To satisfy customers' personalized delivery demands and expand the application of hub location models, this paper introduces a hub location problem with multiple service levels and mixture uncertainties, in which two or more allowed delivery times are provided for shipment. In an expected cost-minimization context with delivery-time restrictions, the researched problem explicitly considers fixed hub number, single-assignment pattern from hubs to demand nodes, diversified transportation modes, and two types of real-life uncertainties (stochastic customer demand and fuzzy transportation cost). To help eliminate the mixture uncertainties in the proposed mathematical model, the expected value model and fuzzy chance constrained programming are utilized. By employing the existing fuzzy arithmetic and the newly-proposed operational law of γ-pessimistic value for functions of regular fuzzy numbers, two novel two-phase solution approaches are designed, respectively. The effectiveness of solution approaches and the necessity of uncertain parameters setting in the model are comprehensively illustrated by a series of numerical experiments based on Turkish data set.
Read full abstract