Fault classification and detection for photovoltaic plants using machine learning algorithms

<span>Using photovoltaic (PV) energy has increased in recently, due to new laws that aim reduce the global use of fossil fuels. The efficiency a PV system relies on many types malfunctions which may cause significant loss during system’s operation, besides ecological factors. Consequently, monitoring (MS) capable measuring both environmental and electrical factors is described order gather real-time historical data estimate plant metrics. Additionally, recursive linear model for detecting problems presented, where input irradiance temperature module, whereas output power, using same MS. achieved fault detection’s accuracy 5-kW power reached 93.09 percent, based 16 days 143 hours failures under various situations. After defect, machine-learning-based algorithm categorizes each defect problem as short circuit, partial shadowing, deterioration, or open-circuit. performance four most prevalent supervised machine learning (ML) approaches this assignment (Naïve Bias, decision tree, LDA, KNN) was evaluated according their results.</span>