×

Mean-variance risk-averse optimal control of systems governed by PDEs with random parameter fields using quadratic approximations. (English) Zbl 1391.93289

Summary: We present a method for optimal control of systems governed by partial differential equations (PDEs) with uncertain parameter fields. We consider an objective function that involves the mean and variance of the control objective, leading to a risk-averse optimal control problem. Conventional numerical methods for optimization under uncertainty are prohibitive when applied to this problem. To make the optimal control problem tractable, we invoke a quadratic Taylor series approximation of the control objective with respect to the uncertain parameter field. This enables deriving explicit expressions for the mean and variance of the control objective in terms of its gradients and Hessians with respect to the uncertain parameter. The risk-averse optimal control problem is then formulated as a PDE-constrained optimization problem with constraints given by the forward and adjoint PDEs defining these gradients and Hessians. The expressions for the mean and variance of the control objective under the quadratic approximation involve the trace of the (preconditioned) Hessian and are thus prohibitive to evaluate. To overcome this difficulty, we employ trace estimators, which only require a modest number of Hessian-vector products. We illustrate our approach with two specific problems: the control of a semilinear elliptic PDE with an uncertain boundary source term, and the control of a linear elliptic PDE with an uncertain coefficient field. For the latter problem, we derive adjoint-based expressions for efficient computation of the gradient of the risk-averse objective with respect to the controls. Along with the quadratic approximation and trace estimation, this ensures that the cost of computing the risk-averse objective and its gradient with respect to the control – measured in the number of PDE solves – is independent of the (discretized) parameter and control dimensions, and depends only on the number of random vectors employed in the trace estimation, leading to an efficient quasi-Newton method for solving the optimal control problem. Finally, we present a comprehensive numerical study of an optimal control problem for fluid flow in a porous medium with an uncertain permeability field.

MSC:

93E20 Optimal stochastic control
93C20 Control/observation systems governed by partial differential equations
60H15 Stochastic partial differential equations (aspects of stochastic analysis)
60H35 Computational methods for stochastic equations (aspects of stochastic analysis)
35Q93 PDEs in connection with control and optimization
35R60 PDEs with randomness, stochastic partial differential equations
65K10 Numerical optimization and variational techniques
PDFBibTeX XMLCite
Full Text: DOI arXiv

References:

[1] A. Alexanderian, P. J. Gloor, and O. Ghattas, {\it On Bayesian A-and D-optimal experimental designs in infinite dimensions}, Bayesian Anal., 11 (2016), pp. 671-695. · Zbl 1359.62315
[2] A. Alexanderian, N. Petra, G. Stadler, and O. Ghattas, {\it A fast and scalable method for A-optimal design of experiments for infinite-dimensional Bayesian nonlinear inverse problems}, SIAM J. Sci. Comput., 38 (2016), pp. A243-A272, . · Zbl 06536072
[3] H. Avron and S. Toledo, {\it Randomized algorithms for estimating the trace of an implicit symmetric positive semi-definite matrix}, J. ACM (JACM), 58 (2011), 8. · Zbl 1327.68331
[4] H.-G. Beyer and B. Sendhoff, {\it Robust optimization–a comprehensive survey}, Comput. Methods Appl. Mech. Engrg., 196 (2007), pp. 3190-3218. · Zbl 1173.74376
[5] L. T. Biegler, O. Ghattas, M. Heinkenschloss, D. Keyes, and B. van Bloemen Waanders, eds., {\it Real-Time PDE-Constrained Optimization}, Comput. Sci. Engrg. 3, SIAM, 2007, . · Zbl 1117.49004
[6] J. R. Birge and F. Louveaux, {\it Introduction to Stochastic Programming}, Springer-Verlag, 1997. · Zbl 0892.90142
[7] F. Bonizzoni and F. Nobile, {\it Perturbation analysis for the Darcy problem with log-normal permeability}, SIAM/ASA J. Uncertainty Quant., 2 (2014), pp. 223-244, . · Zbl 1311.76130
[8] A. Borz\`\i and V. Schulz, {\it Computational Optimization of Systems Governed by Partial Differential Equations}, Comput. Sci. Engrg. 8, SIAM, 2012, . · Zbl 1240.90001
[9] A. Borz\`\i, V. Schulz, C. Schillings, and G. Von Winckel, {\it On the treatment of distributed uncertainties in PDE-constrained optimization}, GAMM-Mitt., 33 (2010), pp. 230-246. · Zbl 1207.49033
[10] A. Borz\`\i and G. von Winckel, {\it Multigrid methods and sparse-grid collocation techniques for parabolic optimal control problems with random coefficients}, SIAM J. Sci. Comput., 31 (2009), pp. 2172-2192, . · Zbl 1196.35029
[11] A. Borz\`\i and G. von Winckel, {\it A POD framework to determine robust controls in PDE optimization}, Comput. Vis. Sci., 14 (2011), pp. 91-103. · Zbl 1242.93025
[12] T. Bui-Thanh, O. Ghattas, J. Martin, and G. Stadler, {\it A computational framework for infinite-dimensional Bayesian inverse problems Part I: The linearized case, with application to global seismic inversion}, SIAM J. Sci. Comput., 35 (2013), pp. A2494-A2523, . · Zbl 1287.35087
[13] P. Chen and A. Quarteroni, {\it Weighted reduced basis method for stochastic optimal control problems with elliptic PDE constraint}, SIAM/ASA J. Uncertainty Quant., 2 (2014), pp. 364-396, . · Zbl 1309.35182
[14] G. Da Prato, {\it An Introduction to Infinite-Dimensional Analysis}, Universitext, Springer, 2006. · Zbl 1109.46001
[15] G. Da Prato and J. Zabczyk, {\it Second-Order Partial Differential Equations in Hilbert Spaces}, Cambridge University Press, 2002. · Zbl 1012.35001
[16] M. Dambrine, C. Dapogny, and H. Harbrecht, {\it Shape optimization for quadratic functionals and states with random right-hand sides}, SIAM J. Control Optim., 53 (2015), pp. 3081-3103, . · Zbl 1326.49070
[17] J. Darlington, C. Pantelides, B. Rustem, and B. Tanyi, {\it Decreasing the sensitivity of open-loop optimal solutions in decision making under uncertainty}, European J. Oper. Res., 121 (2000), pp. 343-362. · Zbl 0961.90109
[18] J. Darlington, C. C. Pantelides, B. Rustem, and B. A. Tanyi, {\it An algorithm for constrained nonlinear optimization under uncertainty}, Automatica, 35 (1999), pp. 217-228. · Zbl 0966.90074
[19] M. Dashti and A. M. Stuart, {\it The Bayesian approach to inverse problems}, in Handbook of Uncertainty Quantification, R. Ghanem, D. Higdon, and H. Owhadi, eds., Springer, 2017, pp. 311-428.
[20] J. C. De Los Reyes, {\it Numerical PDE-constrained optimization}, Springer, 2015. · Zbl 1312.65100
[21] M. Diehl, H. G. Bock, and E. Kostina, {\it An approximation technique for robust nonlinear optimization}, Mathematical Programming, 107 (2006), pp. 213-230. · Zbl 1134.90039
[22] Y. Ermoliev, {\it Stochastic quasigradient methods and their application to system optimization}, Stochastics, 9 (1983), pp. 1-36. · Zbl 0512.90079
[23] X. Fernique, {\it Intégrabilité des vecteurs gaussiens}, C. R. Acad. Sci. Paris Sér. A-B, 270 (1970), pp. A1698-A1699. · Zbl 0206.19002
