In this paper, a statistical analysis method is proposed to research life characteristics of products based on the partially accelerated life test. We discuss the statistical analysis for constant-stress partially accelerated life tests with Lomax distribution based on interval censored samples. The EM algorithm is used to obtain the maximum likelihood estimations(MLEs) and interval estimations for the shape parameter and acceleration factor.The average relative errors(AREs), mean square errors(MSEs), the confidence intervals for the parameters, and the influence of the sample size are discussed. The results show that the AREs and MSEs of the MLEs decrease with the increase of sample size. Finally, a simulation sample is used to estimate the reliability under different stress levels.
Chinese Quarterly Journal of Mathematics
National Natural Science Foundation of China(11271039).
LONG Bing(1973-),male,native of Jingmen,Hubei,an associate professor of Jingchu Univer-sity of Technology,engages in reliability and application;ZHANG Zhong-zhan(1963-),male,native of Cangzhou,Hebei,a professor of Beijing University of Technology,engages in application of probability and statistics.