Asymptotically Optimal Control of Make-to-Stock Systems

• 成果类型: Article
• 署名作者: Nie, Jiawang; Tang, Xindong
• 署名单位: University of California System; University of California San Diego; Hong Kong Polytechnic University
• 刊物名称: MATHEMATICS OF OPERATIONS RESEARCH
• ISSN/ISSBN: 0364-765X
• 发表日期: 2024
• 页码: 1065-1090
• 关键词: multiclass queuing-networks; heavy-traffic limits; convex delay costs; INVENTORY CONTROL; STABILITY

摘要

We consider multiclass make-to-stock production/inventory systems in which the manager makes three decisions, including pricing, outsourcing, and scheduling, to maximize the long-run average profit. For a sequence of systems in the heavy-traffic regime, with linear or strictly convex holding/waiting cost functions, we propose a sequence of policies and establish its asymptotic optimality. Our proof combines the lower bound approach and a thorough steady-state analysis of the systems. We also establish general results on the existence and tightness of the stationary distributions of the state processes under a more general family of policies.