Abstract
Uncertainties in rotating machines are unavoidable, which affect their parameters and dynamic response. So, instead of employing deterministic models, data-driven meta-modeling techniques which incorporate unpredictability and randomness are necessary for the response variation analysis of rotating systems. The performance of the meta-model relies heavily on the quality and amount of the training dataset. In reality, however, only a tiny amount of high-fidelity data is obtainable from high-dimensional finite element simulation or experimental investigation, although low-cost low-fidelity data may be numerous. The objective of this paper is to develop a novel neural network model for multi-level response prediction by obtaining a high number of low-fidelity data quickly through model order reduction and a limited amount of high-fidelity data correctly from a full-order model. The accuracy of the meta-model is demonstrated by comparing against a classical deep neural network. Two different types of meta-model are established by using two model reduction techniques: Guyan reduction and modified system equivalent reduction expansion process. The performance of the model is demonstrated by employing frequency response variation characterization of a complex rotor as a case example. The results reveal that the multi-fidelity neural network performs better than the low-fidelity frequency response curves alone, which is observed to have a lot of inaccuracies. The deep neural network, on the other hand, is unable to reflect on the dynamic response of the full model. A regression of more than 90% shows that the meta-model has high effectiveness in properly predicting the frequency responses. The mean squared error values for the meta-model are found to be less than 0.1, which is typically regarded as acceptable. Frequency response curves of four test samples are selected at random for comparison. It is observed that the meta-model frequency response moves much closer to the full model than compared to that of the low-fidelity model reduction. The performance resilience of the model is tested by using five different training runs with random data splits. Minor changes in the values of logarithm mean absolute error and logarithm root mean squared error under different training runs show appropriate curve fitting and signify superior accuracy. It is concluded that the multi-fidelity neural network can reach a higher level of accuracy with a limited amount of high-fidelity data. The model effectively identifies both the linear and complex nonlinear correlation between the high-and low-fidelity data, resulting in enhanced efficacy in contrast to state-of-the-art methods.
Access this article
We’re sorry, something doesn't seem to be working properly.
Please try refreshing the page. If that doesn't work, please contact support so we can address the problem.
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig1_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig2_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig3_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig4_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig5_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig6_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig7_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig8_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig9_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig10a_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig10b_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig11_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig12_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig13_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig14_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig15a_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig15b_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig16_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs40430-023-04521-2/MediaObjects/40430_2023_4521_Fig17_HTML.png)
Similar content being viewed by others
Abbreviations
- \([M_{e}^{T} ]_{S}\) :
-
Translatory mass matrix of the shaft
- \([M_{e}^{R} ]_{S}\) :
-
Rotary mass matrix of the shaft
- \({[}G_{e}{]}_{S}\) :
-
Gyroscopic matrix of the shaft
- \([K_{e} ]_{S}\) :
-
Stiffness matrix of the shaft
- \(\{ F_{e} \}_{S}\) :
-
Generalized force vector of the shaft element
- \([M_{e} ]_{D}\) :
-
Mass matrix of the disk
- \({[}G_{e} {]}_{D}\) :
-
Gyroscopic matrix of the disk
- \(\{ F_{e} \}_{D}\) :
-
Generalized force vector of the disk element
- \(q\) :
-
Displacement vector
- \(u\) :
-
Displacement of node in X direction
- \(v\) :
-
Displacement of node in X direction
- \(\xi\) :
-
Rotational Displacement of node about X direction
- \(\psi\) :
-
Rotational Displacement of node about Y direction
- \({[}C{]}_{B}\) :
-
Damping matrix of the bearing
- \([K]_{B}\) :
-
Stiffness matrix of the bearing
- \(\{ F_{e} \}_{B}\) :
-
Generalized force vector of the bearing
- \(\left[ M \right]\) :
-
Global mass matrix
- \(\left[ C \right]\) :
-
Global damping matrix
- \(\left[ K \right]\) :
-
Global stiffness matrix
- \(\{ F\}\) :
-
Global force vector
- \(\omega\) :
-
Angular velocity
- \(\Phi\) :
-
Shear deformation coefficient
- T:
-
Transformation matrix
- \(\varphi\) :
-
Eigenvector matrix
- \(p\) :
-
Number of predominate mode
- \(m\) :
-
Number of master degree of freedom
References
Taplak H, Parlak M (2012) Evaluation of gas turbine rotor dynamic analysis using the finite element method. Measurement 45(5):1089–1097
Brehm M, Deraemaeker A (2015) Uncertainty quantification of dynamic responses in the frequency domain in the context of virtual testing. J Sound Vib 342:303–329
Mao Z, Todd M (2013) Statistical modeling of frequency response function estimation for uncertainty quantification. Mech Syst Signal Process 38(2):333–345
Kroese DP, Taimre T, Botev ZI (2013) Handbook of monte carlo methods. John Wiley & Sons
Liao H (2014) Global resonance optimization analysis of nonlinear mechanical systems: application to the uncertainty quantification problems in rotor dynamics. Commun Nonlinear Sci Numer Simul 19(9):3323–3345
Zhou K, Hegde A, Cao P, Tang J (2017) Design optimization toward alleviating forced response variation in cyclically periodic structure using Gaussian process. J Vib Acoust 139(1):1–14
Dackermann U, Li J, Samali B (2013) Identification of member connectivity and mass changes on a two-storey framed structure using frequency response functions and artificial neural networks. J Sound Vib 332(16):3636–3653
Padil KH, Bakhary N, Abdulkareem M, Li J, Hao H (2020) Non-probabilistic method to consider uncertainties in frequency response function for vibration-based damage detection using artificial neural network. J Sound Vib 467:115069
Acerbi, M., Malvermi, R., Pezzoli, M., Antonacci, F., Sarti, A. and Corradi, R., 2021, June. Interpolation of irregularly sampled frequency response functions using convolutional neural networks. In ICASSP 2021–2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (pp. 950–954). IEEE.
