Abstract
Nearly 45% of the energy consumed in residential buildings goes for Heating, Ventilation, and Air Conditioning (HVAC). Typical HVAC systems are mostly controlled by one thermostat that is usually located in the living room. This means that the HVAC system is running without consideration of the thermal conditions in the other rooms (zones), which might be colder (or warmer). Colder zones in the summer indicate that a lot of energy is wasted, and warmer zones mean that the HVAC system can’t produce a good comfort level. Therefore, zoning was introduced into relatively recent HVAC systems. Zoning in HVAC systems uses motorized dampers to regulate the airflow rate in specific home zones. This should help to deliver more air to the occupied zones for comfort or to reduce the air share for the unoccupied zones for ultimately reducing energy consumption. The main challenge that faces zoning systems is how to find the air flow rate for each zone that will satisfy both dynamic conditions of comfort and low energy consumption. This project aims to design and test a smart Zoning system that can find the optimal air volume flow rate for each zone. The Zoning system is designed to operate autonomously based on real-time measurements from sensors. Model predictive control (MPC) is used as a basis for this work to evaluate the required airflow rate for each zone given conditions such as the time during the day, number of occupants, etc. MPC is an optimal control strategy that allows calculating the airflow rate values based on the minimization of the temperature error (difference between actual and desired value) combined with minimizing energy consumption.
In order to implement an MPC, a thermal model of the zones was developed. In the development process, the main focus was to design a model that provides acceptable results, while being simple enough to minimize the workload for the controller. Since a higher complexity of the model will result in an increased susceptibility to disturbances and error, and a slower controller response. For that reason, the model was developed by focusing on the main sources of energy such as conduction from adjacent zones, outside conditions, and estimated internal loads. Additionally, this work included the investigation of the use of Artificial Intelligence (AI) algorithms like Reinforcement Learning (RL) to design smart zoning control systems as an enhanced strategy that would allow to take into consideration more factors without compromising the controller’s performance The results show that MPC zoning can reduce energy consumption by 40% on a typical summer day. Moreover, the temperature distribution in all zones is guaranteed to be as desired.
Date of publication
Summer 8-1-2024
Document Type
Thesis
Language
english
Persistent identifier
http://hdl.handle.net/10950/4747
Committee members
Nael Barakat, Mohammad Biswas, Hayder Abdul-Razzak
Degree
Master of Science in Mechanical Engineering
Recommended Citation
Gomes Diaz, Octavio, "SMART ZONING CONTROL FOR AIR CONDITIONING SYSTEMS" (2024). Mechanical Engineering Theses. Paper 37.
http://hdl.handle.net/10950/4747