Research on Short-term Load Forecasting Method Based on ANFIS-BPNN-LSSVM

In order to improve the basis of safe and orderly dispatching of power system, A BP neural network (back propagation), adaptive neuro fuzzy inference system (ANFIS) and least squares support vector machine (LEast-squares) are proposed Load forecasting method (ANFIS-BPNN-LSSVM) combined with supportvector machines. Through the simulation experiment of load data in a certain area, the weight coefficient model optimized by BP neural network, adaptive neural network, least square support vector machine and combined prediction method is compared and analyzed. The results show that the ANFIS-BPNN-LSSVM model has better prediction accuracy than the single method, and also improves the robustness of the combined model, which provides method basis for energy scheduling.