Observer-based Event-triggered Secure Collaborative Control Under Denial Of Service Attack

In recent years,the cooperative control problem of multi-agent systems has attracted the attention and research interest of scholars due to the wide application.In addition,with the development of the networked control,multi-agent systems begin to use networked control structure.However,the networked control inevitably brings some issues,such as packet dropouts,time delays and network attacks,among which network attacks have the worst influence on the system.Therefore,completing all kinds of control tasks while ensuring the safety of multi-agent systems is a problem worthy of in-depth study.This thesis mainly studies the security cooperative control problem of multi-agent systems under denial-of-service attacks,and provides the security control schemes for linear and nonlinear multi-agent systems,respectively.The observer-based event-triggered cooperative control problem for multi-agent systems under denial-of-service attacks is investigated.To deal with situations where the system state is unmeasurable,an observer is designed to estimate the system states.Furthermore,to save communication resources and achieve the objective of the cooperative control,an event-triggered security control scheme is designed based on the estimated states,which can enable the multi-agent systems to resist the influence of network attacks.Additionally,the Zeno behavior can be eliminated effectively by introducing a positive constant into the designed event-triggered mechanism.The observer-based fully distributed cooperative control problem for multi-agent systems under denial-of-service attacks is studied.Firstly,a switched fully distributed control framework is established for a class of network attacks constrained by the attack duration.Then,based on the established control framework,an attack-resilient security control scheme is proposed to achieve the cooperative control objective,in which an observer-based attack compensator is designed to alleviate the influence of network attacks.The problem of the data-driven security cooperative control for nonlinear multi-agent systems under denial-of-service attacks is studied.Firstly,an improved dynamic linear method is proposed to transform the nonlinear multi-agent systems into an equivalent linear data model.Moreover,in order to alleviate the influence of network attacks,an attack compensation mechanism is designed.Based on the equivalent linear data model and the established attack compensation mechanism,a distributed model-free adaptive control scheme based on the system input and output data is proposed,which can provide a unified framework to realize three cooperative control tasks.