Dynamic Optimisation of Air Handling Units (AHU) using Physics-guided Machine Learning

I&T Solution Dynamic Optimisation of Air Handling Units (AHU) using Physics-guided Machine Learning
(REF : S-0567)
Solution Feature
  • Dynamic optimisation of AHU model is built upon the basis of self-trained physics-guided machine learning (PGML) models using actual operating data
  • Optimal control settings for AHUs are determined in real-time based on outdoor air conditions, cooling load as well as occupancy level
  • Optimal control settings include fresh air/return air damper position, supply air temperature set point and cooling coil valve position
  • Level of occupancy can be determined using a wide range of approaches, including CCTV facial recognition and radar technologies
Trial Application and Expected Outcome
  • Model calibration is conducted using actual operating data to ensure AHU optimisation performance
  • Based on the real-time data including outdoor air conditions, cooling load as well as occupancy level, the optimisation search is performed to determine optimal control settings
  • Optimal control settings are sent back to the building management system (BMS) for optimisation
  • Key indicators, including levels of thermal comfort, CO2 and energy reduction, are established to evaluate the effectiveness of a project
Additional Solution Information EMSD_AI FA Energy Saving Control with People Counting.pptx
Info on I&T Solution Provider
Solution Provider:ATAL Building Services Engineering Ltd
Address:13/F Island Place Tower 510 King's Road North Point Hong Kong
Contact Person:Yeung Yau Kwan, Edmond
Position:Senior Sales Manager
Tel: 2565 3478
Email: edmondyeung@atal.com

If you have interest in trial application of the I&T solutions, please click HERE.