File(s) under embargo

Reason: I haven't published my work yet. I need at least two years to be able to publish my manuscripts.

1

year(s)

6

month(s)

30

day(s)

until file(s) become available

PICKUP AND DELIVERY PROBLEM WITH TRANSFERS AND ELECTRIC VEHICLES

thesis
posted on 26.04.2022, 23:49 authored by Cansu Agrali OnerCansu Agrali Oner

Online retail sales and grocery/food orders have been breaking records every year. As a result, third-party delivery companies have found an opportunity to get their share from the growing transportation network. Electric vehicles (EVs) are becoming a preferable choice for such large delivery systems due to their environmental benefits. However, EVs have limited-service ranges; therefore, intra-route facilities are needed for EVs to stay operational. These facilities offer charging stations for EVs and storage areas for requests, e.g., food and packages. In this dissertation, we propose a novel Pickup and Delivery Problem (PDP) with EVs and transfers. There are requests to be picked up and delivered. EVs leave their origin depot, serve requests, and return to their destination depot. Unlike the generic PDP, intra-route facilities allow EVs to exchange requests. Thus, a request can be transported by more than one vehicle. In this dissertation, three new problems are introduced, and the following research questions are investigated: 1) "How valuable is to include intra-route facilities and allow transfers in a pickup and delivery network with EVs?", 2) "What is the cost of locating intra-route facilities randomly rather than finding the best locations while creating the routes for EVs?", and 3) "How much can drones improve the delivery speed in a pickup and delivery network with EVs and transfers?". A Mixed-integer Linear Programming (MILP) model and a Simulated Annealing (SA) algorithm are developed and compared with each other to answer the first question. For the second question, a MILP model is formulated; however, due to unreasonable computational runtimes, a SA algorithm and an Adaptive Large Neighborhood Search (ALNS) algorithm are proposed. Finally, a MILP model is developed for the hybrid-fleet problem. The overall results highlight that intra-route facilities shorten the total traveled distance in the PDP network by allowing exchanges and recharging.

History

Degree Type

Doctor of Philosophy

Department

Industrial Engineering

Campus location

West Lafayette

Advisor/Supervisor/Committee Chair

Seokcheon Lee

Additional Committee Member 2

Hua Cai

Additional Committee Member 3

Mario Ventresca

Additional Committee Member 4

Byung-Cheol Min