Vehicle‐localization‐based and DSRC‐based autonomous vehicle rear‐end collision avoidance concerning measurement uncertainties

Abstract

Rear‐end collision is one of the most serious accidents for vehicles, some rear end collision avoidance systems have been developed to protect vehicles from rear‐end collisions. Current rear‐end collision avoidance systems mainly utilize vehicle‐relative measurement sensors to acquire information on surrounding vehicles, considering limited detection ranges and impaired performance in bad weathers of these vehicle sensors, currently used collision avoidance systems cannot guarantee vehicle safety in all environments. Vehicle‐localization‐based and intervehicle communication‐based rear‐end collision avoidance approaches can be possible solutions to improve vehicle safety compared with present vehicle sensor‐based collision avoidance systems. Dedicated Short‐Range Communication (DSRC) is introduced to transfer information on surrounding vehicles instead of using vehicle‐relative measurement sensors. Vehicle dynamic model and brake system dynamics are used to generate precise vehicle states that would be used in rear‐end collision avoidance. Required safe distance is used as the safety index when using different vehicle‐localization methods with a predefined safety probability of 99.99%. Monte Carlo simulations are executed to evaluate the influence of measurement uncertainties. Simulation results show that vehicle‐localization‐based and DSRC‐based rear‐end collision avoidance methods can guarantee vehicles free from rear‐end collision and expand the service capabilities of velocity compared with the traditional vehicle radar‐based rear‐end collision avoidance method. Evaluation outcomes encourage us to adopt a more precise absolute vehicle‐localization approach in rear‐end collision avoidance, because required safe distance and safe distance margin can be decreased when using a high‐precision vehicle‐localization method compared with standalone GNSS‐based localization. Some further aspects are discussed in this work that need to be tackled with to make vehicle‐localization‐based and intervehicle communication‐based rear‐end collision avoidance systems into practice, including real‐time brake system dynamics acquirement, benefit of DSRC to curve roads, and so on. © 2019 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

Analyzing Factors Influencing Situation Awareness in Autonomous Vehicles—A Survey

“…Ref. [94] proposed a modern Dedicated-Short-Range-Communication (DSRC) intervehicle communication system along with vehicle localization-based rear-end collision avoidance methods, to help the AVs avoid rear-end collisions. Hongliang Yuan et al [147] suggested a decision-making strategy for autonomous driving to avoid traffic collisions.…”

Multi-Sensor Fusion in Automated Driving: A Survey

“…In reference [81], DSRC is applied to communicate with surrounding vehicles, and the real-time state of vehicles determined by the vehicle dynamics model and braking system dynamics, which prevents rear-end collision with other vehicles and illustrates the reliability of DSRC communication. However, AD requires safety redundancy in the event of communication interference, and integration with other sensors is essential.…”

Real-Time Inter-Vehicle Data Fusion Based on a New Metric for Evidence Distance in Autonomous Vehicle Systems

“…Moreover, DS is effective in managing uncertainty in multisensor environments [15]. We begin by summarizing the mathematical framework of DS [16,19,20,40,41].…”

“…This problem is due to combining in (3) which ignores the conflicts in evidence. In order to illustrate this problem, let us consider the following example [19,40,43]. Given 3 events Z = {A, B, C} and 2 Observers, Observer 1 and 2 assign m 1 (A) = 0.9, m 1 (B) = 0.1, m 1 (C) = 0.0 and m 2 (A) = 0.0, m 2 (B) = 0.1, m 2 (C) = 0.9 as confidences in Z, respectively.…”