A Multi-view FTA Based Multi-Objectives Algorithm for Substation RCM Period Determination
As substation consists of various kind of devices, their importance, reliability characteristics, aging curve, and maintenance cycle vary from one to another. By delving into sis years of fault statistics under management of Zhengzhou Power Supply Company, this paper establishes a set of FTAs (Fault Tree Analysis) categorizing fault data by device type, season and device life time, calling Multi-view FTA(MFTA). This set of FTAs describes fault data from different point of view, so they form a multidimension view of substation faults. A hybrid algorithm combining Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) is then used to search in this set of FTAs. Using lifetime, maintenance property, financial data, season and other indicator, this hybrid algorithm is used to optimize the Reliability Centered Maintenance period toward cost effective direction. As a problem of multi-objectives optimization, GA is used to search throughout the problem space, PSO is used to enhance local search ability, and the hybrid algorithm called Hybrid Genetic and Particle Swarm Optimization Algorithm (HGPSO) is proposed. As useless and unimportant branches in FTA would be coarsely treated, the calculation cost could be reduced tremendously. The implementation result from the recent year reveals the feasibility of this algorithm.
FTA Genetic Algorithm Particle Swarm Optimization RCM
Zheng Yan Kuang Shi
Zheng Zhou Power Supply Company Zheng Zhou, HenanProvince, China Zheng Zhou Power Supply Company Zheng Zhou, Henan Province, China
国际会议
深圳
英文
107-110
2010-11-20（万方平台首次上网日期，不代表论文的发表时间）