考虑避险动态的行人电动自行车冲突风险评估

王琛; 王帅; 肖长江; 段松志; 钟京洋

doi:10.13637/j.issn.1009-6094.2025.083810.13637/j.issn.1009-6094.2025.0838

Chinese

您当前的位置：

首页 >

文章列表页 >

考虑避险动态的行人电动自行车冲突风险评估

安全评价安全工程 | 更新时间：2026-04-17

- 考虑避险动态的行人电动自行车冲突风险评估
- 考虑避险动态的行人电动自行车冲突风险评估
- 安全与环境学报 2026年26卷第4期页码：1245-1254
- 作者机构：
  
  长安大学工程机械学院
- 作者简介：
- 基金信息：
- DOI：10.13637/j.issn.1009-6094.2025.083810.13637/j.issn.1009-6094.2025.0838
  中图分类号： U491.225;U491.31
- 纸质出版：2026
- 稿件说明：
移动端阅览
[1]王琛,王帅,肖长江,等.考虑避险动态的行人电动自行车冲突风险评估[J].安全与环境学报,2026,26(04):1245-1254.
[1]王琛,王帅,肖长江,等.考虑避险动态的行人电动自行车冲突风险评估[J].安全与环境学报,2026,26(04):1245-1254. DOI： 10.13637/j.issn.1009-6094.2025.083810.13637/j.issn.1009-6094.2025.0838.

DOI：10.13637/j.issn.1009-6094.2025.0838.

摘要

针对城市非机动车道中行人横穿行为频发引发的交通冲突问题，提出了一种融合行人避险行为动态特征与运动学指标的冲突严重性评估方法。通过无人机与地面摄像设备采集的10 h视频数据，识别并提取843组行人与电动自行车冲突事件。根据避险反应时机将冲突划分为提前、紧急与无避险三类，设计基于人体关键点角度与角速度特征的身体变化幅度(Body Change Magnitude

BCM)指标，量化行人行为反应强度。结合后侵入时间(Post Encroachment Time

PET)与安全减速度(Deceleration to Safety Time

DST)构建三维冲突评估指标体系，并采用改进K均值(K-means)聚类算法和反向传播(Back Propagation

BP)神经网络模型实现冲突等级识别与预测。结果表明：紧急避险冲突中严重等级占比达50.4%

BCM与相对速度是影响冲突严重性的主要因素，重要性贡献分别达92.8%与90.3%;模型预测准确率为81.7%。研究成果可为城市混行交通环境下高风险行为的识别与事故预防提供方法支持与实用依据。

Abstract

This study introduces a novel method for assessing the severity of conflicts between pedestrians and e-bicycles by integrating dynamic behavioral indicators with traditional motion-based safety metrics. A total of 843 conflict events were identified from 10 hours of aerial and ground video footage collected across three urban road segments in Xi'an

China. Based on the timing of pedestrian evasive responses

the conflicts were classified into three categories: early evasion

emergency evasion

and no evasion. To quantify the intensity of pedestrian movements during conflicts

a keypoint-based Body Change Magnitude(BCM) index was developed. The BCM integrates angula

r displacement and angular velocity derived from 17 human body keypoints detected using the Yolov8n-pose model. A dynamic weighting scheme

based on sigmoid functions

was employed to merge these two components. Additionally

the Post Encroachment Time(PET) and Deceleration to Safety Time(DST) were calculated using trajectory and speed data for both pedestrians and e-bikes. Together

these three indicators formed the input features for risk assessment. A modified K-means clustering algorithm—utilizing probabilistic initialization and Euclidean distance minimization—was employed to classify conflict severity into three levels: minor

moderate

and severe. Subsequently

a Back Propagation Neural Network(BPNN) with one hidden layer containing four neurons and a sigmoid activation function was used to predict these severity levels. The model was trained on 80% of the samples and tested on the remaining 20%

successfully converging within 1 000 iterations at a learning rate of 0.1. The proposed model achieved an overall prediction accuracy of 81.7%

