A new control method for uncertain switched systems under replay attacks

For uncertain switched systems subjected to unknown but bounded disturbances and replay attacks, a robust predictive control technique based on multi-constraint event triggering is proposed. Firstly, a hybrid switching mechanism incorporating attack detection information is constructed. Dwell-time and state-dependent switching are employed to prevent frequent switching, while the integration of detection information avoids malicious switching triggered by replay attacks. By using multiple trigger conditions to establish an event-triggered strategy, under the premise of alleviating the shortage of communication resources, the system can be updated in time after the end of the abnormal state. Secondly, according to the characteristics of unknown but bounded disturbances, the ellipsoid set-membership estimation theory is leveraged to encrypt and decrypt the trigger information to realise the detection of replay attacks. Robust predictive control algorithms, together with linear matrix inequalities and dynamic output feedback controllers, are used to derive sufficient conditions for the asymptotic stability of the closed-loop system. Finally, the presented approach is validated through simulation case studies to effectively cope with replay attacks characterised by different duration periods.