Management of Traffic Light by Evolutionary algorithm

Abstract

Traffic congestion on main streets and intersections is caused by rapid urban expansion, increased automobile use, speeding, and a lack of driver awareness. One of the solution of reducing the congestion is the traffic lights. In this study, we optimized the traffic light timing by maximizing green light duration and minimizing red light duration, we developed an application to study timing at 16 intersections (64 traffic lights total). Each traffic light's initial red and green light values were randomly assigned. We implemented and compared two optimization methods within our application., genetic algorithm (GA) and Bio-inspired meta-heuristic, we selected the differential evolution (DE) algorithm from a suite of algorithms and compared its results to those of the genetic algorithm (GA). The DE method yielded better solutions, as shown in the tables and graphs.