Short-term solar power forecasting via extreme gradient boosting (XGBoost): experimental validation using real-time data from the Heipang hybrid solar–wind test rig

Domaine:

environment and energy

Type de record:

paper

Créateur:

IshJosBil

Éditeur:

GSC

Hôte:

Accurate short-term solar power forecasting is essential for ensuring efficient operation, stability, and intelligent energy management in hybrid renewable energy systems. This study presents an Extreme Gradient Boosting (XGBoost)-based framework for short-term solar power forecasting using real-time experimental data acquired from the Heipang Hybrid Solar–Wind Test Rig, located in Heipang community, Barkin Ladi Local Government Area of Plateau State, Nigeria. The proposed model utilizes meteorological and operational variables, including solar irradiance, ambient temperature, relative humidity, wind speed, and historical photovoltaic power output, to predict short-term solar energy generation under dynamic environmental conditions. Data preprocessing techniques such as normalization, outlier filtering, and feature engineering were employed to improve model robustness and predictive capability. The XGBoost algorithm was optimized through hyperparameter tuning to effectively capture nonlinear relationships and complex interactions among weather-dependent variables influencing photovoltaic performance. Model evaluation was conducted using standard statistical performance metrics, including Root Mean Square Error (RMSE), Mean Absolute Error (MAE) and coefficient of determination (R²). Experimental results demonstrated that the developed XGBoost model achieved high forecasting accuracy, fast computational performance, and strong generalization capability under fluctuating meteorological conditions. The findings indicate that gradient boosting techniques provide a reliable and scalable solution for real-time solar power forecasting and smart grid applications. This study contributes to the advancement of data-driven renewable energy forecasting approaches for sustainable hybrid energy system optimization and improved power dispatch planning in developing regions.

Visit

doi.org

Similaires

Short Term Power Demand Prediction Using Stochastic Gradient BoostingExtreme Gradient Boosting: A Machine Learning Technique for Daily Global Solar Radiation Forecasting on Tilted SurfacesDaily Solar Radiation Forecasting for Northwest Nigeria Using Long Short-Term MemoryShort-term load and wind power forecasting using neural network-based prediction intervalsShort-Term PV Power Forecasting Using a Hybrid TVF-EMD-ELM StrategyShort-Term Wind Speed Prediction via Sample Entropy: A Hybridisation Approach against Gradient Disappearance and Explosion

Short Term Power Demand Prediction Using Stochastic Gradient Boosting

Power prediction demand is vital in power system and delivery engineering fields. By efficiently pre

Extreme Gradient Boosting: A Machine Learning Technique for Daily Global Solar Radiation Forecasting on Tilted Surfaces

Enhancing solar irradiance and accurate forecasting is required for improved performance of photovol

Daily Solar Radiation Forecasting for Northwest Nigeria Using Long Short-Term Memory

In order to ensure energy security and environmental sustainability, transition to renewable energy

Short-term load and wind power forecasting using neural network-based prediction intervals

Electrical power systems are evolving from today's centralized bulk systems to more decentralized sy

Short-Term PV Power Forecasting Using a Hybrid TVF-EMD-ELM Strategy

This paper discusses the efficient implementation of a new hybrid approach to forecasting short-term

Short-Term Wind Speed Prediction via Sample Entropy: A Hybridisation Approach against Gradient Disappearance and Explosion

High-variant wind speeds cause aberrations in wind power systems and compromise the effective operat