Statistical Analysis Of The Sequential Addition And Sequence Life Reduction Tests Under The Improved Chen's Distribution

In order to understand the performance and reliability of the product,it is often necessary to carry out the life test of the product.With the continuous development of modern social science and technology,products with high reliability and long life have springing up.In order to estimate the reliability index of this kind of product,people generally adopt the accelerated life test method of increasing stress in view of the long time and high expenses of the life test.Accelerated life test was introduced into China in the 70 s of last century.It plays an important role in reliability research.It shortens the test cycle by increasing test stress and saves expenses.It has been widely applied in testing and testing the service life of enterprise products.The commonly used acceleration tests include constant stress,step stress and progressive stress,etc.In recent years,accelerated life test has been widely applied in the field of civil industry because it can expose the weak links of the products ahead of time,and it also has many applications in the reliability research and life assessment of military weapons.The failure mechanism of different products is different,and its life distribution is different.The study shows that the bathtub-shaped failure rate function can better fit the real life data of a part of the product.Chen Zhenmin put forward a life distribution function of bathtub products containing two parameters in 2000,and gave the confidence interval of shape parameters and the joint confidence interval of two parameters.However,the distribution lacks important scale parameters,and it is easy to miss some statistical characteristics.Some scholars have improved it and added a scale parameter on the basis of Chen's distribution to make the model more flexible.Based on the modified Chen's distribution,this paper makes a statistical analysis of the life test under different stress models.It mainly discusses the failure distribution mode under the precondition of Nelson assumption and the fixed number of truncated cases.When the stress is humidity or voltage,the inverse power law model is used as an acceleration model.Arrhenius Equation is used as an acceleration model when the stress is the temperature.And then the corresponding life distribution and failure rate function are obtained under the progressive stress.Conducting descending stress accelerated life test,the corresponding failure rate functions are derived according to the inverse power rate and the Arrhenius Equation.The maximum likelihood estimation of the unknown parameters in the failure rate function is obtained in each case.Finally,the Monte Carlo simulation method is used to simulate the experimental data and the maximum likelihood estimation of the previous parameters is calculated.