基于马尔科夫毯和隐朴素贝叶斯的驾驶行为险态辨识

doi:10.13229/j.cnki.jdxbgxb201606013

吉林大学学报(工学版) ›› 2016, Vol. 46 ›› Issue (6): 1851-1857.doi: 10.13229/j.cnki.jdxbgxb201606013

基于马尔科夫毯和隐朴素贝叶斯的驾驶行为险态辨识

严利鑫^{1, 2, 3}, 黄珍⁴, 朱敦尧^{1, 2}, 陈志军^{1, 2}, 冉斌³

1.武汉理工大学智能交通系统研究中心,武汉430063;
2.国家水运安全工程技术研究中心,武汉430063;
3.威斯康星大学麦迪逊分校土木与环境工程学院, 麦迪逊 53705;
4.武汉理工大学自动化学院,武汉430070

收稿日期:2016-03-29 出版日期:2016-11-20 发布日期:2016-11-20
通讯作者: 黄珍(1974-),女,副教授,博士.研究方向:智能交通,交通安全.E-mail:h-zhen@whut.edu.cn
作者简介:严利鑫(1988-),男,博士研究生.研究方向:智能车路关键技术,驾驶行为.E-mail:yanlixinits@126.com
基金资助:
国家自然科学基金项目(51178364,61104158); 武汉理工大学教学研究项目(2011180)

Driving risk status identification based on Markov blanket hidden Naive Bayes

YAN Li-xin^{1, 2, 3}, HUANG Zhen⁴, ZHU Dun-yao^{1, 2}, CHEN Zhi-jun^{1, 2}, RAN Bin³

1.Intelligent Transportation Systems Center, Wuhan University of Technology, Wuhan 430063, China;
2.National Engineering Research Center for Water Transport Safety, Wuhan 430063, China;
3.Transportation Engineering Laboratory, University of Wisconsin-Madison, Madison 53705, USA;
4.School of Automation, Wuhan University of Technology, Wuhan 430070, China

Received:2016-03-29 Online:2016-11-20 Published:2016-11-20

摘要/Abstract

摘要： 为了实现对驾驶行为险态的有效辨识,以实时采集的多源信息为依据,通过融合驾驶人心率变化率及违法行为将驾驶行为险态分为4级。采用马尔科夫毯特征抽取算法提取出速度、纵向加速度、前轮转角变化率、车道偏离量以及车辆位置作为构建驾驶行为险态辨识的特征集,基于隐朴素贝叶斯(HNB)构建驾驶行为险态辨识模型。十折交叉验证结果表明,该模型的辨识精度(90.6%)比朴素贝叶斯(NB)、贝叶斯网络(BN)及径向基函数(RBF)神经网络分别提高14.1%、13.9%和13%。此外,ROC曲线验证结果表明该模型对不同险态等级都具有良好的预测效果。

关键词: 道路工程, 险态辨识, 马尔科夫毯, 隐朴素贝叶斯, 交通安全

Abstract: In order to effectively identify risk status while driving, a driving risk status identification model is proposed based on the information of driver operation and vehicle status. According to the rate of electrocardiogram (ECG) and traffic violation behaviors, the driving risk status is classified into four levels. Using Markov blanket algorithm, five factors are selected as the feature set, including the speed, the longitudinal acceleration, the rate of front wheel angle, the vehicle position and lane departure. Then, the algorithm of Hidden Naive Bayes (HNB) is employed to establish the driving risk status identification model. The results show that the accuracy of HNB is 90.6%, increasing 14.1%, 13.9 and 13.0% compared with Bayesian Network (BN), Naive Bayes (NB) and Radial Basis Function Neural Network, respectively. The results of ROC curve indicates that the model presents high predictive power. The conclusions can provide theoretical support for designing dangerous driving status recognition equipment based on vehicle and driver operation information.

Key words: road engineering, risk status identification, Markov blanket, Hidden Naive Bayes, traffic safety

中图分类号:

U491

严利鑫, 黄珍, 朱敦尧, 陈志军, 冉斌. 基于马尔科夫毯和隐朴素贝叶斯的驾驶行为险态辨识[J]. 吉林大学学报(工学版), 2016, 46(6): 1851-1857.

