This dissertation represents a significant advancement of indoor environmental control, focusing on developing a Stratified Micro-Environment (SME) system designed to cater to differentiated thermal comfort and Indoor Air Quality (IAQ) needs while optimizing energy efficiency. The research is underpinned by the recognition that individuals within the same indoor space can experience a wide range of thermal preferences due to age, gender, health status, cultural background, etc. The traditional model of centralized ventilation, which often provides a uniform thermal environment, is ill-suited to address this diversity, leading to discomfort and inefficiencies. The proposed SME system aims to revolutionize this paradigm by offering a flexible, adaptable, and energy-efficient solution that aligns with individual needs without compromising the communal nature of shared spaces.

The research methodology is robust and multifaceted, incorporating controlled experimental studies, Computational Fluid Dynamics (CFD) simulations, jet theory analysis, and the development of predictive models using both empirical and data-driven approaches. The experimental data serves as a critical benchmark for validating the CFD models, which are then employed to simulate airflow patterns and contaminant dispersion within the indoor environment. The application of jet theory is particularly insightful for understanding the behaviour of air jets, which is essential for designing ventilation systems capable of creating stratified environments. The research also leverages Artificial Neural Networks (ANN) and Support Vector Machines (SVM) to develop predictive models that accurately forecast thermal comfort and IAQ indicators.

A major focus is to establish stratified environments with sidewall air supply in a consultation room. The study evaluates the thermal comfort, contaminant removal efficiency, and energy utilization performance under various supply configurations, including single-sided and double-sided air supply. The findings suggest that sidewall air supply can effectively achieve Stratum Ventilation (SV) and Displacement Ventilation (DV), offering a viable solution for accommodating individual thermal preferences and improving IAQ while enhancing energy efficiency.

The study extends to the ANN models for predicting energy use, thermal comfort, and IAQ in stratified environments. The ANN models developed demonstrate a heightened ability to capture complex, non-linear correlations between air distribution and indoor environment indicators, significantly improving over traditional regression models. Integrating a genetic algorithm further refines the ANN models, particularly for predicting air age under heating conditions.

The thesis proposes a control scheme for differentiated thermal comfort in shared spaces, utilizing Predicted Mean Vote (PMV) and Draft Rate (DR) as indicators. The empirical jet theory models, along with ANN and SVM models, are validated against experimental data, providing a robust foundation for developing personalized microclimate control strategies. The comparative analysis of the proposed SME system with traditional ventilation methods such as Mixing Ventilation (MV), DV, and SV reveals that the SME system significantly outperforms other methods in maintaining individual comfort levels with minimal energy consumption.

The research concludes with a discussion of the limitations and potential areas for future research. It acknowledges the specific room setup and thermal conditions as constraints that may limit the broader application of the findings. Recommendations for future research include improving the accuracy of thermal comfort models across various scenarios and conducting detailed cost analyses to understand better the financial implications of implementing the SME system. The study also suggests the need for further investigation into the impact of diverse occupant distributions and non-adiabatic walls on the performance of the SME system.

In conclusion, this dissertation presents a comprehensive and innovative approach to enhancing indoor environments through differentiated thermal comfort and IAQ management. The proposed SME system, supported by rigorous experimental and computational research, offers a promising solution for energy-efficient and occupant-centric building design. The predictive models developed in this study provide a valuable tool for HVAC design improvements, supporting precision and ease of application in real-world settings.