ADP ajst-29929 87 19 natural natural ADJ ajst-29929 87 20 language language NOUN ajst-29929 87 21 processing processing NOUN ajst-29929 87 22 . . PUNCT ajst-29929 88 1 its its PRON ajst-29929 88 2 core core NOUN ajst-29929 88 3 innovation innovation NOUN ajst-29929 88 4 lies lie VERB ajst-29929 88 5 in in ADP ajst-29929 88 6 the the DET ajst-29929 88 7 use use NOUN ajst-29929 88 8 of of ADP ajst-29929 88 9 self self NOUN ajst-29929 88 10 - - PUNCT ajst-29929 88 11 attention attention NOUN ajst-29929 88 12 mechanisms mechanism NOUN ajst-29929 88 13 to to PART ajst-29929 88 14 establish establish VERB ajst-29929 88 15 direct direct ADJ ajst-29929 88 16 connections connection NOUN ajst-29929 88 17 between between ADP ajst-29929 88 18 different different ADJ ajst-29929 88 19 parts part NOUN ajst-29929 88 20 of of ADP ajst-29929 88 21 the the DET ajst-29929 88 22 input input NOUN ajst-29929 88 23 sequence sequence NOUN ajst-29929 88 24 , , PUNCT ajst-29929 88 25 thereby thereby ADV ajst-29929 88 26 capturing capture VERB ajst-29929 88 27 global global ADJ ajst-29929 88 28 dependencies dependency NOUN ajst-29929 88 29 within within ADP ajst-29929 88 30 the the DET ajst-29929 88 31 data datum NOUN ajst-29929 88 32 . . PUNCT ajst-29929 89 1 compared compare VERB ajst-29929 89 2 to to ADP ajst-29929 89 3 traditional traditional ADJ ajst-29929 89 4 temporal temporal ADJ ajst-29929 89 5 models model NOUN ajst-29929 89 6 , , PUNCT ajst-29929 89 7 the the DET ajst-29929 89 8 transformer transformer NOUN ajst-29929 89 9 demonstrates demonstrate VERB ajst-29929 89 10 superior superior ADJ ajst-29929 89 11 capability capability NOUN ajst-29929 89 12 in in ADP ajst-29929 89 13 capturing capture VERB ajst-29929 89 14 long long ADJ ajst-29929 89 15 - - PUNCT ajst-29929 89 16 range range NOUN ajst-29929 89 17 dependencies dependency NOUN ajst-29929 89 18 and and CCONJ ajst-29929 89 19 complex complex ADJ ajst-29929 89 20 features feature NOUN ajst-29929 89 21 through through ADP ajst-29929 89 22 its its PRON ajst-29929 89 23 multi multi ADJ ajst-29929 89 24 - - ADJ ajst-29929 89 25 head head ADJ ajst-29929 89 26 attention attention NOUN ajst-29929 89 27 mechanism mechanism NOUN ajst-29929 89 28 . . PUNCT ajst-29929 90 1 additionally additionally ADV ajst-29929 90 2 , , PUNCT ajst-29929 90 3 it it PRON ajst-29929 90 4 achieves achieve VERB ajst-29929 90 5 higher high ADJ ajst-29929 90 6 parallel parallel ADJ ajst-29929 90 7 computing computing NOUN ajst-29929 90 8 efficiency efficiency NOUN ajst-29929 90 9 , , PUNCT ajst-29929 90 10 which which PRON ajst-29929 90 11 significantly significantly ADV ajst-29929 90 12 enhances enhance VERB ajst-29929 90 13 the the DET ajst-29929 90 14 model model NOUN ajst-29929 90 15 's 's PART ajst-29929 90 16 feature feature NOUN ajst-29929 90 17 extraction extraction NOUN ajst-29929 90 18 and and CCONJ ajst-29929 90 19 representation representation NOUN ajst-29929 90 20 abilities ability NOUN ajst-29929 90 21 . . PUNCT ajst-29929 91 1 the the DET ajst-29929 91 2 model model NOUN ajst-29929 91 3 proposed propose VERB ajst-29929 91 4 in in ADP ajst-29929 91 5 this this DET ajst-29929 91 6 paper paper NOUN ajst-29929 91 7 employs employ VERB ajst-29929 91 8 a a DET ajst-29929 91 9 standard standard ADJ ajst-29929 91 10 transformer transformer ADJ ajst-29929 91 11 encoder encoder NOUN ajst-29929 91 12 architecture architecture NOUN ajst-29929 91 13 , , PUNCT ajst-29929 91 14 with with ADP ajst-29929 91 15 each each DET ajst-29929 91 16 encoder encoder NOUN ajst-29929 91 17 layer layer NOUN ajst-29929 91 18 integrating integrate VERB ajst-29929 91 19 multi multi ADJ ajst-29929 91 20 - - ADJ ajst-29929 91 21 head head ADJ ajst-29929 91 22 self self NOUN ajst-29929 91 23 - - PUNCT ajst-29929 91 24 attention attention NOUN ajst-29929 91 25 mechanisms mechanism NOUN ajst-29929 91 26 and and CCONJ ajst-29929 91 27 a a DET ajst-29929 91 28 feedforward feedforward ADJ ajst-29929 91 29 neural neural ADJ ajst-29929 91 30 network network NOUN ajst-29929 91 31 . . PUNCT ajst-29929 92 1 specifically specifically ADV ajst-29929 92 2 , , PUNCT ajst-29929 92 3 the the DET ajst-29929 92 4 multi multi ADJ ajst-29929 92 5 - - ADJ ajst-29929 92 6 head head ADJ ajst-29929 92 7 selfattention selfattention NOUN ajst-29929 92 8 mechanism mechanism NOUN ajst-29929 92 9 consists consist VERB ajst-29929 92 10 of of ADP ajst-29929 92 11 two two NUM ajst-29929 92 12 attention attention NOUN ajst-29929 92 13 heads head NOUN ajst-29929 92 14 , , PUNCT ajst-29929 92 15 each each PRON ajst-29929 92 16 with with ADP ajst-29929 92 17 a a DET ajst-29929 92 18 dimension dimension NOUN ajst-29929 92 19 of of ADP ajst-29929 92 20 128 128 NUM ajst-29929 92 21 . . PUNCT ajst-29929 93 1 the the DET ajst-29929 93 2 feed feed NOUN ajst-29929 93 3 - - PUNCT ajst-29929 93 4 forward forward NOUN ajst-29929 93 5 neural neural ADJ ajst-29929 93 6 network network NOUN ajst-29929 93 7 has have VERB ajst-29929 93 8 a a DET ajst-29929 93 9 hidden hide VERB ajst-29929 93 10 layer layer NOUN ajst-29929 93 11 of of ADP ajst-29929 93 12 128 128 NUM ajst-29929 93 13 dimensions dimension NOUN ajst-29929 93 14 and and CCONJ ajst-29929 93 15 employs employ VERB ajst-29929 93 16 the the DET ajst-29929 93 17 relu relu NOUN ajst-29929 93 18 activation activation NOUN ajst-29929 93 19 function function NOUN ajst-29929 93 20 . . PUNCT ajst-29929 94 1 during during ADP ajst-29929 94 2 the the DET ajst-29929 94 3 preprocessing preprocessing NOUN ajst-29929 94 4 stage stage NOUN ajst-29929 94 5 , , PUNCT ajst-29929 94 6 feature feature NOUN ajst-29929 94 7 data datum NOUN ajst-29929 94 8 are be AUX ajst-29929 94 9 processed process VERB ajst-29929 94 10 through through ADP ajst-29929 94 11 a a DET ajst-29929 94 12 series series NOUN ajst-29929 94 13 of of ADP ajst-29929 94 14 convolutional convolutional ADJ ajst-29929 94 15 and and CCONJ ajst-29929 94 16 pooling pool VERB ajst-29929 94 17 layers layer NOUN ajst-29929 94 18 to to PART ajst-29929 94 19 extract extract VERB ajst-29929 94 20 frequency frequency NOUN ajst-29929 94 21 - - PUNCT ajst-29929 94 22 domain domain NOUN ajst-29929 94 23 features feature NOUN ajst-29929 94 24 from from ADP ajst-29929 94 25 the the DET ajst-29929 94 26 cnn cnn PROPN ajst-29929 94 27 . . PUNCT ajst-29929 95 1 these these DET ajst-29929 95 2 features feature NOUN ajst-29929 95 3 are be AUX ajst-29929 95 4 then then ADV ajst-29929 95 5 fed feed VERB ajst-29929 95 6 into into ADP ajst-29929 95 7 the the DET ajst-29929 95 8 transformer transformer NOUN ajst-29929 95 9 encoder encoder NOUN ajst-29929 95 10 layer layer NOUN ajst-29929 95 11 . . PUNCT ajst-29929 96 1 similarly similarly ADV ajst-29929 96 2 , , PUNCT ajst-29929 96 3 the the DET ajst-29929 96 4 temporal temporal ADJ ajst-29929 96 5 features feature NOUN ajst-29929 96 6 processed process VERB ajst-29929 96 7 by by ADP ajst-29929 96 8 bilstm bilstm NOUN ajst-29929 96 9 are be AUX ajst-29929 96 10 also also ADV ajst-29929 96 11 input input VERB ajst-29929 96 12 into into ADP ajst-29929 96 13 the the DET ajst-29929 96 14 transformer transformer NOUN ajst-29929 96 15 encoder encoder NOUN ajst-29929 96 16 layer layer NOUN ajst-29929 96 17 . . PUNCT ajst-29929 97 1 within within ADP ajst-29929 97 2 the the DET ajst-29929 97 3 encoder encoder NOUN ajst-29929 97 4 layer layer NOUN ajst-29929 97 5 , , PUNCT ajst-29929 97 6 the the DET ajst-29929 97 7 global global ADJ ajst-29929 97 8 temporal temporal ADJ ajst-29929 97 9 dependencies dependency NOUN ajst-29929 97 10 and and CCONJ ajst-29929 97 11 feature feature NOUN ajst-29929 97 12 interactions interaction NOUN ajst-29929 97 13 of of ADP ajst-29929 97 14 the the DET ajst-29929 97 15 signal signal NOUN ajst-29929 97 16 are be AUX ajst-29929 97 17 fully fully ADV ajst-29929 97 18 modeled model VERB ajst-29929 97 19 . . PUNCT ajst-29929 98 1 by by ADP ajst-29929 98 2 incorporating incorporate VERB ajst-29929 98 3 the the DET ajst-29929 98 4 transformer transformer ADJ ajst-29929 98 5 encoder encoder NOUN ajst-29929 98 6 layer layer NOUN ajst-29929 98 7 , , PUNCT ajst-29929 98 8 the the DET ajst-29929 98 9 model model NOUN ajst-29929 98 10 is be AUX ajst-29929 98 11 able able ADJ ajst-29929 98 12 to to PART ajst-29929 98 13 more more ADV ajst-29929 98 14 comprehensively comprehensively ADV ajst-29929 98 15 capture capture VERB ajst-29929 98 16 both both CCONJ ajst-29929 98 17 global global ADJ ajst-29929 98 18 and and CCONJ ajst-29929 98 19 local local ADJ ajst-29929 98 20 features feature NOUN ajst-29929 98 21 within within ADP ajst-29929 98 22 the the DET ajst-29929 98 23 vibration vibration NOUN ajst-29929 98 24 signals signal NOUN ajst-29929 98 25 , , PUNCT ajst-29929 98 26 enhancing enhance VERB ajst-29929 98 27 the the DET ajst-29929 98 28 feature feature NOUN ajst-29929 98 29 extraction extraction NOUN ajst-29929 98 30 and and CCONJ ajst-29929 98 31 fusion fusion NOUN ajst-29929 98 32 capabilities capability NOUN ajst-29929 98 33 , , PUNCT ajst-29929 98 34 and and CCONJ ajst-29929 98 35 thereby thereby ADV ajst-29929 98 36 improving improve VERB ajst-29929 98 37 the the DET ajst-29929 98 38 accuracy accuracy NOUN ajst-29929 98 39 and and CCONJ ajst-29929 98 40 generalization generalization NOUN ajst-29929 98 41 ability ability NOUN ajst-29929 98 42 of of ADP ajst-29929 98 43 fault fault NOUN ajst-29929 98 44 diagnosis diagnosis NOUN ajst-29929 98 45 . . PUNCT ajst-29929 99 1 2.4 2.4 NUM ajst-29929 99 2 . . PUNCT ajst-29929 100 1 cross cross ADJ ajst-29929 100 2 - - ADJ ajst-29929 100 3 attention attention ADJ ajst-29929 100 4 mechanism mechanism NOUN ajst-29929 100 5 ( ( PUNCT ajst-29929 100 6 crossattention crossattention NOUN ajst-29929 100 7 ) ) PUNCT ajst-29929 100 8 the the DET ajst-29929 100 9 cross cross ADJ ajst-29929 100 10 - - ADJ ajst-29929 100 11 attention attention ADJ ajst-29929 100 12 mechanism mechanism NOUN ajst-29929 100 13 is be AUX ajst-29929 100 14 an an DET ajst-29929 100 15 attention attention NOUN ajst-29929 100 16 - - PUNCT ajst-29929 100 17 based base VERB ajst-29929 100 18 approach approach NOUN ajst-29929 100 19 used use VERB ajst-29929 100 20 to to PART ajst-29929 100 21 fuse fuse VERB ajst-29929 100 22 features feature NOUN ajst-29929 100 23 from from ADP ajst-29929 100 24 different different ADJ ajst-29929 100 25 sources source NOUN ajst-29929 100 26 or or CCONJ ajst-29929 100 27 of of ADP ajst-29929 100 28 different different ADJ ajst-29929 100 29 types type NOUN ajst-29929 100 30 . . PUNCT ajst-29929 101 1 its its PRON ajst-29929 101 2 fundamental fundamental ADJ ajst-29929 101 3 idea idea NOUN ajst-29929 101 4 is be AUX ajst-29929 101 5 to to PART ajst-29929 101 6 compute compute VERB ajst-29929 101 7 attention attention NOUN ajst-29929 101 8 weights weight NOUN ajst-29929 101 9 between between ADP ajst-29929 101 10 two two NUM ajst-29929 101 11 distinct distinct ADJ ajst-29929 101 