Research On SVM For Condition Assessment Of Insulators

In power transmission lines, the event of insulator Contamination flashover accident often results in a large area and extended power outage, it causes huge losses for the national economy.Surface discharge is the important aspect of pollution level and operating condition of insulator. Real-time monitoring contamination discharge condition plays an important role in the prevention of flashover accident. This paper analyzes the whole course of contaminated insulator on transmission line. It includes wetting, producing dry belt and local electric arc and finally reaching the apex of local electric arc development to be flashover. According to working voltage and time-frequency characteristics of insulator strings in operation, seven characteristic values to symbolize discharge condition is obtained. In accordance with leakage current waveforms in different zones of discharge development process, a dynamic division method based on K-two-end clustering is presented. Exercise K-two-end clustering for discharge processes division of six insulators with the same figure number. As a result, five stages insulators’operating condition, including safety zone, yellow forecasting zone, red forecasting zone, danger zone and the flashover.Support vector machines (SVM) are a class of popular classification algorithms for high generalization ability. However they mainly solve with two-classification problem, while, in practice, there are lots of multi-classification problem still. Based on advantages and shortcomings of existing multi-classification, a kind of neighborhood based SVM multi-classification method is proposed in this paper. To classify Samples in K classes just need to construct K.-1SVM classifiers with this method. And classifiers at prediction stage are chosen according to neighborhood of test samples. By using neighborhood based SVM, pair-wise SVM and BP neural networks to train these five stages and forecast respectively, the usefulness of this multi-classification method and more efficient is proved. In addition, an insulator condition valuation model based on neighborhood is obtained.