To achieve precise prediction of surrounding vehicle trajectories for autonomous driving systems, a spatio-temporal feature fusion prediction model based on Transformer and Graph Attention Network (GAT) was proposed. Non-euclidean data input is supported by this model, and dynamic adaptation to variations in the number of surrounding vehicles is realized. A dual-branch feature extraction mechanism is adopted in the model architecture. Temporal dependency features of vehicles' historical trajectories are captured by the Transformer encoder, and spatial interaction relationships between vehicles are modeled by the GAT network. Subsequently, deep fusion of spatio-temporal features is implemented by the Transformer decoder, and prediction results of future trajectories are output by the fully-connected layer decoder. Experiments are conducted on the public NGSIM dataset. Compared with the LSTM-GAT baseline model, the trajectory root mean square errors (RMSE) of the proposed model are reduced by 13.0%, 10.2%, 14.3%, 18.2% and 21.6% at the prediction horizons of 1 s, 2 s, 3 s, 4 s and 5 s respectively. In comparison with the Transformer-CNN model, remarkable improvements in long-horizon prediction performance at 3 s, 4 s and 5 s are achieved, with corresponding error reductions of 27.2%, 44.3% and 49.4%. The effectiveness and superiority of the proposed model for vehicle trajectory prediction tasks are verified by the experimental results.
HuangY, DuJ, YangZ, et al. A survey on trajectory-prediction methods for autonomous driving[J]. IEEE Transactions on Intelligent Vehicles, 2022, 7(3): 652-674.
[2]
LytrivisP, ThomaidisG, AmditisA. Cooperative path prediction in vehicular environments[C]∥2008 11th International IEEE Conference on Intelligent Transportation Systems, Beijing, China, 2008: 803-808.
JiXue-wu, FeiCong, HeXiang-kun, et al. Intention recognition and trajectory prediction for vehicles using LSTM network[J]. China Journal of Highway and Transport, 2019, 32(6): 34-42.
FangHua-zhen, LiuLi, Xiaoxiao-feng, et al. Vehicle trajectory prediction based on mixed teaching force long short-term memory[J]. Journal of Transportation Systems Engineering and Information Technology, 2023, 23(4): 80-87.
GaoZhen-hai, BaoMing-xi, GaoFei, et al. The method of probabilistic multi-modal expected trajectory prediction based on LSTM[J]. Automotive Engineering, 2023, 45(7): 1145-1152, 1162.
[9]
DeoN, TrivediM M. Convolutional social pooling for vehicle trajectory prediction[C]∥IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Salt Lake City, UT, USA, 2018: 1468-1476.
WuYi-kai, HuQi-zhou, WuXiao-yu. Vehicle trajectory prediction model in the context of Internet of vehicles[J]. Journal of Southeast University(Natural Science Edition), 2022, 52(6): 1199-1208.
[12]
MoX, XingY, LvC. Graph and recurrent neural network-based vehicle trajectory prediction for highway driving[C]∥International Intelligent Transportation Systems Conference (ITSC), Indianapolis, IN, USA, 2021: 1934-1939.
[13]
VaswaniA, ShazeerN, ParmarN, et al. Attention is all you need[C]∥Advances in Neural Information Processing Systems, Long Beach California, USA, 2017: 5998-6008.
WuZ, PanS, ChenF, et al. A comprehensive survey on graph neural networks[J]. IEEE Transactions on Neural Networks and Learning Systems, 2020, 32(1): 4-24.
MaShuai, LiuJian-wei, ZuoXin. Survey on graph neural network[J]. Journal of Computer Research and Development, 2022, 59(1): 47-80.
[18]
AlahiA, GoelK, RamanathanV, et al. Social LSTM: human trajectory prediction in crowded spaces[C]∥IEEE Conference on Computer Vision And Pattern Recognition, Las Vegas, NV, USA, 2016: 961-971.
TianYan-tao, HuangXin, LuHui-qiu,et al. Multi⁃mode behavior trajectory prediction of surrounding vehicle based on attention and depth interaction[J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(5): 1474-1480.
HuangJing, LiuXiang-zhen, DengXiao-yang, et al. Research on intelligent vehicle trajectory planning based on multimodal trajectory prediction[J]. Automotive Engineering,2024,46(6): 965-974, 1024.
[23]
ChenX, ZhangH, ZhaoF, et al. Intention-aware vehicle trajectory prediction based on spatial-temporal dynamic attention network for internet of vehicles[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(10): 19471-19483.
ChengYing, GaoLi, GaoXian-ping. Changes in attention allocation characteristics of drivers driving on expressway[J]. China Safety Science Journal, 2014,24(10): 71-76.