AI driven coupled optical thermal electrical optimization of climate responsive colored bipvt façades for energy efficient buildings.

This study investigates the optimization of colored BIPV/T facades using a coupled optical-thermal-electrical model and artificial intelligence (AI) in two climates of Addis Ababa (humid-temperate) and Riyadh (hot-arid). In this study, a nonlinear model for predicting electrical and thermal efficiency was presented that combines EnergyPlus simulations with the Transformer model. The study specifically focuses on the development of climate-responsive BIPV/T systems that can optimize their performance in response to changes in climatic conditions and solar irradiation. The framework was validated against multiple experimental datasets. The results showed that light-colored modules in Addis Ababa and Riyadh have higher thermal efficiency than dark-colored modules, but this increase is more noticeable in Riyadh due to more intense radiation and higher temperatures. While in Addis Ababa, dark colored modules have better electrical performance due to milder conditions. Furthermore, the results obtained from the Transformer model were confirmed to be highly accurate in predicting the efficiency of the systems (with an RMSE value of less than 3%). Pareto front analysis indicates an optimal balance at 20.8% electrical and 74.0% thermal efficiency in Addis Ababa, shifting to a higher maximum thermal efficiency of 80.8% (at 17.1% electrical) in Riyadh.