Advancing Renewable Energy Systems through Explainable Artificial Intelligence: A Comprehensive Review and Interdisciplinary Framework
Ahmmod Musa Sazib, Joynul Arefin, Sabab Al Farabi, Fozlur Rayhan, Margaret J. Karim, Shamima Akhter
Abstract
Explainable Artificial Intelligence (XAI) plays a pivotal role in advancing transparency, reliability, and informed decision-making in renewable energy systems. This review provides a comprehensive analysis of state-of-the-art XAI methodologies—including Shapley Additive Explanations (SHAP), Local Interpretable Model-agnostic Explanations (LIME), Deep Learning Important FeaTures (DeepLIFT), and rule-based models—by critically evaluating their applications, advantages, and limitations within renewable energy research. Despite notable progress, significant challenges persist, including computational inefficiencies, the absence of standardized evaluation metrics, and the inherent trade-off between model accuracy and interpretability. This study proposes a novel interdisciplinary framework that integrates domain-specific XAI methodologies, standardized benchmarking protocols, and collaborative efforts between AI researchers and energy experts. By addressing these challenges, this review aims to facilitate the broader adoption of interpretable and reliable AI-driven solutions for the sustainable advancement of renewable energy systems.