A proposal of CMAC with foresight knowledge layer for heterogeneous multi-agent system

The purpose of our research is the acquisition of cooperative behaviors in a heterogeneous multi-agent system. In this paper, we used the fire panic problem and we have tried to be learned how agents escape from the fire. Also, we have observed agents behavior and checked a character of agent types. In the fire panic problem, a fire exists in a field, and progresses step by step this fire spreads. The objective of the agent is to reach a goal position without touching the fire. The fire is able to grown three power levels during a few steps in a field. The agent has to run away from a fire and get a goal: exit. Our purpose in this paper is the acquisition of behavior to reach a goal in a changing environment, in this paper, we observe how agents escape from the fire by cooperating with other agents. For this problem, we propose a unique CMAC-based Q-learning system for heterogeneous multi-agent systems.