An adaptive and efficient greedy procedure for the optimal training of parametric reduced-order models. (English) Zbl 1352.65217
Summary: An adaptive and efficient approach for constructing reduced-order models (ROMs) that are robust to changes in parameters is developed. The approach is based on a greedy sampling of the underlying high-dimensional model (HDM) together with an efficient procedure for exploring the configuration space and identifying parameters for which the error is likely to be high. Because this exploration is based on a surrogate model for an error indicator, it is amenable to a fast training phase. Furthermore, a model for the exact error based on the error indicator is constructed and used to determine when the greedy procedure reaches a desired error tolerance. An efficient procedure for updating the reduced-order basis constructed by proper orthogonal decomposition is also introduced in this paper, avoiding the cost associated with computing large-scale singular value decompositions. The proposed procedure is then shown to require less evaluations of a posteriori error estimators than the classical procedure in order to identify locations of the parameter space to be sampled. It is illustrated on the training of parametric ROMs for three linear and nonlinear mechanical systems, including the realistic prediction of the response of a V-hull vehicle to underbody blasts.
MSC:
| 65L99 | Numerical methods for ordinary differential equations |
Keywords:
projection-based model reduction; error estimation; adaptive POD-greedy procedure; adaptive sampling; surrogate modelingReferences:
| [1] | LeGresleyPA, AlonsoJJ. Airfoil design optimization using reduced order models based on proper orthogonal decomposition. AIAA Paper 2000‐2545 Fluids 2000 Conference and Exhibit, Denver, CO, 2000; 1-14. |
| [2] | ZahrMJ, AmsallemD, FarhatC. Construction of parametrically‐robust CFD‐based reduced‐order models for PDE‐constrained optimization. In 21st AIAA Computational Fluid Dynamics Conference. American Institute of Aeronautics and Astronautics: Reston, Virginia, 2013; 1-11. |
| [3] | AmsallemD, ZahrMJ, ChoiY, FarhatC. Design optimization using hyper‐reduced‐order models. submitted for publication. |
| [4] | ZahrMJ, FarhatC. PDE‐constrained optimization using progressively‐constructed reduced‐order models. submitted for publication. |
| [5] | AmsallemD, CortialJ, FarhatC. Toward real‐time computational‐fluid‐dynamics‐based aeroelastic computations using a database of reduced‐order information. AIAA Journal2010; 48(9):2029-2037. |
| [6] | CuiT, MarzoukYM, WillcoxK. Data‐driven model reduction for the Bayesian solution of inverse problems., March 2014. arXiv.org. |
| [7] | FahlM, SachsEW. Reduced order modelling approaches to PDE‐constrained optimization based on proper orthogonal decomposition. In Large‐Scale PDE‐Constrained Optimization, BieglerLT (ed.), GhattasO (ed.), HeinkenschlossM (ed.), vanBloemen WaandersB (ed.) (eds). Springer: Berlin, 2003; 268-280. · Zbl 1060.49023 |
| [8] | HömbergD, VolkweinS. Control of laser surface hardening by a reduced‐order approach using proper orthogonal decomposition. Mathematical and Computer Modelling2003; 38(10):1003-1028. · Zbl 1044.49028 |
| [9] | BrülsO, DuysinxP, GolinvalJC. The global modal parameterization for non‐linear model‐order reduction in flexible multibody dynamics. International Journal for Numerical Methods in Engineering2007; 69(5):948-977. · Zbl 1194.70011 |
| [10] | HayA, BorggaardJT, PelletierD. Local improvements to reduced‐order models using sensitivity analysis of the proper orthogonal decomposition. Journal of Fluid Mechanics2009; 629:41. · Zbl 1181.76045 |
| [11] | DihlmannM, HaasdonkB. Certified PDE‐constrained parameter optimization using reduced basis surrogate models for evolution problems. Submitted to Journal of Computers Optimization and Application2013; 1-27. |
| [12] | YueY, MeerbergenK. Accelerating optimization of parametric linear systems by model order reduction. SIAM Journal on Optimization2013; 23(2):1344-1370. · Zbl 1273.35279 |
