Fault Diagnosis Of Subway ATP Vehicle Equipment Based On SA-BP Neural Network

With the rapid development of China's economy and the acceleration of urbanization,subway,as an important infrastructure of urban rail transit,is also expanding.Subway trains are operated with high density and short interval,which will inevitably lead to failures,thus affecting the transport efficiency and safety of trains.Therefore,this thesis takes TBS-100 subway ATP vehicle-mounted equipment as the research object,and based on the fault data in the fault record table,puts forward the method of combining the statistical results of fault data with SA-BP neural network to establish a fault diagnosis model for ATP vehicle-mounted equipment,and verify the reliability of the model.At the same time,the simulation software of subway ATP vehicle equipment fault diagnosis system is designed and implemented by using this diagnosis model.The specific work of this paper is as follows:(1)Study and analyze the system structure and function of each device structure of TBS-100 subway ATP vehicle-mounted equipment,mainly analyzing the working principle of the equipment and the working principle and mechanism between the functional modules.The common fault types and phenomena of ATP on-board equipment were summarized,and the fault causes were analyzed.After the fault data in the fault record table were processed,the sample data was established.(2)Establishment of fault diagnosis model.Starting from the theory of BP neural network and the SA algorithm,using the sample data,the EIC fault code of the ATP on-board equipment,for example,choose to train operation time,recovery times after emergency braking,emergency stop after the drop number as input,equipment module fault number of times as the output to establish the fault diagnosis model based on SA-BP neural network.(3)Model validation and optimization.The factors influencing the training speed and accuracy of the network model were analyzed,and the fault diagnos is model was optimized from the BP neural network structure,BP neural network algorithm and simulated annealing SA algorithm parameters,and its accuracy was verified.(4)Adopt GUI application program of MATLAB to design and realize fault diagnosis process and result visualization display of subway ATP vehicle equipment fault diagnosis system,providing strong technical support for the safe and efficient operation of subway trains.