Data-Driven Time Parallelism via Forecasting
Abstract
This work proposes a data-driven method for enabling the efficient, stable time-parallel numerical solution of systems of ordinary differential equations (ODEs). The method assumes that low-dimensional bases that accurately capture the time evolution of the dynamical-system state are available. The method adopts the parareal framework for time parallelism, which is defined by an initialization method, a coarse propagator that advances solutions on a coarse time grid, and a fine propagator that operates on an underlying fine time grid. Rather than employing usual approaches for initialization and coarse propagation, we propose novel data-driven techniques that leverage the available time-evolution bases. The coarse propagator is defined by a forecast (proposed in [K. Carlberg, J. Ray, and B. van Bloemen Waanders, Comput. Methods Appl. Mech. Engrg., 289 (2015), pp. 79--103]) applied locally within each coarse time interval, which comprises the following steps: (1) apply the fine propagator for a small number of time steps, (2) approximate the state over the entire coarse time interval using gappy proper orthogonal decomposition (POD) with the local time-evolution bases, and (3) select the approximation at the end of the time interval as the propagated state. We also propose both local-forecast and global-forecast initialization techniques. The method is particularly well suited for POD-based reduced-order models (ROMs), as the time-evolution bases can be extracted from readily available data, i.e., the right singular vectors arising during POD computation. In addition to performing analyses related to the method's accuracy, speedup, stability, and convergence, we also numerically demonstrate the method's performance. Here, numerical experiments on ROMs for a nonlinear convection--reaction problem demonstrate the method's ability to realize near-ideal speedups.
Keywords
MSC codes
Get full access to this article
View all available purchase options and get full access to this article.
Supplementary Material
PLEASE NOTE: These supplementary files have not been peer-reviewed.
Index of Supplementary Materials
Title of paper: Data-driven time parallelism via forecasting
Authors: K. Carlberg, L. Brencher, B. Haasdonk, A. Barth
File: M117436SupMat.pdf
Type: PDF
Contents: Proofs of the analytical results presented in the main manuscript (Section S1), a numerical investigation of the dependence of the quantities appearing in Lemma 4.9 on the time discretization (Section S2), theoretical performance improvements when forecasting is also employed for defining the initial guess in the Newton solver (Section S3), supporting convergence-analysis results (Section S4), an illustration in the particular case of of parameterized linear ODEs in which the proposed method yields an ideal predictive coarse propagator (Section S5), and figures that provide additional interpretation of the numerical experiments (Section S6). Additionally, Section S7 provides a list of notation of the main article file.
References
1.
A. Axelsson and J. Verwer, Boundary value techniques for initial value problems in ordinary differential equations, Math. Comp., 45 (1985), pp. 153--171.
2.
L. Baffico, S. Bernard, Y. Maday, G. Turinici, and G. Zérah, Parallel-in-time molecular-dynamics simulations, Phys. Rev. E, 66 (2002), 057701.
3.
G. Bal, On the convergence and the stability of the parareal algorithm to solve partial differential equations, in Domain Decomposition Methods in Science and Engineering, Springer, New York, 2005, pp. 425--432.
4.
G. Bal and Y. Maday, A “parareal” time discretization for non-linear PDEs with application to the pricing of an American put, in Recent Developments in Domain Decomposition Methods, Springer, New York, 2002, pp. 189--202.
5.
M. Barrault, Y. Maday, N. C. Nguyen, and A. T. Patera, An `empirical interpolation' method: Application to efficient reduced-basis discretization of partial differential equations, C. R. Math. Acad. Sci. Paris, 339 (2004), pp. 667--672.
6.
A. Bellen and M. Zennaro, Parallel algorithms for initial-value problems for difference and differential equations, J. Comput. Appl. Math., 25 (1989), pp. 341--350.
7.
A. Blouza, L. Boudin, and S. M. Kaber, Parallel in time algorithms with reduction methods for solving chemical kinetics, Commun. Appl. Math. Comput. Sci., 5 (2011), pp. 241--263.
8.
