Study On Optimization Design Model & Complexity Of Supply Chain

In the highly competitive market,traditional manufacturing industry faces the challenge of reducing product development time,improving quality,reducing cost and lead-time for production. Many successful industrial applications demonstrate that their successes in overcoming these challenges depend to a great extent on their ability to integrate individual entities into a tightly coupled supply chain.Supply chain extends a single organization's boundaries to a network composed by suppliers,manufacturers,distributors and customers whereby material flows from the suppliers through manufacturers and distributors to the customers. It has been introduced to address the integration of organizational functions ranging from the ordering and receipt of raw materials throughout the manufacturing processes,to the distribution and delivery of products to the customer.This dissertation mainly deals with the central research issues of improving the design and measuring the complexity of supply chain,which is helpful for manufacturing industry to design & implement supply chain and important for ASP & ISP of supply chain. The major contents and contributions are described as follows:Based on investigating literatures and applications of supply chain,this dissertation proposes the definition of supply chain. The meaning & extension and goals of supply chain management are discussed. Some related techniques and applications on supply chain are described. Through studying the order-decomposing process simulation,this dissertation defines the two agent classes:coordination & resource agent,maps the order-decomposing process into a real-time control process formed by a group of agents,proposes Finite State Machine as a kind of ideal modeling approach used to analyze the real-time control process,describes the conversation rules of agent conversation classes,identifies agents & conversation rules and develops the conversation models by using the framework provided by JAFMAS,JDK1.1 integrated development environment & Java.According to the structural characteristics,this dissertation transforms theoptimization of series supply chain design into multi-stage decision-making problem and develops an optimization model for it. Algorithms for the model are proposed. A network based approach,called Extended Petri-Nets that represents the design of the supply chain as an abstracted network with "AND" and "OR" nodes,is presented. Based on their substantial disadvantages of Linear Programming,Integer Programming,Nonlinear Programming and Genetic Algorithm in dealing with a typical supply chain,a Constraint Based Genetic Algorithm as part of the Extended Petri-Nets approach is introduced to search for improvements in the design that satisfies the constraints imposed on the system. A prototype testing system is constructed,which is followed by an example together with a description of the small automobile supply chain system implementation.Supply chain is a complex nonlinear system composed by suppliers,manufacturers,distributors and customers. The complex structure makes its states very uncertain & unpredictable. The measuring problem of complexity has not been studied in supply chain literatures to date. This dissertation brings forward the definition of supply chain complexity and constructs a supply chain communication model. This dissertation introduces discrete sender & united sender concept from information theory into the research and transforms the complexity measuring issue into the uncertainty measuring category of dual-role united senders utilizing the supply chain communication model in order to measuring complexity in the supply chain.This dissertation presents an example together with a specific implementation of a dry-clean machine manufacturer & its parts supplier and obtains some significant and valuable conclusions through demand analysis,concept design,logical design,data collection & analytical calculation phases.