Cooperative behavior rule acquisition for multi-agent systems using a genetic algorithm

In autonomous agents systems, each agent must behave independently according to its states and environments, and, if necessary, must cooperate with other agents in order to perform a given task. Therefore, each agent must incorporate learning and evolution in order to adapt to a, dynamic environment. At present, in the field of multi-agent systems, methods by which to acquire the behavior rule from both expert knowledge and perception for an autonomous agent are generating a great deal of interest. In the present paper, we focused on the problem of "trash collection" by a multi-agent system and simulated the cooperative behavior of agents. Therefore, we investigated methods by which to learn the rules of cooperative behavior of multi-agents so as to solve problems effectively. We also used genetic algorithms (GA) as a method of acquiring the rules of an agent. Individual coding (definition of the rule) methods are performed, and the learning efficiency is evaluated.