Path Planning For Multi-robot Systems Based On Behavioral Dynamics Method

Path planning is one of the most basic fundamental problems of robot research,and also is the fundamental guarantee of robot to complete the task.With the continuous development of science and technology,the robot control technology research has developed from single robot system work to multi-robot systems.Compared with single robot,multi-robot systems has significant advantages,such as broader applications field,better fault tolerance,lower operating cost in complex tasks,more efficient,better scalable,and easier to study and development.So the research on the path planning for multi-robot systems has an important academic and practical meaning.The purpose of multi-robot systems path planning is to search a collision free smooth path from the starting point to the target point for every robot.In this thesis,the multi-robot systems path planning is researched with the improved behavioral dynamics methods in a static environment.The main contents of this thesis include as follows:Firstly,chapter one mainly introduces the background and current situation development of the multi-robot systems path planning technology,and the commonly path planning methods of single robot system and multi-robot systems.The key technical problems in the process of the multi-robot systems path planning and the solution of these problems are introduced in chapter two.In addition,the brief introduction and the principle analysis of the three basic behaviors,namely,towards the target behavior,obstacle avoidance behavior and collision avoidance behavior are given in chapter two.Secondly,in chapter three,the behavioral dynamics method is improved by the reciprocal velocity obstacles.The improved method fully considers the velocity information and the "autonomy" of the mobile robot,making the robot get the "intelligent" which the obstacle doesn't have.The vice target point is determined using sensor model to improve the adaptation of the dynamic path planning method,and the along-wall behavior is designed by the fuzzy controller to solve the "dead lock" phenomenon which is caused when the mobile robot is located in the concave obstacle environment.At the end of this chapter,the feasibility and effectiveness of the improved behavior dynamics method is verified through simulation.Once again,the static weight immune planning algorithm is proposed to fuse behaviors in chapter four,which not only has the good abilities of seeking the global optimization result,but also get faster searching speed and better convergence performance through making two improvements on the basis of the traditional methods.And the simulation of this method proves the effectiveness and the superiority of this method,and can effectively reduce the"shock" phenomenon of competition dynamics fusion method through simulation.Lastly,concluding remarks are presented at the end of this thesis.Some open problems are also pointed out,which deserve further study.