Dynamic speed harmonization for mixed traffic flow on the freeway using deep reinforcement learning

Abstract

AbstractIn the vicinity of weaving areas, freeway congestion is nearly unavoidable due to their negative effects on the continuous freeway mainline flow. The adverse impacts include increased collision risks, extended travel time, and excessive emissions and fuel consumption. Dynamic Speed Harmonization (DSH) has the potential to dampen traffic oscillation during congestion. However, its effectiveness is typically limited by the low compliance rates of drivers and delays in information access. The integration of Connected and Automated Vehicles (CAVs) into intelligent transportation systems aims to enhance various measures of effectiveness. This research investigates the effects of DSH in mixed traffic flow involving human‐driven vehicles and CAVs on the freeway. A deep reinforcement learning‐based strategy is developed to better understand how CAVs can improve operational performance. A holistic evaluation is conducted to quantify the impacts under different penetration rates of CAVs in multiple simulated scenarios. The results reveal that the proposed method can enhance mobility and achieve co‐benefits with safety, and environmental sustainability could be improved under higher penetration rates. Spatiotemporal features of bottleneck speed demonstrate that DSH powered by CAVs can smooth speed variations for partial areas. Sensitivity analysis of headways indicates that high‐level CAVs can further improve performance.