基于LSTM概率多模态预期轨迹预测方法

doi:10.19562/j.chinasae.qcgc.2023.07.005

摘要/Abstract

摘要：

针对单模态轨迹预测无法充分表示未来预测空间以及解决轨迹预测固有的不确定性问题，本文构建了驾驶行为意图识别及交通车辆预期轨迹预测模型。驾驶行为意图识别模块识别被预测车辆车道保持、左换道、右换道、左加速换道和右加速换道的概率；交通车辆预期轨迹预测模块采用编码器-解码器架构，输出被预测车辆未来6 s内可能发生的多种行为和轨迹。通过HighD数据集对模型进行训练、验证与测试。试验结果表明：基于意图识别的预期轨迹预测模型生成的多模态概率分布可提高本车行驶安全性，与其他方法相比显著提高轨迹预测精度，在预测长时域轨迹上具有明显的优势。

关键词: 轨迹预测, 行为意图识别, LSTM, 交互式行为

Abstract:

To address the problem that unimodal trajectory prediction cannot adequately represent the future prediction space and solve the inherent uncertainty of trajectory prediction， this paper constructs a driving behavior intention recognition and traffic vehicle expectation trajectory prediction model. The driving behavior intention recognition module identifies the probability of lane keeping， left lane change， right lane change， left acceleration lane change and right acceleration lane change of the predicted vehicle； the traffic vehicle expected trajectory prediction module uses an encoder-decoder architecture to output multiple behaviors and trajectories of the predicted vehicle that may occur within the next 6 seconds. The model is trained， validated and tested with the HighD dataset. The test results show that the multi-modal probability distribution generated by the intention recognition-based expected trajectory prediction model can improve the driving safety of the vehicle， significantly improve the trajectory prediction accuracy compared with other methods， and have obvious advantages in predicting long time domain trajectories.

Key words: trajectory prediction, behavioral intent recognition, LSTM, interactive behavior

高镇海, 鲍明喜, 高菲, 唐明弘. 基于LSTM概率多模态预期轨迹预测方法[J]. 汽车工程, 2023, 45(7): 1145-1152.

Zhenhai Gao, Mingxi Bao, Fei Gao, Minghong Tang. The Method of Probabilistic Multi-modal Expected Trajectory Prediction Based on LSTM[J]. Automotive Engineering, 2023, 45(7): 1145-1152.

图/表 10

图1

图2

图3

图4

图5

图6

表1

图7

图8

图9

参考文献 25

1	HUANG Z， MO X， LV C. Multi-modal motion prediction with transformer-based neural network for autonomous driving［C］. 2022 International Conference on Robotics and Automation （ICRA）. IEEE， 2022： 2605-2611.
2	LUO C， SUN L， DABIRI D， et al. Probabilistic multi-modal trajectory prediction with lane attention for autonomous vehicles［C］. 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems （IROS）. IEEE， 2020： 2370-2376.
3	CASAS S， GULINO C， SUO S， et al. The importance of prior knowledge in precise multimodal prediction［C］. 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems （IROS）. IEEE， 2020： 2295-2302.
4	HUANG Y， DU J， YANG Z， et al. A survey on trajectory-prediction methods for autonomous driving［J］. IEEE Transactions on Intelligent Vehicles， 2022.
5	张晓宁. 基于LSTM网络的车辆轨迹预测研究［J］. 汽车实用技术， 2020， 45（22）：3.
	ZHANG X N. Research on vehicle trajectory prediction based on LSTM network ［J］. Practical Technology of Automobile， 2020， 45（22）：3.
6	张一恒. 基于驾驶风格辨识和运动预测的换道风险评估及决策方法［D］.西安：西安理工大学，2020.
	ZHANG Y H. Lane change risk assessment and decision method based on driving style recognition and motion prediction ［D］. Xi'an：Xi 'an University of Technology，2020.
7	李亚秋，吴超仲，马晓凤，等. 基于EKF学习方法的BP神经网络汽车换道意图识别模型研究［J］. 武汉理工大学学报：交通科学与工程版， 2013， 37（4）：843-847.
	LI Y Q， WU C Z， MA X F，et al. Research on vehicle lane change intention recognition model based on BP neural network based on EKF learning method ［J］. Journal of Wuhan University of Technology： Transportation Science and Engineering，2013，37（4）：843-847.
8	XIE D F， FANG Z Z， JIA B， et al. A data-driven lane-changing model based on deep learning［J］. Transportation Research Part C： Emerging Technologies， 2019， 106： 41-60.
9	LIN L， LI W， BI H， et al. Vehicle trajectory prediction using lstms with spatial–temporal attention mechanisms［J］. IEEE Intelligent Transportation Systems Magazine， 2021， 14（2）： 197-208.
10	XIAO H， WANG C， LI Z， et al. UB-LSTM： a trajectory prediction method combined with vehicle behavior recognition［J］. Journal of Advanced Transportation， 2020： 1-12.
11	TANG C， SALAKHUTDINOV R R. Multiple futures prediction［J］. Advances in Neural Information Processing Systems， 2019， 32.
12	LEE N， CHOI W， VERNAZA P， et al. Desire： distant future prediction in dynamic scenes with interacting agents［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition， 2017： 336-345.
13	YUAN Y， KITANI K. Diverse trajectory forecasting with determinantal point processes［J］. arXiv preprint arXiv：， 2019.
14	GUPTA A， JOHNSON J， FEI-FEI L， et al. Social gan： socially acceptable trajectories with generative adversarial networks［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition， 2018： 2255-2264.
15	XIA C， WENG C Y， ZHANG Y， et al. Vision-based measurement and prediction of object trajectory for robotic manipulation in dynamic and uncertain scenarios［J］. IEEE Transactions on Instrumentation and Measurement， 2020， 69（11）： 8939-8952.
16	GUO H Y， MENG Q Y， CAO D P， et al. Vehicle trajectory prediction method coupled with ego vehicle motion trend under dual attention mechanism［J］. IEEE Transactions on Instrumentation and Measurement， 2022， 71： 1-16.
17	PHAN-MINH T， GRIGORE E C， BOULTON F A， et al. Covernet： multimodal behavior prediction using trajectory sets［C］. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition， 2020： 14074-14083.
18	RHINEHART N， MCALLISTER R， KITANI K， et al. Precog： prediction conditioned on goals in visual multi-agent settings［C］.Proceedings of the IEEE/CVF International Conference on Computer Vision， 2019： 2821-2830.
19	BIKTAIROV Y， STEBELEV M， RUDENKO I， et al. Prank： motion prediction based on ranking［J］. Advances in Neural Information Processing Systems， 2020， 33： 2553-2563.
20	DEO N， TRIVEDI M M. Convolutional social pooling for vehicle trajectory prediction［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops， 2018： 1468-1476.
21	DEO N， TRIVEDI M M. Multi-modal trajectory prediction of surrounding vehicles with maneuver based lstms［C］. 2018 IEEE Intelligent Vehicles Symposium （IV）. IEEE， 2018： 1179-1184.
22	DEO N， WOLFF E， BEIJBOM O. Multimodal trajectory prediction conditioned on lane-graph traversals［C］. Conference on Robot Learning. PMLR， 2022： 203-212.
23	CUI H， RADOSAVLJEVIC V， CHOU F C， et al. Multimodal trajectory predictions for autonomous driving using deep convolutional networks［C］. 2019 International Conference on Robotics and Automation （ICRA）. IEEE， 2019： 2090-2096.
24	GERS F A， SCHMIDHUBER J， CUMMINS F. Learning to forget： continual prediction with LSTM［J］. Neural Computation， 2000， 12（10）： 2451-2471.
25	KRAJEWSKI R， BOCK J， KLOEKER L， et al. The high dataset： a drone dataset of naturalistic vehicle trajectories on german highways for validation of highly automated driving systems［C］. 2018 21st International Conference on Intelligent Transportation Systems （ITSC）. IEEE， 2018： 2118-2125.

