A Proactive Traffic Responsive Control using State Space Neural Network

M. Abbas and Y. Li (USA)


State space neural networks, traffic responsive control,timing plans, traffic state prediction


Traffic responsive plan selection (TRPS) control is an advanced operation mode on coordinated actuated traffic signals. TRPS is able to improve traffic signal operation performance by continuously changing the signal timing plans according to real-time traffic conditions. However, one of the major drawbacks of TRPS control is the fact that it is “responsive,” rather than adaptive. Therefore, the change in timing plans is always lagging the change in traffic conditions. In this paper, we use state space neural network (SSNN) to directly predict the optimal timing plan that correspond to the futuristic traffic condition in real-time. Our SSNN’s network topology is derived from the geometry of a test case urban arterial route. The design of SSNN explicitly reflects the relationships that exist in physical traffic systems and was able to closely match the optimal timing plans in futuristic states.

Important Links:

Go Back