基于EMD-SG和信息熵矩阵束的同步电机参数辨识

为了解决工程实际中准确获得同步电机瞬态及超瞬态参数问题,提出经验模态分解与矩阵束算法相结合的新型同步电机参数识别法。该方法借助EMD对采集到的含噪短路电流信号进行分解,采用Savitzky-Golay滤波器对高频分量部分进行平滑降噪预处理,借此提高其信噪比;为较好识别短路电流模态阶数,将信息熵引入矩阵束并将此改进矩阵束算法用以提取预处理后的短路电流各分量的频率和阻尼,进而识别出同步电机的瞬态参数。同步发电机三相突然短路仿真与试验参数辨识结果均表明,该方法在信噪比低于24 d B时,仍能快速、精确地辨识同步电机参数。

Parameter identification of synchronous generator based on EMD-SG and information entropy matrix pencil algorithm

In order to obtain accurate value of the generator parameters ,especially transient and sub-tran-sient parameters to satisfy the increasingly sophisticated simulation requirements of the power system ,an empirical mode decomposition ,savitzky-golay filtering and information entropy matrix pencil based method is proposed .It pretreated the short-circuit current by noise with the empirical mode decomposition to im-prove its signal to noise ratio .Then information entropy matrix pencil algorithm was used to extract the fre-quency and damping of each component of short-circuit current .Meanwhile , the transient parameters of synchronous generator were determined with higher accuracy by some simple calculations .Simulation and experiment results show that the proposed method has advantages of high parameter identification accuracy and strong anti-interference ability .