主办：陕西省汽车工程学会
ISSN 1671-7988  CN 61-1394/TH
创刊：1976年
汽车实用技术

汽车实用技术 ›› 2022, Vol. 47 ›› Issue (14): 109-112.DOI: 10.16638/j.cnki.1671-7988.2022.014.024

• 测试试验 • 上一篇    

基于高速公路的驾驶员换道意图识别

郎悦茹   

  1. 长安大学 汽车学院
  • 发布日期:2022-09-15
  • 通讯作者: 郎悦茹
  • 作者简介:郎悦茹（1998—），女，硕士研究生，研究方向为载运工具运用工程，E-mail: 358021948@qq.com。

Lane Change Intention Identification of Motorists Based on Expressway

LANG Yueru   

  1. School of Automobile, Chang’an University
  • Published:2022-09-15
  • Contact: LANG Yueru

PDF

摘要： 近年来，车辆安全问题成为社会关注的焦点。高速公路存在两类典型的驾驶员行车意 图，分别为车道保持和意图换道。在意图换道的过程中，对驾驶人控制车辆提出了更高的要 求。基于 BP 神经网络提出了一个模型，用于检测高速公路上驾驶员的车道变更意图。预测结 果表明，在换道时刻，预测精度可达到 99.3%，以 0.5 s 为间隔向前推，换道前 1 s 的准确率为 98.0%，换道前 2 s 的准确率为 84.8%，换道前 3 s 的准确率为 70.7%；随机选取样本对模型的 准确率进行验证，换道时刻准确率为 96.7%。为深入研究驾驶员的输入特征奠定基础。

关键词: 高速公路；驾驶意图；车道变换；BP 神经网络

Abstract: In recent years, vehicle safety has become the focus of social attention. There are two kinds of typical drivers' driving intentions in expressway, which are lane keeping and lane changing. In the process of intending to change lanes, higher requirements are put forward for the driver to control the vehicle. In this paper, a model based on BP neural network is proposed to detect the driver's lane change intention on expressway. The prediction results show that at the time of lane change, the prediction accuracy can reach 99.3%, pushing forward at an interval of 0.5 s, the accuracy of 1 s before lane change is 98.0%, the accuracy of 2 s before lane change is 84.8%, and the accuracy of 3 s before lane change is 70.7%: Random samples are selected to verify the accuracy of the model, and the accuracy of lane changing time is 96.7%. It lays a foundation for in-depth study of driver input characteristics.

Key words: Expressway; Driving intentions; Lane change; BP neural network