Predicting Photovoltaic Power Generation in Pakistan Using Machine Learning Algorithms

Photovoltaic (PV) power forecasting is central to energy management and integrating renewable energy into the electricity grid network. This study considered environmental data, namely, Global Horizontal Irradiance (GHI), temperature, relative humidity, wind speed, and rainfall data obtained from NASA's Prediction of Worldwide Energy Resources (POWER) project to estimate the PV power generation in Pakistan. The PV generation dataset of Pakistan is trained using three machine learning algorithms: XGBoost, LightGBM, and Random Forest. The performance of the proposed machine learning algorithm is assessed by Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and R-squared (R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>). The evaluations suggest that LightGBM has the smallest RMSE (0.7116), the lowest MAE (0.3922), and the highest R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> (0.99997), which proves its enhanced accuracy and robustness. The study reveals that environmental factors are crucial in solar PV power forecasting and identified GHI and temperature as the most important predictors of PV power output.