Research On Adaptive Finite-time Cooperative Control Of Nonlinear Multi-agent Systems

With the development of artificial intelligence,the application prospects of distributed coordinated control of multi-agents in robot cooperative,aircraft formation and flexible manufacturing are promising,and it has attracted more and more attention.This paper studies the bipartite consensus tracking and containment control of nonlinear multi-agent systems under non-strict feedback.Aiming at the coopetition multi-agent systems,a control strategy that can achieve bipartite consensus in finite time is proposed under the condition of input saturation.And an adaptive finite-time containment control strategy based on neural state observers is proposed for multi-agent systems under unmeasured states.The main research contents of this paper are as follows:1.For the non-strict feedback nonlinear coopetition multi-agent systems with input saturation,by combining the adaptive neural control and the finite-time command filter backstepping,an adaptive neural finite-time bipartite consensus tracking control algorithm is proposed.During each backstepping process,the Radical Basis Function Neural Network is used to approximate the unknown nonlinear dynamics and the finite-time command filter is used to obtain intermediate signals and their derivatives.Moreover,the filtering errors are eliminated by using error compensation signals.By using the finite-time Lyapunov stability theory,it can be proved that the bipartite consensus tracking errors can converge to a sufficient small region of the origin in finite-time,and all signals in the closed-loop systems are bounded in finite-time although there exists the input saturation.The proposed control strategy is realized by Matlab/Simulink,and its effectiveness is verified.2.The neural state observer based adaptive finite-time containment control strategy for non-strict feedback nonlinear multi-agent systems is investigated.The finite-time command filter is used to overcome the explosion of complexity problem and the established fractional power based error compensation signal compensates the filtering error caused by the filter.Then,the finite-time command filtered backstepping based distributed control method combines the adaptive control technology and neural state observer is given,which ensures containment control errors reach to the desired neighborhood of the origin in finite-time in the presence of uncertain dynamics and unmeasurable states in the system.Finally,the given numerical simulations through Matlab/Simulink show the effectiveness of the proposed control strategy.