Robust Tracking Control with Neural Network Dynamic Models under Input Perturbations

TL;DR

This paper develops a robust tracking control approach for neural network dynamic models under input perturbations, introducing reachability analysis and feedback policies to ensure system safety and performance.

Contribution

It introduces a novel reachability analysis tool and a feedback policy framework specifically designed for neural network dynamic models in robust tracking tasks.

Findings

The proposed method effectively bounds reachable sets over an infinite horizon.

Numerical simulations show improved robustness compared to standard tube MPC.

The approach successfully manages input perturbations in neural network models.

Abstract

Robust control problems have significant practical implications since external disturbances can significantly impact the performance of control methods. Existing robust control methods excel at control-affine systems but fail at neural network dynamic models. Developing robust control methods for such systems remains a complex challenge. In this paper, we focus on robust tracking methods for neural network dynamic models. We first propose a reachability analysis tool designed for this system and then introduce how to reformulate a robust tracking problem with reachable sets. In addition, we prove the existence of a feedback policy that bounds the growth of reachable sets over an infinite horizon. The effectiveness of the proposed approach is validated through numerical simulations of the tracking task, where we compare it with a standard tube MPC method.