基于博弈论的自动驾驶驶离专用车道换道决策模型

陆春意; 何赏璐; 高彬彬; 曹从咏; 范钲文

doi:10.3963/j.jssn.1674-4861.2024.04.016

基于博弈论的自动驾驶驶离专用车道换道决策模型

doi: 10.3963/j.jssn.1674-4861.2024.04.016

南京理工大学自动化学院南京 210094

基金项目:

国家自然科学基金项目 52102380

中国博士后科学基金特别资助项目 2021T140325

南京市重大科技专项项目 202209008

江苏省科技计划项目 BZ2023023

详细信息

作者简介:
陆春意(1998—)，硕士研究生. 研究方向：交通运输. E-mail：angranlu@njust.edu.cn

通讯作者:
何赏璐(1987—)，博士，副教授.研究方向：智能网联交通系统. E-mail：slhemickey@126.com

中图分类号: U491
计量
- 文章访问数: 57
- HTML全文浏览量: 25
- PDF下载量: 6
- 被引次数: 0
出版历程
- 收稿日期: 2024-03-09
- 网络出版日期: 2024-11-25

A Game Theory-based Lane Change Decision Model for Automated Vehicle Leaving the Freeway Dedicated Lane for Automated Driving

School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China

摘要

摘要: 在高速公路内侧车道设置自动驾驶专用车道的场景下，自动驾驶车辆驶离专用车道，并换道至匝道的过程中，频繁的换道行为会带来自动驾驶车和人工驾驶车交互冲突，对分流区产生交通扰动，进而带来安全风险以及影响交通效率。面向不同车道属性和自动驾驶混合车流，构建了基于博弈论和模型预测控制的自动驾驶换道决策模型。根据自动驾驶车辆到高速公路主线驶离点的距离和换道途径车道上可接受换道间隙的分布，引入换道危机感以量化自动驾驶车辆换道的难易程度。利用加速度变化率来量化混合车流中人工驾驶车辆风格，构建自动驾驶车和人工驾驶车不同类型车辆的模型预测控制成本函数。通过当前时刻的交通状况预测求解自动驾驶车辆下1个时刻的最优加速度，用斯塔克尔伯格博弈描述自动驾驶车辆在换道过程中与人工驾驶车辆的交互，以人工驾驶车辆的收益最大化为前提，选择最大的自动驾驶车辆收益，求解获得自动驾驶车辆的最优换道时间。搭建了基于Python和SUMO的联合仿真实验平台，设置4种不同车流密度的专用车道和混合车流交通场景，并与SUMO默认换道模型等2类模型进行了对比。结果表明：在所有不同车流密度的场景下，所研究的模型会选择合适的换道策略以保证速度损失的最小化，在速度方面均优于所对比的换道模型，换道过程的平均速度分别增加1.44%和11.81%，有效提升了自动驾驶车驶离专用车道的效率。
- 智能交通 /
- 混合车流 /
- 换道决策模型 /
- 斯塔克尔伯格博弈 /
- 自动驾驶专用车道
Abstract: Under the condition where there is dedicated lane for autonomous vehicles on the freeway, autonomous vehicles need to conduct a consequence lane-changing maneuvers to leave the dedicated lane as well as the freeway mainline. These series of actions would bring the conflicts between autonomous vehicles and human-driving vehicles, which would increase the risk of collisions but decrease the traffic efficiency of diversion area. This study developed a lane-changing decision model for autonomous vehicles based on game theory and model predictive control, considering the impact of different types of lane and mixed traffic conditions. A measure, lane changing crisis, is proposed for autonomous vehicles, which is calculated based on the distance from automated vehicle to the off-ramp and the distribution of acceptable lane change gap in adjacent lanes. The driving style of human-driving vehicle is valued by the variation of acceleration. The cost functions in the model predictive control method for different types of vehicles are formulated. The optimal acceleration of autonomous vehicles at the next time interval could be predicted based on the current traffic state. The interaction between autonomous vehicle and human-driving vehicle is described by the Stackelberg game. The lane changing decision is made to maximize the benefit of autonomous vehicle when the benefit of human-driving vehicle is the largest. And the optimal acceleration of autonomous vehicle at the next time interval is also achieved. A simulation platform integrating Python and SUMO is constructed. Four experiment scenarios with different traffic densities under different lane types and mixed traffic conditions are set. And the proposed model has been also compared with the other two lane-changing models, e.g. the default model within SUMO. The results indicate that the proposed model could consistently make optimal lane-changing decision to minimize the loss of speed. And compared with the other lane-changing models, the proposed model could increase the average speed during lane changes increasing by 1.44% and 11.81%, respectively, which validate the outperformance of the proposed method.
- Intelligent transportation /
- Mixed traffic /
- Lane change decision model /
- Stackelberg game /
- Freeway dedicated lane

