Analysis of Wind Characteristics using ARMA & Weibull Distribution

Abstract

Rapid growth of population demands huge increase of electrical power that can’t be fulfilled with the expansion of conventional generation due to environmental concern. To meet the aggregating power demand, alternative generation like wind energy is getting more importance. Wind energy conversion system (WECS) transforms available speed at a location into electricity. But the problem associated with WECS is the uncertainty in the speed of the wind. Thus, efforts have to be made to showcase the randomness of the wind speed. Two famous methods time based auto regressive moving average (ARMA) and frequency based Weibull distribution are generally followed to fulfill the purpose. Both methods are used to develop models to fit to the observed speed. The accuracy of fitting in case of ARMA is checked through Box-Jenkins guidelines and F-Criterion. But in case of Weibull distribution, the parameters are determined with the estimation of statistical errors.