Structural Damage Identification Using Improved RBF Neural Networks in Frequency Domain

doi:10.1260/1369-4332.15.10.1689

Home
/
Papers
/
Structural Damage Identification Using Improved RBF Neural Networks in Frequency Domain

Journal Article•DOI•

Structural Damage Identification Using Improved RBF Neural Networks in Frequency Domain

Rajendra Machavaram¹, K. Shankar¹•Institutions (1)

Indian Institute of Technology Madras¹

01 Oct 2012-Advances in Structural Engineering (SAGE Publications)-Vol. 15, Iss: 10, pp 1689-1703

TL;DR: The novel improved RBF network is shown to be a good damage identification strategy for multiple member structures compared to conventional RBF and existing hybrid methods in terms of accuracy and computational effort.

read less

Abstract: This paper presents a novel two stage improved Radial basis function (RBF) neural network for the damage identification of multimember structures in the frequency domain. The improvement of the proposed RBF network is carried out in two stages, viz. (i) first stage damage prediction by conventional RBF network trained with effective input-output patterns and (ii) in the second stage, minimization of the prediction error below the predefined error tolerance (3%) by training the network with patterns from reduced search space located after the first stage prediction. The network effective input patterns are fractional frequency change ratios (FFCs) and damage signature indices (DSIs), and the corresponding output patterns are stiffness values or damage severity of the structure at different damage levels. A Latin hypercube search (LHS) technique is used for finding the effective input-output patterns from the search space to improve the training efficiency. The numerical simulation of structural damage iden...

...read moreread less

Citations

PDF

Open Access

More filters

Journal Article•DOI•

Machine learning for structural engineering: A state-of-the-art review

[...]

Huu-Tai Thai

01 Apr 2022-Structures

TL;DR: An overview of ML techniques for structural engineering is presented in this article with a particular focus on basic ML concepts, ML libraries, open-source Python codes, and structural engineering datasets.

...read moreread less

89 citations

Journal Article•DOI•

Vibration suppression of printed circuit boards using an external particle damper

[...]

P. Veeramuthuvel¹, K.K. Sairajan¹, K. Shankar²•Institutions (2)

Indian Space Research Organisation¹, Indian Institute of Technology Madras²

31 Mar 2016-Journal of Sound and Vibration

TL;DR: In this article, the use of particle damper capsule on a Printed Circuit Board (PCB) and the development of Radial Basis Function neural network to accurately predict the acceleration response is presented.

...read moreread less

28 citations

Journal Article•DOI•

Improved Complex-valued Radial Basis Function (ICRBF) neural networks on multiple crack identification

[...]

M. Rajendra¹, K. Shankar¹•Institutions (1)

Indian Institute of Technology Madras¹

01 Mar 2015

TL;DR: The results proved that, the proposed ICRBF and real-valued Improved RBF (IRBF) neural networks have identified the single and multiple cracks with less than 1% absolute mean percentage error as compared to conventional CRBF and RBF neural networks, mainly because of their second stage reduced search space moving technique.

...read moreread less

Abstract: Absolute mean percentage error (AMPE) of single crack and multiple crack identification using different RBF networks. A two-stage ICRBF neural network is developed for multiple crack identification.Conventional CRBF neural network is used in the first stage of ICRBF.Reduced search space moving technique is used in the second stage of ICRBF.Crack location and depth are identified using frequency domain vibration signals.ICRBF is more efficient followed by IRBF, CRBF, RBF and MLP neural networks. This paper introduces a new class of neural networks in complex space called Complex-valued Radial Basis Function (CRBF) neural networks and also an improved version of CRBF called Improved Complex-valued Radial Basis Function (ICRBF) neural networks. They are used for multiple crack identification in a cantilever beam in the frequency domain. The novelty of the paper is that, these complex-valued neural networks are first applied on inverse problems (damage identification) which come under the category of function approximation. The conventional CRBF network was used in the first stage of ICRBF network and in the second stage a reduced search space moving technique was employed for accurate crack identification. The effectiveness of proposed ICRBF neural network was studied first on a single crack identification problem and then applied to a more challenging problem of multiple crack identification in a cantilever beam with zero noise as well as 5% noise polluted signals. The results proved that, the proposed ICRBF and real-valued Improved RBF (IRBF) neural networks have identified the single and multiple cracks with less than 1% absolute mean percentage error as compared to conventional CRBF and RBF neural networks, mainly because of their second stage reduced search space moving technique. It appears that IRBF neural network is a good compromise considering all factors like accuracy, simplicity and computational effort.

