EO-HVAC Framework for Energy Optimization in HVAC Systems

Abstract

The paper presents the EO-HVAC framework (energy-oriented heating, ventilation and air conditioning), which is a smart control system designed to increase energy efficiency and living comfort in the HVAC electro-mechanical system. Artificial Intelligence (AI), Deep Strengthening Learning (DRL), Graff Attention Network (GATS), and Federated Learning, EO-HVAC framework duly united using real time sensor data and predictive analytics for environmental and occupancy changes. The proposed system creates on recent progress in intelligent HVAC optimization, in which transformer-based in-order to learn reinforcement learning and staging methods of dress, to achieve strong generalization and rapid deployment in diverse construction environment. Comparative studies display energy savings up to 45% and improve thermal comfort indices significantly. Framework addresses the limitations of traditional rules-based and model-ecological controllers by employing data-operated, decentralized teaching models that ensure scalability and privacy. This research contributes an integrated, scalable and adaptable solution for smart building energy management, which has been aligned with a increasing demand for intelligent control in global stability goals and HVAC operations.