Identification of time-varying nonlinear structural physical parameters by integrated WMA and UKF/UKF-UI

Yang, Ning; Li, Jun; Lei, Ying; Hao, Hong

doi:10.1007/s11071-021-06682-y

Cited by 16 publications

(7 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…8 A two-step identification method using wavelet multiresolution analysis integrated with unscented Kalman filter was further developed to successfully identify the time-varying physical parameters of nonlinear systems. 9 An unsupervised clustering method on structural frequency using acceleration has been investigated to detect damage of joints in an actual bridge. 10 A support vector machine classifier was employed to perform crack detection and condition assessment using the acceleration of the Sydney Harbour Bridge.…”

Section: Introductionmentioning

confidence: 99%

Unsupervised deep learning method for bridge condition assessment based on intra-and inter-class probabilistic correlations of quasi-static responses

Tian

2022

Structural Health Monitoring

View full text Add to dashboard Cite

Data-driven methods for structural condition assessment have been extensively investigated using deep learning (DL). However, studies on quasi-static response data-based structural health diagnoses are relatively insufficient. The difficulty is that quasi-static response data contain coupled effects of structural parameters and external loads. Considering that the correlation between quasi-static responses subjected to identical external loads is only a function of structural parameters and independent from the external loads, the correlation can therefore be employed as an indicator of the structural condition. This study proposes a condition assessment approach for cable-stayed bridges based on correlation modeling between the deflection of girders and tension in cables. The correlation is modeled by an unsupervised DL network comprising two variational autoencoders (AE) and two generative adversarial networks (GANs). The input and output are marginal probability density functions (PDFs). The DL network is trained as the reconstruction and translation processes to model the intra-class and inter-class correlations. Assumptions of shared latent space and cycle consistency are taken to ensure mutual modeling capacity. The Wasserstein distance between the predicted and ground-truth PDFs of tension in cables is used as an indicator of the structural condition. Using probabilistic correlation of quasi-static responses only requires the PDF of external loads to be identical and does not need the external loads to be precisely identical at any moment, thus relieving time-synchronization restrictions for different sensors. The results show that the predicted PDFs agree well with the ground-truth values under normal conditions. Furthermore, the Wasserstein distance is sensitive to damage and shows noticeable variations when the damage of the stay cable occurs.

show abstract

Section: Introductionmentioning

confidence: 99%

Unsupervised deep learning method for bridge condition assessment based on intra-and inter-class probabilistic correlations of quasi-static responses

Tian

2022

Structural Health Monitoring

View full text Add to dashboard Cite

show abstract

“…Wang et al [17,18] proposed an instantaneous frequency identification method based on the continuous wavelet transform (WT) and applied it to the identification of time-varying cable force. However, the frequency resolution of WT needs to be improved and the boundary effect of WT is problematic [24,25]. Hou et al [19] provided a method based on the variational mode decomposition (VMD).…”

Section: Introductionmentioning

confidence: 99%

Real-Time Identification of Time-Varying Cable Force Using an Improved Adaptive Extended Kalman Filter

Yang

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

The real-time identification of time-varying cable force is critical for accurately evaluating the fatigue damage of cables and assessing the safety condition of bridges. In the context of unknown wind excitations and only one available accelerometer, this paper proposes a novel cable force identification method based on an improved adaptive extended Kalman filter (IAEKF). Firstly, the governing equation of the stay cable motion, which includes the cable force variation coefficient, is expressed in the modal domain. It is transformed into a state equation by defining an augmented Kalman state vector with the cable force variation coefficient concerned. The cable force variation coefficient is then recursively estimated and closely tracked in real time by the proposed IAEKF. The contribution of this paper is that an updated fading-factor matrix is considered in the IAEKF, and the adaptive noise error covariance matrices are determined via an optimization procedure rather than by experience. The effectiveness of the proposed method is demonstrated by the numerical model of a real-world cable-supported bridge and an experimental scaled steel stay cable. Results indicate that the proposed method can identify the time-varying cable force in real time when the cable acceleration of only one measurement point is available.

show abstract

“…Additionally, more low‐frequency scale coefficients need be retained to ensure the accuracy of reconstructing the gradually changing parameters, which also leads to the growth in the number of scale coefficients. Furthermore, the determination of appropriate mother function and decomposition level in WM is a pending issue that is not yet well resolved 24 . Besides, the boundary effect of WM is also inevitable.…”

Section: Introductionmentioning

confidence: 99%

“…Furthermore, the determination of appropriate mother function and decomposition level in WM is a pending issue that is not yet well resolved. 24 Besides, the boundary effect of WM is also inevitable.…”

Section: Introductionmentioning

confidence: 99%

Identification of gradually varying physical parameters based on discrete cosine transform using partial measurements

Yang

Lei

et al. 2022

Structural Contr & Hlth

View full text Add to dashboard Cite

Structural physical parameters often vary gradually due to the degradation of material properties or effects of environment. In this paper, two novel approaches are proposed to identify the gradually varying physical parameters based on the discrete cosine transform (DCT) using partial measurements of structural responses. Approach I is proposed for the circumstance of known excitations. The gradually varying physical parameters are first located by the fading-factor extended Kalman filter (FEKF) and then identified by the proposed DCT integrated with Kalman filter (KF) method. Approach II is proposed for the identification of gradually varying physical parameters under unknown excitations. The gradually varying physical parameters are first localized by the proposed fading-factor extended Kalman filter under unknown input (FEKF-UI) and then identified by the proposed DCT integrated with Kalman filter under unknown input (KF-UI). Numerical examples demonstrate that the proposed approaches can identify the gradually varying physical parameters accurately with incomplete measurement data. Moreover, the identification of time-varying cable force in cable-stayed bridge is also discussed as a case study of the proposed approach I. Experimental verification shows that it provides a new path to identify the time-varying cable force by only using one acceleration response measurement of the cable.

show abstract

Identification of time-varying nonlinear structural physical parameters by integrated WMA and UKF/UKF-UI

Cited by 16 publications

References 46 publications

Unsupervised deep learning method for bridge condition assessment based on intra-and inter-class probabilistic correlations of quasi-static responses

Unsupervised deep learning method for bridge condition assessment based on intra-and inter-class probabilistic correlations of quasi-static responses

Real-Time Identification of Time-Varying Cable Force Using an Improved Adaptive Extended Kalman Filter

Identification of gradually varying physical parameters based on discrete cosine transform using partial measurements

Contact Info

Product

Resources

About