Power Quality Disturbance Identification Method Based On LMD Energy Entropy And GK Fuzzy Clustering

With massive sensitive disruptive power electronic products are widely used in electric power industry, power quality disturbance problem has become one hot topic in the research of many fields of science. The power quality disturbance signal detection and analysis, realize the disturbance feature extraction and classification recognition is to the necessary premise of power quality disturbance monitoring and governance. In this article, therefore, the traditional method of local mean decomposition(LMD) combined with energy entropy, on the basis of the power quality disturbance signal feature extraction, and to locate and testing, and using the GK fuzzy clustering method for classification.First, in this paper introduces the research background and significance of power quality on the basis of discussing the specific research contents and analysis method of power quality disturbance. Discusses the definition of power quality and its basic content, the basic concept of power quality indicators, and the main factors influencing the power quality indicators and the main harm, expounds the standards of power quality and its significance.Second, LMD and energy entropy method of combining the characteristics of power quality disturbance signal are extracted, so that the type of power quality disturbance classification provides feature vector. But because the LMD method in the process of decomposition of the signal is false component, at the same time, in order to reduce the amount of calculation, simplified calculation results, choose selection method selection PF Shannon entropy weight. As a standard, disturbance signal contains the original PF component that most of the information as feature vector.Third, By comparing the local mean decomposition method and Hilbert Huang Transform(Hilbert Huang Transform, HHT) method for the effect of power quality disturbance detection and location, analysis their respective advantages and disadvantages, and put forward combined LMD and Hilbert Transform method, the effect and avoid the edge effect of HHT method and endpoint LMD method on the basis of the shortcomings on the positioning, about the detection of power quality disturbance signals are more accurate positioning analysis.Finally, using the GK clustering algorithm combined with LMD energy entropy, the power quality disturbance signal after normalization processing feature vector for fuzzy clustering classification recognition. At the same time, through with noise and without noise were analyzed under different conditions, GK fuzzy clustering method can accurately identify power disturbance signal effectively, and whether it has good noise resistance. Then, will identify the type of disturbance to establish the standard sample, through the principle of minimum average close to choose close to test the classification results for the method analysis, eventually reasonably accurate evaluation conclusion.