with the moderate severity category attaining a precision of 82.43%. Emergency evasion cases demonstrated the highest average BCM at 0.622 rad

the shortest PET at 1.440 s

and the highest DST at 5.310 m/s

; notably

50.4% of these cases were classified as severe. Importance analysis indicated that BCM and relative velocity were the two most influential factors

with contribution rates of 92.8% and 90.3%

respectively. The primary innovation of this study is the introduction of BCM—a dynamic behavior indicator that effectively captures both the intensity and timing of evasive actions—thereby enhancing conflict severity prediction and providing a new perspective for pedestrian risk assessment in complex urban traffic environments.

关键词

Keywords

references

李英帅，张旭，王卫杰，等.基于随机森林的电动自行车骑行者事故伤害程度影响因素分析[J].交通运输系统工程与信息，2021,21(1):196-200.LI Y S,ZHANG X,WANG W J,et al.Factors affecting electric bicycle rider injury in accident based on random forest model[J].Journal of Transportation Systems Engineering and Information Technology,2021,21(1):196-200.

何霞，张道文，车瑶栎，等.基于行人横穿场景的AEB触发宽度优化研究[J].中国安全科学学报，2020,30(4):141-146.HE X,ZHANG D W,CHE Y L,et al.Optimization of AEB trigger width based on pedestrian crossing scenario[J].China Safety Science Journal,2020,30(4):141-146.

张名芳，马艳华，马勇.无信号交叉口人车冲突严重程度影响因素分析[J].中国安全科学学报，2023,33(8):190-197.ZHANG M F,MA Y H,MA Y.Influencing factors on severity of vehicle-pedestrian conflict at unsignalized intersections[J].China Safety Science Journal,2023,33(8):190-197.

VIGNALI V,CUPPI F,ACERRA E,et al.Effects of median refuge island and flashing vertical sign on conspicuity and safety of unsignalized crosswalks[J].Transportation Research Part F:Traffic Psychology and Behaviour,2019,60:427-439.

常玉林，吴照允，孙超，等.无信控人行横道多类型冲突严重程度模型[J].交通运输系统工程与信息，2022,22(5):205-214.CHANG Y L,WU Z Y,SUN C,et al.Multi-type conflict severity model for uncontrolled pedestrian crosswalks[J].Journal of Transportation Systems Engineering and Information Technology,2022,22(5):205-214.

王伟杰.礼让行人背景下机动车让行行为研究及三方演化博弈分析[D].西安：长安大学，2022.WANG W J.Research on vehicle yield behavior and tripartite evolutionary game analysis under the background of yield to pedestrians[D].Xi'an:Chang'an University,2022.

JIANG R X,ZHU S Y,CHANG H G,et al.Determining an improved traffic conflict indicator for highway safety estimation based on vehicle trajectory data[J].Sustainability,2021,13(16):9278.

ZHENG L,SAYED T,ESSA M.Validating the bivariate extreme value modeling approach for road safety estimation with different traffic conflict indicators[J].Accident Analysis and Prevention,2019,123:314-323.

张开冉，邱谦谦.平面交叉口电动自行车交通冲突模型研究[J].安全与环境学报，2017,17(2):422-425.ZHANG K R,QIU Q Q.Innovative traffic jam-dissolving model for the motor-bicycles in the cross-road sections[J].Journal of Safety and Environment,2017,17(2):422-425.

ZHENG L,WEN C,GUO Y Y,et al.Investigating consecutive conflicts of pedestrian crossing at unsignalized crosswalks using the bivariate logistic approach[J].Accident Analysis and Prevention,2021,162:106402.

曹红丽.基于车辆轨迹数据的高速公路交通事件检测及时空影响范围估计研究[D].成都：西南交通大学，2023.CAO H L.Research on highway traffic incident detection and spatio-temporal influence range estimation based on vehicle trajectory data[D].Chengdu:Southwest Jiaotong University,2023.

