The Application And Research Of Robust Optimization Methodology In Supply Chain Management

Robust optimization method as a powerful methodology for multi-period stochastic operations management problem has been applied in many areas. Many factors will directly affect the normal operation of supply chain system. Some comes from supplying, manufacturing, selling and other aspects of the internal operation. Some comes from earthquakes, floods, fires, industrial accidents, the international economic environment and the external uncertainty of sudden emergency incident. For these uncertain factors, robustness can ensure the sustainability of the benefits and operations. Many supply chain decisions face demand and price uncertainty. Such as Retailer Supplier Flexible Commitments Contracts, multi-item newsboy models with a budget constraint. Robust optimization methodology has been applied to supply chain management and has got perfect performances.Demand uncertainty has a significant impact on the objective function. In this paper we will show that how to change those uncertainty. The problem will be the robustness of this preliminary study. We study the problem of Retailer Supplier Flexible Commitment Contracts (RSFC). Another mode is about the classic newsboy problem. Following is a brief introduction.In this paper we establish multi-product Retailer Supplier Flexible Commitment model and Retailer Supplier Flexible Commitment model with financial commitment. And establish their affine adjustable robust counterpart respectively. We select the "box" uncertainty to respect the demand uncertainty and use the methodology which proposed by Ben-Tal and Nemirovski. Finally we will make an effective verification. In this paper we extend the classic newsboy problem. We establish the model of multi-product newsboy problem with financial constraint to retailer and supplier respectively. We consider the case of inventory centralization problem of robustness, and compare the corresponding optimal order quantity of absolute robustness, deviation robustness and relative robustness.