Traffic Accident Severity Prediction Research and Application Based on Ensemble Learning

doi:10.19678/j.issn.1000-3428.0067241

Abstract

Abstract:

At present, autonomous driving technology focuses on how to proactively avoid collisions. However, when it comes to scenarios of inevitable collisions caused by the intrusion of other traffic participants, few studies have explored how to reduce the severity of the accidents by predicting such severity under different vehicle driving modes. To address this issue, a two-layer stacking accident severity prediction model is proposed; it was trained on the NASS-CDS dataset, which is composed of real-world accident data. The model takes accident-related features that can be perceived by vehicle sensors as input and outputs the highest level of injury to passengers. In the first layer, different learner combinations are trained through experiments, and K-Nearest Neighbor(KNN), adaptive boosting, and extreme gradient boosting are eventually selected as base learners considering both predictive performance and algorithm time consumption. In the second layer, logistic regression is selected as a meta learner to reduce overfitting. Experimental results show that the classification accuracy of the proposed model reaches 85.01%, which is superior to other individual and integrated models in terms of precision, recall, and F1 value. The predicted results can be used as a priori information for the decision-planning module of intelligent vehicles to help vehicles make correct decisions and mitigate accident damage. Finally, the application of the model in L₂ assisted driving and L₄ autonomous driving vehicles is reported in detail. The proposed approach further improves the safety of vehicles based on conventional protection.

Key words: traffic safety, traffic accident severity prediction, intelligent vehicle, ensemble learning, K-Nearest Neighbor(KNN), Adaptive Boosting(AdaBoost) tree, eXtreme Gradient Boosting(XGBoost) tree, logistic regression

摘要：

目前自动驾驶技术重点是关注如何主动避免碰撞，然而在面对其他交通参与者入侵而导致不可避免的碰撞事故场景时，预测车辆在不同行驶模式下的碰撞严重程度来降低事故严重程度的研究却很少。为此，提出一种双层Stacking事故严重程度预测模型。基于真实交通事故数据集NASS-CDS完成训练，模型输入为车辆传感器可感知得到的事故相关特征，输出为车内乘员最高受伤级别。在第1层中，通过实验对不同学习器组合进行训练，最终综合考虑预测性能以及耗时挑选K近邻、自适应提升树、极度梯度提升树作为基学习器；在第2层中，为降低过拟合，采用逻辑回归作为元学习器。实验结果表明，该方法准确率达到85.01%，在精确率、召回率和F1值方面优于其他个体模型和集成模型，该预测结果可作为智能车辆决策规划模块先验信息，帮助车辆做出正确的决策，减缓事故损害。最后阐述了模型在L₂辅助驾驶与L₄自动驾驶车辆中的应用，在常规车辆安全防护的基础上进一步提升车辆的安全性。

关键词: 交通安全, 交通事故严重程度预测, 智能车辆, 集成学习, K近邻, 自适应提升树, 极度梯度提升树, 逻辑回归

Yonghang SHAN, Xi ZHANG, Chuan HU, Taojun DING, Yuan YAO. Traffic Accident Severity Prediction Research and Application Based on Ensemble Learning[J]. Computer Engineering, 2024, 50(2): 33-42.

单永航, 张希, 胡川, 丁涛军, 姚远. 基于集成学习的交通事故严重程度预测研究与应用[J]. 计算机工程, 2024, 50(2): 33-42.

