采用Gabor变换的局部放电信号时频分析

Time-frequency Analysis of Partial Discharge Signal by Gabor Transform

局部放电信号频率是时变的,为更好地反映它的时频特性,采用一种时频分析方法Gabor变换,将它应用于局部放电信号分析中,并利用不确定原理选择高斯窗函数作为Gabor变换的基函数,提高了分析结果的时频聚集性。通过对实际局部放电信号进行分析处理,结果表明在选择合适的窗函数参数的情况下,基于Gabor变换的时频谱能够细致的刻画信号在时频平面上所发生的变化过程,准确反映了局部放电信号的时频特性,可有效满足对局部放电信号进行时频分析的要求,有利于对信号特征的提取。

In order to understand more details on the time-frequency characteristics of the partial discharge signal, the time-frequency distribution is an efficient tool. The Gabor transform is a time-frequency transform, in which sigrials are expressed as the sum of wavelets of sinusoidal form and Gaussian amplitude variation, it is very suitable for analyzing non-stationary time-varying signals. The Gabor transform was applied to time-frequency analysis of partial discharge signal in this paper, because different window functions in regard to their mean square time and frequency widths used in the Gabor transform would produce different results, so the Gaussian window function which can improve joint time-frequency localization of the results was selected by using the Heisenberg inequality. Experimental results show that the time-frequency analysis of partial discharge signal based on the Gabor transform can describe the change of the signal details on the time-frequency plane, and it also can reflect the characteristics of time and frequency accurately of the partial discharge signal. The results can effectively meet the requirements of choosing suitable parameters, it can be used in characteristic extraction of signals, and it also offers a kind of new method for pattern recognition of PD signals.