Research On Replenishment And Dispatching Policy In VMI-based Integrated Environment

Late in the previous century, information technology has been rapidly developed, andglobal integration has been realized gradually. In this circumstance, supply chainmanagement (SCM), as a new managerial thought, has attracted world-wide researchingand industrial interest. The success of supply chain operation will depend on itsmanagerial ability to integrate and coordinate the complicated relation among supplychain members. Vendor-managed inventory (VMI) is a kind of inventory planning andintegrating technology. It implies that vendors must take on extra responsibility tomaintain inventory at retailers' side make appropriate continuous replenishment planaccording to demand forecast. VMI is not merely a brand-new inventory managementparadigm. More importantly, it can establish an integrated circumstance with mutual trustand win-win partnership. In a VMI system, a VMI supplier has the liberty of controllingthe downstream resupply decisions rather than filling orders as they are placed. Thisoffers a framework of synchronizing inventory and transportation decisions. What thestock replenishment and shipment scheduling (SRSS) problem focuses on is how to makeefficient and effective joint decisions of these two aspects to achieve global optimum.This dissertation has advanced the SRSS-related research. Based on renewal theory in Stochastic Process and precise method of formalization, SRSS problems under deterioration as well as lead time considerations are discussed respectively. For these two cases, four different analytical models are established, which provide feasible theoretical foundation for a VMI vendor to make replenishment and dispatching joint decisions. Therefore, this paper can be partitioned into two parts:The 1st part focuses on SRSS problem for deteriorating item. The discussion lies in Chapter 3. The main purpose is to find out optimal hybrid replenishment and dispatching policy under a typical Poisson demand process for this case. However, it's difficult to obtain a precise expression of the expected dispatching number per replenishment cycle. Nevertheless, a numerical range or interval, whose upper bound is only 1 larger than the lower bound, is easy to be derived. Two different methods are proposed: middle value method and correction factor method. The former utilizes the middle value of the derived interval to approximate the expected dispatching number, while the latter introduces a correction factor that varies in (0, 1] . Then based on these two methods, the corresponding analytical models are established. Furthermore, in order to optimize the correction factor based model, a jointly solving approach named OSJCA is suggested.The 2nd part focuses on SRSS with non-zero lead time. The discussions lie in Chapters 4 and 5. Two different cases are taken into account respectively: (1) Replenishment lead time is a fully exogenous random variable; (2) Replenishment lead time is random but controllable. It's reducible by adding extra crashing cost. Under these assumptions, two time-based replenishment and dispatching models are built up.The innovative works of this dissertation are summarized as follows.(1) Innovation on research method: each research of this paper employs formalization method to establish precise analytical model and further optimize it. Besides typical numerical examples are provided to illuminate the application of our models, simulation method is also adopted to validate the models. The simulation programs are developed directly from a certain policy implementing scheme and irrelative to the modeling process. Make comparison between results though model-based optimization and computer-aided simulation. If they are highly coincident with each other, the validity of our models can be confirmed;(2) Innovation on theoretical contribution: This dissertation has considered SRSS with deteriorating item as well as lead time, which are unexplored but fairly worth discussing. In a sense, the paper advances the SRSS-related research, and therefore has both theoretical value and practical significance.