Manufacturing Industry Multi-stage Supply Chain Inventory Decision Model

Manufacturing Industry is the base of all industries. Since the 21st century, with the continuous developing of science and technology as well as people's constant pursuit of the living quality, the enterprises in all fields accelerate the upgrading of their products more and more frequently in order to meet the increasingly harsh demands of customers to ensure that their companies can survive from the changeable social trends. Under this trend, the link among enterprises are no longer as simple as a sense of competition and cooperation, but become more and more focused on the collaboration and the wisdom combat in the new environment, on the strengthening of the effective management of supply chain which will be implemented as a strategic objective. Its importance embodying in the manufacturing industry lies in the inventory control, for the market demand is capricious and the communion of information between different nodes on the supply chain is not complete or in time, which always leads to deviation of demand forecast. The most obviously phenomenon is the "bullwhip effect." Since the beginning of this century, the research on supply chain management has become a hotspot of various scholars which give birth to numerous research results. In practice, the movement of supply chain is always presented in multi-level forms. For this character, the paper designs a multilevel supply chain model as the study objective to do research into coordination problems of multilevel supply chain.This paper analyzes the research result of supply chain inventory management and supply chain contract, including the methods of supply chain coordination, methods of supply chain optimization, and discussion of supply chain contract coordination theory. Then it assumes a multilevel supply chain transaction model under a stochastic market demand, describing the trading models and setting up a mathematical model accordingly based on the two stage supply chain, namely, decentralized and centralized control. By the mathematical methods, the overall yield and specific decision-making are concluded and the differences of those two are compared. The next part introduces the revenue sharing contract. Through the assumption of revenue sharing factors and the design of transaction methods, it calculates both the yields of a node and the overall yields in the supply chain after the application of revenue sharing contract in the distributed multi-stage supply chain. Compared through the use of mathematical methods and numerical experiments, it can be obviously observed that by using of the revenue sharing contract, the yield of every node, compared with the non-contract model, has gained a Pareto improving in the decentralized multi-stage supply chain and the overall revenue also reaches the ideal state of the overall yields of multi-stage supply chain.At last, the paper summarizes the research results and expresses the expectations on the research which can still be carried on in future.