1,354,387 research outputs found
Structural health monitoring by a new hybrid feature extraction and dynamic time warping methods under ambient vibration and non-stationary signals
No full textFeature extraction and classification are crucial steps of a data-driven structural health monitoring strategy. One of the major issues in feature extraction is to extract damage-sensitive features from non-stationary signals under unknown ambient vibration. Furthermore, the use of high-dimensional features in damage detection is the other challenging issue, which may make a difficult and time-consuming process. This article is initially intended to propose a hybrid algorithm as a combination of EEMD technique and ARARX model for feature extraction. Subsequently, correlation-based dynamic time warping method is proposed to detect damage by using randomly high-dimensional multivariate features. Due to the importance of damage localization, dynamic time warping is eventually applied to locate damage. Experimental datasets of the IASC-ASCE benchmark structure are utilized to validate the accuracy of proposed methods. Results suggest that the proposed methods are effective tools for damage detection and localization under ambient vibration and non-stationary and/or stationary signals
Improving feature extraction via time series modeling for structural health monitoring based on unsupervised learning methods
No full textFeature extraction by time series modeling based on statistical pattern recognition is a powerful approach to Structural Health Monitoring (SHM). Determination of an adequate order and identi cation of an appropriate model play prominent roles in extracting sensitive features to damage from time series representations. Early damage detection under statistical decision-making via high-dimensional features is another signi cant issue. The main objectives of this study were to improve a residual-based feature extraction method by time series modeling and to propose a multivariate data visualization approach to early damage detection. A simple graphical tool based on Box-Jenkins methodology was adopted to identify the most compatible time series model with vibration time-domain measurements. Furthermore, k-means and Gaussian Mixture Model (GMM) clustering techniques were utilized to examine the performance of the residuals of the identi ed model in damage detection. A numerical concrete beam and an experimental benchmark model were applied to verifying the improved and proposed methods along with comparative analyses. Results showed that the approaches were successful and superior to a state-of-the-art order determination technique in obtaining a sufficient order, generating uncorrelated residuals, extracting sensitive features to damage, and accurately detecting early damage by high-dimensional data
Damage localization under ambient excitations and non-stationary vibration signals by a new hybrid algorithm for feature extraction and multivariate distance correlation methods
No full text Ambient excitations applied to structures may lead to non-stationary vibration responses. In such circumstances, it may be difficult or improper to extract meaningful and significant damage features through methods that mainly rely on the stationarity of data. This article proposes a new hybrid algorithm for feature extraction as a combination of a new adaptive signal decomposition method called improved complete ensemble empirical mode decomposition with adaptive noise and autoregressive moving average model. The major contribution of this algorithm is to address the important issue of feature extraction under ambient vibration and non-stationary signals. The improved complete ensemble empirical mode decomposition with adaptive noise method is an improvement on the well-known ensemble empirical mode decomposition technique by removing redundant intrinsic mode functions. In addition, a novel automatic approach is presented to select the most relevant intrinsic mode functions to damage based on the intrinsic mode function energy level. Fitting an autoregressive moving average model to each selected intrinsic mode function, the model residuals are extracted as the damage-sensitive features. The main limitation is that such features are high-dimensional multivariate time series data, which may make a difficult and time-consuming decision-making process for damage localization. Multivariate distance correlation methods are introduced to cope with this drawback and locate structural damage using the multivariate residual sets of the normal and damaged conditions. The accuracy and robustness of the proposed methods are validated by a numerical shear-building model and an experimental benchmark structure. The effects of sampling frequency and time duration are evaluated as well. Results demonstrate the effectiveness and capability of the proposed methods to extract sufficient and reliable features, identify damage location, and quantify damage severity under ambient excitations and non-stationary signals. </jats:p
Application of supervised learning to validation of damage detection
