Case-based Reasoning for Structural Health Monitoring

Bernhard Freudenthaler, Reinhard Stumptner, Ernst Forstner, Josef Küng

Research output: Chapter in Book/Report/Conference proceedingConference proceedingspeer-review

Abstract

Case-based Reasoning (CBR) is a cyclic problem solving process whereby knowledge is stored in form of cases which consist of a problem and an appropriate solution. The main objectives of CBR are to avoid the development of new solutions for new problems, to reuse solutions of similar problems and consequently to create rapid and cost-effective solutions. The industrial sector of Structural Health Monitoring (SHM) is gaining in importance whereby safety assessment and life-time prediction are main obstacles. Numerous structures are monitored by now what produces a flood of measurement data. The manual interpretation of this data done by engineers is becoming too time-consuming and raises a strong need for automation what should be fulfilled by a Case-based System. Case-based Reasoning for Structural Health Monitoring can be used to interpret measuring data for periodic measurement relying on measuring data of similar structures. A Case-based-System can compare these actual measuring data with former measuring data. It should be possible to predict risk levels and suggest essential actions which are based on the solutions of former such cases. Another opportunity of CBR for SHM is the provision of information to support structure design where past measuring data can be used as an indication for the design of future structures. Finally one can use Case-based Reasoning for an integrated alert system for permanent monitoring where historic measuring data can be used for the interpretation of future measuring data. Partly automated interpretation of measurement data of permanently monitored structures may be the most urgent need of SHM industry.
Original languageEnglish
Title of host publicationFourth Europen Workshop on Structural Health Monitoring 2008 - EWSHM 2008
Number of pages9
Publication statusPublished - 2008

Fields of science

  • 102 Computer Sciences
  • 102001 Artificial intelligence

Cite this