基于多头注意力的CNN-LSTM的换道意图预测

doi:10.3901/JME.2022.22.369

机械工程学报 ›› 2022, Vol. 58 ›› Issue (22): 369-378.doi: 10.3901/JME.2022.22.369

扫码分享

基于多头注意力的CNN-LSTM的换道意图预测

高凯^1,2, 李勋豪¹, 胡林¹, 陈彬¹, 杜荣华^1,2

1. 长沙理工大学汽车与机械工程学院长沙 410114;
2. 长沙理工大学智能道路与车路协同湖南省重点试验室长沙 410114

收稿日期:2021-11-30 修回日期:2022-05-20 出版日期:2022-11-20 发布日期:2023-02-07
通讯作者: 胡林(通信作者),男,1978年出生,博士,教授,博士研究生导师。主要研究方向为车辆智能化,车辆安全。E-mail:hulin@csust.edu.cn
作者简介:高凯,男,1985年出生,博士,副教授,硕士研究生导师。主要研究方向为自动驾驶汽车感知与控制,智能交通与车联网应用。E-mail:kai_g@csust.edu.cn;李勋豪,男,1997年出生,硕士研究生。主要研究方向为自动驾驶,机器学习。E-mail:lxhms07@163.com
基金资助:
国家自然科学基金(52172399);湖南省自然科学基金(2021JJ40575);湖南省研究生科研创新项目(QL20210194)资助项目

Lane Change Intention Prediction of CNN-LSTM Based on Multi-head Attention

GAO Kai^1,2, LI Xun-hao¹, HU Lin¹, CHEN Bin¹, DU Rong-hua^1,2

1. College of Automotive and Mechanical Engineering, Changsha University of Science & Technology, Changsha 410114;
2. Hunan Key Laboratory of Smart Roadwasy and Cooperative Vehicle-Infrastructure Systems, Changsha University of Science & Technology, Changsha 410114

Received:2021-11-30 Revised:2022-05-20 Online:2022-11-20 Published:2023-02-07

摘要/Abstract

摘要： 自动驾驶车辆与传统车辆混行的交通环境中,车辆的换道意图预测能够为自动驾驶车辆安全行驶提供有效保证。为了更准确地预测车辆的换道意图,将多头注意力与卷积神经网络(Convolution neural network,CNN)和长短时记忆(Long-short term memory,LSTM)网络结合,提出一种新型车辆换道意图预测算法。首先对NGSIM(Next generation Simulaion)数据集进行处理,提取车辆横向位置信息和周围环境信息。然后输入基于多头注意力(Multi-headattention)的CNN-LSTM模型,提高对输入序列特征的提取能力和预测精度。最后在NGSIM数据集验证该模型的有效性。试验结果表明,该模型能够从大量数据中提取到重要特征,同时通过特征对比试验发现,横向位置信息作为预测的主要特征,而周围环境信息作为预测的辅助特征。最后通过模型的对比试验得出,该模型的换道意图预测准确率在换道前1s、2s、3s相比于LSTM、CNN、CNN-LSTM模型具有更好的预测精度,可以为自动驾驶汽车设计先进的意图预测算法提供帮助和参考。

关键词: 自动驾驶, 换道意图, 卷积神经网络, 长短期记忆网络, 多头注意力

Abstract: In the mixed traffic environment of autonomous vehicles and traditional vehicles, the prediction of vehicle lane change intention can provide an effective guarantee for the safe driving of autonomous vehicles. In order to more accurately predict the vehicle lane change intention, a novel vehicle lane change intention prediction algorithm is proposed by combining the multi head attention with convolution neural network(CNN) and long-short term memory(LSTM) network. Firstly, the NGSIM(Next generation simulaion) data set is processed to extract the vehicle lateral position information and surrounding environment information. Then, it is input into the CNN-LSTM model based on multi-head attention to improve the feature extraction ability and prediction accuracy of the input sequence. Finally, the effectiveness of the proposed model is verified on the NGSIM dataset. The experimental results show that the model can extract important features from a large number of data. At the same time, through the feature comparison experiment, it is found that the lateral position information is the main feature of prediction, and the surrounding environment information is the auxiliary feature of prediction. Finally, through the comparison experiment of the model, it is concluded that the prediction accuracy of the model is better than that of LSTM, CNN and CNN-LSTM models before 1s, 2s and 3s. It can provide help and reference for the design of advanced prediction algorithms for autopilot cars.

