Distributed Coordinated Tracking Control For Multiple Euler-Lagrange Systems

In recent years,network communication and computer technology have achieved a rapid development so that it expands the application range of multi-agent systems.There has been an increasing research interest in the coordinated control of multi-agent systems,which are widely used in the fields of robot systems,spacecraft formations,and sensor networks.In the research of leader-following control for multi-agent systems,the behavior of the leader is generally independent of the followers.If we make a reasonably good motion plan for the leader,it will achieve global leader-following consensus.Therefore,it is necessary to investigate the leader-following coordinated control problems.Most practical systems have the inherent nonlinear dynamics.Thus most existing work in distributed control for linear multi-agent systems can not be applied directly.Therefore,it has more theory and practical significance to investigate the coordinated control problems for nonlinear systems.Many nonlinear physical systems can be described by Euler-Lagrange(EL)equations.Therefore,distributed coordinated control for multiple EL systems has wide application prospects.The coordinated control problem for multiple EL systems has achieved many initial results.This dissertation reviews resent results of multi-agent system and give a further study on the distributed coordinated tracking control for multiple EL systems.Main contributions of this dissertation can be summarized as follows:The distributed coordinated tracking problem under directed topology is investigated for multiple EL systems in the presence of model uncertainties and external disturbances.Only a subset of followers can get the information from the dynamic leader.First,when the structure and parameter information of nominal systems can be obtained,with the upper bounds of the general disturbances and the leader acceleration,a distributed sliding-mode tracking scheme is proposed.When considering the high conservation of the distributed sliding-mode control,an improved distributed sliding-mode tracking algorithm with adaptation laws for the upper bounds is developed.both proposed methods can make all followers converge to the dynamic leader asymptotically.Then,when the structure and parameter information of nominal systems can not be obtained,two adaptive coordinated tracking control strategies by using backstepping technique for multiple EL systems are proposed.A distributed coordinated adaptive control scheme is given in the presence of parametric uncertainties.Subject to nonlinear uncertainties and external disturbances,an improved algorithm is developed by using neural-network approximation of nonlinear functions.Both proposed methods can make tracking errors for each follower ultimately bounded.The distributed coordinated tracking problem under directed topology is investigated for multiple EL systems in the presence communication constraints.First,considering that the relative velocity information between agents in unmeasured and communication delays do not exist,a distributed adaptive neural-network tracking algorithm is investigated for the dynamic leader case.Relative velocities are estimated through the low-pass filters.Neural-network method is used to approximate and compensate the nonlinear uncertainties of systems.It is proved that the tracking errors between leader and followers are uniformly ultimately bounded.Then,considering communication delays and the static leader case,the distributed observer is designed such that the state information of the leader can be estimated by each follower.A distributed adaptive control algorithm based on neural-network is proposed to ensure tracking errors for each follower can be asymptotically convergent.Based on the first two control laws and considering communication delays and the dynamic leader case,an improved distributed observer is designed so that each follower can obtain the state information of the leader.Three distributed adaptive neural-network tracking algorithms are proposed,which not only ensure the tracking errors to be bounded,but also improve the tracking accuracy and solve the chattering problem caused by discontinuous functions.The finite-time coordinated tracking problem under directed topology is investigated for multiple EL systems in the presence of model uncertainties and external disturbances.First,when the structure and parameter information of nominal systems can be obtained,a strictly distributed finite-time tracking control method is designed by the combination of matrix properties of graph theory and terminal sliding-mode based design techniques.To solve the singular problem,an improved distributed finite-time tracking algorithm is developed based on non-singular terminal sliding surface.Then,when the structure and parameter information of nominal systems can not be obtained,a distributed finite-time adaptive tracking control strategy is proposed based on backstepping and neural-network techniques for multiple EL systems in the presence of nonlinear uncertainties,external disturbances,and actuator nonlinearities.When considering that there exists incomplete known state,a modified distributed finite-time tracking control strategy is designed based on high-gain observers.Simulation examples of attitude synchronization for spacecraft formation flying are conducted to validate the effectiveness of all the proposed control methods.