Abstract:Aiming at the problem of regional traffic congestion, a gating control method is proposed based on macroscopic fundamental diagram (MFD) and model prediction control (MPC). First, the congestion area is divided into several control sub-regions through clustering method. Then, the optimal control model based on MFD and MPC is established for those sub-regions, in which the objective is to minimize the total travel time of the road network, and the genetic algorithm is adopted to solve this model to obtain the proportion of vehicles under control. In a case study, the proposed methodology is applied to the central urban network of Kunshan City, China. The results indicate that the total travel time in the network decreases significantly after implementing our proposed method, which means that the algorithm is effective in improving network efficiency of the congestion area.

Gating Control for Congested Network with Multiple Sub-Regions and Dynamic Boundaries