12 feature feature NOUN ajst-29929 101 13 sequences sequence NOUN ajst-29929 101 14 , , PUNCT ajst-29929 101 15 thereby thereby ADV ajst-29929 101 16 effectively effectively ADV ajst-29929 101 17 integrating integrate VERB ajst-29929 101 18 them they PRON ajst-29929 101 19 . . PUNCT ajst-29929 102 1 specifically specifically ADV ajst-29929 102 2 , , PUNCT ajst-29929 102 3 during during ADP ajst-29929 102 4 the the DET ajst-29929 102 5 operation operation NOUN ajst-29929 102 6 of of ADP ajst-29929 102 7 the the DET ajst-29929 102 8 cross cross ADJ ajst-29929 102 9 - - ADJ ajst-29929 102 10 attention attention ADJ ajst-29929 102 11 mechanism mechanism NOUN ajst-29929 102 12 , , PUNCT ajst-29929 102 13 one one NUM ajst-29929 102 14 feature feature NOUN ajst-29929 102 15 sequence sequence NOUN ajst-29929 102 16 is be AUX ajst-29929 102 17 regarded regard VERB ajst-29929 102 18 as as ADP ajst-29929 102 19 the the DET ajst-29929 102 20 query query NOUN ajst-29929 102 21 , , PUNCT ajst-29929 102 22 while while SCONJ ajst-29929 102 23 the the DET ajst-29929 102 24 other other ADJ ajst-29929 102 25 feature feature NOUN ajst-29929 102 26 sequence sequence NOUN ajst-29929 102 27 is be AUX ajst-29929 102 28 treated treat VERB ajst-29929 102 29 as as ADP ajst-29929 102 30 both both CCONJ ajst-29929 102 31 the the DET ajst-29929 102 32 key key ADJ ajst-29929 102 33 and and CCONJ ajst-29929 102 34 value value NOUN ajst-29929 102 35 . . PUNCT ajst-29929 103 1 figure figure NOUN ajst-29929 103 2 4 4 NUM ajst-29929 103 3 . . PUNCT ajst-29929 103 4 cross cross ADJ ajst-29929 103 5 - - ADJ ajst-29929 103 6 attention attention ADJ ajst-29929 103 7 mechanism mechanism NOUN ajst-29929 103 8 during during ADP ajst-29929 103 9 the the DET ajst-29929 103 10 fusion fusion NOUN ajst-29929 103 11 process process NOUN ajst-29929 103 12 , , PUNCT ajst-29929 103 13 the the DET ajst-29929 103 14 spatial spatial ADJ ajst-29929 103 15 features feature NOUN ajst-29929 103 16 extracted extract VERB ajst-29929 103 17 by by ADP ajst-29929 103 18 the the DET ajst-29929 103 19 cnn cnn PROPN ajst-29929 103 20 - - PUNCT ajst-29929 103 21 transformer transformer NOUN ajst-29929 103 22 , , PUNCT ajst-29929 103 23 denoted denote VERB ajst-29929 103 24 as as ADP ajst-29929 103 25 fspace fspace NOUN ajst-29929 103 26 , , PUNCT ajst-29929 103 27 are be AUX ajst-29929 103 28 used use VERB ajst-29929 103 29 as as ADP ajst-29929 103 30 the the DET ajst-29929 103 31 query query NOUN ajst-29929 103 32 sequence sequence NOUN ajst-29929 103 33 , , PUNCT ajst-29929 103 34 while while SCONJ ajst-29929 103 35 the the DET ajst-29929 103 36 temporal temporal ADJ ajst-29929 103 37 features feature NOUN ajst-29929 103 38 extracted extract VERB ajst-29929 103 39 by by ADP ajst-29929 103 40 the the DET ajst-29929 103 41 bilstm bilstm NOUN ajst-29929 103 42 - - PUNCT ajst-29929 103 43 transformer transformer NOUN ajst-29929 103 44 , , PUNCT ajst-29929 103 45 denoted denote VERB ajst-29929 103 46 as as ADP ajst-29929 103 47 ftime ftime NOUN ajst-29929 103 48 , , PUNCT ajst-29929 103 49 are be AUX ajst-29929 103 50 used use VERB ajst-29929 103 51 as as ADP ajst-29929 103 52 the the DET ajst-29929 103 53 keyvalue keyvalue PROPN ajst-29929 103 54 pair pair NOUN ajst-29929 103 55 sequences sequence NOUN ajst-29929 103 56 . . PUNCT ajst-29929 104 1 these these DET ajst-29929 104 2 features feature NOUN ajst-29929 104 3 are be AUX ajst-29929 104 4 first first ADV ajst-29929 104 5 linearly linearly ADV ajst-29929 104 6 mapped map VERB ajst-29929 104 7 to to ADP ajst-29929 104 8 the the DET ajst-29929 104 9 same same ADJ ajst-29929 104 10 dimension dimension NOUN ajst-29929 104 11 through through ADP ajst-29929 104 12 learnable learnable ADJ ajst-29929 104 13 weight weight NOUN ajst-29929 104 14 matrices matrix NOUN ajst-29929 104 15 wq wq NOUN ajst-29929 104 16 , , PUNCT ajst-29929 104 17 wk wk PROPN ajst-29929 104 18 , , PUNCT ajst-29929 104 19 and and CCONJ ajst-29929 104 20 wv wv PROPN ajst-29929 104 21 , , PUNCT ajst-29929 104 22 as as SCONJ ajst-29929 104 23 shown show VERB ajst-29929 104 24 in in ADP ajst-29929 104 25 equations equation NOUN ajst-29929 104 26 ( ( PUNCT ajst-29929 104 27 1 1 NUM ajst-29929 104 28 ) ) PUNCT ajst-29929 104 29 , , PUNCT ajst-29929 104 30 ( ( PUNCT ajst-29929 104 31 2 2 NUM ajst-29929 104 32 ) ) PUNCT ajst-29929 104 33 , , PUNCT ajst-29929 104 34 and and CCONJ ajst-29929 104 35 ( ( PUNCT ajst-29929 104 36 3 3 NUM ajst-29929 104 37 ) ) PUNCT ajst-29929 104 38 , , PUNCT ajst-29929 104 39 respectively respectively ADV ajst-29929 104 40 : : PUNCT ajst-29929 104 41 spatial spatial ADJ ajst-29929 104 42     SPACE ajst-29929 104 43 feature feature NOUN ajst-29929 104 44 time time NOUN ajst-29929 104 45     SPACE ajst-29929 104 46 domain domain NOUN ajst-29929 104 47     SPACE ajst-29929 104 48 feature feature NOUN ajst-29929 104 49 v v ADP ajst-29929 104 50 k k PROPN ajst-29929 104 51 q q PROPN ajst-29929 104 52 195 195 NUM ajst-29929 104 53 𝑄 𝑄 NOUN ajst-29929 104 54 𝑊 𝑊 PROPN ajst-29929 104 55 ⋅ ⋅ NOUN ajst-29929 104 56 𝐹 𝐹 PROPN ajst-29929 104 57 1 1 NUM ajst-29929 104 58 in in ADP ajst-29929 104 59 the the DET ajst-29929 104 60 equations equation NOUN ajst-29929 104 61 : : PUNCT ajst-29929 104 62 𝑄 𝑄 PROPN ajst-29929 104 63 is be AUX ajst-29929 104 64 the the DET ajst-29929 104 65 query query NOUN ajst-29929 104 66 sequence sequence NOUN ajst-29929 104 67 , , PUNCT ajst-29929 104 68 dimensionle dimensionle PROPN ajst-29929 104 69 ss;𝑊 ss;𝑊 VERB ajst-29929 104 70 is be AUX ajst-29929 104 71 the the DET ajst-29929 104 72 weight weight NOUN ajst-29929 104 73 matrix matrix NOUN ajst-29929 104 74 , , PUNCT ajst-29929 104 75 dimensionless dimensionless NOUN ajst-29929 104 76 ; ; PUNCT ajst-29929 104 77 𝐹 𝐹 PROPN ajst-29929 104 78 is be AUX ajst-29929 104 79 the the DET ajst-29929 104 80 s s X ajst-29929 104 81 patial patial ADJ ajst-29929 104 82 feature feature NOUN ajst-29929 104 83 parameter parameter NOUN ajst-29929 104 84 , , PUNCT ajst-29929 104 85 dimensionless dimensionless NOUN ajst-29929 104 86 . . PUNCT ajst-29929 105 1 𝐾 𝐾 PROPN ajst-29929 105 2 𝑊 𝑊 PROPN ajst-29929 105 3 ⋅ ⋅ PROPN ajst-29929 105 4 𝐹 𝐹 PROPN ajst-29929 105 5 2 2 NUM ajst-29929 105 6 in in ADP ajst-29929 105 7 the the DET ajst-29929 105 8 equations equation NOUN ajst-29929 105 9 : : PUNCT ajst-29929 105 10 𝐾 𝐾 PROPN ajst-29929 105 11 is be AUX ajst-29929 105 12 the the DET ajst-29929 105 13 key key ADJ ajst-29929 105 14 sequence sequence NOUN ajst-29929 105 15 , , PUNCT ajst-29929 105 16 dimensionless dimensionless NOUN ajst-29929 105 17 ; ; PUNCT ajst-29929 105 18 𝑊 𝑊 PROPN ajst-29929 105 19 is be AUX ajst-29929 105 20 the the DET ajst-29929 105 21 weight weight NOUN ajst-29929 105 22 matrix matrix NOUN ajst-29929 105 23 , , PUNCT ajst-29929 105 24 dimensionless dimensionless NOUN ajst-29929 105 25 ; ; PUNCT ajst-29929 105 26 𝐹 𝐹 PROPN ajst-29929 105 27 is be AUX ajst-29929 105 28 the the DET ajst-29929 105 29 temporal temporal ADJ ajst-29929 105 30 feature feature NOUN ajst-29929 105 31 parameter parameter NOUN ajst-29929 105 32 , , PUNCT ajst-29929 105 33 dimensionless dimensionless NOUN ajst-29929 105 34 . . PUNCT ajst-29929 106 1 𝑉 𝑉 PROPN ajst-29929 106 2 𝑊 𝑊 PROPN ajst-29929 106 3 ⋅ ⋅ PROPN ajst-29929 106 4 𝐹 𝐹 PROPN ajst-29929 106 5 3 3 NUM ajst-29929 106 6 in in ADP ajst-29929 106 7 the the DET ajst-29929 106 8 equations equation NOUN ajst-29929 106 9 : : PUNCT ajst-29929 106 10 𝑉 𝑉 PROPN ajst-29929 106 11 is be AUX ajst-29929 106 12 the the DET ajst-29929 106 13 value value NOUN ajst-29929 106 14 sequence sequence NOUN ajst-29929 106 15 , , PUNCT ajst-29929 106 16 dimensionless dimensionless NOUN ajst-29929 106 17 ; ; PUNCT ajst-29929 106 18 𝑊 𝑊 PROPN ajst-29929 106 19 is be AUX ajst-29929 106 20 the the DET ajst-29929 106 21 weight weight NOUN ajst-29929 106 22 matrix matrix NOUN ajst-29929 106 23 , , PUNCT ajst-29929 106 24 dimensionless dimensionless NOUN ajst-29929 106 25 ; ; PUNCT ajst-29929 106 26 𝐹 𝐹 PROPN ajst-29929 106 27 is be AUX ajst-29929 106 28 the the DET ajst-29929 106 29 temporal temporal ADJ ajst-29929 106 30 feature feature NOUN ajst-29929 106 31 parameter parameter NOUN ajst-29929 106 32 , , PUNCT ajst-29929 106 33 dimensionless dimensionless NOUN ajst-29929 106 34 . . PUNCT ajst-29929 107 1 the the DET ajst-29929 107 2 attention attention NOUN ajst-29929 107 3 weights weight NOUN ajst-29929 107 4 are be AUX ajst-29929 107 5 obtained obtain VERB ajst-29929 107 6 by by ADP ajst-29929 107 7 calculating calculate VERB ajst-29929 107 8 the the DET ajst-29929 107 9 similarity similarity NOUN ajst-29929 107 10 between between ADP ajst-29929 107 11 the the DET ajst-29929 107 12 query query NOUN ajst-29929 107 13 and and CCONJ ajst-29929 107 14 the the DET ajst-29929 107 15 keys key NOUN ajst-29929 107 16 , , PUNCT ajst-29929 107 17 and and CCONJ ajst-29929 107 18 then then ADV ajst-29929 107 19 these these DET ajst-29929 107 20 weights weight NOUN ajst-29929 107 21 are be AUX ajst-29929 107 22 used use VERB ajst-29929 107 23 to to PART ajst-29929 107 24 perform perform VERB ajst-29929 107 25 a a DET ajst-29929 107 26 weighted weighted ADJ ajst-29929 107 27 sum sum NOUN ajst-29929 107 28 of of ADP ajst-29929 107 29 the the DET ajst-29929 107 30 values value NOUN ajst-29929 107 31 to to PART ajst-29929 107 32 obtain obtain VERB ajst-29929 107 33 the the DET ajst-29929 107 34 fused fuse VERB ajst-29929 107 35 feature feature NOUN ajst-29929 107 36 representation representation NOUN ajst-29929 107 37 , , PUNCT ajst-29929 107 38 as as SCONJ ajst-29929 107 39 shown show VERB ajst-29929 107 40 in in ADP ajst-29929 107 41 equation equation NOUN ajst-29929 107 42 ( ( PUNCT ajst-29929 107 43 4 4 NUM ajst-29929 107 44 ): ): PUNCT ajst-29929 107 45 attention attention NOUN ajst-29929 107 46 𝑄 𝑄 PROPN ajst-29929 107 47 , , PUNCT ajst-29929 107 48 𝐾 𝐾 PROPN ajst-29929 107 49 , , PUNCT ajst-29929 107 50 𝑉 𝑉 PROPN ajst-29929 107 51 softmax softmax NOUN ajst-29929 107 52 𝑄𝐾 𝑄𝐾 PROPN ajst-29929 107 53 𝑑 𝑑 PROPN ajst-29929 107 54 𝑉 𝑉 PROPN ajst-29929 107 55 4 4 NUM ajst-29929 107 56 in in ADP ajst-29929 107 57 the the DET ajst-29929 107 58 equations equation NOUN ajst-29929 107 59 , , PUNCT ajst-29929 107 60 dk dk PROPN ajst-29929 107 61 represents represent VERB ajst-29929 107 62 the the DET ajst-29929 107 63 dimension dimension NOUN ajst-29929 107 64 of of ADP ajst-29929 107 65 the the DET ajst-29929 107 66 key key NOUN ajst-29929 107 67 , , PUNCT ajst-29929 107 68 which which PRON ajst-29929 107 69 is be AUX ajst-29929 107 70 dimensionless dimensionless ADJ ajst-29929 107 71 and and CCONJ ajst-29929 107 72 used use VERB ajst-29929 107 73 for for ADP ajst-29929 107 74 scaling scale VERB ajst-29929 107 75 the the DET ajst-29929 107 76 dot dot NOUN ajst-29929 107 77 product product NOUN ajst-29929 107 78 to to PART ajst-29929 107 79 prevent prevent VERB ajst-29929 