×

An optimized dynamic mode decomposition model robust to multiplicative noise. (English) Zbl 1515.37093

Summary: Dynamic mode decomposition (DMD) is an efficient tool for decomposing spatio-temporal data into a set of low-dimensional modes, yielding the oscillation frequencies and the growth rates of physically significant modes. In this paper, we propose a novel DMD model that can be used for dynamical systems affected by multiplicative noise. We first derive a maximum a posteriori (MAP) estimator for the data-based model decomposition of a linear dynamical system corrupted by certain multiplicative noise. Applying penalty relaxation to the MAP estimator, we obtain the proposed DMD model whose epigraphical limits are the MAP estimator and the conventional optimized DMD model. We also propose an efficient alternating gradient descent method for solving the proposed DMD model and analyze its convergence behavior. The proposed model is demonstrated on both the synthetic data and the numerically generated one-dimensional combustor data and is shown to have superior reconstruction properties compared to state-of-the-art DMD models. Considering that multiplicative noise is ubiquitous in numerous dynamical systems, the proposed DMD model opens up new possibilities for accurate data-based modal decomposition.

MSC:

37M10 Time series analysis of dynamical systems
49M37 Numerical methods based on nonlinear programming
49R05 Variational methods for eigenvalues of operators
65P99 Numerical problems in dynamical systems

Software:

RobustDMD; tvreg

References:

