Virtual and intelligent traffic signs in rescue simulation system: Imitation of human society in agent society

Robocup rescue simulation system is a suitable test-bed for test and evaluation of multi-agent system´s related ideas and techniques. Hence, the world robocup competitions are held each year and the used ideas and techniques are evaluated in the form of different teams. In rescue simulation system, at the start of simulation, police agents should search and explore the earthquake area and open the blocked roads. In this paper, which is the first application of distributed learning automata in rescue simulation system, the agents imitate the ways which real people use for solving traffic problems in their societies. Like the real peoples´ solutions for traffic problem, the agents use virtual police and street signs in their virtual environment junctions. The results indicate the fact that the proposed method gives more exploration power to rescue agents.