Application of Gaussian Process to wind speed forecasting for wind power generation

This paper proposes a new kernel machine method for short-term wind speed forecasting. Renewable energy is attractive to protect environment. As renewable energy, wind power generation, solar energy generation, geothermal energy generation, etc. are spread in the world. In Japan, wind power generation is of main concern due to the execution of the Renewable Portfolio Standard (RPS). However, it is difficult to deal with wind power generation due to the uncertainty of the wind power output. The power market players are interested in the prediction of short-term wind speed. In this paper, a new method is proposed to estimate the upper and the lower bounds of wind speed as well as the average. The Gaussian Process (GP) based method is proposed to forecast the uncertainty of wind speed. It is extended to consider the kernel machine technique and Bayesian estimation. The proposed method is successfully applied to real data of the Muroto Cape in Japan.