Degradation-aware energy management system for PV-integrated multi-zone buildings under diverse climatic conditions

초록

Rapid urbanization has significantly increased building energy demand, highlighting the critical need for efficient integration of renewable energy systems. However, existing energy management strategies neglect degradation of photovoltaic (PV) modules, compromising long-term performance, distorting economic projections, and reducing system reliability over the system lifetime. This study proposes a Degradation-Aware Demand Management System (DA-DMS) for multi-zone buildings, integrating forecasting and multi-objective optimization under photovoltaic degradation conditions. Initially, a novel Encoder-Decoder T-Net (ED-T-Net) architecture is developed for day-ahead PV module temperature forecasting to improve predictive accuracy and reduce degradation risk. Subsequently, the proposed framework employs NSGA-III to optimally size renewable energy storage systems, minimizing cost and emissions while maximizing energy self-sufficiency (SS). A validated 25-year comparative assessment demonstrates that the DA-DMS achieves superior forecasting accuracy, leading to enhanced PV lifetime performance by limiting cumulative degradation to approximately 22% over 25 years, compared with nearly 32% for the conventional baseline. Consequently, the optimized system achieves higher long-term power retention, achieves levelized cost of energy (LCOE) of 0.116/kWh under dynamic conditions. Overall, the results demonstrate that the proposed DA-DMS provides an effective and scalable framework for improving building energy management through enhanced PV reliability, lower lifecycle costs, and reduced carbon emissions.

키워드