YAN Li-xin, HUANG Zhen, ZHU Dun-yao, CHEN Zhi-jun, RAN Bin. Driving risk status identification based on Markov blanket hidden Naive Bayes[J]. 吉林大学学报(工学版), 2016, 46(6): 1851-1857.

参考文献

[1] 石京,肖遥. 驾驶心理对交通安全的影响[J]. 交通信息与安全,2014,32(5):65-70.
Shi Jing, Xiao Yao. Effects of driver's phycology on traffic safety[J]. Journal of Transport Information and Safety,2014, 32(5):65-70.
[2] 严新平,张晖,吴超仲, 等. 道路交通驾驶行为研究进展及展望[J]. 交通信息与安全,2013,31(1):45-50.
Yan Xin-ping, Zhang Hui, Wu Chao-zhong, et al. Research progress and prospect of road traffic driving behavior[J]. Journal of Transport Information and Safety,2013,31(1):45-50.
[3] Siordia O S, de Diego I M, Conde C, et al. Combining traffic safety knowledge for driving risk detection[C]∥IEEE Conference on Intelligent Transportation Systems, Washington, DC, USA, 2011:564-569.
[4] de Diego I M, Siordia O S, Conde C, et al. Optimal experts' knowledge selection for intelligent driving risk detection systems[C]∥IEEE Intelligent Vehicles Symposium, Spain,2012:896-901.
[5] de Diego I M, Siordia O S, Crespo R, et al. Analysis of hands activity for automatic driving risk detection[J]. Transportation Research Part C: Emerging Technologies,2013,26:380-395.
[6] 马国忠,李燕,郭孜政,等. 驾驶行为险态辨识方法[J]. 公路交通科技,2013,30(7):113-118.
Ma Guo-zhong, Li Yan, Guo Zi-zheng, et al. A method for identifying risk status of driving behavior[J]. Journal of Highway and Transportation Research and Development,2013,30(7):113-118.
[7] Wu Zhi-zhou, Chen Tian-zi, Zhang Jian-qiao. Reckless status identification of bus driving behavior based on hidden Markov model[C]∥Eighth China Intelligent Transport Annual Meeting,Hefei,China,2013:178-186.
[8] 郭孜政. 驾驶行为险态辨识理论与方法[D]. 成都:西南交通大学交通运输与物流学院,2009.
Guo Zi-zheng. Theories and methods on driving risk status identification[D]. Chengdu: School of Transportation and Logistics, Southwest Jiaotong University,2009.
[9] 郭孜政,陈崇双,闫伟,等. 驾驶危险感知评估方法[J]. 吉林大学学报:工学版,2012,42(1):46-50.
Guo Zi-zheng. Chen Chong-shuang, Yan Wei, et al. Assessment method for driving threat perception[J]. Journal of Jilin University(Engineering and Technology Edition),2012,42(1):46-50.
[10] 陈雪梅,田赓,苗一松. 面向智能驾驶行为的机器学习[J]. 道路交通与安全,2014,14(6):60-64.
Chen Xue-mei, Tian Geng, Miao Yi-song. A survey on machine learning algorithms for intelligent driving behavior[J]. Road Transportation and Safety,2014,14(6):60-64.
[11] 吴晓端,吴志周. 汽车驾驶模拟器在交通安全中的应用综述[J]. 交通信息与安全,2015,33(2):10-19.
Wu Xiao-duan, Wu Zhi-zhou. A review of vehicle driving simulator application on traffic safety[J]. Journal of Transport Information and Safety,2015,33(2):10-19.
[12] Yan Li-xin, Zhu Dun-yao, Wu Chao-zhong, et al. Ranking and causal relationship analysis of incentive factors of driving anger: a case study from an on-road experiment in china[C]∥14th COTA International Conference of Transportation Professionals, Changsha, China,2014: 2534-2547.
[13] 秦大同,詹森,漆正刚,等. 基于K-均值聚类算法的行驶工况构建方法[J]. 吉林大学学报:工学版,2016,46(2):383-389.
Qin Da-tong, Zhan Sen, Qi Zheng-gang, et al. Driving cycle construction using K-means clustering method[J]. Journal of Jilin University Engineering and Technology (Engineering and Technology Edition),2016,46(2):383-389.
[14] Peña J M, Nilsson R, Björkegren J, et al. Towards scalable and data efficient learning of Markov boundaries[J]. International Journal of Approximate Reasoning,2007,45(2):211-232.
[15] Zhuang Hong-lei, Tang Jie, Tang Wen-bin, et al. Actively learning to infer social ties[J]. Data Mining and Knowledge Discovery,2012,25(2):270-297.
[16] Hsu W H. Genetic wrappers for feature selection in decision tree induction and variable ordering in Bayesian network structure learning[J]. Information Sciences,2004,163(1-3):105-120.

