一种抑制PD白噪干扰的有效复合信息技术

根据局部放电(partial discharge,PD)信号与白噪干扰经复小波变换后的不同特点,利用复小波变换后实部、虚部的简单信息构造出一种抑制白噪干扰的实、虚部组合序列的有效复合信息,分析该复合信息中序列幂指数对PD去噪性能的影响,并用去噪前后描述信号波形相似参数(normalized correlation coefficient,NCC),结合信噪比(signal to noise ratio,SNR),对用复小波变换去噪后的效果进行综合评价,给出不同的局部放电波形对应的复合信息序列幂指数值,最后,用所构造的db4复小波对仿真和实测PD信号进行了去噪。研究结果证明了在众多的复合信息中,复小波变换实、虚部组合序列的复合信息能有效抑制PD白噪干扰,畸变小,在-7 dB的强噪声中,提取PD信号的NCC和SNR仍高达0.918 6和17.42 dB,有利于PD检测。

An effective technique of combining information for reducing the white noise of partial discharge signals

Based on the differences between partial discharge(PD) signals and white noise after complex wavelet transformation,an effective information combining technique that included both simple real and imaginary partial information was used to suppress white noise.The influence of the combined information series index on denoising capability of PD was investigated.A comprehensive study of the denoising effect of complex wavelet transformation was carried out by the introducting a normalized correlation coefficient(NCC) and signal-to-noise-ratio(SNR).The constructed db4 complex wavelet was used to denoise simulated and real PD signals.The results indicate that this special information combining composed of real and imaginary parts of complex wavelet transformation,has the best ability to effectively suppress PD white noise in a number of information combinations.Even if the SNR of PD is-7 dB,the NCC and SNR of the denoising signal can still reach 0.918 6 and 17.42 dB,respectively. This result is beneficial for PD detection.