Research On Individuated Cooperation Of Multi-Robot System Based On Game Theory

With the high development of computer science and artificial intelligence, robots have been applied widely in all fields such as social production, defense industry, service trade, intellectual education and so on. For the reason of more and more complex and difficult task, the study of multiple-robot systems naturally extends research on single-robot systems. Many experts and engineers from different countries are achieving systems of multiple mobile robots engaged in collective behavior.For the cooperative systems, the architecture concludes all the structural models and their relationships of the multi-robot system and it is the foundation of the whole system. Based on the architecture, the robots organize according to some strategy, cooperating mechanism, and control rules. They collaborate with each other to get the whole view of the system actions and adopt the effective strategy to accomplish the local control.The paper integrates the multi-agent system and evolution control, and establishes the hybrid architecture of multi-robot system based on individuation level. The architecture combines the virtue of functional architecture and behavioral architecture and integrates the behavior intelligence and real-time response. The definition of individuation level is related to the evolution control theory, which enables the system to evolution. The robot is defined as an agent in the system, which insure the system of the open exploitation. We introduced the main idea, the formalized description and the basic flow of game theory, and explored the Nash Equilibrium and Iterated Games Theory. We look upon one collaboration action as one game round, and the result of the collaboration is to get the set of the best strategies, which is equivalent to achieve the Nash Equilibrium in a game. So we can apply the game theory into the multi-robot collaborative system.In conclusion, we give the results of experiment, meanwhile , we summarize our work and put up our goals for the further research.