计算机科学

计算机科学 ›› 2024, Vol. 51 ›› Issue (6A): 230500118-10.doi: 10.11896/jsjkx.230500118

• 交叉&应用 • 上一篇    下一篇

动态路网下城市交通事故风险预测模型研究与实现

董婉青1, 赵子榕2, 廖惠敏3, 肖晖4, 张晓亮4   

  1. 1 北京市交通运输综合执法总队 北京 100044
    2 北京小马智行科技有限公司 北京 100094
    3 北京市交通运输综合执法总队执法保障中心 北京 100044
    4 中路高科交通科技集团有限公司 北京 100088
  • 发布日期:2024-06-06
  • 通讯作者: 张晓亮(zhangxiaoliang@hstg.com.cn)
  • 作者简介:(dongwanqing@jtw.beijing.gov.cn)
  • 基金资助:
    北京市交通行业科技项目(0686-2241B1251414Z)

Research and Implementation of Urban Traffic Accident Risk Prediction in Dynamic Road Network

DONG Wanqing1, ZHAO Zirong2, LIAO Huimin3, XIAO Hui4, ZHANG Xiaoliang4   

  1. 1 Beijing Transportation Comprehensive Enforcement Corps,Beijing 100044,China
    2 Beijing Pony.Ai Science and Technology Co.,Ltd.,Beijing 100094,China
    3 Support Center of Beijing Transportation Comprehensive Enforcement Corps,Beijing 100044,China
    4 RIOH High Science and Technology Group,Ltd.,Beijing 100088,China
  • Published:2024-06-06
  • About author:DONG Wanqing,born in 1989,master,enior engineer.Her main research interests include intelligent transportation and big data.
    ZHANG Xiaoliang,born in 1983,Ph.D.Her main research interests include intelligent transportation technology,transportation big data mining and applications.
  • Supported by:
    Beijing Transportation Industry Science and Technology Project(0686-2241B1251414Z).

PDF (PC)

329

摘要： 通过图卷积神经网络对交通事故进行风险预测是交通领域的研究热点。然而,现有的使用图卷积神经网络对交通事故进行风险预测的研究存在着缺乏语义邻接性的构造、无法进行图权重的自适应学习的问题。针对以上问题,文中基于多源交通大数据,构建了数据驱动的多粒度、多视角的时空拓扑图,实现了交通网络中时空关联性和依赖性的精准建模。图上的结点从时间和空间两个维度对路段结点的交通状态进行综合描述,边则从地理邻接性和语义邻接性两个视角表现了路段之间的抽象邻接关系。在时空拓扑图的基础上,文中设计了基于动态时空图网络的交通事故风险预测模型,实现了路段级交通事故风险的准确预测。该模型引入了具有多头注意力机制的空间图网络层对空间关联性进行学习,同时采用了基于一维扩张卷积的时间学习单元捕获短时依赖性与长时周期性。在北京地区的实际交通数据集上进行大规模实验,所提方法的召回率达到0.899,F1-Score达到0.860,其他指标与主流方法相比也均有所提升。

关键词: 交通事故风险预测, 图神经网络, 时空数据挖掘

Abstract: Accident risk prediction of traffic accidents through graph convolution networks is a research hotspot in the transportation field.However,the existing researches on using graph convolution networks for accident risk prediction lack semantic adjacency in graph construction and unable to perform adaptive learning of graph weights.To address these problems,a data-driven,multi-granularity and multi-view spatio-temporal topology graph is constructed based on multi-source traffic big data to realize the accurate modeling of spatio-temporal correlation and dependency in traffic network.The nodes on the graph provide a comprehensive description of the traffic state from time and space two dimensions,while the edges show the abstract adjacency relationship between roadways from geography and semantics two perspectives.Then,a dynamic spatio-temporal graph network based on the spatio-temporal topology graph is designed to achieve accurate prediction of roadway-level traffic accident risk.The model introduces spatial graph network layers with multi-headed attention mechanism to learn spatial correlations,while temporal learning units based on 1-D dilated convolution are used to capture short-time dependencies and long-time periodicity.According to large-scale experiments carried out on real traffic data in Beijing area,our method achieves the recall of 0.899 and the F-1 Score of 0.860.Meanwhile,there are also improvements in other indicators comparing to mainstream methods.

