The research on the reliability prediction method of distribution system

The power supply reliability was an important index to measure a city’s development. So the prediction of the power supply reliability become an important research. The traditional methods to predict power supply reliability of distribution system such as BP neural network and SVM neural network had low accuracy.Based on the above questions, this paper provided a PSO-BFO-KELM method to predict power supply reliability of distribution system, whichwas more precise. The kernel limit learning machine was used for modeling, which avoided the linear inseparability of traditional limit learning machines when processing low-dimensional data. In order to aviod the limitations of a single prediction model, this paper provided the combination prediction model based on VC score algorithm. Regarding the actual city power supply reliability data as example, the simulation verified the PSO-BFO-KELM method was accurate and feasible, the combination prediction model was stable and reliable.