New fuzzy logic system for controlling multiple traffic intersections with dynamic phase selection and pedestrian crossing signal

Author

Muhammad Aria Rajasa Pohan, M.T

Abstrak

The purpose of this paper was to demonstrate a distributed fuzzy inference system for controlling traffic lights signal from multiple intersections by selecting dynamic phase and controlling signals for pedestrian crossings. The controller adaptively adjusted the phase sequences and phase lengths to optimize the vehicle's movement to go straight or turn right and to arranged opportunities for pedestrians to cross the street. At each intersection, there was a fuzzy system that can communicate with other fuzzy systems in the connecting intersection. Each fuzzy system consisted of three main modules: The Green phase, red phase, and decision modules. A traffic intersection simulator was developed to test the performance of the fuzzy controller. The proposed method was compared to two other traffic light controllers: Preset-Cycle-Times (PCT) and Vehicle-Actuated (VA) controllers. The measured performance index is the average delay time of vehicles and pedestrians. The simulation results revealed better performance of the proposed controller compared to the PCT and VA controllers. However, in heavy traffic conditions, only a few improvements have been made. This finding indicates that the fuzzy controller provides better traffic control performance than the VA and PCT controller that allows vehicles and pedestrians to have shorter waiting time at the intersection that applies this fuzzy controller