Modeling And Simulating For Distribution System In Supply Chain Under Uncertain Demand

With the incremental development of world economic integration and market globalization, and the quick development of information technology, supply chain management (SCM) has been the important management mode for modern enterprises and also been a hot research field for academic study and practical application since 1990. Under supply chain management mode, individual firms no longer compete as independent entities, but rather as integral parts of supply chain linkage. A good supply chain with competition force not only could provide best service and make products into markets with most reasonable cost, but also have the capability to coordinate other members in supply chain. In supply chain, distribution system is the most important linkage, which is not only the sign to measure the level of supply chain management and is the most key factors to determine the competition capability of supply chain. However, in today's complex new marketplace, the inferior position of traditional distribution system is increasingly in evidence, especially on the conflicts of efficiency, cost and controllability. Furthermore, the non-cooperation relation between vendors and buyers in traditional distribution system has greatly restricted the communication between manufacturers and consumers so as to influence the efficiency of distribution system. Thus, the cost study and the coordination problem study for distribution system will undoubtedly be important for theoretical study and practical application.Based on the review of the overseas and domestic study on supply chain management, this dissertation classify supply chain management model from two layers. The first layer is to determine the research object and modeling methods and the second layer is to determine the concrete research content in SCM modeling. Also the outside situation and research progress of the main problems in SCM are reviewed, which conclude the inventory problem in SCM, the information flow and its value in SCM, supply chain coordination and contract problems, supply chain network designing problems and supply chain performance evaluating problems. This dissertation mainly focus on the issue of modeling and simulating for distribution system under uncertain demand, whose most work are presented in the following four aspects, i.e.(1) The issue, of modeling and simulating for distribution system for single product under uncertain demand. Firstly, under continuous inventory review control policy, by synthesizing logistics costs, a comprehensive cost model for distribution system under the restriction of service level provided by the dealers is established. Hence, the cost minimization problem is solved, and the optimal ordering quantities are calculated for the aforementioned scenarios. Furthermore, simulation results of this model on the freight rate parameter and the ordering proportions parameter are proposed. Secondly, under periodic inventory review control policy, a regional distribution system is considered with one regional distribution center and multiple retailers. The cost models are respectively presented for the regional distribution center and the retailers. Based on them, an independent decision and a joint decision cost optimization model are developed for regionaldistribution systems;thereafter, solving scheme and a numerical example for the models are given.(2) The issue of modeling and simulating for distribution system for multiple products under uncertain demand. Firstly, under continuous inventory review control policy, the distribution system with two manufacturers and multiple dealers is considered. Assume that each distributor distributes the two products from the two manufacturers and the demands of the two products can be partly substituted. The total profit model for dealers and manufacturers are given. And based on them, the profit optimization model for this distribution system is established, whose goal is to maximize the total profit of the entire distribution system and whose constraint is the service level at dealers. Furthermore, simulation results of this model on the pricing parameter, the loyal proportions parameter and the variance coefficients are proposed. Secondly, under periodic inventory review control policy, beginning with the structure of distribution system, two types of multi-product distribution systems is considered. One is a traditional distribution and the other is a distribution system with regional distribution center (RDC). And we compare the performance of the two types of distribution systems via total profit of entire distribution system. The computational results show that the multi-product distribution system with RDC will attain higher profit and service level.(3) The issue of distribution system coordination in supply chain. Firstly, the price discount policy is researched under the demand with price elasticity. The problem of optimal price discount could be changed into developing a manufacturer's (leader) profit maximization model, which constraints simultaneously consider the retailer's (follower) individual rational constraint and manufacturer's (leader) incentive compatible constraint. Furthermore, simulation approach is utilized to analyze optimization problems. Secondly, the distribution system with one manufacturer and one distributor is considered. Under demand with price elasticity and demand with price elasticity and service sensitivity, the total profit functions for the distributor and the manufacturer are given. Then the decentralized coordination model for distribution system is established based on them. Furthermore, simulation results of this model on the different demand mode, demand variance, goal profit rate of the distributor, price discount and different coordination means are proposed.(4) The issue of the cost model and simulation for distribution systems based on demand forecasting. Firstly, fuzzy systems model is applied to the field of demand forecasting under irregular demand mode, that is, the retailers use fuzzy systems to forecast customer demand. Secondly, based on them, cost models are respectively presented for the center warehouse and the retailers. Finally, a simple numerical example is given to compare the difference of forecasting accuracy and total cost as well as service level in distribution system between the fuzzy system-forecasting model and statistic-forecasting model. The computational results show that using fuzzy system model to forecast demand is superior to using the statistic model in this distribution system.In the end, this dissertation is summarized and further research directions are shown.