Development of intelligent air-conditioning control algorithms and models
Author(s)Tse, Wai-leung (謝偉良)
Full recordShow full item record
AbstractThe objective of an air-conditioning system is to provide thermal comfort to occupants that enable them to attain maximum productivity. Besides the climatic control offered by the system, its operating cost is also a concerned factor. A good controller for an air handling unit (AHU) is always desirable from both human comfort and energy saving points of view. Quicker response rate and lower energy consumption are the required capabilities of a controller. Three kinds of new intelligent controllers were developed to cater for the above requirements. The first one was to employ a classical system identification technique in its control algorithms. As it made use of a linear model to identify the system, it was called a linear model based control. The next one was fuzzy logic based control. It was an expert system and was able to deduce an optimal control action based on experiences of human experts. The last one was the most innovative one utilizing the concept of an artificial neural network for control actions. Self-learning features were incorporated in the control to ensure continuous improvement on the response rate and energy consumption. Detailed algorithms of these controls were mentioned in Chapter 4, 5 and 6 respectively. As thermal comfort is the main target of an air-conditioning system, a completely new control concept employing human thermal comfort was developed in Chapter 7. which was called comfort-based control. Unlike other previous control, it guaranteed a comfort environment to all occupants all the time regardless of any change inside or outside the air-conditioned space. As claimed in Section 7.7 of Chapter 7, this control should operate with an ice storage plant to maximize the efficiency of the chiller plant. Thus, a weather information based intelligent control method for an ice storage plant was proposed. It had an ability to predict the cooling load profile so as to generate an optimal amount of ice for the usage on the following day. Together with smart strategies based on cloud tracking techniques, a much lower operating cost was achieved. The details of this control were presented in Chapter 8. Finally, a system-wise approach was adopted and acted on a building automation system (BAS) of a building, which was a system to manage all building facilities remotely, including an air-conditioning system using Internet technologies. It was able to further reduce the operating cost of the whole system significantly by utilizing the concept that a single control centre could monitor and control any facilities of buildings around the world simultaneously. Its details were described in Chapter 9. Through this research work, several contributions have been achieved in the field of air-conditioning control. Besides the new controllers and new control algorithms being developed, several simplified arithmetical models for the built environment and components of air-conditioning plants were developed as well. With these simplified models, tailor-made computer programs could be compiled for simulations instead of relying on expensive software package on the market. Self-compiled computer programs also offered a great advantage of flexibility to incorporate different control algorithms. Five technical papers were published in international journals and one on intelligent control of an ice storage plant is under review. Reference to these six papers can be found in the Bibliography Section.
CityU Call Number: TH7687.5.T75 2000
Includes bibliographical references (leaves 251-260).
Thesis (Ph.D.)--City University of Hong Kong, 2000
xiii, 284 leves : ill. ; 30 cm.