Windowed least-squares model reduction for dynamical systems

This work proposes a windowed least-squares (WLS) approach for model-reduction of dynamical systems. The proposed approach sequentially minimizes the time-continuous full-order-model residual within a low-dimensional space-time trial subspace over time windows. The approach comprises a generalization of existing model reduction approaches, as particular instances of the methodology recover Galerkin, least-squares Petrov-Galerkin (LSPG), and space-time LSPG projection. In addition, the approach addresses key deficiencies in existing model-reduction techniques, e.g., the dependence of LSPG and space-time LSPG projection on the time discretization and the exponential growth in time exhibited by a posteriori error bounds for both Galerkin and LSPG projection. We consider two types of space-time trial subspaces within the proposed approach: one that reduces only the spatial dimension of the full-order model, and one that reduces both the spatial and temporal dimensions of the full-order model. For each type of trial subspace, we consider two different solution techniques: direct (i.e., discretize then optimize) and indirect (i.e., optimize then discretize). Numerical experiments conducted using trial subspaces characterized by spatial dimension reduction demonstrate that the WLS approach can yield more accurate solutions with lower space-time residuals than Galerkin and LSPG projection.

[1]  Andrea Walther,et al.  Getting Started with ADOL-C , 2009, Combinatorial Scientific Computing.

[2]  Rémi Abgrall,et al.  Model reduction usingL1-norm minimization as an application to nonlinear hyperbolic problems: Model reduction for hyperbolic problems , 2018 .

[3]  Anil V. Rao,et al.  ( Preprint ) AAS 09-334 A SURVEY OF NUMERICAL METHODS FOR OPTIMAL CONTROL , 2009 .

[4]  C. G. Broyden.  A Class of Methods for Solving Nonlinear Simultaneous Equations , 1965 .

[5]  Juan J. Alonso,et al.  Investigation of non-linear projection for POD based reduced order models for Aerodynamics , 2001 .

[6]  Matthew F. Barone,et al.  Reduced Order Modeling for Prediction and Control of Large-Scale Systems. , 2014 .

[7]  Karen Willcox,et al.  Parametric reduced-order models for probabilistic analysis of unsteady aerodynamic applications , 2007 .

[8]  Karen Willcox,et al.  Model Reduction for Large-Scale Systems with High-Dimensional Parametric Input Space , 2008, SIAM J. Sci. Comput..

[9]  David D. Morrison,et al.  Multiple shooting method for two-point boundary value problems , 1962, CACM.

[10]  Karsten Urban,et al.  A new error bound for reduced basis approximation of parabolic partial differential equations , 2012 .

[11]  Youngsoo Choi,et al.  Space-time least-squares Petrov-Galerkin projection for nonlinear model reduction , 2017, SIAM J. Sci. Comput..

[12]  C. Farhat,et al.  Dimensional reduction of nonlinear finite element dynamic models with finite rotations and energy‐based mesh sampling and weighting for computational efficiency , 2014 .

[13]  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..

[14]  Juan J. Alonso,et al.  Airfoil design optimization using reduced order models based on proper orthogonal decomposition , 2000 .

[15]  Cheng Huang,et al.  Closure of Reacting Flow Reduced-Order Models via the Adjoint Petrov-Galerkin Method , 2019, AIAA Aviation 2019 Forum.

[16]  D. Rovas,et al.  A Posteriori Error Bounds for Reduced-Basis Approximation of Parametrized Noncoercive and Nonlinear Elliptic Partial Differential Equations , 2003 .

[17]  Kevin T. Carlberg,et al.  Online adaptive basis refinement and compression for reduced-order models , 2019, Computer Methods in Applied Mechanics and Engineering.

[18]  Lawrence Sirovich,et al.  Karhunen–Loève procedure for gappy data , 1995 .

[19]  R. Murray,et al.  Model reduction for compressible flows using POD and Galerkin projection , 2004 .

[20]  Qian Wang,et al.  Recurrent neural network closure of parametric POD-Galerkin reduced-order models based on the Mori-Zwanzig formalism , 2020, J. Comput. Phys..

[21]  O. Pironneau,et al.  Characteristic-Galerkin and Galerkin/least-squares space-time formulations for the advection-diffusion equation with time-dependent domains , 1992 .

[22]  N. Nguyen,et al.  An ‘empirical interpolation’ method: application to efficient reduced-basis discretization of partial differential equations , 2004 .

[23]  Laslo T. Diosady,et al.  Development of a High-Order Space-Time Matrix-Free Adjoint Solver , 2016 .

[24]  Youngsoo Choi,et al.  Conservative model reduction for finite-volume models , 2017, J. Comput. Phys..

[25]  V. Rusanov,et al.  On difference schemes of third order accuracy for nonlinear hyperbolic systems , 1970 .

[26]  Clifford T. Mullis,et al.  Synthesis of minimum roundoff noise fixed point digital filters , 1976 .

[27]  K. Duraisamy,et al.  The Adjoint Petrov–Galerkin method for non-linear model reduction , 2018, Computer Methods in Applied Mechanics and Engineering.

[28]  Thomas J. R. Hughes,et al.  A space-time formulation for multiscale phenomena , 1996 .

[29]  Bui-Thanh Tan,et al.  Model-Constrained Optimization Methods for Reduction of Parameterized Large-Scale Systems , 2007 .

[30]  Benjamin Peherstorfer,et al.  Online Adaptive Model Reduction for Nonlinear Systems via Low-Rank Updates , 2015, SIAM J. Sci. Comput..

