.png){kind=icon}
.png){kind=icon}
Return
Risk-based merging decisions for autonomous vehicles
.png){kind=icon}
PRE
AI
DOI：10.1016/j.jsr.2022.08.003.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Introduction: The safe freeway merging operation for fully Autonomous Vehicles (AVs) in mixed traffic (i.e., the presence of AVs and non-AVs in a traffic stream) is a challenging task. Under a mixed traffic envi-ronment, an AV merging operation could significantly increase conflict risks and reduce operational effi-ciency. Method: This study quantifies the freeway merging conflict risk and develops a freeway merging decision strategy based on conflict risk assessment for an AV attempting to merge to a traffic stream with non-AVs on the freeway. The performance of the risk-based merging decision strategy is evaluated in uncongested, near-congested, and congested traffic conditions. Results: The analyses show that the risk-based merging strategy causes less abrupt deceleration of an AV's immediate upstream vehicle in the target lane on the freeway compared to the based models (i.e., two models based on gap acceptance concepts and a safe gap model based on a surrogate measure, 'Time-to-Collision (TTC)'). The risk-based merging strategy meets the minimum safe gap between an AV intending to merge and its immediate downstream vehicle in the target lane. The risk-based merging strategy produces lower conflict risk in terms of 'Time Exposed Time-to-Collision (TET)' and 'Time Integrated Time-to-Collision (TIT)' compared to the base models. Moreover, the risk-based merging strategy has a lower impact on the average speed of traffic in the target lane compared to the base models considered in this study. Conclusions: The risk -based merging strategy shows higher safety benefits for an AV's merging operation compared to base models. Practical Applications: The findings of this research would help design AV controllers for improv-ing the safety of an AV merging operation in a mixed traffic stream.(c) 2022 National Safety Council and Elsevier Ltd. All rights reserved.
Keywords:
Autonomous vehicle
Merging decision
Risk assessment
Mixed traffic
Journal
IF:
4.4
Papers:
2.7K
Citations:
6.7K
View more organizations