Research On Fixed-time Consoring Sheme Skills Under Weibull Distribution And Reliability Analysis With Its Zero-failture Data

In the reliability qualification test,mostly we used to carry out fixed-time censoring experiment on products which performance index is Life or MTBF.The relative research work,which is relatively mature,on Fixed-time censoring experiment technology at home and abroad,is used to deal with the tested product whose life is subject to exponential distribution.Nowadays,the level of science and technology is increasingly developing.High-reliability and long-life-type products are found in all walks of life.In the course of time-tested truncations,there are usually no-failure data.In the current reliability field at home and abroad,the relevant academic results on the processing and reliability analysis of Weibull distribution failure-free data are mainly based on the point and interval estimates of the distribution parameters,or the Estimation method of nonfailure data proposed for the exponential distribution.This article aims at the characteristics of related products that the repairable system and non-repairable system's failure data under Weibull distribution,and designs a reasonable and complete Fixed-time censoring experiment program,and gives different types through a combination of practical and literature research methods.The empirical values of the distribution function shape parameters of the products whose life is subject to Weibull distribution under the working conditions compensate for the lack of solutions has been put forward.With using the method of transforming Weibull distribution into non-failure data of Fixed-time censoring test,the function of the transformed function can be used to obtain the function relationship of the failure probability at each time of truncation,and then the Bayesian estimation of failure probability at each ending time point will be calculated by combining the actual application information.Finally,the point estimates of the failure probability at every point are fitted to the failure probability function model by the least squares method,and then the product reliability index point value is obtained.With the point estimates of the distribution function parameter,the rank distribution theory can be used to calculated the confidence interval of the reliability index after the Weibull distribution is transformed into an exponential distribution,and the interval estimation of the product reliability index under a certain confidence level is obtained through corresponding calculation and derivation.In this paper,under the premise of research on the existing reliability appraisal tests and the application of Chinese military standard,we combine the characteristics and research of Weibull distribution products with the typical life span,and comprehensively consider the risk of producers,the risk of users,the ratio of identification,and MTBF.Expected value and other information are subject to a time-tailored test plan design technique study,and verify the rationality of the program.In addition,the reliability analysis method of non-failure data presented in this paper can not only fully utilize the life-span information of no-failure data,but also can take advantage of the fact that the Weibull distribution can have no failure data.At the same time,the calculation method of this paper is simple and easy to apply to engineering.It has great theoretical significance for the implementation of reliability improvement.