基于危险场景聚类分析的前车随机运动状态预测研究*

doi:10.19562/j.chinasae.qcgc.2020.07.001

汽车工程 ›› 2020, Vol. 42 ›› Issue (7): 847-853.doi: 10.19562/j.chinasae.qcgc.2020.07.001

• • 下一篇

基于危险场景聚类分析的前车随机运动状态预测研究^*

郭景华¹, 李克强², 王进¹, 陈涛³, 李文昌¹, 王班¹

1.厦门大学机电工程系,厦门 361005;
2.清华大学车辆与运载学院,北京 100084;
3.中国汽车工程研究院股份有限公司,重庆 401122

收稿日期:2019-06-24 出版日期:2020-07-25 发布日期:2020-08-14
通讯作者: 郭景华,副教授,博士,E-mail:guojh@xmu.edu.cn。
基金资助:
*国家重点研发计划(2016YFB0100900)和中央高校基本科研业务费专项资金(20720190015)资助

Study on Prediction of Preceding Vehicle's Stochastic Motion Based on Risk Scenarios Clustering Analysis

Guo Jinghua¹, Li Keqiang², Wang Jin¹, Chen Tao³, Li Wenchang¹, Wang Ban¹

1. Department of Mechanical and Electrical Engineering, Xiamen University, Xiamen 361005;
2. School of Vehicle and Mobility, Tsinghua University, Beijing 100084;
3. China Automotive Engineering Research Institute Co., Ltd., Chongqing 401122

Received:2019-06-24 Online:2020-07-25 Published:2020-08-14

摘要/Abstract

摘要： 我国道路情况复杂多变,构建合适的测试场景对汽车智能驾驶系统测试和评价的有效性起着决定性作用。本文中从我国实际道路出发,从自然驾驶数据中提取具有中国特色的典型场景,筛选出自然驾驶数据中危险工况数据片段,基于此提取出汽车智能驾驶系统综合测试的场景特征要素,并利用聚类分析法得到3类典型危险场景。采用马尔可夫链理论表征前车人类驾驶员驾驶车辆的随机运动特性,将聚类得到各场景下的自车数据作为前车历史工况数据,归纳学习得出马氏链转移概率,并通过马尔可夫链蒙特卡洛模拟预测未来时刻的状态,基于此得到危险场景中前车随机运动预测模型,通过对比原始工况数据验证预测模型的有效性,有效解决了由于采集设备精度低导致的前车数据不准、在测试场景中不能准确表征前车人类驾驶员驾驶车辆随机运动的问题。

关键词: 智能驾驶, 自然驾驶数据, 危险场景聚类, 马尔科夫蒙特卡洛模拟, 预测

Abstract: Due to complex and changeable road situation in China, the construction of a suitable test scenario plays a decisive role in the validity of intelligent driving system test and evaluation. Based on the actual road traffic environment in China, typical scenarios with Chinese characteristics are extracted and the data fragments of risk working condition are selected from the naturalistic driving data. Then, the scenario feature elements of the comprehensive tests of intelligent driving system are extracted, and three types of typical risk scenarios are obtained by clustering analysis. Then the stochastic motion of the vehicle driven by the human driver in front of the subject vehicle is featured by the Markov chain theory and the subject vehicle data in each scenario obtained by clustering is used as the historical working condition data of the preceding vehicle. The Markov chain transfer probability is concluded and learned, and the state of the future time is predicted by Markov chain Monte Carlo simulation. Based on this, the stochastic motion prediction model of the preceding vehicle in the risk scenarios is obtained. The validity of the proposed model is verified by comparing the original working condition data, which effectively solves the problem that the preceding vehicle data is in accurate due to the low accuracy of the acquisition equipment and the stochastic motion of the preceding vehicle driven by the human driver cannot be accurately featured in test scenarios.

Key words: intelligent drive, naturalistic driving data, clustering risk scenarios, Markov Chain Monte Carlo, prediction

郭景华, 李克强, 王进, 陈涛, 李文昌, 王班. 基于危险场景聚类分析的前车随机运动状态预测研究^*[J]. 汽车工程, 2020, 42(7): 847-853.

Guo Jinghua, Li Keqiang, Wang Jin, Chen Tao, Li Wenchang, Wang Ban. Study on Prediction of Preceding Vehicle's Stochastic Motion Based on Risk Scenarios Clustering Analysis[J]. Automotive Engineering, 2020, 42(7): 847-853.

