| Author: | Zhang, Xuange |
| Title: | Impact of the interaction between thermal environment and cognitive load on thermal comfort, task cognitive performance domain and stress - experimental and data-driven approach |
| Advisors: | Lee, Minhyun (BRE) Seo, Joonoh (BRE) |
| Degree: | Ph.D. |
| Year: | 2026 |
| Department: | Department of Building and Real Estate |
| Pages: | xxvi, 385 pages : color illustrations |
| Language: | English |
| Abstract: | This thesis investigates how thermal environments and task difficulty jointly shape human thermal sensation, thermal satisfaction, task performance, workload perception, and stress level, and develops a multimodal predictive framework for modelling such interactions. The study responds to a long-standing gap in thermal-comfort research—its limited integration of cognitive and affective processes—by combining controlled experimentation with statistical and machine-learning approaches to reveal the mechanisms of human adaptation under concurrent thermal and mental load. Because the thermal experience of office occupants is influenced not only by temperature but also by cognitive demand, the research employed a controlled laboratory design with three PMV-defined thermal states (slightly cool, neutral, slightly warm) and three task-difficulty levels (easy, moderate, hard). Thirty-two subjects conducted computer-based cognitive tasks while their subjective ratings, physiological indicators (skin temperature, electrodermal activity, heart rate), and thermal-facial images were recorded. Statistical analyses using Friedman tests and post-hoc comparisons assessed the significance and interaction of environmental and cognitive effects, whereas a progressive modelling pipeline—ranging from single-target algorithms to multimodal CNN-MLP architectures—translated these empirical findings into predictive form. The results revealed clear main and interaction effects between thermal environment and task difficulty. Higher task demand increased workload and stress while moderating thermal sensitivity; neutral temperature conditions produced the best overall performance and well-being. Comfort, workload, and stress were strongly correlated, confirming that thermal and cognitive stressors operate through a shared adaptive system. Predictive modelling further demonstrated that task-related variables and physiological responses substantially enhance the explainability of comfort and performance. While early multi-output models captured inter-response relationships with limited precision, the inclusion of physiological parameters greatly improved stability, and the final CNN-MLP multimodal fusion model, incorporating thermal-image features, achieved the highest accuracy and generalisability. This model effectively represented the intertwined variations among comfort, performance, and psychological strain. The findings make three principal contributions. Theoretically, they extend conventional thermal-comfort theory by establishing thermal-cognitive interaction as a dynamic regulatory process: comfort is not a static physical equilibrium but a context-dependent balance among environmental, cognitive, and affective factors. Methodologically, they demonstrate the feasibility of combining non-parametric statistics with multimodal deep learning to link descriptive and predictive perspectives on human responses. Practically, the study offers a foundation for human-centred intelligent environmental control, enabling adaptive thermal management that aligns energy efficiency with occupant well-being. In conclusion, this research reframes thermal comfort as a multidimensional, embodied, and cognitively moderated phenomenon. It provides both conceptual and computational evidence that human performance and mental state are inseparable from thermoregulatory adaptation. By bridging environmental science, cognitive psychology, and data-driven analytics, the thesis contributes an integrated understanding of how people experience, evaluate, and adapt to their thermal surroundings—paving the way for adaptive, empathetic, and sustainable built environments. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14574