Collaborative Task Planning And Coordinate Collision Of Robot Swarm

This paper does some research on robot swarm, such as framework, mapping, searching, coordinate collision and etc. The robot swarm works in unknown environment, explores a large range of world and makes the task of target searching. The research focuses on the coordination of the robot swarm.Firstly, the thesis designs a hybrid layered framework, which is based on the traditional distributed framework. The design enhances the coordinate ability of every single robot of the swarm.Secondly, considering the searching range is very large and has some targets, the thesis firstly decomposed the task into subtasks, then assign the tasks to appropriate robot in the principle of auction. In addition, the thesis divides the whole search field into some parts by the size of the search field and the members of the robot swarm. The robots of the swarm is assigned a specific part by the position of the robot. At last, the robots equably distribute in the searching field, which ensure the real-time of the searching.Thirdly, the thesis does some simulation of field explore and target searching with the MoboSim software. The simulation shows that the method of this thesis is fit to explore a large range of field in some ways.Lastly, the thesis does some research on path planning and coordinate collision avoidance. When the robot swarm is searching a field, the robots must avoid colliding with other robots and the static and mobile obstacles of the environment. In addition, the path of the search should be optimized. In oder to make the targets mentioned above, this thesis hybrids the method of thread degree and artificial potential field and makes a new method of collision avoiding and path planning. The method increases the reaction speed and the safety of the robot swarm.