基于信息熵的非机动车超越轨迹分段方法

张蕊; 王子轩; 孔令争; 侯先磊

doi:10.3963/j.jssn.1674-4861.2024.02.012

基于信息熵的非机动车超越轨迹分段方法

doi: 10.3963/j.jssn.1674-4861.2024.02.012

张蕊^{1, 2, ,},
王子轩^{1, 3},
孔令争^{1, 4},
侯先磊^{1, 2}

1.
北京建筑大学土木与交通工程学院北京 100044
2.
北京建筑大学首都世界城市顺畅交通协同创新中心北京 100044
3.
北京建筑大学北京市城市交通基础设施建设工程技术研究中心北京 100044
4.
中冶京诚工程技术有限公司北京 100176

基金项目:

国家重点研发计划项目 2022YFB2601900

详细信息

通讯作者:
张蕊(1971—)，博士，教授. 研究方向：交通规划、行人交通等. E-mail：zhangrui@bucea.edu.cn

中图分类号: U491
计量
- 文章访问数: 43
- HTML全文浏览量: 26
- PDF下载量: 0
- 被引次数: 0
出版历程
- 收稿日期: 2023-04-21
- 网络出版日期: 2024-09-14

Segmentation of Overtaking Trajectories for Non-motor Vehicles Based on Information Entropy

ZHANG Rui^{1, 2
, ,},
WANG Zixuan^{1, 3},
KONG Lingzheng^{1, 4},
HOU Xianlei^{1, 2}

1.
School of Civil and Transportation Engineering, Beijing University of Civil Engineering and Architecture, Beijing 100044, China
2.
Beijing Collaborative Innovation Center for Metropolitan Transportation, Beijing University of Civil Engineering and Architecture, Beijing 100044, China
3.
Beijing Urban Transportation Infrastructure Engineering and Architecture, Beijing University of Civil Engineering and Architecture, Beijing 100044, China
4.
MCC Capital Engineering & Research Incorporation Limited, Beijing, 100176, China

摘要

摘要: 通过自行车轨迹识别超越行为是评价非机动车交通服务水平的重要工作之一。针对基于阈值分段方法中需对不同轨迹确定不同的阈值问题，引入信息熵对非机动车超越轨迹进行分段。根据实测视频提取了780条非机动车超越轨迹数据，包括了在视频中可能存在的11种超越轨迹情形，并通过对超越过程中各阶段的特征参数分析，最终选取横向加速度、横向偏移距离、偏移角度作为基于信息熵分段的特征参数，通过引入信息熵理论，提出基于信息熵的非机动车超越轨迹分段方法和分段判断条件。根据信息熵理论中，分段后的2段子轨迹中的特征参数概率密度相较分割前更接近时熵增的定律，同时考虑非机动车超越轨迹的特征参数特征，提出适用于非机动车超越轨迹的信息熵分段标准。以实测路段非机动车超越轨迹数据为实验样本，将基于各特征参数的信息熵分段结果与基于时间、速度阈值的分段结果分别带入K最近邻（K-nearest neighbor，KNN）分类算法中进行超越轨迹识别，并利用轨迹覆盖度指标评价不同分段方法的超越轨迹分段效果。实验结果表明：基于信息熵超越轨迹分段方法的超越轨迹覆盖度平均为83.0%，优于基于阈值分段方法的轨迹覆盖度平均值79.7%，且基于横向加速度信息熵分段法的平均轨迹覆盖度为85.1%，分段效果相较于其他特征参数信息熵分段方法效果最优。
- 交通工程 /
- 轨迹分段 /
- 非机动车超越行为 /
- 信息熵 /
- 特征参数选取 /
- 轨迹覆盖度
Abstract: Identifying overtaking behavior through bicycle trajectories is essential in evaluating the service level of non-motor vehicle transportation. Threshold-based segmentation methods require setting different thresholds for various trajectories, this paper introduces information entropy theory to segment overtaking trajectories of non-motorized vehicle. Using video data, 780 non-motor vehicle overtaking trajectories are extracted, and 11 potential overtaking scenarios are covered. By analyzing the characteristic parameters of each stage of the overtaking process, lateral acceleration, lateral offset distance, and offset angle are identified as the characteristic parameters based on information entropy segmentation. A method for segmenting overtaking trajectory of non-motor vehicles is developed using information entropy theory, and the segmentation judgment criteria is proposed based on this theory. According to the information entropy theory, the law of entropy increase indicates that the probability density of characteristic parameters in two sub-trajectories after segmentation is closer than before segmentation. Besides, considering the features of characteristic parameters of non-motorized vehicle overtaking trajectories, the information entropy segmen-tation standard is proposed for non-motorized vehicle overtaking trajectories. Taking the real trajectory data as experimental samples, trajectory segmentation is carried out using the information entropy segmentation method, and baseline methods with time and speed threshold, respectively. K-nearest neighbor (KNN) classification is adopted for recognizing overtaking trajectories based on the results of trajectory segmentation. Moreover, the trajectory coverage index is used to evaluate the effectiveness of different segmentation methods. The experimental results show that the information entropy based segmentation method has an average coverage of 83.0% for overtaking trajectories, compared to a coverage of 79.7% for the threshold based segmentation method. The information entropy based trajectory segmentation method outperforms the threshold based trajectory segmentation method. Furthermore, the average coverage of lateral acceleration of information entropy based segmentation method is 85.1%, achieving the best performance among the information entropy segmentation methods with different features.
- traffic engineering /
- trajectory segmentation /
- non-motorized vehicle overtaking behavior /
- information entropy /
- feature parameter selection /
- trajectory coverage

