面向多障碍物场景的车辆紧急避撞耦合决策与轨迹规划方法

doi:10.3969/j.issn.1674-8484.2025.06.014

摘要/Abstract

摘要： 为提高车辆在多障碍物高速场景下的紧急避撞能力，解决决策规划因计算复杂而难以实时响应的问题，提出了一种耦合决策与轨迹规划的一体化框架。通过建立多车非合作博弈模型描述动态交互行为，并设计基于威胁评估的顺序决策机制，将高维博弈问题简化为序列化的单障碍物交互过程；基于深度学习框架PyTorch实现图形处理器（GPU）加速轨迹优化算法，在满足车辆动力学约束的同时生成安全、舒适的避撞轨迹。结果表明：该方法在典型高速场景下的决策平均计算时间为20~50 ms，轨迹规划耗时33.1~149.1 ms，优于传统模型预测控制（MPC）方法；规划轨迹横向速度与加速度分别控制在4.0 m/s和4.0 m/s²以内，符合安全性和舒适性要求。对规划轨迹进行跟踪，横向跟踪误差和速度误差最大分别为0.22 m和0.59 m/s，满足高速紧急避撞要求，在CARLA仿真中，所有场景均成功避撞。所提框架能有效平衡决策最优性与实时性，为复杂场景下的车辆主动安全提供可靠解决方案。

关键词: 汽车主动安全, 复杂场景, 紧急避撞决策, 轨迹规划与跟踪, 深度学习框架PyTorch

Abstract:

An integrated framework coupling decision-making with trajectory planning was proposed to enhance the emergency collision avoidance capability of vehicles in high-speed multi-obstacle scenarios and address the challenge of real-time responsiveness in decision-making and planning due to computational complexity. The high-dimensional game problem was simplified into a sequence of single-obstacle interaction processes by establishing a multi-vehicle non-cooperative game model to describe dynamic interactions and designing a sequential decision-making mechanism based on threat assessment. A graphics processing unit (GPU)-accelerated trajectory optimization algorithm was implemented using the open source machine learning framework PyTorch, generating safe and comfortable collision avoidance trajectories while satisfying vehicle dynamic constraints. The results show that the average decision-making computation time of the proposed method in typical high-speed scenarios is 20~50 ms, and trajectory planning takes 33.1~149.1 ms, outperforming traditional model predictive control (MPC) methods. The lateral velocity and acceleration of the planned trajectories are controlled within 4.0 m/s and 4.0 m/s², respectively, meeting safety and comfort requirements. When tracking the planned trajectories, the maximum lateral tracking error and speed error are 0.22 m and 0.59 m/s, respectively, fulfilling the requirements for high-speed emergency collision avoidance. In CARLA simulations, successful collision avoidance is achieved in all scenarios. The conclusion demonstrates that the proposed framework effectively balances decision-making optimality and real-time performance, providing a reliable solution for vehicle active safety in complex scenarios.

Key words: automotive active safety, complex scenarios, emergency collision avoidance decision-making, trajectory planning and tracking, open source machine learning framework PyTorch

中图分类号:

关永学, 刘森海, 韩勇, 徐莉, 舒伟斌, 樊晨旭. 面向多障碍物场景的车辆紧急避撞耦合决策与轨迹规划方法[J]. 汽车安全与节能学报, 2025, 16(6): 945-954.

GUAN Yongxue, LIU Senhai, HAN Yong, XU Li, SHU Weibin, FAN Chenxu. A coupled decision-making and trajectory planning approach for vehicle emergency collision avoidance in multi-obstacle scenarios[J]. Journal of Automotive Safety and Energy, 2025, 16(6): 945-954.