Key words: autonomous driving, lane change intention, CNN, LSTM, multi-head attention

中图分类号:

U461

高凯, 李勋豪, 胡林, 陈彬, 杜荣华. 基于多头注意力的CNN-LSTM的换道意图预测[J]. 机械工程学报, 2022, 58(22): 369-378.

GAO Kai, LI Xun-hao, HU Lin, CHEN Bin, DU Rong-hua. Lane Change Intention Prediction of CNN-LSTM Based on Multi-head Attention[J]. Journal of Mechanical Engineering, 2022, 58(22): 369-378.

参考文献

[1] GONG Y,ZHANG L,LIU R,et al. Nonlinear MIMO for industrial internet of things in cyber-physical systems[J].IEEE Transactions on Industrial Informatics, 2020,17(8):5533-5541.
[2] ZHEN L,BASHUR A K,YU K,et al. Energy-efficient random access for LEO satellite-assisted 6G internet of remote things[J]. IEEE Internet of Things Journal,2021,8(7):5114-5128.
[3] ZHAO L,ZHAO W,HAWBANI A,et al. Novel online sequential learning-based adaptive routing for edge software-defined vehicular networks[J]. IEEE Transactions on Wireless Communications,2020,20(5):2991-3004.
[4] ZHAO L, YANG K, TAN Z, et al. A novel cost optimization strategy for SDN-enabled UAV-assisted vehicular computation offloading[J]. IEEE Transactions on Intelligent Transportation Systems,2021,22(6):3664-3674
[5] ZHAO L, HAN G, LI Z, et al. Intelligent digital twin-based software-defined vehicular networks[J]. IEEE Network,2020,34(5):178-184.
[6] ZHAO L,LIU Y,Al-DUBAI A Y,et al. A novel generation-adversarial-network-based vehicle trajectory prediction method for intelligent vehicular networks[J].IEEE Internet of Things Journal,2021,8(3):2066-2077.
[7] 黄晶,李勇,胡林.基于数据关联和改进统计模型的激光雷达目标跟踪研究[J].汽车工程,2018,40(3):356-362.HUANG Jing,LI Yong,HU Lin. A research on target tracking by LIDAR based on data association and improved statistical model[J]. Automotive Engineering,2018,40(3):356-362.
[8] XIE D,FANG 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] SU S, MUELLING K, DOLAN J, et al. Learning vehicle surrounding-aware lane-changing behavior from observed trajectories[C]//2018 IEEE Intelligent Vehicles Symposium(IV). China:IEEE,2018:1412-1417.
[10] 张海伦,付锐.高速场景相邻前车驾驶行为识别及意图预测[J].交通运输系统工程与信息,2020,20(1):40-46.ZHANG Hailun, FU Rui. Driving behavior recognition and intention prediction of adjacent preceding vehicle in highway scene[J]. Journal of Transportation Systems Engineering and Information Technology,2020,20(1):40-46.
[11] YU K,LIN L,ALAZAB M,et al. Deep learning-based traffic safety solution for a mixture of autonomous and manual vehicles in a 5G-enabled intelligent transportation system[J]. IEEE Transactions on Intelligent Transportation Systems,2021,22(7):4337-4347.
[12] PENG J,GUO Yingshi,FU Rui,et al. Multi-parameter prediction of drivers'lane-changing behaviour with neural network model[J]. Applied Ergonomics, 2015, 50:207-217.
[13] IZQUIERDO R,QUINTANAR A,PARRA I,et al.Experimental validation of lane-change intention prediction methodologies based on CNN and LSTM[C]//2019 IEEE Intelligent Transportation Systems Conference(ITSC). New Zealand:IEEE,2019:3657-3662.
[14] MOUSSAID A,BERRADA I,El KAMILI M,et al.Predicting driver lane change maneuvers using driver's face[C]//2019 International Conference on Wireless Networks and Mobile Communications(WINCOM).Morocco:IEEE,2019:1-7.
[15] ORDÓÑEZ F J,ROGGEN D. Deep convolutional and lstm recurrent neural networks for multimodal wearable activity recognition[J]. Sensors,2016,16(1):115.
[16] ABEDINIA O,AMJADY N,ZAREIPOUR H. A new feature selection technique for load and price forecast of electrical power systems[J]. IEEE Transactions on Power Systems,2016,32(1):62-74.
