Development of Wind Speed Forecast Models Using Artificial Intelligence in Mexicali

This paper studies the application of artificial intelligence (AI) techniques for wind prediction in renewable energy systems using Genetic Algorithms (GA), Particle Swarm Optimization (PSO), and Transformers. The GA model, using data from an on-campus meteorological station, achieved low Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) values, indicating high accuracy. PSO and Transformer models were trained with online datasets. The PSO model optimized ARMA coefficients with p = 8 and q = 6, while the Transformer model captured long-term dependencies. Due to different data sources, a direct comparison with GA was not feasible. The results provide detailed performance metrics and insights into the applicability of PSO and Transformers for real-time wind prediction.