基于相空间重构的电压缺口检测及特征参数辨识

Voltage Notch Detection and Characteristic Parameter Identification Based on Phase Space Reconstruction

电压缺口是由静电放电、电力电子设备换相等引起的电网电压周期性扰动。电压缺口扰动的频率非常高,用常规信号分析方法很难准确检出。该文提出一种利用相空间重构理论对电压缺口进行检测和特征参数辨识的方法。利用坐标延迟法重构吸引子,采用互信息函数法确定最优时延,通过计算序列的关联维数确定最佳嵌入维。电压缺口信号重构相空间后的相轨迹被分为常规和扰动两部分,据此对电压缺口进行检测并提取出缺口发生角度、宽度、均方根深度和最大深度等特征参量。仿真结果表明提出的方法可准确检测电压缺口并实现特征参数的有效辨识,并将所提方法与小波变换检测方法进行对比。该文所提方法避免了传统方法运算量大、难以进行实时分析的缺点。

Voltage notch was a periodic voltage disturbance appeared in power grids caused by electrostatic discharging and power electronic device commutation. As the frequency of voltage notch disturbances was very high, traditional signal analysis methods were difficult to check it out. A method was proposed for detecting voltage notch and identifying the characteristic parameters based on phase space reconstruction. The attractor was reconstructed using coordinate delay method, and the optimal delay was determined with mutual information function. Sequence of correlation dimension was calculated to determine the optimal embedding dimension. The phase trajectory of voltage notch signal in reconstruction phase space was divided into regular component and disturbance. The voltage notch was detected, the characteristic parameters such as angle, width, RMS depth and maximum depth of the voltage notch was extracted. The simulation results show that the proposed method can achieve the task of voltage .notch detection and characteristic parameters identification. Compared with the wavelet transform detection algorithm, this method has the advantages of small amount of calculation and good real-time applications.