[1]崔建明,蔺繁荣,张 迪,等.基于有向图的强化学习自动驾驶轨迹预测[J].郑州大学学报(工学版),2023,44(05):53-61.[doi:10.13705/j.issn.1671-6833.2023.05.002]
　CUI Jianming,LIN Fanrong,ZHANG Di,et al.Reinforcement Learning Autonomous Driving Trajectory Prediction Based on Directed Graph[J].Journal of Zhengzhou University (Engineering Science),2023,44(05):53-61.[doi:10.13705/j.issn.1671-6833.2023.05.002]
点击复制

基于有向图的强化学习自动驾驶轨迹预测()
分享到：

《郑州大学学报(工学版)》[ISSN:1671-6833/CN:41-1339/T]

卷:
44
期数:
2023年05期
页码:
53-61
栏目:
出版日期:
2023-08-20

文章信息/Info

Title:
Reinforcement Learning Autonomous Driving Trajectory Prediction Based on Directed Graph
作者:
崔建明1 ;  蔺繁荣1 ;  张 迪1 ;  张路宁1 ;  刘 铭2
1. 长安大学 信息工程学院,陕西 西安 710018;2. 国家计算机网络应急技术处理协调中心,北京 100029
Author(s):
CUI Jianming1;  LIN Fanrong1;  ZHANG Di1 ;  ZHANG Luning1;  LIU Ming2
1. School of Information Engineering,Chang′an University, Xi′an 710018, China; 2. National Computer Network Emergency Response Technical Team / Coordination Center of China, Beijing 100029, China
关键词:
自动驾驶;  轨迹预测;  有向图;  强化学习;  GAIL;  注意力机制;  多模态预测
Keywords:
autonomous driving;  trajectory prediction;  directed graph;  reinforcement learning;  GAIL;  attention mechanism;  multimodal predictio
分类号:
O211. 62；TP183
DOI:
10.13705/j.issn.1671-6833.2023.05.002
文献标志码:
A
摘要:
轨迹预测作为自动驾驶中的重要组成部分,旨在对车辆进行行驶估计,以便车辆根据行驶估计进行路径 规划,从而做出安全准确的决策。 首先,为提升车辆轨迹预测精度,采用有向图方法构建高清驾驶场景地图,有向 图方法将地图信息矢量化,以便有效提取地图拓扑结构;其次,采用生成对抗模仿学习( GAIL) 通过生成器与判别 器的对抗博弈学习数据集驾驶策略,从而根据当前状态采取对应驾驶行为;最后,通过采样遍历得到多模态预测轨 迹方案。 在 nuScenes 运动预测数据集上进行仿真,量化结果显示相比于其他方法,K = 5 时,最小最终位移误差 MinFDE5 提高了 10. 8%;K = 10 时,最小最终位移误差 MinFDE10 提高了 17. 53%,最小平均位移误差 MinADE10 提高了 9. 52%,失误率 MissRate10 减少了 28. 26%。 评估结果表明:生成的轨迹多模态符合场景基本结构,且准确度得到提高。
Abstract:
As an important part of autonomous driving, trajectory prediction aimed to forcast the vehicle′s driving path, so that the vehicle could make path planning according to the driving estimation, so as to make safe and accurate decisions. Firstly, in order to improve the accuracy of vehicle trajectory prediction, the directed graph method was used to construct a high-definition driving scene map, and the directed graph method vectorized the map information to effectively extract the map topology. Secondly, GAIL was used to learn the driving strategy of the dataset through the confrontation game between the generator and the discriminator, so as to adopt the corresponding driving behavior according to the current state. Finally, the multimodal prediction trajectory scheme was obtained by sampling traversal. Simulation was carried out on the nuScenes motion prediction dataset. The quantitative results showed that compared with other methods, when K = 5, the minimum final displacement error MinFDE5 was increased by 10. 8%; when K = 10, the minimum fianl displacement error MinFDE10 increased by 17. 53%, the minimum average displacement error MinADE10 increased by 9. 52%, and the error rate MissRate10 decreased by 28. 26%. The evaluation showed that the generated trajectories were multimodal, could conform to the basic structure of the scene, with improved accuracy.

参考文献/References:

[1] 中华人民共和国国家发展和改革委员会. 智能汽车创 新发展战略[EB / OL] . (2020-02-24) [ 2022-12-24] . https: ∥ www. ndrc. gov. cn / xxgk / zcfb / tz/ 202002 / t20200224_1235917. html. 

