Optimization and Simulation of Green Light Duration at Intersection with Particle Swarm Optimization and Cellular Automata

Usability of traffic light is considered not to be good enough because it uses fixed time type which is not able to set the light duration based on traffic density. Optimization is needed to maximize the outgoing vehicle from intersection and reduce cycle time. Traffic light optimization is applied at isolated intersection using particle swarm optimization algorithm to obtain the better green light duration. Before being implemented, Optimization experiment should be run first but it is hard to implement directly in real intersection. The best alternative is through the simulator then applied in the real world. The optimization result is to find the better green light duration of each phase. Optimized green light duration is evaluated using microscopic simulator. Cellular automata is developed to build microscopic simulation and NaSch model is applied in the movement of vehicles. Our experiments have shown that in general optimized duration performs better than fixed duration. The right combination of optimization parameters provides better optimization result The result can be used as a recommendation to improve the quality of traffic light duration at intersection.