基于BP算法的AT牵引供电系统状态识别

BP Based Algorithm for AT Traction Power Supply System Status Identification

使用ATP电磁暂态仿真软件对AT牵引供电系统进行数学建模,仿真分析了高速动车组正常运行状态、T-R短路故障、F-R短路故障、T-F短路故障和异相短路故障;重点分析了高速动车组在过电分相时产生的非线性谐振(铁磁谐振)和线性谐振,采用MATLAB仿真软件为工具,应用神经网络中BP算法对AT牵引供电系统的5种状态进行识别。

The mathematical model for traction power supply system,established by means of ATP electro-magnetic transient simulation software,simulates and analyzes the normal operation status of high speed railway EMU,failures of T-R short circuit,F-R short circuit,T-F short circuit and out of phase short circuit;analyzes emphatically the non-linear resonance（ferro-resonance） and linear resonance generated during high speed EMU passing over the phase break,by means of the MATLAB simulation software,with application of BP algorithm in neural network,it identifies 5 status of the AT traction power supply system.