Department of Civil and Structural Engineering,
Hong Kong Polytechnic University,
Hung Hom,
Kowloon,
Hong Kong
Abstract
The modelling of unsteady open channel flow, which is an important area in fluvial engineering, involves a complicated heuristic decision-making process. Numerical modelling is an economical and powerful technique being widely used nowadays. However, engineers usually face many obstacles in the application of conventional numerical software, due to the complex process of task definition, model calibration and result evaluation involved. Expert system, an active subject of artificial intelligence, is suitable to treat the above enginnering problem. This paper describes the methodology of knowledge acquisition and representation for fluvial hydodynamics, which play a key role in the development of a practical knowledge based system. The experience gained through the knowlegde acquisition process in this project is discussed. The construction of the knowledge bases for the development of a microcomputer-based expert system for unsteady open channel flow in a river network is also presented.