Siliang Lu will present the defense for her dissertation, "An interactive hybrid cooling system featuring personalized controls with personal thermal comfort and non-intrusive sensing techniques," to obtain the PhD in Building Performance & Diagnostics (PhD-BPD) on Friday 30 August in MMCH 415.
Title: “An interactive hybrid cooling system featuring personalized controls with personal thermal comfort and non-intrusive sensing techniques”
By Siliang Lu, PhD Candidate
Date: Friday, 30 August 2019
Location: MMCH IW 415
PhD Advisory Committee
Dr. Erica Cochran Hameen, Carnegie Mellon University (CMU); Chair
Dr. Omer Tugrul Karaguzel, CMU
Dr. Berangere Lartigue, Paul Sabatier University
The Heating, Ventilation and Air-Conditioning (HVAC) system plays a key role in shaping the building’s performance. Effective and efficient HVAC operations not only achieves energy savings but also creates a more comfortable indoor environment for occupants. Moreover, compared to the private office environment, the open-plan office environment has become a trend in most office buildings since it not only creates opportunities for employees to communicate with one another and improve productivity but also reduces construction cost. However, the open-plan office building layout is also faced with problems such as interruptions from occupants and coworkers and unsatisfactory shared indoor temperature and humidity levels. Therefore, it is of great importance to develop a new paradigm for the HVAC system framework so that everyone can work under their preferred thermal environment while also achieving improved energy performance. But how can we achieve individualized thermal comfort and energy efficiency without being intrusive?
This dissertation proposes a new integrative hybrid cooling system featuring an adaptive personalized system with non-intrusive sensing techniques for open-plan office spaces. The research mainly consists of four parts:
Development of a personalized cooling system to create a comfortable local thermal environment automatically with non-intrusive sensing techniques and machine learning algorithms. The sensing system consists of an indoor air temperature sensor, relative humidity sensor DHT22, and an infrared temperature sensor AMG8833.
Quantification of the energy savings of the proposed hybrid cooling system and optimization of the cooling set-point of the ambient conditioning system to further save energy.
Development of a data-driven approach with CFD simulator to analyze the benefits of energy savings in a typical office space while maintaining acceptable thermal comfort with the proposed hybrid cooling system.
Development of an energy co-simulation with the proposed hybrid cooling system to analyze the benefits of energy savings in a typical shared office space while maintaining acceptable thermal comfort with comfort database I & II.
As a result, in terms of energy savings, five 3-hour sessions in the field study have shown that the proposed system can achieve 9.6% in HVAC energy savings on average compared with the baseline. Moreover, based on energy co-simulation models, the energy performances with the proposed hybrid cooling system could also be optimized to save HVAC electric demand power by 5.3% on average.
Additionally, in terms of thermal comfort, the performance in the field study has shown the recall scores of the thermal sensation model and satisfaction model with the data from all female subjects are 84.7% and 76.5%, respectively. Meanwhile, the recall scores of the thermal sensation and satisfaction models with the data from all male subjects are 87% and 82.5%, respectively. Furthermore, an automated feedback collection mechanism was implemented to update occupant thermal comfort models by collecting override actions by occupants (frequency that occupants manually changed the fan speed overriding the programmed automation system). As a result of machine learning algorithms developed in this research, 60% of subjects had fewer override actions with the updated thermal sensation models compared with the initialized thermal sensation models.
Overall, the proposed hybrid cooling system featuring personalized cooling system not only optimizes energy performance, but also provides a more comfortable thermal environment in open plan office spaces.