基于BP神经网络的轨道电路故障诊断

Track Circuit Fault Diagnosis Based BP Neural Network

针对高压脉冲轨道电路的故障预测问题,利用BP神经网络模型结构简单、非线性拟合能力强、容错性能良好等特点,构造了一种基于BP神经网络的高压脉冲轨道电路的故障诊断系统。该系统将直接构造出轨道电路发码器电压波峰峰值等相关参量与轨道电路相关故障的非线性映射关系,继而实现对轨道电路有关故障的识别分类。经分析研究表明,采取BP神经网络模型处理轨道电路故障,相比传统预测故障的方法不仅切实可行而且更加高率。

For the track circuit fault forecast problems,using the characteristics of BP neural network model with simple structure,strong nonlinear fitting,good fault-tolerant performance a kind of track circuit fault diagnosis system based on BP neural network was put forward.The system will be constructed directly refers to the track circuit sending secondary side voltage of related parameters,which related to the track circuit fault and the nonlinear mapping relationship,so as to realize the recognition and classification of track circuit related faults.The results show that the BP neural network model processing method is feasible,and more advantage than traditional prediction methods.