Co-Simulation Strategy for Photovoltaic Power Prediction and Validation of Digital Twin

The growing energy demand, driven by expanding population and industrial sectors, has created a need to integrate low-cost renewable energy sources (RESs) into the power network. Solar energy is considered a viable means of generating electricity among RESs. However, the inherent intermittency of such energy sources presents various challenges to the power system. Therefore it requires accurate forecasting methods to ensure seamless power flow in energy networks. This paper investigates a data-driven digital twin of a PV plant, validated by incorporating a TRNSYS simulation model of PV using time-ahead forecasted weather data. Various PV power prediction models have been evaluated based on particular key performance characteristics, and a normalized boosting ensemble approach has been chosen to generate a potential digital twin of a PV plant.