Paper Info
Title
Feedback Linearization Based on Gaussian Processes With Event-Triggered Online Learning
Abstract
Combining control engineering with nonparametric modeling techniques from machine learning allows for the control of systems without analytic description using data-driven models. Most of the existing approaches separate <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">learning</italic> , i.e., the system identification based on a fixed dataset, and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">control</italic> , i.e., the execution of the model-based control law. This separation makes the performance highly sensitive to the initial selection of training data and possibly requires very large datasets. This article proposes a learning feedback linearizing control law using online closed-loop identification. The employed Gaussian process model updates its training data only if the model uncertainty becomes too large. This event-triggered online learning ensures high data efficiency and thereby reduces computational complexity, which is a major barrier for using Gaussian processes under real-time constraints. We propose safe forgetting strategies of data points to adhere to budget constraints and to further increase data efficiency. We show asymptotic stability for the tracking error under the proposed event-triggering law and illustrate the effective identification and control in simulation.
Year
DOI
Venue
2020
10.1109/TAC.2019.2958840
IEEE Transactions on Automatic Control
Keywords
DocType
Volume
Autoregressive processes,Adaptation models,Control systems,Computational modeling,Noise measurement,Gaussian processes,Training data
Journal
65
Issue
ISSN
Citations 
10
0018-9286
1
PageRank 
References 
Authors
0.35
13
2
Name
Order
Citations
PageRank
Jonas Umlauft145.14
Sandra Hirche220.71