[17] 车畅畅,王华伟,倪晓梅,等.基于1D-CNN和Bi-LSTM的航空发动机剩余寿命预测[J].机械工程学报,2021,57(14):304-312.CHE Changchang,WANG Huawei,NI Xiaomei. Residual life prediction of aeroengine based on 1D-CNN and Bi-LSTM[J]. Journal of Mechanical Engineering,2021,57(14):304-312.
[18] 李梅,宁德军,郭佳程.基于注意力机制的CNN-LSTM模型及其应用[J].计算机工程与应用,2019,55(13):20-27.LI Mei, NING Dejun, GUO Jiacheng. Attention mechanism-based CNN-LSTM model and its application[J]. Computer Engineering and Applications,2019,55(13):20-27.
[19] VASWANI A,SHAZEER N,PARMAR N,et al. Attention is all you need[C]//Advances in Neural Information Processing Systems. USA:NIPS,2017:5998-6008.
[20] GIULIARI F, HASAN I, CRISTANI M, et al.Transformer networks for trajectory forecasting[C]//202025th International Conference on Pattern Recognition(ICPR). Italy:IEEE,2021:10335-10342.
[21] ZHOU H,ZHANG S,PPENG J,et al. Informer:Beyond efficient transformer for long sequence time-series forecasting[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Virtual:AAAI,2021:11106-11115.
[22] LIU Y,ZHANG J,FANG L,et al. Multimodal motion prediction with stacked Transformers[C]//2021IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). USA:IEEE,2021:7573-7582.
[23] LIU S,WANG B,DENG X,et al. Self-attentive graph convolution network with latent group mining and collaborative filtering for personalized recommendation[J/OL]. IEEE Transactions on Network Science and Engineering, 2021.[2021-10-11].https://ieeexplore.ieee.org/abstract/document/9531473.
[24] 宋晓琳,曾艳兵,曹昊天,等.基于长短期记忆网络的换道意图识别方法[J].中国公路学报,2021,11(8):1-11.SONG Xiaolin,ZENG Yanbing,CAO Haotian,et al. Lane change intention recognition method based on LSTM network[J]. China Journal of Highways,2021,11(8):1-11.
[25] XU K,BA J,KIROS R,et al. Show,attend and tell:Neural image caption generation with visual attention[C]//International Conference on Machine Learning. France:PMLR,2015:2048-2057.
[26] 李海超,刘景风,谢吉兵,等.基于卷积神经网络的GTAW熔透预测[J].机械工程学报,2019,55(17):22-28.LI Haichao,LIU Jingfeng,XIE Jibing,et al. GTAW penetration prediction model based on convolution neural network algorithm[J]. Journal of Mechanical Engineering,2019,55(17):22-28.
[27] IOFFE S,SZEGEDY C. Batch normalization:accelerating deep network training by reducing internal covariate shift[C]//International Conference on Machine Learning.France:PMLR,2015:448-456.
[28] HOCHREITER S,SCHMIDHUBER J. Long short-term memory[J]. Neural Computation,1997,9(8):1735-1780.
[29] FEDERAL HIGHWAY ADMINISTRATION. NGSIMnext generation simulation[EB/OL].[2021-10-11].https://ops.fhwa.dot.gov/trafficanalysistools/ngsim.htm.
[30] THIEMANN C,TREIBER M,KESTING A. Estimating acceleration and lane-changing dynamics from next generation simulation trajectory data[J]. Transportation Research Record,2008(1):90-101.
[31] DOROUDGAR S,CHUANG H M,PERRY P J,et al.Driving performance comparing older versus younger drivers[J]. Traffic Injury Prevention,2017,18(1):41-46.
[32] 陈灵,陈郑平,李军良,等.基于深度卷积神经网络和合作博弈的多微网实时能量管理策略[J].电气工程学报,2022,17(2):215-225.CHEN Ling,CHEN Zhengping,LI Junliang,et al.Real-time energy management strategy for micro-grid clusters based on deep convolutional neural network and cooperative game[J]. Journal of Electrical Engineering,2022,17(2):215-225.
[33] 李宇,杨柳林.基于卷积神经网络的配电网单相接地故障识别[J].电气工程学报,2020,15(3):22-30.LI Yu, YANG Liulin. Identification of single-phaseto-earth fault in distribution network based on convolutional neural network[J]. Journal of Electrical Engineering,2020,15(3):22-30.

