Optimisation of traffic signal plan for isolated intersection using genetic algorithm

Traffic signals at a junction play an essential role in granting the right to proceed to different collections of non-conflicting vehicular movements according to different periods. Well-timed traffic signals are critical in providing smooth traffic flow. In this study, the ability of the Genetic Algorithm (GA) in the optimization of the traffic signal plan is investigated. The GA is implemented using Python Programming Language. Also, SUMO, a traffic simulation package is used to simulate the traffic scenario of the solution to calculate the mean time loss. The investigation on the efficiency of GA suggests that the selection fraction of 0.8 and the mutation rate of 0.75 are more efficient. Using these figures, GA is employed in finding the traffic signal plan for low, moderate and high levels of traffic demand. The results are compared with the algorithm based on Webster’s formula. GA outperforms the traditional method based on Webster’s formula. Also, GA performs the best at high level of traffic demand with 38.46% lower in mean time loss, followed by 37.41% lower and 3.48% lower at low and moderate traffic levels respectively.