Storage Reliability Models Of Systems And Their Statistical Analysis

The statistical inference of storage reliability for product systems and its maintenance strategies are studied in this thesis.In today's society, a variety of emergency installation and some electronic devices, such as fire prevention system of building, rescue system in watercraft and aircraft, mine safety assembly and mechanical system in medical treatment etc., are usually in storage for a long time during their total life cycles. When these products need to be used after a long storage time, their reliability must be very high for being able to finish their missions.The theory of storage reliability is far away from complement compared with the operational reliability. There are little discussions on the statistical methods, maintainability and its optimization theory in the literature. Based on the exponential distribution and Weibull distribution, the ML estimation and least squares estimation of storage reliability are only shown in the literature [41-42]; In papers [67-68], the model based on binomial data is considered, and its estimation methods are provided. However, the statistical analysis based on periodic inspection data of products in storage is rarely discussed in the literature. In this thesis, the following research topics are considered:1. In order to make good use of the prior information to make up the deficiency of storage failure data, Bayes method is given here to estimate the confidence limit of storage life upon the failure-success data.2. The storage residual life and reliable life are deduced simply. The reliability based on the performance parameter is considered, another the dummy sample analysis is given to analyze the censoring data based on the exponential distribution and Weibull distribution.3. Based on the exponential distribution and the different product system, three types of storage reliability models based on the periodic inspection and maintenance with initial failure are given:? Like new after maintenance but increasing in failure rate;? Not like new after maintenance and failure rate unchanged;? Not like new after maintenance and increasing in failure rate.Through the data modification and models statistical analysis, the Bayes estimate of the reliability after modification is super to ML estimate;4. In order to keep its high reliability, the periodic inspection and maintenance areInstitute of Industrial Economical Engineering Guizhou University...