Feedback Linearization based on Gaussian Processes with event-triggered Online Learning

TL;DR

This paper introduces an event-triggered online learning approach using Gaussian processes for feedback linearization control, improving data efficiency and real-time applicability in data-driven control systems.

Contribution

It presents a novel event-triggered online learning method with safe forgetting strategies for Gaussian process-based feedback linearization control.

Findings

Ensures high data efficiency through event-triggered updates.

Maintains asymptotic stability of tracking error.

Demonstrates effective control in simulation.

Abstract

Combining control engineering with nonparametric modeling techniques from machine learning allows to control systems without analytic description using data-driven models. Most existing approaches separate learning, i.e. the system identification based on a fixed dataset, and control, 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 the 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 constraint 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.