Foundations of population-based SHM, Part III: Heterogeneous populations – Mapping and transfer

Gardner, Paul A.; Bull, Larry; Gosliga, Julian; Dervilis, Nikolaos; Worden, Keith

doi:10.1016/j.ymssp.2020.107142

Cited by 102 publications

(92 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to clarify these questions and their potential answers, it will be useful to introduce some terminology to guide the reader through the remainder of the paper without the mathematical complexities that can be found in the general framework described in Bull et al (2020a), Gosliga et al (2020a), and Gardner et al (2020a).…”

Section: The Potential Advantages Of Pbshm; Why It Is Neededmentioning

confidence: 99%

“…The basic concepts in the framework will be summarized here in a compact and comprehensive manner. Following the first steps discussed in the last section, many of the ideas were published in Bull et al (2020b), Gosliga et al (2020b), Gosliga et al (2020c), Gardner and Worden (2020), Worden (2020), and Lin et al (2020), and were then consolidated into (Bull et al, 2020a;Gardner et al, 2020a;Gosliga et al, 2020a).…”

Section: A General Framework For Pbshmmentioning

confidence: 99%

“…One approach that is used, and will be the one illustrated here, is Joint Domain Adaption (JDA) (Pan and Yang, 2009;Pan et al, 2010;Long et al, 2013;Wang et al, 2019;Gardner et al, 2020a,b). A number of other algorithms are applicable to SHM algorithms, and some of the others are illustrated in Gardner et al (2020b) and Gardner et al (2020a). JDA is one of the techniques that assumes the joint distributions between feature and label spaces are different; it works by mapping the source and target features into a latent space where the distributions of data are harmonized.…”

Section: Heterogeneous Populationsmentioning

confidence: 99%

“…While SS did provide some useful ideas relating to spatio-temporal novelty detection (Hensman et al, 2011), it did not provide a rich enough framework for diagnostics in general populations of structures. Such a general framework was first outlined in the sequence of papers (Bull et al, 2020b;Gardner and Worden, 2020;Gosliga et al, 2020b,c;Lin et al, 2020;Worden, 2020), which were subsequently refined and consolidated into the series: (Bull et al, 2020a;Gardner et al, 2020a;Gosliga et al, 2020a). These papers pointed out that there are differing requirements for population-based SHM, depending on whether the populations are homogenous (i.e., composed of nominallyidentical structures) or heterogeneous.…”

Section: Introductionmentioning

confidence: 99%

“…To reiterate, the aim of this short paper is not to give an overview of SHM in general, or even PBSHM in particular, as there is not enough space; more extensive and detailed discussions can be found in Bull et al (2020a), Gosliga et al (2020a), and Gardner et al (2020a), including a range of challenging case studies and experimental investigations. Furthermore, the emphasis of the work presented here is on the general concepts, emphasizing and illustrating the practical benefits of PBSHM, explaining why it is needed in non-expert terms as far as possible.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A Brief Introduction to Recent Developments in Population-Based Structural Health Monitoring

Worden

Bull

Gardner

et al. 2020

Front. Built Environ.

View full text Add to dashboard Cite

Section: The Potential Advantages Of Pbshm; Why It Is Neededmentioning

confidence: 99%

Section: A General Framework For Pbshmmentioning

confidence: 99%

Section: Heterogeneous Populationsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A Brief Introduction to Recent Developments in Population-Based Structural Health Monitoring

Worden

Bull

Gardner

et al. 2020

Front. Built Environ.

View full text Add to dashboard Cite

On the use of domain adaptation techniques for bridge damage detection in a changing environment

Giglioni,

Poole,

Venanzi

et al. 2023

ce papers

View full text Add to dashboard Cite

Structural Health Monitoring of civil infrastructures often suffers from the limited availability of damage labelled data. The work here seeks to overcome this issue by using Transfer Learning approaches, in the form of Domain Adaptation, for leveraging information from a source structure with determined health‐state labels to make inferences on an unlabeled monitored structure. The idea is to exploit source data to train a Machine Learning algorithm and achieve improved early‐stage damage detection capabilities across a population of bridges. To account for differences in the underlying distributions of each structure, Transfer Learning is seen as a strategy enabling population‐level bridge SHM. In this paper, the natural frequencies obtained from multiple vibration measurements are extracted to characterise different domains during pristine and abnormal conditions. Such damage‐sensitive features are aligned via Domain Adaptation and used to train a standard classifier within a shared feature space. The methodology is validated on the heterogeneous population composed of the Z24 and S101 bridges. The results prove the effectiveness to successfully exchange damage labels, thus increasing available information for health‐state inference for SHM applications with sparce datasets.

show abstract

Challenges for SHM from Structural Repairs: An Outlier-Informed Domain Adaptation Approach

Gardner

Bull

Dervilis

et al. 2021

Data Science in Engineering, Volume 9

Self Cite

View full text Add to dashboard Cite

Data-based approaches to structural health monitoring are typically constructed on the assumption that the underlying data distributions in training will be the same as those experienced when the method is deployed. However, structural repairs alter the physical properties of the system, leading to a change in structural response. This change in response leads to a shift in the data distributions from the pre-to post-repair states -known as domain shift -invalidating the assumption that training, and subsequent operational data, come from the same underlying distribution. As a result, structural repairs represent a significant challenge to data-based approaches to structural health monitoring (SHM). Not only will domain shift cause an algorithm trained on the pre-repair data to fail to generalise, it will also make labels acquired from the pre-repair state redundant for building conventional data-based methods on the post-repair data. Transfer learning, in the form of domain adaptation, provides a solution to this problem, allowing knowledge from the pre-repair labels to be transferred to the post-repair dataset by forming a shared latent space where the pre-and post-repair dataset distributions are approximately equal. This paper presents a novel modification of a domain adaptation technique -joint domain adaptation -in creating outlier-informed joint domain adaptation, which can be used in transferring knowledge from pre-to post-repair states, forming a post-repair classifier that utilises all the pre-repair knowledge and generalises to post-repair data. The algorithm is demonstrated on an experimental dataset from a Gnat aircraft wing, where it is shown to outperform conventional data-based approaches and existing domain adaptation techniques.

show abstract

Foundations of population-based SHM, Part III: Heterogeneous populations – Mapping and transfer

Cited by 102 publications

References 19 publications

A Brief Introduction to Recent Developments in Population-Based Structural Health Monitoring

A Brief Introduction to Recent Developments in Population-Based Structural Health Monitoring

On the use of domain adaptation techniques for bridge damage detection in a changing environment

Challenges for SHM from Structural Repairs: An Outlier-Informed Domain Adaptation Approach

Contact Info

Product

Resources

About