[1] Amin, M. F., Amin, M. I., Al-Nuaimi, A. Y. H., and Murase, K., Wirtinger calculus based gradient descent and Levenberg-Marquardt learning algorithms in complex-valued neural networks, in International Conference on Neural Information Processing, , Springer, New York, 2011, pp. 550-559.
[2] Askham, T. and Kutz, J. N., Variable projection methods for an optimized dynamic mode decomposition, SIAM J. Appl. Dyn. Syst., 17 (2018), pp. 380-416. · Zbl 1384.37122
[3] Askham, T., Zheng, P., Aravkin, A., and Kutz, J. N., Robust and scalable methods for the dynamic mode decomposition, SIAM J. Appl. Dyn. Syst., 21 (2022), pp. 60-79. · Zbl 1490.37100
[4] Aubert, G. and Aujol, J.-F., A variational approach to removing multiplicative noise, SIAM J. Appl. Math., 68 (2008), pp. 925-946. · Zbl 1151.68713
[5] Beck, A., On the convergence of alternating minimization for convex programming with applications to iteratively reweighted least squares and decomposition schemes, SIAM J. Optim., 25 (2015), pp. 185-209. · Zbl 1358.90094
[6] Beck, A. and Teboulle, M., A fast iterative shrinkage-thresholding algorithm for linear inverse problems, SIAM J. Imaging Sci., 2 (2009), pp. 183-202. · Zbl 1175.94009
[7] Beck, A. and Tetruashvili, L., On the convergence of block coordinate descent type methods, SIAM J. Optim., 23 (2013), pp. 2037-2060. · Zbl 1297.90113
[8] Bolte, J., Sabach, S., and Teboulle, M., Proximal alternating linearized minimization for nonconvex and nonsmooth problems, Math. Program., 146 (2014), pp. 459-494. · Zbl 1297.90125
[9] Brand, H. R., Kai, S., and Wakabayashi, S., External noise can suppress the onset of spatial turbulence, Phys. Rev. Lett., 54 (1985), 555.
[10] Bredies, K., Kunisch, K., and Pock, T., Total generalized variation, SIAM J. Imaging Sci, 3 (2010), pp. 492-526. · Zbl 1195.49025
[11] Brezis, H., Functional Analysis, Sobolev Spaces and Partial Differential Equations, Springer, New York, 2010. · Zbl 1218.46002
[12] Brunton, B. W., Johnson, L. A., Ojemann, J. G., and Kutz, J. N., Extracting spatial-temporal coherent patterns in large-scale neural recordings using dynamic mode decomposition, J. Neurosci. Methods, 258 (2016), pp. 1-15.
[13] Calatroni, L. and Chambolle, A., Backtracking strategies for accelerated descent methods with smooth composite objectives, SIAM J. Optim., 29 (2019), pp. 1772-1798. · Zbl 1427.90215
[14] Chan, T. F. and Esedoḡlu, S., Aspects of total variation regularized \({L}^1\) function approximation, SIAM J. Appl. Math., 65 (2005), pp. 1817-1837. · Zbl 1096.94004
[15] Chen, K. K., Tu, J. H., and Rowley, C. W., Variants of dynamic mode decomposition: Boundary condition, Koopman, and Fourier analyses, J. Nonlinear Sci., 22 (2012), pp. 887-915. · Zbl 1259.35009
[16] Clavin, P., Kim, J. S., and Williams, F. A., Turbulence-induced noise effects on high-frequency combustion instabilities, Combust. Sci. Technol., 96 (1994), pp. 61-84.
[17] Dawson, S. T. M., Hemati, M. S., Williams, M. O., and Rowley, C. W., Characterizing and correcting for the effect of sensor noise in the dynamic mode decomposition, Exp. Fluids, 57 (2016), 42.
[18] Fox, R. F., James, G. E., and Roy, R., Laser with a fluctuating pump: Intensity correlations of a dye laser, Phys. Rev. Lett., 52 (1984), pp. 1778-1781.
[19] Getreuer, P., Rudin-Osher-Fatemi total variation denoising using split Bregman, IPOL J. Image Process. Online, 2 (2012), pp. 74-95.
[20] Golub, G. H. and Pereyra, V., The differentiation of pseudo-inverses and nonlinear least squares problems whose variables separate, SIAM J. Numer. Anal., 10 (1973), pp. 413-432. · Zbl 0258.65045
[21] Golub, G. H. and Van Loan, C. F., Matrix Computations, Johns Hopkins University Press, Baltimore, 2013. · Zbl 1268.65037
[22] Granwehr, J., Multiplicative or \(t_1\) noise in NMR spectroscopy, Appl. Magn. Reson., 32 (2007), pp. 113-156.
[23] Guan, Y., Gupta, V., Wan, M., and Li, L. K. B., Forced synchronization of quasiperiodic oscillations in a thermoacoustic system, J. Fluid Mech., 879 (2019), pp. 390-421. · Zbl 1430.76420
[24] Guéniat, F., Mathelin, L., and Pastur, L. R., A dynamic mode decomposition approach for large and arbitrarily sampled systems, Phys. Fluids, 27 (2015), 025113.
[25] Gupta, V., Saurabh, A., Paschereit, C. O., and Kabiraj, L., Numerical results on noise-induced dynamics in the subthreshold regime for thermoacoustic systems, J. Sound Vib., 390 (2017), pp. 55-66.
[26] Hemati, M. S., Rowley, C. W., Deem, E. A., and Cattafesta, L. N., De-biasing the dynamic mode decomposition for applied Koopman spectral analysis of noisy datasets, Theor. Comput. Fluid Dyn., 31 (2017), pp. 349-368.
[27] Holmes, P., Lumley, J. L., Berkooz, G., and Rowley, C. W., Turbulence, Coherent Structures, Dynamical Systems and Symmetry, 2nd ed., Cambridge University Press, Cambridge, UK, 2012. · Zbl 1251.76001
[28] Horsthemke, W., Noise induced transitions, in Non-equilibrium Dynamics in Chemical Systems, Springer, New York, 1984, pp. 150-160.
[29] Ilak, M. and Rowley, C. W., Modeling of transitional channel flow using balanced proper orthogonal decomposition, Phys. Fluids, 20 (2008), 034103. · Zbl 1182.76341
[30] Kreutz-Delgado, K., The Complex Gradient Operator and the CR-calculus, preprint, https://arxiv.org/abs/0906.4835, 2009.
