Reliability Research Of Traction Supply System Based On Bayesian Network With Fusion Of CCF

As an important basis of electrified railway, the failure of traction supply system can not only lead to a paralysis of itself, but also can cause the unsafety problem of passenger and great many losses of property. Therefore, a high attention has been paid to the reliability of traction supply system, but it is still on the early stage at home and abroad. As a result, the reliability research of traction supply system has an important significance in theory and engineering to ensure a stable operation of traction supply system.Based on general description of reliability theory, the combination of common cause failure(CCF) with Bayesian network are applied in reliability research of the traction supply system in the dissertation. Firstly, the basic principles of the CCF and Bayesian network are introduced. Secondly, the composition of traction supply system is described briefly and the basic framework for reliability analysis of the whole system is divided into three relatively independent hierarchies according to the operation modes and functions of power supply system. Then, in terms of the structural characteristics of traction substation in AT power supply system, Bayesian network is applied in reliability analysis of a main electrical connection for the traction substation after taking the CCF factor between primary equipments into account. Thirdly, the computational formulas for reliability indices of series system that takes CCF into consideration are deduced from the Markov model with fusion of CCF. Therefore, the reliability analysis of overhead contact system is realized by these formulas and Bayesian network. At the same time, Bayesian network’s quantitative analysis results for the traction substation or overhead contact system are compared between considering CCF or not respectively. Besides this, the weak links which have an influence on traction substation or overhead contact are determined by the qualitative analysis. Finally, Bayesian network is used for modeling the traction supply system under the three different modes such as unilateral power supply mode, bilateral power supply mode and over-zone power supply mode, and the reliability assessment for the whole system is achieved and a comparison of computational results under the three different operation mode is made.The reliability analysis results for the three hierarchies of traction supply system indicates that the reliability analysis of traction supply system are finished quickly and accurately by Bayesian network regardless of solution to the minimal cut set. Besides this, because the CCF are brought into the relation of electrical equipments, the reliability analysis results of traction supply system come closer to the actual operational condition.