...read moreread less

14 citations

Journal Article•DOI•

Application of RBF neural network in prediction of particle damping parameters from experimental data

[...]

P. Veeramuthuvel¹, P. Veeramuthuvel², K. Shankar¹, K.K. Sairajan²•Institutions (2)

Indian Institute of Technology Madras¹, Indian Space Research Organisation²

01 Apr 2017-Journal of Vibration and Control

TL;DR: In this article, a radial basis function (RBF) neural network was used to predict the modal damping ratio of a particle damping system using system input parameters such as particle size, particle density, packing ratio, and their effect at different modes of vibration.

...read moreread less

Abstract: Particle damping is one of the recent passive damping methods and its relevance in space structural applications is increasing. This paper presents the novel application of a radial basis function (RBF) neural network to accurately predict the modal damping ratio of a particle damping system using system input parameters such as particle size, particle density, packing ratio, and their effect at different modes of vibration. The prediction of particle damping using the RBF neural network is studied in comparison with the back propagation neural (BPN) network on an aluminum alloy beam structure with extensive experimental tests. The prediction accuracy of the RBF neural network is significant with 9.83% error compared to 12.22% obtained by the BPN network for a best case. Limited experiments were also carried out on a mild steel beam to study and compare the trends predicted in earlier studies. The relationships obtained by the proposed method readily provide useful guidelines in the design of particle dam...

...read moreread less

12 citations

Cites background or methods from "Structural Damage Identification Us..."

...The Euclidean distance is calculated from xm ci k k(2) (Machavaram and Shankar, 2012)....
[...]
...where Wni is the weight vector in the output layer for the nth output node and i is the radial basis function of the ith node (Machavaram and Shankar, 2012)....
[...]
...The output from the i th Gaussian neuron from an input xm is calculated using i ¼ exp xm cik k2 2i ð1Þ The Euclidean distance is calculated from xm cik k2 (Machavaram and Shankar, 2012)....
[...]
...The output of the hidden neuron is computed as: yn¼ X i Wni i ð2Þ where Wni is the weight vector in the output layer for the nth output node and i is the radial basis function of the ith node (Machavaram and Shankar, 2012)....
[...]

Journal Article•DOI•

Experimental and numerical studies on a test method for damage diagnosis of stay cables

[...]

Yonghui An¹, Zhong Yue¹, Yanbin Tan¹, Jinping Ou¹•Institutions (1)

Dalian University of Technology¹

01 Feb 2017-Advances in Structural Engineering

TL;DR: In this article, a vibration-based model-free damage diagnosis method of stay cables using the changes in natural frequencies is further proposed and validated, and a structural model is used to diagnose the state of stay cable.

...read moreread less

Abstract: To diagnose the state of stay cables, a vibration-based model-free damage diagnosis method of stay cables using the changes in natural frequencies is further proposed and validated. The structural ...

...read moreread less

10 citations

Cites background from "Structural Damage Identification Us..."

...Changes in the physical properties will cause changes in the modal properties (Machavaram and Shankar, 2012)....
[...]

References

PDF

Open Access

More filters

Journal Article•DOI•

Real time detection of stiffness change using a radial basis function augmented observer formulation

[...]

Michael T. Contreras¹, Satish Nagarajaiah¹, Sriram Narasimhan²•Institutions (2)

Rice University¹, University of Waterloo²

01 Mar 2011-Smart Materials and Structures

TL;DR: In this paper, the authors employ the computational efficiency of single layer radial basis function approximaters to create a subspace capable of isolating faults in multi-degree of freedom systems which involve coupled and uncoupled stiffness changes in real time.

