Strategy Determination by Statistical Technique in Autonomous Conveyance System

Abstract

In recent years, an aim of conveyance systems is to provide suitably for various demands and to treat a wide variety of products, therefore autonomous decentralization is needed in conveyance system. Because conventional conveyance systems are designed for efficiency and it takes much time and resource to change specification or cope with breakdowns. Many research works on autonomous distributed conveyance systems have been proposed so far. However, it is difficult to use conventional methods in dynamic environment with environmental change. Therefore, this paper proposes a strategy determination method by statistical technique. In this method, each robot estimates an environmental state with queueing theory in order to decide strategy autonomously. Moreover, each robot recognizes environmental change and expire the stored environmental state in order to decide a new strategy suitable for a new environment.