摘要
故障分类器是输电线路保护中的基本模块。文中以一个实际500kV输电网络为模型，提出了基于BP前馈网络和Kohonen自组织特征映射网络的高压输电线路故障分类方法，并进行了仿真研究。结果表明：这种方法快速、可靠，对输入信号容错性强，尤其在高阻接地故障时具有较好的分类性能，可以用于支持高速保护装置。
A new HV transmission line fault classifier. based on BP neural network and Kohonen self-organization network,is designed in this paper. Simulation results show its effectiveness on an actual 500 kV transmission network. Feature extraction. signal processing. the organization of neural network training and testing patterns are mainly studied. Then a quick. reliable. correct fault classifier is obtained.
出处
《电力系统自动化》
EI
CSCD
北大核心
1998年第11期28-31,共4页
Automation of Electric Power Systems
关键词
高压输电线路
故障分类器
神经网络
BP neural network Kohonen self-organization network pattern recognition fault classification
