Destination-Aware Coordinated Ramp Metering for Preventing Off-Ramp Queue Spillover and Mainstream Congestion
Cheng Zhang,Wei Ma,Jing Zhao,Chengyuan Ma,Yuelong Su,Xiaoguang Yang
DOI: https://doi.org/10.1109/mits.2023.3323029
IF: 5.293
2024-01-01
IEEE Intelligent Transportation Systems Magazine
Abstract:Freeway congestion usually occurs in on-ramp areas; however, it also occurs around off-ramp areas and can even spill back onto the main line, which is common on many congested urban freeways during peak hours. Nevertheless, traditional ramp metering strategies, one of the most successful approaches to reducing freeway congestion, have not considered the off-ramp traffic dynamics. Therefore, this study proposes a destination-aware coordinated ramp metering model for preventing off-ramp queue spillover and mainstream congestion. The model is established based on a model predictive control, which contains a traffic evolution module and an optimization module. The traffic evolution module is described in a destination-aware manner with explicit consideration of the impact on mainstream traffic due to off-ramp queue spillover. The optimization module especially takes into account the off-ramp queue constraints so as to prevent the spillover. A modified feasible direction algorithm is designed for solving the optimization problem. A comprehensive case study is conducted on a congested urban freeway network in Shanghai, China, during peak hours. A sensitivity analysis demonstrates its capabilities and robustness in preventing off-ramp queue spillover and freeway congestion. The results show that the proposed strategy reduces the total time spent by 8.3% with respect to the no-control case, outperforming the traditional ramp metering strategies without consideration of off-ramp traffic dynamics.