吉林大学学报(工学版) ›› 2023, Vol. 53 ›› Issue (5): 1474-1480.doi: 10.13229/j.cnki.jdxbgxb.20210904

• 通信与控制工程 • 上一篇    

基于注意力与深度交互的周车多模态行为轨迹预测

田彦涛1,2(),黄兴1,卢辉遒1,王凯歌1,许富强1   

  1. 1.吉林大学 通信工程学院，长春 130022
    2.吉林大学 工程仿生教育部重点实验室，长春 130022
  • 收稿日期:2021-09-09 出版日期:2023-05-01 发布日期:2023-05-25
  • 作者简介:田彦涛（1958-），男，教授，博士.研究方向：复杂系统建模，优化与控制.E-mail：tianyt@jlu.edu.cn
  • 基金资助:
    国家自然科学基金区域创新发展联合基金项目(U19A2069)

Multi⁃mode behavior trajectory prediction of surrounding vehicle based on attention and depth interaction

Yan-tao TIAN1,2(),Xing HUANG1,Hui-qiu LU1,Kai-ge WANG1,Fu-qiang XU1   

  1. 1.College of Communication Engineering，Jilin University，Changchun 130022，China
    2.Key Laboratory of Bionic Engineering，Ministry of Education，Jilin University，Changchun 130022，China
  • Received:2021-09-09 Online:2023-05-01 Published:2023-05-25

RICH HTML

 10   

PDF (PC)

370

摘要：

设计了一种车辆深度交互编码并结合基于注意力机制的解码器模型，该模型同时输出车辆多模态行为预测结果和未来轨迹预测分布。使用公开的NGSIM US-101和I-80数据集评估所提出的模型，并且对模型多模态行为机动预测进行了定性分析。结果表明：该模型具有较好的均方根误差值（RMSE），在提升了计算效率的基础上获得了更高的轨迹预测精度。

关键词: 车辆工程, 轨迹预测, 多模态预测, 注意力机制, 门控循环单元

Abstract:

A vehicle deep-interaction coding model combined with a decoder based on the attention-mechanism to solve this problem was presented in this work. The model's multi-modal behavior predictions and trajectory predictions were output. The proposed model is evaluated by the public NGSIM US-101 and I-80 data sets. The results show that the model has a better root mean square error and achieves higher trajectory prediction accuracy while improving computational efficiency. This paper also shows a qualitative analysis of the prediction of multi-modal behavior maneuvering.

Key words: vehicle engineering, trajectory prediction, multimodal prediction, attention mechanism, gated recurrent unit

中图分类号: 

  • U495

图1

参考坐标系与驾驶行为集"

图2

基于注意力机制的深度交互GRU模型"

图3

GRU的单元结构"

图4

NGSIM数据集I-80路段"

表1

不同模型的RMSE值 (m)"

模 型预测时间/s
12345
V-LSTM0.681.652.914.466.27
C-VGMM+VIM0.661.562.754.246.68
CS-LSTM0.651.342.403.564.54
DCS-LSTM0.581.262.213.464.43
ADI-DCS-GRU0.551.242.103.124.23
ADI-DCS-GRU（M）0.571.262.153.234.35

图5

负对数似然误差"

图6

换道过程的预测分析"

图7

周围车辆对多模态行为轨迹预测的影响"

