Robust Model Predictive Control for Asynchronously Switched Linear Systems With Intermittent Controller Failures

• 成果类型: Article
• 署名作者: Tan, Tianyu; Zhuang, Songlin; Shi, Yang
• 署名单位: University of Victoria; Yongjiang Laboratory
• 刊物名称: IEEE TRANSACTIONS ON AUTOMATIC CONTROL
• ISSN/ISSBN: 0018-9286
• DOI: 10.1109/TAC.2024.3422871
• 发表日期: 2025
• 关键词: DISTURBANCE INVARIANT-SETS; Dwell-time; nonlinear-systems; STABILITY; computation; stabilization; constraints; DESIGN

摘要

Switched systems, as an exceptional modeling tool, may face threats to unexpected variations from the environment (e.g., external disturbances) or unreliable networks (e.g., desired controllers lagged to enabled subsystems or even controller disconnection). This article aims to study the robust model predictive control problem for a class of disturbed asynchronously switched systems with occasional controller disconnection. To mitigate the adverse effect of additive disturbances, a tube-based switched model predictive control strategy is designed by properly tightening original constraints. For the feasibility concern, the minimum mode-dependent dwell time is computed offline so as to ensure that reachable sets from an initial region are included in a common feasible set. Furthermore, a nonconservative stability condition is proposed for switched systems from a set-theoretic perspective. Based on this superior result, two stability strategies with distinct converging speeds are proposed to guarantee the closed-loop system to be uniformly asymptotically stable with constructed terminal constraints. Effectiveness of the theoretical results is validated via a simulation of time-varying communication networks.