Comprehensive review of estimating system-reliability confidence-limits from component-test data

Abstract

A common problem of practical importance in reliability analysis is the determination of the reliability of a complex system where the only data available are reliability estimate of testing data on various components or subsystems that make up the system. In the literature, there have been many papers that have dealt with this problem. This paper, based on an extensive review of literature, summarizes and compares methods and models on system reliability confidence limits estimation. The advantages, disadvantages, limitations and application issues of these methods are analyzed and compared. It is shown that analytical methods are good for some specific application cases and Monte Carlo methods are the most powerful technique for very complex systems with different kinds of failure data. The results of this paper can be used as a guideline in choosing an appropriate method.