Research And Implement Of Distributed Multi-agents Supply Chain Negotiation Model

The current rapid economic development becomes more and more linked with computer technology's development. Obtain trade information through the network, use software to analyze and manage business operations have become a very important tool. It is also true in operation and management of the supply chain, and how to coordinate the use of computer technology and supply chain has become a common concern of business and academic circles. This paper summarizes the supply chain management and simulation modeling methods of research and application, then proposes a distributed multi agent supply chain negotiation simulation model to research the organization and management of the supply chain.Firstly, the paper introduce the theory of supply chain management and modeling, point out the usefulness of distributed supply chain model and the trend of organization and management supply chain with intelligent. At the same time, multi-Agent system is also an important research direction of the computer artificial intelligence; it can be widely applied to various fields. This paper combines multi agent system with supply chain management and modeling theory, applies multi-agent system to the supply chain to achieve its intelligent management.Secondly, with the support of management theory, agent technology, and JADE technology platform, this paper build a distributed multi-agent based supply chain simulation system in apparels which make garment industry as the core. Analyze and simulate the planning, control, procurement, production and sales behavior of the enterprises in the supply chain, then design and implementation the supply chain simulation.Thirdly, the paper supports a fairness-based bilateral negotiation algorithm, which described negotiation in three aspects, that is evaluation of issues, learning and prediction of the other acts and compromise and cooperation, to make both parties satisfied with the results of the consultations as close as possible, to achieve negotiated fairness. The algorithm provides support for procurement and sales of the enterprise in the supply chain, to ensure the balance of interests of all parties in the supply chain, so that the supply chain can be long-lasting and orderly run down. Finally, validate the algorithm based on historical information relative to the bayesian learning algorithm to enhance consultation and improvement through the implement of model and algorithm, could well meet the time and the fairness requirements. The model can reflect the self-organization of the inter-enterprise and the orderliness of the supply chain, with the enterprise agents'interaction with each other time and again; it also supports data for research of supply chain management.