J.-S. Yang (USA)
Signal Timing Control; Optimization; Stochastic Approximation; Neural Networks; Measure of-Effectiveness.
This paper presents a pilot study of the development and evaluation of a traffic signal timing control for a small scale road network in downtown Duluth, Minnesota. The study mainly focuses on the alleviation of a sudden traffic flow surge following special events (e.g., concerts, conventions, hockey games, etc.) held at the city's Entertainment and Convention Center. A practical approach for signal timing control that eliminates the need of using traffic flow model is used to optimize the intersection traffic light split times. Our approach is based on neural networks (NNs) with the weight estimation via the Simultaneous Perturbation Stochastic Approximation (SPSA) algorithm. Based on the traffic data collected and the measure-of-effectiveness (MOE) criterion used, the NN weights are adjusted by use of the SPSA algorithm that minimizes the MOE criterion. The performance evaluation using the existing signal timing plan and the one generated by the SPSA algorithm is also compared. The results show the potential to apply this method to control the signal timing over a large-scale road network.
Important Links:
Go Back
