This study investigates the potential applications of Knowledge-Based Expert Systems in transportation engineering. The three major topics covered are: (1) A description of the fundamental concepts of Knowledge-Based Expert Systems and how they relate to transportation engineering. The general differences between expert systems and conventional computer programs are summarized, and the process of transferring knowledge from human experts to computers is described. The architecture of the Knowledge-Based Expert Systems is shown to separate the knowledge of the problem domain (knowledge base) from general problem solving (inference engine). Different approaches to each of these tasks are described and recent applications to several transportation engineering problems are presented. (2) The selection of appropriate representation techniques for different classes of transportation engineering problems. Three of the most commonly used techniques for representing knowledge in (the knowledge base of) Knowledge-Based Expert Systems are described, and the highlights, shortcomings, and applications of each technique is explained. This section focuses on an examination of the significance of the means of representing the knowledge base in the development of Knowledge-Based Expert Systems, with special reference to five classes of transportation engineering problems--planning, operations and control, design, construction and maintenance, and management and finance. Under each class, one or two transportation and traffic problems, suitable for Knowledge-Based Expert Systems development are briefly described, and the most suitable representation technique(s) for their development is recommended. (3) A description of the development of one transportation engineering problem, chosen from the previous section. The stages required in the development of a Knowledge-Based Expert System--problem identification, conceptual analysis, formalization, implementation, and evaluation--for the chosen problem are described in detail.
展开▼
