Transport researchers and practitioners have long been seeking capable solutions to deal with the traffic oscillations caused by freeway merging. Although existing approaches based on ramp metering have improved the overall efficiency of on-ramps, their performance is still far below the theoretical capacity. The recently proposed detecting technology of autonomous vehicles (AVs) provides an alternative for maximizing the merging efficiency by developing and using appropriate controllers for AVs. In this paper, we develop a cooperative intelligent driver model in order to examine the system performance under different proportions of AVs. The results show that, with a proper vehicle-to-vehicle controlling mechanism, an increasing percentage of AVs will reduce the total travel time and smooth traffic oscillations.

Objective: This article is to assess the performance of 3 macroscopic safety indicators (speed, speed dispersion, and volume) and two microscopic potential crash risks (time to collision and deceleration rate to avoid crash) on safety evaluation for expressways. Methods: Field data were collected at 3 locations for 4 different time periods on an expressway in Beijing, China. The speed of each vehicle, headway time, and vehicle length were recorded by a traffic management system. The 5 safety indicators were thus calibrated on the basis of the collected data. Further, consistency and comparative analyses were applied to assess the performance of indicators. Results: According to the analyses, speed dispersion was a better predictor of the two microscopic potential risks compared to the two macroscopic indicators. Conclusions: Speed dispersion is recommended to proactively assess road safety because (1) it provides consistent risk evaluation with microscopic potential risks and (2) it makes data collection easier.