[31]  Weimin Han,et al.  Convergence of the forward-backward sweep method in optimal control , 2012, Comput. Optim. Appl..

[32]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[33]  Traian Iliescu,et al.  A numerical investigation of velocity-pressure reduced order models for incompressible flows , 2014, J. Comput. Phys..

[34]  Omer San,et al.  Neural network closures for nonlinear model order reduction , 2017, Adv. Comput. Math..

[35]  A. Patera,et al.  Reduced basis approximation and a posteriori error estimation for affinely parametrized elliptic coercive partial differential equations , 2007 .

[36]  Harbir Antil,et al.  Galerkin v. least-squares Petrov-Galerkin projection in nonlinear model reduction , 2015, J. Comput. Phys..

[37]  Paul T. Boggs,et al.  Preserving Lagrangian Structure in Nonlinear Model Reduction with Application to Structural Dynamics , 2014, SIAM J. Sci. Comput..

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

[39]  Zhu Wang,et al.  Reduced-Order Modeling of Complex Engineering and Geophysical Flows: Analysis and Computations , 2012 .

[40]  D. Rovas,et al.  Reduced--Basis Output Bound Methods for Parametrized Partial Differential Equations , 2002 .

[41]  Karsten Urban,et al.  A space-time hp-interpolation-based certified reduced basis method for Burgers' equation , 2014 .

[42]  Donald E. Kirk,et al.  Optimal control theory : an introduction , 1970 .

[43]  Traian Iliescu,et al.  Variational multiscale proper orthogonal decomposition: Navier‐stokes equations , 2012, 1210.7389.

[44]  Earl H. Dowell,et al.  Minimal subspace rotation on the Stiefel manifold for stabilization and enhancement of projection-based reduced order models for the compressible Navier-Stokes equations , 2015, J. Comput. Phys..

[45]  John T. Workman,et al.  Optimal Control Applied to Biological Models , 2007 .

[46]  T. Hughes,et al.  Space-time finite element methods for elastodynamics: formulations and error estimates , 1988 .

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

[48]  Karen Willcox,et al.  A Survey of Projection-Based Model Reduction Methods for Parametric Dynamical Systems , 2015, SIAM Rev..

[49]  Pavel B. Bochev,et al.  Least-Squares Finite Element Methods , 2009, Applied mathematical sciences.

[50]  Qiqi Wang,et al.  Residual Minimizing Model Interpolation for Parameterized Nonlinear Dynamical Systems , 2010, SIAM J. Sci. Comput..

[51]  D. Rovas,et al.  Reliable Real-Time Solution of Parametrized Partial Differential Equations: Reduced-Basis Output Bound Methods , 2002 .

[52]  S. Rebay,et al.  A High-Order Accurate Discontinuous Finite Element Method for the Numerical Solution of the Compressible Navier-Stokes Equations , 1997 .

[53]  T. Hughes,et al.  The Galerkin/least-squares method for advective-diffusive equations , 1988 .

[54]  Charles-Henri Bruneau,et al.  Enablers for robust POD models , 2009, J. Comput. Phys..

[55]  Chi-Wang Shu,et al.  Strong Stability-Preserving High-Order Time Discretization Methods , 2001, SIAM Rev..

[56]  A. Patera,et al.  Certified real‐time solution of the parametrized steady incompressible Navier–Stokes equations: rigorous reduced‐basis a posteriori error bounds , 2005 .

[57]  Saifon Chaturantabut,et al.  Structure-preserving model reduction for nonlinear port-Hamiltonian systems , 2011, IEEE Conference on Decision and Control and European Control Conference.

[58]  Nguyen Ngoc Cuong,et al.  Certified Real-Time Solution of Parametrized Partial Differential Equations , 2005 .

[59]  J. Marsden,et al.  Structure-preserving Model Reduction of Mechanical Systems , 2000 .

[60]  Bruce A. Conway,et al.  A Survey of Methods Available for the Numerical Optimization of Continuous Dynamic Systems , 2011, Journal of Optimization Theory and Applications.

[61]  Jesse Chan,et al.  Entropy stable reduced order modeling of nonlinear conservation laws , 2019, J. Comput. Phys..

[62]  Kookjin Lee,et al.  Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders , 2018, J. Comput. Phys..

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

[64]  Pavel B. Bochev,et al.  Finite Element Methods of Least-Squares Type , 1998, SIAM Rev..

[65]  Y. Saad,et al.  GMRES: a generalized minimal residual algorithm for solving nonsymmetric linear systems , 1986 .

[66]  Traian Iliescu,et al.  Proper orthogonal decomposition closure models for fluid flows: Burgers equation , 2013, 1308.3276.

[67]  Peter Benner,et al.  Space-Time Galerkin POD with Application in Optimal Control of Semilinear Partial Differential Equations , 2018, SIAM J. Sci. Comput..

[68]  Traian Iliescu,et al.  A stabilized proper orthogonal decomposition reduced-order model for large scale quasigeostrophic ocean circulation , 2014, Adv. Comput. Math..

[69]  Juan J. Alonso,et al.  Dynamic Domain Decomposition and Error Correction for Reduced Order Models , 2003 .

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

[71]  Serkan Gugercin,et al.  H2 Model Reduction for Large-Scale Linear Dynamical Systems , 2008, SIAM J. Matrix Anal. Appl..

[72]  Zlatko Drmac,et al.  A New Selection Operator for the Discrete Empirical Interpolation Method - Improved A Priori Error Bound and Extensions , 2015, SIAM J. Sci. Comput..

[73]  B. Moore.  Principal component analysis in linear systems: Controllability, observability, and model reduction , 1981 .