Research On Predefined-time Consensus Of Heterogeneous Multi-agent Systems

With the evolution of science and technology and network telecommunications,multiagent systems have received extensive attention in interdisciplinary fields such as control,biology,and artificial intelligence.Due to the independence of its own agents and the coordination and cooperation between multiple individuals,it has been widely used in aircraft collaboration,intelligent transportation,service robots,smart grid systems and other fields.However,in practical scenarios,multi-agent individuals are often heterogeneous systems and have clear limits on system convergence time.Therefore,based on many related theoretical research results,this article develops an intensive study on heterogeneous multi-agent systems based on predefined-time consensus,aiming to overcome the communication problem of heterogeneous multi-agent systems and clarify the system convergence time.The main research results of this paper are as follows:The heterogeneous multi-agent systems consider include first-order,second-order,highorder followers and high-order leaders,and this paper studies the problem of predefined-time scaled consensus under directed topology.Firstly,a distributed predefined-time observer is designed for the single-leader systems that can overcome the asymmetry of the Laplace matrix under the directed communication topology,and the estimated time of the leader states can be effectively controlled by the given predefined observation time parameters.Secondly,a class of time-varying functions with generalized properties is defined for individual followers,and then the sliding mode auxiliary function and controller protocol are constructed based on this function to ensure that the trajectory of the multi-agent systems is constrained on the sliding mode surface,and the predefined-time is introduced into the proportional consistency issue of heterogeneous multi-agent systems.Using algebraic graph theory and Lyapunov stability theory,it is shown that with the control protocol,high-order heterogeneous multiagent systems can achieve scaled consensus at predefined-time.Finally,the effectiveness of the control protocol is demonstrated by numerical simulations.Based on the predefined-time consensus of the third chapter,the predefined-time containment control question of nonlinear heterogeneous multi-agent systems is researched further.For first-order and second-order nonlinear heterogeneous multi-agent systems composed of multiple leaders and multiple followers,a distributed predefined-time observer is designed to estimate information about the states of the convex hull formed by multiple leaders under the undirected topology graph,which translates containment control into successful tracking of the observer states by multiple followers,and the designed observer is able to observe the leader states within a predefined-time.Then,based on the time-varying function and adaptive neural network,the controller protocol is designed to approximate the uncertain nonlinear function by means of a neural network,and at the same time introduce the predefined-time into the heterogeneous multi-agent systems containment control question.Using algebraic graph theory,and Lyapunov stability theory,it is shown that with the control protocol,a nonlinear heterogeneous multi-agent system can achieve containment control at a predefined time.Finally,the effectiveness of the control protocol is demonstrated by numerical simulations.