ROBUST TUNING OF ORDER-UP-TO POLICY IN GOODS DISTRIBUTION NETWORKS WITH LEAD-TIME PERTURBATIONS

The paper addresses the issues of inventory control in distribution networks with lead-time perturbations. Mesh topology with unidirectional interconnections between network nodes is considered. The stock at the controlled nodes used to satisfy uncertain customer demand is replenished from other nodes and external sources. The goods distribution process is managed by a networked order-up-to strategy, realized in a centralized way. A heuristic that provides robustness to arbitrary shipment delay between the nodes is proposed. It allows one to determine reference stock levels using nominal system parameters and statistical data of demand, thus being efficient in practical implementation. The proposed method is examined in multiple numerical tests covering various network structures, demand patterns, and delivery perturbations.