...read moreread less

Abstract: Existing methods for structural health monitoring pose a formidable challenge to real time implementation due to the significantly large computational loads. The proposed algorithm is suitable for online applications because it maintains good pattern recognition capabilities while possessing a computationally compact network topology. This study employs the computational efficiency of single layer radial basis function (RBF) approximaters to create a subspace capable of isolating faults in multi-degree of freedom systems which involve coupled and uncoupled stiffness changes in real time. The RBF network transforms the displacement–time history of the varying plant into a decoupled output space which is then compared to a baseline healthy observer which undergoes the same decoupling transformation. The online comparison of the output of the time varying plant and the healthy observer in a decoupled subspace comprises the observer based error function. The error function is shown to not only detect the existence of faults, but also isolate these faults in real time in the presence of base excitation. The method is validated for systems that experience earthquake induced damage, as well as an experimental system using a semi-active independent variable stiffness device which is capable of varying system stiffness in real time. By simply observing the displacement–time history responses, the RBF augmented observer formulation is capable identifying changes in the stiffness at each degree of freedom.

...read moreread less

18 citations

"Structural Damage Identification Us..." refers methods or result in this paper

...…first stage prediction results in this study were considered as the conventional RBF network (Okafor and Dutta 2001; Reddy and Ganguli 2003; Zang et al. 2007; EI-Shafie et al. 2010; Contreras et al. 2011; Zheng et al. 2011) results to compare the performance of the improved two stage RBF network....
[...]
...…the applicability of the improved RBF methodology, and its performance was compared with conventional RBF method i.e., first stage of the proposed RBF (Contreras et al. 2011; Zheng et al. 2011) and CPN-BPN hybrid method (Prashant and Shankar 2008) in terms of accuracy and computational effort....
[...]
...The first stage prediction results in this study were considered as the conventional RBF network (Okafor and Dutta 2001; Reddy and Ganguli 2003; Zang et al. 2007; EI-Shafie et al. 2010; Contreras et al. 2011; Zheng et al. 2011) results to compare the performance of the improved two stage RBF network....
[...]
...…effort and compared with the conventional RBF method (Okafor and Dutta 2001; Reddy and Ganguli 2003; Zang et al. 2007; EI-Shafie et al. 2010; Contreras et al. 2011; Zheng et al. 2011); and with the other traditional methods, particularly the work published by Prashant and Shankar (2008)....
[...]
...The performance of the proposed improved RBF method is examined in terms of accuracy and computational effort and compared with the conventional RBF method (Okafor and Dutta 2001; Reddy and Ganguli 2003; Zang et al. 2007; EI-Shafie et al. 2010; Contreras et al. 2011; Zheng et al. 2011); and with the other traditional methods, particularly the work published by Prashant and Shankar (2008)....
[...]

Journal Article•

Damage Assessment of Bridge Structures by System Identification

[...]

Chung-Bang Yun, Won Jong Kim, Alfredo H. S. Ang

01 Jan 1990-Structural Safety and Reliability

TL;DR: In this paper, a method for the damage estimation of bridge structures under seismic ground excitation is developed based on system identification approach, where the dynamic behavior of a damaged structure is represented by a nonlinear hysteresis model.

...read moreread less

Abstract: A method for the damage estimation of bridge structures under seismic ground excitation is developed based on system identification approach. The dynamic behavior of a damaged structure is represented by a nonlinear hysteresis model. Estimation of the nonlinear structural parameters is carried out by the extended Kalman filter.

...read moreread less

16 citations

Journal Article•DOI•

Optimal ultrasonic pulse repetition rate for damage detection in plates using neural networks

[...]

Anthony Chukwujekwu Okafor¹, Amitabha Dutta¹•Institutions (1)

Missouri University of Science and Technology¹

01 Oct 2001-Ndt & E International

TL;DR: In this paper, an experimental study for the determination of the optimal pulse repetition rate frequency (PRF) for damage detection in aluminum and composites is presented, and a method for predicting the damage size and depth from C-Scan results using neural networks is also presented.

