Reliable extraction of fault-related features from high-noise vibration signals is crucial for diagnosing incipient defects in large-scale industrial bearings. However, the empirical wavelet transform (EWT) is highly sensitive to boundary definition and mode selection, which limits its robustness in complex operating conditions. This article proposes a boundary-optimized EWT (BEWT) framework that introduces a globally optimized boundary selection strategy: a mechanism-based frequency band boundary filter, designed from the target bearing’s vibration characteristics, is used to determine and adaptively fuse EWT subbands, thereby improving decomposition accuracy. In addition, a mechanism-informed cyclic frequency optimization strategy is developed for second-order cyclostationary blind deconvolution (CYCBD), which strengthens its noise suppression capability and guides the design of internal and external filters within BEWT. By integrating BEWT and CYCBD, the proposed method achieves accurate identification of resonance modulation bands and robust fault feature extraction under strong background noise. Simulation studies and real-world applications, including deployment in a hot rolling mill production line for early detection of rolling element defects, demonstrate that the proposed approach outperforms traditional EWT-based methods in noise resistance and diagnostic accuracy, particularly for high-speed, heavy-load bearing systems. After applying CYCBD-BEWT on the production line, the optimal frequency band achieved an SNR improvement of more than 5 dB.
LeiYJiaFLinJ, et al. An intelligent fault diagnosis method using unsupervised feature learning toward mechanical big data. IEEE Trans Ind Electron2016; 63(5): 3137–3147.
2.
ZhaoDLiJChengW, et al. Bearing multifault diagnosis with iterative generalized demodulation guided by enhanced rotational frequency matching under time-varying speed conditions. ISA Trans2023; 133: 518–528.
3.
YuYLiYYangP, et al. TFC-TFECR feature extraction and state recognition of acoustic emission signal of cylindrical roller bearing. Nondestruct Test Eval2024; 46: 1–19.
GaoJ.Machine fault diagnosis and treatment and self-healing, vol. 42. Beijing: Higher Education Press, 2012, pp. 76–89.
6.
ZhangCKangZLvC, et al. Photovoltaic DC series arc fault identification method based on precise noise reduction algorithm. IEEE Sens J2024; 24(4): 4746–4757.
7.
ChenJPanJLiZ, et al. Generator bearing fault diagnosis for wind turbine via empirical wavelet transform using measured vibration signals. Renew Energy2016; 89: 80–92.
8.
DengWZhangSZhaoH, et al. A novel fault diagnosis method based on integrating empirical wavelet transform and fuzzy entropy for motor bearing. IEEE Access2018; 6: 35042–35056.
9.
GillesJ.Empirical wavelet transform. IEEE Trans Signal Process2013; 61(16): 3999–4010.
10.
XuYZhangKMaC, et al. Adaptive kurtogram and its applications in rolling bearing fault diagnosis. Mech Syst Signal Process2019; 130: 87–107.
11.
GeMWangJRenX.Fault diagnosis of rolling bearings based on EWT and KDEC. Entropy2017; 19(12): 601–633.
12.
CheginiSNBagheriANajafiF.Application of a new EWT-based denoising technique in bearing fault diagnosis. Measurement2019; 144: 275–297.
13.
PengDZhuXTengW, et al. Use of generalized Gaussian cyclostationarity for blind deconvolution and its application to bearing fault diagnosis under non-Gaussian conditions. Mech Syst Signal Process2023; 196: 110302–110351.
14.
LiJWangHLiM, et al. Rolling bearing fault diagnosis based on improved adaptive nonreferential empirical wavelet transform. J Metrol2020; 41(6): 710–716.
15.
YuKLinTRTanJW, et al. An adaptive sensitive frequency band selection method for empirical wavelet transform and its application in bearing fault diagnosis. Measurement2022; 134: 375–384.
16.
XueJXuHLiuX, et al. Application of enhanced empirical wavelet transform and correlation kurtosis in bearing fault diagnosis. Meas Sci Technol2023; 34(3): 23–35.
17.
FeiCChoyYBaiG, et al. Multi-feature entropy distance approach with vibration and acoustic emission signals for process feature recognition of rolling element bearing faults. Struct Health Monit2018; 17(2): 156–168.
18.
InturiVBalajiSVGyanamP, et al. An integrated condition monitoring scheme for health state identification of a multi-stage gearbox through Hurst exponent estimates. Struct Health Monit2023; 22(1): 730–745.
19.
InturiVSabareeshGRPenumakalaPK.Bearing fault severity analysis on a multi-stage gearbox subjected to fluctuating speeds. Exp Tech2020; 44: 541–552.
20.
InturiVSabareeshGRSupradeepanK, et al. Integrated condition monitoring scheme for bearing fault diagnosis of a wind turbine gearbox. J Vib Control2019; 25(12): 1852–1865.
21.
InturiVTalwarRSabareeshGR.Combined vibro-acoustic analysis to identify the health state of a multi-stage gearbox through Hjorth’s parameters. J Vib Eng Technol2025; 13: 543.
22.
MaPLiangWZhangH, et al. Multiscale permutation entropy based on natural visibility graph and its application to rolling bearing fault diagnosis. Struct Health Monit2025; 24(1): 313–326.
