• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2024, Vol. 60 ›› Issue (20): 240-250.doi: 10.3901/JME.2024.20.240

• 运载工程 • 上一篇    下一篇

扫码分享

融合灰色预测和卡尔曼滤波的车辆侧向碰撞预警

张志勇1,2, 王宇翔1, 黄彩霞2,3, 吴悠1, 杜荣华1   

  1. 1. 长沙理工大学汽车与机械工程学院 长沙 410114;
    2. 长沙理工大学机械装备高性能智能制造关键技术湖南省重点实验室 长沙 410114;
    3. 湖南工程学院汽车动力与传动系统湖南省重点实验室 湘潭 411104
  • 收稿日期:2023-10-08 修回日期:2024-04-18 出版日期:2024-10-20 发布日期:2024-11-30
  • 通讯作者: 杜荣华,男,1973年出生,博士,教授。主要研究方向为智能汽车主动安全控制,智能交通与车路协同技术。E-mail:csdrh@163.com
  • 作者简介:张志勇,男,1976年出生,博士,副教授。主要研究方向为智能汽车主动安全控制,车辆动力学及控制。E-mail:zzy04@163.com
  • 基金资助:
    国家自然科学基金(52472399);湖南省自然科学基金(2022JJ50020,2021JJ30182);湖南省教育厅科学研究(20A018)和机械装备高性能智能制造关键技术湖南省重点实验室(长沙理工大学)开放基金(2020YB02)资助项目。

Vehicle Lateral Collision Warning Based on Grey Prediction and Kalman Filter

ZHANG Zhiyong1,2, WANG Yuxiang1, HUANG Caixia2,3, WU You1, DU Ronghua1   

  1. 1. College of Automobile and Mechanical Engineering, Changsha University of Science and Technology, Changsha 410114;
    2. Hunan Province Key Laboratory of Intelligent Manufacturing Technology for High-performance Mechanical Equipment, Changsha University of Science and Technology, Changsha 410114;
    3. Hunan Provincial Key Laboratory of Automotive Power and Transmission System, Hunan Institute of Technology, Xiangtan 411104
  • Received:2023-10-08 Revised:2024-04-18 Online:2024-10-20 Published:2024-11-30

摘要: 车辆碰撞预警是车辆主动安全控制关键技术,常需要利用准确的车辆状态信息进行运动轨迹的高精度预测,再由运动轨迹预测模块判断安全碰撞预警时间内是否存在碰撞风险。为提高碰撞预警精度,建立以恒定转弯率和加速度模型为状态转移方程,平方根容积卡尔曼滤波器为估计算法的状态估计方法,提高目标车相对运动状态的估计精度;提出融合灰色预测的车辆运动轨迹预测方法,通过灰色预测模型对量测变量进行多步预测,再由平方根容积卡尔曼滤波器进行校正,提高目标车相对运动轨迹的预测精度;最后,考虑路面附着系数对安全碰撞预警时间的影响,提出一种车辆侧向碰撞预警方法。数字仿真结果表明,在不同路面附着条件下,提出的车辆侧向碰撞预警方法都准确预测车辆的碰撞时间,提前的预警时间大于安全碰撞预警时间,确保驾驶员或主动避障控制系统能及时操控车辆,提高车辆行驶的安全性。

关键词: 碰撞预警, 轨迹预测, 状态估计, 灰色预测, 卡尔曼滤波

Abstract: Vehicle collision warning is the core technology of vehicle active safety control. It is often necessary to use accurate vehicle state information to precisely predict the motion trajectory, and then determine whether there is a collision risk within the safe collision warning time by the motion trajectory prediction module. To improve the accuracy of collision warning, a state estimation method using constant turning rate and acceleration model as the state transition equation, and square root volume Kalman filter as the estimation algorithm is established first, which is beneficial to improving the estimation accuracy of the relative motion state of the target vehicle. A relative motion trajectory prediction method integrated with grey prediction is then proposed. The measured variables are predicted in multiple steps through the grey prediction model and are corrected by the square root volume Kalman filter to improve the prediction accuracy of the relative motion trajectory of the target vehicle. Considering the influence of road adhesion coefficient on the safe collision warning time, a vehicle lateral collision warning method is proposed at last. The numerical simulation results show that the warning methods proposed can accurately predict the collision time of the vehicle under different road conditions, and the early warning time is greater than the safe collision warning time, so as to ensure that the driver or the active obstacle avoidance control system can control the vehicle timely and improve the safety of vehicle driving.

Key words: collision warning, trajectory prediction, state estimation, grey prediction, Kalman filter

中图分类号: