Accurate Forecasting of Power Output in Floating Photovoltaic Systems Enhances Renewable Energy Generation

Strategic Insights -

A team of researchers from the University of Malaysia Pahang has developed a novel hierarchical prediction framework that significantly improves the accuracy of power output forecasting in Floating Photovoltaic (FPV) systems. This breakthrough is crucial for optimizing renewable energy generation and enhancing energy management strategies. The study, funded by the Malaysian government and the Universiti Malaysia Pahang Al-Sultan Abdullah, introduced a systematic approach to modeling energy output at three levels: maximum power point tracking, phase-wise, and total system levels. The researchers collected a high-resolution dataset from an operational FPV system and trained and validated five machine learning models within the hierarchical framework.

Key Takeaways:

  • The researchers developed a novel hierarchical prediction framework for accurate forecasting of power output in FPV systems, which is essential for optimizing renewable energy generation and improving energy management strategies.
  • The framework consists of three levels: maximum power point tracking, phase-wise, and total system levels, which capture the interdependencies between different operational levels and improve prediction accuracy and interpretability.
  • The study used a high-resolution dataset collected from an operational FPV system at the Universiti Malaysia Pahang Al-Sultan Abdullah, which comprises meteorological parameters and electrical characteristics.
  • Five machine learning models, including Feedforward Neural Network (FFNN), Random Forest (RF), Extreme Learning Machine (ELM), Support Vector Machine (SVM), and eXtreme Gradient Boosting (XGBoost), were evaluated within the hierarchical framework, with FFNN outperforming all other models.
  • The hierarchical structure improves predictive robustness, reduces error propagation across levels, and enhances real-time monitoring by facilitating localized performance analysis.
  • The framework offers a scalable and adaptable solution for FPV forecasting, contributing to enhanced grid stability and more effective energy management.

Statistics:

  • The hierarchical framework achieves an RMSE (Root Mean Square Error) of 0.0125 and MAE (Mean Absolute Error) of 0.0024 at the system level.
  • The R2 (coefficient of determination) value is 1 at the system level, indicating perfect prediction accuracy.
  • The framework provides a scalable and adaptable solution for FPV forecasting, contributing to enhanced grid stability and more effective energy management.

Sources:

  • VerticalNews, "Investigators at University of Malaysia Pahang Discuss Findings in Renewable Energy (Hierarchical Power Output Prediction for Floating Photovoltaic Systems)", May 23, 2025.
  • Energy, "Hierarchical Power Output Prediction for Floating Photovoltaic Systems", 2025;323.
  • Pekan, Malaysia, Asia, Electronics, Energy, Oil & Gas, Photovoltaic, Renewable Energy, University of Malaysia Pahang.