Machine learning paradigm for structural health monitoring

2020 ◽  
pp. 147592172097241
Author(s):  
Yuequan Bao ◽  
Hui Li
Keyword(s):  
Machine Learning ◽  
Structural Health Monitoring ◽  
Health Monitoring ◽  
Degrees Of Freedom ◽  
Monitoring Data ◽  
Learning Paradigm ◽  
Structural Health ◽  
Diagnosis And Prognosis ◽  
Cable Force ◽  
Health Diagnosis

Structural health diagnosis and prognosis is the goal of structural health monitoring. Vibration-based structural health monitoring methodology has been extensively investigated. However, the conventional vibration–based methods find it difficult to detect damages of actual structures because of a high incompleteness in the monitoring information (the number of sensors is much fewer with respect to the number of degrees of freedom of a structure), intense uncertainties in the structural conditions and monitoring systems, and coupled effects of damage and environmental actions on modal parameters. It is a truth that the performance and conditions of a structure must be embedded in the monitoring data (vehicles, wind, etc.; acceleration, displacement, cable force, strain, images, videos, etc.). Therefore, there is a need to develop completely novel structural health diagnosis and prognosis methodology based on the various monitoring data. Machine learning provides the advanced mathematical frameworks and algorithms that can help discover and model the performance and conditions of a structure through deep mining of monitoring data. Thus, machine learning takes an opportunity to establish novel machine learning paradigm for structural health diagnosis and prognosis theory termed the machine learning paradigm for structural health monitoring. This article sheds light on principles for machine learning paradigm for structural health monitoring with some examples and reviews the existing challenges and open questions in this field.

scholarly journals Structural Health Monitoring Using Machine Learning and Cumulative Absolute Velocity Features

2021 ◽  
Vol 11 (12) ◽  
pp. 5727
Author(s):  
Sifat Muin ◽  
Khalid M. Mosalam
Keyword(s):  
Machine Learning ◽  
Logistic Regression ◽  
Structural Health Monitoring ◽  
Health Monitoring ◽  
Degree Of Freedom ◽  
Absolute Velocity ◽  
Support Vector ◽  
Damage State ◽  
Structural Health ◽  
Cumulative Absolute Velocity

Machine learning (ML)-aided structural health monitoring (SHM) can rapidly evaluate the safety and integrity of the aging infrastructure following an earthquake. The conventional damage features used in ML-based SHM methodologies face the curse of dimensionality. This paper introduces low dimensional, namely, cumulative absolute velocity (CAV)-based features, to enable the use of ML for rapid damage assessment. A computer experiment is performed to identify the appropriate features and the ML algorithm using data from a simulated single-degree-of-freedom system. A comparative analysis of five ML models (logistic regression (LR), ordinal logistic regression (OLR), artificial neural networks with 10 and 100 neurons (ANN10 and ANN100), and support vector machines (SVM)) is performed. Two test sets were used where Set-1 originated from the same distribution as the training set and Set-2 came from a different distribution. The results showed that the combination of the CAV and the relative CAV with respect to the linear response, i.e., RCAV, performed the best among the different feature combinations. Among the ML models, OLR showed good generalization capabilities when compared to SVM and ANN models. Subsequently, OLR is successfully applied to assess the damage of two numerical multi-degree of freedom (MDOF) models and an instrumented building with CAV and RCAV as features. For the MDOF models, the damage state was identified with accuracy ranging from 84% to 97% and the damage location was identified with accuracy ranging from 93% to 97.5%. The features and the OLR models successfully captured the damage information for the instrumented structure as well. The proposed methodology is capable of ensuring rapid decision-making and improving community resiliency.

Acceleration data quality assessment for bridge structural health monitoring via statistical and deep-learning approach

2021 ◽  
Author(s):  
Huaqiang Zhong ◽  
Limin Sun ◽  
José Turmo ◽  
Ye Xia
Keyword(s):  
Structural Health Monitoring ◽  
Data Quality ◽  
Health Monitoring ◽  
Structural Damage ◽  
Quality Evaluation ◽  
Evaluation Method ◽  
Operating Conditions ◽  
Monitoring Data ◽  
Statistical Characteristics ◽  
Structural Health

<p>In recent years, the safety and comfort problems of bridges are not uncommon, and the operating conditions of in-service bridges have received widespread attention. Many large-span key bridges have installed structural health monitoring systems and collected massive amounts of data. Monitoring data is the basis of structural damage identification and performance evaluation, and it is of great significance to analyze and evaluate its quality. This paper takes the acceleration monitoring data of the main girder and arch rib of a long-span arch bridge as the research object, analyzes and summarizes the statistical characteristics of the data, summarizes 6 abnormal data conditions, and proposes a data quality evaluation method of convolutional neural network. This paper conducts frequency statistics on the acceleration vibration amplitude of the bridge in December 2018 in hours. In order to highlight the end effect of frequency statistics, the whole is amplified and used as network input for training and data quality evaluation. The results are good. It provides another new method for structural monitoring data quality evaluation and abnormal data elimination.</p>

Sign in / Sign up

Export Citation Format

Share Document