Machine Learning approach for Short Term Wind Speed Forecasting

Due to depletion of conventional energy resources, the exploration of renewable energy resources has gained a lot of significance. Better forecasting models for the forthcoming supply of renewable energy resources are necessary to reduce the energy consumption from conventional power plants. Wind is a fluctuating kind of energy and accurately predicting the output power of wind energy is important to obtain optimal energy utilization in today's grid operation, dealing with the power load and pollution free atmosphere. This paper presents two machine learning tactics for short term wind power forecasting and that are Support Vector Regression (SVR) and Random Forest Regression (RFR). For this we take number of past observations of a series and that we have used to form the input pattern to train our both the models with which forecasts can be made for a present data point.