The accelerated development of global network economy has generated newchallenges to today's supply chain management. Geographically distant supply chainmembers can be connected via information technologies, so that the distributed natureof supply chains becomes more and more obvious. This distributed nature increasesthe dynamics and heterogeneity of supply chain systems, therefore, negotiationsbetween supply chain members turn out to be quite significant. In accordance with thedistributed nature of supply chains, Multi-Agent System methodology is introduced inthis paper to build the Multi-Agent framework of distributed supply chains andprovide an intelligent negotiation policy.Firstly, this paper analyses the characteristics of distributed supply chains andMulti-Agent System as well as their compatibility in order to clarify the cut-in pointof the problem. Then, according to the Multi-Agent modeling methodology andSupply Chain Operations Reference (SCOR), an Ontology based Multi-Agent supplychain framework is developed with the illustration of how to express supply chainOntology, define Agent roles and construct the whole framework. Furthermore,adopting the Agent development platform JADE, the implement platform structureand software framework of Multi-Agent supply chain systems are discussed.Moreover, based on the Multi-Agent framework of distributed supply chains, thefollowing part of this paper primarily focuses on the coordination and negotiationmechanisms between Agents. The concept of multi-contract negotiation which isdeveloped from the Contract Net Protocol (CNP) is proposed with the description oftask decomposing mechanism-material requirement specification. Afterwards, theCNP based Multi-Agent negotiation model is built with the detailed illustration of theAgent structure, utility function and fuzzy inference based negotiation strategy andthe information transfer procedure. Finally, an assumed case study is given to realizethe whole course of the negotiation model. |