Recently, the research and development of the highly intelligent agent systems, such as ITS (Intelligent Transport Systems), have been performed with the progress of the information infrastructure. In such an intelligent agent system, a man-machine system is very important.
In this paper, we pay attention to the display function that becomes an interface between the human and the machine. This display function needs to present the suitable display for the user.
Therefore, we propose a method which is able to acquire the display knowledge with a few learning times by the interaction with the user. We realize this method based on Q-Learning, and we also use the “Chaotic Evolution method” to improve Q-Learning that requires many learning times.
In this paper, the proposed “Q-Learning with Chaotic Evolution” is applied to the driver's support display agent in the commuter car system, which is one of man-machine systems of ITS. The effectiveness is verified by simulation and experiment.
抄録全体を表示