Research On Multi-USV Hunting Based On Game Theory

Unmanned Surface Vessel,as one kind of small and intelligent offshore platform,is playing an increasingly important role in the civil and military fields,and have become a research hotspot for experts from all over the world.Among them,the coordinated operation of multiple unmanned boats can complete more complex and difficult tasks,and improve the execution efficiency and success rate.For this reason,this article focuses on the hunting behavior in multi-vessel missions.The specific content is as follows:First of all,this article proposes a multi-USV target search method based on the threat degree function for the search problem of a single unknown target USV.After rasterized modeling of the search environment and unmanned surface vessels’ movement,the threat degree function is designed to select the staged search point,so that the searching boat is more inclined to select the point with better search effect,such as more inclined to search the undetected area,and each boat should be dispersed as much as possible while preferentially searching for areas that are closer to themselves;at the same time,the fitness function is designed based on the potential field method to make the search boat continuously approach the staged search points and avoid obstacles;and establish rules to replace the staged search points.Finally,the target search is realized;finally,the effectiveness of the method is verified through comparative simulation experiments.Secondly,this paper proposes a method of hunting multiple unmanned boats based on improved grid game to realize the hunting demonstration of the searched target boats.First,build a grid environment model.According to the game theory,select players in the game,design their feasible strategy set and profit function;after constructing the game model,solve the single-step game,and expand to multiple steps to finally achieve hunting;The method optimizes the problem of unsmooth trajectory and inability to avoid obstacles,and optimizes the team distribution income weight to improve the efficiency of hunting;secondly,design the corresponding anti-hunting strategy for the target boat that can escape intelligently;finally use the hunting method proposed in this chapter for simulation The experiment results show that the method can better realize the hunting demonstration of multiple unmanned boats,and Simultaneously verify the influence of escape boat speed,number of hunting boats and formation on the hunting effect at the same time.Combined with the target search in Chapter2,a multi-unmanned boat search-hunting algorithm is formed to realize the hunting of target boats at unknown locations,among which the target boats escape at a constant speed or intelligently way after mutually detection,and considers the situation that the target boat disappears again during the hunting process.Finally,this paper proposes a back-stepping sliding mode trajectory tracking controller based on disturbance observer compensation.Corresponding disturbance observers are designed to estimate the environmental disturbance of the hunting boat during navigation,and the controller is compensated accordingly,so as to reduce the influence of disturbance on the control effect and achieve more accurate tracking.First,processed the heading error and position error of the unmanned boat to obtain the virtual control law of longitudinal velocity and turning angular velocity.Then,combined with proportional integral sliding mode control,the control law of longitudinal thrust and turning torque is designed,and Prove the stability of the unmanned boat system.Finally,one uses the cubic spline interpolation method to interpolate and fit the discrete sequence points of the captured trajectory to obtain the desired trajectory that can be used,and conduct a simulation test.The results show that this method can better track the trajectory of the hunting.