The problem of multi-robots round-up should not only solve the contradiction between multirobots,but also solve the real-time trajectory tracking problem of each robot to the target.With the advantages of simple algorithm structure,high real-time performance and small amount of calculation,the artificial potential field method is used as a path planning method to solve the problem of multi-robot rounding up.However,the algorithm has limitations: the target is unreachable and the local minimum problem exists.Therefore,this paper improves the artificial potential field method and applies it to the research of multi-robot round-up task.The main research contents are as follows:Aiming at the tracking path problem of the target in static environment,a simplified artificial potential field method for obstacle prediction collision is proposed.Firstly,the idea of collision prediction is introduced to help the robot predict in advance in the unknown environment and avoid the robot falling into local optimum.Secondly,the influence range of the robot by obstacles is simplified,so that it only considers the influence of obstacle repulsion on one side of the target point;Then,by adding virtual target points,the repulsion function of the influence range of repulsion factors is improved to guide the robot to reach the target point quickly,smoothly and without collision.Finally,by testing the problem,the effectiveness of the proposed algorithm is verified by the evaluation indexes such as path length,pursuit steps,path turning times and running time.Aiming at the real-time tracking problem of dynamic unknown target state,an adaptive Kalman filter-dynamic artificial potential field algorithm in double-layer mode of prediction and tracking is proposed to realize real-time tracking of dynamic trajectory.Firstly,the Kalman filter algorithm with adaptive error factor is used to predict the unknown target information in the upper layer,and then a virtual target point is set in the lower layer according to the prediction result of the upper layer and the updated obstacle type in the environment.By using the repulsive force of the dynamic artificial potential field,the optimal target is solved between the virtual target point and the prediction value of the upper layer for tracking.Finally,the experimental simulation is carried out according to the test questions,and the efficiency of the algorithm is visually verified by evaluation indexes such as tracking time,path length,safety and tracking stability.Aiming at the cooperative round-up task of multi-robots,a cooperative round-up strategy of multi-mobile robots based on game theory model is proposed.Firstly,this strategy proposes a fixed-point negotiation round-up strategy to get the expected round-up point.Then,according to the game theory,taking the expected round-up point and the rounded-up mobile robot as players,a feasible strategy set and a round-up revenue function to realize the minimum round-up path are established.The first two parts of this paper are used as path planning algorithms to achieve realtime and efficient round-up.Finally,the effectiveness of this algorithm is verified by simulation experiments in the MATLAB test platform and the simulation environment of Webots. |