| 772 | 14 | 1991 |
| 下载次数 | 被引频次 | 阅读次数 |
自动驾驶汽车需要具备预测周围车辆轨迹的能力，以便做出合理的决策规划，提高行驶中的安全性和舒适性。为了能准确地预测汽车行驶的未来轨迹，本文运用神经网络的方法，设计了一种基于改进Informer模型的多意图轨迹预测模型。该模型使用编码器-解码器结构，输入数据为交通场景中的历史时域信息，输出为车辆的多意图预测轨迹。模型的编码器使用交互信息提取网络，根据特征间的依赖关系提取车辆交互信息，解码器根据编码器的输出特征向量预测表征多种驾驶意图的多意图轨迹。通过使用真实高速公路轨迹HighD数据集对模型进行训练、验证和测试，试验结果表明，本文提出的多意图轨迹预测模型能准确地预测出目标车辆的未来可能轨迹，并且在预测精度上优于基于长短时记忆网络的轨迹预测模型，增加交互信息提取网络使模型预测具有更高的准确率，输出多条表征不同驾驶意图的轨迹有利于反映客观真实轨迹分布，提高车辆主动安全性。本文还进行了通过预测轨迹判断换道意图的补充实验，通过本文模型预测轨迹判断换道意图准确率在换道前3 s达到97%，从侧面反映出本文提出的模型预测轨迹的性能优异。
Abstract：Autonomous vehicles must accurately predict the trajectories of nearby vehicles to enhance safety and comfort and facilitate rational decision-making. This study utilized a neural network approach to develop a multi-intention trajectory prediction model based on the Informer model and achieve accurate vehicle trajectory prediction. The model employs an encoder-decoder structure,using historical information from the traffic scene as input data and generating the vehicle's multi-intention-predicted trajectory as output. The encoder of the model utilizes an interactive information extraction network to extract vehicle interaction information by considering the dependency relationships. Subsequently, the decoder predicts a multi-intention trajectory that reflects various driving intentions based on the output of the encoder. Through training, validation, and testing on the HighD dataset of highway trajectories, the experimental results demonstrated the accurate prediction capability of the proposed multi-intention trajectory prediction model for target vehicles. Moreover, it outperformed the long short-term memory network-based trajectory prediction model in terms of accuracy. The inclusion of an interactive information extraction network enhanced the accuracy of the model predictions. Furthermore, generating multiple trajectories embodying distinct driving intentions contributes to objectively capturing the actual trajectory distribution, thereby enhancing vehicle safety. In addition, a supplementary experiment was conducted to assess the lane change intention using predicted trajectories. The proposed model achieved 97% accuracy in predicting lane changes 3 s prior, indirectly indicating its outstanding trajectory prediction performance.
[1]李松，张开碧，李永福，等.理想诱导环境下的网联车与网联自动驾驶车混合交通流建模研究[J].交通运输工程与信息学报, 2023, 21(3):31-58.LI Song, ZHANG Kaibi, LI Yongfu, et al. Modeling a mixed traffic flow of connected vehicles and connected autonomous vehicles in an ideal induction environment[J]. Journal of Transportation Engineering and Information, 2023, 21(3):31-58.
[2] BAHARI M, SAADATNEJAD S, RAHIMI A, et al. Vehicle trajectory prediction works, but not everywhere[C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). New Orleans:IEEE, 2022:17102-17112.
[3] YIN H, WEN Y, LI J. A survey of vehicle trajectory prediction based on deep-learning[C]//2023 3rd International Conference on Neural Networks, Information and Communication Engineering(NNICE). Guangzhou:IEEE,2023:140-144.
[4] XIE G, GAO H, QIAN L, et al. Vehicle trajectory prediction by integrating physics-and maneuver-based approaches using interactive multiple models[J]. IEEE Transactions on Industrial Electronics, 2018, 65(7):5999-6008.
[5]邹智军，杨东援.微观交通仿真中的车道变换模型[J].中国公路学报, 2002, 15(2):105-108.ZOU Zhijun, YANG Dongyuan. Lane changing model for micro traffic simulation[J]. China Journal of Highway and Transport, 2002, 15(2):105-108.
[6] YU Y, LUO X, SU Q, et al. A dynamic lane-changing decision and trajectory planning model of autonomous vehicles under mixed autonomous vehicle and human-driven vehicle environment[J]. Physica A:Statistical Mechanics and Its Applications, 2023, 609:128361.
[7] HOUENOU A, BONNIFAIT P, CHERFAOUI V, et al. Vehicle trajectory prediction based on motion model and maneuver recognition[C]//2013 IEEE/RSJ International Conference on Intelligent Robots and Systems. Tokyo:IEEE,2013:4363-4369.
