Research Of The MAS Application In Partner Selection For Dynamic Supply Chain

Dynamic supply chain's competitive advantage stems from the complementarity and integration ability of the various components.Partner selection is the key of a dynamic supply chain's formation and operation.This thesis proposes a model of dynamic supply chain's partner selection based on target decomposition and optimization.This model is based on Multi-Agent based dynamic supply chain system, combined the multi-objective selection model with the tender mode and add in the concept of multi-agent.This model decomposes the overall objective of dynamic supply chain into a set of sub-targets at first.Through describes and decompose target system of dynamic supply chain,express the problem of dynamic supply chain's partner selection as distributed constraint satisfaction and optimization.The principal agent releases information on business needs to federal agent.Then,the principal agent makes a conference with the federal agent which want to join in tender,and evaluate all the enterprises which are suitable for selection by multi-attributive utility theory.Solve multi-attributive utility function by combine agent and traditional integral linear programming which is also often called the "ch and bound algorithm".Lead to the specific a method of solving based on agent model.When all the candidates do not meet the requirements,the system will change goals system and constraints of dynamic supply chain,and choose the partners again. Finally,for example,take the TV design partner selection,describes the specific application for the partner selection model.The method of goal description and decomposition of dynamic supply chain provides popular language for the negotiation among the different enterprises.This thesis formalizes the problem with distributed constraint satisfaction and optimization, which can provide language region and decision-making for the negotiation among the different enterprises.The goal of this thesis is that use distributed constraint satisfaction and optimization solutions to provide a new thought for the dynamic supply chain partner selection problem.