K. Carlberg, Model Reduction of Nonlinear Mechanical Systems via Optimal Projection and Tensor Approximation, Ph.D. thesis, Stanford University, Stanford, CA, 2011.
9.
K. Carlberg, M. Barone, and H. Antil, Galerkin v. least-squares Petrov--Galerkin projection in nonlinear model reduction, J. Comput. Phys., 330 (2017), pp. 693--734.
10.
K. Carlberg, C. Farhat, and C. Bou-Mosleh, Efficient non-linear model reduction via a least-squares Petrov--Galerkin projection and compressive tensor approximations, Internat. J. Numer. Methods Engrg., 86 (2011), pp. 155--181.
11.
K. Carlberg, C. Farhat, J. Cortial, and D. Amsallem, The GNAT method for nonlinear model reduction: Effective implementation and application to computational fluid dynamics and turbulent flows, J. Comput. Phys., 242 (2013), pp. 623--647.
12.
K. Carlberg, J. Ray, and B. van Bloemen Waanders, Decreasing the temporal complexity for nonlinear, implicit reduced-order models by forecasting, Comput. Methods Appl. Mech. Engrg., 289 (2015), pp. 79--103.
13.
S. Chaturantabut and D. C. Sorensen, Nonlinear model reduction via discrete empirical interpolation, SIAM J. Sci. Comput., 32 (2010), pp. 2737--2764, https://doi.org/10.1137/090766498.
14.
F. Chen, J. S. Hesthaven, and X. Zhu, On the use of reduced basis methods to accelerate and stabilize the parareal method, in Reduced Order Methods for Modeling and Computational Reduction, Springer, New York, 2014, pp. 187--214.
15.
Y. Choi and K. Carlberg, Space--Time Least-Squares Petrov--Galerkin Projection for Nonlinear Model Reduction, preprint, https://arxiv.org/abs/1703.04560, 2017.
16.
M. Drohmann, B. Haasdonk, and M. Ohlberger, Reduced basis approximation for nonlinear parametrized evolution equations based on empirical operator interpolation, SIAM J. Sci. Comput., 34 (2012), pp. A937--A969, https://doi.org/10.1137/10081157X.
17.
R. Everson and L. Sirovich, Karhunen--Loève procedure for gappy data, J. Optical Soc. Amer. A, 12 (1995), pp. 1657--1664.
18.
R. D. Falgout, S. Freidhoff, T. V. Kolev, S. P. MacLachlan, and J. B. Schroder, Parallel time integration with multigrid, SIAM J. Sci. Comput., 36 (2014), pp. C635--C661, https://doi.org/10.1137/130944230.
19.
C. Farhat and M. Chandesris, Time-decomposed parallel time-integrators: Theory and feasibility studies for fluid, structure, and fluid--structure applications, Internat. J. Numer. Methods Engrg., 58 (2003), pp. 1397--1434.
20.
C. Farhat, J. Cortial, C. Dastillung, and H. Bavestrello, Time-parallel implicit integrators for the near-real-time prediction of linear structural dynamic responses, Internat. J. Numer. Methods Engrg., 67 (2006), pp. 697--724.
21.
P. F. Fischer, F. Hecht, and Y. Maday, A parareal in time semi-implicit approximation of the Navier--Stokes equations, in Domain Decomposition Methods in Science and Engineering, Springer, New York, 2005, pp. 433--440.
22.
M. J. Gander, Overlapping Schwarz for parabolic problems, in Proceedings of the 9th International Conference on Domain Decomposition Methods (Boulder, CO, 1997), AMS, Providence, RI, 1998, pp. 425--431.
23.
M. J. Gander and E. Hairer, Nonlinear convergence analysis for the parareal algorithm, in Domain Decomposition Methods in Science and Engineering XVII, Springer, Berlin, 2008, pp. 45--56.
24.
M. J. Gander and S. Vandewalle, Analysis of the parareal time-parallel time-integration method, SIAM J. Sci. Comput., 29 (2007), pp. 556--578, https://doi.org/10.1137/05064607X.
25.
