A method of predicting failures of 10KV distribution networks of countries every month by Markov chain with weights

摘要

The failures in 10KV distribution networks of countries every month are affected by various factors such as weather, load. There are much uncertainty and imprecise characteristics in predicting the failures of 10KV distribution networks every month. A method called Markov chain applied in predicting the future failures of rural distribution networks is presented in this paper. The number of fault is calculated according to the history data of the failures sequence. The self-coefficients of the historical data sequence is calculated as weight number by applying the scheme of blur and cluster to constitute the standard of classification. The number of fault in the next month is predicted by applying the model of the Markov chain. Other disturbance is avoided by applying the scheme presented in this paper, the precision is improved and model of prediction is simplified at the same time. The scheme is validated at last the paper by applying the scheme to predict a real failures of 10KV distribution networks in nearly 40 months, and an ideal result is obtained.