Contractivity of Runge-Kutta methods for convex gradient systems
Abstract
We consider the application of Runge-Kutta (RK) methods to gradient systems , where, as in many optimization problems, is convex and (globally) Lipschitz-continuous with Lipschitz constant . Solutions of this system behave contractively, i.e. the Euclidean distance between two solutions and is a nonincreasing function of . It is then of interest to investigate whether a similar contraction takes place, at least for suitably small step sizes , for the discrete solution. Dahlquist and Jeltsch results' imply that (1) there are explicit RK schemes that behave contractively whenever is below a scheme-dependent constant and (2) Euler's rule is optimal in this regard. We prove however, by explicit construction of a convex potential using ideas from robust control theory, that there exists RK schemes that fail to behave contractively for any choice of the time-step .
Keywords
Cite
@article{arxiv.1909.09971,
title = {Contractivity of Runge-Kutta methods for convex gradient systems},
author = {J. M. Sanz-Serna and Konstantinos C. Zygalakis},
journal= {arXiv preprint arXiv:1909.09971},
year = {2021}
}
Comments
13 pages, 2 figures