23.
ZhouXLiAHanG.An intelligent multi-local model bearing fault diagnosis method using small sample fusion. Sensors2023; 23(17): 67–75.
24.
BorghesaniPPennacchiPChattertonS.The relationship between kurtosis- and envelope-based indexes for the diagnostic of rolling element bearings. Mech Syst Signal Process2014; 43(2): 25–43.
25.
MoZLZhangHShenY, et al. Conditional empirical wavelet transform with modified ratio of cyclic content for bearing fault diagnosis. ISA Trans2023; 133: 597–611.
26.
LiuCHeCHanT, et al. Low-variance version of the RCC index and form factor index for machine condition monitoring. Mech Syst Signal Process2023; 200: 110578–110614.
27.
JiangTZhangQ.Bearing failure impulse enhancement method using multiple resonance band center positioning and envelope integration. Measurement2022; 220: 111597–111623.
28.
HeJShiPXuX, et al. Rolling mill fault diagnosis under limited datasets. Knowl Based Syst2024; 291: 75–97.
29.
SunJGuoHGuoX, et al. The simulation analysis and experimental study on the temperature field of four-row rolling bearings of rolling mill under non-uniform load conditions. Sensors2024; 24(3): 874–914.
30.
DanielLLNLeiteDFQGomesACA, et al. Sound-based diagnosis of failure in a vertical roller mill. J Braz Soc Mech Sci Eng2024; 46(4): 102–151.
31.
BuzzoniMAntoniJDeliaG.Blind deconvolution based on cyclostationarity maximization and its application to fault identification. J Sound Vib2018; 432: 569–601.
ZhangYLvYGeM.Time-frequency analysis via complementary ensemble adaptive local iterative filtering and enhanced maximum correlation kurtosis deconvolution for wind turbine fault diagnosis. Energy Rep2021; 7: 2418–2435.
34.
XiongMMLvYDangZ, et al. Early fault diagnosis of rolling bearings based on parameter-adaptive multipoint optimal minimum entropy deconvolution adjusted and dynamic mode decomposition. Meas Sci Technol2022; 33(12): 27–41.
35.
ChenBZhangWSongD, et al. Squared envelope sparsification via blind deconvolution and its application to railway axle bearing diagnostics. Struct Health Monit2023; 22(6): 3637–3658.
36.
HuoWJiangZShengZ, et al. Cyclostationarity blind deconvolution via eigenvector screening and its applications to the condition monitoring of rotating machinery. Mech Syst Signal Process2025; 222: 111721–111782.
37.
YangJZhouCLiX, et al. A fault feature extraction method based on improved VMD multi-scale dispersion entropy and TVD-CYCBD. Entropy2023; 25(2): 241–277.
38.
WangZZhouJDuW, et al. Bearing fault diagnosis method based on adaptive maximum cyclostationarity blind deconvolution. Mech Syst Signal Process2022; 162: 108–181.
39.
ZhangCZhangXXuZ, et al. An improved denoising method for fault vibration signals of wind turbine gearbox bearings. Machines2023; 11(11): 971–1004.
40.
ChenRHuangYXuX, et al. Rolling bearing fault feature extraction method using adaptive maximum cyclostationarity blind deconvolution. IEEE Sens J2023; 23(15): 17761–17770.
41.
WangXTangGHeY.Application of RSSD-OCYCBD strategy in enhanced fault detection of rolling bearing. Complexity2020; 2020: 1–24.
42.
SunHLiangFLiuY, et al. Application of a novel improved adaptive CYCBD method in gearbox compound fault diagnosis. IEEE Access2021; 9: 133835–133848.
43.
LiJLiuYXiangJ.Optimal maximum cyclostationary blind deconvolution for bearing fault detection. IEEE Sens J2023; 23(14): 15975–15987.
44.
YuYQianMChenT, et al. A new method to select frequency band for vibration signal demodulation and condition estimation of rolling bearings. ISA Trans2023; 133: 575–596.
45.
DongYZhangSZhangH, et al. Chaotic evolution optimization: a novel metaheuristic algorithm inspired by chaotic dynamics. Chaos Solitons Fractals2025; 192: 116049.
46.
ZhaoCZioEShenW.Domain generalization for cross-domain fault diagnosis: an application-oriented perspective and a benchmark study. Reliab Eng Syst Saf2024; 187: 7–15.
47.
ZhuoSBaiXHanJ, et al. A novel rolling bearing fault diagnosis method based on the NEITD-ADTL-JS algorithm. Sensors2025; 25(3): 857–873.
48.
XEWangWYuanH. Bearing fault feature extraction method based on adaptive time-varying filtering empirical mode decomposition and singular value decomposition denoising. Machines2025; 13(1): 3–50.
49.
LuJJiaBLiS, et al. A noise reduction method of rolling bearing based on empirical wavelet transform and adaptive time frequency peak filtering. Meas Sci Technol2023; 34(12): 125107–125146.
50.
LiKWuHLiuX, et al. Intelligent diagnosis of rolling bearing based on ICEEMDAN-WTD of noise reduction and multi-strategy fusion optimization SCNs. IEEE Access2024; 12: 36908–36923.
