The New Jersey Institute of Technology's
Electronic Theses & Dissertations Project

Title:	Optimizing integrated service for a transit route with heterogeneous demand
Author:	Ulusoy, Yavuz Yusuf
View Online:	njit-etd2010-098 (xvii, 170 pages ~ 8.3 MB pdf)
Department:	Department of Civil and Environmental Engineering
Degree:	Doctor of Philosophy
Program:	Civil Engineering
Document Type:	Dissertation
Advisory Committee:	Chien, I-Jy Steven (Committee chair) Bladikas, Athanassios K. (Committee member) Daniel, Janice Rhoda (Committee member) Lutin, Jerome M. (Committee member) Spasovic, Lazar (Committee member)
Date:	2010-05
Keywords:	Bus service Transit service Optimization
Availability:	Unrestricted
Abstract:	The methodology developed in this dissertation attempts to optimize integrated service that minimizes the total cost, including user and supplier costs, of a transit route with heterogeneous demand. While minimizing total cost, a set of practical constraints, such as capacity, operable fleet size and frequency conservation, are considered. The research problem is presented in three scenarios, consisting of various service patterns (e.g., all-stop, short-turn and express) under heterogeneous demand. A logit-based model was used to estimate passenger transfer demand. An exhaustive search method was developed to find the optimal solutions for a simplified transit route with six stops, and a Genetic Algorithm (GA) was developed to find the optimal solution for a real-world, large scale transit route. The optimized variables include the combination of service patterns, the associated service frequencies, and stops skipped by the express service. A six-stop transit route was designed and analyzed via a proof-of-concept demonstration to ensure that the developed models are capable of finding the optimal solutions. A sensitivity analysis was conducted, which enables transit planners to quantify the impact of various model parameters (e.g., user value of time, vehicle capacity, operating cost, etc.) to the decision variables and the objective function. Finally, the developed models and solution algorithm were applied to optimize integrated service for a real world bus route in New Jersey.