/ / Recommend / Download Citations

URL: http://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0067241

http://www.ecice06.com/EN/Y2024/V50/I2/33

Figures/Tables 22

Fig.1 SMOTE algorithm diagram

Fig.2 Schematic drawing of cross validation

Fig.3 Implementation steps of Stacking algorithm

Fig.4 Prediction performance of each base learners combination

Fig.5 Time consumption of each base learners combination

Fig.6 Importance distribution of KNN feature

Fig.7 Importance distribution of AdaBoost feature

Fig.8 Importance distribution of XGBoost feature

Fig.9 Importance distribution of feature average

Fig.10 Structure of accident severity prediction mitigation system

Fig.11 Schematic diagram of vehicle following

Fig.12 Traffic scenario

References 27

1	WISMANS J, SKOGSMO I, NILSSON-EHLE A, et al. Commentary: status of road safety in Asia. Traffic Injury Prevention, 2016, 17(3): 217- 225. doi: 10.1080/15389588.2015.1066498
2	ZHANG J, LI Z B, PU Z Y, et al. Comparing prediction performance for crash injury severity among various machine learning and statistical methods. IEEE Access, 2018, 6, 60079- 60087. doi: 10.1109/ACCESS.2018.2874979
3	YU R J, ABDEL-ATY M. Using hierarchical Bayesian binary probit models to analyze crash injury severity on high speed facilities with real-time traffic data. Accident Analysis & Prevention, 2014, 62, 161- 167.
4	ZHANG Y, LI Z B, LIU P, et al. Exploring contributing factors to crash injury severity at freeway diverge areas using ordered probit model. Procedia Engineering, 2011, 21, 178- 185. doi: 10.1016/j.proeng.2011.11.2002
5	XIE Y C, ZHANG Y L, LIANG F M. Crash injury severity analysis using Bayesian ordered probit models. Journal of Transportation Engineering, 2009, 135(1): 18- 25. doi: 10.1061/(ASCE)0733-947X(2009)135:1(18)
6	SHANKAR V, MANNERING F, BARFIELD W. Statistical analysis of accident severity on rural freeways. Accident Analysis & Prevention, 1996, 28(3): 391- 401.
7	PEI X, WONG S C, SZE N N. A joint-probability approach to crash prediction models. Accident Analysis & Prevention, 2011, 43(3): 1160- 1166.
8	YE X, PENDYALA R M, WASHINGTON S P, et al. A simultaneous equations model of crash frequency by collision type for rural intersections. Safety Science, 2009, 47(3): 443- 452. doi: 10.1016/j.ssci.2008.06.007
9	MA J M, KOCKELMAN K M, DAMIEN P. A multivariate Poisson-lognormal regression model for prediction of crash counts by severity, using Bayesian methods. Accident Analysis & Prevention, 2008, 40(3): 964- 975.
10	ZENG Q, HUANG H L. A stable and optimized neural network model for crash injury severity prediction. Accident Analysis & Prevention, 2014, 73, 351- 358.
11	DE OÑA J, LÓPEZ G, ABELLÁN J. Extracting decision rules from police accident reports through decision trees. Accident Analysis & Prevention, 2013, 50, 1151- 1160.
12	CHEN C, ZHANG G H, QIAN Z, et al. Investigating driver injury severity patterns in rollover crashes using support vector machine models. Accident Analysis & Prevention, 2016, 90, 128- 139.
13	HARB R, YAN X D, RADWAN E, et al. Exploring precrash maneuvers using classification trees and random forests. Accident Analysis & Prevention, 2009, 41(1): 98- 107.
14	MAURO R, DE LUCA M, DELL'ACQUA G. Using a K-means clustering algorithm to examine patterns of vehicle crashes in before-after analysis. Modern Applied Science, 2013, 7(10): 11- 22.
15	MENAHEM E, ROKACH L, ELOVICI Y. Troika: an improved stacking schema for classification tasks. Information Sciences, 2009, 179(24): 4097- 4122. doi: 10.1016/j.ins.2009.08.025
16	SCHAFFER C. Cross-validation, stacking and Bi-level stacking: meta-methods for classification learning. New York, USA: ACM Press, 1994.
17	BERG F A, WALZ F, MUSER M, et al. Implications of velocity change delta-v and energy equivalent speed ees for injury mechanism assessment in various collision configurations[C]//Proceedings of IRCOBIʼ98. Washington D. C., USA: IEEE Press, 1998: 1682-595.
18	JI A, LEVINSON D. An energy loss-based vehicular injury severity model. Accident Analysis & Prevention, 2020, 146, 105730.
19	NHTSA Research & Data. National Automotive Sampling Syatem(NASS)[EB/OL]. [2023-02-10]. https://www.nhtsa.gov/research-data/national-automotive-sampling-system-nass.
20	RAGLAND C, DALRYMPLE G. Overlap car-to-car tests compared to car-to-half barrier and car-to-full barrier tests. Auto & Traffic Safety, 1991, 1(2): 213- 224. URL
21	孙轶轩, 邵春福, 岳昊, 等. 基于SVM灵敏度的城市交通事故严重程度影响因素分析. 吉林大学学报(工学版), 2014, 44(5): 1315- 1320.
	SUN Y X, SHAO C F, YUE H, et al. Urban traffic accident severity analysis based on sensitivity analysis of support vector machine. Journal of Jilin University(Engineering and Technology Edition), 2014, 44(5): 1315- 1320.
22	贺小娟, 潘文捷, 程宏. 基于集成学习方法的点击率预估模型研究. 计算机工程与科学, 2019, 41(12): 2278- 2284.
	HE X J, PAN W J, CHENG H. An advertisement click-through rate prediction model based on ensemble learning. Computer Engineering & Science, 2019, 41(12): 2278- 2284.
23	周星, 丁立新, 万润泽, 等. 分类器集成算法研究. 武汉大学学报(理学版), 2015, 61(6): 503- 508.
	ZHOU X, DING L X, WAN R Z, et al. Research on classifier ensemble algorithms. Journal of Wuhan University (Natural Science Edition), 2015, 61(6): 503- 508.
24	SHAH A H, AFZAL M M. Voting classification method for network traffic prediction. International Journal of Innovative Technology and Exploring Engineering, 2020, 9(7): 510- 515. doi: 10.35940/ijitee.E3104.059720
25	WOLPERT D H. Stacked generalization. Neural Networks, 1992, 5(2): 241- 259. doi: 10.1016/S0893-6080(05)80023-1
26	CUI S Z, YIN Y Q, WANG D J, et al. A stacking-based ensemble learning method for earthquake casualty prediction. Applied Soft Computing, 2021, 101, 107038. doi: 10.1016/j.asoc.2020.107038
27	SRIMANEEKARN N, HAYTER A, LIU W, et al. Binary response analysis using logistic regression in dentistry. International Journal of Dentistry, 2022, 22, 5358602.

