• 대한안전경영과학회지
• /
• 제16권4호
• /
• pp.323-330
• /
• 2014

Various inventory control theories have tried to modelling and analyzing supply chains by using quantitative methods and characterization of optimal control policies. However, despite of various efforts in this research filed, the existing models cannot afford to be applied to the realistic problems. The most unrealistic assumption for these models is customer demand. Most of previous researches assume that the customer demand is stationary with a known distribution, whereas, in reality, the customer demand is not known a priori and changes over time. In this paper, we propose a reinforcement learning based adaptive echelon base-stock inventory control policy for a multi-stage, serial supply chain with non-stationary customer demand under the service level constraint. Using various simulation experiments, we prove that the proposed inventory control policy can meet the target service level quite well under various experimental environments.

• 한국경영과학회지
• /
• 제31권3호
• /
• pp.27-40
• /
• 2006

Today's customer demands in supply chains tend to change quickly, variously even in a short time Interval. The uncertainties of customer demands make it difficult for supply chains to achieve efficient inventory replenishment, resulting in loosing sales opportunity or keeping excessive chain wide inventories. Un this paper, we propose an adaptive vendor managed inventory (VMI) model for a two-echelon supply chain with non-stationary customer demands using the action-reward learning method. The Purpose of this model is to decrease the inventory cost adaptively. The control Parameter, a compensation factor, is designed to adaptively change as customer demand pattern changes. A simulation-based experiment was performed to compare the performance of the adaptive VMI model.

• 대한산업공학회지
• /
• 제37권3호
• /
• pp.216-228
• /
• 2011

A logistics system involving a supplier who produces and delivers a single product and a buyer who receives and sells the product to the final customers is analyzed. In this system, the supplier and the buyer establish a contract which specifies that the supplier will deliver necessary amount of the product to raise inventory up to a specified position at the beginning of each period. A new periodic order-up-to-level inventory control policy specifically designed for nonstationary end customer's demand is proposed for the system. Simulations are used to test the efficiency of the proposed policy. An analysis of the test results reveals that the proposed policy performs much better than does the existing order-up-to-level policy, especially when the demand is nonstationary.

• 산업경영시스템학회지
• /
• 제17권31호
• /
• pp.59-72
• /
• 1994

This paper deals with ordering policies of consumable goods which have large demand rates in a multi-level distribution system. The system we are concerned consists of one Central Distribution Center(CDC) and N non-identical Regional Distribution Centers(RDCs) which have different demand rates, minimum fillrates, leadtimes, etc. The customer demand on the RDC is stationary poisson and the RDCs demand on the CDC is superposition of Q-stage Erlang distributions. We approximate the RDCs and CDC demand distribution to nomal in order to enhance the efficiency of algorithm. The relevant costs include a fixed ordering cost and inventory holding cost, and backorder cost. The objective is to find a continuous-review ordering policy that minimizes the expected average costs under constraints of minimum fill rates of RDCs and maximum allowable mean delay of CDC. We developed an algorithm for determining the optimal ordering policies of the CDC and the RDCs. We verified and compared the performance of the algorithm through the simulation using the algorithm result as the input parameters.