Oct
18
An advanced AI-enabled HVAC control system: Towards greener and healthier indoor facilities in Vietnam
Principal Investigators & Key Members:
Le Duy Dung
– Using mobile LiDAR sensor technology to create detailed 3D models of multi-user indoor spaces. And, utilizing physics-informed neural network (PINN) and Computational Fluid Dynamics (CFD) techniques to model the airflow and heat transfer within the spaces.
– Implementing sensors to continuously monitor air quality parameters such as temperature, humidity, and particulate matter.
– Developing an artificial intelligence-based management system that dynamically adjusts HVAC settings based on real-time measurement data.
– Assessing the practical effectiveness and efficiency of the V-IndoorSMART framework in controlled- and real-world indoor spaces, focusing on energy conservation and air quality maintenance.
WP1: Establishment of instruments, data infrastructure, and software: Purchase and install sensors; Establish centralized data repository to store and manage data; Develop user-friendly interface.
WP2: Implementation of 3D indoor space modelings: Develop tool to leverage mobile LiDAR; Create computational fluid dynamics CFD model: Establish direct API connections between simulation software, 3D modeling tools, and virtual platform.
WP3: Designing an AI-based HVAC management system: Objective quantification; Reinforced Learning (RL) problem formulation; Iteratively Learning and Model Training the RL model.
WP4: Validation of the whole system: Lab-scale evaluation in controlled settings; Onsite evaluations in actual operational settings;