基于小波变换的局放信号检测与提取

A Method Based on Wavelet Transform for Detictioning Partial Discharge Singal

为提高局放信号检测的准确度,确保电力设备故障诊断的有效性和可靠性,利用连续小波变换对背景噪声较强的的局放信号进行多尺度分解,在某些尺度下,使局放信号明显增强,用阀值比较就可以有效地检测并消除噪声干扰。实验和仿真结果表明,选取合适的小波函数就能有效地抑制噪声干扰并完成对局放信号的检测和特征提取。

In order to improve the detetion accuracy of partial discharge signal,and ensure the validity and reliablitly of false diagnosis for the enquipment,in this paper the continues wavelet transform is applied to decompose partial discharge pulse at multi-scales,which contains the stark noise. In some scalesof wavelet transfom,the characteristics of the pulse is obviously enhanced so that we can accurately detect the puls,and eliminate noise interference by using the thrshold method. The results of simulation and experiment show that optimail wavelet function is an effective approach to restrain noise interference,and detect partial discharge plus.