Multi-Supply Order Optimization and Allocation with Markov Decision Making

Articles in Press

Document Type : Research/ Original/ Regular Article

Authors

1 Student

2 Department of Industrial Engineering, Yazd University, P.O. BOX 89195-741, Pejoohesh Street, Safa-ieh, Yazd, Iran

3 Technical and Engineering Campus Building 1, Room 223. Yazd University. Yazd, Iran

4 Department of Industrial Engineering, Yazd University,, Pejoohesh Street, Safa-ieh, Yazd, Iran

Abstract

In this paper, an inventory management model based on a Markov decision framework for finite time horizon and discrete periods is developed. The main objective of this model is to reduce the overall inventory management costs by determining the optimal ordering quantities and allocating them to suppliers. Ordering costs are modeled as random variables and holding costs as linear functions. Using the dynamic backward programming method, optimal policies are designed that minimize the costs associated with ordering and holding. To evaluate the model, a case study is conducted in a manufacturing company that receives its polypropylene raw material from two suppliers. The results show that optimal allocation of orders can significantly reduce the overall supply chain costs by the end of the planning period. This cost reduction is due to optimization in ordering quantities and appropriate selection of suppliers based on the policies provided by the model. The proposed model, considering uncertainty in ordering costs, has the potential to be applied in real environments. This model provides effective tools for improving decision-making and reducing costs in the supply chain and can be used as a practical approach for manufacturing companies with multiple suppliers.

Keywords

Main Subjects

Supply Chain Management

Articles in Press, Accepted Manuscript
Available Online from 30 August 2025

Files

History

  • Receive Date: 05 January 2025
  • Revise Date: 21 May 2025
  • Accept Date: 30 August 2025
  • Publish Date: 30 August 2025

Share

How to cite

Statistics