[8] JIANG Y, ZHU B, YANG S, et al. Vehicle trajectory prediction considering driver uncertainty and vehicle dynamics based on dynamic Bayesian network[J]. IEEE Transactions on Systems, Man, and Cybernetics:Systems, 2023,53(2):689-703.
[9] XIE D F, FANG Z Z, JIA B, et al. A data-driven lanechanging model based on deep learning[J]. Transportation Research Part C:Emerging Technologies, 2019, 106:41-60.
[10]季学武，费聪，何祥坤，等.基于LSTM网络的驾驶意图识别及车辆轨迹预测[J].中国公路学报, 2019, 32(6):34-42.JI Xuewu, FEI Cong, HE Xiangkun, et al. Intention recognition and trajectory prediction for vehicles using LSTM network[J]. China Journal of Highway and Transport, 2019, 32(6):34-42.
[11] DAI S, LI L, LI Z. Modeling vehicle interactions via modified LSTM models for trajectory prediction[J].IEEE Access, 20199, 7:38287-38296.
[12] SHI K, WU Y, SHI H, et al. An integrated car-following and lane changing vehicle trajectory prediction algorithm based on a deep neural network[J]. Physica A:Statistical Mechanics and Its Applications, 2022, 599:127303.
[13] HAO Z, HUANG X, WANG K, et al. Attention-based GRU for driver intention recognition and vehicle trajectory prediction[C]//2020 4th CAA International Conference on Vehicular Control and Intelligence(CVCI).Hangzhou:IEEE, 2020:86-91.
[14] QIE T, WANG W, YANG C, et al. Trajectory prediction method using deep learning for intelligent and connected vehicles[C]//2023 IEEE 6th International Conference on Industrial Cyber-Physical Systems(ICPS). Wuhan:IEEE, 2023:1-5.
[15] ZHANG X, SUN J, QI X, et al. Simultaneous modeling of car-following and lane-changing behaviors using deep learning[J]. Transportation Research Part C:Emerging Technologies, 2019, 104:287-304.
[16] ZHOU H, ZHANG S, PENG J, et al. Informer:beyond efficient transformer for long sequence time-series forecasting[EB/OL].(2020-12-12)[2023-08-01]. https://arxiv.org/abs/2012.07436.
[17] MOHD KAMARUDIN J A, ABDULLAH A, SALLEHUDDIN R. A review of deep learning architectures and their application[M]//Mohamed Ali M, WahidH, Mohd Subha N, et al. Modeling, design and simulation of systems. Singapore:Springer, 2017:83-94.
[18] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[EB/OL](2017-06-20)[2023-08-01].https://arxiv.org/abs/1706.03762.
[19]赵懂宇,王志建,宋程龙.基于Informer算法的网联车辆运动轨迹预测模型[J].计算机应用研究, 2023, 41(4)1-6.ZHAO Dongyu, WANG Zhijian, SONG Chenglong.Model of predicting motion trajectory of connected vehicles based on Informer algorithm[J]. Application Research of Computers, 2023, 41(4):1-6.
[20] KRAJEWSKI R, BOCK J, KLOEKER L, et al. The highD 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). Maui:IEEE, 2018:2118-2125.
[21] ALAHI A, GOEL K, RAMANATHAN V, et al. Social LSTM:human trajectory prediction in crowded spaces[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR). Las Vegas:IEEE, 2016:961-971.
[22] DEO N, TRIVEDI M M. Convolutional social pooling for vehicle trajectory prediction[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops(CVPRW). Salt Lake City:IEEE, 2018:1549-15498.
[23] CAI Y, WANG Z, WANG H, et al. Environment-attention network for vehicle trajectory prediction[J]. IEEE Transactions on Vehicular Technology, 2021, 70(11):11216-11227.
[24] MESSAOUD K, YAHIAOUI I, VERROUST-BLONDET A, et al. Attention based vehicle trajectory prediction[J].IEEE Transactions on Intelligent Vehicles, 2021, 6(1):175-185.
[25] WANG R, GAN B. Modeling and prediction of vehicle trajectory in the Internet of vehicles[C]//2022 4th International Conference on Artificial Intelligence and Advanced Manufacturing(AIAM). Hamburg:IEEE, 2022:36-40.
基本信息:
DOI：10.19961/j.cnki.1672-4747.2023.11.034
中图分类号:TP18;U463.6;U495
引用信息:
[1]吴红兰,胡德富,郭旭周.基于Informer的车辆多意图运动轨迹预测[J].交通运输工程与信息学报,2024,22(03):68-79.DOI:10.19961/j.cnki.1672-4747.2023.11.034.
基金信息:
国家自然科学基金项目（U2033201）; 航空科学基金项目（20181352009）
2024-08-21
2024-08-21
2024-08-21