Abstract

The primary aim of this study is to develop a hybrid system that maximizes thermal and electrical energy outputs, reduces operational costs, improves environmental performance, and ensures reliable operation. The experimental setup involved installing the hybrid system in a residential test facility equipped with instrumentation for real-time monitoring and control. Performance testing was conducted under various operating conditions, measuring the thermal and electrical outputs of the PVT collectors with embedded PV cells, GSHP performance, and overall system efficiency. The system's performance was evaluated in both heating and cooling modes, with results indicating a GSHP Coefficient of Performance (COP) of 4.2 during winter heating and a PVT collector efficiency of 18 %. Significant reductions were observed in annual heating loads and grid-purchased electricity compared to traditional systems. Optimization was achieved using a hybrid approach that combined Genetic Algorithms (GA) and machine learning (ML) techniques, which iteratively improved system design and operational strategies. The conclusion highlights that the integrated GSHP and PVT system substantially enhances energy efficiency, reduces greenhouse gas emissions, and offers a payback period of approximately 4.67 years. Future work will focus on further optimization and real-time operational refinements to ensure adaptability to varying industrial conditions.