D. Guibert and D. Tromeur-Dervout, Adaptive parareal for systems of ODEs, in Domain Decomposition Methods in Science and Engineering XVI, Springer, Berlin, 2007, pp. 587--594.
26.
W. Hackbusch, Parabolic multigrid methods, in Computing Methods in Applied Sciences and Engineering, VI (Versailles, France, 1983), R. Glowinski and J. Lions, eds., North--Holland, Amsterdam, 1984, pp. 189--197.
27.
J. Lions, Y. Maday, and G. Turinici, A “parareal” in time discretization of PDE's, C. R. Acad. Sci. Paris Sér. I Math., 332 (2001), pp. 661--668 (in French).
28.
C. Lubich and A. Ostermann, Multigrid dynamic iteration for parabolic equations, BIT, 27 (1987), pp. 216--234.
29.
Y. Maday and E. M. Rønquist, Parallelization in time through tensor-product space--time solvers, C. R. Math. Acad. Sci. Paris, 346 (2008), pp. 113--118.
30.
Y. Maday and G. Turinici, Parallel in time algorithms for quantum control: Parareal time discretization scheme, Internat. J. Quantum Chem., 93 (2003), pp. 223--228.
31.
M. Minion, A hybrid parareal spectral deferred corrections method, Commun. Appl. Math. Comput. Sci., 5 (2011), pp. 265--301.
32.
W. L. Miranker and W. Liniger, Parallel methods for the numerical integration of ordinary differential equations, Math. Comp., 21 (1967), pp. 303--320.
33.
A. S. Nielsen, Feasibility Study of the Parareal Algorithm, Masters thesis, Technical University of Denmark, DTU Informatic, Kongens Lyngby, Denmark, 2012.
34.
J. Nievergelt, Parallel methods for integrating ordinary differential equations, Comm. ACM, 7 (1964), pp. 731--733.
35.
M. J. Rewienski, A Trajectory Piecewise-Linear Approach to Model Order Reduction of Nonlinear Dynamical Systems, Ph.D. thesis, Massachusetts Institute of Technology, Cambridge, MA, 2003.
36.
P. Saha, J. Stadel, and S. Tremaine, A parallel integration method for solar system dynamics, Astron. J., 114 (1997), pp. 409--415.
37.
D. A. D. Subcommittee, Synergistic Challenges in Data-Intensive Science and Exascale Computing, DOE Office of Science, Washington, DC, 2013.
38.
S. Vandewalle, Parallel Multigrid Waveform Relaxation for Parabolic Problems, Vieweg+Teubner Verlag, Berlin, 1993.
39.
N. Vannieuwenhoven, R. Vandebril, and K. Meerbergen, A new truncation strategy for the higher-order singular value decomposition, SIAM J. Sci. Comput., 34 (2012), pp. A1027--A1052, https://doi.org/10.1137/110836067.
40.
D. E. Womble, A time-stepping algorithm for parallel computers, SIAM J. Sci. Statist. Comput., 11 (1990), pp. 824--837, https://doi.org/10.1137/0911049.
Information & Authors
Information
Published In
SIAM Journal on Scientific Computing
Pages: B466 - B496
DOI: 10.1137/18M1174362
ISSN (online): 1095-7197
Copyright
© 2019, Society for Industrial and Applied Mathematics.
History
Submitted: 13 March 2018
Accepted: 22 March 2019
Published online: 23 May 2019
Keywords
MSC codes
Authors
Funding Information
Sandia National Laboratories https://doi.org/10.13039/100006234 : DE-NA0003525
Funding Information
Deutsche Forschungsgemeinschaft https://doi.org/10.13039/501100001659 : EXC 310/2
Metrics & Citations
Metrics
Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
Cited By
- Projection-based model reduction of dynamical systems using space–time subspace and machine learningComputer Methods in Applied Mechanics and Engineering, Vol. 389 | 1 Feb 2022
- An adaptive parareal algorithmJournal of Computational and Applied Mathematics, Vol. 377 | 1 Oct 2020
- Memory embedded non-intrusive reduced order modeling of non-ergodic flowsPhysics of Fluids, Vol. 31, No. 12 | 23 December 2019