107 80 large large ADJ ajst-29929 107 81 values value NOUN ajst-29929 107 82 from from ADP ajst-29929 107 83 causing cause VERB ajst-29929 107 84 gradient gradient ADJ ajst-29929 107 85 vanishing vanishing NOUN ajst-29929 107 86 . . PUNCT ajst-29929 108 1 the the DET ajst-29929 108 2 transpose transpose NOUN ajst-29929 108 3 of of ADP ajst-29929 108 4 k k PROPN ajst-29929 108 5 ( ( PUNCT ajst-29929 108 6 i.e. i.e. X ajst-29929 108 7 , , PUNCT ajst-29929 108 8 kt kt X ajst-29929 108 9 ) ) PUNCT ajst-29929 108 10 helps help VERB ajst-29929 108 11 avoid avoid VERB ajst-29929 108 12 excessively excessively ADV ajst-29929 108 13 large large ADJ ajst-29929 108 14 values value NOUN ajst-29929 108 15 that that PRON ajst-29929 108 16 could could AUX ajst-29929 108 17 lead lead VERB ajst-29929 108 18 to to ADP ajst-29929 108 19 gradient gradient ADJ ajst-29929 108 20 vanishing vanishing NOUN ajst-29929 108 21 . . PUNCT ajst-29929 109 1 in in ADP ajst-29929 109 2 rolling rolling ADJ ajst-29929 109 3 bearing bear VERB ajst-29929 109 4 fault fault NOUN ajst-29929 109 5 diagnosis diagnosis NOUN ajst-29929 109 6 , , PUNCT ajst-29929 109 7 vibration vibration NOUN ajst-29929 109 8 signals signal NOUN ajst-29929 109 9 contain contain VERB ajst-29929 109 10 rich rich ADJ ajst-29929 109 11 time time NOUN ajst-29929 109 12 - - PUNCT ajst-29929 109 13 frequency frequency NOUN ajst-29929 109 14 features feature NOUN ajst-29929 109 15 . . PUNCT ajst-29929 110 1 features feature NOUN ajst-29929 110 2 extracted extract VERB ajst-29929 110 3 by by ADP ajst-29929 110 4 fast fast ADJ ajst-29929 110 5 fourier fourier NOUN ajst-29929 110 6 transform transform NOUN ajst-29929 110 7 ( ( PUNCT ajst-29929 110 8 fft fft PROPN ajst-29929 110 9 ) ) PUNCT ajst-29929 110 10 and and CCONJ ajst-29929 110 11 variational variational ADJ ajst-29929 110 12 mode mode NOUN ajst-29929 110 13 decomposition decomposition NOUN ajst-29929 110 14 ( ( PUNCT ajst-29929 110 15 vmd vmd NOUN ajst-29929 110 16 ) ) PUNCT ajst-29929 110 17 represent represent VERB ajst-29929 110 18 the the DET ajst-29929 110 19 frequency frequency NOUN ajst-29929 110 20 - - PUNCT ajst-29929 110 21 domain domain NOUN ajst-29929 110 22 and and CCONJ ajst-29929 110 23 time time NOUN ajst-29929 110 24 - - PUNCT ajst-29929 110 25 domain domain NOUN ajst-29929 110 26 information information NOUN ajst-29929 110 27 of of ADP ajst-29929 110 28 the the DET ajst-29929 110 29 signal signal NOUN ajst-29929 110 30 , , PUNCT ajst-29929 110 31 respectively respectively ADV ajst-29929 110 32 . . PUNCT ajst-29929 111 1 the the DET ajst-29929 111 2 cross cross ADJ ajst-29929 111 3 - - ADJ ajst-29929 111 4 attention attention ADJ ajst-29929 111 5 mechanism mechanism NOUN ajst-29929 111 6 can can AUX ajst-29929 111 7 effectively effectively ADV ajst-29929 111 8 integrate integrate VERB ajst-29929 111 9 these these DET ajst-29929 111 10 features feature NOUN ajst-29929 111 11 from from ADP ajst-29929 111 12 different different ADJ ajst-29929 111 13 domains domain NOUN ajst-29929 111 14 . . PUNCT ajst-29929 112 1 specifically specifically ADV ajst-29929 112 2 , , PUNCT ajst-29929 112 3 the the DET ajst-29929 112 4 frequencydomain frequencydomain ADJ ajst-29929 112 5 features feature NOUN ajst-29929 112 6 extracted extract VERB ajst-29929 112 7 by by ADP ajst-29929 112 8 fft fft NOUN ajst-29929 112 9 are be AUX ajst-29929 112 10 used use VERB ajst-29929 112 11 as as ADP ajst-29929 112 12 the the DET ajst-29929 112 13 query query NOUN ajst-29929 112 14 , , PUNCT ajst-29929 112 15 while while SCONJ ajst-29929 112 16 the the DET ajst-29929 112 17 time time NOUN ajst-29929 112 18 - - PUNCT ajst-29929 112 19 domain domain NOUN ajst-29929 112 20 features feature NOUN ajst-29929 112 21 extracted extract VERB ajst-29929 112 22 by by ADP ajst-29929 112 23 vmd vmd PROPN ajst-29929 112 24 are be AUX ajst-29929 112 25 used use VERB ajst-29929 112 26 as as ADP ajst-29929 112 27 the the DET ajst-29929 112 28 key key NOUN ajst-29929 112 29 and and CCONJ ajst-29929 112 30 value value NOUN ajst-29929 112 31 . . PUNCT ajst-29929 113 1 the the DET ajst-29929 113 2 attention attention NOUN ajst-29929 113 3 weights weight NOUN ajst-29929 113 4 are be AUX ajst-29929 113 5 calculated calculate VERB ajst-29929 113 6 between between ADP ajst-29929 113 7 these these DET ajst-29929 113 8 two two NUM ajst-29929 113 9 sets set NOUN ajst-29929 113 10 of of ADP ajst-29929 113 11 features feature NOUN ajst-29929 113 12 , , PUNCT ajst-29929 113 13 thereby thereby ADV ajst-29929 113 14 fusing fuse VERB ajst-29929 113 15 the the DET ajst-29929 113 16 frequencydomain frequencydomain ADJ ajst-29929 113 17 features feature NOUN ajst-29929 113 18 with with ADP ajst-29929 113 19 multi multi ADJ ajst-29929 113 20 - - ADJ ajst-29929 113 21 scale scale ADJ ajst-29929 113 22 time time NOUN ajst-29929 113 23 - - PUNCT ajst-29929 113 24 domain domain NOUN ajst-29929 113 25 features feature NOUN ajst-29929 113 26 . . PUNCT ajst-29929 114 1 this this DET ajst-29929 114 2 approach approach NOUN ajst-29929 114 3 fully fully ADV ajst-29929 114 4 utilizes utilize VERB ajst-29929 114 5 the the DET ajst-29929 114 6 diverse diverse ADJ ajst-29929 114 7 characteristics characteristic NOUN ajst-29929 114 8 of of ADP ajst-29929 114 9 the the DET ajst-29929 114 10 signal signal NOUN ajst-29929 114 11 , , PUNCT ajst-29929 114 12 thereby thereby ADV ajst-29929 114 13 enhancing enhance VERB ajst-29929 114 14 the the DET ajst-29929 114 15 accuracy accuracy NOUN ajst-29929 114 16 and and CCONJ ajst-29929 114 17 robustness robustness NOUN ajst-29929 114 18 of of ADP ajst-29929 114 19 fault fault NOUN ajst-29929 114 20 diagnosis diagnosis NOUN ajst-29929 114 21 . . PUNCT ajst-29929 115 1 3 3 X ajst-29929 115 2 . . X ajst-29929 115 3 experimental experimental ADJ ajst-29929 115 4 results result NOUN ajst-29929 115 5 and and CCONJ ajst-29929 115 6 analysis analysis NOUN ajst-29929 115 7 3.1 3.1 NUM ajst-29929 115 8 . . PUNCT ajst-29929 116 1 data datum NOUN ajst-29929 116 2 source source VERB ajst-29929 116 3 the the DET ajst-29929 116 4 dataset dataset NOUN ajst-29929 116 5 used use VERB ajst-29929 116 6 in in ADP ajst-29929 116 7 this this DET ajst-29929 116 8 experiment experiment NOUN ajst-29929 116 9 was be AUX ajst-29929 116 10 obtained obtain VERB ajst-29929 116 11 from from ADP ajst-29929 116 12 the the DET ajst-29929 116 13 experimental experimental ADJ ajst-29929 116 14 testing testing NOUN ajst-29929 116 15 system system NOUN ajst-29929 116 16 of of ADP ajst-29929 116 17 case case NOUN ajst-29929 116 18 western western PROPN ajst-29929 116 19 reserve reserve PROPN ajst-29929 116 20 university university PROPN ajst-29929 116 21 ( ( PUNCT ajst-29929 116 22 cwru cwru PROPN ajst-29929 116 23 ) ) PUNCT ajst-29929 116 24 in in ADP ajst-29929 116 25 the the DET ajst-29929 116 26 united united PROPN ajst-29929 116 27 states states PROPN ajst-29929 116 28 , , PUNCT ajst-29929 116 29 as as SCONJ ajst-29929 116 30 shown show VERB ajst-29929 116 31 in in ADP ajst-29929 116 32 figure figure NOUN ajst-29929 116 33 5 5 NUM ajst-29929 116 34 . . PUNCT ajst-29929 117 1 this this DET ajst-29929 117 2 dataset dataset NOUN ajst-29929 117 3 has have AUX ajst-29929 117 4 been be AUX ajst-29929 117 5 widely widely ADV ajst-29929 117 6 used use VERB ajst-29929 117 7 by by ADP ajst-29929 117 8 scholars scholar NOUN ajst-29929 117 9 both both CCONJ ajst-29929 117 10 domestically domestically ADV ajst-29929 117 11 and and CCONJ ajst-29929 117 12 internationally internationally ADV ajst-29929 117 13 , , PUNCT ajst-29929 117 14 making make VERB ajst-29929 117 15 it it PRON ajst-29929 117 16 highly highly ADV ajst-29929 117 17 authoritative authoritative ADJ ajst-29929 117 18 and and CCONJ ajst-29929 117 19 representative representative NOUN ajst-29929 117 20 . . PUNCT ajst-29929 118 1 for for ADP ajst-29929 118 2 this this DET ajst-29929 118 3 experiment experiment NOUN ajst-29929 118 4 , , PUNCT ajst-29929 118 5 we we PRON ajst-29929 118 6 selected select VERB ajst-29929 118 7 the the DET ajst-29929 118 8 drive drive NOUN ajst-29929 118 9 - - PUNCT ajst-29929 118 10 end end NOUN ajst-29929 118 11 bearing bear VERB ajst-29929 118 12 data datum NOUN ajst-29929 118 13 with with ADP ajst-29929 118 14 a a DET ajst-29929 118 15 sampling sample VERB ajst-29929 118 16 frequency frequency NOUN ajst-29929 118 17 of of ADP ajst-29929 118 18 12 12 NUM ajst-29929 118 19 khz khz NOUN ajst-29929 118 20 , , PUNCT ajst-29929 118 21 a a DET ajst-29929 118 22 motor motor NOUN ajst-29929 118 23 speed speed NOUN ajst-29929 118 24 of of ADP ajst-29929 118 25 1797 1797 NUM ajst-29929 118 26 r r PROPN ajst-29929 118 27 / / SYM ajst-29929 118 28 min min NOUN ajst-29929 118 29 , , PUNCT ajst-29929 118 30 and and CCONJ ajst-29929 118 31 data datum NOUN ajst-29929 118 32 collected collect VERB ajst-29929 118 33 at at ADP ajst-29929 118 34 a a DET ajst-29929 118 35 rate rate NOUN ajst-29929 118 36 of of ADP ajst-29929 118 37 12,000 12,000 NUM ajst-29929 118 38 samples sample NOUN ajst-29929 118 39 per per ADP ajst-29929 118 40 second second NOUN ajst-29929 118 41 . . PUNCT ajst-29929 119 1 the the DET ajst-29929 119 2 bearing bearing NOUN ajst-29929 119 3 type type NOUN ajst-29929 119 4 used use VERB ajst-29929 119 5 was be AUX ajst-29929 119 6 skf6205 skf6205 NOUN ajst-29929 119 7 . . PUNCT ajst-29929 120 1 the the DET ajst-29929 120 2 dataset dataset NOUN ajst-29929 120 3 includes include VERB ajst-29929 120 4 three three NUM ajst-29929 120 5 different different ADJ ajst-29929 120 6 fault fault NOUN ajst-29929 120 7 modes mode NOUN ajst-29929 120 8 : : PUNCT ajst-29929 120 9 inner inner ADJ ajst-29929 120 10 race race NOUN ajst-29929 120 11 fault fault NOUN ajst-29929 120 12 , , PUNCT ajst-29929 120 13 outer outer ADJ ajst-29929 120 14 race race NOUN ajst-29929 120 15 fault fault NOUN ajst-29929 120 16 , , PUNCT ajst-29929 120 17 and and CCONJ ajst-29929 120 18 rolling rolling ADJ ajst-29929 120 19 element element ADJ ajst-29929 120 20 fault fault NOUN ajst-29929 120 21 . . PUNCT ajst-29929 121 1 each each DET ajst-29929 121 2 fault fault NOUN ajst-29929 121 3 mode mode NOUN ajst-29929 121 4 is be AUX ajst-29929 121 5 further far ADV ajst-29929 121 6 divided divide VERB ajst-29929 121 7 into into ADP ajst-29929 121 8 three three NUM ajst-29929 121 9 levels level NOUN ajst-29929 121 10 of of ADP ajst-29929 121 11 severity severity NOUN ajst-29929 121 12 : : PUNCT ajst-29929 121 13 mild mild ADJ ajst-29929 121 14 , , PUNCT ajst-29929 121 15 moderate moderate ADJ ajst-29929 121 16 , , PUNCT ajst-29929 121 17 and and CCONJ ajst-29929 121 18 severe severe ADJ ajst-29929 121 19 . . PUNCT ajst-29929 122 1 including include VERB ajst-29929 122 2 the the DET ajst-29929 122 3 normal normal ADJ ajst-29929 122 4 condition condition NOUN ajst-29929 122 5 , , PUNCT ajst-29929 122 6 there there PRON ajst-29929 122 