HTML全文

图 1 各超越轨迹形式示意图

Figure 1. Schematic diagram of each transcending trajectory

下载: 全尺寸图片幻灯片

图 2 基于信息熵轨迹分段流程图

Figure 2. Segmentation flow chart based on information entropy trajectory

下载: 全尺寸图片幻灯片

图 3 不同K取值下轨迹分段前后平均准确率误差百分比

Figure 3. Average accuracy error percentage before and after trajectory segmentation under different K values

下载: 全尺寸图片幻灯片

图 4 各超越轨迹划分情形轨迹覆盖度对比图

Figure 4. Trajectory coverage comparison chart of each

下载: 全尺寸图片幻灯片

图 5 各种分段方法轨迹覆盖度对比箱型图

Figure 5. Comparative box plot of trajectory coverage of various segmentation methods

下载: 全尺寸图片幻灯片

表 1 调查地点及调查时段情况

Table 1. Investigation location and investigation period

路段名称	隔离类型	非机动车道宽度/m	调查时段	调查轨迹最大长度/m
车公庄大街	分隔带	3	07：00—09：00 11：00—13：00 17：00—19：00	86
中关村南大街	分隔带	3.5		75
东三环南路辅路	划线	4.0		101

下载: 导出CSV

表 2 非机动车平均超越时长统计表

Table 2. Statistical Table of Average Overtaking Time of Non-motorized Vehicles 单位: s

平均超越时长/s
车型	换道超越	对向超越	返回超越	平均超越时长
自行车	2.36	3.15	2.19	7.7
电动车	1.87	2.72	1.81	6.4

下载: 导出CSV

表 3 超越各阶段特征参数调查数据统计表

Table 3. Survey data statistics beyond the characteristic parameters of each stage

特征参数	车型	换道超越	差值比值/%	对向超越	差值比值/%	返回超越	差值比值/%	正常骑行
横向平均加速度/（m/s²）	自行车	2.122	73.2	0.602	5.5	1.975	71.2	0.569
横向平均加速度/（m/s²）	电动车	2.312	67.7	0.963	22.5	2.236	66.6	0.746
平均横向偏移距离/m	自行车	1.436	66.0	0.342	42.7	1.132	56.9	0.488
平均横向偏移距离/m	电动车	1.663	65.5	0.553	3.8	1.465	60.8	0.574
平均偏移角度/rad	自行车	0.775	89.2	0.079	6.3	0.665	87.4	0.084
平均偏移角度/rad	电动车	0.612	89.4	0.054	20.4	0.642	89.9	0.065
横向平均速度/（m/s）	自行车	0.234	48.3	0.145	16.6	0.215	43.7	0.121
横向平均速度/（m/s）	电动车	0.325	48.6	0.185	9.7	0.331	49.5	0.167
纵向平均速度/（m/s）	自行车	5.643	3.7	5.631	3.5	5.693	4.6	5.433
纵向平均速度/（m/s）	电动车	6.754	8.3	6.616	6.4	6.535	5.2	6.194
横向速度标准差/（m/s）	自行车	0.028	32.1	0.031	38.7	0.025	24.0	0.019
横向速度标准差/（m/s）	电动车	0.033	36.4	0.037	43.2	0.031	32.3	0.021
纵向速度标准差/（m/s）	自行车	0.154	16.9	0.184	30.4	0.133	3.80	0.128
纵向速度标准差/（m/s）	电动车	0.163	13.5	0.192	26.6	0.214	34.1	0.141

