Adaptive Event-triggered Control For Strict-feedback Systems With Time-varying Parameters

Yan Tan,Liucang Wu,Wenqi Liu

2023-12-11

Abstract:In this article, we develop a new adaptive event-triggered asymptotic control scheme for strict-feedback systems with fast time-varying parameters. To deal with time-varying parameters with unknown variation boundaries in the feedback path and the input path, we construct three adaptive laws for parameter estimation, two for the uncertain parameters in the feedback path and one for the uncertain parameters in the input path. In particular, two sets of tuning functions are introduced to avoid over-parametrization. Additionally, an event-triggering mechanism is embedded in this adaptive control framework to reduce the data transmission from the controller to the actuator. We also introduce a soft sign function to handle the perturbations caused by sampling errors to achieve asymptotic stability and avoid the so-called parameter drift. The stability analysis shows that the closed-loop system is globally uniformly asymptotically stable and the Zeno behavior can be excluded. Simulation results verify the effectiveness and performance of the proposed adaptive scheme.

Systems and Control

What problem does this paper attempt to address?

The paper attempts to address the problem of achieving adaptive event-triggered control in strict-feedback systems with time-varying parameters. Specifically, the paper focuses on how to handle time-varying parameters with unknown changing boundaries in the feedback path and input path to achieve zero-error convergence; how to deal with disturbance terms arising from the coupling of unknown input gains with system states, which cannot be compensated by the usual design of damping terms; and how to integrate event-triggered mechanisms into the developed adaptive framework to reduce high sampling frequency and save communication resources.

By combining Adaptive Backstepping Design and Nonlinear Damping Design, the paper proposes a novel adaptive event-triggered asymptotic control scheme to handle strict-feedback systems with completely unknown time-varying parameters. This scheme introduces unknown constants and variables to avoid time derivatives and constructs three adaptive laws for parameter estimation, two of which are used to estimate the time-varying parameters in the feedback path, and one to estimate the time-varying parameters in the input path. Additionally, two sets of tuning functions are introduced to avoid over-parameterization. Furthermore, the paper utilizes soft sign functions to handle disturbances caused by sampling errors, achieving asymptotic stability and avoiding the so-called parameter drift.

The main contributions of the paper include:
1. Proposing an event-triggered adaptive control method without the Zeno phenomenon, which integrates a sampling error compensation mechanism to ensure global convergence of all closed-loop signals while avoiding parameter drift.
2. Compared to the assumptions in the literature, the assumptions in this paper are more general, removing the requirement for prior knowledge of the parameter set radius.
3. Using a two-layer estimation method to handle time-varying parameters, effectively eliminating the need for prior knowledge.

The effectiveness and performance of the proposed controller are verified through numerical simulations.