Review of AI Methods for Fault Diagnosis and Predictive Maintenance in Solar Panels

Solar energy has emerged as a cornerstone of renewable energy, driving global efforts to expand photovoltaic (PV) solar power generation. Ensuring the optimal performance of PV solar plants requires robust fault diagnosis and predictive maintenance systems. Artificial intelligence (AI) has proven to be a transformative tool in this domain, enabling advanced monitoring and maintenance strategies. This paper, conducted as part of the ZERODEFECT4PV project, presents a comprehensive review of AI-based methods for fault diagnosis and predictive maintenance in solar panels. The study explores the capabilities and limitations of current AI techniques in identifying and addressing faults at the panel level, including Electrical-Based Methods (EBMs), Visual and Thermal Methods (VTMs), and hybrid approaches. Key AI methods, such as machine learning, deep learning, and neural network, are examined in the context of their applications, efficiency, and scalability. By consolidating insights from existing literature, this review identifies gaps in current methodologies and highlights opportunities for innovation within the ZERODEFECT4PV framework. The findings of this review provide a foundation for developing advanced AI-driven solutions aimed at enhancing operational efficiency and reliability in PV solar plants.