Optical-Electrical-Thermal Performance Modelling of Semi-transparent PV Windows Based on Grasshopper

Semi-transparent photovoltaic (STPV) windows demonstrate excellent energy performance, generating electricity for buildings while maintaining indoor daylighting. However, the optical, electrical, and thermal properties of STPV windows are inherently coupled, making it challenging to simultaneously optimize their daylighting, power generation, and heat transfer performance. This paper proposed a simulation model for the optical-electrical-thermal performance of STPV windows based on Grasshopper, enabling parametric design optimization. The model employed a co-simulation approach to separately simulate the thermal transfer, daylighting, and power generation performance of photovoltaic windows. An experimental setup for CdTe semi-transparent photovoltaic windows was constructed in this study to validate the model using empirical data. Results indicate that the optical model exhibits a Normalized Mean Bias Error (NMBE) of −2.98% and a Coefficient of Variation of Root Mean Square Error (CV(RMSE)) of 25.09% at the vertical plane (1.2 m height), while showing −5.99% NMBE and 25.23% CV(RMSE) at the horizontal plane (0.75 m height). The electrical and thermal models demonstrate NMBE values of −2.44% and 0.94%, with CV(RMSE) values of 17.66% and 2.22% respectively. All model errors fall within acceptable thresholds, establishing a technical pathway for the comprehensive optimization of optical-electrical-thermal performance in semi-transparent photovoltaic windows.