Statistical Analysis Of An Accelerated Life Test With A Two-parameter Half Logistic Distribution

In recent years,China's comprehensive national strength has been developed rapidly.And all kinds of products have been updated and replaced quickly,once they can't keep up with them.The pace of technology development is very likely to be eliminated from the market,so it is particularly important to speed up the research and development of product life.At present,China's military science and technology has been in the forefront of the world and can be quantified by the high quality products of long life.As the demand is increasing,the accelerated life test is one of the important ways to solve this kind of demand,and it has been widely used in the project of setting up longevity and longevity of weapons and equipment.Life expectancy extends the life of the subject.Accelerated life test uses accelerated stress to test the life of the specimen.It can not only effectively reduce the test time,improve the efficiency of the test,but also reduce the cost of the test.The research makes the reliability evaluation of the high reliability and long life possible.The tests are usually divided into constant stress test,step stress test and sequential stress test.In 2005,Zhang Chunhua and so on proposed a step-down test,after that some researchers extended the step-down test to order-drop test.Because stress levels loaded by progressive stress tests(sequential stress tests)continuously increases(decreases)with time,more time is saved than with other stress tests.In addition,some foreign scholars have proved that the introduction of the scaled half-logistic distribution as the life distribution of the failure data in the life test can better fit some life data such as aircraft equipment,electricity and insulation material.etc.Therefore,in view of the present status of the research,the location scale distribution and Weibull distribution are used for reference,we introduce the shape parameters into the half-logistic distribution of only one scale parameter to obtain the half-logistic distribution of two parameters,namely SSHL(m,?).Based on this distribution,under the cumulative damage model(CEM)in the sequence plus test and the descending test,the inverse power-law acceleration model and the arrhenius acceleration model are respectively discussed.The failure distribution function of the product life is deduced and the maximum likelihood estimation.Finally,the Monte-Carlo method was used to simulate and give the result of the parameter estimation.The result of the parameter estimation is compared with the true value of the parameter,and the deviation is small.This proves the effectiveness of the test under the two-parameter half-logistic distribution.