Multiplicative Seasonality Prophet Model for PV Energy Forecasting and Anomaly Detection

With the increasing adoption of renewable energy, accurate and efficient monitoring of PV systems is essential. Traditional diagnostic methods are often insufficient for large-scale plants. This work employs Prophet, a forecasting model developed by Meta, to detect performance deviations and identify potential faults. Using a multiplicative seasonality configuration, the model accounts for long-term degradation and seasonal trends reaching R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.979. The proposed methodology combines data preprocessing, model training, and residual-based anomaly detection. A real-case study validates the approach, confirming Prophet’s suitability for predictive maintenance by offering accurate forecasts with low computational complexity.