Design of Deep Neural Networks as Add-on Blocks for Improving Impromptu Trajectory Tracking

TL;DR

This paper presents a method to enhance feedback control systems with deep neural networks as add-on blocks, improving trajectory tracking accuracy across different platforms by providing design guidelines and effectiveness conditions.

Contribution

It introduces platform-independent design guidelines for DNN-enhanced control, including feature selection, effectiveness conditions, and training efficiency improvements.

Findings

Achieved 43% average reduction in tracking error on unseen trajectories.

Provided guidelines for DNN feature selection and effectiveness conditions.

Identified cases where training efficiency can be increased.

Abstract

This paper introduces deep neural networks (DNNs) as add-on blocks to baseline feedback control systems to enhance tracking performance of arbitrary desired trajectories. The DNNs are trained to adapt the reference signals to the feedback control loop. The goal is to achieve a unity map between the desired and the actual outputs. In previous work, the efficacy of this approach was demonstrated on quadrotors; on 30 unseen test trajectories, the proposed DNN approach achieved an average impromptu tracking error reduction of 43% as compared to the baseline feedback controller. Motivated by these results, this work aims to provide platform-independent design guidelines for the proposed DNN-enhanced control architecture. In particular, we provide specific guidelines for the DNN feature selection, derive conditions for when the proposed approach is effective, and show in which cases the training efficiency can be further increased.