Short term wind speed forecasting and wind energy estimation: A case study of Rajasthan

Abstract

Wind power has increased in recent years to meet the growing energy demand. Accurate short term wind energy forecast is important for optimal scheduling of the wind farms. Increasing the accuracy can help the power system operators to increase the reliability of energy supply. A Nonlinear Autoregressive Network with Exogenous Inputs (NARX) Network is applied to predict hourly wind speed and power. Meteorological variables are taken as exogenous variables for wind speed prediction. Geographical area under study is Jaisalmer in Rajasthan. Wind data is obtained for this location from National Renewable Energy Laboratory (NREL) for one year. Comparison of the wind speed prediction of NARX model with linear regression and persistence model is also carried out. The performance of the test results is done using statistical error measurements like MAE, RMSE and MAPE. The results obtained with NARX network are found to be promising.