7 are be VERB ajst-29929 122 8 a a DET ajst-29929 122 9 total total NOUN ajst-29929 122 10 of of ADP ajst-29929 122 11 10 10 NUM ajst-29929 122 12 data datum NOUN ajst-29929 122 13 types type NOUN ajst-29929 122 14 . . PUNCT ajst-29929 123 1 figure figure NOUN ajst-29929 123 2 5 5 NUM ajst-29929 123 3 . . PUNCT ajst-29929 123 4 case case NOUN ajst-29929 123 5 western western PROPN ajst-29929 123 6 reserve reserve PROPN ajst-29929 123 7 university university PROPN ajst-29929 123 8 experimental experimental ADJ ajst-29929 123 9 test test NOUN ajst-29929 123 10 platform platform NOUN ajst-29929 123 11 the the DET ajst-29929 123 12 dataset dataset NOUN ajst-29929 123 13 can can AUX ajst-29929 123 14 be be AUX ajst-29929 123 15 segmented segment VERB ajst-29929 123 16 in in ADP ajst-29929 123 17 various various ADJ ajst-29929 123 18 ways way NOUN ajst-29929 123 19 , , PUNCT ajst-29929 123 20 and and CCONJ ajst-29929 123 21 the the DET ajst-29929 123 22 segmentation segmentation NOUN ajst-29929 123 23 stride stride NOUN ajst-29929 123 24 also also ADV ajst-29929 123 25 varies vary VERB ajst-29929 123 26 . . PUNCT ajst-29929 124 1 in in ADP ajst-29929 124 2 this this DET ajst-29929 124 3 experiment experiment NOUN ajst-29929 124 4 , , PUNCT ajst-29929 124 5 the the DET ajst-29929 124 6 data data NOUN ajst-29929 124 7 segmentation segmentation NOUN ajst-29929 124 8 was be AUX ajst-29929 124 9 performed perform VERB ajst-29929 124 10 with with ADP ajst-29929 124 11 fixed fix VERB ajst-29929 124 12 parameter parameter NOUN ajst-29929 124 13 settings setting NOUN ajst-29929 124 14 . . PUNCT ajst-29929 125 1 the the DET ajst-29929 125 2 stride stride NOUN ajst-29929 125 3 was be AUX ajst-29929 125 4 set set VERB ajst-29929 125 5 to to ADP ajst-29929 125 6 512 512 NUM ajst-29929 125 7 , , PUNCT ajst-29929 125 8 with with ADP ajst-29929 125 9 a a DET ajst-29929 125 10 sliding slide VERB ajst-29929 125 11 overlap overlap NOUN ajst-29929 125 12 rate rate NOUN ajst-29929 125 13 of of ADP ajst-29929 125 14 0.5 0.5 NUM ajst-29929 125 15 . . PUNCT ajst-29929 126 1 the the DET ajst-29929 126 2 total total ADJ ajst-29929 126 3 number number NOUN ajst-29929 126 4 of of ADP ajst-29929 126 5 fault fault NOUN ajst-29929 126 6 samples sample NOUN ajst-29929 126 7 was be AUX ajst-29929 126 8 2330 2330 NUM ajst-29929 126 9 , , PUNCT ajst-29929 126 10 including include VERB ajst-29929 126 11 9 9 NUM ajst-29929 126 12 types type NOUN ajst-29929 126 13 of of ADP ajst-29929 126 14 fault fault NOUN ajst-29929 126 15 samples sample NOUN ajst-29929 126 16 and and CCONJ ajst-29929 126 17 1 1 NUM ajst-29929 126 18 type type NOUN ajst-29929 126 19 of of ADP ajst-29929 126 20 normal normal ADJ ajst-29929 126 21 sample sample NOUN ajst-29929 126 22 . . PUNCT ajst-29929 127 1 after after ADP ajst-29929 127 2 segmentation segmentation NOUN ajst-29929 127 3 , , PUNCT ajst-29929 127 4 the the DET ajst-29929 127 5 samples sample NOUN ajst-29929 127 6 were be AUX ajst-29929 127 7 divided divide VERB ajst-29929 127 8 into into ADP ajst-29929 127 9 training training NOUN ajst-29929 127 10 , , PUNCT ajst-29929 127 11 validation validation NOUN ajst-29929 127 12 , , PUNCT ajst-29929 127 13 and and CCONJ ajst-29929 127 14 testing testing NOUN ajst-29929 127 15 sets set NOUN ajst-29929 127 16 in in ADP ajst-29929 127 17 a a DET ajst-29929 127 18 ratio ratio NOUN ajst-29929 127 19 of of ADP ajst-29929 127 20 7:2:1 7:2:1 NUM ajst-29929 127 21 . . PROPN ajst-29929 128 1 3.2 3.2 NUM ajst-29929 128 2 . . PUNCT ajst-29929 129 1 experimental experimental ADJ ajst-29929 129 2 results result NOUN ajst-29929 129 3 after after SCONJ ajst-29929 129 4 the the DET ajst-29929 129 5 one one NUM ajst-29929 129 6 - - PUNCT ajst-29929 129 7 dimensional dimensional ADJ ajst-29929 129 8 fault fault NOUN ajst-29929 129 9 signals signal NOUN ajst-29929 129 10 were be AUX ajst-29929 129 11 transformed transform VERB ajst-29929 129 12 and and CCONJ ajst-29929 129 13 decomposed decompose VERB ajst-29929 129 14 , , PUNCT ajst-29929 129 15 they they PRON ajst-29929 129 16 were be AUX ajst-29929 129 17 fed feed VERB ajst-29929 129 18 into into ADP ajst-29929 129 19 the the DET ajst-29929 129 20 model model NOUN ajst-29929 129 21 for for ADP ajst-29929 129 22 training training NOUN ajst-29929 129 23 . . PUNCT ajst-29929 130 1 the the DET ajst-29929 130 2 training training NOUN ajst-29929 130 3 results result NOUN ajst-29929 130 4 are be AUX ajst-29929 130 5 shown show VERB ajst-29929 130 6 in in ADP ajst-29929 130 7 figures figure NOUN ajst-29929 130 8 6 6 NUM ajst-29929 130 9 , , PUNCT ajst-29929 130 10 7 7 NUM ajst-29929 130 11 , , PUNCT ajst-29929 130 12 8 8 NUM ajst-29929 130 13 , , PUNCT ajst-29929 130 14 and and CCONJ ajst-29929 130 15 9 9 NUM ajst-29929 130 16 . . X ajst-29929 130 17 196 196 NUM ajst-29929 130 18 figure figure NOUN ajst-29929 130 19 6 6 NUM ajst-29929 130 20 . . PUNCT ajst-29929 131 1 training training NOUN ajst-29929 131 2 loss loss NOUN ajst-29929 131 3 figure figure NOUN ajst-29929 131 4 7 7 NUM ajst-29929 131 5 . . PUNCT ajst-29929 131 6 training training NOUN ajst-29929 131 7 accuracy accuracy NOUN ajst-29929 131 8 figure figure NOUN ajst-29929 131 9 8 8 NUM ajst-29929 131 10 . . PUNCT ajst-29929 132 1 validation validation NOUN ajst-29929 132 2 loss loss NOUN ajst-29929 132 3 197 197 NUM ajst-29929 132 4 figure figure NOUN ajst-29929 132 5 9 9 NUM ajst-29929 132 6 . . PUNCT ajst-29929 132 7 validation validation NOUN ajst-29929 132 8 accuracy accuracy NOUN ajst-29929 132 9 figures figure NOUN ajst-29929 132 10 6 6 NUM ajst-29929 132 11 , , PUNCT ajst-29929 132 12 7 7 NUM ajst-29929 132 13 , , PUNCT ajst-29929 132 14 8 8 NUM ajst-29929 132 15 , , PUNCT ajst-29929 132 16 and and CCONJ ajst-29929 132 17 9 9 NUM ajst-29929 132 18 illustrate illustrate VERB ajst-29929 132 19 the the DET ajst-29929 132 20 training training NOUN ajst-29929 132 21 loss loss NOUN ajst-29929 132 22 , , PUNCT ajst-29929 132 23 training training NOUN ajst-29929 132 24 accuracy accuracy NOUN ajst-29929 132 25 , , PUNCT ajst-29929 132 26 validation validation NOUN ajst-29929 132 27 loss loss NOUN ajst-29929 132 28 , , PUNCT ajst-29929 132 29 and and CCONJ ajst-29929 132 30 validation validation NOUN ajst-29929 132 31 accuracy accuracy NOUN ajst-29929 132 32 of of ADP ajst-29929 132 33 the the DET ajst-29929 132 34 model model NOUN ajst-29929 132 35 across across ADP ajst-29929 132 36 different different ADJ ajst-29929 132 37 epochs epoch NOUN ajst-29929 132 38 ( ( PUNCT ajst-29929 132 39 training training NOUN ajst-29929 132 40 iterations iteration NOUN ajst-29929 132 41 ) ) PUNCT ajst-29929 132 42 . . PUNCT ajst-29929 133 1 in in ADP ajst-29929 133 2 these these DET ajst-29929 133 3 figures figure NOUN ajst-29929 133 4 , , PUNCT ajst-29929 133 5 the the PRON ajst-29929 133 6 x x NOUN ajst-29929 133 7 - - NOUN ajst-29929 133 8 axis axis NOUN ajst-29929 133 9 represents represent VERB ajst-29929 133 10 the the DET ajst-29929 133 11 number number NOUN ajst-29929 133 12 of of ADP ajst-29929 133 13 training training NOUN ajst-29929 133 14 iterations iteration NOUN ajst-29929 133 15 , , PUNCT ajst-29929 133 16 while while SCONJ ajst-29929 133 17 the the DET ajst-29929 133 18 y y NOUN ajst-29929 133 19 - - PUNCT ajst-29929 133 20 axis axis NOUN ajst-29929 133 21 reflects reflect VERB ajst-29929 133 22 the the DET ajst-29929 133 23 changes change NOUN ajst-29929 133 24 in in ADP ajst-29929 133 25 loss loss NOUN ajst-29929 133 26 values value NOUN ajst-29929 133 27 or or CCONJ ajst-29929 133 28 accuracy accuracy NOUN ajst-29929 133 29 . . PUNCT ajst-29929 134 1 observations observation NOUN ajst-29929 134 2 reveal reveal VERB ajst-29929 134 3 that that SCONJ ajst-29929 134 4 as as ADP ajst-29929 134 5 training training NOUN ajst-29929 134 6 progresses progress NOUN ajst-29929 134 7 , , PUNCT ajst-29929 134 8 the the DET ajst-29929 134 9 training training NOUN ajst-29929 134 10 loss loss NOUN ajst-29929 134 11 rapidly rapidly ADV ajst-29929 134 12 decreases decrease VERB ajst-29929 134 13 , , PUNCT ajst-29929 134 14 while while SCONJ ajst-29929 134 15 the the DET ajst-29929 134 16 validation validation NOUN ajst-29929 134 17 loss loss NOUN ajst-29929 134 18 and and CCONJ ajst-29929 134 19 validation validation NOUN ajst-29929 134 20 accuracy accuracy NOUN ajst-29929 134 21 remain remain VERB ajst-29929 134 22 relatively relatively ADV ajst-29929 134 23 stable stable ADJ ajst-29929 134 24 . . PUNCT ajst-29929 135 1 although although SCONJ ajst-29929 135 2 the the DET ajst-29929 135 3 performance performance NOUN ajst-29929 135 4 metrics metric NOUN ajst-29929 135 5 on on ADP ajst-29929 135 6 the the DET ajst-29929 135 7 training training NOUN ajst-29929 135 8 set set NOUN ajst-29929 135 9 provide provide VERB ajst-29929 135 10 feedback feedback NOUN ajst-29929 135 11 on on ADP ajst-29929 135 12 the the DET ajst-29929 135 13 model model NOUN ajst-29929 135 14 's 's PART ajst-29929 135 15 learning learning NOUN ajst-29929 135 16 process process NOUN ajst-29929 135 17 , , PUNCT ajst-29929 135 18 these these DET ajst-29929 135 19 indicators indicator NOUN ajst-29929 135 20 do do AUX ajst-29929 135 21 not not PART ajst-29929 135 22 comprehensively comprehensively ADV ajst-29929 135 23 reflect reflect VERB ajst-29929 135 24 the the DET ajst-29929 135 25 model model NOUN ajst-29929 135 26 's 's PART ajst-29929 135 27 generalization generalization NOUN ajst-29929 135 28 ability ability NOUN ajst-29929 135 29 in in ADP ajst-29929 135 30 real real ADJ ajst-29929 135 31 - - PUNCT ajst-29929 135 32 world world NOUN ajst-29929 135 33 applications application NOUN ajst-29929 135 34 . . PUNCT ajst-29929 136 1 to to PART ajst-29929 136 2 precisely precisely ADV ajst-29929 136 3 evaluate evaluate VERB ajst-29929 136 4 the the DET ajst-29929 136 5 model model NOUN ajst-29929 136 6 's 's PART ajst-29929 136 7 generalization generalization NOUN ajst-29929 136 8 performance performance NOUN ajst-29929 136 9 , , PUNCT ajst-29929 136 10 further further ADJ ajst-29929 136 11 assessment assessment NOUN ajst-29929 136 12 will will AUX ajst-29929 136 13 be be AUX ajst-29929 136 14 conducted conduct VERB ajst-29929 136 15 on on ADP ajst-29929 136 16 an an DET ajst-29929 136 17 independent independent ADJ ajst-29929 136 18 test test NOUN ajst-29929 136 19 set set NOUN ajst-29929 136 20 . . PUNCT ajst-29929 137 1 the the DET ajst-29929 137 2 test test NOUN ajst-29929 137 3 set set NOUN ajst-29929 137 4 , , PUNCT ajst-29929 137 5 which which PRON ajst-29929 137 6 was be AUX ajst-29929 137 7 predefined predefine VERB ajst-29929 137 8 and and CCONJ ajst-29929 137 9 not not PART ajst-29929 137 10 used use VERB ajst-29929 137 11 during during ADP ajst-29929 137 12 model model NOUN ajst-29929 137 13 training training NOUN ajst-29929 137 14 , , PUNCT ajst-29929 137 15 ensures ensure VERB ajst-29929 137 16 the the DET