Distributed Learning Agents with Motivation for Cellular Warehouse Problem
Katsumi Hama*, Sadayoshi Mikami**, Keiji Suzuki**, and Yukinori Kakazu***
*Department of Mechanical Engineering, Hakodate National College of Technology, 14-1 Tokura-cho, Hakodate, 042-8501, Japan Oak
**Department of Media Architecture, Future University-Hakodate, 116-2 Kameda-nakano-cho, Hakodate, 041-8655, Japan
***Hokkaido University - Computing Center; Dept. of Eng., Lab. of Complex Systems Eng., Hokkaido University, North 13, West 8, Kita-ku, Sapporo, Hokkaido 060-8628, Japan
In an approach to resolve motion conflicts of transport pallets for cellular warehouse problems, pallets are considered autonomous agents and built-in behavior provided by ANN and problem-oriented connection weights evolved using Evolutionary Programming navigate agents to goals. To determine agents to be moved, priority is introduced and the measure of each agent changes based on the results of its motion and interaction with other agents. The solution of the problem is the motion sequence of agents. The effectiveness of the approach is demonstrated by computer simulation.
This article is published under a Creative Commons Attribution-NoDerivatives 4.0 International License.
Copyright© 2002 by Fuji Technology Press Ltd. and Japan Society of Mechanical Engineers. All right reserved.