论文部分内容阅读
针对实际电力系统中存在的电能质量复合扰动问题,采用以往单一的方法很难准确识别出复合扰动中所包含的每一类扰动,所以提出了基于动态树和支持向量机相结合的方法分类识别复合扰动。电能质量扰动分类过程分为特征提取和分类器两个阶段。在特征提取阶段,分析了通过d-q变换、小波包分解和S变换对扰动信号进行特征提取,综合3种分析方法得到特征量,得到特征组合,更好地反映出了扰动信号的特征。在分类器设计中,利用了聚类分析中的类距离概念构造出了二叉树结构的支持向量机分类器,以更快、更准确地识别扰动类型。在此基础上,提出了基于动态树的分类方法来识别复合扰动信号中所包含的所有扰动类型。测试结果表明,所提出的复合扰动分类方法可以有效分类出复合扰动中的各类扰动。
Aiming at the problems of complex power quality disturbance existing in the real power system, it is difficult to accurately identify each kind of disturbances contained in the complex disturbance using a single method in the past. Therefore, a method based on dynamic tree and support vector machine is proposed to classify and identify Compound disturbance. Power quality disturbance classification process is divided into two stages of feature extraction and classifier. In the feature extraction stage, the feature extraction of disturbance signals by d-q transform, wavelet packet decomposition and S transform is analyzed. The three kinds of analysis methods are used to obtain the feature quantities, and the feature combinations are obtained to better reflect the characteristics of the disturbance signals. In classifier design, a binary tree-structured SVM classifier is constructed by using the class distance concept in cluster analysis to identify disturbance types faster and more accurately. On this basis, a classification method based on dynamic tree is proposed to identify all perturbation types contained in the composite disturbance signal. The test results show that the proposed composite disturbance classification method can effectively classify all kinds of disturbance in complex disturbance.