Performance assessment of a bio-inspired anomaly detection algorithm for unsupervised SHM: application to a Manueline masonry church
by Alberto Barontini; Maria Giovanna Masciotta; Paulo Amado-Mendes; Luís F. Ramos
International Journal of Masonry Research and Innovation (IJMRI), Vol. 5, No. 4, 2020

Abstract: Vibration-based techniques are commonly used in structural health monitoring (SHM) to assess the condition of structural systems and identify the presence of damage. Negative selection algorithms (NSAs) are bio-inspired methods which allow to automatise the damage detection process by classifying the monitored system's features as normal or abnormal. In this paper, an NSA with a non-random strategy for detector generation is tested on the monitoring data of a remarkable masonry church in Portugal. The work aims to make users aware of NSA potential, contributing to a diligent application of the method in terms of best algorithm instance definition. Different setting approaches for the algorithm parameters are discussed and compared, exploiting artificial outliers of the features distribution to assess the NSA performance. Such a strategy allows the optimisation of the algorithm in most of the civil engineering applications where no information about the features belonging to unhealthy scenarios is available.

Online publication date: Tue, 15-Dec-2020

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Masonry Research and Innovation (IJMRI):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com