Intelligent Urban Road Traffic Management at a Network Crossroads using Genetic Algorithm

Abstract

Given the fact that the traffic is significantly increasing, this would undoubtedly affect the delays at the traffic lights. On the face of it crossroads management should be innovatively handled and has to consider the emerging variables related to the environmental road infrastructure and the security of its users. Having this concern in mind, this study aims at developing a mathematical model that relies on acyclic nonlinear programming. According to this model we assume that the lights switching decisions could be made in real time minimizing, thereby, the traffic waiting time. We assume that although there have been attempts to provide mathematical models of traffic management, these models were mostly if not all cyclic assigning constant phase durations and cycles. The present study shows the possibility to design acyclic mathematical models that meet real time traffic management requirements.This mathematical framework proves efficient when implemented on isolated crossroads and on two adjacent crossroadswith the aim of optimizing the flow and the waiting time of traffic in a complex environment.In the present work we aim at testing the feasibility of the proposed model on a complex network. Due to the various independent variables and the non-linearity of criteria the study resorts to a metaheuristic. The final results show that the genetic algorithms successfully carried out the computations and yield satisfactory outcomes.