矿用电缆局部放电监测的最优二代小波包基降噪  

Second Generation Wavelet Packet Based Mining Power Cables Partial Discharge On-Line Data Denoising

作　　者：唐明[1] 梁得亮[1] 王青山[1] 刘骧[1] 

机构地区：[1]西安交通大学电力设备电气绝缘国家重点实验室,西安710049

出　　处：《西安交通大学学报》2013年第12期32-37,76,共7页Journal of Xi'an Jiaotong University

摘　　要：为了提高局部放电信号的信噪比,以便更准确地提取局部放电脉冲,在二代小波基础上,构造基于插值细分法的二代小波包,对矿用电缆在线监测中染噪的局部放电信号进行降噪处理。在搜索最优小波包基时,以均方根误差、信噪比与峰值误差作为评价指标,比较了阈值熵、香农熵、log energy熵、SURE熵以及NORM熵的降噪效果。实验结果表明,从以上各种熵所搜索的最优小波包基去噪情况分析,阈值熵的综合性能最好,其均方根误差与峰值误差分别为8.792 4×10-4及0.021 1,细小毛刺干扰信号已被滤除,信号较为光滑,且保留了原始信号的大多数特征信息。To improve the signal-to-noise ratio （SNR） of the partial discharge （PD） signal and accurately extract the PD pulses, second generation wavelet transform packet is proposed for effective denoising of the partial discharge data obtained from on-line monitoring of mining power cables. Noting the advantage of statistical properties of white noise, the variance is taken as the criterion to search the best wavelet packet tree. Compared with Shannon entropy, log energy entropy, SURE entropy and NORM entropy, the threshold entropy is able to restore PD signal from noise polluted data with very low SNR, the root mean square error and amplitude error approach to 8. 792 4 X 10.4 and 0. 021 1, respectively. In the denoised PD signal remaining the original characteristics, the slight interferences are filtered out.

关 键 词：矿用电缆 在线监测 局部放电 最优小波包 信噪比 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]