Module defect detection and diagnosis for intelligent maintenance of solar photovoltaic plants: Techniques, systems and perspectives

Abstract

The energy production efficiency of photovoltaic (PV) systems can be degraded due to the complicated operating environment. Given the huge installed capacity of large-scale PV farms, intelligent operation and maintenance techniques and strategies are required to keep the healthy operation of the photovoltaic system. A complete inspection system, which is a key part of the intelligent operation and maintenance system, should focus on the following issues: defects types and mechanisms, defects detection methods, IoT techniques and UAV-based inspection methods. In this review, a comprehensive study is proposed to review and conclude the research advance and the prospects. In particular, given the complicated operation condition, we first review the environmental factor causing the defects and the corresponding possible degradation for PV modules. Then, the defect type and detection techniques are discussed and analyzed. Due to the strong ability for feature extraction, deep learning is a useful tool for defect detection of PV modules. Considering the location and geographical characteristics, conventional manual inspection is inefficient and even infeasible in practice. IoT techniques and UAV-based systems are utilized more and more popular, which are also discussed and summarized in this review. Due to the limit of the I/V sensors in the PV plants, this work reviewed the UAV-based system in detail, which has high efficiency for inspection and is widely used in industry, especially for visible and IR image-based systems. With technological advances in image sensors, the UAV-based system mounted with an Electroluminescence (EL) camera also presents huge potential. Finally, the conclusion and future direction for intelligent inspection and defect detection are provided.