Residual life estimation based on bivariate Wiener degradation process with measurement errors

Abstract

An adaptive method of residual life estimation for deteriorated products with two performance characteristics (PCs) was proposed, which was sharply different from existing work that only utilized one-dimensional degradation data. Once new degradation information was available, the residual life of the product being monitored could be estimated in an adaptive manner. Here, it was assumed that the degradation of each PC over time was governed by a Wiener degradation process and the dependency between them was characterized by the Frank copula function. A bivariate Wiener process model with measurement errors was used to model the degradation measurements. A two-stage method and the Markov chain Monte Carlo (MCMC) method were combined to estimate the unknown parameters in sequence. Results from a numerical example about fatigue cracks show that the proposed method is valid as the relative error is small.

Keywords

residual life / performance characteristics / bivariate Wiener process / Frank copula / MCMC method

References

[1] SiX-s, WangW-b, HuC-h, ZhouD-hua. Remaining useful life estimation-A review on the statistical data driven approaches [J]. European Journal of Operational Research, 2011, 213(1): 1-14

[2] GebraeelN Z, LawleyM A, LiR, RyanJ K. Residual-life distributions from component degradation signals: A Bayesian approach [J]. IIE Transactions, 2005, 37(6): 543-557

[3] WangW-b, CarrM, XuW-j, KobbacyA K. A model for residual life prediction based on Brownian motion with an adaptive drift [J]. Microelectronics Reliability, 2010, 51(2): 285-293

[4] NoortwijkJ M. A survey of the application of Gamma processes in maintenance [J]. Reliability Engineering and System Safety, 2009, 94(1): 2-21

[5] KharoufehJ P. Explicit results for wear processes in a Markovian environment [J]. Operations Research Letters, 2003, 31(3): 237-244

[6] SariJ K, NewbyM J, BrombacherA C, TangL C. Bivariate constant stress degradation model: LED lighting system reliability estimation with two-stage modelling [J]. Quality and Reliability Engineering International, 2009, 25(8): 1067-1084

[7] PanZ-q, BalakrishnanN, SunQuan. Bivariate constant-stress accelerated degradation model and inference [J]. Communications in Statistics — Simulation and Computation, 2011, 40(2): 259-269

[8] PanZ-q, BalakrishnanN, SunQ, ZhouJ-lun. Bivariate degradation analysis of products based on Wiener processes and copulas [J]. Journal of Statistical Computation and Simulation, 20121-14

[9] WangY-p, PhamH. Modeling the dependent competing risks with multiple degradation processes and random shock using time-varying copulas [J]. IEEE Transaction on Reliability, 2012, 61(1): 13-22

[10] NelsenR BAn Introduction to Copulas [M], 2006New YorkSpringer Science109-150

[11] PengC Y, TsengS T. Mis-specification analysis of linear degradation models [J]. IEEE Transaction on Reliability, 2009, 58(3): 444-455

[12] WangXiao. Wiener processes with random effects for degradation data [J]. Journal of Multivariate Analysis, 2010, 101(2): 340-351

[13] TsaiC C, TsengS T, BalakrishnanN. Mis-specification analyses of gamma and Wiener degradation processes [J]. Journal of Statistical Planning and Inference, 2011, 141(12): 3725-3735

[14] WangX-l, GuoB, ChengZ-jun. Real-time reliability evaluation of equipment based on separated-phase Wiener-Einstein process [J]. Journal of Central South University: Science and Technology, 2012, 43(2): 534-540

[15] SiX-s, WangW-b, HuC-h, ZhouD-h, PechtM G. Remaining useful life estimation based on a nonlinear diffusion degradation process [J]. IEEE Transaction on Reliability, 2012, 61(1): 50-67

[16] ZhouD-h, FrankP M. Strong tracking filtering of nonlinear time-varying stochastic systems with colored noise: Application to parameter estimation and empirical robustness analysis [J]. International Journal of Control, 1996, 65(2): 295-307

[17] MeekerW Q, EscobarL AStatistical methods for reliability data [M], 1998New YorkJohn Wiley & Sons639-640