Damage Detection in Structures Using Artificial Neural Networks

Zhang, Shilei; Huan-ding, Wang; Wang, Wei; Chen, Shaofeng

doi:10.1109/aici.2010.50

Cited by 13 publications

(3 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An improved optimization procedure based on the adaptive random search algorithm is also proposed by Masri et al . In some more recent works, Zhang et al presents the Levenberg–Marquardt backpropagation algorithm and describes it as the harmonization between the Gauss–Newton method and the steepest descent method. Furthermore, Dee et al study and compare the performance of various neural network training algorithms and suggest that the Levenberg–Marquardt backpropagation algorithm provides the best results for the purposes of vibration‐based damage detection in structural systems.…”

Section: Development Of Reduced‐order Nonlinear Computational Modelsmentioning

confidence: 99%

Development and validation of nonlinear computational models of dispersed structures under strong earthquake excitation

Derkevorkian

Masri

Fujino

et al. 2013

Earthq Engng Struct Dyn

View full text Add to dashboard Cite

SUMMARYStructural health monitoring of large multispan flexible bridges is particularly important because of their important role in civil infrastructure and transportation systems. In this study, the response of the Yokohama Bay Bridge (YBB), a three‐span cable‐stayed bridge, to the 2011 Great East Japan Earthquake is used to perform multi‐input multi‐output system identification studies. The extensive multicomponent measurements are also used to develop and validate data‐driven nonlinear mathematical models that can predict the response of YBB to various earthquake records and can accurately estimate its damping characteristics when the system is driven into the nonlinear response range. A combination of least‐square (parametric) and neural network (nonparametric) approaches is used to develop the mathematical models, along with time‐marching techniques for dynamic response calculations. It is shown that the nonlinear mathematical models perform better than the equivalent linear models, both for response prediction and damping estimation. The importance of having an accurate approach for quantifying the damping due to the variety of nonlinear features in the YBB response is shown. This study demonstrates the significance of constructing robust mathematical models that can capture the correct physics of the underlying system and that can be used for computational purposes to augment experimental studies. Given the lack of suitable data sets for full‐scale structures under extreme loads, the availability of the long‐duration measurements from the 2011 Great East Japan Earthquake and its many strong aftershocks provides an excellent opportunity to perform the analyses presented in this study. Copyright © 2013 John Wiley & Sons, Ltd.

show abstract

Section: Development Of Reduced‐order Nonlinear Computational Modelsmentioning

confidence: 99%

Development and validation of nonlinear computational models of dispersed structures under strong earthquake excitation

Derkevorkian

Masri

Fujino

et al. 2013

Earthq Engng Struct Dyn

View full text Add to dashboard Cite

show abstract

“…Three distinct algorithms are employed in stochastic subspace techniques: principal component, canonical variate analysis algorithms, and the unweighted principal component; in all cases, random data analysis and operational modal analysis constitute the primary fields of investigation for the recorded accelerograms [7][8][9]; (c) the "modal time-histories method" [10], which is based on the aforementioned techniques; this method is well suited for structures exposed to earthquake ground excitation or structures experiencing significant wind pressure. Using the "modal time-histories method", eigenfrequencies, mode shapes, and modal damping ratios have been calculated within the linear domain for a variety structures [11]; (d) the "minimum rank perturbation theory" (MRPT), as proposed by Zimmerman and Kaouk [12,13], interprets a non-zero entry in the damage vector as an indicator of the damage location; (e) a technique developed by Domaneschi et al [14,15], which involves utilizing the discontinuity of mode shape forms; (f) the concept of the damage stiffness matrix is explored in notable works, including those by Peeters [3], Amani et al [16], and Zhang et al [17]; (g) techniques that integrate structural health monitoring with pushover analysis are employed for the detection of damage in both individual structural elements [18] and frame structures [19]; (h) several artificial neural network techniques that were developed by Nazari and Baghalian [20] for simple symmetric beams. It is noteworthy to mention the recent research contributions of Reuland et al [21], who conducted an extensive review of data-driven damage indicators for rapid seismic structural health monitoring.…”

Section: Introductionmentioning

confidence: 99%

The “M and P” Technique for Damage Identification in Reinforced Concrete Bridges

Bakalis,

Makarios,

Lekidis

2024

Preprint

View full text Add to dashboard Cite

The seismic damage in reinforced concrete bridges is identified in this study using the "M and P" hybrid technique initially developed for planar frames. The proposed methodology involves a series of pushover and instantaneous modal analyses with a progressively increasing target deck displacement along the longitudinal direction of the bridge. From the results of these analyses, the diagram of the instantaneous eigenfrequency of the bridge, ranging from the health state to near collapse, is plotted against the inelastic seismic deck displacement. By pre-determining the eigen-frequency of an existing bridge along its longitudinal direction through "monitoring and frequency identification", the target deck displacement corresponding to the damage state can directly be found from this diagram. Subsequently, the damage can be identified by examining the results of the pushover analysis at the step where the target deck displacement is indicated. The effectiveness of this proposed technique is evaluated in the context of multiple span bridges with unequal pier heights, illustrated through an example of a four-span bridge. The findings demonstrate that the damage potential in bridge piers can be successfully identified by combining the results of a monitoring process and pushover analysis.

show abstract

“…One more technique, which uses an artificial neural network, was developed by Nazari and Baghalian [16] for simple symmetric beams. Moreover, the idea of the damage stiffness matrix is presented in interesting works, such those of Peeters [3], Amani et al [17], and Zhang et al [18]. It is also worth mentioning the recent research efforts by Reuland et al [19], which led to a comprehensive review of data-driven damage indicators for rapid seismic structural health monitoring, as well as those by Martakis et al [20], which considered a combination of traditional structural health monitoring techniques with novel machine learning tools.…”

Section: Introductionmentioning

confidence: 99%

Identification of Damage in Planar Multi-Storey Reinforced Concrete Frames Developing a Beam-Sway Plastic Mechanism Using the “M and P” Technique

Makarios,

Bakalis

2023

Preprint

View full text Add to dashboard Cite

The effectiveness of a recently proposed methodology in the identification of damage in planar, multi-storey, Reinforced Concrete (RC) moment-frames, which develop a plastic-yield mechanism on their beams, is showcased here via the examining of a group of such existing multi-storey frames with three or more unequal spans. According to the methodology, the diagram of the instantaneous Eigen-Frequencies of the frame in the nonlinear regime is drawn as a function of the inelastic seismic roof displacement by performing a sequence of pushover and instantaneous modal analyses with gradually increasing target displacement. Using this key-diagram, the locations of severe seismic damage in an existing moment-frame can be evaluated if the instantaneous fundamental eigen-frequency of the damaged frame, at an analysis step within the nonlinear area, is known in advance by “the monitoring and the identification of frequencies” using a local network of uniaxial accelerometers. This is a hybrid technique because both procedures, the instrumental Monitoring of the structure and the Pushover analysis on the frame (M and P technique), are combined. Moreover, the damage image of the planar multi-storey moment-frame is illustrated, and the lateral stiffness matrix of the damaged frame is calculated with high accuracy.

show abstract

Damage Detection in Structures Using Artificial Neural Networks

Cited by 13 publications

References 20 publications

Development and validation of nonlinear computational models of dispersed structures under strong earthquake excitation

Development and validation of nonlinear computational models of dispersed structures under strong earthquake excitation

The “M and P” Technique for Damage Identification in Reinforced Concrete Bridges

Identification of Damage in Planar Multi-Storey Reinforced Concrete Frames Developing a Beam-Sway Plastic Mechanism Using the “M and P” Technique

Contact Info

Product

Resources

About