Research On Cooperative Pursuit Strategy Of Multiple Unmanned Surface Vehicle

As a vehicle suitable for performing dangerous tasks,intercepting or pursuing escape ships at sea is one of important application scenes of unmanned surface vehicle(USV).Under the complex obstacles such as marine environmental disturbance,offshore islands,dynamic ships,it is difficult for a single USV to complete the task of pursuing multiple escape targets.Therefore,the cooperative system consisting of multiple USVs with stronger applicability and extensibility has gradually become the current research trend.However,at present,there are still some problems in cooperative pursuit strategies for evade targets,such as simple pursuitevasion game motion model,loss of contact among capture members,inapplicability of static task allocation in situations of falling behind/joining the swarm,idealization of pursuit scenarios,and insufficient flexibility in pursuit formation.This thesis focuses on the interception and pursuit tasks of USVs in the unknown complex environment,including the capture escape game model,the allocation of multiple USVs tasks,the interception of single pursuit USV and USVs swarm against evade targets and field tests.For the pursuit-evasion game scene in the open environment,Apollonius circle is introduced into the USVs filed to solve the problem that grid scene cannot reasonably show the complexity of the actual game process,and improve the applicability of the pursuit formation in the continuous environment.For the unwise escape path caused by the simplification escape motion model in traditional strategies such as straight line,regular curve or random direction,a more practical model of pursuit-evasion game is established.The intelligent escape strategy is designed to simulate the flexible escape direction in the real scene,the reality of the pursuitevasion game and the applicability of the pursuit algorithm are improved.Based on the established pursuit-evasion model,the dynamic game situation between the pursuier and evader is represented,providing a theoretical basis for the interception/pursuit strategy in this paper.The research on task allocation for the encirclement task between multiple pursuit USVs and multiple escape USVs was carried out.Based on particle method,the modeling of multiple USVs with complex constraints based on location,performance,value and the pursuit alliance balance is established.In view of the problem that static task allocation is not applicable due to the variability and uncertainty of the battlefield environment,various situations such as lost communication,members falling behind/joining the swarm in the process of encirclement are considered.In order to improve the efficiency of cooperative encirclement,a dynamic task allocation model for the execution of the encirclement task is established.The designed task allocation method can meet the goal of optimizing the pursuit revenue under constraint conditions,and improve the efficiency of cooperative pursuit.The interception algorithm of single interceptor against single escaping target is studied.In order to improve the accuracy of the target interception guidance algorithm,an improved augmented proportional guidance law is designed for the extended term based on interception time.For the control of the USVs in the pursuit task under the uncertain disturbance of the practical marine environment,a real-time estimated disturbance observer is designed to improve the robustness of the control and achieve the control compensation for the disturbance.The interception effect is analyzed in simulation.The simulation results show that the improved augmented proportional navigation can improve the interception efficiency of pursuit USV against escaping targets.By combining the swarm control algorithm with the pursuit pattern based on Apollonius,a swarm cooperative pursuit algorithm for USV is proposed.The escape targets can be captured only by adjusting swarm pattern and swarm movement direction,which solves the inflexibility rigid structure formation and assignment conflict problem between multiple pursuers and capture pursuit in in traditional algorithms.Based on the point element strategy of discretizing the boundary of irregular obstacles,and considering the problems of collision avoidance and obstacle avoidance in the pursuit task under the environment of multiple obstacles,a collision avoidance area based on speed,size and turning radius is proposed.For dynamic obstacle avoidance,a two-level collision avoidance strategy based on collision regular rules and artificial potential field method is designed.The effectiveness of the aforementioned task allocation strategy,swarm cooperation strategy and obstacle avoidance strategy are verified by simulations in complex obstacle environment,and it improves the adaptability of control methods in USVs swarm under various obstacle environments,achieving satisfactory cooperative pursuit effects.USVs test platform of the pursuit-evasion game was built and interception experiments were carried out in various obstacle environments.The effectiveness of interception guidance laws and collision avoidance algorithms were verified.In addition,the swarm cooperative encirclement experiment was carried out,realizing the cooperative encirclement task of US Vs,which further verified the effectiveness of the swarm cooperative encirclement strategy in practical environments.