A pro-active real-time control approach for dynamic vehicle routing problems dealing with the delivery of urgent goods

This paper proposes a new pro-active real-time control approach for dynamic vehicle routing problems in which the urgent delivery of goods is of utmost importance. Without assuming any distribution, stochastic knowledge about future requests is generated using past request information. The generated knowledge is integrated into the transportation process, which is controlled by a Tabu Search algorithm, in order to actively guide vehicles to request-likely areas before requests arrive there. By analyzing the results attained for various test settings, we identify structural diversity as a crucial criterion for classifying the quality of stochastic knowledge attainable from past request information. This criterion provides a promising starting point for assessing the quality of past request information in order to efficiently use the derived stochastic knowledge in real-time control approaches. We prove the efficiency of our approach by a direct comparison with a deterministic approach on test scenarios with varying structural diversity. Thanks to the proposed classification of structural diversity, differences in results obtained among the tested scenarios become explainable.