Reliability Research Of Subway Traction Power Supply System

With the rapid development of the economy and the continuous acceleration of urbanization of our country,urban rail transit system,as a public transport to alleviate urban traffic congestion,plays an increasingly important role in today's urban development.Moreover,the metro power supply system is an important link of urban rail transit system to provide the power for metro trains.Once the power supply system breaks down,it will not only disrupt the operation of metro lines,but also endanger the safety of passengers and have a serious social impact.Therefore,it is of great significance to study and improve the reliability of metro traction power system.In this paper,the principal methods of reliability analysis and the power supply system for urban rail transit are introduced firstly,and then the Fault Tree models according to the main electrical connection of the typical traction substation and the components of the overhead line are established respectively.In addition,the transformation modeling method of Fault Tree-Bayesian Network model is proposed,and the optimization method of multiple series system is studied with the two-state model of circuit breaker is modified based on the polymorphism model of Bayesian Network(BN)together.Then Junction tree inference algorithm is used to get the posterior probability rankings of each nodes of traction substatio n under different working conditions,which can help us to find out the weak parts of the system.Based on the BN models of traction substation and overhead line,the reliability model of a certain line of the metro traction power system is established,and the reliability parameters of the system under different power supply modes are compared,which provides theoretical reference for the design and maintenance of the metro traction power system.In order to calculate steady-state reliability indexes of metro traction power system under bilateral power supply mode,the Dynamic Bayesian Network(DBN)model is established.After that,by the inference of DBN,the posterior probability rankings of the steady-state nodes is obtained,and the main factors affecting the steady-state reliability of the system are found.At last,based on the Fuzzy Comprehensive Evaluation method and the inference of DBN model,the reliability allocation method of the metro traction power system is proposed,and the feasibility of this method is verified according to the reliability distribution result.