摘要
利用小波变换在信号处理方面的时频分析能力和神经网络对任意非线性函数的普遍的逼近能力 ,提出了一个基于小波神经网络的电力系统故障段辨别方法。故障诊断系统依据保护继电器和断路器的采样信息估计电力系统中故障段的位置。仿真结果显示 ,小波神经网络故障诊断系统能正确估计电力系统单一故障和多重故障的位置 ,即使在电力系统中存在保护继电器和断路器误动或拒动的情况下 ,小波神经网络也能给出合理的结果。测试结果表明 。
A method of fault section estimation of power system base on wavelet neural network (WNN), which inspired by both wavelet transform and artificial neural network(ANN) that reveals readily properties of a signal in localized regions of the joint time-frequency space, is presented. The diagnostic system can be applicable to power system control center for fault section estimation using information of relays and circuit breakers. The simulation results show that the WNN based method can estimate fault section for single or multiple fault, and give the reasonable results even in the case of the failure of protective relay and circuit breakers, which is suitable for complex fault diagnosis. The test results suggest that the wavelet neural network method is promising in the development of fault diagnostic systems.
出处
《电力系统及其自动化学报》
CSCD
2003年第2期80-83,102,共5页
Proceedings of the CSU-EPSA
基金
天津自然科学基金资助 :0 2 3 80 12 11
