Comparisons of agent approaches with centralized alternatives based on logistical scenarios

摘要

The research work described in this paper pusues two goals. Firstly, it deals with how Distributed Artificial Intelligence (DAI) techniques are used to solve “real-world” logistical problems. Secondly, it compares decentralized solutions (agent systems) with alternative centralized approaches. The results of most of the cases indicate that concerning the main objectives the agent systems can out-perform alternative approaches. A general statement in favour of the DAI-approaches is not possible, though.