64
reads

Tsai-Hung Fan

2011-01-14
12:45:00 - 14:45:00

R440 , Astronomy and Mathematics Building

In order to quickly extract information on the life of a product, the accelerated life test (ALT) is usually employed. There is a vast and growing literature on statistical methods for ALT. In this talk, I start with an introduction to accelerated testing, including basic ideas, terminology and practical engineering considerations. Related references on ALTs as well as their optimum testing plans based on maximum likelihood (ML) and Bayesian perspectives are briefly presented. Next, I will talk about some of our recent work on step-stress ALTs. The first topic discusses a step-stress ALT with M-stress variables when the underlying data are progressively Type-I group censored from the exponential lifetime distribution with a link function that relates the failure rate and the stress variables in a linear way under the Box-Cox transformation, and a cumulative exposure model for modeling the effect of stress changes. The second topic considers a system of independent and non-identical components connected in series. In a series system, the system fails if any of the components fails. It is often to include masked data in which the component that causes failure of the system is not observed. For inference on models developed in the above topics, computational techniques are mainly based on the classical ML method by conducting the Fisher-scoring or EM-type algorithms. A fully Bayesian method based on the Markov chain Monte Carlo (MCMC) technique is also provided. Numerical outcomes from simulation and data analysis are given to demonstrate the proposed methodologies.