图/表 12

图1 场景工况

场景	v₀ km·h^-1	v₁ km·h^-1	v₂ km·h^-1	a₁ m·s^-2	L₁ m	L₂ m
前车匀速	120	60	60	0	50	90
前车紧急制动	120	120	90	-4	8	70
道路变窄	120	90	0	0	50	80
相邻车紧急切入	120	100	50	0	30	80
前车紧急切出	120	100	50	0	30	70

表1 各场景工况参数

图2 2种场景下的轨迹规划

图3 2种场景下的主车横向速度和加速度变化

图4 3种场景下的轨迹规划

图5 3种场景下的主车横向速度和加速度变化

图6 轨迹跟踪

图7 横向跟踪误差和速度跟踪误差

场景	L_em / m	E_vm / ( m·s^-1)
前车匀速	0.22	0
前车紧急制动	0.18	0
道路变窄	0.17	0
相邻车紧急切入	0.17	0
前车紧急切出	0.13	0.59

表2 各场景工况下的跟踪误差

图8 主车转向角和加速度变化

图9 前车匀速场景避撞

图10 车道变窄和前车切出场景避撞

参考文献 26

[1]	朱西产, 魏昊舟, 马志雄. 基于自然驾驶数据的跟车场景潜在风险评估[J]. 中国公路学报, 2020, 33(4): 169-181. doi: 10.19721/j.cnki.1001-7372.2020.04.017
	ZHU Xichan, WEI Haozhou, MA Zhixiong. Potential risk assessment of car-following scenarios based on naturalistic driving data[J]. Chin J Highw Transport, 2020, 33(4): 169-181. (in Chinese) doi: 10.19721/j.cnki.1001-7372.2020.04.017
[2]	郑修磊. 自动驾驶车辆紧急主动避撞控制算法研究[D]. 长春: 吉林大学, 2022.
	ZHENG Xiulei. Research on emergency active collision avoidance control algorithm for autonomous vehicles[D]. Changchun: Jilin University, 2022. (in Chinese)
[3]	王璐瑶. 交互场景下智能车运动预测与规划一体化方法研究[D]. 长春: 吉林大学, 2024.
	WANG Luyao. Integrated Motion prediction and planning for intelligent vehicles in interactive scenarios[D]. Changchun: Jilin University, 2024. (in Chinese)
[4]	DENG Weiwen, ZHAO Qingrong, SUN Hao, et al. Dynamic trajectory planning for vehicle autonomous driving[C]// 2013 IEEE Int’l Conf Syst, Man, Cybe, IEEE, 2013: 4161-4166.
[5]	《中国公路学报》编辑部. 中国汽车工程学术研究综述· 2023[J]. 中国公路学报, 2023, 36(11): 1-192. doi: 10.19721/j.cnki.1001-7372.2023.11.001
	Editorial Department of China Journal of Highway and Transport. Review of academic research on automotive engineering in China 2023[J]. Chin J Highw Transport, 2023, 36(11): 1-192. (in Chinese)
[6]	National Highway Traffic Safety Administration (NHTSA). Fatality analysis reporting system (FARS) analytic reference guide: 1975-2021[R]. Washington, DC: U.S. Department of Transportation, 2023.
[7]	田浩东, 赵树廉, 古昕昱, 等. 基于驾驶行为基元的典型驾驶场景识别与提取方法[J]. 汽车工程学报, 2023, 13(5): 676-686.
	TIAN Haodong, ZHAO Shulian, GU Xinyu, et al. Typical driving scenario recognition and extraction method based on driving behavior primitives[J]. Chin J Autom Engi, 2023, 13(5): 676-686. (in Chinese)
[8]	DA Jiang, DU Ling, LI Shuhui, et al. An improved dynamic window approach based on reinforcement learning for the trajectory planning of automated guided vehicles[J]. IEEE Access, 2024, 12: 36016-36025. doi: 10.1109/ACCESS.2024.3373446 URL
[9]	林汐鸿, 聂枝根, 黄井轩, 等. 考虑前车插入的车辆队列非合作博弈决策[J]. 计算机仿真, 2025, 42(5): 176-181.
	LIN Xihong, NIE Zhigen, HUANG Jingxuan, et al. Non-cooperative game decision making for vehicle platoons considering preceding vehicle cut-in[J]. Comput Simul, 2025, 42(5): 176-181. (in Chinese)
