Limited‐memory adaptive snapshot selection for proper orthogonal decomposition

Summary Reduced order models are useful for accelerating simulations in many-query contexts, such as optimization, uncertainty quantification, and sensitivity analysis. However, offline training of reduced order models (ROMs) can have prohibitively expensive memory and floating-point operation costs in high-performance computing applications, where memory per core is limited. To overcome this limitation for proper orthogonal decomposition, we propose a novel adaptive selection method for snapshots in time that limits offline training costs by selecting snapshots according an error control mechanism similar to that found in adaptive time-stepping ordinary differential equation solvers. The error estimator used in this work is related to theory bounding the approximation error in time of proper orthogonal decomposition-based ROMs, and memory usage is minimized by computing the singular value decomposition using a single-pass incremental algorithm. Results for a viscous Burgers' test problem demonstrate convergence in the limit as the algorithm error tolerances go to zero; in this limit, the full-order model is recovered to within discretization error. A parallel version of the resulting method can be used on supercomputers to generate proper orthogonal decomposition-based ROMs, or as a subroutine within hyperreduction algorithms that require taking snapshots in time, or within greedy algorithms for sampling parameter space. Copyright © 2016 John Wiley & Sons, Ltd.

[1]  David Amsallem,et al.  An adaptive and efficient greedy procedure for the optimal training of parametric reduced‐order models , 2015 .

[2]  David J. Lucia,et al.  Reduced Order Modeling for a One-Dimensional Nozzle Flow with Moving Shocks , 2001 .

[3]  William Gropp,et al.  Efficient Management of Parallelism in Object-Oriented Numerical Software Libraries , 1997, SciTools.

[4]  L. Sirovich.  Turbulence and the dynamics of coherent structures. I. Coherent structures , 1987 .

[5]  Stefan Volkwein,et al.  Galerkin Proper Orthogonal Decomposition Methods for a General Equation in Fluid Dynamics , 2002, SIAM J. Numer. Anal..

[6]  Paul Van Dooren,et al.  Recursive Calculation of Dominant Singular Subspaces , 2003, SIAM J. Matrix Anal. Appl..

[7]  A. Patera,et al.  A posteriori error bounds for reduced-basis approximations of parametrized parabolic partial differential equations , 2005 .

[8]  Zhu Wang,et al.  Are the Snapshot Difference Quotients Needed in the Proper Orthogonal Decomposition? , 2013, SIAM J. Sci. Comput..

[9]  P. Holmes,et al.  The Proper Orthogonal Decomposition in the Analysis of Turbulent Flows , 1993 .

[10]  D. Keyes,et al.  Jacobian-free Newton-Krylov methods: a survey of approaches and applications , 2004 .

[11]  Oliver Lass,et al.  Adaptive POD basis computation for parametrized nonlinear systems using optimal snapshot location , 2014, Comput. Optim. Appl..

[12]  R. H. W. HOPPE,et al.  Snapshot location by error equilibration in proper orthogonal decomposition for linear and semilinear parabolic partial differential equations , 2014, J. Num. Math..

[13]  Muruhan Rathinam,et al.  A New Look at Proper Orthogonal Decomposition , 2003, SIAM J. Numer. Anal..

[14]  Danny C. Sorensen,et al.  Nonlinear Model Reduction via Discrete Empirical Interpolation , 2010, SIAM J. Sci. Comput..

[15]  Charbel Farhat,et al.  The GNAT method for nonlinear model reduction: Effective implementation and application to computational fluid dynamics and turbulent flows , 2012, J. Comput. Phys..

[16]  M. Brand,et al.  Fast low-rank modifications of the thin singular value decomposition , 2006 .

[17]  Stefan Volkwein,et al.  Galerkin proper orthogonal decomposition methods for parabolic problems , 2001, Numerische Mathematik.

[18]  Matthew Brand,et al.  Incremental Singular Value Decomposition of Uncertain Data with Missing Values , 2002, ECCV.

[19]  Adrian Sandu,et al.  Comparison of POD reduced order strategies for the nonlinear 2D shallow water equations , 2014, International Journal for Numerical Methods in Fluids.

[20]  Danny C. Sorensen,et al.  A Posteriori Error Estimation for DEIM Reduced Nonlinear Dynamical Systems , 2014, SIAM J. Sci. Comput..

[21]  Kevin Carlberg,et al.  The ROMES method for statistical modeling of reduced-order-model error , 2014, SIAM/ASA J. Uncertain. Quantification.

[22]  Danny C. Sorensen,et al.  A State Space Error Estimate for POD-DEIM Nonlinear Model Reduction , 2012, SIAM J. Numer. Anal..

[23]  Paul Van Dooren,et al.  Model Reduction of Time-Varying Systems , 2005 .

[24]  C. Farhat,et al.  Efficient non‐linear model reduction via a least‐squares Petrov–Galerkin projection and compressive tensor approximations , 2011 .

[25]  B. Haasdonk,et al.  REDUCED BASIS METHOD FOR FINITE VOLUME APPROXIMATIONS OF PARAMETRIZED LINEAR EVOLUTION EQUATIONS , 2008 .

[26]  Bernard Haasdonk,et al.  A training set and multiple bases generation approach for parameterized model reduction based on adaptive grids in parameter space , 2011 .

[27]  Linda R. Petzold,et al.  Approved for public release; further dissemination unlimited Error Estimation for Reduced Order Models of Dynamical Systems ∗ , 2003 .

[28]  John R. Singler,et al.  New POD Error Expressions, Error Bounds, and Asymptotic Results for Reduced Order Models of Parabolic PDEs , 2014, SIAM J. Numer. Anal..

[29]  A. Hindmarsh,et al.  CVODE, a stiff/nonstiff ODE solver in C , 1996 .

[30]  Carol S. Woodward,et al.  Enabling New Flexibility in the SUNDIALS Suite of Nonlinear and Differential/Algebraic Equation Solvers , 2020, ACM Trans. Math. Softw..

[31]  Charbel Farhat,et al.  Construction of Parametrically-Robust CFD-Based Reduced-Order Models for PDE-Constrained Optimization , 2013 .

[32]  Charbel Farhat,et al.  Nonlinear Model Reduction for CFD Problems Using Local Reduced Order Bases , 2012 .

[33]  D. Ryckelynck,et al.  A priori hyperreduction method: an adaptive approach , 2005 .

[34]  Kevin Carlberg,et al.  Adaptive h‐refinement for reduced‐order models , 2014, ArXiv.

[35]  K. Kunisch,et al.  Optimal snapshot location for computing POD basis functions , 2010 .

[36]  Sivasankaran Rajamanickam,et al.  EFFICIENT ALGORITHMS FOR SPARSE SINGULAR VALUE DECOMPOSITION , 2009 .

[37]  Liqian Peng,et al.  An Online Manifold Learning Approach for Model Reduction of Dynamical Systems , 2012, SIAM J. Numer. Anal..

[38]  Paul Van Dooren,et al.  Low-rank incremental methods for computing dominant singular subspaces☆ , 2012 .

[39]  C. Farhat,et al.  A low‐cost, goal‐oriented ‘compact proper orthogonal decomposition’ basis for model reduction of static systems , 2011 .

[40]  David J. Lucia,et al.  Reduced Order Modeling For High Speed Flows with Moving Shocks , 2001 .

[41]  P. Sagaut,et al.  Towards an adaptive POD/SVD surrogate model for aeronautic design , 2011 .