A Distribution Network Identification of Cause of Line Loss Method Based on Markov Random Field

The distribution network structure is huge and complex, and the system operates in a variety of environments. It displays differentiated operation characteristics in different regions, different environments and different time. With the application of advanced technologies such as distributed generation and micro grid, the operation of distribution system is more and more random. The problem of distribution network line loss is particularly obvious, especially because of the wide distribution of distribution network, it is difficult to find the position of high loss. In order to strengthen the continuous power supply capacity of the distribution network and improve the control and repair of all kinds of high loss positions, this paper puts forward a new method of identification of various types of distribution network line loss. Based on Markov random field method, the identification model of line loss causes in distribution network is constructed. The line loss identification is realized by using the fuzzy and uncertainty of various priori conditions related to the position of line loss. This method is better than the classical support vector machine and neural network classification algorithm to deal with the aliasing of the classified edge better, obviously reduces the interference noise, and also has a certain improvement in the accuracy of the algorithm. Finally, the proposed algorithm is applied to the identification of line loss causes of power distribution network and has achieved good results.