No full textUnsupervised learning methods are effective and suitable tools for damage detection. The main reason for the popularity of these methods in structural health monitoring originates from the fact that the process of learning can be implemented by information of the only normal condition called training data. In contrast, supervised learning methods require information of both normal and current conditions for the process of interest. Because civil engineering structures are expensive and complex, it is not reasonable and economical to impose intentional damage on providing training data. Hence, it is not simple to directly exploit supervised learning techniques in structural health monitoring. To deal with this limitation, this article proposes a novel two-level strategy including three algorithms for using the concepts of both unsupervised learning and supervised learning. The major contribution of this strategy is to consider supervised learning as a validation tool for damage detection. First, the results of damage detection are obtained from two unsupervised learning methods developed by Mahalanobis squared distance and a deep autoencoder neural network in the first two algorithms of the proposed strategy. The main objective is to separate accurate and confusing results of damage detection based on Type I and Type II errors. Second, the confusing results are fed into the third algorithm to train a classifier and compute their classification margins for making the final decision and validating damage detection. The effectiveness and applicability of the proposed strategy are assessed by a numerical concrete beam and an experimental laboratory frame. Results show that this strategy with the aid of the Naïve Bayes classifier enables the unsupervised learning methods to make accurate decisions
New sensitivity-based methods for structural damage diagnosis by least square minimal residual techniques
No full textThis paper presents new sensitivity-based methods for detection of structural damage using incomplete noisy modal data. These methods are based on the first-order derivative of modal parameters. Changes of natural frequency do not usually provide spatial information on the structural damage. They are also not sensitive to the local damage. In this paper, a new sensitivity function is proposed using method of Lagrange multipliers in order to deal with these weaknesses when applying natural frequency in the sensitivity-based damage diagnosis. Mode shape is the other vibrational data which leads to better results in comparison with natural frequency. However, usually some mode shape's sensitivities require all modes to obtain exact sensitivity functions. Thus, an improved sensitivity of mode shape is presented to constitute an applicable formulation based on using incomplete modes. To determine the damage quantity, a powerful iterative method named Least-Square Minimal Residual (LSMR) technique is proposed in the condition of incomplete modes. Subsequently, Regularized Least-Square Minimal Residual (RLSMR) method is presented to detect structural damage when the incomplete modal parameters are contaminated by noise. Applicability and effectiveness of the proposed methods are numerically verified using two practical examples consisting of a six-story shear building and a planner truss. Eventually, numerical results indicate that the LSMR and RLSMR are influential algorithms for precisely determining the damage severity. Furthermore, obtained results of damage diagnosis process in the free-noise data show that the proposed sensitivities of natural frequency and mode shape can provide reliable and accurate results for structural damage diagnosis
An unsupervised learning approach by novel damage indices in structural health monitoring for damage localization and quantification
No full text The aim of this article is to propose novel damage indices for damage localization and quantification based on time series modeling. In order to extract damage-sensitive features from time series models, it is essential to choose adequate and robust orders in such a way that the models are able to extract uncorrelated residuals. On this basis, a new iterative order determination method is proposed to select robust orders of time series models under residual analysis by Ljung–Box Q-test. The damage-sensitive features are the parameters and residuals of an AutoRegressive (AR) model obtained from current feature extraction approaches. In this study, the AR model is identified as the most compatible time series model with measured vibration time-domain responses using Box–Jenkins methodology and Leybourne–McCabe hypothesis test. The proposed damage indices are the parametric assurance criterion and the residual reliability criterion that exploit the parameters and residuals of AR models, respectively. The main idea behind locating a damage is to define threshold limits for both damage indices using the features of undamaged conditions based on an unsupervised learning way. The major contributions of this article are to propose an iterative order determination method for time series models and two novel damage indices for locating and quantifying damage. The accuracy and performance of the proposed methods are experimentally demonstrated on a three-story laboratory frame and a model-scale steel structure. Results show that the proposed iterative approach leads to uncorrelated residuals, and the proposed parametric assurance criterion and the residual reliability criterion methods are promising and efficient tools in damage detection problems under varying operational and environmental conditions. </jats:p
Damage detection in structural systems by improved sensitivity of modal strain energy and Tikhonov regularization method
