Abstract:Many dynamical systems -- from robots interacting with their surroundings to large-scale multiphysics systems -- involve a number of interacting subsystems. Toward the objective of learning composite models of such systems from data, we present i) a framework for compositional neural networks, ii) algorithms to train these models, iii) a method to compose the learned models, iv) theoretical results that bound the error of the resulting composite models, and v) a method to learn the composition itself, when it is not known a priori. The end result is a modular approach to learning: neural network submodels are trained on trajectory data generated by relatively simple subsystems, and the dynamics of more complex composite systems are then predicted without requiring additional data generated by the composite systems themselves. We achieve this compositionality by representing the system of interest, as well as each of its subsystems, as a port-Hamiltonian neural network (PHNN) -- a class of neural ordinary differential equations that uses the port-Hamiltonian systems formulation as inductive bias. We compose collections of PHNNs by using the system's physics-informed interconnection structure, which may be known a priori, or may itself be learned from data. We demonstrate the novel capabilities of the proposed framework through numerical examples involving interacting spring-mass-damper systems. Models of these systems, which include nonlinear energy dissipation and control inputs, are learned independently. Accurate compositions are learned using an amount of training data that is negligible in comparison with that required to train a new model from scratch. Finally, we observe that the composite PHNNs enjoy properties of port-Hamiltonian systems, such as cyclo-passivity -- a property that is useful for control purposes.

Port-Hamiltonian Neural ODE Networks on Lie Groups For Robot Dynamics Learning and Control

Hamiltonian Dynamics Learning from Point Cloud Observations for Nonholonomic Mobile Robot Control

Safe Autonomous Navigation for Systems with Learned SE(3) Hamiltonian Dynamics

Symplectic ODE-Net: Learning Hamiltonian Dynamics with Control

Learning Complex Motion Plans using Neural ODEs with Safety and Stability Guarantees

Neural Port-Hamiltonian Models for Nonlinear Distributed Control: An Unconstrained Parametrization Approach

GEM: Group Enhanced Model for Learning Dynamical Control Systems

Model-free tracking control of complex dynamical trajectories with machine learning

Adaptive Control of SE(3) Hamiltonian Dynamics with Learned Disturbance Features

Cooperative Deterministic Learning Control for a Group of Homogeneous Nonlinear Uncertain Robot Manipulators

General Hamiltonian Neural Networks for Dynamic Modeling: Handling Sophisticated Constraints Automatically and Achieving Coordinates Free

Compositional Learning of Dynamical System Models Using Port-Hamiltonian Neural Networks

Learning on Manifolds: Universal Approximations Properties using Geometric Controllability Conditions for Neural ODEs

Dynamic Modeling of Robotic Manipulator Via an Augmented Deep Lagrangian Network

Composite Learning Enhanced Neural Control for Robot Manipulator With Output Error Constraints

Neural Distributed Controllers with Port-Hamiltonian Structures

Dynamic Analysis of Industrial Robot Using Lie Groups

Exploring Model Structures to Reduce Data Requirements for Neural ODE Learning in Control Systems

Learning Networked Dynamical System Models with Weak Form and Graph Neural Networks

Nonlinear Model Predictive Control of a 3D Hopping Robot: Leveraging Lie Group Integrators for Dynamically Stable Behaviors