评价指标		精确率 p	召回率 r	F1-score	准确率 a
预测意图	直线行驶	0.989 0	0.992 4	0.990 7	0.983 4
	左换道	0.973 0	0.972 7	0.972 8
	右换道	0.973 2	0.981 8	0.977 5
	左加速换道	0.923 0	0.815 7	0.866 0
	右加速换道	0.920 7	0.817 6	0.872 0

[1]	赵高士,陈龙,蔡英凤,廉玉波,王海,刘擎超,滕成龙. 融合复杂网络和记忆增强网络的轨迹预测技术[J]. 汽车工程, 2023, 45(9): 1608-1616.
[2]	刘卫国,项志宇,刘锐,李国栋,王子旭. 基于深度学习的端到端车辆运动规划方法研究[J]. 汽车工程, 2023, 45(8): 1343-1352.
[3]	连静,李硕贤,刘一荻,杨东方,李琳辉. 基于车道目标引导的车辆轨迹预测[J]. 汽车工程, 2023, 45(8): 1353-1361.
[4]	胡启慧,蔡英凤,王海,陈龙,董钊志,刘擎超. 基于层次图注意的异构多目标轨迹预测方法[J]. 汽车工程, 2023, 45(8): 1448-1456.
[5]	韩勇,林旭洁,黄红武,蔡鸿瑜,罗金镕,李燕婷. 典型汽车碰撞事故场景中行人运动轨迹预测方法[J]. 汽车工程, 2023, 45(6): 1022-1030.
[6]	高翔,陈龙,王歆叶,熊晓夏,李祎承,陈月霞. 基于轨迹预测的智能汽车行驶风险评估方法[J]. 汽车工程, 2023, 45(4): 588-597.
[7]	汪梓豪,蔡英凤,王海,陈龙,熊晓夏. 基于单目视觉运动估计的周边多目标轨迹预测方法[J]. 汽车工程, 2022, 44(9): 1318-1326.
[8]	张哲雨,吕超,李景行,熊光明,吴绍斌,龚建伟. 基于车辆视角数据的行人轨迹预测与风险等级评定[J]. 汽车工程, 2022, 44(5): 675-683.
[9]	郭景华,何智飞,罗禹贡,李克强. 人机混驾环境下基于深度学习的车辆切入轨迹预测[J]. 汽车工程, 2022, 44(2): 153-160.
[10]	梁冠群,赵通,王岩,危银涛. 基于LSTM网络的路面不平度辨识方法[J]. 汽车工程, 2021, 43(4): 509-517.
[11]	吴晓建,燕冬,王爱春,黄菊花,伍磊,周兵. 融合前车轨迹预测的改进人工势场轨迹规划研究[J]. 汽车工程, 2021, 43(12): 1752-1761.
[12]	张一鸣, 周兵, 吴晓建, 崔庆佳, 柴天. 基于前车轨迹预测的高速智能车运动规划^*[J]. 汽车工程, 2020, 42(5): 574-580.
[13]	谢枫, 娄静涛, 赵凯, 齐尧. 基于行为识别和曲率约束的车辆轨迹预测方法研究^*[J]. 汽车工程, 2019, 41(9): 1036-1042.
[14]	焦新宇, 杨殿阁, 江昆, 曹重, 谢诗超, 王思佳. 基于端到端学习机制的高速公路行驶轨迹曲率预测^*[J]. 汽车工程, 2018, 40(12): 1494-1499.