Optimal management adaptive of two crossroads by genetic algorithm

Abstract

This paper addresses the issue of managing the traffic flow at traffic-light controlled junctions adopting a mixed, nonlinear programming model. More precisely, we adopt an adaptive, acyclic control system based on the genetic algorithm (GA), a system that allows to accurately decide in real time on the light switching pattern that minimizes the traffic waiting time. Equally importantly, the inherent crossroads constraints, namely, the security of their users and the structure thereof, also referred to as the antagonism principle, are considered in the proceeding of this method. The implementation of this model on two crossroads significantly decreases the waiting time and regulates the traffic flow. Therefore, the adaptive control system based on the GA proves efficient compared to the nonadaptive system.