A homogeneous agent architecture for robot navigation

A homogeneous agent architecture consists of agents having the same or similar structures in one system. The design introduces the advantage of reducing the development time. In this architecture, the engineer would only design one or few templates for agents, and every agent is constructed by one of these templates. Two categories of agents in the proposed system are the primitive agent and the behavioral agent. The behavioral agents play the role of leaders, so that they can produce output signals through combining the signals of those primitive agents. Hence, we could increase new capability of the whole system flexibly through adding new primitive agents with different capability and adding new behavioral agents to process their opinions. In this paper, we use this homogeneous agent architecture to control a autonomous mobile robot, and it performs well in simulation.