Reliability analysis of cold-standby systems using a self-stratifying subset simulation method

Abstract

Cold-standby systems are often featured with small failure probability and widely used as critical safety systems in current engineering fields. The existing analysis methods for cold standby systems have the problems with low-computional efficiency either for small failure probability events or for the components with mixed distribution types. To solve these problems, we propose a self-stratifying subset simulation (4S) method based on MCMC simulation and the performance function of cold-standby system. Numerical case studies are performed to demonstrate the reasonability and feasibility of the proposed 4S method. The case studies indicate that the results obtained by the proposed 4S method agree well with those obtained by the existing methods and the computation efficiency is greatly improved especially for small probability events. Moreover, the proposed 4S method was applicable for the cold-standby system cold-standby system with not only the exponentially or non-exponentially distributed components but components with more mixed distribution types.