Research On Event-triggered Secure Consensus Control Of Multi-UAV Systems

The consensus control of multiple unmanned aerial vehicle(multi-UAV)systems has attracted the attention of scholars as a hot issue in collaborative control.At the same time,as the research foundation of multi-UAV systems,the improvement of consensus control is of great significance for the research of clustering,tracking,formation and other issues.However,in traditional consensus control protocols,achieving consensus in multi-UAV systems requires a large amount of information exchange between individuals and requires the system to be configured with high-performance hardware devices.In order to reduce information exchange between UAVs and diminish the requirements for hardware devices,this thesis introduces eventtriggered control strategy into the research of consensus issues in multi-UAV systems.However,the UAV systems are exposed to various network threats due to its dependence on public network transmission.When multi-UAV systems are subject to unknown interference or cyber attack,more data packets need to be transmitted to achieve better consensus control performance.On the basis of the existing research work,this thesis applies knowledge such as algebraic graph theory,matrix theory and stability theory to study the security consensus of multi-UAV system under event-triggered scheme.The main contents are summarized as follows.To address the issue of event-triggered security consensus for fuzzy multi-UAV system under the leader-following framework,a memory-type event-triggered scheme is designed by using the historic data packets of the following UAVs for consensus control.While reducing redundant data transmission,it makes the multi-UAV system more sensitive to deception attacks,and the data releasing rate is improved so that the controller can receive more information to ensure control performance.At the same time,it can increase the event-transmission frequency when the system state response is at the peak or trough,ensuring the integrity and high reliability of the system under the influence of deception attacks.Further,comparative simulation experiments verify the feasibility and effectiveness of the proposed control strategy.Aiming at the problem that the memory event-triggered control cannot adjust the data releasing rate in real time according to the system state,the secure consensus tracking control of multi-UAV system is studied considering the deception attack and actuator saturation constraint.Considering the wind disturbance,the uncertain disturbance inside the actuator,the actuator saturation and the deception attack injected into the system during the flight,a novel adaptive data transmission scheme is proposed under a unified framework.By introducing measurement error between UAVs,the designed threshold can be adjusted online according to the fluctuation of system state,and the data with little change in the latest released data packet can be actively discarded.By using a designed adaptive event-triggered mechanism,the data releasing rate can be reduced while increasing sensitivity to malicious attacks for the system.In addition,with the help of a new Lyapunov functional,sufficient conditions and system security criteria are provided to ensure the consensus between all followers and the leader UAV.Based on the analysis of memory-based and adaptive event-triggered consensus control for multi-UAV system under Do S attacks,an adaptive memory-based event-triggered scheme is studied to reduce the occupation of network communication bandwidth and improve the consensus performance of UAVs.The number of historic transmission packets stored in the sequence buffer can be dynamically changed according to the proposed triggering scheme,and the threshold can be adjusted adaptively based on the state error.Under this communication scheme,segmented controllers are constructed for UAVs with different sequence buffer lengths.Then,in view of the fact that the end of Do S attack occurs at the non-sampling instant are not considered in the existing research,an effective Do S attack(EDA)period model is proposed.In addition,a consecutive transmission attempt mechanism is designed during the EDA to update the latest transmitted control signal at the end of the Do S attack.In exchange for a lower data releasing rate of the overall system,this control strategy sacrifices the computing power of the UAV itself before the system stabilizes,thereby improving the endurance of the UAV and allowing the network to accommodate more UAVs.Finally,the effectiveness of the consensus control strategy is verified by simulation analysis.To address the issue of multi-UAV system subject to deception attacks,an event-triggered consensus control strategy based on historical data is studied to further simplify the design of event-triggered algorithms and eliminate the impact of attacks on the system.By integrating historic data information and current triggered data into the design of adaptive thresholds,the system can be more sensitive to the data close to the peak/trough,thereby avoiding information loss.Further,by replacing the input of traditional event-triggered scheme with the average value of historic data information within a specific time interval,this transmission scheme can eliminate data mutations and adverse data-releasing caused by instantaneous random jitter and deception attacks.With the application of the attack estimation observer,a distributed consensus controller is designed for each UAV to achieve the control objective and eliminate the impact of the attack on the system.Sufficient conditions for global consensus are given by stability theory.