Learning self-triggered controllers with Gaussian processes

TL;DR

This paper presents a method for designing self-triggered controllers for unknown plant dynamics in networked control systems using Gaussian process regression and reinforcement learning, enabling joint learning and control policy optimization.

Contribution

It introduces a novel approach combining Gaussian processes and reinforcement learning to design self-triggered controllers for systems with unknown dynamics.

Findings

Effective learning of plant dynamics using Gaussian processes.

Joint optimization of control and communication policies.

Numerical simulation demonstrates approach effectiveness.

Abstract

This paper investigates the design of self-triggered controllers for networked control systems (NCSs), where the dynamics of the plant is \textit{unknown} apriori. To deal with the unknown transition dynamics, we employ the Gaussian process (GP) regression in order to learn the dynamics of the plant. To design the self-triggered controller, we formulate an optimal control problem, such that the optimal control and communication policies can be jointly designed based on the GP model of the plant. Moreover, we provide an overall implementation algorithm that jointly learns the dynamics of the plant and the self-triggered controller based on a reinforcement learning framework. Finally, a numerical simulation illustrates the effectiveness of the proposed approach.