Statistical Analysis Of Lifespan Data Under Two Testing Models

System reliability analysis is an important branch of mathematical statistics, andreliability analysis is mainly carried out through analysis of life data. Statisticalanalysis of life data is an important quantitative understanding means of lifecharacteristics of systems or components. In order to improve the reliability, reducethe cost, arrange for repair and replacement reasonably, the obtained informationfeedback to the design, manufacture and use, which can make better use value andeconomic effect. Statistical analysis of life data has become a hot topic of engineering,medicine, biological science and so on.This paper promotes two groups of mixed life test model proposed by Balakrishnanand Rasouli in2008to N groups. We consider that the products of N assembly linesmix together, and assume that the life of each line obeys exponential distribution butdifferences, which means that the distribution parameters is different, so the model iscalled type-II hybrid censoring model. We consider a priori information and themaximum likelihood estimation of gradual normality fully by Bayes method and theJeffrey criterion to discuss the maximum likelihood estimation (MLE), the Bayesestimation and confidence interval estimation of exponential distribution parametersbased on mixed type II truncation model.At the same time, based on progressively type-II censoring, this paper alsodiscusses that according to the proportion of risk model and overall obeys weibulldistribution parameters of maximum likelihood estimation (MLE) and the Bayesestimation, and the failure rate of H (t) and reliability function R (t) estimates, andnumerical simulation is carried out on the estimated results.