A New Safe Lane-Change Trajectory Model and Collision Avoidance Control Method for Automatic Driving Vehicles.
Tao Peng,Lili Su,Ronghui Zhang,Zhiwei Guan,Honglin Zhao,Zhijun Qiu,Changfu Zong,Hongguo Xu
DOI: https://doi.org/10.1016/j.eswa.2019.112953
IF: 8.5
2019-01-01
Expert Systems with Applications
Abstract:Lane change maneuvers, are important contributors to road traffic accidents on highway. In this paper, we propose a new safe lane change trajectory model and collision avoidance control method for Society of Automotive Engineers (SAE)-level 2 automatic driving vehicles. First, a Gaussian distribution is used to describe the new trajectory model that uses pure steering and combined braking. According to regional and progressive states, a new safe lane change meaning is defined. Second, we design a new four-level automatic driving mode and an effective decision mechanism that considers safety and ergonomics. Moreover, a new trajectory tracking controller combined with a decision mechanism was designed using feedback linearization, verified using typical lane-change scenarios. Finally, based on a physical simulation platform, PreScan, the hardware-in-the-loop simulation result demonstrate the feasibility and effectiveness of our method. This paper provides a valuable reference on an expert and intelligent system methodology for automatic driving vehicles, which will be helpful for improving highway traffic safety and efficiency. (C) 2019 Elsevier Ltd. All rights reserved.