scholarly journals Automated Structural Damage Identification Using Data Normalization and 1-Dimensional Convolutional Neural Network

2021 ◽  
Vol 11 (6) ◽  
pp. 2610
Author(s):  
Jongbin Won ◽  
Jong-Woong Park ◽  
Soojin Jang ◽  
Kyohoon Jin ◽  
Youngbin Kim
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Damage Detection ◽  
Structural Damage ◽  
Input Data ◽  
Damage Identification ◽  
Beam Model ◽  
Data Normalization ◽  
Structural Damage Detection ◽  
Structural Damage Identification

In the field of structural-health monitoring, vibration-based structural damage detection techniques have been practically implemented in recent decades for structural condition assessment. With the development of deep-learning networks that make automatic feature extraction and high classification accuracy possible, deep-learning-based structural damage detection has been gaining significant attention. The deep-learning neural networks come with fixed input and output size, and input data must be downsampled or cropped to the predetermined input size of the networks to obtain desired output of the network. However, the length of input data (i.e., sensing data) is associated with the excitation quality of a structure, adjusting the size of the input data while maintaining the excitation quality is critical to ensure high accuracy of the deep-learning-based structural damage detection. To address this issue, natural-excitation-technique-based data normalization and the use of 1-D convolutional neural networks for automated structural damage detection are presented. The presented approach converts input data to predetermined size using cross-correlation and uses convolutional network to extract damage-sensitive feature for automated structural damage identification. Numerical simulations were conducted on a simply supported beam model excited by random and traffic loadings, and the performance was validated under various scenarios. The proposed method successfully detected the location of damage on a beam under random and traffic loadings with accuracies of 99.90% and 99.20%, respectively.

Vibration-based damage detection for bridges by deep convolutional denoising autoencoder

2020 ◽  
pp. 147592172094283 ◽  
Author(s):  
Zhiqiang Shang ◽  
Limin Sun ◽  
Ye Xia ◽  
Wei Zhang
Keyword(s):  
Neural Networks ◽  
Damage Detection ◽  
Correlation Functions ◽  
Structural Damage ◽  
Cross Correlation ◽  
Deep Neural Networks ◽  
Beam Model ◽  
Structural Damage Detection ◽  
Denoising Autoencoder ◽  
Vibration Data

One of the main challenges for structural damage detection using monitoring data is to acquire features that are sensitive to damages but insensitive to noise (e.g. sensor measurement noise) as well as environmental and operational effects (e.g. temperature effect). Inspired by the capabilities of deep learning methods in representation learning, various deep neural networks have been developed to obtain effective damage features from raw vibration data. However, most of the available deep neural networks are supervised, resulting in practical difficulties owing to the lack of damage labels. This article proposes a damage detection strategy based on an unsupervised deep neural network, referred to as deep convolutional denoising autoencoder, which accepts multi-dimensional cross-correlation functions as input. The strategy aims to extract damage features from field measurements of undamaged structures under the influence of noise and temperature uncertainties. In the proposed strategy, cross-correlation functions of vibration data are first calculated as basic features; then deep convolutional denoising autoencoder is developed to reconstruct cross-correlation functions from their noise-corrupted versions to extract desired features; exponentially weighted moving average control charts are finally established for these features to identify minor structural damages. The strategy is evaluated through a numerical simply supported beam model and an experimental continuous beam model. The mechanism of deep convolutional denoising autoencoder to extract damage features is interpreted by visualizing feature maps of convolutional layers in the encoder. It is found that these layers perform rough estimations of modal properties and preserve the damage information as the general trend of these properties in multiple extra frequency bands. The results show that the proposed strategy is competent for structural damage detection under the exposed environment and worth further exploring its capabilities in applications of real bridges.

Structural damage identification based on autoencoder neural networks and deep learning

2018 ◽  
Vol 172 ◽  
pp. 13-28 ◽  
Author(s):  
Chathurdara Sri Nadith Pathirage ◽  
Jun Li ◽  
Ling Li ◽  
Hong Hao ◽  
Wanquan Liu ◽  
...  
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Structural Damage ◽  
Damage Identification ◽  
Structural Damage Identification

Application of an Implicit Redundant Genetic Algorithm for Structural Damage Identification of Flexible Structures

2004 ◽  
Author(s):  
T. R. Liszkai
Keyword(s):  
Damage Detection ◽  
Structural Damage ◽  
Damage Identification ◽  
Optimization Problems ◽  
Flexible Structures ◽  
Nuclear Industry ◽  
Superior Performance ◽  
Structural Damage Detection ◽  
Measuring Device ◽  
Structural Damage Identification

Detecting structural damage is critical in assessing current condition, calculating remaining life, and developing rehabilitation strategies for existing structures. Many structural damage identification methods (SDIM) use vibration data to localize and identify deterioration of structural members. Due to practical constraints, such as cost, number of input channels of the measuring device, or lack of access of parts of the structure, the actual number of sensors used to collect measurement data is much smaller then the number of possible sensor locations. Therefore, the inverse problem associated with structural damage identification is ill formulated and often difficult to solve explicitly. This research addresses the problem of structural damage detection using the linear vibration information contained in frequency response functions (FRF). A structural damage identification method (SDIM) is proposed, which minimizes the error between the analytically computed and measured vibration signatures of structures. The SDIM is formulated as an unconstrained optimization problem, which is solved using genetic algorithms (GA). The implicit redundant representation (IRR) of genes allows the formulation of unstructured optimization problems in which the number of unknown variables is indefinite. The IRR GA efficiently exploits the unstructured nature of structural damage detection by allowing the number of assumed damaged elements to change throughout the optimization. The accuracy and efficiency of SDIM is increased when the IRR GA is used instead of the simple fixed representation GA. The procedure is applied to flexible structures to show that the proposed SDIM is capable of identifying damages in structures often used in the nuclear industry. Noisy measurements are also considered in the simulations to investigate their effect on the proposed SDIM accuracy. Test case results using different measurement noise levels show that the IRR GA has superior performance over the standard fixed representation GA in correctly identifying both the location and extent of damages.

scholarly journals Structural Damage Identification Based on the Minimum System Realization and Sensitivity Analysis

2014 ◽  
Vol 2014 ◽  
pp. 1-12
Author(s):  
W. R. Li ◽  
Y. F. Du ◽  
S. Y. Tang ◽  
L. J. Zhao
Keyword(s):  
Sensitivity Analysis ◽  
Damage Detection ◽  
Structural Damage ◽  
Damage Identification ◽  
Structural Information ◽  
Structural Response ◽  
Frame Structure ◽  
Structural Damage Detection ◽  
System Realization ◽  
Minimum System

On the basis of the thought that the minimum system realization plays the role as a coagulator of structural information and contains abundant information on the structure, this paper proposes a new method, which combines minimum system realization and sensitivity analysis, for structural damage detection. The structural damage detection procedure consists of three steps: (1) identifying the minimum system realization matrixes A, B, and R using the structural response data; (2) defining the mode vector, which is based on minimum system realization matrix, by introducing the concept of the measurement; (3) identifying the location and severity of the damage step by step by continuously rotating the mode vector. The proposed method was verified through a five-floor frame model. As demonstrated by numerical simulation, the proposed method based on the combination of the minimum realization system and sensitivity analysis is effective for the damage detection of frame structure. This method not only can detect the damage and quantify the damage severity, but also is not sensitive to the noise.

Sign in / Sign up

Export Citation Format

Share Document