Diagnosis algorithms for indirect structural health monitoring of a bridge model via dimensionality reduction

Jingxiao Liu, Siheng Chen, Mario Bergés, Jacobo Bielak, James H. Garrett, Jelena Kovačević, Hae Young Noh

Research output: Contribution to journalArticle

Abstract

We present a data-driven approach based on physical insights to achieve damage diagnosis of bridges using only vibration signals collected on board the vehicles passing over the bridge. Though data-driven models have been shown to produce promising results in this context, they generally require labeled examples to fit the models (i.e., supervised learning) and make it difficult to interpret the physical mechanisms. We posit that these shortcomings can be alleviated by studying the physical relationship between damage and the distribution of the resulting acceleration signals, and then choosing an appropriate model to invert this process. To help guide the development of appropriate damage diagnosis algorithms, we first make use of the theoretical formulation of the vehicle-bridge interaction system in the frequency domain and conduct a finite element simulation of this system. From the derived numerical solution, we observe that not only is the trend of the acceleration signals of a passing vehicle with different damage severity non-linear, but also that both the low- and high-frequency responses of a passing vehicle contain information about damage severity. Guided by these observations, we use several dimensionality reduction methods to extract representative features from the vehicle's vibration response. We then propose an unsupervised damage severity comparison model and a semi-supervised damage severity estimation model aiming at indirect monitoring of bridges. We apply the algorithms to diagnose changes that occur in a laboratory bridge model to which a concentrated mass of gradually changing magnitude is attached at mid-span. The experimental results of the damage severity comparison and estimation show that a non-convex and non-linear dimensionality reduction technique (stacked autoencoders) outperforms other linear and/or convex dimensionality reduction techniques. Overall, our results provide evidence for the applicability of indirect structural health monitoring in bridge models and suggest the feasibility of extending this approach to actual structures.

Original languageEnglish (US)
Article number106454
JournalMechanical Systems and Signal Processing
Volume136
DOIs
StatePublished - Feb 2020

Fingerprint

Structural health monitoring
Supervised learning
Frequency response
Monitoring

Keywords

  • Damage diagnosis
  • Dimensionality reduction
  • Indirect SHM
  • Vehicle-bridge interaction

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Signal Processing
  • Civil and Structural Engineering
  • Aerospace Engineering
  • Mechanical Engineering
  • Computer Science Applications

Cite this

Diagnosis algorithms for indirect structural health monitoring of a bridge model via dimensionality reduction. / Liu, Jingxiao; Chen, Siheng; Bergés, Mario; Bielak, Jacobo; Garrett, James H.; Kovačević, Jelena; Noh, Hae Young.

In: Mechanical Systems and Signal Processing, Vol. 136, 106454, 02.2020.

Research output: Contribution to journalArticle

Liu, Jingxiao ; Chen, Siheng ; Bergés, Mario ; Bielak, Jacobo ; Garrett, James H. ; Kovačević, Jelena ; Noh, Hae Young. / Diagnosis algorithms for indirect structural health monitoring of a bridge model via dimensionality reduction. In: Mechanical Systems and Signal Processing. 2020 ; Vol. 136.
@article{8291061fb7ff4e2fa4ccfbc6531065e7,
title = "Diagnosis algorithms for indirect structural health monitoring of a bridge model via dimensionality reduction",
abstract = "We present a data-driven approach based on physical insights to achieve damage diagnosis of bridges using only vibration signals collected on board the vehicles passing over the bridge. Though data-driven models have been shown to produce promising results in this context, they generally require labeled examples to fit the models (i.e., supervised learning) and make it difficult to interpret the physical mechanisms. We posit that these shortcomings can be alleviated by studying the physical relationship between damage and the distribution of the resulting acceleration signals, and then choosing an appropriate model to invert this process. To help guide the development of appropriate damage diagnosis algorithms, we first make use of the theoretical formulation of the vehicle-bridge interaction system in the frequency domain and conduct a finite element simulation of this system. From the derived numerical solution, we observe that not only is the trend of the acceleration signals of a passing vehicle with different damage severity non-linear, but also that both the low- and high-frequency responses of a passing vehicle contain information about damage severity. Guided by these observations, we use several dimensionality reduction methods to extract representative features from the vehicle's vibration response. We then propose an unsupervised damage severity comparison model and a semi-supervised damage severity estimation model aiming at indirect monitoring of bridges. We apply the algorithms to diagnose changes that occur in a laboratory bridge model to which a concentrated mass of gradually changing magnitude is attached at mid-span. The experimental results of the damage severity comparison and estimation show that a non-convex and non-linear dimensionality reduction technique (stacked autoencoders) outperforms other linear and/or convex dimensionality reduction techniques. Overall, our results provide evidence for the applicability of indirect structural health monitoring in bridge models and suggest the feasibility of extending this approach to actual structures.",
keywords = "Damage diagnosis, Dimensionality reduction, Indirect SHM, Vehicle-bridge interaction",
author = "Jingxiao Liu and Siheng Chen and Mario Berg{\'e}s and Jacobo Bielak and Garrett, {James H.} and Jelena Kovačević and Noh, {Hae Young}",
year = "2020",
month = "2",
doi = "10.1016/j.ymssp.2019.106454",
language = "English (US)",
volume = "136",
journal = "Mechanical Systems and Signal Processing",
issn = "0888-3270",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - Diagnosis algorithms for indirect structural health monitoring of a bridge model via dimensionality reduction