ajst-29929 137 17 objectivity objectivity NOUN ajst-29929 137 18 and and CCONJ ajst-29929 137 19 reliability reliability NOUN ajst-29929 137 20 of of ADP ajst-29929 137 21 the the DET ajst-29929 137 22 evaluation evaluation NOUN ajst-29929 137 23 results result NOUN ajst-29929 137 24 . . PUNCT ajst-29929 138 1 when when SCONJ ajst-29929 138 2 assessing assess VERB ajst-29929 138 3 the the DET ajst-29929 138 4 performance performance NOUN ajst-29929 138 5 on on ADP ajst-29929 138 6 the the DET ajst-29929 138 7 test test NOUN ajst-29929 138 8 set set NOUN ajst-29929 138 9 , , PUNCT ajst-29929 138 10 key key ADJ ajst-29929 138 11 metrics metric NOUN ajst-29929 138 12 such such ADJ ajst-29929 138 13 as as ADP ajst-29929 138 14 accuracy accuracy NOUN ajst-29929 138 15 , , PUNCT ajst-29929 138 16 recall recall NOUN ajst-29929 138 17 , , PUNCT ajst-29929 138 18 f1 f1 NOUN ajst-29929 138 19 score score NOUN ajst-29929 138 20 , , PUNCT ajst-29929 138 21 and and CCONJ ajst-29929 138 22 the the DET ajst-29929 138 23 confusion confusion NOUN ajst-29929 138 24 matrix matrix NOUN ajst-29929 138 25 are be AUX ajst-29929 138 26 considered consider VERB ajst-29929 138 27 . . PUNCT ajst-29929 139 1 table table NOUN ajst-29929 139 2 1 1 NUM ajst-29929 139 3 . . PUNCT ajst-29929 139 4 partial partial ADJ ajst-29929 139 5 performance performance NOUN ajst-29929 139 6 metrics metric NOUN ajst-29929 139 7 overview overview VERB ajst-29929 139 8 precision precision NOUN ajst-29929 139 9 recall recall NOUN ajst-29929 139 10 f1 f1 NOUN ajst-29929 139 11 - - PUNCT ajst-29929 139 12 score score NOUN ajst-29929 139 13 support support NOUN ajst-29929 139 14 0 0 NUM ajst-29929 139 15 1.0 1.0 NUM ajst-29929 139 16 1.0 1.0 NUM ajst-29929 139 17 27 27 NUM ajst-29929 139 18 1 1 NUM ajst-29929 139 19 1.0 1.0 NUM ajst-29929 139 20 1.0 1.0 NUM ajst-29929 139 21 28 28 NUM ajst-29929 139 22 2 2 NUM ajst-29929 139 23 1.0 1.0 NUM ajst-29929 139 24 1.0 1.0 NUM ajst-29929 139 25 24 24 NUM ajst-29929 139 26 3 3 NUM ajst-29929 139 27 1.0 1.0 NUM ajst-29929 139 28 1.0 1.0 NUM ajst-29929 139 29 20 20 NUM ajst-29929 139 30 4 4 NUM ajst-29929 139 31 1.0 1.0 NUM ajst-29929 139 32 1.0 1.0 NUM ajst-29929 139 33 28 28 NUM ajst-29929 139 34 5 5 NUM ajst-29929 139 35 1.0 1.0 NUM ajst-29929 139 36 1.0 1.0 NUM ajst-29929 139 37 13 13 NUM ajst-29929 139 38 6 6 NUM ajst-29929 139 39 1.0 1.0 NUM ajst-29929 139 40 1.0 1.0 NUM ajst-29929 139 41 22 22 NUM ajst-29929 139 42 7 7 NUM ajst-29929 139 43 1.0 1.0 NUM ajst-29929 139 44 1.0 1.0 NUM ajst-29929 139 45 21 21 NUM ajst-29929 139 46 8 8 NUM ajst-29929 139 47 1.0 1.0 NUM ajst-29929 139 48 1.0 1.0 NUM ajst-29929 139 49 22 22 NUM ajst-29929 139 50 9 9 NUM ajst-29929 139 51 1.0 1.0 NUM ajst-29929 139 52 1.0 1.0 NUM ajst-29929 139 53 19 19 NUM ajst-29929 139 54 figure figure NOUN ajst-29929 139 55 10 10 NUM ajst-29929 139 56 . . PUNCT ajst-29929 140 1 confusion confusion NOUN ajst-29929 140 2 matrix matrix NOUN ajst-29929 140 3 198 198 NUM ajst-29929 140 4 the the DET ajst-29929 140 5 accuracy accuracy NOUN ajst-29929 140 6 , , PUNCT ajst-29929 140 7 recall recall NOUN ajst-29929 140 8 , , PUNCT ajst-29929 140 9 and and CCONJ ajst-29929 140 10 f1 f1 ADJ ajst-29929 140 11 score score NOUN ajst-29929 140 12 of of ADP ajst-29929 140 13 the the DET ajst-29929 140 14 model model NOUN ajst-29929 140 15 are be AUX ajst-29929 140 16 sh sh NOUN ajst-29929 140 17 own own ADJ ajst-29929 140 18 in in ADP ajst-29929 140 19 table table NOUN ajst-29929 140 20 1 1 NUM ajst-29929 140 21 , , PUNCT ajst-29929 140 22 all all PRON ajst-29929 140 23 of of ADP ajst-29929 140 24 which which PRON ajst-29929 140 25 are be AUX ajst-29929 140 26 1.0000 1.0000 NUM ajst-29929 140 27 . . PUNCT ajst-29929 141 1 the the DET ajst-29929 141 2 confusion confusion NOUN ajst-29929 141 3 matrix matrix NOUN ajst-29929 141 4 generated generate VERB ajst-29929 141 5 by by ADP ajst-29929 141 6 running run VERB ajst-29929 141 7 the the DET ajst-29929 141 8 model model NOUN ajst-29929 141 9 in in ADP ajst-29929 141 10 this this DET ajst-29929 141 11 paper paper NOUN ajst-29929 141 12 is be AUX ajst-29929 141 13 sh sh PROPN ajst-29929 141 14 own own ADJ ajst-29929 141 15 in in ADP ajst-29929 141 16 figure figure NOUN ajst-29929 141 17 10 10 NUM ajst-29929 141 18 . . PUNCT ajst-29929 142 1 in in ADP ajst-29929 142 2 this this DET ajst-29929 142 3 matrix matrix NOUN ajst-29929 142 4 , , PUNCT ajst-29929 142 5 each each DET ajst-29929 142 6 row row NOUN ajst-29929 142 7 corresponds correspond VERB ajst-29929 142 8 t t NOUN ajst-29929 142 9 o o NOUN ajst-29929 142 10 the the DET ajst-29929 142 11 true true ADJ ajst-29929 142 12 class class NOUN ajst-29929 142 13 , , PUNCT ajst-29929 142 14 while while SCONJ ajst-29929 142 15 each each DET ajst-29929 142 16 column column NOUN ajst-29929 142 17 corresponds correspond VERB ajst-29929 142 18 to to ADP ajst-29929 142 19 the the DET ajst-29929 142 20 pr pr NOUN ajst-29929 142 21 edicted edicte VERB ajst-29929 142 22 class class NOUN ajst-29929 142 23 . . PUNCT ajst-29929 143 1 the the DET ajst-29929 143 2 numbers number NOUN ajst-29929 143 3 within within ADP ajst-29929 143 4 the the DET ajst-29929 143 5 matrix matrix NOUN ajst-29929 143 6 represent represent VERB ajst-29929 143 7 th th X ajst-29929 143 8 e e NOUN ajst-29929 143 9 counts count NOUN ajst-29929 143 10 of of ADP ajst-29929 143 11 instances instance NOUN ajst-29929 143 12 where where SCONJ ajst-29929 143 13 the the DET ajst-29929 143 14 true true ADJ ajst-29929 143 15 class class NOUN ajst-29929 143 16 matches match VERB ajst-29929 143 17 the the DET ajst-29929 143 18 pr pr NOUN ajst-29929 143 19 edicted edicte VERB ajst-29929 143 20 class class NOUN ajst-29929 143 21 at at ADP ajst-29929 143 22 a a DET ajst-29929 143 23 specific specific ADJ ajst-29929 143 24 position position NOUN ajst-29929 143 25 . . PUNCT ajst-29929 144 1 for for ADP ajst-29929 144 2 example example NOUN ajst-29929 144 3 , , PUNCT ajst-29929 144 4 the the DET ajst-29929 144 5 num num ADJ ajst-29929 144 6 ber ber PROPN ajst-29929 144 7 19 19 NUM ajst-29929 144 8 in in ADP ajst-29929 144 9 the the DET ajst-29929 144 10 first first ADJ ajst-29929 144 11 row row NOUN ajst-29929 144 12 and and CCONJ ajst-29929 144 13 first first ADJ ajst-29929 144 14 column column NOUN ajst-29929 144 15 indicates indicate VERB ajst-29929 144 16 that that SCONJ ajst-29929 144 17 the the DET ajst-29929 144 18 model model NOUN ajst-29929 144 19 correctly correctly ADV ajst-29929 144 20 predicted predict VERB ajst-29929 144 21 19 19 NUM ajst-29929 144 22 instances instance NOUN ajst-29929 144 23 when when SCONJ ajst-29929 144 24 the the DET ajst-29929 144 25 actual actual ADJ ajst-29929 144 26 cl cl NOUN ajst-29929 144 27 ass ass NOUN ajst-29929 144 28 was be AUX ajst-29929 144 29 a. a. NOUN ajst-29929 144 30 similarly similarly ADV ajst-29929 144 31 , , PUNCT ajst-29929 144 32 each each DET ajst-29929 144 33 cell cell NOUN ajst-29929 144 34 in in ADP ajst-29929 144 35 the the DET ajst-29929 144 36 matrix matrix NOUN ajst-29929 144 37 provides provide VERB ajst-29929 144 38 inf inf ADJ ajst-29929 144 39 ormation ormation NOUN ajst-29929 144 40 on on ADP ajst-29929 144 41 the the DET ajst-29929 144 42 model model NOUN ajst-29929 144 43 's 's PART ajst-29929 144 44 accuracy accuracy NOUN ajst-29929 144 45 in in ADP ajst-29929 144 46 predicting predict VERB ajst-29929 144 47 different different ADJ ajst-29929 144 48 c c NOUN ajst-29929 144 49 lasses lass NOUN ajst-29929 144 50 . . PUNCT ajst-29929 145 1 the the DET ajst-29929 145 2 results result NOUN ajst-29929 145 3 show show VERB ajst-29929 145 4 that that SCONJ ajst-29929 145 5 the the DET ajst-29929 145 6 model model NOUN ajst-29929 145 7 achieved achieve VERB ajst-29929 145 8 an an DET ajst-29929 145 9 accur accur ADJ ajst-29929 145 10 acy acy NOUN ajst-29929 145 11 of of ADP ajst-29929 145 12 100 100 NUM ajst-29929 145 13 % % NOUN ajst-29929 145 14 in in ADP ajst-29929 145 15 detecting detect VERB ajst-29929 145 16 each each DET ajst-29929 145 17 fault fault NOUN ajst-29929 145 18 category category NOUN ajst-29929 145 19 . . PUNCT ajst-29929 146 1 3.3 3.3 NUM ajst-29929 146 2 . . PUNCT ajst-29929 146 3 comparison comparison NOUN ajst-29929 146 4 with with ADP ajst-29929 146 5 different different ADJ ajst-29929 146 6 models model NOUN ajst-29929 146 7 to to PART ajst-29929 146 8 validate validate VERB ajst-29929 146 9 the the DET ajst-29929 146 10 efficiency efficiency NOUN ajst-29929 146 11 and and CCONJ ajst-29929 146 12 advantages advantage NOUN ajst-29929 146 13 of of ADP ajst-29929 146 14 the the DET ajst-29929 146 15 propos propos PROPN ajst-29929 146 16 ed ed PROPN ajst-29929 146 17 optimized optimize VERB ajst-29929 146 18 cnn cnn PROPN ajst-29929 146 19 - - PUNCT ajst-29929 146 20 bilstm bilstm NOUN ajst-29929 146 21 - - PUNCT ajst-29929 146 22 transformer transformer NOUN ajst-29929 146 23 fault fault NOUN ajst-29929 146 24 diagnosis diagnosis NOUN ajst-29929 146 25 model model NOUN ajst-29929 146 26 in in ADP ajst-29929 146 27 terms term NOUN ajst-29929 146 28 of of ADP ajst-29929 146 29 precision precision NOUN ajst-29929 146 30 and and CCONJ ajst-29929 146 31 execution execution NOUN ajst-29929 146 32 speed speed NOUN ajst-29929 146 33 , , PUNCT ajst-29929 146 34 we we PRON ajst-29929 146 35 com com VERB ajst-29929 146 36 pared pare VERB ajst-29929 146 37 it it PRON ajst-29929 146 38 with with ADP ajst-29929 146 39 the the DET ajst-29929 146 40 cnn cnn PROPN ajst-29929 146 41 - - PUNCT ajst-29929 146 42 transformer transformer NOUN ajst-29929 146 43 and and CCONJ ajst-29929 146 44 cnn cnn PROPN ajst-29929 146 45 - - PUNCT ajst-29929 146 46 bilstm bilstm NOUN ajst-29929 146 47 fa fa PROPN ajst-29929 146 48 ult ult PROPN ajst-29929 146 49 diagnosis diagnosis NOUN ajst-29929 146 50 models model NOUN ajst-29929 146 51 using use VERB ajst-29929 146 52 the the DET ajst-29929 146 53 same same ADJ ajst-29929 146 54 dataset dataset NOUN ajst-29929 146 55 as as ADP ajst-29929 146 56 in in ADP ajst-29929 146 57 this this DET ajst-29929 146 58 stu stu PROPN ajst-29929 146 59 dy dy NOUN ajst-29929 146 60 . . PUNCT ajst-29929 146 61 to to PART ajst-29929 146 62 further far ADV ajst-29929 146 63 examine examine VERB ajst-29929 146 64 the the DET ajst-29929 146 65 generalization generalization NOUN ajst-29929 146 66 ability ability NOUN ajst-29929 146 67 and and CCONJ ajst-29929 146 68 robu robu VERB ajst-29929 146 69 stness stness NOUN ajst-29929 146 70 of of ADP ajst-29929 146 71 the the DET ajst-29929 146 72 models model NOUN ajst-29929 146 73 and and CCONJ ajst-29929 146 74 eliminate eliminate VERB ajst-29929 146 75 the the DET ajst-29929 146 76 influence influence NOUN ajst-29929 146 77 of of ADP ajst-29929 146 78 other other ADJ ajst-29929 146 79 random random ADJ ajst-29929 146 80 factors factor NOUN ajst-29929 146 81 , , PUNCT ajst-29929 146 82 we we