2020
DOI: 10.1504/ijmri.2020.111798
|View full text |Cite
|
Sign up to set email alerts
|

Performance assessment of a bio-inspired anomaly detection algorithm for unsupervised SHM: application to a Manueline masonry church

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
4
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
2
1
1

Relationship

1
3

Authors

Journals

citations
Cited by 4 publications
(4 citation statements)
references
References 0 publications
0
4
0
Order By: Relevance
“…The classifier is a function trained through a dataset from the reference state of the system. Among the algorithms suitable to address such problems, the authors have recently investigated the applicability of Negative Selection Algorithms (NSAs), developing a deterministic generation based version that analyses the evolution of pairs of features at the same time [7][8][9]. The feature selection strategy that will be described in Section 2 is tailored to this new version of the NSA and it is based on the pairwise correlation of structural properties between themselves and with non-structural factors.…”
Section: Introductionmentioning
confidence: 99%
See 1 more Smart Citation
“…The classifier is a function trained through a dataset from the reference state of the system. Among the algorithms suitable to address such problems, the authors have recently investigated the applicability of Negative Selection Algorithms (NSAs), developing a deterministic generation based version that analyses the evolution of pairs of features at the same time [7][8][9]. The feature selection strategy that will be described in Section 2 is tailored to this new version of the NSA and it is based on the pairwise correlation of structural properties between themselves and with non-structural factors.…”
Section: Introductionmentioning
confidence: 99%
“…To establish whether or not the sample set is large enough, irrespective of the goals of the monitoring, specific performance metrics are tested in Section 3, to rate the approximation. Given the necessity of a large validating dataset, the tests presented hereafter are carried out by using quite a unique data set collected from the long-term monitoring of the Church of the Jerónimos Monastery in Lisbon [10,11], which have also benchmarked a previous experience of the authors of this work [8]. An extensive discussion of the main results is provided in Section 4 with focus on the integration of all the proposed strategies into a single methodology and on the analysis of the potential advantages for stakeholders operating in the civil engineering SHM field.…”
Section: Introductionmentioning
confidence: 99%
“…In fact, due to the inability of commonly used analysis techniques to consider several factors and quantify uncertainties related to the analysis, longer monitoring periods are often required to establish a satisfactory level of confidence on resulting conclusions. Although recent research has focused more on data analysis for dynamic monitoring Barontini et al, 2018;Khoa et al, 2018;Mario Azzara et al, 2018;, presumably due to the fact that this monitoring strategy enables the extraction of useful information about the structure as a whole in a short time period, it must be said that masonry heritage structures are most often affected by slow ongoing deterioration mechanisms that are not easily identifiable. As such, static monitoring appears to be particularly appealing.…”
Section: Static Shm For Masonry Constructionsmentioning
confidence: 99%
“…In fact, this is also a requirement of many other sophisticated analysis methods that have been applied to damage detection in the presence of environmental variability from dynamic SHM data. These include negative selection (Barontini et al, 2018), other machine learning techniques (Worden et al, 2007;Khoa et al, 2018) as well as those based on linear and kernel principal component analysis (PCA) (Reynders;Mario Azzara et al, 2018; or on the Mahalanobis squared-distance . This is a difficult requirement when it comes to static SHM systems for masonry heritage structures since the damage phenomena of interest very often relate to very slow and long processes which have begun long before any decision on monitoring could be taken.…”
Section: Combined With Dynamic Tests?mentioning
confidence: 99%