AU - Liu, Jingxiao

AU - Chen, Siheng

AU - Bergés, Mario

AU - Bielak, Jacobo

AU - Garrett, James H.

AU - Kovačević, Jelena

AU - Noh, Hae Young

PY - 2020/2

Y1 - 2020/2

N2 - We present a data-driven approach based on physical insights to achieve damage diagnosis of bridges using only vibration signals collected on board the vehicles passing over the bridge. Though data-driven models have been shown to produce promising results in this context, they generally require labeled examples to fit the models (i.e., supervised learning) and make it difficult to interpret the physical mechanisms. We posit that these shortcomings can be alleviated by studying the physical relationship between damage and the distribution of the resulting acceleration signals, and then choosing an appropriate model to invert this process. To help guide the development of appropriate damage diagnosis algorithms, we first make use of the theoretical formulation of the vehicle-bridge interaction system in the frequency domain and conduct a finite element simulation of this system. From the derived numerical solution, we observe that not only is the trend of the acceleration signals of a passing vehicle with different damage severity non-linear, but also that both the low- and high-frequency responses of a passing vehicle contain information about damage severity. Guided by these observations, we use several dimensionality reduction methods to extract representative features from the vehicle's vibration response. We then propose an unsupervised damage severity comparison model and a semi-supervised damage severity estimation model aiming at indirect monitoring of bridges. We apply the algorithms to diagnose changes that occur in a laboratory bridge model to which a concentrated mass of gradually changing magnitude is attached at mid-span. The experimental results of the damage severity comparison and estimation show that a non-convex and non-linear dimensionality reduction technique (stacked autoencoders) outperforms other linear and/or convex dimensionality reduction techniques. Overall, our results provide evidence for the applicability of indirect structural health monitoring in bridge models and suggest the feasibility of extending this approach to actual structures.

AB - We present a data-driven approach based on physical insights to achieve damage diagnosis of bridges using only vibration signals collected on board the vehicles passing over the bridge. Though data-driven models have been shown to produce promising results in this context, they generally require labeled examples to fit the models (i.e., supervised learning) and make it difficult to interpret the physical mechanisms. We posit that these shortcomings can be alleviated by studying the physical relationship between damage and the distribution of the resulting acceleration signals, and then choosing an appropriate model to invert this process. To help guide the development of appropriate damage diagnosis algorithms, we first make use of the theoretical formulation of the vehicle-bridge interaction system in the frequency domain and conduct a finite element simulation of this system. From the derived numerical solution, we observe that not only is the trend of the acceleration signals of a passing vehicle with different damage severity non-linear, but also that both the low- and high-frequency responses of a passing vehicle contain information about damage severity. Guided by these observations, we use several dimensionality reduction methods to extract representative features from the vehicle's vibration response. We then propose an unsupervised damage severity comparison model and a semi-supervised damage severity estimation model aiming at indirect monitoring of bridges. We apply the algorithms to diagnose changes that occur in a laboratory bridge model to which a concentrated mass of gradually changing magnitude is attached at mid-span. The experimental results of the damage severity comparison and estimation show that a non-convex and non-linear dimensionality reduction technique (stacked autoencoders) outperforms other linear and/or convex dimensionality reduction techniques. Overall, our results provide evidence for the applicability of indirect structural health monitoring in bridge models and suggest the feasibility of extending this approach to actual structures.

KW - Damage diagnosis

KW - Dimensionality reduction

KW - Indirect SHM

KW - Vehicle-bridge interaction

UR - http://www.scopus.com/inward/record.url?scp=85074995744&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85074995744&partnerID=8YFLogxK

U2 - 10.1016/j.ymssp.2019.106454

DO - 10.1016/j.ymssp.2019.106454

M3 - Article

AN - SCOPUS:85074995744

VL - 136

JO - Mechanical Systems and Signal Processing

JF - Mechanical Systems and Signal Processing

SN - 0888-3270

M1 - 106454

ER -