Ensemble Learning Techniques-Based Monitoring Charts for Fault Detection in Photovoltaic Systems

Over the past few years, there has been a significant increase in interest and adoption of solar energy all over world. However, despite ongoing efforts to protect photovoltaic (PV) plants, they are continuously exposed numerous anomalies. If not detected accurately timely manner, anomalies PV plants may degrade desired performance result severe consequences. Hence, developing effective flexible methods capable early detection is essential for enhancing their management. This paper proposes data-driven techniques detect DC side plants. Essentially, this approach amalgamates desirable characteristics ensemble learning approaches (i.e., boosting (BS) bagging (BG)) sensitivity Double Exponentially Weighted Moving Average (DEWMA) chart. Here, we employ exploit capability enhance modeling accuracy DEWMA monitoring chart uncover potential In models, values parameters selected with assistance Bayesian optimization algorithm. BS BG adopted obtain residuals, which then monitored by Kernel density estimation utilized define decision thresholds proposed learning-based charts. The schemes illustrated via actual measurements from 9.54 kW plant. Results showed superior BG-based charts non-parametric threshold uncovering different types anomalies, including circuit breaker faults, inverter disconnections, short-circuit faults. addition, compared that BS-based EWMA parametric thresholds.