A combined forecasting method for renewable generations and loads in power systems

Abstract

Accurate forecasting for "net load", i.e., the difference between the renewable generations and loads, are important for economical and secure dispatch of power systems. Of course, it is significant to ensure sufficient levels of ancillary service, in particular regulation service. Previously, wind power, photovoltaic generation (PV) and loads are forecasted separately. In contrast, in this paper, a direct and adaptive combined forecasting method is proposed for wind power, PV and load which is regardless of market structure (centralized planning/dispatch vs. market outcomes). Compared with the traditional forecasting methods such as support vector machine (SVM), it can online adjust model parameters to improve the forecasting accuracy. A contrastive analysis is performed between the separate forecasting model for wind power, PV and load, the offline combined forecasting model and the proposed approach. The results show that the proposed method can be self-adaptive to the fluctuation of renewable energy and is able to make the forecasting more accurate.