Multi-depot vehicle routing problem with drones in emergency logistics

Abstract

In order to expand the application of truck-drone cooperative delivery in emergency logistics, this paper investigates the multi-depot vehicle routing problem with drones for emergency logistics scheduling. A mixed-integer programming model with the objective of minimizing the maximum expected rescue time from urban distribution centres to temporary distribution stations is proposed. To address this problem, a new two-stage hybrid heuristic algorithm is proposed. In the first stage, we employ an improved K-means clustering to cluster the temporary distribution stations. In the second stage, a combination of Tabu search (TS), enhanced genetic algorithm (GA), and simulated annealing (SA) is applied to optimize the routes of both trucks and drones. The numerical study involves the validation of the model using real-world cases. Through comparison experiments with the Gurobi solver, we demonstrate the significant advantages of the proposed heuristic algorithm in terms of solution quality and efficiency.