HTML全文

图 1 所研究交通场景示意图

Figure 1. The studied traffic scene

下载: 全尺寸图片幻灯片

图 2 算法总体流程图

Figure 2. The flowchart of proposed algorithm

下载: 全尺寸图片幻灯片

图 3 AV换道效益算法流程图

Figure 3. AV lane change benefit algorithm flow chart

下载: 全尺寸图片幻灯片

图 4 换道间隙选择示意图

Figure 4. Schematic diagram of lane change gap selection

下载: 全尺寸图片幻灯片

图 5 HV受换道影响效益算法流程图

Figure 5. Flow chart of HV algorithm affected by lane change

下载: 全尺寸图片幻灯片

图 6 仿真场景图

Figure 6. Simulation scene diagram

下载: 全尺寸图片幻灯片

图 7 仿真结果对比图

Figure 7. Comparison chart of simulation results

下载: 全尺寸图片幻灯片

表 1 实验参数设置

Table 1. Experimental parameters

参数名	取值	参数名	取值
P_v	3	Δa_min/（m/s²）	-1
P_a	1	Δa_max/（m/s²）	1
P_N'	1	v_esp/（(m/s）	33.33
P_cN'	1	P_SS	1.4
N	20	P_Sf	1.4
a_min/（m/s²）	-2.5	P_acc	0.256
a_max/（m/s²）	-2.5	x_exit/m	2 000
t_max	20	R	0.8

下载: 导出CSV

参考文献(23)