| [13] | HovlandS, GravdahlJT, WillcoxK. Explicit model predictive control for large‐scale systems via model reduction. Journal of Guidance, Control, and Dynamics2008; 31(4):1-23. |
| [14] | AmsallemD, DeolalikarS, GurrolaF, FarhatC. Model predictive control under Coupled Fluid‐Structure Constraints Using a Database of Reduced‐Order Models on a Tablet, AIAA Paper 2013-2588, AIAA Fluid Dynamics and Co‐located Conferences and Exhibit, San Diego, California, 24-27 June 2013; 1-12. |
| [15] | Bui‐ThanhT, WillcoxK, GhattasO. Parametric reduced‐order models for probabilistic analysis of unsteady aerodynamic applications. AIAA Journal2008; 46(10):2520-2529. |
| [16] | AmsallemD, FarhatC. On‐line framework for real‐time linearized parametric CFD computations based on model reduction. manuscript in preparation. |
| [17] | ItoK, RavindranSS. A reduced‐order method for simulation and control of fluid flows. Journal of Computational Physics June 1998; 143(2):403-425. · Zbl 0936.76031 |
| [18] | TaylorJA, GlauserMN. Towards practical flow sensing and control via POD and LSE based low‐dimensional tools. Journal of Fluids Engineering2004; 126(3):337-345. |
| [19] | SirovichL. Turbulence and the dynamics of coherent structures. Part I: coherent structures. Quarterly of Applied Mathematics1987; 45(3):561-571. · Zbl 0676.76047 |
| [20] | BaurU, BeattieC, BennerP, GugercinS. Interpolatory projection methods for parameterized model reduction. SIAM Journal on Scientific Computing2011; 33(5):2489-2518. · Zbl 1254.93032 |
| [21] | AmsallemD, FarhatC. Interpolation method for adapting reduced‐order models and application to aeroelasticity. AIAA Journal2008; 46(7):1803-1813. |
| [22] | AmsallemD, FarhatC. An online method for interpolating linear parametric reduced‐order models. SIAM Journal on Scientific Computing2011; 33(5):2169-2198. · Zbl 1269.65059 |
| [23] | AmsallemD, ZahrMJ, FarhatC. Nonlinear model order reduction based on local reduced‐order bases. International Journal for Numerical Methods in Engineering2012; 92(10):891-916. · Zbl 1352.65212 |
| [24] | VeroyK, PateraAT. Certified real‐time solution of the parametrized steady incompressible Navier‐Stokes equations: rigorous reduced‐basis a posteriori error bounds. International Journal for Numerical Methods in Fluids2005; 47(8‐9):773-788. · Zbl 1134.76326 |
| [25] | GreplMA, PateraAT. A posteriori error bounds for reduced‐basis approximations of parametrized parabolic partial differential equations. ESAIM: Mathematical Modelling and Numerical Analysis March 2005; 39(1):157-181. · Zbl 1079.65096 |
| [26] | PateraAT, RozzaG. Reduced basis approximation and a posteriori error estimation for parametrized partial differential equations. Version 1.0, Copyright MIT 2006, to appear in (tentative rubric) MIT Pappalardo Graduate Monographs in Mechanical Engineering. |
| [27] | Bui‐ThanhT, WillcoxK, GhattasO. Model reduction for large‐scale systems with high‐dimensional parametric input space. Technical Report ACDL Technical Report TR‐07‐2: MIT, Massachusetts, 2007; 1-19. |
| [28] | HaasdonkB, DihlmannMarkus, OhlbergerMario. A training set and multiple bases generation approach for parameterized model reduction based on adaptive grids in parameter space. Mathematical and Computer Modelling of Dynamical Systems2011; 17(4):423-442. · Zbl 1302.65221 |
| [29] | EftangJL, KnezevicDJ, PateraAT. An hp certified reduced basis method for parametrized parabolic partial differential equations. Mathematical and Computer Modelling of Dynamical Systems2011; 17(4):395-422. · Zbl 1302.65223 |
| [30] | EftangJL, StammB, ZhangS. Efficient greedy algorithms for high‐dimensional parameter spaces with applications to empirical interpolation and reduced basis methods. ESAIM: Mathematical Modelling and Numerical Analysis2014; 48(1):259-283. · Zbl 1292.41001 |