Key words: Traffic accident risk prediction, Graph neural network, Spatio-Temporal data mining

中图分类号: 

  • TP391
[1]CHEN Q,SONG X,YAMADA H,et al.Learning deep representation from big and heterogeneous data for traffic accident inference[C]//AAAI Conference on Artificial Intelligence,2016:338-344.
[2]CHEN C,FAN X,ZHENG C,et al.SDCAE:stack denoisingconvolutional autoencoder model for accident risk prediction via traffic big data[C]//International Conference on Advanced Cloud and Big Data.2018:328-333.
[3]REN H.A deep learning approach to the citywide traffic accident risk prediction[C]//2018 21st International Conference on Intelligent Transportation Systems(ITSC).Maui,United States,2018:3346-3351.
[4]BAO J,LIU P,UKKUSURI S V.A spatiotemporal deep learning approach for citywide shortterm crash risk prediction with multi-source data[J].Accident Analysis & Prevention,2019,122:239-254.
[5]SCARSELLI F,GORI M,TSOI A C,et al.The graph neural network model[J].IEEE Transactions on Neural Networks,2008,20(1):61-80.
[6]ZHOU Z,WANG Y,XIE X,et al.RiskOracle:a minutelevel citywide traffic accident forecasting framework[C]//AAAI Conference on Artificial Intelligence.2020.
[7]YU L,DU B,HU X,et al.Deep spatio-temporal graph convolutional network for traffic accident prediction[J].Neurocomputing,2021,423:135-147.
[8]KIPF T N,WELLING M.Semi-supervised classification withgraph convolutional networks[J].arXiv:1609.02907,2016.
[9]MIKOLOV T,KARAFI Á T M,BURGET L,et al.Recurrent neural network based language model[C]//Interspeech.2010:1045-1048.
[10]BAI S,KOLTER J Z,KOLTUN V.An empirical evaluation of generic convolutional and recurrent networks for sequence modeling[J].arXiv:1803.01271,2018.
[11]VASWANI A,SHAZEER N,PARMAR N,et al.Attention isall you need[J].Advances in Neural Information Processing Systems,2017,30.
[12]HOCHREITER S,SCHMIDHUBER J.Long short-term memory[J].Neural Computation,1997,9(8):1735-1780.
[13]OORD A,DIELEMAN S,ZEN H,et al.Wavenet:A generative model for raw audio[J].arXiv:1609.03499,2016.
[14]BOROVYKH A,BOHTE S,OOSTERLEEC W.Conditionaltime series forecasting with convolutional neural networks[J].arXiv:1703.04691,2017.
[15]DEVLIN J,CHANG M W,LEE K,et al.Bert:Pre-training of deep bidirectional transformers for language understanding[J].arXiv:1810.04805,2018
[16]LI Y,YU R,SHAHABI C,et al.Diffusion convolutional recurrent neural network:Data-driven traffic forecasting[J].arXiv:1707.01926,2017.
[17]YU B,YIN H,ZHU Z.Spatio-temporal graph convolutional networks:A deep learning framework for traffic forecasting[J].arXiv:1709.04875,2017.
[18]CALIENDO C,GUIDA M,PARISI A.A crash-prediction model for multilane roads[J].Accident Analysis & Prevention,2007,39(4):657-670.
[19]BERGEL-HAYAT R,DEBBARH M,ANTONIOU C,et al.Explaining the road accident risk:Weather effects[J].Accident Analysis & Prevention,2013,60:456-465.
[20]EISENBERG D.The mixed effects of precipitation on trafficcrashes[J].Accident Analysis & Prevention,2004,36(4):637-647.
[21]VELIČKOVIĆ P,CUCURULL G,CASANOVA A,et al.Graph attention networks[J].arXiv:1710.10903,2017.
[22]WALKER S H,DUNCAN D B.Estimation of the probability of an event as a function of several independent variables[J].Biometrika,1967,54(1/2):167-179.
[23]CORTES C,VAPNIK V.Support-vector networks[J].Machine Learning,1995,20(3):273-297.
[24]QUINLANJ R.Induction of decision trees[J].Machine Lear-ning,1986,1(1):81-106.
[25]CHEN T,GUESTRIN C.Xgboost:A scalable tree boosting system[C]//Proceedings of the 22nd ACM Sigkdd International Conference on Knowledge Discovery and Data Mining.2016:785-794.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发