基于互信息最小化盲源分离准则的电压闪变检测方法

Detection of voltage flicker based on blind source separation of minimum mutual information

为了准确测量电网电压闪变的参数,提出了一种新的基于最小化互信息准则的盲源分离方法。将含有电压闪变的混合信号视为盲信号,对采样得到的混合信号序列分别延时若干点,构造输入信号矩阵,基于互信息最小化盲源分离准则,用Fast-ICA算法从混合信号中分离出闪变的包络波形,再进行频率和幅值的计算,就可得到短时闪变值。设计了由信号发生电路、功率放大电路及采样电路等组成的实验电路,实验结果表明此方法对平稳和瞬时闪变信号的幅值、频率测量精度高,特别是在含非整次调制谐波时,优于传统的快速傅里叶变换算法,取得了较好的效果。

To measure the parameters of voltage flicker accurately,blind source separation algorithm is presented based on the criterion of minimum mutual information.Taking the mixed signal containing voltage flicker as blind signal,the paper samples the order of mixed signals and delays them a number of points,and constructs the input signal matrix.The voltage flicker envelope curve is separated from mixed signal with Fast-ICA algorithm based on blind source separation of minimum mutual information.Then frequency and amplitude of the envelope curve are calculated and short-term flicker value is acquired.An experimental circuit containing signal generation circuit,power amplifier circuit and sampling circuit is designed.Experiment results show that the method is more effective in stationary and instantaneous voltage flicker detection.The method is better than conventional FFT algorithm when envelope curve contains inter-harmonics.