A freeway vehicle early warning method based on risk map: Enhancing traffic safety through global perspective characterization of driving risk

Chuang Cui; Bocheng An; Linheng Li; Xu Qu; Huhe Manda; Bin Ran

doi:10.1016/j.aap.2024.107611

A freeway vehicle early warning method based on risk map: Enhancing traffic safety through global perspective characterization of driving risk

Accid Anal Prev. 2024 May 10:203:107611. doi: 10.1016/j.aap.2024.107611. Online ahead of print.

Authors

Chuang Cui¹, Bocheng An¹, Linheng Li², Xu Qu¹, Huhe Manda³, Bin Ran¹

Affiliations

¹ School of Transportation, Southeast University, Nanjing, Jiangsu Province 211189, China; Institute on Internet of Mobility, Southeast University and University of Wisconsin-Madison, Southeast University, Nanjing, Jiangsu Province 211189, China; Jiangsu Province Collaborative Innovation Center of Modern Urban Traffic Technologies, Southeast University, Nanjing, Jiangsu 211189, China.
² School of Transportation, Southeast University, Nanjing, Jiangsu Province 211189, China; Institute on Internet of Mobility, Southeast University and University of Wisconsin-Madison, Southeast University, Nanjing, Jiangsu Province 211189, China; Jiangsu Province Collaborative Innovation Center of Modern Urban Traffic Technologies, Southeast University, Nanjing, Jiangsu 211189, China. Electronic address: leelinheng@seu.edu.cn.
³ Ordos New Energy Development and Utilization Co., Ltd, Ordos, Inner Mongolia Autonomous Region, 017000, China.

PMID: 38733809
DOI: 10.1016/j.aap.2024.107611

Abstract

In the era of rapid advancements in intelligent transportation, utilizing vehicle operating data to evaluate the risk of freeway vehicles and study on vehicle early warning methods not only lays a theoretical foundation for improving the active safety of vehicles, but also provides the technical support for reducing accident rate. This paper proposes a freeway vehicle early warning method based on risk map to enhance vehicle safety. Firstly, Modified Time-to-Collision (MTTC), a two-dimensional indicator that describes the risk of inter-vehicle travel, is used as an indicator of road traffic risk. This paper designs a transformation function to probabilistically transform MTTC into Risk Indicators (RI). The single-vehicle risk map is generated based on the mapping relationship between the risk values and the corresponding roadway segments. These single-vehicle risk maps of all vehicles on the road are superimposed to construct the risk map, which is used to describe the risk distribution in the freeway. Then, a vehicle early warning framework is built based on the risk map. The risk values in the risk map are compared with predefined early warning thresholds to alert the vehicle when it enters a high-risk state. Finally, VISSIM is used to carry out simulation experiments. The experiment simulates a freeway accident stopping situation. This scenario includes sub-scenarios such as unplanned stopping and lane-changing, continuous lane-changing, and adjacent lane overtaking. We analyze the risk map and vehicle warning results in different sub-scenarios, evaluate the risk changes of the vehicles before and after receiving the warning, and compare the warning results of the method in this paper with other alternative methods. The method is applied to 17 vehicles in the simulation to adjust their motion states. The results show that the total warning time is reduced by 29.6% and 73.3% of vehicles change lanes away from the accident vehicle. The overall results validate the effectiveness of the vehicle early warning method based on risk map proposed in this paper.

Keywords: Freeway; Risk characterization; Risk map; Traffic safety; Vehicle early warning.