Statistical Analysis Of A Degradation Failure Model Under Some Poisson Shock

Products may suffer outer cumulative or independent shock damage during their natural degradation processes. In this paper, statistical models are proposed systematically and their inference methods are proposed. In chapter 1, damage models and existing methods for degradation test are briefly introduced. In chapter 2, the normal degradation process is characterized by random coefficient model. Based on different ways of influence, a composite failure model and a competition failure model are built respectively. Models are also discussed in constant stress accelerated degradation tests. In chapter 3, we use Wiener process to describe its normal degradation, and then take shock damage into consideration. Mixed models of poisson shock and its natural degradation under constant stress and step stress are studied successively. In addition, we use Gibbs sampling to estimate the parameters and Monte-Carlo simulation to obtain the reliability character such as survival probility. Finally, some concise conclusion is drawn in the end of this paper.