Stochastic optimization methods for the simultaneous control of parameter-dependent systems

We address the application of stochastic optimization methods for the simultaneous control of parameter-dependent systems. In particular, we focus on the classical Stochastic Gradient Descent (SGD) approach of Robbins and Monro, and on the recently developed Continuous Stochastic Gradient (CSG) algorithm. We consider the problem of computing simultaneous controls through the minimization of a cost functional defined as the superposition of individual costs for each realization of the system. We compare the performances of these stochastic approaches, in terms of their computational complexity, with those of the more classical Gradient Descent (GD) and Conjugate Gradient (CG) algorithms, and we discuss the advantages and disadvantages of each methodology. In agreement with well-established results in the machine learning context, we show how the SGD and CSG algorithms can significantly reduce the computational burden when treating control problems depending on a large amount of parameters. This is corroborated by numerical experiments.

[1]  Y. Nesterov.  A method for unconstrained convex minimization problem with the rate of convergence o(1/k^2) , 1983 .

[2]  Lei Li,et al.  Random Batch Methods (RBM) for interacting particle systems , 2018, J. Comput. Phys..

[3]  Stephen P. Boyd,et al.  Convex Optimization , 2004, Algorithms and Theory of Computation Handbook.

[4]  T. N. Stevenson,et al.  Fluid Mechanics , 2021, Nature.

[5]  Warren B. Powell,et al.  Adaptive Stochastic Control for the Smart Grid , 2011, Proceedings of the IEEE.

[6]  R. Durrett.  Probability: Theory and Examples , 1993 .

[7]  Anthony Widjaja,et al.  Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond , 2003, IEEE Transactions on Neural Networks.

[8]  H. Robbins.  A Stochastic Approximation Method , 1951 .

[9]  Emmanuel Trélat,et al.  Contrôle optimal : théorie & applications , 2005 .

[10]  A. Patera,et al.  A PRIORI CONVERGENCE OF THE GREEDY ALGORITHM FOR THE PARAMETRIZED REDUCED BASIS METHOD , 2012 .

[11]  Nathan Srebro,et al.  SVM optimization: inverse dependence on training set size , 2008, ICML '08.

[12]  L. Eon Bottou.  Online Learning and Stochastic Approximations , 1998 .

[13]  Karen Veroy,et al.  Certified Reduced Basis Methods for Parametrized Distributed Elliptic Optimal Control Problems with Control Constraints , 2016, SIAM J. Sci. Comput..

[14]  Roland Glowinski,et al.  Variational methods for the numerical solution of nonlinear elliptic problems , 2015, CBMS-NSF regional conference series in applied mathematics.

[15]  K. ITO,et al.  Reduced Basis Method for Optimal Control of Unsteady Viscous Flows , 2001 .

[16]  Petros Koumoutsakos,et al.  Machine Learning for Fluid Mechanics , 2019, Annual Review of Fluid Mechanics.

[17]  Michael Stingl,et al.  CSG: A new stochastic gradient method for the efficient solution of structural optimization problems with infinitely many states , 2020 .

[18]  Antoine Guitton,et al.  Fast Full Waveform Inversion With Random Shot Decimation , 2011 .

[19]  Enrique Zuazua Iriondo,et al.  Controllability of star-shaped networks of strings , 2001 .

[20]  Mark Kärcher,et al.  A POSTERIORI ERROR ESTIMATION FOR REDUCED ORDER SOLUTIONS OF PARAMETRIZED PARABOLIC OPTIMAL CONTROL PROBLEMS , 2014 .

[21]  Enrique Zuazua,et al.  A Stochastic Approach to the Synchronization of Coupled Oscillators , 2020, Frontiers in Energy Research.

[22]  Tong Zhang,et al.  Accelerated proximal stochastic dual coordinate ascent for regularized loss minimization , 2013, Mathematical Programming.

[23]  Enrique Zuazua,et al.  Greedy optimal control for elliptic problems and its application to turnpike problems , 2018, Numerische Mathematik.

[24]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[25]  Franck Boyer,et al.  ON THE PENALISED HUM APPROACH AND ITS APPLICATIONS TO THE NUMERICAL APPROXIMATION OF NULL-CONTROLS FOR PARABOLIC PROBLEMS , 2013 .

[26]  K. Kunisch,et al.  Control of the Burgers Equation by a Reduced-Order Approach Using Proper Orthogonal Decomposition , 1999 .

[27]  Yurii Nesterov,et al.  Introductory Lectures on Convex Optimization - A Basic Course , 2014, Applied Optimization.

[28]  Enrique Zuazua,et al.  Model predictive control with random batch methods for a guiding problem , 2020, Mathematical Models and Methods in Applied Sciences.

[29]  Michael Jahrer,et al.  Collaborative Filtering Applied to Educational Data Mining , 2010 .

[30]  Albert Cohen,et al.  Kolmogorov widths under holomorphic mappings , 2015, ArXiv.

[31]  Léon Bottou,et al.  Large-Scale Machine Learning with Stochastic Gradient Descent , 2010, COMPSTAT.

[32]  Luca Dedè,et al.  Reduced Basis Method and A Posteriori Error Estimation for Parametrized Linear-Quadratic Optimal Control Problems , 2010, SIAM J. Sci. Comput..

[33]  Yoram Singer,et al.  Adaptive Subgradient Methods for Online Learning and Stochastic Optimization , 2011, J. Mach. Learn. Res..

[34]  Jiming Chen,et al.  Load scheduling with price uncertainty and temporally-coupled constraints in smart grids , 2015, 2015 IEEE Power & Energy Society General Meeting.

[35]  P. Ciarlet,et al.  Mathematical elasticity, volume I: Three-dimensional elasticity , 1989 .

[36]  Enrique Zuazua,et al.  Greedy controllability of finite dimensional linear systems , 2016, Autom..

[37]  Vladimir Vapnik,et al.  Principles of Risk Minimization for Learning Theory , 1991, NIPS.

[38]  Jorge Nocedal,et al.  Optimization Methods for Large-Scale Machine Learning , 2016, SIAM Rev..

[39]  Yousef Saad,et al.  Iterative methods for sparse linear systems , 2003 .

[40]  J. C. Jaeger,et al.  Conduction of Heat in Solids , 1952 .

[41]  S. Ravindran.  A reduced-order approach for optimal control of fluids using proper orthogonal decomposition , 2000 .

[42]  J. Coron.  Control and Nonlinearity , 2007 .

[43]  Enrique Zuazua,et al.  From averaged to simultaneous controllability , 2016 .

[44]  L. Rosenhead.  Conduction of Heat in Solids , 1947, Nature.

[45]  Belinda B. King,et al.  Proper orthogonal decomposition for reduced basis feedback controllers for parabolic equations , 2001 .

[46]  Eric Moulines,et al.  Non-Asymptotic Analysis of Stochastic Approximation Algorithms for Machine Learning , 2011, NIPS.

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

[48]  Léon Bottou,et al.  The Tradeoffs of Large Scale Learning , 2007, NIPS.

[49]  G. Weiss,et al.  Observation and Control for Operator Semigroups , 2009 .

[50]  Werner Römisch,et al.  Optimal Power Generation under Uncertainty via Stochastic Programming , 1998 .