Comparative Analysis of Data-Driven Models for AC Power Prediction in PV Inverters

Accurately forecasting the AC power output of photovoltaic (PV) inverters is essential for optimizing energy management, grid integration, and predictive maintenance of PV systems. This study presents a comparative analysis of different data-driven models for AC power prediction, focusing on the Sandia, PVWatts, and Driess models. These models employ various mathematical and empirical approaches to characterize inverter performance based on input power, voltage, and efficiency parameters. The comparative assessment, conducted using simulation-based datasets, reveals that the Sandia model demonstrates superior accuracy and efficiency, providing AC power forecast that closely match measured values. This highlights its robustness for real-world applications in PV system monitoring and performance evaluation. The findings offer valuable insights for researchers and industry professionals in selecting appropriate forecasting models for PV inverters, ultimately contributing to enhanced reliability and efficiency in renewable energy systems.