On the use of the GP-NARX model for predicting hysteresis effects of bolted joint structures

Teloli, Rafael de Oliveira; Villani, Luis Gustavo; Silva, Samuel da; Todd, Michael D.

doi:10.1016/j.ymssp.2021.107751

Cited by 12 publications

(9 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…When the random process is stationary, let 𝑡 𝑗 = 𝑡 𝑖 + 𝜏, and the 2-dimensional spectrum is following as 𝑆 𝑋𝑋 (2) (𝜔 𝑖 , 𝜔 𝑗 ) = ℱ [𝑅 𝑥𝑥 (2) (𝑡 𝑖 , 𝑡 𝑗 )] = 1 (2𝜋) 2 ∬ 𝑅 𝑥𝑥 (2) (𝑡 𝑖 , 𝑡 𝑗 )e −i(𝜔 𝑖 𝑡 𝑖 +𝜔…”

Section: Generalized Pem For the Analysis Of Response Mpsdmentioning

confidence: 99%

“…Linear dynamic models can be used only for small deformations with small amplitude vibrations or the inapparent nonlinear characteristics. More often, nonlinear dynamic models must be used to describe many practical situations, such as thin plate buckling problems [1], hysteresis effects from nonlinear constitutive relation [2], contact nonlinearity in bolted structures [3,4], and geometric nonlinearity from large deflection beams [5]. Nonlinear systems typically exhibit complex mechanical behaviour, such as the dependency of the natural frequency and dissipation with the input excitation amplitude, discontinuities in the frequency response function, and multi-harmonic response via a single harmonic *Corresponding author, E-mail: yzhao@dlut.edu.cn random vibrations, it will be a very interesting topic.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Volterra-PEM approach for random vibration spectrum analysis of nonlinear systems

Zhao

2022

Preprint

View full text Add to dashboard Cite

Based on Volterra series theory, a Volterra-pseudo excitation method (Volterra-PEM) is developed to compute the response power spectral density (PSD) for nonlinear systems under random excitation. Firstly, within the framework of Volterra series theory, an improved pseudo excitation method (PEM) is established for multi-dimensional power spectral density (MPSD) analysis of nonlinear systems. As a generalized PEM for linear random vibration analysis, the Volterra-PEM is used to analyse the response MPSD, which also has a very concise expression. Secondly, in the case of computation difficulties when multi-dimensional integrating from MPSD to PSD, the computational efficiency is improved by converting the multi-dimensional integral into a matrix operation. Finally, as numerical examples, the Volterra-PEM is used to estimate the response PSD for stationary random vibration of a nonlinear spring-damped oscillator and a non-ideal boundary beam with geometrical nonlinearity. Compared with Monte Carlo simulation (MCS), the results show that the proposed method can predict the response PSD for nonlinear systems quickly and accurately under appropriate excitation intensities.

show abstract

Section: Generalized Pem For the Analysis Of Response Mpsdmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Volterra-PEM approach for random vibration spectrum analysis of nonlinear systems

Zhao

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…The application of this regression model has been extensively investigated in SHM as it determines a robust model capable of describing the uncertainties involved in the process. 27 Teimouri et al 28 highlighted the benefits of the GPR model by comparing its performance with neural network models to determine the extension and location of damage in aircraft structures. Different damaged scenarios on a composite airfoil were tested to describe the impact of uncertainties.…”

Section: Introductionmentioning

confidence: 99%

“…The application of this regression model has been extensively investigated in SHM as it determines a robust model capable of describing the uncertainties involved in the process. 27…”

Section: Introductionmentioning

confidence: 99%

Damage quantification using transfer component analysis combined with Gaussian process regression

Yano

Silva

Figueiredo

et al. 2022

Structural Health Monitoring

View full text Add to dashboard Cite

Machine learning methods used in Structural Health Monitoring applications still have generalization difficulties among structures, even when structures are nominally and topologically similar. The data sets present divergences between their probability distributions that do not allow the model’s generalization for damage detection. This issue is even more complex in situations where one wants to quantify damage levels through data sets collected from different structures. Transfer learning methods offer a solution to overcome those limitations, using relevant information from a labeled structure (source domain) to assist the analysis of another structure (target domain) under unknown conditions. Therefore, this paper proposes the use of transfer component analysis to mitigate divergences between the model/structure’s features, and the label consistency requirement is applied in combination with a Gaussian process regression model for damage quantification. The effectiveness of the estimated model improves when the labels consistency between domains is achieved, indicating the current damage level in the structure when the regression model achieves its best performance (lowest error). The proposed methodology is applied on the benchmark data of a three-story building structure from the Los Alamos National Laboratory using the knowledge from its numerical model under several conditions, where the complete information of its behavior is available. The results compare the analysis in the original space and after applying the proposed methodology, demonstrating an improvement of the performance in the damage detection and quantification steps.

show abstract

“…Despite numerous researches in this area, there are still many unanswered questions: which method is the best, how to automate all processes, and so on. According to the application type, there are two main directions of technical diagnostics: monitoring and diagnostics of rotating machinery, [1][2][3][4] and structural health monitoring (SHM), which is defined as a process of implementing a damage detection and characterisation strategy for various engineering structures. [5][6][7][8] According to the used technique, diagnostic methods can be divided into: destructive technique when maximum load or fatigue test of samples are performed 9,10 ; and non-destructive technique when the state of the structure is evaluated without causing damage to it.…”

Section: Introductionmentioning

confidence: 99%

Multi-Bolt looseness detection using a new acoustic emission strategy

Wang

2022

Structural Health Monitoring

View full text Add to dashboard Cite

In mechanical and aerospace engineering, different components are usually integrated together via bolted connections. Compared to the rivet joint and welding joint, the bolted connection is preferred in some cases due to its easy-to-operation and low-cost. However, the bolt self-loosening caused by vibration or other issues (e.g., improper installation and chemical corrosion) may induce severe accidents. Therefore, in this paper, the author proposes a new strategy based on the acoustic emission (AE) technique to detect bolt looseness. To the best of the author’s knowledge, this research is the first attempt to identify multi-bolt looseness via the AE-based method. Particularly, the main contribution is that the author proposes a new shapelet-enhanced AE method that employs a newly developed dual-shapelet networks classifier to discriminate AE waves. The dual-shapelet networks classifier consists of sample-specific shapelets, which is sensitive to the difference among various categories, and category-specific shapelets derived from auxiliary binary classifiers. The objective of category-specific shapelets is to address the imbalanced classification task, that is, discriminating minority categories. Then, the sample-specific shapelets and category-specific shapelets are combined to extract features from AE signals under different multi-bolt looseness cases, and the final classification is achieved by feeding the extracted features into a softmax layer. Finally, the author conducts an experiment to verify the effectiveness of the proposed method. Moreover, by comparing the proposed method’s performance with two baselines, the advantages of the shapelet-enhanced AE method can be demonstrated. Overall, this research demonstrates that the AE technique is valid to characterize friction and collision between asperities on the bolted interface, thus providing a new direction for multi-bolt looseness detection, and the proposed shapelet-enhanced AE method has substantial potential in the field of structural health monitoring (SHM).

show abstract

On the use of the GP-NARX model for predicting hysteresis effects of bolted joint structures

Cited by 12 publications

References 36 publications

A Volterra-PEM approach for random vibration spectrum analysis of nonlinear systems

A Volterra-PEM approach for random vibration spectrum analysis of nonlinear systems

Damage quantification using transfer component analysis combined with Gaussian process regression

Multi-Bolt looseness detection using a new acoustic emission strategy

Contact Info

Product

Resources

About