摘要
分析了局部放电在线监测系统中局部放电信号及白噪干扰小波变换特征 ,局部放电信号在大部分尺度上具有较大的幅度 ,而白噪声信号随尺度的增加而迅速趋于零 ,从而对信号在各个尺度上的小波系数设定一个门限 ,如果某尺度上的系数大于门限 ,则认为它对应于局部放电信号 ,否则对应于噪声信号。提出了用基于门限值的小波包变换方法从白噪中提取局部放电信号的方法。实测数据分析结果表明 ,经此法处理后信噪比明显提高 。
This paper analyzes the different characteristics of white-noise interference in the signals of partial discharge (PD) after wavelet transform. There is high value in lots of scales for PD and white-noise interference is to zero with increasing scale. The threshold is set for wavelet coefficient in all scales. If the coefficient of signal is higher than the threshold, it is PD signal. Otherwise it is noise interference. A threshold-based wavelet packet transform (WPT) algorithm is put forward to suppress white noise interference in PD signals. The results testifies that it has a favorable adaptability to extract PD signals using WPT.
出处
《重庆大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2003年第6期53-55,共3页
Journal of Chongqing University
关键词
局部放电
白噪干扰
小波包变换
门限
partial discharge
white noise interference
wavelet packet decomposition
threshold