Real-time traffic signal settings at an isolated signal control intersection

In this paper, we present a method for traffic signal control including simultaneously signal plan design and signal timing optimization with real-time information on the network dynamics. The problem is formulated so as to find the signal plan design, the green and inter-green time for each signal group in response to recurrent traffic flow demand at an intersection. The approach used was the group-based Akcelik method for determining critical path, cycle length and green time split for each possible signal plan design. The signal plan design selection was formulated to minimize the total delay. The resulting algorithm was coded in Java and used TraSMAPI to dynamically link it to Aimsun's API, which allows an automatic change of the signal settings at an isolated signal control intersection. A theoretical three arms intersection with a time varying origin-destination demand example is tested to demonstrate the proposed method. The paper contributes to the development of an integrated design of signal plan design and signal timings. The proposed method aims to be sufficiently general for its application in different networks, with few inputs dependent on the local geometry layout. This strategy is able to react to non-schedulable events or unpredictable events without requiring human manipulation.