何炎东，张鹏，刘丽华，等.基于Logit模型的过街行人：非机动车的交通冲突模型研究[J].黑龙江工程学院学报，2025,39(3):10-14,32.HE Y D,ZHANG P,LIU L H,et al.Research on traffic conflict model of pedestrians crossing the street and non-motorized vehicles based on Logit model[J].Journal of Heilongjiang Institute of Technology,2025,39(3):10-14,32.

王磊，吕璞，林永杰.高速公路交通事故影响因素分析及伤害估计[J].中国安全科学学报，2016,26(3):86-90.WANG L,LÜ P,LIN Y J.Traffic accidents on freeways:influencing factors analysis and injury severity evaluation[J].China Safety Science Journal,2016,26(3):86-90.

彭勇，蒋沛，沙晓宇，等.行人过街人车冲突等级评价及影响分析[J].交通运输系统工程与信息，2019,19(2):175-181,188.PENG Y,JIANG P,SHA X Y,et al.Evaluation and impact analysis of pedestrian-vehicle conflict severity[J].Journal of Transportation Systems Engineering and Information Technology,2019,19(2):175-181,188.

ZHANG C B,CHEN F,WEI Y Y.Evaluation of pedestrian crossing behavior and safety at uncontrolled mid-block crosswalks with different numbers of lanes in China[J].Accident Analysis and Prevention,2019,123:263-273.

张盈盈，刘帅，邱志军，等.行人车辆冲突参数分析与安全评价[J].哈尔滨工业大学学报，2012,44(12):123-128.ZHANG Y Y,LIU S,QIU Z J,et al.Pedestrian-vehicle conflict parameter analysis and safety evaluation[J].Journal of Harbin Institute of Technology,2012,44(12):123-128.

KUMAR A,PAUL M,GHOSH I.Analysis of pedestrian conflict with right-turning vehicles at signalized intersections in India[J].Journal of Transportation Engineering,2019,145(6):04019018.

戢晓峰，耿昭师，曹瑞，等.严重冲突场景下山区双车道公路货车跟驰风险预测[J].安全与环境学报，2023,23(9):2975-2983.JI X F,GENG Z S,CAO R,et al.Real-time prediction of accident risk of truck following on two-lane mountain highway in severe conflict scenes[J].Journal of Safety and Environment,2023,23(9):2975-2983.

朱顺应，蒋若曦，王红，等.机动车交通冲突技术研究综述[J].中国公路学报，2020,33(2):15-33.ZHU S Y,JIANG R X,WANG H,et al.Review of research on traffic conflict techniques[J].China Journal of Highway and Transport,2020,33(2):15-33.

温惠英，何梓琦，李秋灵，等.高速公路货车换道冲突预测及其影响因素分析[J].吉林大学学报(工学版),2024,54(10):2827-2836.WEN H Y,HE Z Q,LI Q L,et al.Traffic conflict prediction and influencing factors analysis of truck lane change on expressway[J].Journal of Jilin University (Engineering and Technology Edition),2024,54(10):2827-2836.

王晓丽.融合统计模型与机器学习的交叉口交通冲突实时预测研究[D].哈尔滨：哈尔滨工业大学，2023.WANG X L.Research on real-time prediction of intersection traffic conflict based on fusion statistical model and machine learning[D].Harbin:Harbin Institute of Technology,2023.

王超.基于轨迹数据提取的交叉口交通冲突分析及预测研究[D].北京：北京交通大学，2022.WANG C.Research on traffic conflict analysis and prediction at intersection based on trajectory extraction[D].Beijing:Beijing Jiaotong University,2022.

浏览量

174

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

面向复杂施工环境的人员防护装备穿戴识别方法对比研究

基于YOLOv10n-knot的电力安全带挂点检测算法

改进型金字塔解析网络在矿区浮石识别中的应用研究

沿空留巷采空区注氮防灭火参数优化

金属粉体爆炸火焰发射光谱与爆炸压力耦合研究