With rising global energy demand and increasing awareness of climate change, the share of renewable energies in the global energy mix continues to grow. Unlike conventional power generation, the output of renewable energy sources is inconsistent due to environmental conditions, making reliable prediction of current and future energy production essential. This paper reports findings from two structured literature reviews on wind turbine power curve modeling and photovoltaic power prediction.
For wind energy, a comprehensive literature review was conducted, while for photovoltaic energy, key findings regarding frequently selected input features were synthesized from existing surveys. The analysis reveals that, despite the abundance of monitoring and environmental variables, feature selection methods remain limited and unsystematic. To address this, we propose Cluster-based Sequential Feature Selection (CSFS), a novel, model-agnostic, clustering-based wrapper method for automatic, efficient, and reliable feature selection in renewable energy prediction pipelines.
To support reproducibility and reuse, we provide an open-source implementation of CSFS on GitHub. We empirically evaluate the proposed approach on both use cases and compare it with established feature selection techniques, such as wrapper-based sequential feature selection (SFS), filter-based methods, and Random Forest's embedded feature importance. Results show that wrapper-based methods generally yield better feature selection performance. CSFS achieves predictive performance comparable to SFS while reducing computational costs by an average of 21%.
Blogger's Review: This paper effectively addresses the feature selection issue in renewable energy prediction by proposing the CSFS method, showcasing the potential of combining clustering with feature selection, and providing significant insights for future research and practical applications.