[31] Le, T., Chartrand, R., and Asaki, T. J., A variational approach to reconstructing images corrupted by Poisson noise, J. Math. Imaging Vis., 27 (2007), pp. 257-263.
[32] Lee, C.-O., Park, E.-H., and Park, J., A finite element approach for the dual Rudin-Osher-Fatemi model and its nonoverlapping domain decomposition methods, SIAM J. Sci. Comput., 41 (2019), pp. B205-B228. · Zbl 1412.65220
[33] Lee, C.-O. and Park, J., Recent advances in domain decomposition methods for total variation minimization, J. Korean Soc. Ind. Appl. Math., 24 (2020), pp. 161-197. · Zbl 1453.65429
[34] Lee, M., System Identification near a Hopf Bifurcation Via the Noise-Induced Dynamics in the Fixed-Point Regime, Ph.D. thesis, Hong Kong University of Science and Technology, 2020.
[35] Lee, M., Early warning detection of thermoacoustic instability using three-dimensional complexity-entropy causality space, Exp. Therm. Fluid Sci., 130 (2021), 110517.
[36] Lee, M., Guan, Y., Gupta, V., and Li, L. K. B., Input-output system identification of a thermoacoustic oscillator near a Hopf bifurcation using only fixed-point data, Phys. Rev. E, 101 (2020), 013102.
[37] Leroux, R. and Cordier, L., Dynamic mode decomposition for non-uniformly sampled data, Exp. Fluids, 57 (2016), 94.
[38] Li, F., Ng, M. K., and Shen, C., Multiplicative noise removal with spatially varying regularization parameters, SIAM J. Imaging Sci., 3 (2010), pp. 1-20. · Zbl 1185.65067
[39] Li, H. and Adalı, T., Complex-valued adaptive signal processing using nonlinear functions, EURASIP J. Adv. Signal Process., 2008, (2008), 765615. · Zbl 1184.94113
[40] Lieuwen, T. and Banaszuk, A., Background noise effects on combustor stability, J. Propuls. Power, 21 (2005), pp. 25-31.
[41] Lindqvist, B. H. and Taraldsen, G., On the proper treatment of improper distributions, J. Statist. Plann. Inference, 195 (2018), pp. 93-104. · Zbl 1383.62016
[42] Lores, M. E. and Zinn, B. T., Nonlinear longitudinal combustion instability in rocket motors, Combust. Sci. Technol., 7 (1973), pp. 245-256.
[43] Luong, H. T., Wang, Y., Sung, H.-G., and Sohn, C. H., A comparative study of dynamic mode decomposition methods for mode identification in a cryogenic swirl injector, J. Sound Vib., 503 (2021), 116108.
[44] Marquardt, D. W., An algorithm for least-squares estimation of nonlinear parameters, J. SIAM, 11 (1963), pp. 431-441. · Zbl 0112.10505
[45] Mezić, I., Spectral properties of dynamical systems, model reduction and decompositions, Nonlinear Dyn., 41 (2005), pp. 309-325. · Zbl 1098.37023
[46] Nikolova, M., A variational approach to remove outliers and impulse noise, J. Math. Imaging Vis., 20 (2004), pp. 99-120. · Zbl 1366.94065
[47] Orchini, A., Rigas, G., and Juniper, M. P., Weakly nonlinear analysis of thermoacoustic bifurcations in the Rijke tube, J. Fluid Mech., 805 (2016), pp. 523-550. · Zbl 1454.80012
[48] Park, J., Fast gradient methods for uniformly convex and weakly smooth problems, Adv. Comput. Math., 48 (2022), 34. · Zbl 1489.90125
[49] Proctor, J. L. and Eckhoff, P. A., Discovering dynamic patterns from infectious disease data using dynamic mode decomposition, Int. Health., 7 (2015), pp. 139-145.
[50] Rathinam, M. and Petzold, L. R., A new look at proper orthogonal decomposition, SIAM J. Numer. Anal., 41 (2003), pp. 1893-1925. · Zbl 1053.65106
[51] Rockafellar, R. T. and Wets, R. J.-B., Variational Analysis, , Springer, Berlin, 2009.
[52] Rowley, C. W., Model reduction for fluids, using balanced proper orthogonal decomposition, Int. J. Bifurcat. Chaos, 15 (2005), pp. 997-1013. · Zbl 1140.76443
[53] Rudin, L. I., Osher, S., and Fatemi, E., Nonlinear total variation based noise removal algorithms, Phys. D, 60 (1992), pp. 259-268. · Zbl 0780.49028
[54] Schmid, P. J., Dynamic mode decomposition of numerical and experimental data, J. Fluid. Mech., 656 (2010), pp. 5-28. · Zbl 1197.76091
[55] Shen, J., Kang, S. H., and Chan, T. F., Euler’s elastica and curvature-based inpainting, SIAM J. Appl. Math., 63 (2003), pp. 564-592. · Zbl 1028.68185
[56] Short, R., Mandel, L., and Roy, R., Correlation functions of a dye laser: Comparison between theory and experiment, Phys. Rev. Lett., 49 (1982), pp. 647-650.
[57] Sirovich, L., Turbulence and the dynamics of coherent structures. I. coherent structuress, Quart. Appl. Math., 45 (1987), pp. 561-571. · Zbl 0676.76047
[58] Taraldsen, G. and Lindqvist, B. H., Improper priors are not improper, Amer. Statist., 64 (2010), pp. 154-158.
[59] Tu, J. H., Rowley, C. W., Luchtenburg, D. M., Brunton, S. L., and Kutz, J. N., On dynamic mode decomposition: Theory and applications, J. Comput. Dyn., 1 (2014), pp. 391-421. · Zbl 1346.37064
[60] Waugh, I. C. and Juniper, M. P., Triggering in a thermoacoustic system with stochastic noise, Int. J. Spray Combust. Dyn., 3 (2011), pp. 225-241.
[61] Wei, M., Perturbation of the least squares problem, Linear Algebra Appl., 141 (1990), pp. 177-182. · Zbl 0711.15005
[62] Yeo, D. and Lee, C.-O., Variational shape prior segmentation with an initial curve based on image registration technique, Image Vis. Comput., 94 (2020), 103865.
[63] Zinn, B. T. and Lores, M. E., Application of the Galerkin method in the solution of non-linear axial combustion instability problems in liquid rockets, Combust. Sci. Technol., 4 (1971), pp. 269-278.
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.