Abstract:
Traffic Control System plays a key role in overall safety of traffic and pedestrians, but congestion problem on road networks led to the increases slower speeds, longer trip times and increased vehicular queuing. These prompt idea of Adaptive Traffic Light Controller based on Neuro-fuzzy techniques which were used in controlling traffic so as to design and simulate a hypothetical Cross(+)-junction intersection traffic scenario to achieve congestion free and smooth traffic condition with special consideration of traffic length, presence and absent of emergency on traffic using priority. Sensors/detectors for vehicle detection were installed and configured on stop-lines behind each traffic light, upstream-lines behind the first sensor at distance S, and corner-lines for right-turning. Detectors will count the number of vehicles through the upstream-line, stop-line and corner-line within a given time. To detect left-turning vehicles, ultrasonic detectors were placed on side of left-turning sides. These sensors were used to detect the vehicle appearance and count the number of vehicles driven into left-turning sides. The system considered five input variables one output variable, All the input, were obtained by detectors/sensors lock, which process and converts the crisp value of inputs into the fuzzy value and they were mapped with three linguistic variables (Low, Average, High), all the variables were fuzzified using Gaussian membership function. The objectives of the work were to design an adaptive neuro-fuzzy based model to control vehicle movement in a double lane traffic intersections, which were simulated using MATLAB integrated development environment and performance evaluation of the system shows highly perfect in achieving the desire objectives compared to conventional traffic system.
