Wind Power Forecasting Algorithm Based on Similarity of Multivariate Time Series

An accurate prediction of wind power plants has great significance to the power leveling and stability of a power grid. While the electricity market gradually introducing competition mechanism, it could maximize the growth of economic benefits for electricity companies. From a complex system’s point of view, this paper introduces the forecasting of wind power by considering the power generation of wind plants as a black box system, using weighted average and analysis of the historical time series similarity between the meteorological forecast data and the relevant factors affecting wind power generation. The experiment has proof this algorithm simpler, more accurate and maneuverable compared than other wind power algorithms.