Parameter Estimation of Accelerated Model for Birnbaum-Saunders Distribution Based on Best Unbiased Integral Estimate Method

Birnbaum-Saunders fatigue life distribution is one of the common distributions, the use of which in the accelerated life test is discussed in this paper. A regression model is established according to its characteristics. Then the parameters of accelerated model are obtained by means of best unbiased integral estimate method, which makes use of censored data in different stress levels as a whole. The present method is better than the traditional group analysis test method in the precision of statistical inference, and a great number of specimens can be saved.