Symplectic Runge-Kutta schemes for adjoint equations, automatic differentiation, optimal control, and more

DOI10.1137/151002769zbMATH Open1339.65243arXiv1503.04021OpenAlexW2137254413MaRDI QIDQ2808265FDOQ2808265

Authors: Jesús María Sanz-Serna 

Publication date: 20 May 2016

Published in: SIAM Review (Search for Journal in Brave)

Abstract: It is well known that symplectic Runge-Kutta and Partitioned Runge-Kutta methods exactly preserve {em quadratic} first integrals (invariants of motion) of the system being integrated. While this property is often seen as a mere curiosity (it does not hold for arbitrary first integrals), it plays an important role in the computation of numerical sensitivities, optimal control theory and Lagrangian mechanics, as described in this paper, which, together with some new material, presents in a unified way a number of results now scattered or implicit in the literature. Some widely used procedures, such as the direct method in optimal control theory and the computation of sensitivities via reverse accumulation imply "hidden" integrations with symplectic Partitioned Runge-Kutta schemes.

Full work available at URL: https://arxiv.org/abs/1503.04021

Recommendations

• Automatic differentiation of explicit Runge-Kutta methods for optimal control
• On the Properties of Runge-Kutta Discrete Adjoints
• Runge-Kutta methods in optimal control and the transformed adjoint system
• Implicit-explicit Runge-Kutta schemes for numerical discretization of optimal control problems
• A class of symplectic partitioned Runge-Kutta methods

zbMATH Keywords

Lagrange multipliersHamiltonian systemsoptimal controlLagrangian mechanicssymplectic integrationvariational equationsadjoint equationsRunge-Kutta methodspartitioned Runge-Kutta methodsautomatic differentiationconstrained controlsdifferential-algebraic problemssymplectic Runge-Kuttacomputation of sensitivitiesreflected and transposed Runge-Kutta schemes

Mathematics Subject Classification ID

Numerical optimization and variational techniques (65K10) Numerical methods for differential-algebraic equations (65L80) Existence theories for optimal control problems involving ordinary differential equations (49J15) Multistep, Runge-Kutta and extrapolation methods for ordinary differential equations (65L06) Numerical methods for Hamiltonian systems including symplectic integrators (65P10) Discretization methods and integrators (symplectic, variational, geometric, etc.) for dynamical systems (37M15)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Runge-Kutta methods in optimal control and the transformed adjoint system
• Computation of order conditions for symplectic partitioned Runge-Kutta schemes with application to optimal control
• Hamiltonian Runge-Kutta type methods and their variational interpretation
• Discrete mechanics and optimal control: An analysis
• Solving Ordinary Differential Equations I
• Discrete mechanics and variational integrators
• Title not available (Why is that?)
• Title not available (Why is that?)
• Second-Order Runge--Kutta Approximations in Control Constrained Optimal Control
• Title not available (Why is that?)
• Geometric Numerical Integration
• Optimal control. Theory and applications
• Numerical Methods for Ordinary Differential Equations
• Runge-Kutta schemes for Hamiltonian systems
• Canonical Runge-Kutta methods
• Simulating Hamiltonian Dynamics
• On the Scope of the Method of Modified Equations
• An introduction to the adjoint approach to design
• Molecular modeling and simulation. An interdisciplinary guide.
• On Order Conditions for Partitioned Symplectic Methods
• Title not available (Why is that?)
• New families of symplectic splitting methods for numerical integration in dynamical astronomy
• A mathematical view of automatic differentiation
• Order Conditions for Canonical Runge–Kutta Schemes
• Partitioned Runge-Kutta Methods for Separable Hamiltonian Problems
• Discrete variational Hamiltonian mechanics
• Adjoint sensitivity analysis of regional air quality models
• Title not available (Why is that?)
• The non-existence of symplectic multi-derivative Runge-Kutta methods
• Stability of Runge-Kutta Methods for Trajectory Problems
• Numerical approximation of exact controls for waves
• On quadratic invariants and symplectic structure
• Reflected and transposed Runge-Kutta methods

Cited In (42)

• High order variational integrators in the optimal control of mechanical systems
• Symplecticness conditions of some low order partitioned methods for non-autonomous Hamiltonian systems
• A general framework for solving differential equations
• A Novel Class of Energy-Preserving Runge-Kutta Methods for the Korteweg-de Vries Equation
• A symplectic indirect approach for a class of nonlinear optimal control problems of differential‐algebraic systems
• Generalization of partitioned Runge-Kutta methods for adjoint systems
• Effective order of partitioned Runge-Kutta methods
• Title not available (Why is that?)
• Analysis of energy and quadratic invariant preserving (EQUIP) methods
• Learning adaptive regularization for image labeling using geometric assignment
• Assignment Flows
• Construction of higher order symplectic schemes by composition
• Deep learning as optimal control problems: models and numerical methods
• Estimation of Ordinary Differential Equation Models with Discretization Error Quantification
• Adjoint-based exact Hessian computation
• Acoustic shape optimization using energy stable curvilinear finite differences
• Efficient delay and parameter estimation for nonlinear systems based on lifted IRK integrators with inexact Jacobians
• On properties of adjoint systems for evolutionary PDEs
• On difference schemes approximating first-order differential equations and defining a projective correspondence between layers
• Geometric methods for adjoint systems
• Contractivity of Runge--Kutta Methods for Convex Gradient Systems
• Pointwise error estimates of numerical solutions to linear quadratic optimal control problems
• A discrete optimality system for an optimal harvesting problem
• Forward and adjoint sensitivity analysis with continuous explicit Runge-Kutta schemes
• Line integral solution of differential problems
• Discrete adjoint computations for relaxation Runge-Kutta methods
• Adjoint-based inversion for stress and frictional parameters in earthquake modeling
• Explicit Stabilized Integrators for Stiff Optimal Control Problems
• An efficient adjoint computational method based on lifted IRK integrator and exact penalty function for optimal control problems involving continuous inequality constraints
• Solving parameter estimation problems with discrete adjoint exponential integrators
• Implicit peer triplets in gradient-based solution algorithms for ODE constrained optimal control
• Classification with Runge-Kutta networks and feature space augmentation
• Optimal control of constrained mechanical systems in redundant coordinates: formulation and structure-preserving discretization
• On the Properties of Runge-Kutta Discrete Adjoints
• Structure-preserving algorithms with uniform error bound and long-time energy conservation for highly oscillatory Hamiltonian systems
• A new Lagrangian approach to control affine systems with a quadratic Lagrange term
• Geometric integrators and the Hamiltonian Monte Carlo method
• Palindromic 3-stage splitting integrators, a roadmap
• Symplectic Runge-Kutta discretization of a regularized forward-backward sweep iteration for optimal control problems
• Exact discrete solutions of boundary control problems for the 1D heat equation
• (Spectral) Chebyshev collocation methods for solving differential equations
• Discrete LQR and ILQR methods based on high order Runge-Kutta discretizations

This page was built for publication: Symplectic Runge-Kutta schemes for adjoint equations, automatic differentiation, optimal control, and more

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2808265)