Study On Power Quality Disturbances Classification Algorithm

With the healthy development of our national economy power load grow fast and input use of various nonlinear load in power system is more and more. The result is voltage and current waveform of grid has significant distortion. Consequences caused by deterioration of power quality are mainly:power equipment facing huge challenges, the increase of network losses and reduce service life of equipment etc. Power quality has gradually become an urgent important technical problem. So, as the first step to improve the quality of power, its particularly important to detect accurately various power quality disturbance type of event.This paper is mainly against the work of power quality disturbances classification algorithm and programs it to complete solve practical problems. Solution is, firstly, using wavelet packet transform to extract the features of disturbance signal. Then by decision tree category classify disturbance signals according to their features. At last, a complete decision tree rule will be formed for the purpose of classification of power quality disturbance events. Meanwhile, this paper studied wavelet packet transform algorithm and data mining areas of decision tree algorithm such as ID3,C4.5 and provides a whole algorithm procedure and program realization process. According to the original decision tree algorithm of some disadvantages this paper improves it, in order to improve the efficiency of the algorithm, shorten the classification of time. The end of the article, taken on algorithms for data validation show that by using wavelet packet transform classification algorithm and decision tree of disturbance classification classic result is more accurate, classification rules are visual concise, and the improved C4.5 algorithm can effectively improve the efficiency of the algorithm and shorten the running time of the program.