基于多头注意力的CNN-LSTM的换道意图预测

Lane Change Intention Prediction of CNN-LSTM Based on Multi-head Attention

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	狄子钧, 袁东风, 李东阳, 梁道君, 周晓天, 信苗苗, 曹凤, 雷腾飞. 基于多尺度-高效通道注意力网络的刀具故障诊断方法[J]. 机械工程学报, 2024, 60(6): 82-90.
[2]	褚端峰, 刘鸿祥, 高博麟, 王金湘, 殷国栋. 车辆队列预测巡航控制研究综述[J]. 机械工程学报, 2024, 60(18): 218-246.
[3]	隗寒冰, 吴化腾, 徐进. 考虑驾驶员NMS特征的自动驾驶汽车人机共驾鲁棒横向控制[J]. 机械工程学报, 2024, 60(16): 280-290.
[4]	袁静, 任港星, 蒋会明, 赵倩, 魏臣隽, 朱骏. 基于多元提升核神经网络的机械故障诊断方法及其特征提取可解释性研究[J]. 机械工程学报, 2024, 60(12): 51-64.
[5]	陈钱, 陈康康, 董兴建, 皇甫一樊, 彭志科, 孟光. 一种面向机械设备故障诊断的可解释卷积神经网络[J]. 机械工程学报, 2024, 60(12): 65-76.
[6]	王振亚, 刘韬, 伍星. 互信息规范的卷积神经网络及其在轴承故障特征提取中的应用[J]. 机械工程学报, 2024, 60(12): 137-146.
[7]	高镇海, 于桐, 孙天骏. 考虑社会性行为的自动驾驶运动规划研究综述[J]. 机械工程学报, 2024, 60(10): 112-128.
[8]	戢杨杰, 张馨雨, 杨紫茹, 周上航, 黄岩军, 曹建永, 熊璐, 余卓平. 多智能网联汽车轨迹规划：现状与展望[J]. 机械工程学报, 2024, 60(10): 129-146.
[9]	梁凯冲, 赵治国, 颜丹姝, 赵坤. 基于动态运动基元的车辆高速公路换道轨迹规划[J]. 机械工程学报, 2024, 60(10): 192-206.
[10]	周洪龙, 裴晓飞, 刘一平, 赵柯帆. 面向动态不确定场景的自动驾驶车辆时空耦合分层轨迹规划研究[J]. 机械工程学报, 2024, 60(10): 222-234.
[11]	曾迪, 郑玲, 李以农, 杨显通. 自动驾驶奖励函数贝叶斯逆强化学习方法[J]. 机械工程学报, 2024, 60(10): 245-260.
[12]	聂士达, 刘辉, 廖志昊, 谢雨佳, 项昌乐, 韩立金, 林思豪. 考虑复杂地形的越野环境无人车辆路径规划研究[J]. 机械工程学报, 2024, 60(10): 261-272.
[13]	杨硕, 李时珍, 赵中原, 黄小鹏, 黄岩军. 基于时序差分学习模型预测控制的一体化自动驾驶换道策略[J]. 机械工程学报, 2024, 60(10): 329-338.
[14]	张志勇, 黄大洋, 黄彩霞, 胡林, 杜荣华. TD3算法改进与自动驾驶汽车并道策略学习[J]. 机械工程学报, 2023, 59(8): 224-234.
[15]	邓磊鑫, 谢清林, 陶功权, 温泽峰. 基于轴箱振动与动力学模型驱动的高速列车车轮失圆状态识别方法[J]. 机械工程学报, 2023, 59(3): 110-121.