A genetic algorithm for the p-hub center problem with stochastic service level constraints


Thesis Type: Postgraduate

Institution Of The Thesis: Middle East Technical University, Faculty of Engineering, Department of Industrial Engineering, Turkey

Approval Date: 2010

Thesis Language: English

Student: Şükran Eraslan Demirci

Supervisor: FATMA SEDEF MERAL

Abstract:

ABSTRACT A GENETIC ALGORITHM FOR p-HUB CENTER PROBLEM WITH STOCHASTIC SERVICE LEVEL CONSTRAINTS Eraslan Demirci, Şükran M.Sc., Department of Industrial Engineering Supervisor: Asst. Prof. Dr. Sedef Meral December 2010, 170 pages The emphasis on minimizing the costs and travel times in a network of origins and destinations has led the researchers to widely study the hub location problems in the area of location theory in which locating the hub facilities and designing the hub networks are the issues. The p-hub center problem considering these issues is the subject of this study. p-hub center problem with stochastic service level constraints and a limitation on the travel times between the nodes and hubs is addressed, which is an uncapacitated, single allocation problem with a complete hub network. Both a mathematical model and a genetic algorithm are proposed for the problem. We discuss the general framework of the genetic algorithm as well as the problem-specific components of algorithm. The computational studies of the proposed algorithm are realized on a number of problem instances from Civil Aeronautics Board (CAB) data set and Turkish network data set. The computational results indicate that the proposed genetic algorithm gives satisfactory results when compared with the optimum solutions and solutions obtained with other heuristic methods.