| In the forthcoming "14th Five-Year Plan" period,"new infrastructure" will become a strategic direction for local governments to focus on.Urban rail transit,such as subways and light rails,as an important part of "new infrastructure",its development momentum is gradually strong.During the operation of the subway,due to the degradation of the insulation performance of the subway to the ground,it will cause a phenomenon of current leakage that is widespread in the main structure of the subway system and the surrounding underground environment,which is called the subway stray current.The stray current of the subway will have a serious electrochemical corrosion effect on the buried metal pipelines around the lines.In severe cases,it can cause the buried metal pipelines to perforate,causing the leakage of dangerous transportation media such as natural gas and coal gas in the pipelines,which will induce explosions,Safety incidents such as fires.Therefore,the subway rail-to-ground transition resistance is an indicator of the insulation performance of the subway,and its test is particularly critical.Dealing with the weak points of the transition resistance plays an important role in preventing the generation of stray currents.The main research work of this paper is as follows:(1)Analyzed the architecture of subway traction return system.Under this background,summarize the causes of stray current and the composition of subway rail-to-ground transition resistance,study the equivalent resistance of insulation fasteners,and the equivalent resistance and subway section and conversion relationship of transition resistance value.(2)Established a resistance network model of rail-earth and rail-drainage network-earth under single and double power supply.At the same time,the three-dimensional topology model of the subway system is established using CDEGS software,and then the two models are simulated and analyzed to verify the three-dimensional topology model can simulate the subway system under actual working conditions more accurately and effectively.(3)Based on the three-dimensional topology model,the existing transition resistance test technology was simulated and analyzed,and the neural network algorithm optimized by particle swarm optimization(PSO-NN)was proposed to predict the transition resistance value,and the PSO-NN-based Prediction algorithm flow and structure framework,a data set obtained by simulation based on a three-dimensional topology model was established,a prediction model of transition resistance value was constructed and neural network training was carried out.The difference in accuracy and stability of the PSO-NN algorithm compared with BPNN and the traditional transition resistance test method is compared,which proves the performance advantage of the PSO-NN algorithm in the transition resistance test problem.(4)Designed and built a transition resistance test system based on PSO-NN,and conducted a test experiment in a subway section to verify the reliability and stability of the system in practical engineering applications.This Thesis has 89 figures,32 tables and 83 references. |
