| Over the last decade,the urban rail transit industry in China has developed rapidly,and urban population is also increasing constantly.The metro,which undertakes more than half of the passenger flow,is also faced with new challenges and opportunities in terms of construction,operation and maintenance.According to relevant literature research,the main manifestations of insufficient power supply capability for metro include DC switch tripping due to overcurrent when multiple trains start,and frequent switching of over voltage protection device due to abnormal increase of rail potential.Therefore,it is of great practical significance to systematically and perfectly establish the power supply capability evaluation system for the safe and stable operation of metro traction power supply system.In this thesis,starting from the key influencing factors of power supply capability simulation,in allusion to traction network parameters and train stop time—two factors which affect the accuracy of power supply capability simulation,the corresponding solutions were put forward respectively.And then,based on simulation results,the power supply capability evaluation system was built for metro traction power supply system,and corresponding auxiliary decision-making means for each evaluation index were given.Finally,the development of simulation software was completed.Firstly,the demand analysis of power supply capability simulation evaluation was carried out,which confirmed the importance of power supply capability evaluation in terms of design and operation of metro.This thesis introduced the modeling and simulation calculation method of traction power supply system.In allusion to the problem that the working conditions frequently change during simulation calculation of multi-section polyline characteristics of rectifier unit,in combination with measured data and the protection set value of output current of rectifier unit,a simplified model of two-section polyline was proposed.Finally,the main factors that may affected the evaluation of power supply capability were analyzed,and the technical route for further study on the uncertain factors was determined.Different catenary resistance and rail resistance have a certain influence on the simulation results of power supply capability.However,the above parameters are relatively stable,while the complex and changeable rail ground parameters have a greater influence on the calculation accuracy of rail potential.Therefore,a black box conceptual model of rail potential was proposed based on the measured data.BP neural network and machine learning framework were adopted respectively to solve the model.Compared with the traditional equivalent circuit method,the fitting accuracy of rail potential of black box model obtained by machine learning framework exceeds 90%.Due to the randomness of stop time,the reliability of the simulation results of power supply capability is poor,which cannot reflect the influence of passenger flow on the train diagram during actual operation.Based on the research results of the existing literature on stop time,the probability model of the stop time was established by fitting the measured stop delay time data with distribution.Considering the composition of the train diagram,the probability model of the train diagram was constructed.Based on Monte-Carlo simulation,the application of probability model of train diagram was realized in power supply capability simulation.By referring to the related concepts of power supply capability in power distribution network,and in combination with the actual operation of metro,the meaning of power supply capability of metro traction power supply system was given.By referring to relevant national standard specifications and coordinating with the settings of relay protection,the evaluation indices of power supply capability were determined.The corresponding index evaluation system was designed by means of analytic hierarchy process method,and the synthetic calculation method of power supply capability health degree was given.Finally,according to the power supply meaning of each index,the auxiliary decision-making library corresponding to each evaluation index was designed.At last,the development of metro power supply simulation software was completed.The software included complete functions such as line parameter management,model construction,simulation calculation and other basic functions.Meanwhile,the rail potential regression model and power supply capability evaluation function were embedded in the software.By comparing the results of simulation software with the measured data of the substation,the results show that the accuracy of the simulation software developed meets the engineering requirements. |
