Dissipative SymODEN: Encoding Hamiltonian Dynamics with Dissipation and Control into Deep Learning

TL;DR

Dissipative SymODEN is a deep learning model that accurately infers physical system dynamics with dissipation by encoding energy and dissipation structures, enabling better predictions and energy-based control design.

Contribution

The paper introduces Dissipative SymODEN, a novel deep learning architecture that encodes port-Hamiltonian dynamics with dissipation, improving prediction accuracy and interpretability.

Findings

Accurately infers dissipative physical dynamics from data.

Reveals key physical parameters like inertia and dissipation.

Facilitates energy-based control design.

Abstract

In this work, we introduce Dissipative SymODEN, a deep learning architecture which can infer the dynamics of a physical system with dissipation from observed state trajectories. To improve prediction accuracy while reducing network size, Dissipative SymODEN encodes the port-Hamiltonian dynamics with energy dissipation and external input into the design of its computation graph and learns the dynamics in a structured way. The learned model, by revealing key aspects of the system, such as the inertia, dissipation, and potential energy, paves the way for energy-based controllers.