摘要
针对实际电力系统行波信号中存在噪声干扰的问题,提出了一种基于屏蔽滤波的行波信号消噪方法,对信号进行极大重叠离散小波包变换,将信号分解在不同的尺度上,利用能量算子得到能量强化后的小波包系数,通过计算小波包相邻尺度的相关度,归一化后得到的相关系数与各个尺度的小波包系数进行比较,获得信号奇异点,形成屏蔽滤波器,将屏蔽滤波器与各个尺度下的小波包系数相乘,从而得到新的小波包系数,利用小波包系数得到行波信号产生的准确位置。理论分析、ATP仿真结果及现场实际数据验证了该方法的有效性。
A method based on mask filtering is proposed for the noise elimination of traveling-wave signals in power system. MODWT（Maximal Overlap Discrete Wavelet Transform） and Teager operator are used to decompose the signals on different scales and strength the wavelet coefficient energy. By computing the correlation degree between adjoining scales and comparing the normalized correlation coefficient with the wavelet coefficient,the signal singularity points are obtained and the mask filter is formed, by which the wavelet transform coefficients are multiplied to get new coefficients and then determine the precise position where the traveling- wave signal is generated. Theory analysis,ATP simulation and filed data prove its validity.
出处
《电力自动化设备》
EI
CSCD
北大核心
2009年第6期35-39,共5页
Electric Power Automation Equipment
