Wind and Solar Power Generation Forecasting Based on LSTM and XGBoost Models: A Case Study for Mangaluru

With the growing global emphasis on renewable energy, the share of clean energy sources like wind and solar power in the overall energy mix is steadily expanding. Enhancing the efficiency of wind and solar power utilization necessitates accurately forecasting their generation. This study evaluates the predictive capabilities of two widely used models - Long Short-Term Memory (LSTM), a model of recurrent neural network (RNN), and Extreme Gradient Boosting (XGBoost), an advanced tree-based ensemble learning technique. A comparative analysis is conducted to assess their forecasting performance, and the results indicate that XGBoost outperforms LSTM in terms of both predictive accuracy and consistency.