Data-Driven Discrete-time Control with H\"{o}lder-Continuous Real-time Learning

TL;DR

This paper introduces a data-driven control framework for discrete-time systems using Holder-continuous learning schemes, enabling stable, robust, and real-time tracking of output trajectories despite uncertainties and disturbances.

Contribution

It proposes a novel Holder-continuous learning scheme as a discrete-time disturbance observer, ensuring finite-time stable convergence and robustness in nonlinear model-free control.

Findings

The framework guarantees bounded tracking error under Lipschitz conditions.

The observer and controller are proven to be finite-time stable and robust.

Numerical experiments demonstrate effective control of a nonlinear second-order system.

Abstract

This work provides a framework for data-driven control of discrete time systems with unknown input-output dynamics and outputs controllable by the inputs. This framework leads to stable and robust real-time control of the system such that a feasible output trajectory can be tracked. This is made possible by rapid real-time stable learning of the unknown dynamics using H\"{o}lder-continuous learning schemes that are designed as discrete-time stable disturbance observers. This observer learns from prior input-output history and it ensures finite-time stable convergence of model estimation errors to a bounded neighborhood of the zero vector if the system is known to be Lipschitz-continuous with respect to outputs, inputs, internal parameters and states, and time. In combination with nonlinearly stable controller designs, this makes the proposed framework nonlinearly stable and robust to disturbances, model uncertainties, and unknown measurement noise. Nonlinear stability and robustness analyses of the observer and controller designs are carried out using discrete Lyapunov analysis. H\"{o}lder-continuous Finite-time stable observer and controller designs in this framework help to prove robustness of these schemes and guaranteed convergence of outputs to a bounded neighborhood of the desired output trajectory. A numerical experiment on a nonlinear second-order system demonstrates the performance of this discrete nonlinear model-free control framework.