Traffic Pattern Modeling Using Clustering Algorithm

Abstract

Traffic congestion is one of the most alarming problems of Dhaka, the capital of Bangladesh. In order to control traffic intensity of this densely populated city we need to find traffic pattern and the factors that influence traffic congestion. In this paper, we have analyzed traffic intensity pattern for different cases with respect to time. For macroscopic modeling, we have used 13 DPZ zones of Dhaka proposed by RAJUK and have analyzed traffic pattern for each zone. For finding similarities among zones, clustering method has been used using the land use information of each zone. For finding the factors that influence traffic congestion we have used regression analysis considering both land use pattern and social infrastructures of zones. Bus routes and road intersections have been also analyzed for finding the reasons behind traffic congestion. Thus the whole analysis will help to take proper steps in urbanization planning in order to minimize traffic congestion of this beautiful Dhaka city.