A decomposition-based iterative optimization algorithm for traveling salesman problem with drone


Eş Yürek E., Özmutlu H. C.

TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, cilt.91, ss.249-262, 2018 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 91
  • Basım Tarihi: 2018
  • Doi Numarası: 10.1016/j.trc.2018.04.009
  • Dergi Adı: TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Sayfa Sayıları: ss.249-262
  • Anahtar Kelimeler: TSP, Drone, Drone delivery, Decomposition, Iterative algorithm, MIP, DELIVERY
  • Bursa Uludağ Üniversitesi Adresli: Evet

Özet

This study investigates a new delivery problem that has emerged after the attempts of several e-commerce and logistics firms to deploy drones in their operations to increase efficiency and reduce delivery times. In this problem, a delivery truck that carries a drone on its roof serves customers in coordination with a drone. The drone is considered to complement the truck due to its cost-efficiency and ability to access difficult terrains and to travel without exposure to congestion. This study presents an iterative algorithm that is based on a decomposition approach to minimize delivery completion time. In the first stage of the proposed methodology, the truck route and the customers assigned to the drone are determined. In the second stage, a mixed integer linear programming model is solved to optimize the drone route by fixing the routing and the assignment decisions that are made in the first stage. Beginning with the shortest truck route, the assignment and the routing decisions are iteratively improved. The solution times of our algorithm are compared with the solution times of the state-of-the-art formulations that are solved by CPLEX. The results demonstrate that our algorithm yields shorter solution times for the instances that we generated with the specified parameters. An optimization-based heuristic algorithm, which obtains solutions for medium-sized instances, is developed by reducing the feasible search area.