Research On Finite-time Consensus Of Multi-agent Systems Subject To Input Saturation

Multi-agents consensus is a basic problem in distributed cooperative control.The salient feature is that the controller is designed through local cooperative information,so that the state of each agent in the system tends to the same value,without relying on the acquisition of global state information of the system.At present,most of the results for consensus are asymptotically stable,that is,the time for the system to converge to consensus is theoretically infinite,and the convergence speed is slow.Compared with the asymptotically stable theory,the finite-time theory has a convergence time related to the initial position,which can improve the convergence speed of the multi-agent systems.At the same time,actuators in actual systems often have input saturation constraints.If they are not considered,they will affect the control accuracy and even make the system unstable.Therefore,it is particularly important to design an anti-saturation finite-time consensus controller for multi-agent systems.The main work of this paper includes:Firstly,the finite-time consensus problem of the second-order multi-agent systems under the input saturation constraint is studied.Under the assumption of the connected graph of the undirected graph,a nested homogeneous finite time consistent controller based on the single saturation function is given.Use the saturation function to limit the output of the controller,and use the nested homogeneous function to scale the state error of the current agent and its neighboring agents.The Lyapunov theory is used to verify the global stability of the control algorithm,and the advantages of the controller design are explained through comparative simulation verification.Secondly,the formation path tracking problem of a group of followers with input saturation constraints is studied.Using the leader-follower agent model,a virtual leader is introduced and the desired tracking path is transformed into the virtual leader's state.The relative state and formation position constructs the formation coordination error,and further extends the finite-time consensus problem to the multi-robots finite-time formation tracking problem.And theoretically proved that under the topology of undirected connected graph,the system is globally stable for finite time.The simulation example verifies that the designed controller can make the follower agent form a fixed formation in a finite time,and can track the target path.Finally,in order to verify the feasibility of the control algorithm in practical applications,a multi-robots test platform based on visual positioning is designed.The test platform uses the E-puck robot as the control object,the label-based visual positioning system and the WIFIbased communication system.Established a test platform for multi-agents related tests.Through the functional test of each system of the multi-mobile robots platform,the accuracy and stability of the platform are verified.In the multi-robots test platform,the feasibility of the limited time formation control algorithm was verified,the formation test effect was demonstrated,and the formation data was analyzed.