Font Size: a A A

Research On Inventory Control Policies For Nonstationary Stochastic Demand

Posted on:2008-08-14Degree:DoctorType:Dissertation
Country:ChinaCandidate:J YangFull Text:PDF
GTID:1119360212498646Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
Nonstationary inventory problems occur in a large number of industrial, distribution, and service contexts. Influenced by technology development, environment changes, promotional mechanisms and management styles, the demand of many products are stochastic and nonstationary, the distribution of which may change rapidly. For nonstationary demand, traditional stationary policies will lead to stockouts with unsatisfied demands or to larger inventories being carried than planned. Yet besides a few publications on nonstationary inventory control methods, not much work has been down in this area. Since the demand distribution for different period is different, it is a multistage stochastic programming problem and is difficult to solve directly. For general inventory control problems with nonstationary demand, there lack of effective inventory control method.The background of the thesis comes from spare part inventory control problem in process industries. We carried out research on nonstationary stochastic demand, try to develop new inventory control policies which can be applied to general problems without any limitation on demand distributions, length of lead time, backlog etc.This paper involves six chapters, and the main contents of this thesis are summarized as follows:The first chapter introduces the background of this thesis, and reviews the inventory control methods for nonstationary stochastic demand. Then the research methods and main contents are given.In chapter two, we first summarize .the characteristics of the spare parts inventory control problems for process industry. Then an integrated forecasting method for nonstationary demand is introduced, which provides a mechanism to integrate the demand autocorrelated process and the relationship between explanatory variables and the demand of spare parts during forecasting the demand distributions. Then illustrative examples of spare parts are used to illustrate the demand nonstationary process.In chapter three, for problems with constant nonzero lead time, time-varied unit purchasing cost and backlogs, a static-dynamic uncertainty strategy is provided. The first stage of the strategy determines the replenishment periods and order-up-to-levels. When the demand realization is known, the second stage of this strategy then determines the order quantity based on order-up-to-level and inventory status. This strategy is superior to the static uncertainty strategy which determines the replenishment periods and order quantities at the beginning of the planning horizon. Compared with current static-dynamic uncertainty strategy, our method has lower costs and is more general.In chapter four, since the static-dynamic uncertainty strategy is not optimal after demand realizes, a new inventory control policy based on rolling horizon approach is suggested. By solving a static-dynamic uncertainty model, the policy first makes decisions on replenishment periods and order-up-to-levels over the planning horizon, but implements only the decisions of the first period. It then uses the rolling horizon approach in the next period when the inventory status is revised and the multi-period inventory control problem is updated. Using simulation datasets, the effectiveness of the proposed policy is verified. By applying this policy to real spare parts inventory control problems, we show the application results.Chapter five develops a new (s_t,S_t) optimal inventory control policy fornonstationary demand. The control parameters in each period follows the traditional (s, S) policy, but parameters for different periods are different. As demand realizes period by period, the inventory control parameters are still optimal for the rest of the planning horizon. The solution algorithm has no limitation on demand distributions, lead time, unit purchasing costs and backlogs, and can be applied to general inventory control problems with nonstationary demand.Chapter six summarizes all the work of this thesis, and gives some useful suggestions for future research. Main contributions of this thesis are briefly summarized as follows:(1) Since the static uncertainty strategy can not adjust the order quantity based on demand realizations, a new general static-dynamic uncertainty strategy is developed, and the solution algorithm is also provided.(2) Because the control parameters of the static-dynamic uncertainty strategy are not optimal after demand realizes, a new inventory control policy based on rolling horizon approach is suggested.(3) Since current researchers have not provided the optimal policy for inventory control problem with nonstationary demand, we provide a new (s_t,S_t) optimalinventory control policy.
Keywords/Search Tags:inventory, nonstationary demand, static uncertainty strategy, static-dynamic uncertainty strategy, dynamic programming
PDF Full Text Request
Related items