Preconditioning Navier–Stokes control using multilevel sequentially semiseparable matrix computations

In this article, we study preconditioning techniques for the control of the Navier–Stokes equation, where the control only acts on a few parts of the domain. Optimization, discretization, and linearization of the control problem results in a generalized linear saddle‐point system. The Schur complement for the generalized saddle‐point system is very difficult or even impossible to approximate, which prohibits satisfactory performance of the standard block preconditioners. We apply the multilevel sequentially semiseparable (MSSS) preconditioner to the underlying system. Compared with standard block preconditioning techniques, the MSSS preconditioner computes an approximate factorization of the global generalized saddle‐point matrix up to a prescribed accuracy in linear computational complexity. This in turn gives parameter independent convergence for MSSS preconditioned Krylov solvers. We use a simplified wind farm control example to illustrate the performance of the MSSS preconditioner. We also compare the performance of the MSSS preconditioner with the performance of the state‐of‐the‐art preconditioning techniques. Our results show the superiority of the MSSS preconditioning techniques to standard block preconditioning techniques for the control of the Navier–Stokes equation.

[1]  Andrew J. Wathen,et al.  Preconditioning for boundary control problems in incompressible fluid dynamics , 2015, Numer. Linear Algebra Appl..

[2]  Sabine Le Borne,et al.  H-matrix Preconditioners in Convection-Dominated Problems , 2005, SIAM J. Matrix Anal. Appl..

[3]  Edmond Chow,et al.  SMASH: Structured matrix approximation by separation and hierarchy , 2017, Numer. Linear Algebra Appl..

[4]  Andrew J. Wathen,et al.  A Preconditioner for the Steady-State Navier-Stokes Equations , 2002, SIAM J. Sci. Comput..

[5]  Joseph E. Pasciak,et al.  A convergence analysis for a sweeping preconditioner for block tridiagonal systems of linear equations , 2015, Numer. Linear Algebra Appl..

[6]  Martin Berggren,et al.  Numerical Solution of a Flow-Control Problem: Vorticity Reduction by Dynamic Boundary Action , 1998, SIAM J. Sci. Comput..

[7]  Jean-Yves L'Excellent,et al.  Improving Multifrontal Methods by Means of Block Low-Rank Representations , 2015, SIAM J. Sci. Comput..

[8]  M. Verhaegen,et al.  Evaluation of multilevel sequentially semiseparable preconditioners on computational fluid dynamics benchmark problems using Incompressible Flow and Iterative Solver Software , 2018 .

[9]  George Biros,et al.  Multigrid Algorithms for Inverse Problems with Linear Parabolic PDE Constraints , 2008, SIAM J. Sci. Comput..

[10]  Howard C. Elman,et al.  IFISS: A Computational Laboratory for Investigating Incompressible Flow Problems , 2014, SIAM Rev..

[11]  W. Hackbusch.  A Sparse Matrix Arithmetic Based on $\Cal H$-Matrices. Part I: Introduction to ${\Cal H}$-Matrices , 1999, Computing.

[12]  S. Börm,et al.  ℋ︁‐LU factorization in preconditioners for augmented Lagrangian and grad‐div stabilized saddle point systems , 2012 .

[13]  Martin Stoll,et al.  A Low-Rank in Time Approach to PDE-Constrained Optimization , 2015, SIAM J. Sci. Comput..

[14]  Jennifer Annoni,et al.  Analysis of axial‐induction‐based wind plant control using an engineering and a high‐order wind plant model , 2016 .

[15]  Alfio Borzì,et al.  Computational Optimization of Systems Governed by Partial Differential Equations , 2012, Computational science and engineering.

[16]  Joachim Schöberl,et al.  Symmetric Indefinite Preconditioners for Saddle Point Problems with Applications to PDE-Constrained Optimization Problems , 2007, SIAM J. Matrix Anal. Appl..

[17]  Karl Kunisch,et al.  Second Order Methods for Optimal Control of Time-Dependent Fluid Flow , 2001, SIAM J. Control. Optim..

[18]  Sabine Le Borne,et al.  Numerische Mathematik Domain decomposition based H-LU preconditioning , 2009 .

[19]  Lei Du,et al.  A block IDR(s) method for nonsymmetric linear systems with multiple right-hand sides , 2011, J. Comput. Appl. Math..

