Reliability analysis of multi-state systems subject to failure mechanism dependence based on a combination method

A multi-state system is a kind of system in which both the system and its components may display multiple performance levels, and it can be utilized to model more complicated and practical systems. As one kind of multi-state object, the performance degradation of a system and its components can be analyzed in terms of failure mechanism dependence. Some coupling relationships of these failure mechanisms exhibit multi-state properties as well. This paper merged the primary failure mechanism correlations proposed in our previous research and provided a combinational method that proposes a modified binary decision diagram (BDD) and a multi-state multi-valued decision diagram (MMDD) models for the state-probability evaluation and reliability analysis of the multi-state system. The method consists of four steps: generation of BDD models for components, calculation of state-probability for components, generation of MMDD models for the system and calculation of reliability for the system. As a study case, the reliability and state-probability of a multi-state sensor system are evaluated using this combinational method. Finally, a comparison between the binary-state and the multi-state are given as well.