Application of the 4-Step Model in Forecasting Private Vehicular Traffic in an African Metropolis with Low Data. The case of Lagos Metropolis, Nigeria

Imuentinyan

Key information

Authors:

Imuentinyan (Imuentinyan Aivinhenyo)

Supervisors:

João António De Abreu e Silva (João António De Abreu e Silva)

Published in

12/14/2012

Abstract

Traffic demand analysis is a vital aspect in the planning process of a transportation system. While most developed countries usually employ disaggregate modelling systems that relies on large amount of data, developing countries that are characterized by lack of robust data still have to rely on the aggregate trip-based approach for strategic level assessment of the transportation system. This study utilised the 4-step process to model private vehicle traffic in the city of Lagos, Nigeria, using GIS-based traffic modelling software, PTV Visum, with only available data being population forecast, Origin-Destination trip matrix and traffic count. The study defined 20 traffic analysis zones comprising the 20 local administrative areas in the city. The 4-step process (Trip generation, Distribution, Modal Choice and Assignment) were carried out without the modal choice step. Trip generation modelling employed log-linear regression method by developing a log-linear relationship between the total car trip production and attraction of each zone and the population of the zone. The Trip distribution and parameter estimation employed the gravity model inbuilt in Visum. In Traffic Assignment step, the Visum-distributed trips were assigned to the network using the equilibrium assignment model. The volume-delay function used was the BPR function, with parameters set at default values of the basic BPR curve. Assignment was followed by calibration of the demand matrix to traffic count. Two matrix calibration techniques, the T-flow fuzzy and the Lohse Calibration approaches were utilized and compared.