Intelligent combinatorial auctions of decentralized task assignment for AGV with multiple loading capacity

The need to have an efficient transportation system has attracted worldwide attention. Although there is increasing demand to implement distributed control system for industrial applications, there is still an unexplored potential of deploying distributed transportation system. This paper focuses on dynamic assignment of transportation requests to a fleet of vehicles in real time. We introduce an improved combinatorial auction methodology to accommodate the distributed task assignment procedure. Based on a multiagent architecture, each vehicle is represented by an intelligent agent that bids for task and plans its own schedule. On the other hand, the auctioneer has the objective of minimizing transportation tardiness. An automated guided vehicle (AGV) has been selected as the case study, and numerical experiments have been carried out. The result obtained shows that the improved task assignment approach is able to produce performance competitive to a conventional task assignment. © 2013 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.