Machine Learning-Based Condition Monitoring of Solar Photovoltaic Systems: A Review

Due to independency from physical models and low cost of implementation, machine learning-based condition monitoring methods for solar photovoltaic (PV) systems have recently gained attentions from academia and industry. In this paper, a literature review is conducted on machine learning applications in condition monitoring of PV systems. Different types of faults in PV systems and the general categorization of PV condition monitoring are firstly introduced. Machine learning-based PV condition monitoring are discussed in three groups, shallow, hybrid, and deep networks. The future research direction in this area is also provided in the paper.