基于灰色理论和PSO的牵引供电系统可靠性研究

采用灰色GM(1,1)模型预测其相关设备中长期的故障数据,增大建模所需数据样本量,提出以威布尔分布模型为基础,基于粒子群算法的智能拟合方法,建立牵引供电系统各设备可靠性分析的数学模型,所建模型全部通过拟合优度K-S检验,证明该方法十分适合于牵引供电系统设备的可靠性建模。运用BDD算法得到用最小割集表示的牵引变电所和接触网故障树模型,综合得到牵引供电系统整体的可靠性模型,并计算出系统的可靠度和平均使用寿命。最后分析可知接触网的可靠性在很大程度上决定了牵引供电系统整体的可靠性,其分析结果为牵引供电系统以后的维护工作提供了理论依据。

Research on the Reliability of Traction Power Supply System Based on Gray Theory and PSO

In order to increase the amount of sample data required to establish models and predict the failure-rate data of the relevant equipment in medium and long terms,gray GM( 1,1) model is used based on particle swarm optimization( PSO) and a reliability modeling method of traction power supply system based on Weibull distribution model is proposed. Such models so established satisfy K-S goodness of fit test,and justify the feasibility of the method. The fault tree analysis( FTA) model of traction substation and catenary system in terms of minimum cut set is established by using binary decision diagram( BBD) algorithm,the reliability model of TPSS is obtained,and the reliability and the average useful time of system are calculated. The results of analysis demonstrate that the reliability of the traction power supply system as a whole depends largely on the reliability of catenary system,and provides a theoretical basis for future maintenance for traction power supply system.