Research Of Multi-Robot Cooperation In The Small-Size League RoboCup

The Cooperation is the key research problem of MRS (multi-robot system), which is that robots are organized by some rules to work together for those tasks that cannot be completed by one single robot. The timely and exact opponent's intention recognition is necessary for MRS under dynamic environments to get predominance in a competition. The real time reaction of robots depends on intention recognition.The dynamic Coalition method based on opponent's intention recognition for MRS adopts by author in order to solve above problem in this paper. This method consists of three parts. The first is an architecture named dynamic coalition structure for robots organization, the second is gift value algorithm for tasks allocation while the third is quadratic estimation algorithm for opponent's intention recognition.Dynamic coalition structure is a profit-drive MRS architecture, which is global distributed while local weakly centralized. In MRS a task executive is produced at first, it is responsible for resolving tasks into several stage tasks, and to a certain stage task among them, a dynamic coalition made up of several robots is jointly responsible for finishing. A union executive is produced in the dynamic coalition; it is responsible for resolving stage tasks into several sub tasks and any sub task can be finished by a certain robot. Union executive assign every sub task to a robot to finish, thus must achieve the realization of the goal task.Tasks allocation adopts the method of electing first and then negotiating. Gift value algorithm for tasks allocation in this paper is two steps artifice. Firstly, estimate the task executing capabilities of all relative robots by gift value and select the one who give max value. Then, one to one negotiation happens between this robot and another robot, which is responsible for task allocation. The optimal solution of task allocation for different global objective can be obtained by adjusting weight vector for different estimating criterions in gift set.Quadratic estimation algorithm for intention recognition is also a two-step artifice. The first step is to estimate the opponent's behavior series form its current behavior. The next step is to recognize the opponents' intention from its behavior series gained in previous step. According to the result of the intention recognition, robots make the dynamic response to the environment. The system produce suddenly task and carry out, thus the Cooperation of MRS based on the intention recognition is realized.Finally, this paper with the aid of figure programming tool, with RoboCup small-size league robot soccer the system is for platform experimental, prove those methods is feasible and effectual.