[20]  Gene H. Golub,et al.  Numerical solution of saddle point problems , 2005, Acta Numerica.

[21]  Peter Benner,et al.  Riccati-based boundary feedback stabilization of incompressible Navier-Stokes flow , 2013 .

[22]  Per-Gunnar Martinsson,et al.  An O(N) algorithm for constructing the solution operator to 2D elliptic boundary value problems in the absence of body loads , 2013, Advances in Computational Mathematics.

[23]  M. Benzi,et al.  INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN FLUIDS Int. J. Numer. Meth. Fluids (2010) Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/fld.2267 Modified augmented Lagrangian preconditioners for the incompressible Navier , 2022 .

[24]  Johannes P. Schlöder,et al.  An efficient multiple shooting based reduced SQP strategy for large-scale dynamic process optimization. Part 1: theoretical aspects , 2003, Comput. Chem. Eng..

[25]  Andrew J. Wathen,et al.  Preconditioning Iterative Methods for the Optimal Control of the Stokes Equations , 2011, SIAM J. Sci. Comput..

[26]  D. Barkley,et al.  The Onset of Turbulence in Pipe Flow , 2011, Science.

[27]  Eric Darve,et al.  A fast, memory efficient and robust sparse preconditioner based on a multifrontal approach with applications to finite‐element matrices , 2016 .

[28]  M. Heinkenschloss.  Numerical Solution of Implicitly Constrained Optimization Problems * , 2008 .

[29]  T. Papanastasiou,et al.  Viscous Fluid Flow , 1999 .

[30]  Edmond Chow,et al.  Accelerating Parallel Hierarchical Matrix-Vector Products via Data-Driven Sampling , 2020, 2020 IEEE International Parallel and Distributed Processing Symposium (IPDPS).

[31]  L. Grasedyck,et al.  Domain decomposition based $${\mathcal H}$$ -LU preconditioning , 2009, Numerische Mathematik.

[32]  John W. Pearson,et al.  Preconditioned iterative methods for Navier-Stokes control problems , 2015, J. Comput. Phys..

[33]  Jacek Gondzio,et al.  Fast interior point solution of quadratic programming problems arising from PDE-constrained optimization , 2017, Numerische Mathematik.

[34]  R. Flemming,et al.  Actuator Disc Methods Applied to Wind Turbines , 2016 .

[35]  Manfred Laumen.  Newton's Method for a Class of Optimal Shape Design Problems , 1996, SIAM J. Optim..

[36]  Lexing Ying,et al.  A fast nested dissection solver for Cartesian 3D elliptic problems using hierarchical matrices , 2014, J. Comput. Phys..

[37]  John W. Pearson,et al.  Refined saddle-point preconditioners for discretized Stokes problems , 2018, Numerische Mathematik.

[38]  L. Grasedyck,et al.  Domain-decomposition Based ℌ-LU Preconditioners , 2007 .

[39]  Michael A. Saunders,et al.  Preconditioners for Indefinite Systems Arising in Optimization , 1992, SIAM J. Matrix Anal. Appl..

[40]  Wolfgang Hackbusch,et al.  A Sparse Matrix Arithmetic Based on H-Matrices. Part I: Introduction to H-Matrices , 1999, Computing.

[41]  Cornelis Vuik,et al.  Eigenvalue analysis of the SIMPLE preconditioning for incompressible flow , 2004, Numer. Linear Algebra Appl..

[42]  Jianlin Xia,et al.  A robust inner–outer hierarchically semi‐separable preconditioner , 2012, Numer. Linear Algebra Appl..

[43]  Johan Meyers,et al.  Sequential Quadratic Programming (SQP) for optimal control in direct numerical simulation of turbulent flow , 2014, J. Comput. Phys..

[44]  Barry Lee,et al.  Finite elements and fast iterative solvers: with applications in incompressible fluid dynamics , 2006, Math. Comput..

[45]  Johan Meyers,et al.  Optimal turbine spacing in fully developed wind farm boundary layers , 2012 .

[46]  Alle-Jan van der Veen,et al.  Some Fast Algorithms for Sequentially Semiseparable Representations , 2005, SIAM J. Matrix Anal. Appl..

[47]  Jianlin Xia,et al.  Superfast Multifrontal Method for Large Structured Linear Systems of Equations , 2009, SIAM J. Matrix Anal. Appl..

