The Study On Consensus Tracking Control Of Multi-Agent Systems With Output Nonlinear Constraints

Generally,whether it is a single robot system or a multi-agent system composed of robots,it is mainly made up of controlled objects,controllers,actuators and sensors.The control performance of the controlled system is not only related to the controlled object,but also affected by the physical characteristics of other devices in the control system.For example,in the actual industrial production process,the feedback path and the forward path of the system are susceptible to some non-smooth nonlinear constraint interference.However,in multi-agent systems,the compensation control of the sensor nonlinear constraints in the feedback channel becomes more difficult than the actuator nonlinear constraints in the forward channel,because it may lead to unknown and time-varying control gain problem.Due to the existence of unknown high-frequency gain,the Nussbaum-type method will cause some unknown transient performance.It is worth noting that it is difficult to achieve the transient and steady-state performance requirements for multi-agent systems by simply adjusting the Nussbaum-type gain.In addition,the convergence rate of the control system is an important indicator to evaluate the performance of such system.In the case of coupling uncertain dynamics between agent and its accessible agent,how to overcome the influence of nonlinear constraints,and ensure that the multi-agent achieves rapid convergence rate in a limited time becomes extremely challenging.In view of this,this paper uses Backstepping technology as a framework,and takes the universal approximation characteristics of tools such as radial basis function neural networks and fuzzy logic systems to carry out the study on adaptive consensus tracking control of multi-agent systems with output nonlinear constraints.The main work can be summarized as follows:In chapter 1,we introduce the background and existing studies on nonlinear multi-agent with output nonlinear constraints,and summarizes the main research contents for this paper.In chapter 2,the preliminary knowledge of this thesis are given.Including graph theory;intelligent control methods of radial basis function neural networks and fuzzy logic systems with universal approximation properties;adaptive dynamic surface backstepping control;finite-time stable control theory and other related knowledge.In chapter 3,for the problem of unknown control direction,in this chapter,two mainstream control methods are studied.One is the Nussbaum gain,which is used to solve the time-varying and unknown high-frequency gain,which ensures that the tracking error ofthe multi-agent system converges to a bounded region or asymptotically converges to the origin.Secondly,in view of the limitation that the Nussbaum-type method can only guarantee that the control error of the closed-loop system gradually converges to zero,a logic switching rule based on Lyapunov is proposed,and a new switching control is designed based on the switching rule strategy.It is worth noting that with the help of this control strategy,the state information of the multi-agent system can quickly converge to the equilibrium point within a finite time range.In Chapter 4,a distributed dynamic surface neural consensus tracking control method is proposed for a class of nonlinear multi-agent system with partial tracking error constraints and unknown output hysteresis.Firstly,since the unknown high-frequency gain caused by output hysteresis is time-varying,a new Nussbaum-type gain function is adopted to successfully solve the unknown multi-gain control problem of multi-agent systems caused by output hysteresis nonlinearity.Secondly,the intervention of Nussbaum-type functions results in the performance of multi-agent systems under transient conditions cannot be guaranteed.Under such circumstances,in this chapter,by combining the new Nussbaum type function and the prescribed performance control(PPC)method,the distributed collaborative control scheme proposed solves the transient performance problem of multi-agent systems with output hysteresis.Under such circumstances,this chapter solves the transient performance problem of multi-agent systems with output hysteresis by combining the new Nussbaum-type function and prescribed performance control(PPC).Simulation experiments show that the PPC-based consensus tracking control strategy designed in this chapter not only ensures that all signals of the closed-loop system are bounded,but also the transient performance of the cooperative tracking controller is always limited to the defined range.In Chapter 5,the finite-time adaptive fuzzy consensus stabilization is investigated for a class of leaderless multi-agent system with unknown output dead-zone nonlinearity.Note that the control gain of each agent caused by the nonlinearity of the output dead-zone is unknown and time-varying.Because the Nussbaum method has the characteristics of progressive convergence,it cannot be used to solve the fast convergence problem of multi-agent systems with output dead-zone.At the same time,unlike the problem of asymptotic stability,the finite-time stability can ensure that the system state variables quickly converge to the system equilibrium point within a finite time.However,due to the existence of the output dead-zone,only the finite-time stability theory cannot overcome theunknown control gain problem of multi-agent systems,which makes the control design and stability analysis of this problem very complicated.In order to solve this control problem,a logic switching rule based on Lyapunov is used to design the switching controller in the proposed finite-time adaptive consensus control design process.Simulation experiments prove that the proposed scheme not only ensures that the output tracking error of the system converges to an area around zero with adjustable size,but also enables the multi-agent system with output dead-zone to achieve state stability within a finite time.