Performance Comparison of Prediction Algorithms for Forecasting of Wind Power Generation

The current situation, in which there is a rapid intensification in the energy crisis that prevails all over the world, necessitates the transformation of power generation towards alternative, sustainable energy-providing resources. Wind energy resource is yet another prominent alternative for generating electrical power, playing a vital role as a clean, renewable resource in reducing air pollution. Forecasting wind power generation turns out to be the need of the hour for balancing out the ever-increasing demand in power generation systems. Therefore, an accurate wind power forecasting model minimizes the errors in scheduling and increases the reliability of power grids in times of need. In this research paper, the performance comparison of ARIMA, SARIMAX, and ARMA algorithms is performed and it is found that ARIMA provides the lowest error with an MSE measure of 523.01. The forecasting of active power with the ARIMA algorithm at different time scales over an interval of 10 minutes is done.