On calculating structural similarity metrics in population-based structural health monitoring

The newly introduced discipline of Population-Based Structural Health Monitoring (PBSHM) has been developed in order to circumvent the issue of data scarcity in “classical” SHM. PBSHM does this by using data across an entire population, in order to improve diagnostics for a single data-poor structur...

Full description

Saved in:
Bibliographic Details
Main Authors: Daniel S. Brennan, Timothy J. Rogers, Elizabeth J. Cross, Keith Worden
Format: Article
Language:English
Published: Cambridge University Press 2025-01-01
Series:Data-Centric Engineering
Subjects:
Online Access:https://www.cambridge.org/core/product/identifier/S2632673624000455/type/journal_article
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849703862035808256
author Daniel S. Brennan
Timothy J. Rogers
Elizabeth J. Cross
Keith Worden
author_facet Daniel S. Brennan
Timothy J. Rogers
Elizabeth J. Cross
Keith Worden
author_sort Daniel S. Brennan
collection DOAJ
description The newly introduced discipline of Population-Based Structural Health Monitoring (PBSHM) has been developed in order to circumvent the issue of data scarcity in “classical” SHM. PBSHM does this by using data across an entire population, in order to improve diagnostics for a single data-poor structure. The improvement of inferences across populations uses the machine-learning technology of transfer learning. In order that transfer makes matters better, rather than worse, PBSHM assesses the similarity of structures and only transfers if a threshold of similarity is reached. The similarity measures are implemented by embedding structures as models —Irreducible-Element (IE) models— in a graph space. The problem with this approach is that the construction of IE models is subjective and can suffer from author-bias, which may induce dissimilarity where there is none. This paper proposes that IE-models be transformed to a canonical form through reduction rules, in which possible sources of ambiguity have been removed. Furthermore, in order that other variations —outside the control of the modeller— are correctly dealt with, the paper introduces the idea of a reality model, which encodes details of the environment and operation of the structure. Finally, the effects of the canonical form on similarity assessments are investigated via a numerical population study. A final novelty of the paper is in the implementation of a neural-network-based similarity measure, which learns reduction rules from data; the results with the new graph-matching network (GMN) are compared with a previous approach based on the Jaccard index, from pure graph theory.
format Article
id doaj-art-5f18d918b4ff45ebbb812c5a11da1ff7
institution DOAJ
issn 2632-6736
language English
publishDate 2025-01-01
publisher Cambridge University Press
record_format Article
series Data-Centric Engineering
spelling doaj-art-5f18d918b4ff45ebbb812c5a11da1ff72025-08-20T03:17:03ZengCambridge University PressData-Centric Engineering2632-67362025-01-01610.1017/dce.2024.45On calculating structural similarity metrics in population-based structural health monitoringDaniel S. Brennan0https://orcid.org/0009-0001-7223-3431Timothy J. Rogers1https://orcid.org/0000-0002-3433-3247Elizabeth J. Cross2https://orcid.org/0000-0001-5204-1910Keith Worden3Dynamics Research Group, Department of Mechanical Engineering, University of Sheffield, Sheffield, UKDynamics Research Group, Department of Mechanical Engineering, University of Sheffield, Sheffield, UKDynamics Research Group, Department of Mechanical Engineering, University of Sheffield, Sheffield, UKDynamics Research Group, Department of Mechanical Engineering, University of Sheffield, Sheffield, UKThe newly introduced discipline of Population-Based Structural Health Monitoring (PBSHM) has been developed in order to circumvent the issue of data scarcity in “classical” SHM. PBSHM does this by using data across an entire population, in order to improve diagnostics for a single data-poor structure. The improvement of inferences across populations uses the machine-learning technology of transfer learning. In order that transfer makes matters better, rather than worse, PBSHM assesses the similarity of structures and only transfers if a threshold of similarity is reached. The similarity measures are implemented by embedding structures as models —Irreducible-Element (IE) models— in a graph space. The problem with this approach is that the construction of IE models is subjective and can suffer from author-bias, which may induce dissimilarity where there is none. This paper proposes that IE-models be transformed to a canonical form through reduction rules, in which possible sources of ambiguity have been removed. Furthermore, in order that other variations —outside the control of the modeller— are correctly dealt with, the paper introduces the idea of a reality model, which encodes details of the environment and operation of the structure. Finally, the effects of the canonical form on similarity assessments are investigated via a numerical population study. A final novelty of the paper is in the implementation of a neural-network-based similarity measure, which learns reduction rules from data; the results with the new graph-matching network (GMN) are compared with a previous approach based on the Jaccard index, from pure graph theory.https://www.cambridge.org/core/product/identifier/S2632673624000455/type/journal_articleCanonical formgraph-matching networkirreducible-element modelsJaccard indexpopulation-based SHM
spellingShingle Daniel S. Brennan
Timothy J. Rogers
Elizabeth J. Cross
Keith Worden
On calculating structural similarity metrics in population-based structural health monitoring
Data-Centric Engineering
Canonical form
graph-matching network
irreducible-element models
Jaccard index
population-based SHM
title On calculating structural similarity metrics in population-based structural health monitoring
title_full On calculating structural similarity metrics in population-based structural health monitoring
title_fullStr On calculating structural similarity metrics in population-based structural health monitoring
title_full_unstemmed On calculating structural similarity metrics in population-based structural health monitoring
title_short On calculating structural similarity metrics in population-based structural health monitoring
title_sort on calculating structural similarity metrics in population based structural health monitoring
topic Canonical form
graph-matching network
irreducible-element models
Jaccard index
population-based SHM
url https://www.cambridge.org/core/product/identifier/S2632673624000455/type/journal_article
work_keys_str_mv AT danielsbrennan oncalculatingstructuralsimilaritymetricsinpopulationbasedstructuralhealthmonitoring
AT timothyjrogers oncalculatingstructuralsimilaritymetricsinpopulationbasedstructuralhealthmonitoring
AT elizabethjcross oncalculatingstructuralsimilaritymetricsinpopulationbasedstructuralhealthmonitoring
AT keithworden oncalculatingstructuralsimilaritymetricsinpopulationbasedstructuralhealthmonitoring