This work studies the effect of hub congestion and time-sensitive demand on a hub-and-spoke location/allocation system. The Hub Location with Congestion and Time-sensitive Demand Problem is introduced, which combines these two main characteristics. On the one hand, hubs can be activated at several service levels, each of them characterized by a maximum capacity, expressed as the amount of flow that may circulate through the hub, which is associated with a hub transit time. On the other hand, alternative levels are available for served commodities, where each demand level is characterized by its amount of demand, unit revenue, and maximum service time. In this problem the efficiency of a hub-and-spoke system is given by the maximum net profit it may produce. To the best of our knowledge this is the first work where hub congestion and time-sensitive demand are jointly considered. Two alternative mixed-integer linear programming formulations are proposed. They include a new set of constraints, which are necessary to guarantee the consistency of the obtained solutions under the presence of the capacity-type constraints derived from hub service levels and served demand levels. The efficiency of the formulations is analyzed through a set of computational experiments. The results of the computational experiments allow to study the structure of the obtained solutions and to analyze how the different parameters affect them.
Modeling Network Congestion under Demand Uncertainty Using Wardrop Principles
Eindhoven University of Technology(埃因霍温理工大学);Istituto di Analisi dei Sistemi ed Informatica “Antonio Ruberti”, Consiglio Nazionale delle Ricerche(意大利国家研究委员会系统与信息分析研究所);Dipartimento di Ingegneria Informatica, Automatica e Gestionale, Sapienza Università di Roma(罗马大学信息与自动化与工商管理系);ESSEC Business School(欧洲高等商学院)