Predictive Traffic Management: Spatiotemporal Analysis and Clustering for Urban Road Networks

Abstract

Communication, smart transportation, and computer developments in recent years have significantly enhanced the potential for intelligent traffic convenience, and efficiency solutions. The importance of intelligent transportation systems (ITS) in alleviating traffic congestion in cities cannot be overstated. A poorly planned road network, high vehicle volumes, and critical congestion areas are the main causes of traffic congestion. The paper presents a congestion avoidance method based on estimating traffic congestion in real-time on urban road networks and predicting alternate shortest routes. Using threshold-based cluster head selection and modified K-means clustering formulation algorithms, the proposed system can estimate the degree of congestion on diverse roads and predict the shortest route. In order to optimize network design and dynamic route planning, the proposed approach demonstrates spatiotemporal regularities of traffic congestion. There is a greater degree of comprehensiveness and objectiveness in the research results than in the existing methods.