Path differentiability of ODE flows

DOI10.1016/J.JDE.2022.07.038zbMATH Open1497.49023arXiv2201.03819OpenAlexW4281756723MaRDI QIDQ2676533FDOQ2676533

Authors: Swann Marx, Edouard Pauwels 

Publication date: 27 September 2022

Published in: Journal of Differential Equations (Search for Journal in Brave)

Abstract: We consider flows of ordinary differential equations (ODEs) driven by path differentiable vector fields. Path differentiable functions constitute a proper subclass of Lipschitz functions which admit conservative gradients, a notion of generalized derivative compatible with basic calculus rules. Our main result states that such flows inherit the path differentiability property of the driving vector field. We show indeed that forward propagation of derivatives given by the sensitivity differential inclusions provide a conservative Jacobian for the flow. This allows to propose a nonsmooth version of the adjoint method, which can be applied to integral costs under an ODE constraint. This result constitutes a theoretical ground to the application of small step first order methods to solve a broad class of nonsmooth optimization problems with parametrized ODE constraints. This is illustrated with the convergence of small step first order methods based on the proposed nonsmooth adjoint.

Full work available at URL: https://arxiv.org/abs/2201.03819

Recommendations

• Conservative set valued fields, automatic differentiation, stochastic gradient methods and deep learning
• A convergence of ODE method in constrained optimization
• scientific article; zbMATH DE number 37929
• An ODE method of solving nonlinear programming
• Generalized derivatives for solutions of parametric ordinary differential equations with non-differentiable right-hand sides

zbMATH Keywords

nonsmooth optimizationpath differentiabilityparameterized ODE constraints

Mathematics Subject Classification ID

Nonsmooth analysis (49J52) Control problems involving ordinary differential equations (34H05) Existence theories for optimal control problems involving ordinary differential equations (49J15) Methods involving semicontinuity and convergence; relaxation (49J45) Derivative-free methods and methods using generalized derivatives (90C56)

Cites Work

• Measure theory and fine properties of functions
• Numerical methods for nonsmooth dynamical systems. Applications in mechanics and electronics
• Title not available (Why is that?)
• Optimization and nonsmooth analysis
• Title not available (Why is that?)
• A Chain Rule for Essentially Smooth Lipschitz Functions
• Asymptotic stability and smooth Lyapunov functions
• Applied functional analysis. Functional analysis, Sobolev spaces and elliptic differential equations
• Difference Methods for Differential Inclusions: A Survey
• Dynamic Data Assimilation
• Real analysis
• Lexicographic differentiation of nonsmooth functions
• Title not available (Why is that?)
• Solution dependence on initial conditions in differential variational inequalities
• Adjoint Sensitivity Analysis for Differential-Algebraic Equations: The Adjoint DAE System and Its Numerical Solution
• Generalized subdifferentials: a Baire categorical approach
• Title not available (Why is that?)
• Generalized derivatives for solutions of parametric ordinary differential equations with non-differentiable right-hand sides
• The Euler-Lagrange differential inclusion
• Generalized derivatives of differential-algebraic equations
• Dependence of solutions of nonsmooth differential-algebraic equations on parameters
• Differential inclusions in Wasserstein spaces: the Cauchy-Lipschitz framework
• Computationally relevant generalized derivatives: theory, evaluation and applications
• Generalisations, examples, and counter-examples in analysis and optimisation. \textit{In honour of Michel Théra at 70}
• Conservative and semismooth derivatives are equivalent for semialgebraic maps

Cited In (5)

• Differentiating Nonsmooth Solutions to Parametric Monotone Inclusion Problems
• The flow of ODEs: formalization of variational equation and Poincaré map
• Subgradient Sampling for Nonsmooth Nonconvex Minimization
• Conservative set valued fields, automatic differentiation, stochastic gradient methods and deep learning
• Flow equations on spaces of rough paths

This page was built for publication: Path differentiability of ODE flows

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2676533)