1 Lefèvre S, Vasquez D, Laugier C. A survey on motion prediction and risk assessment for intelligent vehicles [J]. Robomech Journal, 2014, 1(1): 1-14.
2 Ammous S, Nashashibi F. Real time trajectory prediction for collision risk estimation between vehicles [C]∥IEEE 5th Internationgal Conference on Intelligent Computer Communication and Proceding, Cluj-Napoca, Romania, 2009: 417-422.
3 Broadhurst A, Baker S, Kanade T. Monte carlo road safety reasoning[C]∥IEEE Intelligent Vehicles Symposium Proceedings,Las Vegas,USA,2005: 319-324.
4 Schreier M, Willert V, Adamy J. An integrated approach to maneuver-based trajectory prediction and criticality assessment in arbitrary road environments[J]. IEEE Transactions on Intelligent Transportation Systems, 2016, 17(10): 2751-2766.
5 Zhang H, Geiger A, Urtasun R. Understanding high-level semantics by modeling traffic patterns[C]∥IEEE International Conference on Computer Vision, Sydney, Australia, 2013: 3056-3063.
6 Lefèvre S, Laugier C, Ibaez-Guzmàn J. Intention-aware risk estimation for general traffic situations, and application to intersection safety[J]. Hal Inria, 2013, 8379: 1-23.
7 Deo N, Rangesh A, Trivedi M M. How would surround vehicles move? a unified framework for maneuver classification and motion prediction[J]. IEEE Transactions on Intelligent Vehicles, 2018, 3(2): 129-140.
8 Lee D, Kwon Y P, Mcmains S. Convolution neural network-based lane change intention prediction of surrounding vehicles for ACC[C]∥IEEE 20th International Conference on Intelligent Transportation Systems, Yokohama, Japan, 2017: 1-6.
9 Khosroshahi A, Ohn-Bar E, Trivedi M M. Surround vehicles trajectory analysis with recurrent neural networks[C]∥IEEE International Conference on Intelligent Transportation Systems, Rio de Janeiro, Cluj-Napoca, Romania, 2016: 2267-2272.
10 Phillps D J, Wheeler T A, Kochenderfer M J. Generalizable intention prediction of human drivers at intersections [C]∥IEEE Intelligent Vehicles Symposium (IV), Redondo Beach, USA, 2017: 1665-1670.
11 Kim B, Kang C M, Kim J. Probabilistic vehicle trajectory prediction over occupancy grid map via recurrent neural network [C]∥IEEE 20th International Conference on Intelligent Transportation Systems, Yokohama, Japan, 2017: 399-404.
12 Lee N, Choi W, Vernaza P. DESIRE: distant future prediction in dynamic scenes with interacting agents[C]∥IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 2165-2174.
13 Deo N, Trivedi M M. Convolutional social pooling for vehicle trajectory prediction[C]∥IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, Salt Lake City, USA, 2018: 1549-1557.
14 Deo N, Trivedi M M. Multi-modal trajectory prediction of surrounding vehicles with maneuver based LSTMs[C]∥IEEE Intelligent Vehicles Symposium, Changshu, China, 2018: 1179-1184.
15 Zhang H, Wang Y, Liu J. A multi-modal states based vehicle descriptor and dilated convolutional social pooling for vehicle trajectory prediction[J]. SAE Technical Papers, arXiv: .
16 Alexiadis V, Colyar J, Halkias J. The next generation simulation program[J]. ITE Journal, 2004, 74(8): 22-26.
17 Thiemann C, Treiber M, Kesting A. Estimating acceleration and lane-changing dynamics based on NGSIM trajectory data[J]. Transportation Research Record Journal of the Transportation Research Board, 2008(2088): 90-101.
[1] 陈磊,王杨,董志圣,宋亚奇. 一种基于转向意图的车辆敏捷性控制策略[J]. 吉林大学学报(工学版), 2023, 53(5): 1257-1263.
[2] 陈鑫,张冠宸,赵康明,王佳宁,杨立飞,司徒德蓉. 搭接焊缝对铝合金焊接结构轻量化设计的影响[J]. 吉林大学学报(工学版), 2023, 53(5): 1282-1288.
[3] 张勇,毛凤朝,刘水长,王青妤,潘神功,曾广胜. 基于Laplacian算法的汽车外流场畸变网格优化[J]. 吉林大学学报(工学版), 2023, 53(5): 1289-1296.
[4] 汪少华,储堃,施德华,殷春芳,李春. 基于有限时间扩张状态观测的HEV鲁棒复合协调控制[J]. 吉林大学学报(工学版), 2023, 53(5): 1272-1281.
[5] 于贵申,陈鑫,武子涛,陈轶雄,张冠宸. AA6061⁃T6铝薄板无针搅拌摩擦点焊接头结构及性能分析[J]. 吉林大学学报(工学版), 2023, 53(5): 1338-1344.
[6] 杨红波,史文库,陈志勇,郭年程,赵燕燕. 基于NSGA⁃II的斜齿轮宏观参数多目标优化[J]. 吉林大学学报(工学版), 2023, 53(4): 1007-1018.
[7] 赵睿,李云,胡宏宇,高镇海. 基于V2I通信的交叉口车辆碰撞预警方法[J]. 吉林大学学报(工学版), 2023, 53(4): 1019-1029.
[8] 陈小波,陈玲. 定位噪声统计特性未知的变分贝叶斯协同目标跟踪[J]. 吉林大学学报(工学版), 2023, 53(4): 1030-1039.
[9] 吕卫,韩镓泽,褚晶辉,井佩光. 基于多模态自注意力网络的视频记忆度预测[J]. 吉林大学学报(工学版), 2023, 53(4): 1211-1219.
[10] 何科,丁海涛,赖宣淇,许男,郭孔辉. 基于Transformer的轮式里程计误差预测模型[J]. 吉林大学学报(工学版), 2023, 53(3): 653-662.
[11] 刘嫣然,孟庆瑜,郭洪艳,李嘉霖. 图注意力模式下融合高精地图的周车轨迹预测[J]. 吉林大学学报(工学版), 2023, 53(3): 792-801.
[12] 田彦涛,季言实,唱寰,谢波. 深度强化学习智能驾驶汽车增广决策模型[J]. 吉林大学学报(工学版), 2023, 53(3): 682-692.
[13] 张建,刘金波,高原,刘梦可,高振海,杨彬. 基于多模交互的车载传感器定位算法[J]. 吉林大学学报(工学版), 2023, 53(3): 772-780.
[14] 谢波,高榕,许富强,田彦涛. 低附着路况条件下人车共享转向系统稳定控制[J]. 吉林大学学报(工学版), 2023, 53(3): 713-725.
[15] 田彦涛,许富强,王凯歌,郝子绪. 考虑周车信息的自车期望轨迹预测[J]. 吉林大学学报(工学版), 2023, 53(3): 674-681.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 《吉林大学学报(工学版)》编辑部
地址：长春市人民大街5988号　电话：+86-431-85095297　传真：+86-431-85094074　E-mail: xbgxb@jlu.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发