Dynamic network traffic control

This study developed a dynamic traffic control formulation designated as dy namic intersection signal control optimization (DISCO). Traffic in DISCO is modeled after the cell-transmission model (CTM), which is a convergent num erical approximation to the hydrodynamic model of traffic flow. It consider s the entire fundamental diagram and captures traffic phenomena such as sho ckwaves and queue dynamics. As a dynamic approach, the formulation derives dynamic timing plans for time-variant traffic patterns. We solved DISCO bas ed on a genetic algorithm (GA) approach and applied it to a traffic black s pot in Hong Kong that is notorious for severe congestion. For performance c omparisons, we also applied TRANSYT to the same scenarios. The Results show ed that DISCO outperformed TRANSYT for all the scenarios tested especially in congested traffic. For the congested scenarios, DISCO could reduce delay by as much as 33% when compared with TRANSYT. Even for the uncongested sce narios, DISCO's delays could be smaller by as much as 23%. (C) 2001 Elsevie r Science Ltd. All rights reserved.