Cooling, Heating and Electrical Load Forecasting Method for Integrated Energy System based on SVR Model

In order to further reduce environmental pressure and promote the integration of renewable generation, integrated energy system (IES) has become a promising way of energy consumption. The economic dispatch and optimal operation of the IES rely on accurate load forecasting. In this paper, a Support Vector Regression (SVR) based multiple load forecasting method for cooling loads, heating loads and electrical loads of integrated energy system is established. First, through Pearson correlation analysis, the correlation between cooling loads, heating loads and electrical loads are investigated. Then, a load forecasting model based on SVR is designed, and Particle Swarm optimization (PSO) is adopted to optimize model parameter setting. Electrical loads, heating loads, cooling loads, day type, and weather data are used as inputs in the prediction model. A case study on a realistic IES of a park in Yunnan Province is implemented to verify the proposed method. Comparing results of the proposed method with that of traditional models show that, the proposed model can effectively consider the coupling of power load, cooling load and heating load, and has better prediction accuracy.