A Heuristic Adaptive Traffic Control Algorithm for Signalized Intersections

It is observed that most of the adaptive traffic control algorithms popular in countries with homogeneous traffic perform sub-optimally in heterogeneous non-lane-following traffic conditions owing to the inaccuracies in predicting demand, sub-optimal solutions, and time-consuming computation. The determination of demand based on discharge headway has also contributed to inaccuracies in demand estimation. Hence, this study proposes a heuristic adaptive traffic control algorithm using demand estimation based on queue length, which is expected to perform better in varying traffic composition, roadway geometry, and the presence of road-side friction. VISSIM 6.0 was used to evaluate the algorithm, with an existing vehicle-actuated algorithm used as a benchmark. In order to evaluate the developed algorithm, two test cases are presented with average stopped delay, average control delay, average queue length, and peak hour discharge. Results show significant improvements to the adaptive traffic control algorithm with respect to the parameters considered for evaluation.