English

Unifying Model-Based and Neural Network Feedforward: Physics-Guided Neural Networks with Linear Autoregressive Dynamics

Systems and Control 2023-03-31 v1 Machine Learning Systems and Control

Abstract

Unknown nonlinear dynamics often limit the tracking performance of feedforward control. The aim of this paper is to develop a feedforward control framework that can compensate these unknown nonlinear dynamics using universal function approximators. The feedforward controller is parametrized as a parallel combination of a physics-based model and a neural network, where both share the same linear autoregressive (AR) dynamics. This parametrization allows for efficient output-error optimization through Sanathanan-Koerner (SK) iterations. Within each SK-iteration, the output of the neural network is penalized in the subspace of the physics-based model through orthogonal projection-based regularization, such that the neural network captures only the unmodelled dynamics, resulting in interpretable models.

Keywords

Cite

@article{arxiv.2209.12489,
  title  = {Unifying Model-Based and Neural Network Feedforward: Physics-Guided Neural Networks with Linear Autoregressive Dynamics},
  author = {Johan Kon and Dennis Bruijnen and Jeroen van de Wijdeven and Marcel Heertjes and Tom Oomen},
  journal= {arXiv preprint arXiv:2209.12489},
  year   = {2023}
}

Comments

Accepted for presentation at the 2022 Conference on Decision and Control (CDC)

R2 v1 2026-06-28T02:04:55.653Z