多N值Db小波多分辨分解的电压闪变检测

The Detection of Voltage Flicker on Different Value Daubechies Wavelet Multiresolution Decomposition

与传统的傅里叶变换相比,小波变换具有时频局域化特性,时频窗口的宽度均可随信号的变化自动调节,在检测突变信号和不平稳信号方面有非常明显的优势。提出一种小波多分辨分析提取电压闪变信号的方法,并根据小波函数的特点和分析的要求,比较选用多N值(消失矩阶数)的Daubechies小波对典型闪变信号进行特征提取和定位。根据调幅波的时频信息获得闪变信号的频率和幅值,实现了对电压闪变发生、恢复时间的精确定位。仿真结果表明,多N值的Daubechies小波结合使用在电压闪变检测和分析中具有更好的效果。

Compared with the traditional Fourier transform, the wavelet transform has the time-frequency localization features, and the width of the window can adapt to the change of the signal automatically, so it has a distinct advantage in the detection of transient signal and unstable signal. A method using the wavelet multi-resolution analysis technology to extract voltage flicker signal is proposed in the paper. According to the characteristics of wavelet function and the demand of analysis, different N values （number of vanishing moments） of the Daubechies wavelet are compared to extract and locate the features of typical transient signal. According to the time-frequency information of amplitude modulation wave, the frequency and the amplitude of transient signal are obtained, and the goal of accurate location of the occurrence and recovery time is achieved. The simulation results show that better performance will be obtained just with the combination of different N values Daubechies wavelet.