Adaptive Control For A Class Of Nonlinear Stochastic Multi-Agent Systems

The multi-agent systems are composed of multiple independent agents,and different agents cooperate with each other to complete complex control tasks.Each agent only uses some partial information to identify the adjacent agents.The distributed control method is used in various fields,such as,unmanned aerial vehicle control,formation control,distributed sensor network,swarming control and so on.In practice,a dynamic system always is affected by various stochastic factors,such as stochastic mechanism vibration,stochastic disturbance,stochastic communication noise,etc.These stochastic factors significantly affect the stability of the dynamic system.Over that last decades,the control problem of stochastic multi-agent systems has received widespread attention.Although there have been some results,in actual engineering,many tasks need to consider that some states are not measurable,many tasks need to consider the existence of competition relationships among agents,and numerous tasks require the system to have a faster convergence rate.The system needs to complete the corresponding requirements in a finite time,and so on.In order to enable the controlled system to operate more smoothly and meet increasingly accurate control requirements,there is still a need to design effective control schemes to cope with these difficulties.Based on the existing research results,combined with the distributed backstepping design approach for a class of nonlinear stochastic multi-agent systems,the main results of this paper are given as follows:Firstly,for a class of high-order stochastic nonlinear multi-agent systems,the adaptive fuzzy leader-following tracking control problem with unknown dead-zone input and unmeasurable state variables is studied,and a reduced-order observer with dynamic gain is designed.To reduce the computational burden of the system,a fuzzy logic system is introduced to approximate random disturbances and unknown nonlinear functions,and an adaptive fuzzy tracking controller for high-order multi-agent systems that can overcome unknown dead-zone input is constructed.It is shown that all signals of the closed-loop system are bounded in probability,and the consensus errors can converge to a small region of origin under Lyapunov stability theory.Some simulation results are given to show the effectiveness of the controller.Then,for a class of stochastic nonlinear multi-agent systems with input saturation,the problem of event-triggered bipartite tracking control is considered.Combining the distributed characteristics of the multi-agent systems,a new distributed reduced-order observer is constructed to estimate the unknown state of the stochastic nonlinear multi-agent systems.A relative threshold event-triggered mechanism is introduced to reduce the communication burden and an adaptive controller with bipartite tracking is proposed to handle the situation with negative weights in the topology,and a numerical simulation example is given to prove the given effectiveness of the method.Finally,the command-filter-based fixed-time bipartite containment control problem is studied for a class of nonlinear stochastic multi-agent systems.Using the characteristics of neural networks,the state coupling problem of stochastic multi-agent systems in the form of non-strict feedback is solved.The control protocol based on the command-filtered backstepping technique is proposed to ensure that the followers can converge to the convex hull formed by the leaders.The closed-loop stability of stochastic multi-agent systems is proved to be semi-global practical fixed-time stability.A numerical example simulation and an actual system simulation of a group of five single-link manipulator systems are presented to verify the effectiveness of the proposed method.