Methods for knowledge-based systems in damage assessment
Abstract
Several Knowledge based systems have been developed for damage assessment using various methods including approximate reasoning, decision making process, heuristic procedure, and automation of knowledge acquisition and reorganization. One critical issue in reasoning is to define a proper measure of the match between knowledge and data when both are represented by fuzzy sets. This is accomplished by introducing the concept of normalized fuzzy set and developing a linguistic match method for ranking fuzzy sets. Damage events are classified into three categories: damage evidence, damage cause, and damage mode. Methods for the estimation of structural reliability of multiple damage events are presented based on decision rules. A heuristic technique based on the alpha-beta procedure and possibility theory is also developed and used to manage the information process. The automation of knowledge acquisition is formalized and implemented following the concepts of conditional knowledge. In addition, a triangular model is created for knowledge reorganization. The implementation of SPERIL-3 is also discussed.
Degree
Ph.D.
Advisors
Chameau, Purdue University.
Subject Area
Civil engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.