| [31] | Prud’hommeC, RovasDV, VeroyK, MadayY, PateraAT, TuriniciG. ASME DC | Journal of Fluids Engineering – reliable real‐time solution of parametrized partial differential equations: reduced‐basis output bound methods. Journal of Fluids Engineering2002; 124:70-80. |
| [32] | VeroyK, Prud’hommeC, RovasDV, PateraAT. A posteriori error bounds for reduced‐basis approximation of parametrized noncoercive and nonlinear elliptic partial differential equations, AIAA Paper 2003-3847. 16th Computational Fluid Dynamics Conference, Orlando Florida, 24-25 June 2003; 1-18. |
| [33] | HaasdonkB, OhlbergerM. Reduced basis method for finite volume approximations of parametrized linear evolution equations. ESAIM: Mathematical Modelling and Numerical Analysis2008; 42(2):277-302. · Zbl 1388.76177 |
| [34] | HaasdonkB, OhlbergerM. Efficient reduced models and a posteriori error estimation for parametrized dynamical systems by offline/online decomposition. Mathematical and Computer Modelling of Dynamical Systems2011; 17(2):145-161. · Zbl 1230.37110 |
| [35] | AmsallemD, HetmaniukU. A posteriori error estimators for linear reduced order models using Krylov‐based integrators. accepted for publication, International Journal for Numerical Methods in Engineering2014; 1-23. |
| [36] | CarlbergK, FarhatC, CortialJ, AmsallemD. The GNAT method for nonlinear model reduction: effective implementation and application to computational fluid dynamics and turbulent flows. Journal of Computational Physics2013; 242(C):623-647. · Zbl 1299.76180 |
| [37] | HinzeM, KunkelM. Residual based sampling in POD model order reduction of drift‐diffusion equations in parametrized electrical networks, March 2010. arXiv.org. |
| [38] | WuY, HetmaniukU. An adaptive training for POD‐based reduced order modeling of incompressible Navier‐Stokes flows. submitted for publication. |
| [39] | BarraultM, MadayY, NguyenNC, PateraAT. An empirical interpolation method: application to efficient reduced‐basis discretization of partial differential equations. Comptes Rendus de l’Academie des Sciences Paris2004; 339:667-672. · Zbl 1061.65118 |
| [40] | DrohmannM, CarlbergK. The ROMES method for statistical modeling of reduced‐order‐model error, May 2014. arXiv.org. |
| [41] | WirtzD, SorensenDC, HaasdonkB. A‐posteriori error estimation for DEIM reduced nonlinear dynamical systems. SIAM Journal of Scientific Computing2014; 36(2):A311-A338. · Zbl 1312.65127 |
| [42] | KrigeDG. A statistical approach to some basic mine valuation problems on the Witwatersrand. Journal of the Chemical, Metallurgical and Mining Society of South Africa1951; 52(6):201-213. |
| [43] | SacksJ, WelchWJ, MitchellTJ, WynnHP. Design and analysis of computer experiments. Statistical Science1989; 4(4):409-423. |
| [44] | JonesDR. A taxonomy of global optimization methods based on response surfaces. Journal of Global Optimization2001; 21:345-383. · Zbl 1172.90492 |
| [45] | KushnerHJ. A new method of locating the maximum point of an arbitrary multipeak curve in the presence of noise. Journal of Basic Engineering1964; 86(1):97-106. |
| [46] | MacQueenJ. Some methods for classification and analysis of multivariate observations. In Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability. University of California Press: Berkeley and Los Angeles, California, 1967; 281-293. · Zbl 0214.46201 |
| [47] | HastieT, TibshiraniR, FriedmanJH. The Elements of Statistical Learning (2nd edn), Data Mining, Inference, and Prediction. Springer: Berlin, 2009. · Zbl 1273.62005 |
| [48] | SmithRC. Uncertainty Quantification, Theory, Implementation, and Applications. SIAM: Philadelphia, 2013. |
| [49] | FarhatC, AveryP, ChapmanT, CortialJ. Dimensional reduction of nonlinear finite element dynamic models with finite rotations and energy‐conserving mesh sampling and weighting for computational efficiency, pages 1-46, December 2013. submitted for publication. |
| [50] | AnSS, KimT, JamesDL. Optimizing cubature for efficient integration of subspace deformations. ACM Transactions on Graphics2008; 27(5):1. |
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.
