Thesis Defense - Yıldız Özdemir (MSIE)

Yıldız  Özdemir - M.Sc.

Asst.Prof. Görkem Yılmaz – Advisor

Date: 15.12.2020

Time: 13:00

Location: This meeting will be held ONLINE. Please send an e-mail to gizem.bakir@ozyegin.edu.tr in order to participate in this defense.

Multi-Trip Periodic Vehicle Routing Problem with Time Windows, Split Delivery for Simultaneous Pick-up and Delivery

Thesis Committee:

Asst.Prof. Görkem Yılmaz, Özyeğin University

Assoc.Prof. Okan Örsan Özener, Özyeğin University

Assoc.Prof. İsmail Karaoğlan, Konya Technical University

Abstract:

Vehicle routing and delivery planning problems are one of the main problems that logistics systems face with.  Given pick-up and demand quantities for each customer, available working hours, shift times, the total distance between two customers, cost parameters and a variety of transportation constraints, planners search for effective delivery and, in some cases, pick-up schedule in order to satisfy customers’ needs.  De-livery planners identify vehicles to be used, the start time of service, the shortest vehicle routes within the determined time window, shifts and number of a tour of a specific vehicle in a period to increase vehicle utilization and also to meet these requirements at minimum cost.  In this paper,  in order to solve such a vehicle routing problem, we propose three mixed-integer programming models that reduce the logistic costs and number of vehicle traffic by combining shipments from different customers.  The first model includes a heterogeneous fleet of vehicles that defines service to customers with time windows, simultaneously pick-up and/or delivery requests, and shift the planning of these vehicles.  The second model tackles the first problem with split pick-up and delivery.  Besides these two models, the third detailed model deals with vehicles performing multiple tours in a given period.  These models have been implemented and tested with Solomon data set up to 25 and 50 customers to verify the models and make comparisons between their performances.  In order to solve large problems, we propose a decomposition-based heuristics which separates periods by considering time windows and total loads of each period and setting an objective that minimizes the difference between load quantity of each period, firstly.  Then, an updated mathematical model does not contain shift parameters but counts multi-trip, customer time windows, split delivery, simultaneous pick-up, and delivery.

The results show that inefficient transportation between two nodes is avoided, numbers of vehicles are decreased, the utilization rate of each vehicle is increased and vehicle traffic is reduced at each of the customer areas.

Bio: 

Yıldız Özdemir graduated from Izmir Institute of Technology with a bachelor's degree in Electronics & Communications Engineering in 2015. She has started her career in a white goods company and changed her scope to the automotive industry. Now, she is working in DAF Trucks N.V. as Embedded Software Test Engineer. In addition, she is highly interested in optimization methods used in the industry and eager to learn new approaches to find solutions close to the optimality. She is agile to adapt new topics and uses her communication skills to reach good solutions. She is a graduate student in the department of Industrial Engineering at Özyeğin University.  Her dissertation in M.Sc.  is ”Multi-Trip Periodic VRP with Time Windows, Split Delivery for Simultaneous Pick-up and Delivery”.