Development of Modal Choice Model For HomeBased Work Trips-The Case of Jemo Area Residents In Addis Ababa A Thesis in Road and Transport Engineering
No Thumbnail Available
Date
2018-08
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
AAU
Abstract
Addis Ababa is one of the cities that are plagued by a continuous increase in traffic congestion
which resulted in longer travel times, increased accidents, more fuel consumption, and
deterioration in air quality. Providing and improving urban public transport service is highly
important in mitigating these challenges as well as creating sustainable transport systems. The
corridor from Jemo to Mexico is one of the most congested corridors in the city and the trips in
the morning peak hour are mostly home-based work trips. The study of modal choice behavior of
residents traveling to work will guide policy makers to improve the service quality of public
transportation systems in a way the majority of the commuters will use them.
Discrete choice models can be used to analyze and predict a decision maker’s choice of one
alternative from a set of alternatives. This study aims to contribute for the improvement of the
service quality of public transport systems in the city by determining the basic factors that affect
the preference of modes and how each factor affect the mode choice of trip makers to travel to
work place. 475 questionnaires were collected in which 4/5
th
of the data were used for calibration
and 1/5
th
for validation of the revealed and stated preference models.
The collected data of variables related to the trip maker and characteristics of the mode are fitted
in special type of multinomial logit model known as conditional logit model using SPSS. Revealed
preference model was developed for preferences among large bus, minibus (taxi) and private cars.
Whereas the stated preference model was developed for preferences among the existing modes
(large bus, minibus (taxi) and private car) and the BRT system.
The developed models for the revealed preference survey proved that the total travel time and total
travel cost of the transport modes are the primary factors that statistically significantly affect in
addition to car ownership, time of leaving home and education level interaction variable with taxi.
The stated preference models indicated that the preference towards the BRT increases with the
decrease in total travel cost, in-vehicle travel time and out of vehicle travel time. The models are
validated to be capable of predicting 80% and 92% of the preference of trip makers among the
available revealed and stated modes, respectively.
Description
Keywords
Mode choice modeling, conditional logit model, stated preference, BRT, Addis Ababa