摘要: |
Most of the developed techniques and models for planning, routing and scheduling in the trucking industry assume 'known' static data as their input. For instance, in the Vehicle Routing Problem (VRP) the customer demands, travel costs, and travel times are known in advance. In this case, the fundamental problem is to determine the optimal route that minimizes a certain objective such as fleet size and travel distance. The built-in assumption of these approaches is that there will be small deviations on the realization of the demand and travel times from the plan so that the pre-determined routes form a basis for either the pickup or delivery schedule. In the real world, however, operations in any traffic network contain a fairly high degree of uncertainties including variable waiting and travel times due to traffic congestion, arrival of new orders and cancellation of existing orders. In a highly dynamic and stochastic environment, the pre-planned optimal routes are no longer of practical use. In this case, most of the research effort has focused on easy to control dispatching rules. The drawback with these techniques is that they do not make use of pre-planned and known information. There is a gap in the routing literature for systems that operate between the two ends of the spectrum, which is the most realistic condition for trucking operations. Our research on partial route development addresses this gap by developing a new approach within an area that has received little attention. In a simulation study, we demonstrate the benefits of the partial routing approach over the pre-planned and dispatching methodologies. |