基于最大流理论和分解技术的多态系统可靠性评估  

Reliability Evaluation of Multi-state Systems Based on the Max-flow Theory and Decomposition Technique

作　　者：牛义锋[1] 韦纯福[1] 赵宁[2] 

机构地区：[1]河南理工大学数学与信息科学学院,河南焦作454000 [2]河南理工大学计算机科学与技术学院,河南焦作454000

出　　处：《模糊系统与数学》2013年第5期182-190,共9页Fuzzy Systems and Mathematics

基　　金：国家自然科学基金资助项目(61040040;71271023);河南省教育厅科学技术研究重点项目(13B120022;13B630034);河南省应用数学重点学科项目

摘　　要：针对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.

关 键 词：多态系统 可靠性 多态极小路 最大流理论 分解技术 

分 类 号：N945[自然科学总论—系统科学]