Research On Obstacle Avoidance And Dynamic Hunting Of Mobile Robot

Mobile robot is an integrative system that can apperceive environment and itsown status through the sensors, realize the object-oriented independent movement andcomplete the scheduled task under the complicated and unknown environment.Mobile robot is being studied by more and more scientists and researchers because ofits wider application space. In this thesis, AS-R robot that can be used for teaching,research, competition and many other fields is regarded as the research platform, andthe single robot’s path plan for obstacle avoidance is studied deeply. Then, theformation problem of multi-robot is researched, the problem how to hunt a dynamictarget through multi-robot’s cooperation is studied based on the research of multiplerobots in formation problem.The research of single mobile robot’s path plan method for obstacle avoidanceincludes static obstacle avoidance and dynamic obstacle avoidance. In the obstacleavoidance for static-obstacle part, the grid method, the artificial potential field methodwith improved repulsion potential function, and the potential grid method for obstacleavoidance are studied respectively. Among them, the Goal Nonreachable withObstacle Nearby (GNRON) problem has been solved through the artificial potentialfield method with improved repulsion potential function, the simultaneous problemsof GNRON and local minimum point has been solved with the potential grid method.In the obstacle avoidance for dynamic-obstacle part, the move vector of obstacle andits position relative to mobile robot are introduced into the obstacle avoidanceinformation. The velocity vector potential field function has been established, canovercome the default of the traditional artificial potential field method effectively, andcan enhance the efficiency of path plan for obstacle avoidance.The research of multi-robot’s formation problem includes formation formed andformation control. On the formation formed respect, the target formation is composedefficiently by bid jointly target point method. On the formation control respect, ahybrid formation control method which is composed of leader-follower method andbehavior based method is proposed, can make up of the shortage of the traditionalleader-follower method.In multi-robot cooperative hunting a dynamic target issue, the movement strategyof target and hunters are designed respectively. According to different situation, targetrobot can use the maximum speed escape or breakout movement strategy to evadetrackers. However, hunters are divided into pursuit group and intercept group, theytrack the target robot with the movement strategy that is compose of intercepting thetarget robot face to face and narrowing the ring of encirclement. Finally, the feasibilityof the proposed cooperative hunting strategy has been proved by the MATLABsimulation result. The summary of the full text and the view of multi-robot system research inrobustness and cooperative hunting many targets have been shown in the end ofthesis.