Study on risk model of highway traffic accidents based on LSTM-BF

doi:10.16265/j.cnki.issn1003-3033.2022.05.1602

Abstract

Abstract:

In order to reduce highway traffic accidents, a traffic risk state prediction model was studied by using LSTM neural network and BF. Firstly, time dependency existing in historical traffic flow risk data was studied through LSTM module. Then, real-time risk prediction performance was improved by BF module integrating LSTM prediction results. Finally, with accident data and traffic flow data of Ningbo roundabout in 2020 as an example, multi-step feature variables were constructed in the form of migration time window based on spatio-temporal data between upper and lower port stations within 20 min prior to accidents, and 5-fold cross-validation was carried out. The results show that the precision and recall rate of LSTM model are higher than that of random forest (RF) algorithm, and F₁ score of the final prediction result is close to 0.80 by adding BF module.

Key words: short and long term memory (LSTM), bayesian filter (BF), highway, traffic accident, risk prediction, traffic flow

XIONG Xiaoxia, LIU Qingchao, SHEN Yujie, CAI Yingfeng, CHEN Long. Study on risk model of highway traffic accidents based on LSTM-BF[J]. China Safety Science Journal, 2022, 32(5): 170-176.

Figures/Tables 9

Fig.1

Tab.1

Fig.2

Fig.3

Tab.2

Fig.4

Tab.3

Fig.5

Fig.6

References 18

[1]	黄西子, 黄淑萍. 高速道路追尾感知PC-Crash仿真与统计建模[J]. 中国安全科学学报, 2020, 30(5):143-148. doi: 10.16265/j.cnki.issn1003-3033.2020.05.022
	HUANG Xizi, HUANG Shuping. Freeway rear-end perception model based on PC-Crash simulation and statistical analysis[J]. China Safety Science Journal, 2020, 30(5):143-148. doi: 10.16265/j.cnki.issn1003-3033.2020.05.022
[2]	HOSSAIN M, MUROMACHI Y. A Bayesian network based framework for real-time crash prediction on the basic freeway segments of urban expressways[J]. Accident Analysis Prevention, 2012, 45: 373-381. doi: 10.1016/j.aap.2011.08.004
[3]	SUN Jie, SUN Jian. A dynamic Bayesian network model for real-time crash prediction using traffic speed conditions data[J]. Transportation Research Part C, 2015, 54:176-186. doi: 10.1016/j.trc.2015.03.006
[4]	ZHAI Ben, LU Jiantao, WANG Yanli, et al. Real-time prediction of crash risk on freeways under fog conditions[J]. International Journal of Transportation Science and Technology, 2020, 9(4):287-298. doi: 10.1016/j.ijtst.2020.02.001
[5]	LYU Yisheng, TANG Shuning, ZHAO Hongxia. Real-Time Highway traffic accident prediction based on the k-nearest neighbor method[C]. International Conference on Measuring Technology & Mechatronics Automation, 2009: 547-550.
[6]	WEN Huiying, LUO Jun, CHEN Xiaolong, et al. Real-time highway accident prediction based on grey relation entropy analysis and probabilistic neural network[C]. 2011 International Conference on Electric Technology and Civil Engineering, 2011: 1420-1423.
[7]	张军, 胡震波, 朱新山. 基于 AdaBoost 分类器的实时交通事故预测[J]. 计算机应用, 2017, 37(1): 284-288. doi: 10.11772/j.issn.1001-9081.2017.01.0284
	ZHANG Jun, HU Zhenbo, ZHU Xinshan. Real-time traffic accident prediction based on AdaBoost classifier[J]. Journal of Computer Applications, 2017, 37(1): 284-288. doi: 10.11772/j.issn.1001-9081.2017.01.0284
[8]	付存勇, 王俊骅. 基于监控数据的高速公路实时事故风险模型[J]. 交通信息与安全, 2017, 35(5):11-17.
	FU Cunyong, WANG Junhua. A rea-time accident risk model on freeways based on monitoring data[J]. Journal of Transport Information and Safety, 2017, 35(5):11-17.
[9]	QU Xu, WANG Wei, WANG Wenfu, et al. Real-time rear-end crash potential prediction on freeways[J]. Journal of Central South University, 2017, 24(11):2664-2673. doi: 10.1007/s11771-017-3679-2
[10]	YANG Kui, ZHAO Wenjing, ANTONIOU C. Utilizing import vector machines to identify dangerous pro-active traffic conditions[C]. IEEE 23^rd International Conference on Intelligent Transportation Systems, 2020:1-6.
[11]	CHEN Quanjun, SONG Xuan, YAMADA H. Learning deep representation from big and heterogeneous data for traffic accident inference[C]. 30^th AAAI Conference on Artificial Intelligence, 2016:338-344.
[12]	SAMEEN M, PRADHAN B. Severity prediction of traffic accidents with recurrent neural networks[J]. Applied Sciences, 2017, 7(6): DOI: 10.3390/app7060476. doi: 10.3390/app7060476
[13]	YU Le, DU Bowen, HU Xiao, et al. Deep spatio-temporal graph convolutional network for traffic accident prediction[J]. Neurocomputing, 2021, 423:135-147. doi: 10.1016/j.neucom.2020.09.043
[14]	ZHENG Haifeng, LIN Feng, FENG Xinxin, et al. A hybrid deep learning model with attention-based conv-lstm networks for short-term traffic flow prediction[J]. IEEE Transactions on Intelligent Transportation Systems, 2020, 22(11):6910-6920. doi: 10.1109/TITS.2020.2997352
[15]	戴健非, 杨鹏, 诸利一, 等. 集成PCA和LSTM神经网络的浸润线预测方法[J]. 中国安全科学学报, 2020, 30(3): 94-101. doi: 10.16265/j.cnki.issn1003-3033.2020.03.015
	DAI Jianfei, YANG Peng, ZHU Liyi, et al. A PCA-LSTM neural network-integrated method for phreatic line prediction[J]. China Safety Science Journal, 2020, 30(3): 94-101. doi: 10.16265/j.cnki.issn1003-3033.2020.03.015
[16]	AOUDE G S, DESARAJU V R, STEPHENS L H, et al. Behavior classification algorithms at intersections and validation using naturalistic data[J]. IEEE Transactions on Intelligent Transportation Systems, 2012, 13(2):724-736. doi: 10.1109/TITS.2011.2179537
[17]	李振明, 牛毅, 樊运晓, 等. 不同区域高速公路货车事故特征研究[J]. 中国安全科学学报, 2020, 30(6): 121-127. doi: 10.16265/j.cnki.issn1003-3033.2020.06.018
	LI Zhenming, NIU Yi, FAN Yunxiao, et al. Research on characteristics of expressway truck accidents in different regions[J]. China Safety Science Journal, 2020, 30(6): 121-127. doi: 10.16265/j.cnki.issn1003-3033.2020.06.018
[18]	ARIANNEZHAD A, KARIMPOUR A, QIN Xiao, et al. Handling imbalanced data for real-time crash prediction: application of boosting and sampling techniques[J]. Journal of Transportation Engineering, Part A: Systems, 2021, 147(3): DOI: 10.1061/JTEPBS.0000499. doi: 10.1061/JTEPBS.0000499

