Key Technologies Of The Multi-robot Collaboration

Multi-robot cooperation is one of the important methods to resolve some problems that are complex, real time and collateral. So people pay more and more attention to this research.This paper summarizes the current research on multi-robot cooperation and sums up some correlational technologies, main research results and problems. The article mainly studies at the next four points:1,Multi-robot cooperative system architecture. To make the robots cooperate with each other , we have to make sure of the signal relation and the control relation at logic and physics. The paper gives an architecture which adapt to multi-robot cooperative hunting system, and descript parts of the architecture briefly. And the paper gives the method of disassemble and distribute jobs.2,Multi-robot cooperative localization. to multi-robot cooperation, robot's apperception has some effects: we can estimate the cooperationist's intent and act effect with apperception, then can reduce the dependence of communication. And apperception is the base of action, with good apperception, system can answer the tangent of environment quickly. The paper gives the method of localization based on relative observations with which the robots can localize each other, and then localize the goal.3,The goal of multi-robot path planning is to plan a path for every robot without hitting obstacle or other robots. This contains global robot path planning and local path planning. In the paper, the global path planning is based on genetic algorithm and the local path planning is based on moving area method. The global shortcut is given and there will be no hitting between robots and obstacle locally.4,To resolve the problem of goal's escape, a hunting method is designed for hunting robot. This method is based on dynamic virtual range, and can quicken the hunting. We lucubrate on and do some experiments of the multi-robot cooperation technologies for the multi-robot hunting formation application.In the end , a conclusion of this paper and prospect of the future work are drawn.