A Comparison of Structural Similarity Metrics within Population-Based Structural Health Monitoring

Abstract

Population-based Structural Health Monitoring (PBSHM) aims to gain additional insights into the health of a structure when using data available across a population of similar structures, as compared to the insight available when using only data from a single structure. Before knowledge can be transferred across structures, the similarity between structures (or substructures), within the population must be established. As a result of recent developments within PBSHM, there are now several methodologies for the comparison of structures, including Graph Matching Networks (GMNs) and the Jaccard index. The recent work on GMNs has highlighted how different methodologies may be required for different structure types or populations. This paper looks at these aforementioned algorithms and evaluates the performance of each algorithm against different toy and real-world datasets and determines which algorithms have the potential for being used within which scenarios.