Coupling RDA-RPR-NSGAII optimization design method for comprehensive performance of Building Integrated Photovoltaics

Building Integrated Photovoltaics (BIPV) can reduce the proportion of power used from the grid, thereby conserving energy and reducing carbon emissions. The design method of traditional passive buildings, aiming at comfortableness and low energy demand, is not suitable for the BIPV performance design requirements considering carbon emissions and power load consumption. In this study, a BIPV comprehensive performance design and optimization method is proposed. Firstly, a universal evaluation index system for BIPV considering high-efficiency PV and passive designment is proposed. Secondly, the BIPV stochastic performance distribution model and sample database are constructed through Monte Carlo simulation. Then, redundancy analysis is used to quantitatively identify the key optimal design variables, and the Random Forest Regression algorithm is used to construct the surrogate model as a fitness function for subsequent optimization. Finally, the Python combined with Non-dominated Sorting Genetic Algorithms II is used to complete the Pareto optimization based on the surrogate model to obtain the optimal design scheme. The case simulation results show that the optimization framework achieves a 21.79 % increase in fitting accuracy. Compared with the original sample database, the optimal design scheme has reduced the total annual carbon emissions by 40 %, the life cycle cost by 13.5 %, and increased the PV power generation by 2.4 %, the building energy flexibility index by 61.4 %. It solves the problem of BIPV in the high-dimensional complex design space, and can be used to guide the comprehensive performance optimization design of BIPV considering passive technology and PV technology. © 2025 Elsevier Ltd.