IoT-based HVAC system modeling

¹ Escuela Superior Politécnica del Litoral, Facultad de Ingeniería en Electricidad y Computación, Guayaquil, Ecuador
² Escuela Superior Politécnica del Litoral, Facultad de Ingeniería Mecánica y Ciencias de la Producción, Guayaquil, Ecuador
³ University of Central Florida, Department of Mechanical Engineering, Florida, United States of America

^* e-mail: jjlainez@espol.edu.ec
^** e-mail: douplaza@espol.edu.ec
^*** e-mail: caralcas@espol.edu.ec

Abstract

One of the greatest challenges for buildings equipped with HVAC systems is achieving a balance between occupant thermal comfort and energy efficiency. Indoor air temperature is typically regulated by a classic thermostat located on site, which serves as a single reference point. However, this approach can result in unsatisfactory thermal conditions in areas beyond the thermostat’s control, failing to provide the desired comfort or efficiency. This article presents a mathematical model of a real HVAC system installed at the Guayaquil Convention Center, a venue hosting diverse activities with a high flow of occupants. The installed HVAC capacity and thermal load are analyzed and compared to validating the mathematical model. The model is based on system identification techniques using real data collected from IoT sensors installed in the convention center over several months at different times. For modeling, polynomial models such as ARX, ARMAX, OE, and BJ were employed. The IoT sensors, together with a web platform, enable online temperature monitoring for future investigations aimed at developing a control system to reduce electrical energy consumption.