National Development and Reform Commission . Innovative development strategy for intelligent vehicles [ EB / OL] . ( 2020 - 02 - 24 ) [ 2022 - 12 - 24 ] https: ∥ www. ndrc. gov. cn / xxgk / zcfb / tz/ 202002 / t20200224_12 35917. html. 
[2] BANSAL M, KRIZHEVSKY A, OGALE A. ChauffeurNet: learning to drive by imitating the best and synthesizing the worst[ EB / OL] . ( 2018 - 12 - 07) [ 2022 - 12 - 24] . https:∥arxiv. org / abs/ 1812. 03079. 
[3] 王丙琛, 司怀伟, 谭国真. 基于深度强化学习的自动 驾驶车控制算法研究[ J] . 郑州大学学报( 工学版) , 2020, 41(4) : 41-45, 80. 
WANG B C, SI H W, TAN G Z. Research on autopilot control algorithm based on deep reinforcement learning [ J] . Journal of Zhengzhou University ( Engineering Science) , 2020, 41(4) : 41-45, 80. 
[4] HO J, ERMON S. Generative adversarial imitation learning[C]∥Proceedings of the 30th International Conference on Neural Information Processing Systems. New York: ACM, 2016: 4572-4580. 
[5] CAESAR H, BANKITI V, LANG A H, et al. nuScenes: a multimodal dataset for autonomous driving [ C] ∥2020 IEEE / CVF Conference on Computer Vision and Pattern Recognition ( CVPR) . Piscataway: IEEE, 2020: 11618 -11628. 
[6] LIANG M, YANG B, HU R, et al. Learning lane graph representations for motion forecasting[ C]∥Computer Vision-ECCV 2020: 16th European Conference. New York: ACM, 2020: 541-556.
 [7] KIPF T N, WELLING M. Semi-supervised classification with graph convolutional networks [ EB / OL]. ( 2016 - 09 - 09)[2022-12-24]. https:∥arxiv. org / abs/ 1609. 02907. 
[8] VELICˇKOVIC ’ P, CUCURULL G, CASANOVA A, et al. Graph attention networks [ EB / OL ] . ( 2017 - 10 - 30 ) [2022-12-24] . https:∥arxiv. org / abs/ 1710. 10903. 
[9] 张三川, 马啸. 基于轨迹加权预测的主动避撞安全距 离模型及算法 [ J] . 郑 州 大 学 学 报 ( 工 学 版) , 2022, 43(3) : 104-110. 
ZHANG S C, MA X. A safe distance model and algorithm for active collision avoidance based on weighted prediction of trajectory[ J] . Journal of Zhengzhou University(Engineering Science) ,2022, 43(3) : 104-110. 
[10] WANG C H, WANG Y C, XU M Z, et al. Stepwise goal-driven networks for trajectory prediction [ C]∥IEEE Robotics and Automation Letters. Piscataway: IEEE, 2022: 2716-2723.
 [11] KIM B, PARK S H, LEE S, et al. LaPred: lane-aware prediction of multi-modal future trajectories of dynamic agents[C]∥2021 IEEE / CVF Conference on Computer Vision and Pattern Recognition ( CVPR ) . Piscataway: IEEE, 2021: 14631-14640. 
[12] CHAI Y N, SAPP B, BANSAL M, et al. MultiPath: multiple probabilistic anchor trajectory hypotheses for behavior prediction[EB / OL] . ( 2019-10- 12) [ 2022- 12 -24] . https:∥arxiv. org / abs/ 1910. 05449. 
[13] DEO N, TRIVEDI M M. Trajectory forecasts in unknown environments conditioned on grid-based plans[ EB / OL] . (2021-04-29) [ 2022-12-24] . https:∥arxiv. org / abs/ 2001. 00735.
 [14] GILLES T, SABATINI S, TSISHKOU D, et al. GOHOME: graph-oriented heatmap output for future motion estimation[C]∥2022 International Conference on Robotics and Automation ( ICRA) . New York: ACM, 2022: 9107-9114. 
[15] MESSAOUD K, DEO N, TRIVEDI M M, et al. Trajectory prediction for autonomous driving based on multi-head attention with joint agent-map representation [ C] ∥2021 IEEE Intelligent Vehicles Symposium ( Ⅳ) . New York: ACM, 2021: 165-170.

相似文献/References:

[1]王丙琛,司怀伟,谭国真.基于深度强化学习的自动驾驶车控制算法研究[J].郑州大学学报(工学版),2020,41(04):41.[doi:10.13705/j.issn.1671-6833.2020.04.002]
　WANG Bingchen,SI Huaiwei,TAN Guozhen.Research on Autopilot Control Algorithms Based on Deep Reinforcement Learning[J].Journal of Zhengzhou University (Engineering Science),2020,41(05):41.[doi:10.13705/j.issn.1671-6833.2020.04.002]
[2]刘明剑,朱云鹤,张思佳,等.基于最大团理论的自治交叉路口控制方法[J].郑州大学学报(工学版),2024,45(02):114.[doi:10.13705/j.issn.1671-6833.2023.05.004]
　LIU Mingjian,ZHU Yunhe,ZHANG Sijia,et al.Autonomous Intersection Control Method Based on Maximum Clique Theory[J].Journal of Zhengzhou University (Engineering Science),2024,45(05):114.[doi:10.13705/j.issn.1671-6833.2023.05.004]
[3]汤林东,云利军,罗瑞林,等.基于改进 YOLOv5s 的复杂道路交通目标检测算法[J].郑州大学学报(工学版),2024,45(03):64.[doi:10. 13705 / j. issn. 1671-6833. 2024. 03. 016]
　TANG Lindong,YUN Lijun,LUO Ruilin,et al.Complex Road Traffic Target Detection Algorithm Based on Improved YOLOv5s[J].Journal of Zhengzhou University (Engineering Science),2024,45(05):64.[doi:10. 13705 / j. issn. 1671-6833. 2024. 03. 016]

更新日期/Last Update: 2023-09-04