基于线调频小波变换的电机故障信号消噪方法

Motor Fault Signals Denosing Based on Chirplet Transform

电机故障信号包含大量白噪声和有色噪声 ,消噪是故障信号分析、诊断、处理的前过程 ,消噪的效果直接影响到后期工作。小波变换在电力系统的故障信号处理中得到广泛地应用 ,然而 ,小波变换只能消除白色噪声 ,对有色噪声不起作用。线调频小波变换是窗口Fourier变换和小波变换的推广 ,它不仅具有小波变换良好的时频局部性特点 ,而且它的时频窗口比小波变换的时频窗口更加灵活。本文应用线调频小波变换对电机故障信号进行消噪 ,效果明显。

A lot of white and colored noises are contained in fault signals of electric machines.Denoising is a necessary pre process before analyzing,diagnosing and processing fault signals.Its efforts influence the later processing.Although wavelet transforms(WTs) are being used to process fault signals of power system widely.Wavelet transforms can only filter white noises and are useless to filter colored noises.Chirplet transform(CT)is extension of short time Fourier transforms(STFTs)and wavelet transforms.It has time frequency locality as wavelet transforms,and its time frequency windows is more flexible than wavelet transform.In this paper,CT has been used to denoise fault signals of electric machines.The results show that denoising of CT is very useful.It is a certainty that CT will be used to process fault signals of power system extensively.