Research On Consensus Control Of Multi-agent System Based On Event-Triggering

In the past decade,the cooperative control problem of multi-agent systems has received extensive attention due to its widespread application in engineering systems,such as smart grid,robot,UAV formation and satellite swarm.Consensus is one of the basic problems of cooperative control.Its goal is to design a suitable distributed control algorithm for each agent so that the states(or outputs)of all agents converge to the same vector.The leader-follower consensus problem is to make the states of all follower agents track the state trajectories generated by leader agents asymptotically,while the problem of non-leader consensus is to converge the state of all agents to the same trajectory asymptotically.We mainly focus on the following aspects:First,for linear multi-agent systems where states cannot be directly obtained,we design a state observer to estimate the unavailable states of the system.Then,taking the time-delay in the information exchange and the designed state observer into consideration,we propose an adaptive sampled-data event-triggered condition to reduce the unnecessary information transmission while ensuring the performance of the system.And the sampled-data event-triggered can naturally avoid the occurrence of the Zeno phenomenon due to its own characteristics and obtain the parameters in the controller and the observer together through the LMI inequation.In addition,the theoretical analysis of linear system can be extended to the nonlinear Lur’e system through simple modifications,which further proves the flexibility and practicability of the designed mechanism.The Matlab simulation examples verify the effectiveness of the design scheme.Compared with other methods,we have also demonstrated the superiority of the design algorithm.Secondly,under the constraint of input saturation in the system,the distributed event-triggered issue for multi-agent systems is studied.Here,we adopt the low gain feedback technique to deal with the saturation constraint and the controller gains with low gain parameters can be obtained by a Riccati equation.The proposed distributed event-triggered control strategy can ensure a positive time interval between two events while reducing the energy consumption of the system,that is,to ensure that the Zeno phenomenon is absolutely excluded.Finally,the feasibility of the controller scheme and event-triggered condition are confirmed by Matlab simulation.Finally,the problem of output consensus for a class of heterogeneous multi-agent systems with actuator failures is studied.Firstly,the observer with adaptive parameters is designed based on the local information of the multi-agent system to estimate the states of the leader.And the designed distributed event-triggered condition can save a lot of energy consumption and prolong the life span of the system.Zeno behavior can be avoided by specifying a strictly positive minimum inter-event time for each agent.For the general failure in the actuator,we design a controller with adaptive compensation mechanism to carry out fault-tolerant control of the system and ensure that the multi-agent system’s consensus protocol can still be realized when actuator failures occur.Simulation experiments further verify the effectiveness of the proposed distributed controller.