用于提升光伏/热能系统效率的智能建模与设计框架

Abstract The increasing global demand for energy, coupled with the depletion of fossil fuels and the growing awareness of environmental issues, has necessitated the transition towards renewable energy sources. Among the various renewable technologies, solar energy stands out as one of the most abundant and sustainable options. In photovoltaic-thermal (PV/T) systems, excess heat is transferred to a cooling medium, such as air or water, to regulate cell temperature. An intelligent and controllable framework is proposed for designing PV/T thermal systems with high levels of efficiency. The framework begins with creating a parametric design of the PV/T unit. Then, the design of experiment (DoE) block is created. In the present model, the parametric design consisted of three variables and the face-centered central composite scheme was utilized. After deriving a predictive model, linking the design variables and objective functions, the functionality of the system under varying conditions is analyzed. In the next step, a machine learning-based algorithm, i.e. non-dominated sorting differential evolution (NSDE), was utilized to find the optimum model, fulfilling adjusted needs such as higher thermal and/or electrical efficiency. It is noticed, that Nusselt number is highly influenced by changes in Reynolds number. The results revealed that the Nusselt number increases by over 80% as Reynolds number rises from 500 to 2200, demonstrating significant heat transfer enhancement. The optimization framework also led to an improvement in electrical efficiency from approximately 12.4% to 13.2%, achieved without the use of nanofluids or phase change materials. Trade-offs between thermal enhancement and pressure drop were managed through a multi-objective optimization approach, ensuring that net power consumption was minimized. The final optimized configurations reflect a balance between performance gains and energy input requirements, providing a practical and scalable design strategy for high-efficiency PV/T systems.