基于贝叶斯网络的联网环境中跟车工况下的前车运动状态预测

详细信息    查看全文 | 推荐本文 |

英文篇名：Prediction of Preceding Car Motion Under Car-following Scenario in the Internet of Vehicle Based on Bayesian Network 作者：张金辉 ; 李克强 ; 罗禹贡 ; 张书玮 ; 李红 英文作者：Zhang Jinhui;Li Keqiang;Luo Yugong;Zhang Shuwei;Li Hong;Tsinghua University, State Key Laboratory of Automotive Safety and Energy; 关键词：智能驾驶 ; 状态预测 ; 车联网 ; 贝叶斯网络 英文关键词：intelligent driving;;state prediction;;Internet of vehicle;;Bayesian network 中文刊名：QCGC 英文刊名：Automotive Engineering 机构：清华大学汽车安全与节能国家重点实验室; 出版日期：2019-03-25 出版单位：汽车工程 年：2019 期：v.41;No.296 基金：国家重点研发计划项目(2016YFB0100900)资助 语种：中文; 页：QCGC201903002 页数：8 CN：03 ISSN：11-2221/U 分类号：10-16+39

摘要

车辆行驶过程中,前方车辆运动状态预测是车辆智能控制系统的重要研究部分。车辆运动状态受其驾驶员驾驶风格、道路状况、交通流、前方车辆运动速度和加速度等因素的影响,使车辆在未来一定时间段内的运动状态具有较大不确定性,给前方车辆的运动状态预测带来困难,因此本文中对跟车工况下前车运动状态预测进行研究。本文中在分析车辆跟车工况时的运动特性,采用贝叶斯网络对前方车辆运动速度进行预测,将获得的车辆跟车工况时的运动状态数据分为训练集和测试集。通过训练集辨识前车速度预测贝叶斯网络参数,通过测试集检验前车速度预测贝叶斯网络的预测效果。对前车未来0.1,0.5,1和2s时的运动速度进行预测,预测结果表明,前车的实际运动速度均在前车速度预测贝叶斯网络预测的95%置信区间内。
        In the process of vehicle driving, the prediction of motion state of vehicle in front is an important research part of intelligent vehicle control system. Influenced by factors such as the driver's driving style, road condition, traffic flow, the speed and acceleration of the vehicle ahead, the motion state of the vehicle in a certain period of time in the future has great uncertainties, which brings difficulties to the prediction of the motion state of the vehicle ahead. This paper studies the prediction of the motion state of the vehicle ahead under the car-following conditions. In this paper, the motion characteristics of the vehicle under car-following conditions are analyzed, and Bayesian network is used to predict the speed of the vehicle ahead. The obtained motion state data of the vehicle under car-following conditions are divided into training set and test set. The training set is used to identify the parameters of the front vehicle speed for prediction of Bayesian network parameters, and the test set is used to verify the prediction effect of Bayesian network for front vehicle speed prediction. The prediction results show that the actual speed of the front vehicle is within 95% confidence interval predicted by Bayesian network for the prediction of the front vehicle speed in the next 0.1,0.5,1 and 2 s.

引文

[1] 高振海,王竣,佟静,等.车载毫米波雷达对前方目标的运动状态估计[J].吉林大学学报:工学版,2014(6):1537-1544.
    [2] 肖献强,任春燕,王其东.基于隐马尔可夫模型的驾驶行为预测方法研究[J].中国机械工程,2013,24(21):2972-2976.
    [3] 葛如海,管军,虞小波,等.车辆姿态感知与运动状态预测技术[J].计算机与现代化,2013(8):38-42.
    [4] 王兴伟,原常青,黄敏.一种基于模糊决策树的运动预测机制[J].计算机科学,2005(9):176-179.
    [5] HAMDAR S. Driver behavior modeling[M]. Handbook of Intelligent Vehicles. Springer London,2012:537-558.
    [6] PIAO J, MCDONALD M. Low speed car following behaviour from floating vehicle data[C]. Intelligent Vehicles Symposium,2003. Proceedings. IEEE,2003:462-467.
    [7] SIMONSSON S O. Car-following as a tool in road traffic simulation[C]. Vehicle Navigation and Information Systems Conference. IEEE,1993:150-153.
    [8] SAIFUZZAMAN M, ZHENG Z, HAQUE M M, et al. Revisiting the task-capability interface model for incorporating human factors into car-following models[J]. Transportation Research Part B: Methodological,2015,82:1-19.
    [9] HAO H, MA W, XU H. A fuzzy logic-based multi-agent car-following model[J]. Transportation Research Part C: Emerging Technologies,2016,69:477-496.
    [10] YU S, ZHAO X, XU Z, et al. An improved car-following model considering the immediately ahead car’s velocity difference[J]. Physica A: Statistical Mechanics and Its Applications,2016,461:446-455.
    [11] 李升波.车辆多目标协调式自适应巡航控制[D].北京,清华大学,2009.