The Reliability Estimation Of Component And System About Random Censoring Under Theaccelerated Test

Reliability analysis is an important branch of statistics, there is a strongapplicability, it has been received highly attention by engineering sector and others.In practical work and life, we often encounter reliability issues. For example, all ofus want products that are not prone to failure and have long life when we buy TVand other appliances. From a reliability standpoint, it means to purchase productswith highly reliability. When a new product is on sale, customers need to beprovided information such as average life expectancy. Products such as light bulbsare more relevant to the life of the information. Reliability analysis is the research offailure law analysis. Reliability is the probability expression of ability that a systemor a device finishes some functions under the specific environment andpredetermined time. If the reliability tests are in accordance with conventionalpractice under normal test conditions, they are often time-consuming, evenimpossible, so we usually take accelerated Life Test. That is to say under the samephysical condition, we make environment worse to shorten the life, and then convertthe reliability index from bad condition to normal condition. There is moreComponent reliability under Constant pressure accelerated tests than the complexsystem reliability. Most data form are based on a set number or timing of censored,and there is lack of the research to random censoring, but random censoring is moregeneral, so it is highly significant to take experiment to the Constant pressureaccelerated test under random censoring.Contents of this paper have the following aspects:1. This paper describes the content about the statistical inference under constantstress accelerated life testing and reliability expressions of complex system based onthe minimum path matrix.2. Under the random Censoring Weibull distribution and two-parameter of theexponential distribution, this paper gives the least squares estimation of acceleration factor and estimation of the mean life and the reliability under normal condition.3. This paper gives reliability estimation of complex systems that described byminimum path or minimal cut sets when the life of subsystem obey the Weibulldistribution and two-parameter of the exponential distribution under the constantstress accelerated life testing.