APPLICABILITY OF INFERENCE SYSTEM WITH KNOWLEDGE REFINEMENT FUNCTION FOR SELECTING RETROFITTING METHOD FOR BRIDGES DAMAGED BY FATIGUE

The authors have proposed a knowledge refinement method for a crack diagnostic inference system. The inference engine was constructed for a reciprocal network based on a min-max composition algorithm. The knowledge refinement function was installed in the engine by applying the concept of the back-propagation algorithm. In this paper, the inference and refinement methods are applied to the rule-base system for selecting the retrofitting method for steel bridges damaged by fatigue. Through some inferences for practical cases, it is confirmed that the inference engine can be applied to this domain.