| With the development of computer science,artificial intelligence and other disciplines,the consensus control of nonlinear multi-agent systems possesses important applications in the fields of wireless sensor network,military reconnaissance,intelligent transportation and so on.It has become a research hotspot of control theory and control engineering.Due to the limitation of sensor technology,the state information of many practical physical systems cannot be available effectively.Therefore,the research on output feedback consensus control of nonlinear multi-agent systems has great theoretical value and practical significance.However,nonlinear multi-agent systems have obvious features,such as numerous state variables,complex and diverse nonlinear functions,and network channels with limited capacity.There are some serious deficiencies and huge challenges in the research of pivotal technologies,such as state estimation,distributed control protocol design,and system transient performance improvement.In this dissertation,the output feedback consensus control problem is studied comprehensively for nonlinear multi-agent systems.The static gain design approach and the time-varying gain design approach are proposed skillfully,and a distributed output feedback control protocol is constructed effectively.Six difficult problems,namely serious waste of communication resources,poor transient performance of systems,serious uncertainty of systems,diverse time delay of state and input,complex modeling error and unknown measurement sensitivity,are overcome successfully.By improving the traditional control design approach and breaking through the existing observer construction technology,this dissertation focuses on the output feedback consensus control of nonlinear multi-agent systems under five cases,that is,certainty,uncertainty,time delay,modeling error and unknown measurement sensitivity.(1)The static gain design approach and the time-varying gain design approach are proposed for feedforward and strict-feedback nonlinear systems(Corresponding to the research content of Chapter 2).By elaborating the structural characteristics of a single feedforward and strict-feedback nonlinear system,the information transmission laws of system states and control input are revealed deeply.By using the forwarding design approach,the backstepping design approach,the static gain design approach and the time-varying gain design approach respectively,the feedback controllers are constructed to solve the global stabilization or regulation control problem for feedforward and strict-feedback nonlinear systems.Different from the forwarding design approach and the backstepping design approach,the static gain design approach and the timevarying gain design approach avoid the tedious construction procedure of feedback controllers,and have the advantages of simplicity and convenience in improving the system transient performance,accelerating the state convergence speed of feedforward nonlinear systems,and saving the control cost of strict-feedback nonlinear systems.(2)The output feedback consensus control problem is solved for certain nonlinear multi-agent systems(Corresponding to the research content of Chapter 3).When the functions of feedforward and strict-feedback nonlinear systems satisfy the Lipschitz condition with known constants,the directed fixed topology theory and the static gain design approach as well as the time-varying gain design approach are used comprehensively.Based on the output consensus error,a new observer can be constructed reasonably.Then,a static gain distributed output feedback control protocol and a bounded time-varying gain distributed output feedback control protocol are designed separately for different control objectives,which not only ensures that the states of the following agents converge to the corresponding states of the leading agent,but also improves the transient performance of systems significantly.Different from the existing relevant literatures,the observer and the controller of each following agent here only need the output information of its neighbor agents,which reduces information transmission effectively and saves communication resources greatly.(3)The output feedback consensus control problem is solved for uncertain nonlinear multi-agent systems(Corresponding to the research content of Chapter 4).When the functions of feedforward and strict-feedback nonlinear systems satisfy the Lipschitz condition with an unknown constant or a time-varying function,the directed fixed topology theory and L ’Hopital’s law are used simultaneously.The output feedback consensus controllers with unbounded time-varying gain are constructed successfully to solve the leader-following consensus control problem of feedforward and strict-feedback nonlinear multi-agent systems.By selecting appropriate time-varying gain design parameters and introducing the time-varying function with slow growth rate,the requirements of system transient performance,the convergence speed of state consensus errors and control cost are considered effectively.(4)The output feedback consensus control problem is solved for nonlinear time-delay multi-agent systems(Corresponding to the research content of Chapter 5).The nonlinear functions of multi-agent systems have not only the time delay of state but also the time delay of input.The form of time delay is more complex,and the existence of discrete delay and distributed delay is allowed.By using the directed fixed topology theory,the static gain design approach and the time-varying gain design approach are proposed.The delay-dependent output feedback consensus controllers are constructed.Based on the Lyapunov-Krasovskii functional theorem and the LyapunovRazumikhin theorem,the operating rules of state consensus errors can be analyzed effectively.The bounded time-varying gain controllers not only significantly accelerate the convergence speed of state consensus errors of feedforward nonlinear systems,but also greatly save the control cost of strict-feedback nonlinear systems.(5)The output feedback consensus control problem is solved for nonlinear multi-agent systems in the presence of modeling error(Corresponding to the research content of Chapter 6).The nonlinear functions of multi-agent systems are provided with a modeling error that depends on system states and the control input,which shows that the concerned systems have a more general structure.Based on the directed fixed topology theory,a static gain design approach as well as a time-varying gain design approach are proposed.By constructing a distributed output feedback control protocol,the states of the following agents converge to the corresponding states of the leading agent,if the strength of modeling error keeps within the specified boundary.Due to the existence of modeling error,the considered nonlinear multi-agent systems are no longer limited to the common upper triangle structure and lower triangle structure.Therefore,the favorable robustness of the constructed output feedback consensus controllers for nonlinear multi-agent systems is demonstrated.(6)The output feedback consensus control problem is solved for nonlinear multi-agent systems subject to unknown measurement sensitivity(Corresponding to the research content of Chapter 7).In this chapter,the system output contains unknown measurement sensitivity,which indicates that the considered systems are more complicated.Based on the undirected fixed topology theory,a novel dual static gain design approach and a new dual time-varying gain design approach are proposed skillfully.Then,a static gain output feedback consensus controller and a time-varying gain output feedback consensus controller are designed respectively.The leader-follower consensus control problem of nonlinear multi-agent systems can be achieved effectively,if the measurement sensitivity keeps within the specified boundary.By introducing two types of time-varying gains that increase slowly and tend to positive infinity,a sufficient condition for two types of time-varying functions are presented for the first time,which avoids the complex determination process of static gain effectively.To sum up the above,a feasible distributed output feedback control protocol is established in this dissertation for nonlinear multi-agent systems.Some innovative research results have been obtained.Specifically,the related research results have published in sixteen academic papers as the first author in the journals,such as IEEE Transactions on Automatic Control and Automatica.These research results enrich and improve the consensus control theory of nonlinear multi-agent systems.Moreover,these research results provide the theoretical support and the technical guarantee for achieving orderly and smooth operation of state consensus errors. |
PDF Full Text Request