[1]	ZHANG J, WU K, CHENG M, et al. Safety evaluation for connected and autonomous vehicles' exclusive lanes considering penetrate ratios and impact of trucks using surrogate safety measures[J]. Journal of Advanced Transportation, 2020, 22(1): 1-16.
[2]	CHEN X, LIN X, HE F, et al. Modeling and control of automated vehicle access on dedicated bus rapid transit lanes[J]. Transportation Research Part C: Emerging Technologies, 2020, 120: 102795. doi: 10.1016/j.trc.2020.102795
[3]	MARTIN G M, ELEFTERIADOU L. Traffic management with autonomous and connected vehicles at single-lane roundabouts[J]. Transportation Research Part C: Emerging Technologies, 2021, 125: 102964. doi: 10.1016/j.trc.2021.102964
[4]	ALOTIBI F, ABDELHAKIM M. Anomaly detection for cooperative adaptive cruise control in autonomous vehicles using statistical learning and kinematic model[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 22(6): 3468-78. doi: 10.1109/TITS.2020.2983392
[5]	张健, 李青扬, 李丹, 等. 基于深度强化学习的自动驾驶车辆专用道汇入引导[J]. 吉林大学学报(工学版), 2023, 53 (9): 2508-2518. ZHANG J, LING D Y, LING D, et al. Merging quidance of exclusive lanes for connected and autonomous vehicles based on deep reinforcement learning[J]. Journal of Jilin University (Engineering and Technology Edition), 2023, 53(9): 2508-2518. (in Chinese)
[6]	杨敏, 王立超, 张健, 等. 面向智慧高速的合流区协作车辆冲突解脱协调方法[J]. 交通运输工程学报, 2020, 20(3): 217-24. YANG M, WANG L C, ZHANG J, et al. Collaborative method of vehicle conflict resolution in merging area for intelligent expressway[J]. Journal of Traffic and Transportation Engineering, 2020, 20(3): 217-24. (in Chinese)
[7]	DING N, MENG X, XIA W, et al. Multivehicle coordinated lane change strategy in the roundabout under internet of vehicles based on game theory and cognitive computing[J]. IEEE Transactions on Industrial Informatics, 2020, 16 (8) : 5435-43. doi: 10.1109/TII.2019.2959795
[8]	NAN D, MARIA A, XIANGHUA M, et al. Mandatory lane change strategy in VANET based on coordinated stackelberg game model[C]. 2020 Chinese Control and Decision Conference, Hefei, China: IEEE, 2020.
[9]	ZHOU S T, ZHANG J, LI L H, et al. Discretionary lane change model for intelligent connected vehicles on expressway[J]. 19^th COTA International Conference of Transportation Professionals, Nanjing, China: COTA, 2019.
[10]	张可琨, 曲大义, 宋慧, 等. 自动驾驶车辆换道博弈策略分析及建模[J]. 复杂系统与复杂性科学, 2023, 20(2): 60-67. ZHANG K K, QUN D Y, SONG H, et al. Analysis and modeling for lane-changing game strategy of autonomous vehicles[J]. Complex Systems and Complexity Science, 2023, 20 (2): 60-67. (in Chinese)
[11]	ZHANG Q, FILEV D, TSENG H E, et al. Addressing mandatory lane change problem with game theoretic model predictive control and fuzzy markov chain[C]. 2018 Annual American Control Conference, Milwaukee, WI, USA: AACC, 2018.
[12]	PENG J, GUO Y, SHAO Y. Lane change decision analysis based on drivers' perception-judgment and game theory[J]. Applied Mechanics and Materials, 2013, 361-363: 1875-1879. doi: 10.4028/www.scientific.net/AMM.361-363.1875
[13]	崔冰艳, 李贺, 崔哲, 等. 智能网联汽车换道决策安全性研究综述[J]. 交通信息与安全, 2023, 41(4): 1-13. doi: 10.3963/j.jssn.1674-4861.2023.04.001 CUI B Y, LI H, CUI Z, et al. A review of safety studies on lane change decision-makings for connected automated vehicles[J]. Journal of Transport Information and Safety, 2023, 41 (4): 1-13. (in Chinese) doi: 10.3963/j.jssn.1674-4861.2023.04.001
[14]	王竹青. 混合驾驶环境下基于博弈论的车辆换道策略研究[D]. 西安: 长安大学, 2020. WANG Z Q. Research on lane-changing strategy of vehicles based on game theory in mixed driving environment[D]. Xi'an: Chang'an University, 2020. (in Chinese)
[15]	刘思源, 喻伟, 刘洁莹, 等. 考虑驾驶风格的车辆换道行为及预测模型[J]. 长沙理工大学学报(自然科学版), 2019, 16 (1): 28-35. LIU X Y, YU W, LIU J Y, et al. Characteristics analysis and prediction model of lane changing behavior under different driving styles[J]. Journal of Changsha University of Science and Technology(Natural Science), 2019, 16(1): 28-35. (in Chinese)
[16]	WANG C, ZHANG J, XU L, et al. A new solution for freeway congestion: cooperative speed limit control using distributed reinforcement learning[J]. IEEE Access, 2019, 7: 41947-41957. doi: 10.1109/ACCESS.2019.2904619
[17]	何廷全, 宋浪, 俞山川. 高速公路主线提前换道与入口匝道协同控制研究[J]. 公路, 2023, 68(3): 288-93. HE T Q, SONG L, YU S C. Research on the cooperative control of the main line of expressway in advance and the entrance ramp[J]. Highway, 2023, 68(3): 288-93. (in Chinese)
[18]	DONG C, WANG H, LI Y, et al. Route control strategies for autonomous vehicles exiting to off-ramps[J]. IEEE Transactions on Intelligent Transportation Systems, 2020, 21(7): 3104-3116. doi: 10.1109/TITS.2019.2925319
[19]	CAO Z, YANG D, XU S, et al. Highway exiting planner for automated vehicles using reinforcement learning[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 22 (2): 990-1000. doi: 10.1109/TITS.2019.2961739
[20]	HAO W, ZHANG Z, GAO Z, et al. Research on mandatory lane-changing behavior in highway weaving sections[J]. Journal of Advanced Transportation, 2020, 2020: 1-9.
[21]	KIM Y S, SAEKOH J, KIM J H. Flexible distance maintenance of autonomous vehicle in accordance with lane change of lateral position vehicle[C]. 11^th International Conference on Ubiquitous Robots and Ambient Intelligence, Kuala Lumpur, Malaysia: IEEE, 2014.
[22]	MURPHEY Y L, MILTON R, KILIARIS L. Driver's style classification using jerk analysis[C]. 2009 IEEE Workshop on Computational Intelligence in Vehicles and Vehicular Systems, Nashville, TN, USA: IEEE, 2009.
[23]	WANG R, LUKIC S M. Review of driving conditions prediction and driving style recognition based cont rol algorithms for hybrid electric vehicles[C]. 2011 IEEE Vehicle Power and Propulsion Conference, Chicago, IL, USA: IEEE, 2011.