A Study on the Impact Mechanism of Human-Machine Mixed Driving Traffic Flow Under Occasional Accident

With the ongoing development of integrated vehicle-road-cloud systems, mixed traffic flow composed of human-driven vehicles (HDVs) and connected and autonomous vehicles (CAVs) is expected to become the dominant form of future transportation. To explore the influence mechanism of CAV human-like driving strategy and sensing information capability on human-machine mixed driving traffic flow under occasional accident, this paper improved the cellular automata rules under the framework of the KKW (Kerner-Klenov-Wolf) model, introduced the synchronization factor to consider the CAV human-like driving strategy, and constructed the HDV and CAV car-following rules for different following modes. Considering the lane-changing demand in accident scenarios, a multi-lane discretionary lane-changing strategy incorporating the lane preference of HDVs and CAVs was constructed, along with a mandatory lane-changing rule for CAVs based on lane-changing pressure. Sensitivity analysis was conducted on different lane-changing pressure parameters. Through numerical simulations, the effects of varying traffic volume, CAV penetration rate, CAV perception range of accident information, and CAV human-like driving strategies on mixed traffic flow were analyzed. The results show that the increase of CAVs can effectively alleviate the congestion of traffic flow after occasional accident and limit the spatial and temporal scope of congestion, and the average speed and average traffic volume of the low traffic volume are increased by 11.74% and 6.32%, respectively, when CAV penetration rate is increased from 0 to 1. The enhancement is lower than that of medium and high traffic volume. In the case of medium and high traffic volume with CAV penetration rate greater than 0.4, with the increase of CAV accident information sensing range, the congestion space in the merging area is gradually dispersed, and traffic efficiency is improved. With the transition of the CAV human-like driving strategy from aggressive to conservative, the flow of the human-machine mixed driving traffic flow is gradually reduced, and the range of slow queues expands, traffic congestion gradually worsens, and the trend of speed fluctuations in each lane gradually converges over time.