Back to search

University of British Columbia

Robust statistical subspace-based damage assessment

Abstract

dc:description

The rapid spread and increasing affordability of sensors, are encouraging the government and stake-holders to instrument important infra-structures and structures. These sensors generate vast amount of data which can be used in real-time health monitoring of the instrumented structures by using damage identification methods. A significant component of structural health monitoring is damage identification methods which process the data with the purpose of detecting damages in the structures. One of these methods with a theoretical background is the statistical subspace damage identification method (SSDI). The overarching goal in this thesis is to close the gap between theory and practice, in order to have a method with a strong theoretical background and a credible applicability at the same time. In order to achieve this goal several contributions are motivated in this thesis, which are presented as follows: Firstly, the effect of two challenges faced in the damage detection of structures under real test conditions, namely the measurement noise and duration (length), are theoretically evaluated. It is demonstrated that the measurement noise and length have considerable influence on the statistical subspace damage detection method and they need to be considered based on these proposed theories. Secondly, the statistical subspace damage localization (SSDL) method, is assessed for the first time, in localizing the damage of a real experimental structure, i.e. the Yellow frame, established on the course of this research at UBC. Several methods and theories are developed in order to enable this method in identifying the damage under real test conditions. It was demonstrated that by employing the proposed theories, the SSDL method can robustly locate the damage in a real structure such as the Yellow frame. Finally, two indexes are proposed in predicting the detectability of damage in each element of a structure. These indexes provide valuable information on the sensitivity of SSDL method to the damage in each element. All the proposed theories and methods are demonstrated theoretically; subsequently, they are verified by simple and sophisticated analytical models, and finally, they are validated by real-test data.

Degree

thesis:*
Name thesis:degree_name
Doctor of Philosophy - PhD
Level thesis:degree_level
doctoral
Discipline thesis:degree_discipline
Civil Engineering
Grantor dc:publisher
University of British Columbia
Year dc:date
2017

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Allahdadian, Saeid

Rights

dc:rights
Statement dc:rights
  • Attribution-NonCommercial-NoDerivatives 4.0 International
Language dc:language
eng

Identifiers

dc:identifier.*
Handle dc:identifier
http://hdl.handle.net/2429/62148

Chain of custody

source
Harvested from
University of British Columbia
Base URL
circle.library.ubc.ca/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Allahdadian, Saeid. Robust statistical subspace-based damage assessment. doctoral thesis, University of British Columbia, 2017. http://hdl.handle.net/2429/62148