[3O4-GS-5] Agents: social-problem solving

[3O4-GS-5-02] Analysis of Factory Automated Guided Vehicles Systems by using Multi-agent system

〇Daimotsu Kato¹, Shohei Yada¹, Setsuya Kurahashi¹ (1. Univ. of Tsukuba)

Keywords:Multi-agent, Factory Automated Guided Vehicles Systems

Automated guided vehicles (AGVs) are used in a flexible job-shop production system. In this type of production system, proper scheduling of tasks is the most important issue. This issue is considered to be non-deterministic polynomial time-hard and difficult to solve in a reasonable time. Moreover, there are constraints that AGV must pass on a predetermined transportation route in the semiconductor manufacturing process. This constraint causes spatial interference between AGVs, which results in congestion of AGV on the transfer task. In this paper, an analytical model is investigated by using the multi-agent system that takes into account the spatial interference model of the AGV. In addition, we applied the contract net protocol. Our analytical model in which the AGV assignment agent and the task manager agent collaborate make decisions using feedback of time-series information of traffic congestion. By using this approach, we found that traffic flow could be improved.