路段		上游路段				下游路段			事故数(2020-07-02— 2020-11-10)
路段		上游车道数		上游卡口日平均交通量		下游车道数		下游卡口日平均交通量	事故数(2020-07-02— 2020-11-10)
K0+900九向-K8+200九向		3		22 739		3		24 861	7.30	30
K1+600姜向-K82+375姜向		3		25 806		3		28 385	4.73	48
K10+100姜向-K1+600姜向	3		23459		3		25806		8.50	47
K11+400姜向-K10+100姜向	2		20 029		3		23 459		1.30	5
K79+420九向-K82+500九向	4		47 062		4		31 749		3.10	11
K8+200九向-K11+400九向	3		24 861		2		19 047		3.20	16
K82+500九向-K0+900九向	4		31 749		3		22 739		3.90	26
总计	—									183

变量名	变量描述
Vol_up	5 min内上游卡口流量/veh
Vol_down	5 min内下游卡口流量/veh
Diff_Lane_Flow_up	5 min内上游卡口第1和第2车道流量差/veh
Diff_Lane_Flow_down	5 min内下游卡口第1和第2车道流量差/veh
Len_up_down	上下游卡口的距离/km
Lane_Num_up	上游卡口车道数
Lane_Num_down	下游卡口车道数
Dens_Road_up	上游卡口交通流平均密度/(veh·km^-1)
Dens_Road_down	下游卡口交通流平均密度/(veh·km^-1)
Dens_Lane_up	上游卡口车道平均流量/veh
Dens_Lane_down	下游卡口车道平均流量/veh
Largeveh_Ratio_up	上游卡口大型/小型车流量比
Largeveh_Ratio_down	下游卡口大型/小型车流量比
Time of Day	一天中的时间(日时间)
Day of Week	一周中的时间(周时间)

Fold编号	事故样本	非事故样本	样本比
Fold1	170	170	1∶1
Fold2	170	170	1∶1
Fold3	170	170	1∶1
Fold4	170	170	1∶1
Fold5	175	175	1∶1
合计	855	855	1∶1