Research on Short-term Forecasting and Uncertainty of Wind Turbine Power Based on Relevance Vector Machine

With the increasingly serious environmental pollution, the development of clean energy has received widespread attention. This paper proposes a short-term wind power forecasting method for a single unit based on a correlation vector machine. This method uses a small amount of historical numerical weather forecasts and historic wind farm power data as training samples. Combining with future numerical weather forecasts, predict the power generated by wind turbines. This paper analyses the influencing factors of wind power forecast uncertainty, and describes the uncertainty of power forecasting through the confidence interval estimation method. The RVM, wavelet and RBF were used to predict and compare the power of three wind farms in a northern wind farm. The simulation results verified the accuracy and efficiency of the RVM method. (C) 2019 The Authors. Published by Elsevier Ltd.