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

TL;DR

This paper explores how symplectic Runge-Kutta methods preserve quadratic invariants, impacting numerical sensitivities, optimal control, and mechanics, and unifies scattered results with new insights.

Contribution

It provides a unified presentation of symplectic Runge-Kutta schemes' properties and their applications in sensitivities and optimal control, including new theoretical insights.

Findings

Symplectic methods preserve quadratic invariants.

Hidden symplectic integrations occur in optimal control procedures.

Unified framework for existing and new results in the literature.

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.