计时现有算法在多态系统可靠性评估方面的缺点和不足,基于最大流理论和分解技术,提出多态系统可靠性评估的实用高效算法.不需要知道系统的极小路和极小割,算法首先通过最大流理论寻找多态极小路;基于已得的多态极小路,算法利用分解技术把状态向量集分解成互不相交的子集,使得多态极小路能更高效地从子集中求得.最后,实例和数值试验表明,提出的算法在系统可靠性评估方面具有更高的效率.
Considering the shortcomings of existing algorithms in computing multi-state system reliability, this paper, based on max-flow theory and decomposition technique, presents a practical and efficient algorithm for reliability evaluation of multi-state systems. Without knowing MCs and MPs,The proposed algorithm first searches for multi-state minimal paths in terms of the max-flow theory. Based on the obtained multi-state minimal paths, the algorithm uses decomposition technique to decompose the set of state vectors into disjoint subsets, such that multi-state minimal paths can be efficiently derived from these subsets. Finally, the illustrative example and numerical experiment show that the presented algorithm is more efficient in system reliability evaluation.