摘要
基于盲信号分离(BSS)能快速准确地实现源信号分离和恢复,提出将盲信号分离运用于电压闪变分析。首先对检出的包络信号进行延时处理,构建一个嵌入矩阵作为信号处理的多通道输入信号;再进行鲁棒预白化处理,可以将数据分解到信号子空间以减小噪声对分离结果精度的影响;再运用二阶盲辨识(SOBI)算法进行盲源分离,得到闪变信号的频率估计。在此基础上求解一超定方程组估算出闪变幅值。仿真结果表明,该算法对闪变信号的频率和幅值检测的有效性。
This paper proposes a blind signal separation method for voltage flicker analysis by taking the advantages of blind signal separation （BBS） which could separate source signals fast and accurately. Firstly, an appropriate embedding matrix is constructed out of a series of delay vectors from measured signal as the input signals. Secondly, robust whitening is used to reduce the noise and project the data onto the signal subspace. At last, the second order blind identification （SOBI） algorithm is used to separate the source signals. The method is proved to be highly effective compared with other methods.
出处
《电工技术学报》
EI
CSCD
北大核心
2007年第3期138-142,共5页
Transactions of China Electrotechnical Society
关键词
电压闪变
盲信号分离
鲁棒预白化
SOBI算法
Voltage flicker, blind signal separation, robust whitening, second order blind identification algorithm
