PLS-SVR Optimized by PSO Algorithm for Electricity Consumption Forecasting

The development of smart grid and electricity market requires more accurate electricity consumption forecasting. The impact of different parameters of Support vector regression (SVR) on electricity consumption forecasting, and the parameters of SVR model were preprocessed through Particle Swarm Optimization (PSO) to get the optimum parameter values. For the input variables of forecasting model are normalized to reduce the influence of different units on SVR model, and using the partial least square method (PLS) can solve the multicollinearity between the independent variable. A actual data is employed to simulate computing, the result shows proposed method could reduce modeling error and forecasting error, and compared with back-propagation artificial neural networks (BP ANN) and single LS-SVR algorithm, PSO-PLS-SVR algorithm can achieve higher prediction accuracy and better generalized performance.