OPTIMIZATION OF DYNAMIC RAMP METERING CONTROL WITH SIMULTANEOUS PERTURBATION STOCHASTIC APPROXIMATION

Congestion is always a serious problem in many transportation commuting corridors. The need to implement effective control in the corridors has long been recognized by transportation professionals. Ramp metering has been discussed as a viable control strategy on freeways in many places of the US, which may alleviate or at least reduce the delay. In this study, a dynamic ramp metering control model is developed to maximize the total throughput with simultaneous perturbation stochastic approximation (SPSA), subject to the constraints of link densities, capacities, and feasible range of metering rates. A case study for a segment of US I-80 in New Jersey is conducted, while the impact of the proposed ramp metering control on traffic behavior is simulated with CORSIM.