摘要
目前对于线路瞬时性故障的最佳重合闸时刻以离线计算为主 ,如利用能量函数法 ,但其计算困难 ,计算时间较长 ,在电力系统中不能满足实际运行条件变化的要求。文中提出了一种基于小波变换和人工神经网络 (ANN)方法的在线寻求瞬时性故障最佳重合闸时刻的方法 ,只需较短时间就能计算出最佳重合闸时刻。首先利用 MATLAB对电力系统故障进行仿真 ,把故障信号通过小波变换分解成不同尺度下的“近似”分量 (approximation)和“详细”分量 (detail) ,并把提取的特征值作为人工神经网络的输入量 ,进行训练 ,从而找到最佳重合闸时刻。
At present. the methods for calculating the optimal reclosing time for transient faults mostly are off--line. Forexample, the usage of transient energy function is difficult in computation and needs long computational time. which can'tmeet the requirements of rapid change of real operation in power system. An on-- line method for obtaining the optimalreclosing time of transient faults is presented based on wavelet transform and a rtificial neural network. which costs less timeto calculate the optimal reclosing time. At first. power system faults are simul ated using MATLAB. and the signals of faultsthat are transformed by wavelet transform are decomposed into 'approximations' a nd 'details' at different scales. and theirfeatures which are extracted through wavelet transform are considered to be the inputs for artificial neural network. Thenthey are trained through artificial neural network to find the optimal reclosing time. The validity and accuracy of this methodis testified by test examples.This project is supported by National Key Basic Research Special Fund of China ( No. G1998020312) and National KeyLab of Tsinghua University.
出处
《电力系统自动化》
EI
CSCD
北大核心
2000年第15期6-10,15,共6页
Automation of Electric Power Systems
基金
国家重点基础研究专项经费!(G1998020312)
清华大学电力系统国家重点实验室资助项目
关键词
电力系统
重合闸时刻
小波变换
人工神经网络
power systems
optimal reclosing time
wavelet transform
artificial n eural network
