基于人工神经网络的电力系统中振荡与短路模式识别的研究  

Identification on oscillation and short circuit modes in electric power system based on ANN

作　　者：郁惟镛[1] 李航[1] 康建洲[1] 

机构地区：[1]上海交通大学电气工程与自动化系,上海200240

出　　处：《继电器》2000年第6期7-9,22,共4页Relay

摘　　要：电力系统中的振荡会影响许多距离保护的性能 ,它是许多学者共同关注的难点。文章在比较了以往几种传统保护的原理之后 ,提出一种新的基于人工神经网络的判别算法 ,它能正确区分故障、振荡及振荡中故障等各种状态。并对BP网络算法进行改进后 ,应用于距离保护之中。EMTP仿真结果表明 ,此算法比以往的保护更可靠 。The oscillation in electric power system would impact on the performances of the distance relay. It is a problem to solve. A new criterion based on ANN is proposed by comparing several traditional relaying principles. Using this new criterion, the modes of fault, oscillation and fault by oscillation can be distinguished correctly. BP algorithm is available for distance relay after innovated. The result from EMTP simulation shows that the new criterion is more reliable and available than that of the traditional one.

关 键 词：电力系统 振荡 人工神经网络 短路 模式识别 

分 类 号：TM713[电气工程—电力系统及自动化]