基于神经网络的三相电缆故障预测定位系统

为实现三相电缆故障准确定位,在大量实测数据的基础上,用L-M数值优化改进的BP算法建立了复杂非线性电缆故障预测定位系统输入信号与故障距离之间的数学模型,设计了基于BP神经网络的电缆故障预测定位系统,实现了三相电缆短路或接地故障的预测定位功能。仿真结果表明,故障距离的预报误差在指定范围内,预报精度可靠。

Prediction and location system of three-phase cable fault based on neural network

To implement accurate location of three-phase cable fault, a complex nonlinear system is designed with BP neural network technique, in which a mathematical model between the input signals and fault distance is constructed with an improved BP algorithm that is the L-M data-optimized method. Consequently, the prediction and location function of short circuit fault or grounded fault of three-phase cable are implemented. It is indicated by the simulation that the prediction error of fault distance is in the specified range, and the precision of prediction is reliable.