Control of air flow in active chilled beam duct systems using general predictive control

• Main Author: Justin Ashish Jerry
• Other Authors: Cai Wenjian
• Format: Theses and Dissertations
• Language: English
• Published: 2015
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation
• Online Access: http://hdl.handle.net/10356/64821
• Institution: Nanyang Technological University

This dissertation deals with the control of Air Flow in an Active Chilled beam duct system using General Predictive Control. Objective of the work is to implement, analyse and compare two novel GPC tuning algorithms to modulate pressure and air flow in a Duct system. An ACB was used in the study because of the inherent advantages over conventional HV AC systems and it also serves as a new domain where implementation of high level advanced control techniques such as GPC could be explored on low level controllers. GPC is a subset of Model Predictive Control which has been widely adapted and used in high level optimizers and economizers. This dissertation aims to extend the application paradigm of GPC by implementing it to low level control systems. First an ACB duct system was assembled and appropriate sensors and data acquisition devices were placed. The processes which have to be controlled were then modelled as transfer functions in their respective operating regions. Controllers were then designed and tuned using two GPC tuning strategies. Process variable and controller actions were recorded for different sampling times and different the fundamental tuning parameter values. This recorded data was then compared and analyzed. Results showed the advantages and flexibility of performance and robustness variability offered by N* Tuning Strategy [1] over that of Shridhar Cooper DMC [2] tuning.