Modelling lane flow distribution on multi-lane highways: Insights from built-up and non-built-up areas

Abstract

This study investigated lane flow distribution on selected sections on the Kumasi – Accra multilane highway, with a specific focus on a section located in a built-up area, characterized by side frictions and driveways, and another in a non-built up area, characterized by flat terrain with no access or side friction. The study sought to (1) explore the lane flow distribution patterns under varying traffic conditions and roadway environmental types (urban and non-urban), and (2) investigate the combined effect of roadway and traffic conditions on lane flow distribution. Traffic volume data was extracted from video recordings made at the two study sites for two days. At the same time, vehicular speeds were measured with the use of a radar speed gun. The data were explored descriptively, after which multiple linear regression was employed to model the lane flow distribution in the median lane. The results indicated pronounced disparities in lane utilization between the two sites with different vehicle categories exhibiting distinct lane preferences. The model suggested that roadway environment type, proportion of motorcycles in the traffic stream, and total link flow significantly determine the proportion of traffic flow that travel in the median lane whereas the proportion of trotros and the average link speed have marginal influence. The findings emphasize the importance of considering local contextual factors and driver behaviors in modelling lane flow distribution, particularly in developing countries with heterogeneous roadway and traffic conditions.