Short-Term Prediction of Power Load for Urban Residents Based on VMD-SVR-PSO Model

Abstract

Electricity load forecasting is the basis for ensuring the balance of power supply and demand, and provides information and basis for the planning and construction of power grids and power sources, as well as the management decisions of grid enterprises and grid users. Load forecasting is divided into long-term, medium-term, short-term and ultra-short-term load forecasting. Accurate load forecasting can economically and reasonably arrange the start and stop of generating units within the power grid, maintain the safety and stability of grid operation, reasonably arrange the unit maintenance schedule, guarantee the normal production and life of society, effectively reduce the cost of power generation, and improve the economic and social benefits. In this paper, an electricity short-term forecasting model based on variational modal decomposition and optimal support vector regression machine is proposed by combining the electricity consumption data of a household in Paris, France, from the UCI database. First, the data are decomposed by using VMD (Variational Modal Decomposition) to reduce the nonsmoothness of the data, thus reducing the impact of nonsmoothness on the prediction performance; then, each decomposed series obtained is predicted by SVR (Support Vector Regression Machine) model, and the parameters in the SVR are optimized by using PSO (Particle Swarm Optimization algorithm) algorithm; finally, the final results are obtained by superimposing all the predicted components. The results show that the prediction model proposed in this paper has high prediction accuracy for short-term power loads.