Surrogate Modeling for Rapid Prediction of Energy Yield from Vehicle-Integrated Photovoltaics

This work demonstrates the development of a ma-chine learning-based surrogate model for rapid prediction of energy yield from vehicle-integrated photovoltaic (VIPV) systems. The surrogate model was trained using results from thermal-electrical simulations that couple a commercial heat transfer solver with temperature-dependent electrical models. The trained model was tested on unseen data also generated from the thermal-electrical simulations. The surrogate model was used to perform a sensitivity analysis of the impact of location and meteorological conditions on VIPV energy production.