No full textIn this article, new methods for detecting damage in structural systems are presented. These methods are categorized as damage localization and damage quantification, respectively. Hence, direct changes of modal strain energy are applied to identify locations of damage. Moreover, some restraints such as incomplete measured modes and simple assumptions in structural modeling may cause failure in the results of damage localization. Therefore, a correlation-based method is utilized to obviate these limitations and precisely detect damage sites. Subsequently, an improved sensitivity of modal strain energy is generated to determine damage severities. To achieve appropriate results in damage quantification, Tikhonov regularization approach is utilized instead of classical methods such as applying penalty function and current inverse problem techniques. Applicability and effectiveness of proposed methods are numerically verified using two practical examples consisting of a planner truss and a portal frame, respectively. Eventually, numerical results indicate that the proposed damage localization approach provides an influential algorithm for precisely identifying damage sites. Furthermore, obtained damage severities show that utilizing the sensitivity of modal strain energy and also solving the damage equation by Tikhonov regularization makes it possible to accurately determine damage extents in the case of incomplete modal data
Damage localization in shear buildings by direct updating of physical properties
No full textThe objective of this article is to present a new method for identifying the damage location in a multi-story shear building by direct model updating method. In this regard, structural perturbation matrices should be determined that are directly defined as the discrepancy between mass and stiffness matrices of undamaged and damaged structures. As a result of expanding the dynamic orthogonality conditions, mass and stiffness perturbation matrices are formulated by the initial information of undamaged structures as well as the structure’s modal parameters before and after the occurrence of damages. These matrices cannot easily detect the damage site. Therefore, a more explicit determination of damage location is performed dividing the amount of change in these matrices’ diagonals by the physical properties of undamaged structure. This modification facilitates the damage localization process and yields precise and preferable results in comparison with applying classical methods such as natural frequencies, mode shapes and structural properties changes. Subsequently, the applicability and effectiveness of the proposed damage detection method are verified numerically and experimentally. For numerical verification of the proposed methods, a six-story shear building is utilized as a discrete system. Then, the experimental verification of proposed methods is conducted detecting the location of damages in a simple laboratory frame. It can be deduced that the proposed damage localization method can reliably detect and also localize the structural damage
A sensitivity-based finite element model updating based on unconstrained optimization problem and regularized solution methods
No full textAn effective and reliable approach to updating finite element (FE) models of real structures is to utilize a sensitivity-based strategy. A challenging issue concerning the sensitivity-based finite element model updating (FEMU) is to create a well-established framework for updating the inherent structural properties of FE models under incomplete noisy modal data. When noise contaminates the measured modal parameters, another challenging issue stems from the ill-posedness of the FEMU inverse problem. This article proposes an innovative sensitivity-based FEMU strategy based on the combination of modal kinetic energy and modal strain energy for simultaneously updating the element mass and stiffness matrices of FE models. The great novelty of this strategy is to get an idea from the unconstrained optimization problem for the establishment of a sensitivity-based FEMU framework. The correction of the element mass and stiffness matrices in a simultaneous way is another novelty of the proposed FEMU strategy. Moreover, new iterative and hybrid regularization methods under the Krylov subspace theory and bidiagonalization process are presented to solve the ill-posed inverse problem of FEMU. The accuracy and reliability of the proposed methods are numerically validated by a two-story concrete frame and a two-span continuous steel truss along with some comparative analyses. Results demonstrate that the suggested sensitivity-based strategy and regularized solution methods are influential and successful in FEMU under incomplete noisy modal data
A Multi-stage Machine Learning Methodology for Health Monitoring of Largely Unobserved Structures Under Varying Environmental Conditions
No full textStructural Health Monitoring (SHM) via data-driven techniques can be based upon vibrations acquired by sensor networks. However, technical and economic reasons may prevent the deployment of pervasive sensor networks over civil structures, thus limiting their reliability in terms of damage detection. Moreover, the effects of environmental (and operational) variability may lead to false alarms. To address these challenges, a multi-stage machine learning (ML) method is here proposed by exploiting autoregressive (AR) spectra as damage-sensitive features. The proposed method is framed as follows: (i) computing the distances between different sets of the AR spectra via the log-spectral distance (LSD), providing also the training and test datasets; (ii) removing the potential environmental variability by an auto-associative artificial neural network (AANN), to set normalized training and test datasets; (iii) running a statistical analysis via the Mahalanobis-squared distance (MSD) for early damage detection. The effectiveness of the proposed approach is assessed in the case of limited vibration data for the laboratory truss structure known as the Wooden Bridge. Comparative studies show that the AR spectrum is a reliable feature, sensitive to damage even in the presence of a limited number of sensors in the network; additionally, the multi-stage ML methodology succeeds in early detecting damage under environmental variability
- …
