Will Autonomous Vehicles Improve Traffic Efficiency and Safety in Urban Road Bottlenecks? The Penetration Rate Matters
Paper i proceeding, 2020

The emerging autonomous vehicles (AVs) are expected to bring pronounced evolutions in transport systems. This study explores the characteristics of mixed traffic flow with both AVs and human drivers in urban bottlenecks. We investigate the influences of penetration rate (PR) of AVs on the performances concerning traffic efficiency and safety in urban bottlenecks with road width reduction. We developed a cellular automata model (CAM) to realize the microscope simulation of the mixed traffic flow with both AVs and traditional vehicles manipulated by human drivers. The divergences in driving behavior of human drivers and AVs in terms of car-following, lane-change and free-driving are fully delineated and integrated in the simulation. The results demonstrate that the traffic flow stability firstly decreases and then increases with the PRs of AVs and in mixed traffic flow. When PR of AVs reaches 100%, the traffic flow is stabilized and shows high travel speed, indicating higher traffic efficiency. The lane-changing frequency increases when PR of AVs increases, reaching the maximum value at the PRs of 15%-25%, and then gradually drops. The lane-changing frequencies under the scenarios of all AVs are found to be smaller than the scenarios of all human drivers. The actual road capacity is reduced when PR of AVs increases at first, reaches lowest at the PR of 15%-25%, and then gradually rebounds. The risk of collision gradually increases with PRs of AVs, and then reaches the maximum value at the PR of 25%-30%. As PR of AVs continues to increase, the risk will keep decreasing to 0. The findings provide a comprehensive investigation of how the AVs will influence traffic efficiency and safety from different aspects, which are basic for the development and planning of AVs in the future.

mixed traffic flow


road width reduction

cellular automata model

autonomous vehicle


Tianshu Zhang

Student vid Chalmers

Kun Gao

Chalmers, Arkitektur och samhällsbyggnadsteknik, Geologi och geoteknik

2020 IEEE 5th International Conference on Intelligent Transportation Engineering, ICITE 2020

366-370 9231360
9781728194097 (ISBN)

5th IEEE International Conference on Intelligent Transportation Engineering, ICITE 2020
Beijing, China,


Transportteknik och logistik





Mer information

Senast uppdaterat