Representation of the perceived environment and acquisition of behavior rule for multi-agent systems by Q-learning

Abstract

Multiple autonomous robotic systems can be represented by multi-agent. In multi-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. In the present study, we focused on the problem of ldquotrash collectionrdquo, in which multiple agents collect all trash as quickly as possible. The goal is for multiple agents to learn to accomplish a task by interacting with the environment and acquiring cooperative behavior rules. Therefore, for a multi-agent system, we discuss how to acquire the rules of cooperative action to solve problems effectively. We construct the learning agent using the Q-learning which is a representative technique of reinforcement learning. Regarding the perceived environment of agent, two representation methods are used. We then observe how the autonomous agents obtain their action rules and examined the influence of the learning situations on the system. Moreover, we discuss how the system was influenced by learning situation and the view information of the agent.