小波包变换和支持向量机在孤岛与扰动识别中的应用

Application of Wavelet Packet Transformation and SVM on Classification of Islanding and Grid Disturbance

为防止孤岛检测受到电网扰动的干扰,提出了小波包变换和支持向量机(SVM)相结合进行孤岛与扰动识别的新方法。分析对象选择公共点电压,将其进行小波包分解及重构。针对公共点电压采用标幺值的特点,为了使得不同类型的特征向量区别更加明显,将各节点重构信号做开方,并与电网正常供电情况下相应节点电压值做比值处理,以此作为特征向量,来实现孤岛与扰动的识别。仿真结果表明,该方法能够准确地鉴别孤岛与扰动情况,分类准确率高,能够有效避免光伏系统误退出系统运行。

In order to differ islanding from grid disturbance, we present a novel method to detect and classify islanding or grid disturbance based on wavelet packet transformation and support vector machine （SVM）. The method adopts the vol- tage of point of common coupling（PCC） as its target, which is analyzed through wavelet packet decomposition and reconstruction. Since the PCC voltage is generally recorded in per unit value, to distinguish different characteristic vectors, we extract the roots of the reconstruction values of each note, and then make their division with the normal voltage of the corresponding notes. The processed vectors are employed as the input vectors for the SVM distinguishing between is- landing and grid disturbance. Simulation results indicate that the method can achieve its goal while being highly accurate, so the photovoltaic system mistakenly withdrawing from power grids can be avoided.