参考文献

[1] 郭景华,李克强,罗禹贡.智能车辆运动控制研究综述[J].汽车安全与节能学报,2016(7):151-159.
[2] ZHAO D. Accelerated evaluation of automated vehicles[D]. Michigan: University of Michigan,2016.
[3] HUANG W, WEN D, GENG J, et al. Task-specific performance evaluation of UGVs: case studies at the IVFC[J]. IEEE Transactions on Intelligent Transportation Systems,2014,15(5):1969-1979.
[4] BAGSCHIK G, MENZEL T, MAURER M. Ontology based scene creation for the development of automated vehicles[C]. 2018 IEEE Intelligent Vehicles Symposium (IV),2018:1813-1820.
[5] MENZEL T, BAGSCHIK G, MAURER M. Scenarios for development, test and validation of automated vehicles[C]. 2018 IEEE Intelligent Vehicles Symposium (IV),2018:1821-1827.
[6] 吴斌,朱西产,沈剑平,等.基于自然驾驶研究的直行追尾危险场景诱导因素分析[J].同济大学学报,2018,46(9):1254-1260.
[7] 胡林,易平,黄晶,等.基于真实事故案例的自动紧急制动系统两轮车测试场景研究[J].汽车工程,2018,40(12):1436-1446.
[8] 宗长富,杨肖,王畅,等.汽车转向时驾驶员驾驶意图辨识与行为预测[J].吉林大学学报(工学版),2009,39(1):28-32.
[9] 彭金栓,付锐,郭应时.自然驾驶条件下驾驶人换道行为实时预测[J].哈尔滨工业大学学报,2015,47(9):120-122.
[10] 熊晓夏,陈龙,梁军,等.基于驾驶员避撞行为的行车风险判别方法的仿真研究[J].汽车工程,2019,41(2):154-160.
[11] 倪捷,刘志强,涂孝军,等.面向驾驶辅助系统的换道安全性预测模型研究[J].交通运输系统工程与信息,2016,16(4):96-100.
[12] 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):39-42.
[13] 张金辉,李克强,罗禹贡,等.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[J].汽车工程,2019,41(3):246-251.
[14] HANKEY J M, PEREZ M A, MCCLAFFERTY J A. Description of the SHRP 2 naturalistic database and the crash, near-crash, and baseline data sets[R]. Blacksburg, VA: Transportation Research Board of The National Academies,2016:1-18.
[15] SHONKWILER W, MENDIVIL F. Explorations in Monte Carlo methods[M]. New York: Springer,2009.

基于危险场景聚类分析的前车随机运动状态预测研究^*

Study on Prediction of Preceding Vehicle's Stochastic Motion Based on Risk Scenarios Clustering Analysis

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

Metrics

本文评价

推荐阅读 10

[1]	杨彪, 韦智文, 倪蓉蓉, 王海, 蔡英凤, 杨长春. 基于动作条件交互的高效行人过街意图预测[J]. 汽车工程, 2024, 46(1): 29-38.
[2]	王军年, 程川泰, 高菲, 付铁军, 任金东. 共享汽车驾驶员适宜H点和座椅参数实验研究[J]. 汽车工程, 2024, 46(1): 39-49.
[3]	秦洪懋,沈国利,周云水,黄圣杰,秦晓辉,谢国涛,丁荣军. 特征稀疏场景下基于标签的车辆视觉SLAM[J]. 汽车工程, 2023, 45(9): 1543-1552.
[4]	赵高士,陈龙,蔡英凤,廉玉波,王海,刘擎超,滕成龙. 融合复杂网络和记忆增强网络的轨迹预测技术[J]. 汽车工程, 2023, 45(9): 1608-1616.
[5]	梅润,褚端峰,高博麟,李克强,丛炜,陈超义. 基于云控系统的队列预测性巡航与换道决策[J]. 汽车工程, 2023, 45(8): 1299-1308.
[6]	连静,李硕贤,刘一荻,杨东方,李琳辉. 基于车道目标引导的车辆轨迹预测[J]. 汽车工程, 2023, 45(8): 1353-1361.
[7]	王明,唐小林,杨凯,李国法,胡晓松. 考虑预测风险的自动驾驶车辆运动规划方法[J]. 汽车工程, 2023, 45(8): 1362-1372.
[8]	高锋,冯德福,胡秋霞. 面向NMPC运动规划系统的数值优化加速技术[J]. 汽车工程, 2023, 45(8): 1438-1447.
[9]	胡启慧,蔡英凤,王海,陈龙,董钊志,刘擎超. 基于层次图注意的异构多目标轨迹预测方法[J]. 汽车工程, 2023, 45(8): 1448-1456.
[10]	陈国强,申正义,孙利,支梦帆,李彤. 基于BP神经网络优化遗传算法的智能座舱感性意象预测[J]. 汽车工程, 2023, 45(8): 1479-1488.
[11]	林程, 汪博文, 吕沛原, 宫新乐, 于潇. 面向变曲率道路的自动驾驶汽车换道博弈运动规划与协同控制研究[J]. 汽车工程, 2023, 45(7): 1099-1111.
[12]	高镇海, 鲍明喜, 高菲, 唐明弘. 基于LSTM概率多模态预期轨迹预测方法[J]. 汽车工程, 2023, 45(7): 1145-1152.
[13]	李军, 周伟, 唐爽. 基于自适应拟合的智能车换道避障轨迹规划[J]. 汽车工程, 2023, 45(7): 1174-1183.
[14]	查云飞,吕小龙,陈慧勤,易迎春,王燕燕. 基于路面附着系数估计的车辆轨迹跟踪控制[J]. 汽车工程, 2023, 45(6): 1010-1021.
[15]	韩勇,林旭洁,黄红武,蔡鸿瑜,罗金镕,李燕婷. 典型汽车碰撞事故场景中行人运动轨迹预测方法[J]. 汽车工程, 2023, 45(6): 1022-1030.