PRON ajst-29929 146 83 randomly randomly ADV ajst-29929 146 84 executed execute VERB ajst-29929 146 85 the the DET ajst-29929 146 86 training training NOUN ajst-29929 146 87 proces proce NOUN ajst-29929 146 88 s s PRON ajst-29929 146 89 for for ADP ajst-29929 146 90 each each DET ajst-29929 146 91 model model NOUN ajst-29929 146 92 10 10 NUM ajst-29929 146 93 times time NOUN ajst-29929 146 94 . . PUNCT ajst-29929 147 1 the the DET ajst-29929 147 2 comparative comparative ADJ ajst-29929 147 3 results result NOUN ajst-29929 147 4 of of ADP ajst-29929 147 5 th th X ajst-29929 147 6 e e NOUN ajst-29929 147 7 models model NOUN ajst-29929 147 8 ' ' PART ajst-29929 147 9 performance performance NOUN ajst-29929 147 10 after after ADP ajst-29929 147 11 training training NOUN ajst-29929 147 12 are be AUX ajst-29929 147 13 presented present VERB ajst-29929 147 14 in in ADP ajst-29929 147 15 tabl tabl ADJ ajst-29929 147 16 e e NOUN ajst-29929 147 17 2 2 NUM ajst-29929 147 18 . . PUNCT ajst-29929 147 19 table table NOUN ajst-29929 147 20 2 2 NUM ajst-29929 147 21 . . PUNCT ajst-29929 147 22 comparison comparison NOUN ajst-29929 147 23 of of ADP ajst-29929 147 24 model model NOUN ajst-29929 147 25 performance performance NOUN ajst-29929 147 26 metrics metric NOUN ajst-29929 147 27 comparison comparison NOUN ajst-29929 147 28 model model NOUN ajst-29929 147 29 average average ADJ ajst-29929 147 30 accuracy accuracy NOUN ajst-29929 147 31 average average ADJ ajst-29929 147 32 runtime runtime NOUN ajst-29929 147 33 ( ( PUNCT ajst-29929 147 34 s s NOUN ajst-29929 147 35 ) ) PUNCT ajst-29929 147 36 cnn cnn PROPN ajst-29929 147 37 - - PUNCT ajst-29929 147 38 transformer transformer NOUN ajst-29929 147 39 93.95 93.95 NUM ajst-29929 147 40 733.78 733.78 NUM ajst-29929 147 41 cnn cnn PROPN ajst-29929 147 42 - - PUNCT ajst-29929 147 43 bilstm bilstm NOUN ajst-29929 147 44 cnn cnn PROPN ajst-29929 147 45 - - PUNCT ajst-29929 147 46 bilstm bilstm NOUN ajst-29929 147 47 - - PUNCT ajst-29929 147 48 transformer transformer NOUN ajst-29929 147 49 94.96 94.96 NUM ajst-29929 147 50 99.99 99.99 NUM ajst-29929 147 51 713.35 713.35 NUM ajst-29929 147 52 634.17 634.17 NUM ajst-29929 147 53 4 4 NUM ajst-29929 147 54 . . PUNCT ajst-29929 147 55 conclusion conclusion NOUN ajst-29929 147 56 to to PART ajst-29929 147 57 achieve achieve VERB ajst-29929 147 58 higher high ADJ ajst-29929 147 59 accuracy accuracy NOUN ajst-29929 147 60 and and CCONJ ajst-29929 147 61 faster fast ADJ ajst-29929 147 62 operational operational ADJ ajst-29929 147 63 efficiency efficiency NOUN ajst-29929 147 64 in in ADP ajst-29929 147 65 rolling roll VERB ajst-29929 147 66 bearing bear VERB ajst-29929 147 67 fault fault NOUN ajst-29929 147 68 diagnosis diagnosis NOUN ajst-29929 147 69 , , PUNCT ajst-29929 147 70 this this DET ajst-29929 147 71 paper paper NOUN ajst-29929 147 72 proposes propose VERB ajst-29929 147 73 a a DET ajst-29929 147 74 fault fault NOUN ajst-29929 147 75 diagnosis diagnosis NOUN ajst-29929 147 76 method method NOUN ajst-29929 147 77 that that PRON ajst-29929 147 78 combines combine VERB ajst-29929 147 79 multiple multiple ADJ ajst-29929 147 80 neural neural ADJ ajst-29929 147 81 networks network NOUN ajst-29929 147 82 with with ADP ajst-29929 147 83 cross cross ADJ ajst-29929 147 84 - - ADJ ajst-29929 147 85 attention attention ADJ ajst-29929 147 86 mechanisms mechanism NOUN ajst-29929 147 87 . . PUNCT ajst-29929 148 1 the the DET ajst-29929 148 2 method method NOUN ajst-29929 148 3 employs employ VERB ajst-29929 148 4 fast fast ADJ ajst-29929 148 5 fourier fourier NOUN ajst-29929 148 6 transform transform NOUN ajst-29929 148 7 ( ( PUNCT ajst-29929 148 8 fft fft PROPN ajst-29929 148 9 ) ) PUNCT ajst-29929 148 10 and and CCONJ ajst-29929 148 11 variational variational ADJ ajst-29929 148 12 mode mode NOUN ajst-29929 148 13 decomposition decomposition NOUN ajst-29929 148 14 ( ( PUNCT ajst-29929 148 15 vmd vmd NOUN ajst-29929 148 16 ) ) PUNCT ajst-29929 148 17 to to PART ajst-29929 148 18 extract extract VERB ajst-29929 148 19 timefrequency timefrequency NOUN ajst-29929 148 20 domain domain NOUN ajst-29929 148 21 features feature NOUN ajst-29929 148 22 from from ADP ajst-29929 148 23 the the DET ajst-29929 148 24 signals signal NOUN ajst-29929 148 25 , , PUNCT ajst-29929 148 26 enabling enable VERB ajst-29929 148 27 the the DET ajst-29929 148 28 extraction extraction NOUN ajst-29929 148 29 of of ADP ajst-29929 148 30 multi multi ADJ ajst-29929 148 31 - - ADJ ajst-29929 148 32 scale scale ADJ ajst-29929 148 33 features feature NOUN ajst-29929 148 34 from from ADP ajst-29929 148 35 fault fault NOUN ajst-29929 148 36 signals signal NOUN ajst-29929 148 37 . . PUNCT ajst-29929 149 1 subsequently subsequently ADV ajst-29929 149 2 , , PUNCT ajst-29929 149 3 the the DET ajst-29929 149 4 cnn cnn PROPN ajst-29929 149 5 - - PUNCT ajst-29929 149 6 transformer transformer NOUN ajst-29929 149 7 network network NOUN ajst-29929 149 8 is be AUX ajst-29929 149 9 utilized utilize VERB ajst-29929 149 10 to to PART ajst-29929 149 11 extract extract VERB ajst-29929 149 12 spatial spatial ADJ ajst-29929 149 13 features feature NOUN ajst-29929 149 14 from from ADP ajst-29929 149 15 the the DET ajst-29929 149 16 preprocessed preprocesse VERB ajst-29929 149 17 multi multi ADJ ajst-29929 149 18 - - ADJ ajst-29929 149 19 scale scale ADJ ajst-29929 149 20 features feature NOUN ajst-29929 149 21 of of ADP ajst-29929 149 22 the the DET ajst-29929 149 23 fault fault NOUN ajst-29929 149 24 signals signal NOUN ajst-29929 149 25 , , PUNCT ajst-29929 149 26 while while SCONJ ajst-29929 149 27 the the DET ajst-29929 149 28 bilstm bilstm NOUN ajst-29929 149 29 - - PUNCT ajst-29929 149 30 transformer transformer NOUN ajst-29929 149 31 network network NOUN ajst-29929 149 32 captures capture VERB ajst-29929 149 33 the the DET ajst-29929 149 34 temporal temporal ADJ ajst-29929 149 35 features feature NOUN ajst-29929 149 36 . . PUNCT ajst-29929 150 1 finally finally ADV ajst-29929 150 2 , , PUNCT ajst-29929 150 3 crossattention crossattention NOUN ajst-29929 150 4 is be AUX ajst-29929 150 5 introduced introduce VERB ajst-29929 150 6 to to PART ajst-29929 150 7 fuse fuse VERB ajst-29929 150 8 these these DET ajst-29929 150 9 features feature NOUN ajst-29929 150 10 , , PUNCT ajst-29929 150 11 thereby thereby ADV ajst-29929 150 12 enhancing enhance VERB ajst-29929 150 13 the the DET ajst-29929 150 14 feature feature NOUN ajst-29929 150 15 representation representation NOUN ajst-29929 150 16 capability capability NOUN ajst-29929 150 17 for for ADP ajst-29929 150 18 fault fault NOUN ajst-29929 150 19 signal signal NOUN ajst-29929 150 20 recognition recognition NOUN ajst-29929 150 21 . . PUNCT ajst-29929 151 1 the the DET ajst-29929 151 2 experimental experimental ADJ ajst-29929 151 3 results result NOUN ajst-29929 151 4 demonstrate demonstrate VERB ajst-29929 151 5 that that SCONJ ajst-29929 151 6 the the DET ajst-29929 151 7 proposed propose VERB ajst-29929 151 8 fault fault NOUN ajst-29929 151 9 diagnosis diagnosis NOUN ajst-29929 151 10 method method NOUN ajst-29929 151 11 achieves achieve VERB ajst-29929 151 12 an an DET ajst-29929 151 13 accuracy accuracy NOUN ajst-29929 151 14 of of ADP ajst-29929 151 15 99 99 NUM ajst-29929 151 16 % % NOUN ajst-29929 151 17 and and CCONJ ajst-29929 151 18 outperforms outperform VERB ajst-29929 151 19 other other ADJ ajst-29929 151 20 compared compare VERB ajst-29929 151 21 models model NOUN ajst-29929 151 22 in in ADP ajst-29929 151 23 terms term NOUN ajst-29929 151 24 of of ADP ajst-29929 151 25 operational operational ADJ ajst-29929 151 26 efficiency efficiency NOUN ajst-29929 151 27 . . PUNCT ajst-29929 152 1 references reference NOUN ajst-29929 152 2 [ [ X ajst-29929 152 3 1 1 NUM ajst-29929 152 4 ] ] X ajst-29929 152 5 guo guo PROPN ajst-29929 152 6 y y PROPN ajst-29929 152 7 , , PUNCT ajst-29929 152 8 zhou zhou PROPN ajst-29929 152 9 j j PROPN ajst-29929 152 10 , , PUNCT ajst-29929 152 11 dong dong PROPN ajst-29929 152 12 z z PROPN ajst-29929 152 13 , , PUNCT ajst-29929 152 14 et et PRON ajst-29929 152 15 al.research al.research VERB ajst-29929 152 16 on on ADP ajst-29929 152 17 bearing bear VERB ajst-29929 152 18 fault fault NOUN ajst-29929 152 19 diagnosis diagnosis NOUN ajst-29929 152 20 based base VERB ajst-29929 152 21 on on ADP ajst-29929 152 22 novel novel ADJ ajst-29929 152 23 mrsvd mrsvd NOUN ajst-29929 152 24 - - PUNCT ajst-29929 152 25 cwt cwt NOUN ajst-29929 152 26 and and CCONJ ajst-29929 152 27 improved improve VERB ajst-29929 152 28 cnnlstm[j].measurement cnnlstm[j].measurement NOUN ajst-29929 152 29 science science NOUN ajst-29929 152 30 and and CCONJ ajst-29929 152 31 technology,2024,35(9 technology,2024,35(9 NOUN ajst-29929 152 32 ): ): PUNCT ajst-29929 152 33 [ [ X ajst-29929 152 34 2 2 X ajst-29929 152 35 ] ] X ajst-29929 152 36 kang kang PROPN ajst-29929 152 37 j j PROPN ajst-29929 152 38 , , PUNCT ajst-29929 152 39 zhu zhu PROPN ajst-29929 152 40 x x PUNCT ajst-29929 152 41 , , PUNCT ajst-29929 152 42 shen shen PROPN ajst-29929 152 43 l l PROPN ajst-29929 152 44 , , PUNCT ajst-29929 153 1 et et NOUN ajst-29929 153 2 al.fault al.fault NOUN ajst-29929 153 3 diagnosis diagnosis NOUN ajst-29929 153 4 of of ADP ajst-29929 153 5 a a DET ajst-29929 153 6 wave wave NOUN ajst-29929 153 7 energy energy NOUN ajst-29929 153 8 converter converter NOUN ajst-29929 153 9 gearbox gearbox NOUN ajst-29929 153 10 based base VERB ajst-29929 153 11 on on ADP ajst-29929 153 12 an an DET ajst-29929 153 13 adam adam PROPN ajst-29929 153 14 optimized optimize VERB ajst-29929 153 15 cnn cnn PROPN ajst-29929 153 16 - - PUNCT ajst-29929 153 17 lstm lstm ADJ ajst-29929 153 18 algorithm[j].renewable algorithm[j].renewable ADJ ajst-29929 153 19 energy,2024,231121022 energy,2024,231121022 NOUN ajst-29929 153 20 - - NOUN ajst-29929 153 21 121022 121022 NUM ajst-29929 153 22 . . PUNCT ajst-29929 154 1 [ [ X ajst-29929 154 2 3 3 NUM ajst-29929 154 3 ] ] PUNCT ajst-29929 154 4 xi xi X ajst-29929 154 5 tao tao PROPN ajst-29929 154 6 , , PUNCT ajst-29929 154 7 yang yang PROPN ajst-29929 154 8 weizhen weizhen PROPN ajst-29929 154 9 . . PUNCT ajst-29929 155 1 research research NOUN ajst-29929 155 2 on on ADP ajst-29929 155 3 gearbox gearbox ADJ ajst-29929 155 4 fault fault NOUN ajst-29929 155 5 diagnosis diagnosis NOUN ajst-29929 155 6 based base VERB ajst-29929 155 7 on on ADP ajst-29929 155 8 optimized optimize VERB ajst-29929 155 9 vmd vmd PROPN ajst-29929 155 10 and and CCONJ ajst-29929 155 11 cnn cnn PROPN ajst-29929 156 1 [ [ X ajst-29929 156 2 j j X ajst-29929 156 3 ] ] X ajst-29929 156 4 . . PUNCT ajst-29929 157 1 mechanical mechanical ADJ ajst-29929 157 2 science science NOUN ajst-29929 157 3 and and CCONJ ajst-29929 157 4 technology technology NOUN ajst-29929 157 5 : : PUNCT ajst-29929 157 6 1 1 NUM ajst-29929 157 7 - - SYM ajst-29929 157 8 11 11 NUM ajst-29929 158 1 [ [ X ajst-29929 158 2 4 4 NUM ajst-29929 158 3 ] ] X ajst-29929 158 4 mao mao NOUN ajst-29929 158 5 chenglong chenglong PROPN ajst-29929 158 6 , , PUNCT ajst-29929 158 7 zhang zhang PROPN ajst-29929 158 8 mei mei PROPN ajst-29929 158 9 , , PUNCT ajst-29929 158 10 zhang zhang PROPN ajst-29929 158 11 jie jie PROPN ajst-29929 158 12 . . PUNCT ajst-29929 159 1 rolling roll VERB ajst-29929 159 2 bearing bear VERB ajst-29929 159 3 fault fault NOUN ajst-29929 159 4 diagnosis diagnosis NOUN ajst-29929 159 5 based base VERB ajst-29929 159 6 on on ADP ajst-29929 159 7 mssa mssa PROPN ajst-29929 159 8 - - PUNCT ajst-29929 159 9 vmd vmd PROPN ajst-29929 159 10 and and CCONJ ajst-29929 159 11 cnn cnn PROPN ajst-29929 159 12 - - PUNCT ajst-29929 159 13 bilstm bilstm NOUN ajst-29929 160 1 [ [ X ajst-29929 160 2 j j X ajst-29929 160 3 ] ] X ajst-29929 160 4 . . PUNCT ajst-29929 161 1 journal journal PROPN ajst-29929 161 2 of of ADP ajst-29929 161 3 chongqing chongqing PROPN ajst-29929 161 4 technology technology PROPN ajst-29929 161 5 and and CCONJ ajst-29929 161 6 business business NOUN ajst-29929 161 7 university university NOUN ajst-29929 161 8 ( ( PUNCT ajst-29929 161 9 natural natural ADJ ajst-29929 161 10 science science NOUN ajst-29929 161 11 edition edition NOUN ajst-29929 161 12 ): ): PUNCT ajst-29929 161 13 1 1 NUM ajst-29929 161 14 - - SYM ajst-29929 161 15 10 10 NUM ajst-29929 162 1 [ [ X ajst-29929 162 2 5 5 NUM ajst-29929 162 3 ] ] PUNCT ajst-29929 162 4 wang wang PROPN ajst-29929 162 5 minjuan minjuan PROPN ajst-29929 162 6 , , PUNCT ajst-29929 162 7 jia jia PROPN ajst-29929 162 8 qian qian PROPN ajst-29929 162 9 , , PUNCT ajst-29929 162 10 wang wang PROPN ajst-29929 162 11 youming youming PROPN ajst-29929 162 12 , , PUNCT ajst-29929 162 13 ding ding ADJ ajst-29929 162 14 wenke wenke NOUN ajst-29929 162 15 . . PUNCT ajst-29929 163 1 rolling roll VERB ajst-29929 163 2 bearing bear VERB ajst-29929 163 3 fault fault NOUN ajst-29929 163 4 diagnosis diagnosis NOUN ajst-29929 163 5 method method NOUN ajst-29929 163 6 based base VERB ajst-29929 163 7 on on ADP ajst-29929 163 8 imse imse ADJ ajst-29929 163 9 and and CCONJ ajst-29929 163 10 parameter parameter NOUN ajst-29929 163 11 - - PUNCT ajst-29929 163 12 optimized optimize VERB ajst-29929 163 13 vmd vmd NOUN ajst-29929 164 1 [ [ X ajst-29929 164 2 j j X ajst-29929 164 3 ] ] X ajst-29929 164 4 . . PUNCT ajst-29929 165 1 journal journal PROPN ajst-29929 165 2 of of ADP ajst-29929 165 3 xi'an xi'an PROPN ajst-29929 165 4 university university PROPN ajst-29929 165 5 of of ADP ajst-29929 165 6 posts post NOUN ajst-29929 165 7 and and CCONJ ajst-29929 165 8 telecommunications telecommunication NOUN ajst-29929 165 9 : : PUNCT ajst-29929 165 10 1 1 NUM ajst-29929 165 11 - - SYM ajst-29929 165 12 10 10 NUM ajst-29929 165 13 . . PUNCT ajst-29929 166 1 [ [ X ajst-29929 166 2 6 6 NUM ajst-29929 166 3 ] ] PUNCT ajst-29929 166 4 cao cao PROPN ajst-29929 166 5 jingsheng jingsheng PROPN ajst-29929 166 6 , , PUNCT ajst-29929 166 7 yu yu PROPN ajst-29929 166 8 yang yang PROPN ajst-29929 166 9 , , PUNCT ajst-29929 166 10 wang wang PROPN ajst-29929 166 11 qi qi PROPN ajst-29929 166 12 , , PUNCT ajst-29929 166 13 dong dong PROPN ajst-29929 166 14 yining yining PROPN ajst-29929 166 15 . . PUNCT ajst-29929 167 1 intelligent intelligent ADJ ajst-29929 167 2 fault fault NOUN ajst-29929 167 3 diagnosis diagnosis NOUN ajst-29929 167 4 of of ADP ajst-29929 167 5 motor motor NOUN ajst-29929 167 6 bearings bearing NOUN ajst-29929 167 7 based base VERB ajst-29929 167 8 on on ADP ajst-29929 167 9 optimized optimize VERB ajst-29929 167 10 vmdcnn vmdcnn NOUN ajst-29929 167 11 - - PUNCT ajst-29929 167 12 bilstm bilstm NOUN ajst-29929 168 1 [ [ X ajst-29929 168 2 j j X ajst-29929 168 3 ] ] X ajst-29929 168 4 . . PUNCT ajst-29929 169 1 modern modern ADJ ajst-29929 169 2 electronics electronic NOUN ajst-29929 169 3 technique technique NOUN ajst-29929 169 4 , , PUNCT ajst-29929 169 5 2024 2024 NUM ajst-29929 169 6 , , PUNCT ajst-29929 169 7 47(12 47(12 NUM ajst-29929 169 8 ): ): PUNCT ajst-29929 169 9 115 115 NUM ajst-29929 169 10 - - SYM ajst-29929 169 11 121 121 NUM ajst-29929 169 12 . . PUNCT ajst-29929 170 1 [ [ X ajst-29929 170 2 7 7 NUM ajst-29929 170 3 ] ] PUNCT ajst-29929 170 4 xie xie PROPN ajst-29929 170 5 fengyun fengyun PROPN ajst-29929 170 6 , , PUNCT ajst-29929 170 7 wang wang PROPN ajst-29929 170 8 gan gan PROPN ajst-29929 170 9 , , PUNCT ajst-29929 170 10 shang shang PROPN ajst-29929 170 11 jiandong jiandong PROPN ajst-29929 170 12 , , PUNCT ajst-29929 170 13 fan fan PROPN ajst-29929 170 14 qiuyang qiuyang PROPN ajst-29929 170 15 , , PUNCT ajst-29929 170 16 zhu zhu PROPN ajst-29929 170 17 haiyan haiyan PROPN ajst-29929 170 18 . . PUNCT ajst-29929 171 1 gearbox gearbox PROPN ajst-29929 171 2 fault fault NOUN ajst-29929 171 3 diagnosis diagnosis NOUN ajst-29929 171 4 based base VERB ajst-29929 171 5 on on ADP ajst-29929 171 6 adaptive adaptive ADJ ajst-29929 171 7 variational variational ADJ ajst-29929 171 8 mode mode NOUN ajst-29929 171 9 decomposition decomposition NOUN ajst-29929 172 1 [ [ X ajst-29929 172 2 j j X ajst-29929 172 3 ] ] X ajst-29929 172 4 . . PUNCT ajst-29929 173 1 journal journal PROPN ajst-29929 173 2 of of ADP ajst-29929 173 3 propulsion propulsion NOUN ajst-29929 173 4 technology technology NOUN ajst-29929 173 5 : : PUNCT ajst-29929 173 6 1 1 NUM ajst-29929 173 7 - - SYM ajst-29929 173 8 11 11 NUM ajst-29929 173 9 . . PUNCT ajst-29929 174 1 [ [ X ajst-29929 174 2 8 8 NUM ajst-29929 174 3 ] ] X ajst-29929 174 4 meng meng PROPN ajst-29929 174 5 jingyu jingyu PROPN ajst-29929 174 6 , , PUNCT ajst-29929 174 7 yang yang PROPN ajst-29929 174 8 liming liming PROPN ajst-29929 174 9 , , PUNCT ajst-29929 174 10 zhang zhang PROPN ajst-29929 174 11 cheng cheng PROPN ajst-29929 174 12 , , PUNCT ajst-29929 174 13 wu wu PROPN ajst-29929 174 14 boyang boyang PROPN ajst-29929 174 15 , , PUNCT ajst-29929 174 16 xu xu PROPN ajst-29929 175 1 guoping guoping PROPN ajst-29929 175 2 , , PUNCT ajst-29929 175 3 yu yu PROPN ajst-29929 175 4 jian jian PROPN ajst-29929 175 5 . . PROPN ajst-29929 176 1 feature feature NOUN ajst-29929 176 2 extraction extraction NOUN ajst-29929 176 3 and and CCONJ ajst-29929 176 4 diagnosis diagnosis NOUN ajst-29929 176 5 of of ADP ajst-29929 176 6 wind wind NOUN ajst-29929 176 7 turbine turbine NOUN ajst-29929 176 8 gearbox gearbox NOUN ajst-29929 176 9 faults fault NOUN ajst-29929 176 10 based base VERB ajst-29929 176 11 on on ADP ajst-29929 176 12 zoom zoom NOUN ajst-29929 176 13 - - PUNCT ajst-29929 176 14 fft fft NOUN ajst-29929 176 15 - - PUNCT ajst-29929 176 16 ceemd ceemd ADJ ajst-29929 176 17 and and CCONJ ajst-29929 176 18 wavelet wavelet NOUN ajst-29929 176 19 packet packet NOUN ajst-29929 176 20 denoising denoising NOUN ajst-29929 177 1 [ [ X ajst-29929 177 2 j j X ajst-29929 177 3 ] ] X ajst-29929 177 4 . . PUNCT ajst-29929 178 1 micro micro PROPN ajst-29929 178 2 motors motors PROPN ajst-29929 178 3 , , PUNCT ajst-29929 178 4 2024 2024 NUM ajst-29929 178 5 , , PUNCT ajst-29929 178 6 52(04 52(04 NOUN ajst-29929 178 7 ): ): PUNCT ajst-29929 178 8 2832=37 2832=37 NUM ajst-29929 178 9 . . PUNCT ajst-29929 179 1 [ [ X ajst-29929 179 2 9 9 NUM ajst-29929 179 3 ] ] X ajst-29929 179 4 gao gao PROPN ajst-29929 179 5 hongwei hongwei PROPN ajst-29929 179 6 , , PUNCT ajst-29929 179 7 li li PROPN ajst-29929 179 8 xincheng xincheng PROPN ajst-29929 179 9 , , PUNCT ajst-29929 179 10 he he PRON ajst-29929 179 11 xiaoning xiaoning PROPN ajst-29929 179 12 , , PUNCT ajst-29929 179 13 liu liu PROPN ajst-29929 179 14 chang'an chang'an PROPN ajst-29929 179 15 , , PUNCT ajst-29929 179 16 wang wang PROPN ajst-29929 179 17 hanchi hanchi PROPN ajst-29929 179 18 , , PUNCT ajst-29929 179 19 li li PROPN ajst-29929 179 20 wencheng wencheng PROPN ajst-29929 179 21 , , PUNCT ajst-29929 179 22 zhang zhang PROPN ajst-29929 179 23 baohui baohui PROPN ajst-29929 179 24 . . PUNCT ajst-29929 180 1 research research NOUN ajst-29929 180 2 on on ADP ajst-29929 180 3 wind wind NOUN ajst-29929 180 4 turbine turbine NOUN ajst-29929 180 5 gearbox gearbox NOUN ajst-29929 180 6 fault fault NOUN ajst-29929 180 7 diagnosis diagnosis NOUN ajst-29929 180 8 method method NOUN ajst-29929 180 9 based base VERB ajst-29929 180 10 on on ADP ajst-29929 180 11 hilbert hilbert NOUN ajst-29929 180 12 transform transform NOUN ajst-29929 180 13 [ [ X ajst-29929 180 14 j j X ajst-29929 180 15 ] ] X ajst-29929 180 16 . . PUNCT ajst-29929 181 1 machine machine NOUN ajst-29929 181 2 tool tool PROPN ajst-29929 181 3 & & CCONJ ajst-29929 181 4 hydraulics hydraulic NOUN ajst-29929 181 5 , , PUNCT ajst-29929 181 6 2024 2024 NUM ajst-29929 181 7 , , PUNCT ajst-29929 181 8 52(09 52(09 NUM ajst-29929 181 9 ): ): PUNCT ajst-29929 181 10 215220 215220 NUM ajst-29929 181 11 . . PUNCT ajst-29929 182 1 [ [ X ajst-29929 182 2 10 10 NUM ajst-29929 182 3 ] ] X ajst-29929 182 4 wei wei PROPN ajst-29929 182 5 hangxin hangxin PROPN ajst-29929 182 6 , , PUNCT ajst-29929 182 7 cheng cheng PROPN ajst-29929 182 8 huan huan PROPN ajst-29929 182 9 , , PUNCT ajst-29929 182 10 wu wu PROPN ajst-29929 182 11 wei wei PROPN ajst-29929 182 12 , , PUNCT ajst-29929 182 13 wang wang PROPN ajst-29929 182 14 xiaorong xiaorong PROPN ajst-29929 182 15 . . PUNCT ajst-29929 183 1 rolling roll VERB ajst-29929 183 2 bearing bear VERB ajst-29929 183 3 fault fault NOUN ajst-29929 183 4 diagnosis diagnosis NOUN ajst-29929 183 5 based base VERB ajst-29929 183 6 on on ADP ajst-29929 183 7 vmd vmd PROPN ajst-29929 183 8 visualization visualization NOUN ajst-29929 183 9 and and CCONJ ajst-29929 183 10 deep deep ADJ ajst-29929 183 11 learning learning NOUN ajst-29929 184 1 [ [ X ajst-29929 184 2 j j X ajst-29929 184 3 ] ] X ajst-29929 184 4 . . PUNCT ajst-29929 185 1 mechanical mechanical ADJ ajst-29929 185 2 design design NOUN ajst-29929 185 3 and and CCONJ ajst-29929 185 4 manufacture manufacture NOUN ajst-29929 185 5 : : PUNCT ajst-29929 185 6 1 1 NUM ajst-29929 185 7 - - SYM ajst-29929 185 8 6 6 NUM ajst-29929 185 9 . . PUNCT ajst-29929 186 1 [ [ X ajst-29929 186 2 11 11 NUM ajst-29929 186 3 ] ] X ajst-29929 186 4 zhu zhu PROPN ajst-29929 186 5 junjie junjie PROPN ajst-29929 186 6 , , PUNCT ajst-29929 186 7 zhang zhang PROPN ajst-29929 186 8 qinghua qinghua PROPN ajst-29929 186 9 , , PUNCT ajst-29929 186 10 zhu zhu PROPN ajst-29929 186 11 guanhua guanhua PROPN ajst-29929 186 12 , , PUNCT ajst-29929 186 13 su su PROPN ajst-29929 186 14 naiquan naiquan PROPN ajst-29929 186 15 . . PUNCT ajst-29929 187 1 rolling roll VERB ajst-29929 187 2 bearing bear VERB ajst-29929 187 3 fault fault NOUN ajst-29929 187 4 diagnosis diagnosis NOUN ajst-29929 187 5 based base VERB ajst-29929 187 6 on on ADP ajst-29929 187 7 eemd eemd PROPN ajst-29929 187 8 and and CCONJ ajst-29929 187 9 cnnsvm cnnsvm NOUN ajst-29929 187 10 [ [ X ajst-29929 187 11 