[24] A. Gaivoronskii, {\it Nonstationary stochastic programming problems}, Cybernet. Systems Anal., 14 (1978), pp. 575-579.
[25] M. D. Gunzburger, {\it Perspectives in Flow Control and Optimization}, Adv. Des. Control 5, SIAM, 2003, . · Zbl 1088.93001
[26] M. D. Gunzburger and J. Ming, {\it Optimal control of stochastic flow over a backward-facing step using reduced-order modeling}, SIAM J. Sci. Comput., 33 (2011), pp. 2641-2663, . · Zbl 1232.93097
[27] M. Hinze, R. Pinnau, M. Ulbrich, and S. Ulbrich, {\it Optimization with PDE Constraints}, Springer, 2009. · Zbl 1167.49001
[28] L. S. Hou, J. Lee, and H. Manouzi, {\it Finite element approximations of stochastic optimal control problems constrained by stochastic elliptic PDEs}, J. Math. Anal. Appl., 384 (2011), pp. 87-103. · Zbl 1227.65011
[29] D. Kouri, {\it An Approach for the Adaptive Solution of Optimization Problems Governed by Partial Differential Equations with Uncertain Coefficients}, Ph.D. thesis, Rice University, 2012.
[30] D. P. Kouri, {\it A multilevel stochastic collocation algorithm for optimization of PDEs with uncertain coefficients}, SIAM/ASA J. Uncertainty Quant., 2 (2014), pp. 55-81, . · Zbl 1307.49026
[31] D. P. Kouri, M. Heinkenschloss, D. Ridzal, and B. van Bloemen Waanders, {\it Inexact objective function evaluations in a trust-region algorithm for PDE-constrained optimization under uncertainty}, SIAM J. Sci. Comput., 36 (2014), pp. A3011-A3029, . · Zbl 1312.49033
[32] D. P. Kouri, M. Heinkenschloss, D. Ridzal, and B. G. van Bloemen Waanders, {\it A trust-region algorithm with adaptive stochastic collocation for PDE optimization under uncertainty}, SIAM J. Sci. Comput., 35 (2013), pp. A1847-A1879, . · Zbl 1275.49047
[33] D. P. Kouri and T. M. Surowiec, {\it Risk-averse PDE-constrained optimization using the conditional value-at-risk}, SIAM J. Optim., 26 (2016), pp. 365-396, . · Zbl 1337.49049
[34] A. Kunoth and C. Schwab, {\it Analytic regularity and GPC approximation for control problems constrained by linear parametric elliptic and parabolic PDEs}, SIAM J. Control Optim., 51 (2013), pp. 2442-2471, . · Zbl 1278.49043
[35] R. Rackwitz, {\it Reliability analysis–a review and some perspectives}, Structural Safety, 23 (2001), pp. 365-395.
[36] H. Robbins and S. Monro, {\it A stochastic approximation method}, Ann. Math. Statist., 22 (1951), pp. 400-407. · Zbl 0054.05901
[37] R. T. Rockafellar and S. Uryasev, {\it Optimization of conditional value-at-risk}, J. Risk, 2 (2000), pp. 21-42.
[38] F. Roosta-Khorasani and U. Ascher, {\it Improved bounds on sample size for implicit matrix trace estimators}, Found. Comput. Math., (2014), pp. 1-26. · Zbl 1323.65043
[39] A. Ruszczyński and W. Syski, {\it A method of aggregate stochastic subgradients with on-line stepsize rules for convex stochastic programming problems}, in Stochastic Programming 84 Part II, Springer, 1986, pp. 113-131. · Zbl 0597.90064
[40] N. V. Sahinidis, {\it Optimization under uncertainty: State-of-the-art and opportunities}, Comput. Chem. Eng., 28 (2004), pp. 971-983.
[41] A. Shapiro, D. Dentcheva, and A. Ruszczyński, {\it Lectures on Stochastic Programming: Modeling and Theory}, MOS-SIAM Ser. Optim. 9, SIAM, 2009, . · Zbl 1183.90005
[42] S. Smolyak, {\it Quadrature and interpolation formulas for tensor products of certain classes of functions}, Soviet Math. Dokl., 4 (1963), pp. 240-243. · Zbl 0202.39901
[43] A. M. Stuart, {\it Inverse problems: A Bayesian perspective}, Acta Numer., 19 (2010), pp. 451-559. · Zbl 1242.65142
[44] H. Tiesler, R. M. Kirby, D. Xiu, and T. Preusser, {\it Stochastic collocation for optimal control problems with stochastic PDE constraints}, SIAM J. Control Optim., 50 (2012), pp. 2659-2682, . · Zbl 1260.60125
[45] F. Tröltzsch, {\it Optimal Control of Partial Differential Equations: Theory, Methods and Applications}, Grad. Stud. Math. 112, American Mathematical Society, 2010. · Zbl 1195.49001
[46] C. Withers, {\it The moments of the multivariate normal}, Bull. Australian Math. Soc., 32 (1985), pp. 103-107. · Zbl 0573.62047
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.