Research On High-order Nonlinear Multiagent System Consensus Control Under The Unreliable And Limited Communication

Multiagent as a typical networked system,its consensus control has received extensive attention with the continuous development control theory,sensor theory,communication theory,etc.In multiagent systems,each agent is vulnerable to suffering from input saturation and external disturbances.According to different control objectives and practical issues,the position,velocity,and outputs need to be constrained.The communication information between agents is transmitted through a communication network.Moreover,actuators,controllers,and sensors also transmit information through the communication network.Therefore,in a complex network environment,the network communication resource is vulnerable to being limited,and the network communication bandwidth is vulnerable to being limited,which will to a limited communication environment.Moreover,the multiagent systems are vulnerable to network attacks due to the traits of openness,which will lead to an unreliable communication environment.Based on the above discussion,this paper has considered the practical problems comprehensively,such as unknown disturbances,input saturation,state constraints,limited communication bandwidth,limited communication resources,and network attacks,this paper has studied the consensus control of nonlinear networked multiagent systems under unreliable and limited communication.The main research work is summarized as follows.(1)For the consensus problems of limited communication resources,input saturation,and external disturbances that exist in multiagent systems,this paper proposes an adaptive command filtered control method based on a dynamic event-triggered mechanism.First,by combining with the adaptive neural network theory,the anti-saturation auxiliary system and disturbance observer are designed to compensate for the effects of input saturation and external disturbances,simultaneously.Second,this paper proposes a novel dynamic eventtriggered mechanism,which can dynamically adjust the update frequency of the control signal to reduce the information transmission from the controller to the actuator and save the network communication resources.To avoid solving complex analytical derivatives,command filters have been introduced,which can reduce the online computational burdens caused by frequently calculating the derivatives of virtual control signals.Finally,based on Lyapunov theory,it is proved that the proposed method can make the consensus tracking error converge to the desired neighborhood under limited communication resources.(2)For the consensus problems of limited communication bandwidth,switching communication topology,and state constraints that exist in multiagent systems,an adaptive time-varying full-state constraints quantized consensus control strategy is proposed based on integral barrier Lyapunov function(IBLF).For the problem of communication bandwidth constraints,hysteresis quantizer is introduced,which can reduce communication burdens in the process of digital signal transmission.By designing time varying IBLF,the full-state constraints control can be achieved for multiagent systems,which can relax the conservative feasibility conditions in state constarints control.This method can directly design the time-varying state constraint parameters on system states under the condition of limited communication bandwidth,which has important theoretical significance for timevarying position and velocity constraints in practical multiagent systems.(3)For the consensus problems of limited communication bandwidth,state constraints,and external disturbances that exist in multiagent systems,this paper proposed an IBLF backstepping command filtered input quantization consensus control approach based on a time-varying state constraints estimator.First,based on IBLF theory,a time-varying state constraints estimator is designed for the disturbances and nonlinear functions.Due to the proposed estimator does not need to update weight online,it can effectively reduce the online computational burdens.Based on Lyapunov it has proved that the feasibility conditions are not required in the proposed estimator,and all the system states will not violate the time-varying bound.Second,based on filtered error,IBLF functions are designed to effectively reduce the computational burdens caused by solving the derivative of virtual control signals under IBLF theory.A quantizer is introduced to solve the limited communication bandwidth,Finally,the nonlinear multiagent system with different orders and dynamics is used to validate that this approach has better control performance under the conditions of limited communication bandwidth and state constraints.(4)For the consensus problems of limited communication resources,input saturation,and Denial of Service(DoS)attacks that exist in multiagent systems,this paper studies an adaptive command filtered resilient event-triggered secure consensus control.First,to handle the problem that the system state cannot be obtained during DoS attack,a novel adaptive neural network switched observer is designed by introducing the switching signals.By combining with the Lyapunov analysis,state conversion,and neural network,which can simplifiy the process of solving the observer gains.Second,the error compensation signal is designed to compensate for the filtering errors during DoS attacks.Third,based on the consensus tracking error and the observer error,a novel event-triggered mechanism is designed to to improve control performances under the limited communication resources and unreliable communication.The proposed method allows DoS attacks to be probabilityindependent and attack both communication channels and control channels.(5)For the consensus problems of limited communication resources between agents and the deception attacks,an adaptive event-triggered communication anti-deception attacks leader-following consensus control approach is proposed.The deception attacks are considered as actuator attacks and sensor attacks.A deception attacks compensator is designed based on adaptive neural network method.Different from previous studies,the problem of communication resources between agents is considered,and an event-triggered communication mechanism is constructed to achieve discontinuous update of communication information,thereby reducing information transmission between agents and effectively saving communication resources.However,the event-triggered communication mechanism leads to the problem of discontinuity.For this problem,a dynamic filter is introduced to obtain the derivative of the virtual control signals.Subsequently,based on the Lyapunov theory,it is proved that the consensus tracking error can convergence to the desired neighborhood.