【摘要】 提出一种基于多监测点电压暂降信息结合BP(back propagation)神经网络的配电网故障定位方法，利用BP神经网络实现对输入数据和输出数据最大程度拟合，最终实现故障定位。由故障引起的电压暂降发生时，电压暂降时域信号含故障位置信息，以监测点获得的电压、电流信息为输入，以故障所在支路和距首端监测点的距离为输出，建立适用故障定位的BP模型。基于Matlab软件对多分支配电网进行仿真验证，结果验证了该方法在故障定位上的优越性，且不受过渡电阻的影响，适用于实际工程。更多还原

【Abstract】 This paper presents a distribution network fault location method based on multi-monitoring point voltage sag information and combined with back propagation(BP) neural network. By using the BP neural network, the maximum fitting of the input data and output data is achieved, ultimately realizing fault location. When a voltage sag caused by a fault occurs, the time-domain signal of the voltage sag contains the fault location information. Taking the voltage and current information obtained from the monitoring points as the input, and the faulty branch and the distance from the head-end monitoring point as the outputs, a BP neural network model suitable for fault location is established. The multi-branch distribution networks are simulated and verified based on the Matlab software. The superiority of the BP neural network method in fault location is analyzed and verified. Moreover, the method is not affected by the transition resistance and is applicable to practical engineering.更多还原