相关文章 15

[1]	李伊,刘黎萍,孙立军. 沥青面层不同深度车辙等效温度预估模型[J]. 吉林大学学报(工学版), 2018, 48(6): 1703-1711.
[2]	臧国帅, 孙立军. 基于惰性弯沉点的刚性下卧层深度设置方法[J]. 吉林大学学报(工学版), 2018, 48(4): 1037-1044.
[3]	念腾飞, 李萍, 林梅. 冻融循环下沥青特征官能团含量与流变参数灰熵分析及微观形貌[J]. 吉林大学学报(工学版), 2018, 48(4): 1045-1054.
[4]	宫亚峰, 申杨凡, 谭国金, 韩春鹏, 何钰龙. 不同孔隙率下纤维土无侧限抗压强度[J]. 吉林大学学报(工学版), 2018, 48(3): 712-719.
[5]	程永春, 毕海鹏, 马桂荣, 宫亚峰, 田振宏, 吕泽华, 徐志枢. 纳米TiO₂/CaCO₃-玄武岩纤维复合改性沥青的路用性能[J]. 吉林大学学报(工学版), 2018, 48(2): 460-465.
[6]	张仰鹏, 魏海斌, 贾江坤, 陈昭. 季冻区组合冷阻层应用表现的数值评价[J]. 吉林大学学报(工学版), 2018, 48(1): 121-126.
[7]	季文玉, 李旺旺, 过民龙, 王珏. 预应力RPC-NC叠合梁挠度试验及计算方法[J]. 吉林大学学报(工学版), 2018, 48(1): 129-136.
[8]	马晔, 尼颖升, 徐栋, 刁波. 基于空间网格模型分析的体外预应力加固[J]. 吉林大学学报(工学版), 2018, 48(1): 137-147.
[9]	罗蓉, 曾哲, 张德润, 冯光乐, 董华均. 基于插板法膜压力模型的沥青混合料水稳定性评价[J]. 吉林大学学报(工学版), 2017, 47(6): 1753-1759.
[10]	尼颖升, 马晔, 徐栋, 李金凯. 波纹钢腹板斜拉桥剪力滞效应空间网格分析方法[J]. 吉林大学学报(工学版), 2017, 47(5): 1453-1464.
[11]	郑传峰, 马壮, 郭学东, 张婷, 吕丹, 秦泳. 矿粉宏细观特征耦合对沥青胶浆低温性能的影响[J]. 吉林大学学报(工学版), 2017, 47(5): 1465-1471.
[12]	于天来, 郑彬双, 李海生, 唐泽睿, 赵云鹏. 钢塑复合筋带挡土墙病害及成因[J]. 吉林大学学报(工学版), 2017, 47(4): 1082-1093.
[13]	蔡氧, 付伟, 陶泽峰, 陈康为. 基于扩展有限元模型的土工布防荷载型反射裂缝影响分析[J]. 吉林大学学报(工学版), 2017, 47(3): 765-770.
[14]	王海玮, 温惠英, 刘敏. 夜间环境驾驶员精神负荷的生理特性评估与实验[J]. 吉林大学学报(工学版), 2017, 47(2): 420-428.
[15]	刘寒冰, 张互助, 王静. 失水干燥对路基压实黏质土抗剪强度特性的影响[J]. 吉林大学学报(工学版), 2017, 47(2): 446-451.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于马尔科夫毯和隐朴素贝叶斯的驾驶行为险态辨识

Driving risk status identification based on Markov blanket hidden Naive Bayes

RICH HTML

PDF (PC)

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