Title data
Grüne, Lars:
Computing Lyapunov functions using deep neural networks.
In: Journal of Computational Dynamics.
Vol. 8
(April 2021)
Issue 2
.
 pp. 131152.
ISSN 21582491
DOI: https://doi.org/10.3934/jcd.2021006
This is the latest version of this item.
Related URLs
Abstract in another language
We propose a deep neural network architecture and a training algorithm for computing approximate Lyapunov functions of systems of nonlinear ordinary differential equations. Under the assumption that the system admits a compositional Lyapunov function, we prove that the number of neurons needed for an approximation of a Lyapunov function with fixed accuracy grows only polynomially in the state dimension, i.e., the proposed approach is able to overcome the curse of dimensionality. We show that nonlinear systems satisfying a smallgain condition admit compositional Lyapunov functions. Numerical examples in up to ten space dimensions illustrate the performance of the training scheme.
Further data
Available Versions of this Item

Computing Lyapunov functions using deep neural networks. (deposited 20 May 2020 08:51)
 Computing Lyapunov functions using deep neural networks. (deposited 07 Jan 2021 13:50) [Currently Displayed]