改进PSO-BP方法在ATP车载设备多故障诊断中的应用研究

针对目前ATP车载设备采用单故障诊断方法存在诊断精度偏低的问题，首先对ATP车载设备的故障原因进行分析，并提取出9种故障特征作为输入，7种故障类型作为输出，同时结合ATP车载设备的结构和故障特点建立了改进PSO-BP的多故障诊断模型；其次，在模型求解过程中，引入遗传算法中的变异思想，通过动态调整粒子群的参数来优化BP网络，采用改进PSO-BP算法对此模型进行求解；最后，以武广线数据进行仿真验证，期望输出与实际输出基本一致，故障识别的正确率达到95%。结果表明，采用改进PSO-BP算法解决ATP车载设备的多故障诊断问题是一种有效的方法，其诊断能力优于传统的BP算法和PSO-BP算法。

Application of Improved PSO-BP Method in Multiple Fault Diagnosis of ATP Vehicle Equipment

In order to solve the problem of the single fault diagnostic method with the low efficiency of ATP vehicle equipment,whose cause of fault was analyzed firstly,and the nine fault features were put as input and seven fault types were put as output.At the same time,a multi-fault diagnosis model of improved PSO-BP was established on the basis of the structure with fault characteristics of the equipment.Secondly,in the process of solving the model,the idea of mutation in the genetic algorithm was introduced to optimize the BP network by dynamically adjusting the parameters of the particle swarm optimization(PSO),so the improved PSO-BP algorithm was used to solve the model.Finally,the simulation was carried out with the data of Wuhan-Guangzhou line,the expected and actual output were the same basically,and the correct rate of fault identification reached 95%.The results show that it is efficient to use the improved PSO-BP algorithm to solve the multi-fault diagnosis of ATP vehicle equipment,and the diagnostic capability of the method is superior to the traditional BP algorithm and PSO-BP algorithm.