Date of Award

3-25-2026

Thesis Type

PhD

Document Type

Thesis

Divisions

Faculty of Built Environment

Department

Department of Building Surveying

Institution

Universiti Malaya

Abstract

Air conditioning mechanical ventilation (ACMV) supply parameters—such as airflow rate, temperature, and humidity—significantly influence indoor thermal and hygrometric conditions, influencing airflow patterns and mould proliferation. In healthcare facilities, mould poses significant health risks, particularly for immunocompromised patients, contributing to respiratory issues, allergic reactions, and infections. While ACMV systems are central to infection control and indoor air quality, conventional designs often fall short in hot and humid climates. Existing practices typically prioritise high air change rates (ACH) without sufficient dehumidification, underestimate humidity control, and assume spatial homogeneity, overlooking microclimates conducive to localised mould growth. To address these limitations, this research develops an integrated optimization framework that combines predictive analytics, adaptive control strategies, and Multiphysics modelling. Machine learning-based prediction models, trained on field measurements and publicly available datasets, capture spatiotemporal variations in temperature, humidity, and airflow to identify high-risk mould zones. Adaptive strategies dynamically optimise ACMV parameters—such as airflow velocity, temperature, and humidity—to enhance indoor environmental quality. The framework comprises four complementary components: systematic literature review, on-site environmental monitoring, mathematical modelling, and physical simulation. This multifaceted approach addresses the nonlinear interdependencies between ACMV settings and mould risks. Validation using computational fluid dynamics (CFD) confirms the framework’s robustness across varied scenarios, including extreme room configurations and airflow conditions. Even under challenging boundary conditions, the predictive model maintains strong accuracy (R2 > 0.9) in mapping parameter distributions. Compared to conventional ACMV optimisation techniques, the proposed data-driven framework enables real-time, intelligent ACMV control, offering precise environmental regulation in healthcare settings.

Initial

khm

Additional Information

Thesis (PhD) – Faculty of Built Environment, Universiti Malaya, 2026.

Download

Share

COinS