摘要
将语音信号处理领域的隐马尔科夫模型(HMM)引入变压器保护,在简要介绍HMM 及其优点的基础上,提出了一种基于离散HMM 的励磁涌流和故障电流的识别方法,利用离散HMM 为PSCAD仿真的励磁涌流和故障电流数据建立了模型。计算结果表明该方法能可靠识别内部故障和励磁涌流。
Hidden Markov m odel(HMM), w hich is initially em ployed in the field of speech signalprocessing, is introduced into transform er relaying protection. HMM is depicted briefly, and a new m ethod to detect transform er internalfault and m agnetizing inrush currentbased on HMM is presented. Models for fault and m agnetizing inrush currentdata generated by PSCAD are trained utilizing discrete HMM. The calculation result shows that this m ethod is able to detect transform er internalfaultand m agnetizing inrush currentcorrectly.
出处
《电力系统自动化》
EI
CSCD
北大核心
2000年第3期23-28,共6页
Automation of Electric Power Systems
关键词
励磁涌流
隐马尔科夫模型
变压器
电力系统
m agnetizing inrush current
discrete
hidden Markov m odel(HMM)
probability
