Selection method of community load coincidence factor based on BP neural network

Selecting the community load coincidence factor reasonably is the premise for load accurate forecasting work of the power system and can effectively guide the smooth development of distribution network planning. Currently, selection of the community load coincidence factor is mainly relied on the planners and the relevant provisions of the electrical conduction, which is not combined with the practical, thus a community load coincidence factor selection method based on BP neural network is presented by the paper. Combining with the actual situation, the main factor influential factors of the community load coincidence factor are identified by the method, and a BP neural network model predicting the community load coincidence factor is established. The actual example results show that the prediction results proposed by the method are more systematic and scientific, and the absolute error can meet the requirements of the actual engineering´s precision, the selection work of the community load coincidence factor can also be carried out scientifically and effectively by using the method.