A Theoretical Study on Knowledge Discovery Technique for Structural Health Monitoring

Abstract

Based on the similarity between knowledge discovery from data bases (KDD) and Structural health monitoring (SHM), and considered the particularity of SHM problems, a four-step framework of SHM is proposed. The framework extends the final goal of SHM from detecting damages to extracting knowledge to facilitate decision making. The purposes and proper methods of each step of this framework are discussed. To demonstrate the proposed SHM framework, a specific SHM method which is consisted by second order structural parameter identification as feature extraction and statistical control chart analysis of identified stiffness for feature analysis is then presented. Through clarifying the goal and hierarchy of extracting useful knowledge of SHM problems, the framework has potential to facilitate the further development of SHM.