[1]	Qingsong AI, Haozhe ZHANG, Junwei YAN. Abnormal Driving Behavior Detection Method Based on Adaptive Gamma Correction [J]. Computer Engineering, 2023, 49(9): 279-286.
[2]	Shanshu BAO, Bo CHE, Linhong DENG. Lung Sound Signal Recognition Based on Dual-Source Domain Transfer Learning [J]. Computer Engineering, 2023, 49(9): 295-302, 312.
[3]	WANG Qihan, PANG Jianmin, YUE Feng, ZHU Di, SHEN Li, XIAO Qian. Implementation and Optimization of Parallel KNN Algorithm for Sunway Architecture [J]. Computer Engineering, 2023, 49(5): 286-294.
[4]	YANG Yuejia, HUA Bei, ZHONG Zhiwei, GAO Mi. Collaborative Computing of Privacy-Preserving Logistic Regression Based on Homomorphic Encryption [J]. Computer Engineering, 2023, 49(4): 23-31.
[5]	TANG Min, ZHANG Yuhao, DENG Guoqiang. An Efficient Non-Interactive and Privacy-Preserving Logistic Regression Model [J]. Computer Engineering, 2023, 49(4): 32-42,51.
[6]	AI Chenghao, GAO Jianhua, HUANG Zijie. Code Smell Detection Driven by Hybrid Feature Selection and Ensemble Learning [J]. Computer Engineering, 2022, 48(7): 168-176,198.
[7]	WANG Mengduo, XU Xinying, YAN Gaowei, SHI Lijuan, GUO Lei. Imbalanced Data Classification Based on Ensemble Weighted Broad Learning System with AdaBoost [J]. Computer Engineering, 2022, 48(4): 99-105,112.
[8]	ZHANG Yao, MA Yingcang, ZHU Hengdong, LI Heng, CHEN Cheng. Multi-label Feature Selection Combining Manifold Learning and Logistic Regression [J]. Computer Engineering, 2022, 48(3): 90-99,106.
[9]	LIU Ziwei, LUO Xi, LI Ke, CHEN Fuqiang. Multi-Scale Hypersphere Convolutional Neural Network Learning Based on k-Nearest Neighbor [J]. Computer Engineering, 2022, 48(11): 111-119.
[10]	WU Jiao, HONG Caifeng, GU Yongchun, GU Xingquan, JIN Shiju. Linear Regression Text Classification Based on Class-wise Nearest Neighbor Dictionary [J]. Computer Engineering, 2021, 47(8): 93-99,108.
[11]	LIANG Xiaomeng, YAN Ming, WU Jie. Tor Traffic Online Identification Method Based on Artificial Bee Colony Algorithm [J]. Computer Engineering, 2021, 47(11): 129-135,143.
[12]	BAO Yintu, LIU Wei, NIU Chaoyang, LI Runsheng, ZHANG Haobo. Scene Classification of Optical Remote Sensing Images Joint Ensemble Learning and EfficientNet [J]. Computer Engineering, 2021, 47(10): 226-235.
[13]	WANG Xu, CHEN Yongle, WANG Qingsheng, CHEN Junjie. Cryptosystem Identification Scheme Combining Feature Selection and Ensemble Learning [J]. Computer Engineering, 2021, 47(1): 139-145,153.
[14]	LI Peiyuan, HUANG Chi. Classification Method for Human Protein Images Based on CSPPNet and Ensemble Learning [J]. Computer Engineering, 2020, 46(8): 235-242.
[15]	ZHOU Sheng, LIU Sanmin. Research on Multi-Source Transfer Learning Data Streams Classification Based on Dynamic Strategy [J]. Computer Engineering, 2020, 46(5): 139-143,149.

Please choose a citation manager

Content to export