Real-Time Density-Based on-Ramp Metering Algorithm Considering Multi-Lane of Mainstream

On-ramp metering is regarded as one of most effective control methods of balancing mainstream traffic flow and releasing congestion on expressway in recent years. The current control methods suffer from several shortcomings, such as lack of consideration of on-ramp queue or lane-change behavior on multi-lane mainstream. This paper optimizes the real-time density-based on-ramp metering algorithm by taking multi-lane traffic flow character into consideration. The error function representing lane difference is defined as the objective, and real-time density considering lane change is used as the control parameter of calculating the metering rate. The micro-simulation is used for testing the performance of the multi-lane real-time density-based on-ramp metering (MRD-RM) model using the field data collected in Chengdu. The simulation result shows that MRD-RM model outperforms ALINEA and non-control in terms of reducing average queue length as well as keeping traffic flow on mainstream close to capacity.