Comparative Analysis of Artificial Intelligence Models for HVAC System Optimization in UNESCO Heritage Buildings

This study explores the integration of advanced machine learning and artificial intelligence technologies in historic buildings to optimize energy consumption management while preserving cultural heritage and ensuring occupant comfort. Focusing on a historically significant church in San Antonio, Texas, two predictive control models, a Feedforward Neural Network (FNN) and an Adaptive Neuro-Fuzzy Inference System (ANFIS), are developed and compared against a conventional controller. Results demonstrate the promising predictive capabilities of both FNN and ANFIS models in regulating HVAC system operations. ANFIS outperforms FNN due to its ability to incorporate fuzzy inference systems (FIS), enabling the learning of hidden rules within the data. Lastly, the study emphasizes the need for robust strategies to balance energy efficiency with heritage preservation and occupant comfort in historic buildings.