基于统计特征矢量符号值和聚类经验模态分解的短时电能质量扰动信号分析

Analysis of short time power quality disturbances based on statistic feature vector symbolic and ensemble empirical mode decomposition

研究了希尔伯特-黄变换(HHT)方法分析电能质量扰动信号的不足,提出了一种统计特征矢量符号化(SFVS)算法与聚类经验模态分解(EEMD)相结合的短时电能质量扰动信号分析方法。该方法采用循环周期比较的欧氏距离的边界检测算法来对电能质量扰动信号的突变时间进行检测,以突变时刻为边界点将原始电能质量信号进行划分,再用EEMD方法对区块划分信号进行分解,有效抑制模态混叠,以改善信号分解性能。测试结果表明,该方法能够实现突变时刻的准确检测,对电能质量扰动信号中的各种成分进行准确分析。

Aiming at the deficiency of the Hibert-Huang transform（ HHT） method for power quality disturbance analysis,a new method for analysis of short time power quality disturbances based on the combination of statistic feaure vector symbolic（ SFVS） and ensemble empirical mode decomposition（ EEMD） is proposed. This method uses the SFVS algorithm for boundary detection,once the instants of transition are found,the short time power quality disturbance signal can be divided into certain segments of stationary signals and the EEMD can be applied to each segment to decompose it to effectively restrain the mode mixing. The testing results show that the proposed method can accurately detect the time of transition and analyse various components in the short time power quality disturbance signal.