AN EXPERT SYSTEM USING CASE-BASED REASONING FOR SELECTING RETROFITTING METHODS OF FATIGUE DAMAGE ON STEEL BRIDGES

The maintenance and administration of existing civil engineering structures are becoming important subjects, and engineers with experience are required for them. Therefore, expert systems may be effectively used in this field. A diagnostic knowledge-based expert system has been developed by the authors for the maintenance and administration of steel bridge. Repair and retrofit methods have been selected for fatigue damage of steel bridges. The systems involve a knowledge-base and an inference engine. Firstly, the knowledge included in the knowledge-base was obtained from 165 cases of fatigue damage. The production rules, causal relations, in the knowledge-base were assumed to have a certainty factor. Secondly, the inference engine has the function of machine learning. But, the expert systems using causal relations have a difficulty of acquiring the knowledge-base. A case-based reasoning mechanism is able to solve the problem of knowledge acquisition. In the present paper, an expert system is built using case-based reasoning to select the repair and retrofit methods of fatigue damage. This system has a case-based database of 205 past cases, and 4 kinds of inference processes (retrieval, modification, restoration, and storing). The inference results obtained from this system which apply to actual cases are discussed. (A) For the covering abstract see IRRD 882313.

Language

  • English

Media Info

  • Features: References;
  • Pagination: p. 17-25

Subject/Index Terms

Filing Info

  • Accession Number: 00729452
  • Record Type: Publication
  • Source Agency: Transport Research Laboratory
  • ISBN: 0-419-21210-8
  • Files: ITRD
  • Created Date: Dec 26 1996 12:00AM