Damage detection using vector auto-regressive models

Zongming Huang; Gang Liu; Michael Todd; Zhu Mao

doi:10.1117/12.2012248

17 April 2013 Damage detection using vector auto-regressive models

Zongming Huang, Gang Liu, Michael Todd, Zhu Mao

Proceedings Volume 8695, Health Monitoring of Structural and Biological Systems 2013; 86953E (2013) https://doi.org/10.1117/12.2012248
Event: SPIE Smart Structures and Materials + Nondestructive Evaluation and Health Monitoring, 2013, San Diego, California, United States

Abstract

This study presents a damage detection method for transmission towers based on vector auto-regressive (VAR) models. The vibration signals obtained from both baseline and unknown conditions of the structure are divided into multiple data segments, respectively, and each segment is then modeled as a VAR time series. The diagonal elements of the VAR coefficient matrices series are extracted, and that vector’s Mahalanobis distance (MD) is used as a damage-sensitive feature. At the sensor locations where damage is introduced, the mean and variance of MD distribution will change from their values under baseline condition. Thus, the area under a receiver operating characteristic (ROC) curve and deflection coefficient of MD distribution are used as the decision metric for damage detection, localization, and severity. The method’s effectiveness is assessed on a 6 degree-of-freedom mass-spring simulation system and a transmission tower model. The results confirm the high potential and effectiveness of this method for data-driven damage assessment.

Citation Download Citation

Zongming Huang, Gang Liu, Michael Todd, and Zhu Mao "Damage detection using vector auto-regressive models", Proc. SPIE 8695, Health Monitoring of Structural and Biological Systems 2013, 86953E (17 April 2013); https://doi.org/10.1117/12.2012248

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available