...read moreread less

Abstract: An experimental study for the determination of the optimal pulse repetition rate frequency (PRF) for damage detection in aluminum and composites is presented in this paper. A method for predicting the damage size and depth from C-Scan results using neural networks is also presented. Two graphite fiber IM7/F5250-4 (Bismaleimid) composite plates and four aluminum plates were used for the study. Damage was fabricated by drilling holes of varying depth and diameter on the test specimens. Ultrasonic transmission tests were carried out on a DIGITALWAVE immersion type C-Scan system. PRF values from 100 to 5000 Hz were investigated for the scan. The defect locations were clearly observed as peaks in the C-Scan mesh. The equivalent hole diameter, depth and the location of the holes with respect to a predetermined edge were calculated from the C-Scan plots and correlated with the actual values to determine the optimal PRF values. A close correlation was found between the calculated diameter obtained from the C-Scan results and the actual hole diameter. Low PRF values (100 Hz) were found best for scanning of aluminum and intermediate values (500 Hz) were best for scanning of composites. Prediction of the actual damage size from the C-Scan calculated damage size was successfully accomplished with radial basis function neural network.

...read moreread less

13 citations

"Structural Damage Identification Us..." refers methods or result in this paper

...The first stage prediction results in this study were considered as the conventional RBF network (Okafor and Dutta 2001; Reddy and Ganguli 2003; Zang et al. 2007; EI-Shafie et al. 2010; Contreras et al. 2011; Zheng et al. 2011) results to compare the performance of the improved two stage RBF network....
[...]
...…of the proposed improved RBF method is examined in terms of accuracy and computational effort and compared with the conventional RBF method (Okafor and Dutta 2001; Reddy and Ganguli 2003; Zang et al. 2007; EI-Shafie et al. 2010; Contreras et al. 2011; Zheng et al. 2011); and with the other…...
[...]
...Okafor and Dutta (2001) used radial basis function neural networks to predict the damage size and depth of aluminium and composite plates as measured from C-Scan results....
[...]
...The performance of the proposed improved RBF method is examined in terms of accuracy and computational effort and compared with the conventional RBF method (Okafor and Dutta 2001; Reddy and Ganguli 2003; Zang et al. 2007; EI-Shafie et al. 2010; Contreras et al. 2011; Zheng et al. 2011); and with the other traditional methods, particularly the work published by Prashant and Shankar (2008)....
[...]
...The first stage prediction results in this study were considered as the conventional RBF network (Okafor and Dutta 2001; Reddy and Ganguli 2003; Zang et al. 2007; EI-Shafie et al. 2010; Contreras et al. 2011; Zheng et al. 2011) results to compare the performance of the improved two stage RBF…...
[...]

Journal Article•DOI•

A hybrid neural network strategy for the identification of structural damage using time domain responses

[...]

Prashanth Pillai¹, K. Shankar¹•Institutions (1)

Indian Institute of Technology Madras¹

20 Jan 2009-The Ies Journal Part A: Civil & Structural Engineering

TL;DR: A hybrid neural network method has been proposed that uses a counterpropagation neural network (CPN) in the first stage for sorting the training data into clusters, giving an approximate guess of the damage extent quickly, showing the computational superiority of the hybrid method compared with the conventional single stage method.

...read moreread less

Abstract: A multistage identification scheme for structural damage detection using time domain acceleration responses is proposed. Previous studies of damage assessment using neural networks mostly involved training a backpropagation neural network (BPN) to learn damage patterns with significant computational effort. A hybrid neural network method has been proposed that uses a counterpropagation neural network (CPN) in the first stage for sorting the training data into clusters, giving an approximate guess of the damage extent quickly. After an approximate estimate is obtained, a new set of training patterns of reduced size is generated using the CPN prediction. In the second stage, a BPN trained with the Levenberg–Marquardt algorithm is used to learn the new training data and predict a more accurate result. A superior convergence and a substantial decrease in central processing unit time have been observed for three numerical examples. These examples show the computational superiority of the hybrid method compared...

...read moreread less

8 citations

"Structural Damage Identification Us..." refers background in this paper

...Prashant and Shankar (2009) developed a hybrid neural network strategy for structural damage detection using time domain acceleration responses....
[...]
...This reduces the prediction efficiency and accuracy of the damage detection in single stage identification (Prashant and Shankar 2008, 2009)....
[...]