摘要
提出一种基于人工神经元网络的距离保护新原理。利用电磁暂态仿真程序（ＡＴＰ）进行大量的仿真计算，证明了所提方法的优越性，对于线路全长的８５％以内的故障，保护所需的数据窗长为２２ｍｓ，动作时间不大于２６ｍｓ。对于线路全长的７０％以内的故障，数据窗长为１２ｍｓ，动作时间不大于１６ｍｓ。
A new principle of distance protection of high voltage
transmission lines based on artificial neural network(ANN) is proposed in this paper. The ANN
distance protection consists of three sub neural networks:one for power swing detection,one for
fanlt detection and fault phase determination and the third for fault location. The resultant action
of the protection is determined by the comprehensive judgement according to the outputs of the
three sub networks. Training and testing results by ATP simulations are given.
出处
《中国电机工程学报》
EI
CSCD
北大核心
1999年第5期67-70,共4页
Proceedings of the CSEE
基金
国家教委博士点基金
