Inference For Constant-stress Accelerated Life Tests With Dependent Compeing Risks From Multivariant Birnbaum-saunders Distribution Based On Adaptive Progrssively Hybrid Consoring

In life testing,competitive risk models are usually discussed under independent assumptions.In this paper,we study the dependent competitive risk model under the multivariate birnbaum-saunders distribution(MBS)based on the adaptive progressive mixed truncation(APHCS)constant stress accelerated life test.The expected failure time is observed by using an adaptive stepwise hybrid truncation method to terminate the life test at a given time.The second chapter briefly introduces the Birnbaum-Saunders distribution,and gives its marginal distribution and conditional distribution.Then,based on the constant stress accelerated life risk data of the adaptive stepwise mixed truncation method,the model is derived by using the approximate maximum likelihood estimation method and the self-help method,and the maximum likelihood estimation and interval estimation of the parameters are given.In the third chapter,in order to consider the independence among competitive risk factors and the shape and scale parameter relations among multiple competitive risk factors,we put forward the hypothesis test,and evaluated the estimated results through simulation research and a practical application example.