[10]	HU Manjiang, XIE Guotao, GAO Hongbo, et al. Manoeuvre prediction and planning for automated and connected vehicles based on interaction and gaming awareness under uncertainty[J]. *IET Intel Trans Syst*, 2019, 13(6): 933-941.
[11]	Fudenberg D, Tirole J. Game Theory[M]. Cambridge, MA: MIT Press, 1991: 1-30.
[12]	Kreps D M. The New Palgrave Dictionary of Economics[M]. London, UK: Palgrave Macmillan, 2018: 9251-9258.
[13]	Werling M, Ziegler J, Kammel S, et al. Optimal trajectory generation for dynamic street scenarios in a frenét frame[C]// 2010 IEEE Int’l Conf Robot Autom. Anchorage: IEEE, 2010: 987-993.
[14]	Arbis D, Dixit V V. Game theoretic model for lane changing: Incorporating conflict risks[J]. Accid Analy Prev, 2019, 125: 158-164.
[15]	姚军. 基于前车换道意图辨识的智能巡航控制算法研究[D]. 长春: 吉林大学, 2020.
	YAO Jun. Research on intelligent cruise control algorithm based on preceding vehicle lane change intention recognition[D]. Changchun: Jilin University, 2020. (in Chinese)
[16]	BAI Guoxing, MENG Yu, LIU Li, et al. Review and comparison of path tracking based on model predictive control[J]. Elect, 2019, 8(10): 1077-1109.
[17]	LI Guoqiang, ZHANG Xudong, GUO Hongliang, et al. Real-time optimal trajectory planning for autonomous driving with collision avoidance using convex optimization[J]. Autom Inno, 2023, 6(3): 481-491.
[18]	Paszke A, Lerer A, Killeen T, et al. PyTorch: An imperative style, high-performance deep learning library[C]// Advan Neur Info Proc Syst 32. Vancr, Curr Asso, Inc, 2019: 8024-8035.
[19]	Paszke A, Chaurasia A, Kim S, et al. ENet: A deep neural network architecture for real-time semantic segmentation[J/OL]. arXiv preprint, 2016: arXiv:1606.02147.
[20]	Euro NCAP. 2026 Euro NCAP test protocol-safe driving (Version 1.0)[R]. Brussels: European New Car Assessment Programme, 2024.
[21]	余志生. 汽车理论(5版)[M]. 北京: 机械工业出版社, 2009: 105-112.
	YU Zhisheng. Automobile Theory (5th edit)[M]. Beijing: China Machine Press, 2009: 105-112. (in Chinese)
[22]	Kingma D P, Ba J. Adam: A method for stochastic optimization[J/OL]. arXiv preprint, 2014: arXiv:1412.6980.
[23]	Kuang M L, Fodor M, Hrovat D, et al. Hydraulic brake system modeling and control for active control of vehicle dynamics[C]// 1999 Amer Contr Conf. San Diego: IEEE, 1999: 4538-4542.
[24]	何川. 基于PreScan的车辆自动紧急制动方法研究[D]. 邯郸: 河北工程大学, 2022.
	HE Chuan. Research on vehicle automatic emergency braking method based on PreScan[D]. Handan: Hebei University of Engineering, 2022. (in Chinese)
[25]	Sledge Jr N H. An investigation of vehicle critical speed and its influence on lane-change trajectories[D]. Austin: The University of Texas at Austin, 1997.
[26]	高绍元, 吴长水. 基于加权变时域MPC的无人车辆轨迹跟踪控制[J]. 上海工程技术大学学报, 2024, 38(2): 123-131.
	GAO Shaoyuan, WU Changshui. Trajectory tracking control of unmanned vehicles based on weighted variable horizon MPC[J]. J Shanghai Univ Engi Sci, 2024, 38(2): 123-131. (in Chinese)