Chen L, Cao J, Wu K, Zhang Z (2022) Application of generalized frequency response functions and improved convolutional neural network to fault diagnosis of heavy-duty industrial robot. Robotics and Comput-Int Manuf 73:102228
Fallahian M, Khoshnoudian F, Talaei S, Meruane V, Shadan F (2018) Experimental validation of a deep neural network—sparse representation classification ensemble method. Struct Design Tall Spec Build 27(15):e1504
Qu, Y., Vogl, G.W. and Wang, Z., (2021) A deep neural network model for learning runtime frequency response function using sensor measurements. In International Manufacturing Science and Engineering Conference (Vol. 85079, p. V002T07A017). American Society of Mechanical Engineers.
Mahmoudabadbozchelou M, Caggioni M, Shahsavari S, Hartt WH, Em Karniadakis G, Jamali S (2021) Data-driven physics-informed constitutive metamodeling of complex fluids: a multifidelity neural network (mfnn) framework. J Rheol 65(2):179–198
Chakraborty S (2021) Transfer learning based multi-fidelity physics informed deep neural network. J Comput Phys 426:109942
De S, Britton J, Reynolds M, Skinner R, Jansen K, Doostan A (2020) On transfer learning of neural networks using bi-fidelity data for uncertainty propagation. Int J Uncertain Quantif 10(6):1–32
Liu D, Wang Y (2019) Multi-fidelity physics-constrained neural network and its application in materials modeling. J Mech Des 141(12):121403
Pichler L, Pradlwarter HJ, Schuëller GI (2009) A mode-based meta-model for the frequency response functions of uncertain structural systems. Comput Struct 87(5–6):332–341
Zhou K, Tang J (2021) Efficient characterization of dynamic response variation using multi-fidelity data fusion through composite neural network. Eng Struct 232:111878
Babaee H, Perdikaris P, Chryssostomidis C, Karniadakis GE (2016) Multi-fidelity modelling of mixed convection based on experimental correlations and numerical simulations. J Fluid Mech 809:895–917
Qu, Z.Q., (2004). Model Order Reduction Techniques with Applications in Finite Element Analysis: With Applications in Finite Element Analysis. Springer Science & Business Media.
Guyan RJ (1965) Reduction of stiffness and mass matrices. AIAA J 3(2):380–380
O'CALLAHAN, J.C., (1989) System equivalent reduction expansion process. In Proc. of the 7th Inter. Modal Analysis Conf., 1989.
Sastry CVS, Mahapatra DR, Gopalakrishnan S, Ramamurthy TS (2003) An iterative system equivalent reduction expansion process for extraction of high frequency response from reduced order finite element model. Comput Methods Appl Mech Eng 192(15):1821–1840
Das AS, Dutt JK (2008) Reduced model of a rotor-shaft system using modified SEREP. Mech Res Commun 35(6):398–407
Saint Martin LB, Mendes RU, Cavalca KL (2020) Model reduction and dynamic matrices extraction from state-space representation applied to rotating machines. Mech Mach Theory 149:103804
Perdikaris P, Raissi M, Damianou A, Lawrence ND, Karniadakis GE (2017) Nonlinear information fusion algorithms for data-efficient multi-fidelity modelling. Proc Royal Soc A 473(2198):20160751
Friswell MI (2010) Dynamics of rotating machines. Cambridge University Press
Khamari DS, Kumar J, Behera SK (2021) Numerical investigation of influence sensitivity of a gas foil bearing parameters on the dynamic coefficients. J Braz Soc Mech Sci Eng 43(3):1–19
Meirovitch L (2010) Fundamentals of vibrations. Waveland Press
Tiwari R (2017) Rotor systems: analysis and identification. CRC Press
Haykin, S., 2010. Neural networks and learning machines, 3/E. Pearson Education India.
Kumar M, Panda D, Sahoo RK, Behera SK (2019) Preliminary design, flow field, and thermal performance analysis of a helium turboexpander: a numerical approach. SN Applied Sciences 1(11):1–30
Fernández-Godino, M.G., Park, C., Kim, N.H. and Haftka, R.T., 2016. Review of multi-fidelity models. arXiv preprint arXiv:1609.07196.
Kennedy MC, O’Hagan A (2000) Predicting the output from a complex computer code when fast approximations are available. Biometrika 87(1):1–13
Meng X, Karniadakis GE (2020) A composite neural network that learns from multi-fidelity data: application to function approximation and inverse PDE problems. J Comput Phys 401:109020
Han ZH, Görtz S (2012) Hierarchical kriging model for variable-fidelity surrogate modeling. AIAA J 50(9):1885–1896
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that there is no conflict of interest to any person or organization.
Additional information
Technical Editor: Aldemir Cavallini Jr.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix
Shaft element matrices
where,
Bearing matrices
Disk matrices
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Khamari, D.S., Behera, S.K. A novel multi-fidelity neural network for response prediction using rotor dynamics and model reduction. J Braz. Soc. Mech. Sci. Eng. 45, 600 (2023). https://doi.org/10.1007/s40430-023-04521-2
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s40430-023-04521-2