Evaluation of Optimal Weibull Parameter for Wind Resource Assessment at Coastal Terrain by Metaheuristic Optimization Algorithms Using Reanalysis Data

This study focuses on evaluating wind resources at a coastal terrain in India by optimizing the parameters of the Weibull distribution. The research utilizes a 22-year dataset from NASA MERRA-2 reanalysis to analyze wind characteristics. The Weibull probability distribution is used, and the shape (k) and scale (A) parameters are determined through three different numerical techniques [modified maximum likelihood method (MMLM), empirical method of Lysen (EML), and method of moment (MOM)] along with two metaheuristic optimization algorithms [particle swarm optimization (PSO) and genetic algorithm (GA)] at a height of 50 m. The accuracy of the methods is assessed using goodness of fit tests, RMSE, and R2. The results indicate that all parameter estimation methods are suitable, but the MMLM is the most accurate for evaluating wind potential. As compared to GA, PSO was shown to be more efficient. These findings are anticipated to contribute to existing knowledge on probabilistic modeling of wind resources and the usefulness of MERRA-2 in assessing wind resources.