j j X ajst-29929 187 12 ] ] X ajst-29929 187 13 . . PUNCT ajst-29929 188 1 machine machine NOUN ajst-29929 188 2 tool tool PROPN ajst-29929 188 3 & & CCONJ ajst-29929 188 4 hydraulics hydraulic NOUN ajst-29929 188 5 : : PUNCT ajst-29929 188 6 1 1 NUM ajst-29929 188 7 - - SYM ajst-29929 188 8 9 9 NUM ajst-29929 188 9 . . PUNCT ajst-29929 189 1 [ [ X ajst-29929 189 2 12 12 NUM ajst-29929 189 3 ] ] X ajst-29929 189 4 zhou zhou PROPN ajst-29929 189 5 ruifeng ruifeng PROPN ajst-29929 189 6 . . PUNCT ajst-29929 190 1 research research NOUN ajst-29929 190 2 and and CCONJ ajst-29929 190 3 implementation implementation NOUN ajst-29929 190 4 of of ADP ajst-29929 190 5 intelligent intelligent ADJ ajst-29929 190 6 fault fault NOUN ajst-29929 190 7 diagnosis diagnosis NOUN ajst-29929 190 8 methods method NOUN ajst-29929 190 9 for for ADP ajst-29929 190 10 rolling roll VERB ajst-29929 190 11 bearings bearing NOUN ajst-29929 191 1 [ [ X ajst-29929 191 2 d d X ajst-29929 191 3 ] ] X ajst-29929 191 4 . . PUNCT ajst-29929 192 1 dalian dalian PROPN ajst-29929 192 2 university university PROPN ajst-29929 192 3 of of ADP ajst-29929 192 4 technology technology NOUN ajst-29929 192 5 , , PUNCT ajst-29929 192 6 2009 2009 NUM ajst-29929 192 7 . . PUNCT ajst-29929 193 1 [ [ X ajst-29929 193 2 13 13 NUM ajst-29929 193 3 ] ] PUNCT ajst-29929 193 4 xu xu PROPN ajst-29929 193 5 jingwen jingwen PROPN ajst-29929 193 6 , , PUNCT ajst-29929 193 7 yang yang PROPN ajst-29929 193 8 ping ping PROPN ajst-29929 193 9 , , PUNCT ajst-29929 193 10 yin yin PROPN ajst-29929 193 11 xiaojun xiaojun PROPN ajst-29929 193 12 . . PUNCT ajst-29929 194 1 gear gear PROPN ajst-29929 194 2 fault fault NOUN ajst-29929 194 3 diagnosis diagnosis NOUN ajst-29929 194 4 based base VERB ajst-29929 194 5 on on ADP ajst-29929 194 6 emd emd PROPN ajst-29929 194 7 decomposition decomposition NOUN ajst-29929 194 8 and and CCONJ ajst-29929 194 9 levy levy NOUN ajst-29929 194 10 - - PUNCT ajst-29929 194 11 ssa ssa NOUN ajst-29929 194 12 - - PUNCT ajst-29929 194 13 bp bp PROPN ajst-29929 194 14 neural neural ADJ ajst-29929 194 15 network network NOUN ajst-29929 195 1 [ [ X ajst-29929 195 2 j j X ajst-29929 195 3 ] ] X ajst-29929 195 4 . . PUNCT ajst-29929 196 1 mechanical mechanical ADJ ajst-29929 196 2 transmission transmission NOUN ajst-29929 196 3 , , PUNCT ajst-29929 196 4 2024 2024 NUM ajst-29929 196 5 , , PUNCT ajst-29929 196 6 48(05 48(05 NUM ajst-29929 196 7 ): ): PUNCT ajst-29929 196 8 152 152 NUM ajst-29929 196 9 - - SYM ajst-29929 196 10 157 157 NUM ajst-29929 196 11 . . PUNCT ajst-29929 197 1 [ [ X ajst-29929 197 2 14 14 NUM ajst-29929 197 3 ] ] X ajst-29929 197 4 long long PROPN ajst-29929 197 5 xiafei xiafei PROPN ajst-29929 197 6 , , PUNCT ajst-29929 197 7 he he PRON ajst-29929 197 8 zhicheng zhicheng PROPN ajst-29929 197 9 , , PUNCT ajst-29929 197 10 zhou zhou PROPN ajst-29929 197 11 ling ling PROPN ajst-29929 197 12 , , PUNCT ajst-29929 197 13 liu liu PROPN ajst-29929 197 14 weiqiang weiqiang PROPN ajst-29929 197 15 , , PUNCT ajst-29929 197 16 liang liang PROPN ajst-29929 197 17 kai kai PROPN ajst-29929 197 18 . . PUNCT ajst-29929 198 1 research research NOUN ajst-29929 198 2 on on ADP ajst-29929 198 3 wind wind NOUN ajst-29929 198 4 turbine turbine NOUN ajst-29929 198 5 gearbox gearbox NOUN ajst-29929 198 6 fault fault NOUN ajst-29929 198 7 diagnosis diagnosis NOUN ajst-29929 198 8 based base VERB ajst-29929 198 9 on on ADP ajst-29929 198 10 koa koa PROPN ajst-29929 198 11 - - PUNCT ajst-29929 198 12 cnn cnn PROPN ajst-29929 198 13 - - PUNCT ajst-29929 198 14 bilstm bilstm NOUN ajst-29929 198 15 - - PUNCT ajst-29929 198 16 am am NOUN ajst-29929 198 17 [ [ X ajst-29929 198 18 j j X ajst-29929 198 19 ] ] X ajst-29929 198 20 . . PUNCT ajst-29929 199 1 machine machine NOUN ajst-29929 199 2 tool tool PROPN ajst-29929 199 3 & & CCONJ ajst-29929 199 4 hydraulics hydraulic NOUN ajst-29929 199 5 : : PUNCT ajst-29929 199 6 1 1 NUM ajst-29929 199 7 - - SYM ajst-29929 199 8 10 10 NUM ajst-29929 199 9 . . PUNCT ajst-29929 200 1 [ [ X ajst-29929 200 2 15 15 NUM ajst-29929 200 3 ] ] X ajst-29929 200 4 wang wang PROPN ajst-29929 200 5 jinhua jinhua PROPN ajst-29929 200 6 , , PUNCT ajst-29929 200 7 liu liu PROPN ajst-29929 200 8 zhengqi zhengqi PROPN ajst-29929 200 9 , , PUNCT ajst-29929 200 10 cao cao PROPN ajst-29929 200 11 jie jie PROPN ajst-29929 200 12 , , PUNCT ajst-29929 200 13 liu liu PROPN ajst-29929 200 14 yunqiang yunqiang PROPN ajst-29929 200 15 , , PUNCT ajst-29929 200 16 chen chen PROPN ajst-29929 200 17 li li PROPN ajst-29929 200 18 . . PROPN ajst-29929 200 19 gearbox gearbox PROPN ajst-29929 200 20 fault fault NOUN ajst-29929 200 21 diagnosis diagnosis NOUN ajst-29929 200 22 based base VERB ajst-29929 200 23 on on ADP ajst-29929 200 24 r r NOUN ajst-29929 200 25 vine vine NOUN ajst-29929 200 26 copula copula NOUN ajst-29929 200 27 - - PUNCT ajst-29929 200 28 dbn dbn NOUN ajst-29929 200 29 [ [ X ajst-29929 200 30 j j X ajst-29929 200 31 ] ] X ajst-29929 200 32 . . PUNCT ajst-29929 201 1 journal journal PROPN ajst-29929 201 2 of of ADP ajst-29929 201 3 beijing beijing PROPN ajst-29929 201 4 university university PROPN ajst-29929 201 5 of of ADP ajst-29929 201 6 aeronautics aeronautics PROPN ajst-29929 201 7 and and CCONJ ajst-29929 201 8 astronautics astronautic NOUN ajst-29929 201 9 : : PUNCT ajst-29929 201 10 1 1 NUM ajst-29929 201 11 - - SYM ajst-29929 201 12 15 15 NUM ajst-29929 201 13 . . PUNCT ajst-29929 202 1 [ [ X ajst-29929 202 2 16 16 NUM ajst-29929 202 3 ] ] X ajst-29929 202 4 wang wang PROPN ajst-29929 202 5 chengbing chengbe VERB ajst-29929 202 6 . . PUNCT ajst-29929 203 1 gearbox gearbox NOUN ajst-29929 203 2 fault fault NOUN ajst-29929 203 3 diagnosis diagnosis NOUN ajst-29929 203 4 based base VERB ajst-29929 203 5 on on ADP ajst-29929 203 6 multisensor multisensor NOUN ajst-29929 203 7 information information NOUN ajst-29929 203 8 fusion fusion NOUN ajst-29929 203 9 and and CCONJ ajst-29929 203 10 dual dual ADJ ajst-29929 203 11 - - PUNCT ajst-29929 203 12 stream stream NOUN ajst-29929 203 13 cnn cnn PROPN ajst-29929 204 1 [ [ X ajst-29929 204 2 j j X ajst-29929 204 3 ] ] X ajst-29929 204 4 . . PUNCT ajst-29929 205 1 construction construction NOUN ajst-29929 205 2 machinery machinery NOUN ajst-29929 205 3 , , PUNCT ajst-29929 205 4 2024 2024 NUM ajst-29929 205 5 , , PUNCT ajst-29929 205 6 55(04 55(04 NUM ajst-29929 205 7 ): ): PUNCT ajst-29929 205 8 69 69 NUM ajst-29929 205 9 - - SYM ajst-29929 205 10 78 78 NUM ajst-29929 205 11 + + NOUN ajst-29929 205 12 10 10 NUM ajst-29929 205 13 - - SYM ajst-29929 205 14 11 11 NUM ajst-29929 205 15 . . PUNCT ajst-29929 206 1 199 199 NUM ajst-29929 207 1 [ [ X ajst-29929 207 2 17 17 NUM ajst-29929 207 3 ] ] X ajst-29929 207 4 xu xu PROPN ajst-29929 207 5 m m PROPN ajst-29929 207 6 , , PUNCT ajst-29929 207 7 yu yu PROPN ajst-29929 207 8 q q PROPN ajst-29929 207 9 , , PUNCT ajst-29929 207 10 chen chen PROPN ajst-29929 207 11 s s PART ajst-29929 207 12 , , PUNCT ajst-29929 207 13 et et NOUN ajst-29929 207 14 al.rolling al.rolle VERB ajst-29929 207 15 bearing bear VERB ajst-29929 207 16 fault fault NOUN ajst-29929 207 17 diagnosis diagnosis NOUN ajst-29929 207 18 based base VERB ajst-29929 207 19 on on ADP ajst-29929 207 20 cnn cnn PROPN ajst-29929 207 21 - - PUNCT ajst-29929 207 22 lstm lstm PROPN ajst-29929 207 23 with with ADP ajst-29929 207 24 fft fft PROPN ajst-29929 207 25 and and CCONJ ajst-29929 207 26 svd svd PROPN ajst-29929 208 1 [ [ X ajst-29929 208 2 j j X ajst-29929 208 3 ] ] X ajst-29929 208 4 . . PUNCT ajst-29929 209 1 information information NOUN ajst-29929 209 2 , , PUNCT ajst-29929 209 3 2024 2024 NUM ajst-29929 209 4 , , PUNCT ajst-29929 209 5 15(7):399 15(7):399 NUM ajst-29929 209 6 - - SYM ajst-29929 209 7 399 399 NUM ajst-29929 209 8 . . PUNCT ajst-29929 210 1 [ [ X ajst-29929 210 2 18 18 NUM ajst-29929 210 3 ] ] PUNCT ajst-29929 210 4 yu yu PROPN ajst-29929 210 5 t t PROPN ajst-29929 210 6 , , PUNCT ajst-29929 210 7 li li PROPN ajst-29929 210 8 c c PROPN ajst-29929 210 9 .fault .fault PUNCT ajst-29929 210 10 diagnosis diagnosis NOUN ajst-29929 210 11 method method NOUN ajst-29929 210 12 of of ADP ajst-29929 210 13 rolling rolling ADJ ajst-29929 210 14 bearing bearing NOUN ajst-29929 210 15 based base VERB ajst-29929 210 16 on on ADP ajst-29929 210 17 mssa mssa ADJ ajst-29929 210 18 - - PUNCT ajst-29929 210 19 vmd vmd NOUN ajst-29929 210 20 algorithm[j].journal algorithm[j].journal PROPN ajst-29929 210 21 of of ADP ajst-29929 210 22 physics physics PROPN ajst-29929 210 23 : : PUNCT ajst-29929 210 24 conference conference NOUN ajst-29929 210 25 series series NOUN ajst-29929 210 26 , , PUNCT ajst-29929 210 27 2024,2752(1 2024,2752(1 NUM ajst-29929 210 28 ): ): PUNCT ajst-29929 210 29 [ [ X ajst-29929 210 30 19 19 NUM ajst-29929 210 31 ] ] PUNCT ajst-29929 210 32 cui cui NOUN ajst-29929 210 33 k k PROPN ajst-29929 210 34 , , PUNCT ajst-29929 210 35 liu liu PROPN ajst-29929 210 36 m m PROPN ajst-29929 210 37 , , PUNCT ajst-29929 210 38 meng meng PROPN ajst-29929 210 39 y y PROPN ajst-29929 210 40 .a .a PROPN ajst-29929 211 1 new new ADJ ajst-29929 211 2 fault fault NOUN ajst-29929 211 3 diagnosis diagnosis NOUN ajst-29929 211 4 of of ADP ajst-29929 211 5 rolling rolling ADJ ajst-29929 211 6 bearing bearing NOUN ajst-29929 211 7 on on ADP ajst-29929 211 8 fft fft NOUN ajst-29929 211 9 image image NOUN ajst-29929 211 10 coding coding NOUN ajst-29929 211 11 and and CCONJ ajst-29929 211 12 l l NOUN ajst-29929 211 13 - - PROPN ajst-29929 211 14 cnn[j].measurement cnn[j].measurement ADJ ajst-29929 211 15 science science NOUN ajst-29929 211 16 and and CCONJ ajst-29929 211 17 technology,2024,35(7 technology,2024,35(7 NUM ajst-29929 211 18 ): ): PUNCT ajst-29929 212 1 [ [ X ajst-29929 212 2 20 20 NUM ajst-29929 212 3 ] ] PUNCT ajst-29929 212 4 zhenzhen zhenzhen PROPN ajst-29929 212 5 j j PROPN ajst-29929 212 6 , , PUNCT ajst-29929 212 7 deqiang deqiang PROPN ajst-29929 212 8 h h PROPN ajst-29929 212 9 , , PUNCT ajst-29929 212 10 zexian zexian PROPN ajst-29929 212 11 w w PROPN ajst-29929 212 12 .intelligent .intelligent PROPN ajst-29929 212 13 fault fault VERB ajst-29929 212 14 diagnosis diagnosis NOUN ajst-29929 212 15 of of ADP ajst-29929 212 16 train train NOUN ajst-29929 212 17 axle axle PROPN ajst-29929 212 18 box box PROPN ajst-29929 212 19 bearing bearing NOUN ajst-29929 212 20 based base VERB ajst-29929 212 21 on on ADP ajst-29929 212 22 parameter parameter PROPN ajst-29929 212 23 optimization optimization PROPN ajst-29929 212 24 vmd vmd PROPN ajst-29929 212 25 and and CCONJ ajst-29929 212 26 improved improve VERB ajst-29929 212 27 dbn[j].engineering dbn[j].engineering PROPN ajst-29929 212 28 applications application NOUN ajst-29929 212 29 of of ADP ajst-29929 212 30 artificial artificial ADJ ajst-29929 212 31 intelligence,2022,110 intelligence,2022,110 NOUN ajst-29929 212 32 . . PUNCT