Reliability Sampling Inspection And Several Hypotheses Test Problems Of The Constant-stress Accelerated Test Under Two Kind Of Distributions

This paper considers the method of the reliability sampling inspection and several hypothesestest problems of the constant-stress accelerated test, for Generalized exponential distributionsand Pareto distributions.Firstly, this paper studies the method of the reliability sampling inspection under Generalizedexponential distributions and Pareto distributions. For Generalized exponential distributions,it is given a design scheme based on the maximum likelihood estimate of the shape parameters,in all the sample. In the truncated sample, it is given three designs based on the number offailure, no failure data and failure time. At last we prove method is feasible through the datasimulation.For Pareto distributions, it is given a design scheme based on the maximum likelihoodestimate, in all the sample.The second this paper studies four hypothesis test problems of constant-stress acceleratedtest under two kinds of distributions. The main research is the hypothesis test problem whichis test the failure mechanism of the constant-stress accelerated test is whether or not changing,under Generalized exponential distributions and Pareto distributions. For the Generalized ex-ponential distributions, this paper improve the method of literature[26], and this method hasthe advantages of itself; For the Pareto distributions, using the theory of generalized p-value,we improve the method of literature[22], and we prove that the method has the excellent prop-erties of itself. Two other problems are the hypothesis test problems for shape parameters ofgeneralized exponential distributions and Pareto distributions, under constant-stress acceleratedtest,using the conception of generalizd p-value interval. The simulated result indicated that themethods carried out in paper efciently controlled the probability on type I error.