Remaining Useful Life Estimation Based On Wiener Degradation Process

Estimation of remaining useful life as one of the important components in prediction and health management has been extensively studied.For modern high-reliability and long-life product equipments,reliability modeling based on product degradation data can be used to more accurately assess the life.Because of the randomness of the performance degradation process of the product,the reliability analysis of the product can be performed by a stochastic process model such as Wiener process,Gamma process,and inverse Gaussian process.The Wiener process is used to do further research on the remaining life of the product.Firstly,a Wiener degradation process process model with a random diffusion coefficient is proposed,which introduces random effects into the diffusion coefficient.The concept of first hitting time is used to define the life and remaining life.Based on the model assumptions,the probability density function and cumulative distribution function of the life and remaining life are derived,and the parameters in the model are estimated using the EM algorithm.Secondly,a Wiener degradation process process model with two random coefficients is proposed which introduces random effects into drift coefficients and diffusion coefficients respectively.Based on this assumption,giving the exact expression of the life and remaining life distribution,and the Bayesian formula and EM algorithm are used to estimate the parameters in the model.Finally,the gallium arsenide laser data set is used to illustrate the two models proposed in this paper.Based on the AIC criterion and model log-likelihood function values,they are compared with the basic Wiener degenerate process model.The results show that the Wiener degenerative process model with two random coefficients has a better fitting effect than the other two models.