Research On Theory And Application Of Coordinated Target Fencing Control With Multiple Unmanned Surface Vessels

In recent years,with the increasing intensity of the maritime conflicts,the naval vessels from the United States and Japan have entered into China’s territorial waters frequently,which seriously threatens the maritime rights and the safety of our strategic islands and reefs.Coordinated target fencing is a common strategy in the defense task of strategic islands and reefs,which could effectively drive off dangerous targets.The utilization of the multiple unmanned surface vessels(USVs)to coordinately fence the dangerous target vessel enhances the advantages of low cost,high precision and great flexibility,which could effectively improve the efficiency and scope of the defense task of the islands and reefs.However,there usually exist complex characteristics such as the varying-velocity target and multiple targets in practical application,which easily lead to the failure of the target-fencing task.Moreover,due to the limitation of cost and usage environment for onboard sensors,the US Vs may only obtain the constrained bearing information,which also bring challenges to the target-fencing problem.Aiming at the aforementioned problems of complex target characteristics and constrained sensing,this thesis have deeply studied the design of distributed coordinated target fencing algorithms under constrains and conducted the coordinated target fencing experiments with multi-USV system.The main research works are given as follows:Firstly,a distributed coordinated fencing control algorithm based on nearest-angle repulsion is proposed.Aiming at the basic fencing problem with a constant-velocity target,a distributed cooperative fencing controller is designed with nearest-angle neighbors,which consists of agent-target regulation and inter-agent angle repulsion to achieve constantvelocity target fencing,collision avoidance among the agents,and regular pattern regulation,simultaneously.The asymptotic convergence of the closed-loop system is attained by the comparison principle and Barbalat’s lemma.Finally,the effectiveness of the controller is demonstrated by numerical simulations.Secondly,a distributed coordinated fencing control algorithm based on input-to-state stability is proposed.Aiming at the complex fencing problem with equal-rotating formation and taking into account the influence of the casually connected topologies,an estimationand-control hierarchical framework is established,which decouples the tasks of target estimation and coordinated target fencing.Then,a distributed cooperative fencing controller is designed with input-to-state stable,where the fencing task with equal-rotating formation is achieved under casually connected topologies.Moreover,asymptotical convergence is proved by Lyapunov stability and input-to-state stability.Finally,the effectiveness of the controller is demonstrated by numerical simulations.Thirdly,a distributed coordinated fencing control algorithm based on output regulation principle is proposed.Aiming at the complex fencing problem with a varying-velocity target,by regarding the varying-velocity target as an exosystem,a dynamic feedback fencing controller is designed with an internal model,which achieves the varying-velocity target fencing,collision avoidance among the agents,and rigid pattern regulation,simultaneously.Moreover,asymptotical convergence of the closed-loop system is proved by LaSalle invariance principle and Lyapunov stability.Finally,the effectiveness of the controller is demonstrated by numerical simulations.Fourthly,A distributed multi-target coordinated fencing control algorithm based on adaptive parameters is proposed.Aiming at the complex fencing problem of multiple targets with different varying velocities,a distributed estimator of the center of moving targets and a distributed fencing controller with adaptive parameters are designed,which overcome the influence of the unknown nonlinear dynamics.Moreover,potential functions are also designed to achieve guaranteed inter-agent collision avoidance as well.Finally,The effectiveness of the proposed fencing controllers with multiple moving targets are verified by real experiments with indoor multi-USV systems.Fifthly,A distributed coordinated fencing control algorithm based on persistent-exciting inputs is proposed.Aiming at the fencing problem under constrained bearing information,a distributed estimator is firstly designed with persistent-exciting inputs to approximate the target state.Moreover,based on the proposed estimator,a distributed hierarchical cooperative fencing controller is designed based on perturbed stability principle,which fulfills the target-fencing mission under time-varying topologies.Significantly,the asymptotic convergence of the closed-loop system is attained with rigorous analysis.Finally,the effectiveness of the proposed controller with bearing-only measurements is verified by various experiments with indoor multi-USV systems and aerial-marine cross-domain unmanned systems.Finally,the coordinated fencing algorithms and experimental results with multiple USVs are summarized,and the future directions of coordinated target fencing with multiple US Vs are exhibited.