下载: 导出CSV

表 4 基于信息熵及阈值分段方法各类超越分段情况示意图

Table 4. Schematic diagram of various types of transcendence segmentation based on information entropy and threshold segmentation method

图号	实际超越轨迹段	基于信息熵分段方法			基于阈值分段方法
图号	实际超越轨迹段	LAIE	LODIE	OAIE	ST	TT
类型1
类型2
类型3
类型4
类型5
类型6
类型7
类型8
类型9
类型10
类型11

下载: 导出CSV

参考文献(21)

[1]	RAHMANOV M, SHISHKIN A, KOMKOV V, et al. Simulation of pedestrian dynamics based withemantic trajectory segmentation[C]. E3S Web of Conferences, Online: INTERAGROMASH 2022
[2]	HIGGS B, ABBAS M. Segmentation and clustering of car-following behavior : recognition of driving patterns[J]. IEEE Transactions on Intelligent Transportation Systems, 2015, 16(1): 81-90. doi: 10.1109/TITS.2014.2326082
[3]	郭炜强. 面向轨迹大数据的轨迹分段方法研究及系统实现[D]. 北京: 北方工业大学, 2021. GUO W Q. Research and implementation of trajectory segmentation method for trajectory big data[D]. Beijing: North China University of Technology, 2021. (in Chinese)
[4]	邬婷. 基于分段和分组滤波的细节保留轨迹总结方法[D]. 天津: 天津大学, 2019. WU T. Detail-Preserving trajectory summarization based on segmentation and group-based filtering[D]. Tianjin: Tianjin university, 2019. (in Chinese)
[5]	KUMAR P, PERROLLAZ M, LEFEVRE S, et al. Learning-based approach for online lane change intention prediction[C]. IEEE Intelligent Vehicles Symposium, Gold Coast City, Australia: IEEE, 2013.
[6]	徐文洁, 赵欣, 酆磊, 等. 基于高精度轨迹数据的车辆换道行为识别研究[J]. 武汉理工大学学报(交通科学与工程版), 2023, 47(2): 239-244, 250. doi: 10.3963/j.issn.2095-3844.2023.02.008 XU W J, ZHAO X, LI F, et al. Research on vehicle lane changing behavior recognition based on high precision trajectory data[J]. Journal of Wuhan University of Technology (Transportation Science and Engineering Edition), 2023, 47(2): 239-244, 250. (in Chinese) doi: 10.3963/j.issn.2095-3844.2023.02.008
[7]	LEE J G, HAN J, WHANG K Y. Trajectory clustering: a partition-and-group framework[C]. Acm Sigmod International Conference on Management of Data, New York, USA: Association for Computing Machinery, 2019.
[8]	YUAN G, XIA S, ZHANG L, et al. An efficient trajectory-clustering algorithm based on an index tree[J]. Transactions of the Institute of Measurement and Control. 2012, 34(7): 850-861. doi: 10.1177/0142331211423284
[9]	宋鑫, 朱宗良, 高银萍, 等. 动态阈值结合全局优化的船舶AIS轨迹在线压缩算法[J]. 计算机科学, 2019, 46(7): 333-338. SONG X, ZHU Z L, GAO Y P, et al. Online compression algorithm of ship AIS trajectory based on dynamic threshold combined with global optimization[J]. Computer Science, 2019, 46(7): 333-338. (in Chinese)
[10]	何爱林, 周德超, 陈萍, 等. 基于轨迹聚类的运动趋势分析[J]. 海军工程大学学报, 2017, 29(5): 103-107. HE A L, ZHOU D C, CHEN P, et al. Cluster-based trajectory overall trend extraction[J]. Journal of Naval University of Engineering, 2017, 29(5): 103-107. (in Chinese)
