In the conceptual design of bridges, it is a general practice to consider carefully the related laws and regulations, topographical features, environmental conditions and economical factors. Besides these considerations, expertise and experience of the designers are essential factors as well as the theoretical background in this field. In the reality in Korea, however, there exist factors which make it difficult to examine and analyse thoroughly the economic factor, the workability, the safety factor and so on. Standards and regulations are not ready which would decide the span, bridge location, superstructure type, construction method and substructure type. These make it necessary for the designers to rely heavily on the past experience and on past drawings of similar design. All these may eventually result in unstandardized bridge design, economical losses at both the designing and construction stages. Therefore it is of utmost importance to set a reasonable and comprehensive standard in the design of bridges.
This research aims to develop a rule based system called EXCON, which will be based on existing theoretical knowledge of various design criteria, mechanics of structural elements and on heuristic knowledge of experts as well. This effort to produce a consistent model of the various stages of bridge design, will prove to be significant because conceptual design of bridge involves active exchange of information between experts. By building a rule based system, a popular field of artificial intelligence, a designer will be able to check most, and sometimes all of the possible cases in rather a short period of time which would be very helpful in designing.
The EXCON is implemented with the expert system tool CLIPS that uses the production system for knowledge representation and provides the mechanism of forward chaining. The regression analysis of the bridge construction cost is specially used for the stage of the economic evaluation of the candidates of superstructures. This system adopts a certainty factor to express the ambiguity of design data at selecting the types of foundation.