[48]  Magne Nordaas,et al.  Robust preconditioners for PDE-constrained optimization with limited observations , 2015 .

[49]  Kathryn E. Johnson,et al.  Evaluating wake models for wind farm control , 2014, 2014 American Control Conference.

[50]  Martin B. van Gijzen,et al.  Algorithm 913: An elegant IDR(s) variant that efficiently exploits biorthogonality properties , 2011, TOMS.

[51]  Michel Verhaegen,et al.  Efficient System Identification of Heterogeneous Distributed Systems via a Structure Exploiting Extended Kalman Filter , 2011, IEEE Transactions on Automatic Control.

[52]  Owe Axelsson,et al.  Preconditioning methods for linear systems arising in constrained optimization problems , 2003, Numer. Linear Algebra Appl..

[53]  Michel Verhaegen,et al.  Efficient Preconditioners for PDE-Constrained Optimization Problems with a Multi-level Sequentially Semi-Separable Matrix Structure , 2014 .

[54]  Julio Hernández,et al.  Survey of modelling methods for wind turbine wakes and wind farms , 1999 .

[55]  D K Smith,et al.  Numerical Optimization , 2001, J. Oper. Res. Soc..

[56]  Michel Verhaegen,et al.  Modeling of the flow in wind farms for total power optimization , 2011, 2011 9th IEEE International Conference on Control and Automation (ICCA).

[57]  Qiqi Wang,et al.  Simultaneous single-step one-shot optimization with unsteady PDEs , 2015, J. Comput. Appl. Math..

[58]  B. Koren,et al.  Review of computational fluid dynamics for wind turbine wake aerodynamics , 2011 .

[59]  L. Hou,et al.  Boundary Value Problems and Optimal Boundary Control for the Navier--Stokes System: the Two-Dimensional Case , 1998 .

[60]  Mario Bebendorf,et al.  Why Finite Element Discretizations Can Be Factored by Triangular Hierarchical Matrices , 2007, SIAM J. Numer. Anal..

[61]  N. Troldborg.  Actuator Line Modeling of Wind Turbine Wakes , 2009 .

[62]  Shivkumar Chandrasekaran,et al.  A Fast Solver for HSS Representations via Sparse Matrices , 2006, SIAM J. Matrix Anal. Appl..

[63]  Johannes P. Schlöder,et al.  Newton-Picard-Based Preconditioning for Linear-Quadratic Optimization Problems with Time-Periodic Parabolic PDE Constraints , 2012, SIAM J. Sci. Comput..

[64]  M. Gijzen,et al.  Convergence analysis of multilevel sequentially semiseparable preconditioners , 2015 .

[65]  Owe Axelsson,et al.  Comparison of preconditioned Krylov subspace iteration methods for PDE-constrained optimization problems , 2016, Numerical Algorithms.

[66]  Kuniyoshi Abe.  Hybrid Bi-CG methods with a Bi-CG formulation closer to the IDR approach , 2012, Appl. Math. Comput..

[67]  Alfio Borzì,et al.  Multigrid Methods for PDE Optimization , 2009, SIAM Rev..

[68]  Fernando Porté-Agel,et al.  The Effect of Free-Atmosphere Stratification on Boundary-Layer Flow and Power Output from Very Large Wind Farms , 2013 .

[69]  Zeng-Qi Wang,et al.  On parameterized inexact Uzawa methods for generalized saddle point problems , 2008 .

[70]  R. Vandebril,et al.  Matrix Computations and Semiseparable Matrices: Linear Systems , 2010 .

[71]  M. Heinkenschloss,et al.  Large-Scale PDE-Constrained Optimization: An Introduction , 2003 .

[72]  J. Meyers,et al.  Optimal control of energy extraction in wind-farm boundary layers , 2015, Journal of Fluid Mechanics.

[73]  S. Börm.  ℋ2-matrices – Multilevel methods for the approximation of integral operators , 2004 .

[74]  Steffen Börm,et al.  Data-sparse Approximation by Adaptive ℋ2-Matrices , 2002, Computing.

[75]  Owe Axelsson,et al.  Comparison of preconditioned Krylov subspace iteration methods for PDE-constrained optimization problems , 2016, Numerical Algorithms.

[76]  R. Plessix,et al.  An MSSS-preconditioned matrix equation approach for the time-harmonic elastic wave equation at multiple frequencies , 2018, Computational Geosciences.