Foundations of Population-based SHM, Part II: Heterogeneous populations – Graphs, networks, and communities

This paper is the second in a series of three which aims to provide basis for Population-Based Structural Health Monitoring (PBSHM); new technology will allow transfer diagnostic information across population structures, augmenting SHM capability beyond that applicable individual structures. The PBSHM can potentially knowledge about normal operating conditions, damage states, and even physics-based models be transferred between first part this considered homogeneous populations nominally-identical theory extended heterogeneous disparate In order achieve aim, introduces an abstract representation structures based on Irreducible Element (IE) models, capture essential structural characteristics, are then converted into Attributed Graphs (AGs). AGs form complex network structure metric used assess similarity; similarity being key measure whether successfully transferred. Once pairwise has been established similar clustered communities. Within these communities, it assumed certain level possible. itself accomplished using machine learning methods discussed third series. ideas introduced define precise terminology both cases.