Anti-asynchrony output regulation for switched systems under switching-Q-learning event-triggering against DoS attacks  

作　　者：Lili LI Yalin CHEN Dan MA Qingjun GUO 

机构地区：[1]College of Marine Electrical Engineering,Dalian Maritime University,Dalian 116026,China [2]State Key Laboratory of Synthetical Automation for Process Industries,Northeastern University,Shenyang 110819,China [3]College of Information Science and Engineering,Northeastern University,Shenyang 110819,China

出　　处：《Science China(Information Sciences)》2024年第11期190-206,共17页中国科学（信息科学）（英文版）

基　　金：supported by National Natural Science Foundation of China(Grant Nos.62273068,U23A203-24);Natural Science Foundation of Liaoning Province(Grant No.2023-MS-120)。

摘　　要：This paper proposes a secure output regulation(SOR)scheme for network switched systems(NSSs)experiencing severely unstable dynamics(SUDs).This scheme aims to address the complex consecutive asynchrony caused by long-duration denial-of-service(LDDoS)attacks.When the denial-of-service(DoS)attack lasts longer than the dwell time of each subsystem,switching to a new subsystem is necessary to manage the unstable dynamic behavior caused by SUDs.The controller cannot immediately obtain information about the new modality.However,owing to the delayed access to new modality information by the controller,consecutive asynchrony occurs.To tackle this,we propose a resilient switching rule that mitigates the impact of consecutive asynchrony by choosing the optimal subsystem once the minimum dwell time is met.Multi-stage event-triggering mechanisms(METMs)are proposed to coordinate with switching signals by incorporating modality matching conditions and a switching-Q-learning algorithm.These mechanisms reduce asynchronous durations by aligning modality matching conditions and terminating attacks using attack parameters.This control scheme for SOR of NSSs with SUDs disturbed by LDDoS attacks provides a detailed analysis of dynamic switching behavior,leveraging the resilient switching rule and METMs.Finally,the feasibility of the proposed methodology is substantiated through simulations with an F-18 aircraft model.

关 键 词：switched systems long-duration DoS attacks consecutive asynchrony switching-Q-learning algorithm multi-stage event-triggering resilient switching rule 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程] TP393.08[自动化与计算机技术—控制科学与工程]