[11]	金佳龙, 周伟, 姜佰辰. 基于行为模式的海上目标轨迹分段算法[J]. 信号处理, 2020, 36(12): 2074-2084. JIN J L, ZHOU W, JIANG B C. Trajectory segmentation algorithm based on behavior pattern[J]. Journal of Signal Processing, 2020, 36(12): 2074-2084. (in Chinese)
[12]	ZHU F X, MIAO L M, LIU W. Research on vessel trajectory multi-dimensional compression algorithm based on Douglas-Peucker theory[J]. Applied Mechanics and Materials, 2014, 694: 59-62. doi: 10.4028/www.scientific.net/AMM.694.59
[13]	YANG X, TANG L. Crowdsourcing big trace data filtering: a partition-and-filter model[C]. XXⅢ ISPRS Congress Prague, Czech Republic: The International Society of Photogrammetry and Remote Sensing, 2016
[14]	王知昊, 元海文, 李维娜, 等. 交汇水域船舶轨迹预测与航行意图识别[J]. 交通信息与安全, 2022, 40(4): 101-109. doi: 10.3963/j.jssn.1674-4861.2022.04.011 WANG Z H, YUAN H W, LI W N, et al. Vessel trajectory prediction and navigational intent recognition in confluence waters[J]. Journal of Transport Information and Safety, 2022, 40(4): 101-109. (in Chinese) doi: 10.3963/j.jssn.1674-4861.2022.04.011
[15]	杨家轩, 马令琪. 基于信息熵的船舶轨迹自适应分段压缩算法[J]. 上海海事大学学报, 2022, 43(2): 7-13, 73. YANG J X, MA L Q. An adaptive segmentation and compression algorithm of ship trajectory based on entropy of information[J]. Journal of Shanghai Maritime University, 2022, 43(2): 7-13, 73. (in Chinese)
[16]	李志斌, 王炜, 赵德, 等. 机非物理分隔道路上自行车超车事件模型[J]. 东南大学学报(自然科学版), 2012, 42(1): 156-161. LI Z B, WANG W, ZHAO D, et al. Modeling bicycle passing events on physically separated roadways[J]. Journal of Southeast University (Natural Science Edition), 2012, 42 (1): 156-161. (in Chinese)
[17]	柴攀. 城市自行车出行者环境感知与行为研究[D]. 西安: 西安建筑科技大学, 2016. CHAI P. Research on environmental perception and behavior of urban bicycle travelers[D]. Xi'an: Xi'an University of Architecture and Technology, 2016. (in Chinese)
[18]	陶思然. 基于自行车与电动自行车的二元混合交通流特性研究[D]. 西安: 长安大学, 2015. TAO S R. Binary mixed traffic characteristics based on bicycle and electric bicycle[D]. Xi'an: Chang'an Univrsity, 2015. (in Chinese)
[19]	王雨楠. 基于轨迹数据的用户行为分析方法研究[D]. 沈阳: 沈阳理工大学, 2020. WANG Y N. Research on user behavior analysis method based on trajectory data[D]. Shenyang: Shenyang University of Science and Technology, 2020. (in Chinese)
[20]	后旗旸. 基于骑行实验的自行车微观行为研究[D]. 北京: 北京交通大学, 2019. HOU Q Y. Study on micro-behavior of cyclist under experiment condition[D]. Beijing: Beijing Jiaotong University, 2019. (in Chinese)
[21]	温惠英, 张伟罡, 赵胜. 基于生成对抗网络的车辆换道轨迹预测模型[J]. 华南理工大学学报(自然科学版), 2020, 48(5): 32-40. WEN H Y, ZHANG W G, ZHAO S. Vehicle lane-change trajectory prediction model based on generative adversarial networks[J]. Journal of South China University of Technology (Natural Science Edition), 2020, 48(5): 32-40. (in Chinese)