Regression analysis method for type-Icensored data

Regression analysis method for type-I censored data was presented. The regression equation for type-I censored data was established. The best unbiased integral estimators of the regression coefficients and the standard deviation, the point estimators and confidence intervals of percentiles were given. The regression analysis methods for type-I censored data from extreme distribution, Weibull distribution, normal distribution and location-scale distribution were discussed in detail. As the experimental information at the censoring time and the cross information among test data in different states were fully exploited, the presented method could reach a high precision. The method provides a feasible and effective approach to life prediction and reliability estimation of products when the test time or expense are limited.