Online residual useful life prediction of large-size slewing bearings—A data fusion method

Abstract

To decrease breakdown time and improve machine operation reliability, accurate residual useful life (RUL) prediction has been playing a critical role in condition based monitoring. A data fusion method was proposed to achieve online RUL prediction of slewing bearings, which consisted of a reliability based RUL prediction model and a data driven failure rate (FR) estimation model. Firstly, an RUL prediction model was developed based on modified Weibull distribution to build the relationship between RUL and FR. Secondly, principal component analysis (PCA) was introduced to process multi-dimensional life-cycle vibration signals, and continuous squared prediction error (CSPE) and its time-domain features were employed as equipment performance degradation features. Afterwards, an FR estimation model was established on basis of the degradation features and relevant FRs using simplified fuzzy adaptive resonance theory map (SFAM) neural network. Consequently, real-time FR of equipment can be obtained through FR estimation model, and then accurate RUL can be calculated through the RUL prediction model. Results of a slewing bearing life test show that CSPE is an effective indicator of performance degradation process of slewing bearings, and that by combining actual load condition and real-time monitored data, the calculation time is reduced by 87.3% and the accuracy is increased by 0.11%, which provides a potential for online RUL prediction of slewing bearings and other various machineries.

Keywords

slewing bearing / life prediction / Weibull distribution / failure rate estimation / data fusion

References

[1] PotocnikR, FlaskerJ, GlodezS. Fatigue analysis of large slewing bearing using strain-life approach [C]//. 12th International Conference on Fracture 2009, 20093891-3898

[2] PotocnikR, GonczP, FlaskerJ, GlodezS. Fatigue life of double row slewing ball bearing with irregular geometry [J]. Procedia Eng, 2010, 2(1): 1877-1886

[3] GonczP, PotocnikR, GlodezSLifetime determination of the raceway of a large three-row roller slewing bearing [C]//, 2012113-116

[4] GaoX-h, HuangX-d, HongR-j, WangH, ChenJie. A rolling contact fatigue reliability evaluation method and its application to a slewing bearing [J]. J Tribol, 2012, 134(1): 011101

[5] ShimizuS. P-S-N/P-F-L curve approach using 3-parameter Weibull distribution for life and fatigue analysis of structural and rolling contact component [C]//. Proceedings of the ASME/STLE International Joint Tribology Conference, 2005973-979

[6] RobinsonE YA universal bivariate Weibull model for static and dynamic fatigue reliability forecasting [C]//, 201226-46

[7] CaesarendraW, KosasihB, TieuA K, MoodieC A S. Circular domain features based condition monitoring for low speed slewing bearing [J]. Mech Syst Signal Pr, 2014, 45(1): 114-138

[8] CaesarendraW, KosasihB, TieuA K, MoodieC A S. Application of the largest Lyapunov exponent algorithm for feature extraction in low speed slew bearing condition monitoring [J]. Mech Syst Signal Pr, 2015, 50-51(1): 116-138

[9] BenkedjouhT, MedjaherK, ZerhouniN, RechakS. Remaining useful life estimation based on nonlinear feature reduction and support vector regression [J]. Eng Appl Artif Intel, 2013, 26(7): 1751-1760

[10] DongS-j, LuoT-hong. Bearing degradation process prediction based on the PCA and optimized LS-SVM model [J]. Measurement, 2013, 46(9): 3143-3152

[11] DongS-j, YinS-r, TangB-p, ChenL-l, LuoT-hong. Bearing degradation process prediction based on the Support Vector Machine and Markov model [J]. Shock Vib, 2014, 2014: 1-15

[12] ChenX-f, ShenZ-j, HeZ-j, SunC, LiuZ-wen. Remaining life prognostics of rolling bearing based on relative features and multivariable support vector machine [J]. Proceedings of the Institution of Mechanical Engineers, Part C: J Mech Eng Sci, 2013, 227(12): 2849-2860

[13] KimH E, TanA C C, MathewJ, ChoiB K. Bearing fault prognosis based on health state probability estimation [J]. Expert Syst Appl, 2012, 39(5): 5200-5213

[14] Mech Syst Signal Pr, 2015, 56-57(150-172

[15] WeibullW. A statistical distribution function of wide applicability [J]. J Appl Mech, 1951, 183): 293-297

[16] ShenZ-j, ChenX-f, HeZ-j, SunC, ZhangX-l, LiuZ-wen. Remaining life predictions of rolling bearing based on relative features and multivariable support vector machine [J]. J Mech Eng, 2013, 49(2): 183-189

[17] CarpenterG A, GrossbergSPattern recognition by self-organizing neural networks [M], 1991CambridgeMIT Press

[18] VenkatesanP, SureshM. Classification of renal failure using simplified fuzzy adaptive resonance theory map [J]. Int J Com Sci Net Secur, 2009, 9(11): 129-134

[19] ZhangM, LiG-x, GongJ-z, WuB-zhong. Predicting configuration performance of modular product family using principal component analysis and support vector machine [J]. J Cent South Univ, 2014, 21: 2701-2711

[20] JohnHPrinciples of tribology [M], 1975

[21] ZvokeljM, ZupanS, PrebilI. Non-linear multivariate and multiscale monitoring and signal denoising strategy using kernel principal component analysis combined with ensemble empirical mode decomposition method [J]. Mech Syst Signal Pr, 2011, 25(7): 2631-2653

[22] ZvokeljM, ZupanS, PrebilI. Multivariate and multiscale monitoring of large-size low-speed bearings using ensemble empirical mode decomposition method combined with principal component analysis [J]. Mech Syst Signal Pr, 2010, 24(4): 1049-1067

[23] YiB K, FaloutsosCFast time sequence indexing for arbitrary Lp norms [C]//, 2000385-394

[24] Chinese Standard No. JB/T2300—1999. Slewing bearing [S]. (in Chinese)

[25] FengY, HuangX-d, ChenJ, WangH, HongR-jing. Reliability-based residual life prediction of large-size low-speed slewing bearings [J]. Mech Mach Theory, 2014, 81: 94-106