摘要
为实现电力电缆故障点在线迅速准确的定位,结合电力电缆的故障特点,提出一种利用小波神经网络对电力电缆故障进行定位的方法。应用小波包分析理论对电力电缆故障信号进行特征提取,并在此基础上利用人工神经网络强大的非线性特征函数拟合能力来实现小波包系数到故障距离的映射,解决故障在线测距问题。选择若干历史故障测距数据进行仿真试验,试验结果表明,利用训练成熟的小波神经网络能够在较小的误差范围内实现故障定位。
For realizing quick and correct online positioning of electric cable fault point, this paper proposes one method based on wavelet neutral network by combining with fault characteristics of electric cable. It uses wavelet packet analysis theory to extract features of electric cable and realizes mapping from wavelet packet factor to fault distance by using strong matching capability of non-linear characteristic function of artificial neutral network to solve online distance measurement for fault. It selects several historical distance measurement data to proceed simulation test and the result shows that it is pos- sible to realize fault positioning in a comparatively small error range by using mature wavelet neutral network.
出处
《广东电力》
2013年第2期49-52,68,共5页
Guangdong Electric Power
关键词
电力电缆
故障定位
小波包理论
小波神经网络
仿真试验
electric cable
fault positioning
wavelet packet theory
wavelet neutral network
simulation test