| 引用本文: | 程国柱,路程程,方丽庄,尤志栋,潘屹磊.高速公路CAV专用车道出口换道意图识别与博弈决策模型[J].哈尔滨工业大学学报,2026,58(7):62.DOI:10.11918/202412043 |
| Cheng Guozhu,Lu Chengcheng,Fang Lizhuang,You Zhidong,Pan Yilei.Lane change intention recognition and game decision-making model for dedicated CAV lanes on freeways[J].Journal of Harbin Institute of Technology,2026,58(7):62.DOI:10.11918/202412043 |
|
| 摘要: |
| 为保证智能网联自动驾驶车辆(connected and autonomous vehicle,CAV)可安全高效进行多次换道,针对高速公路上CAV在专用车道出口这一特定场景,分析CAV的换道行为,以车道效用和换道迫切度之和超过阈值作为触发换道意图的条件,构建了高速公路CAV专用车道出口的强制换道意图识别模型。运用博弈理论,将车辆多次换道过程根据冲突点位置的不同,划分成两阶段,形成每个阶段的行动策略,将安全、速度、舒适度三方面作为博弈决策的评价指标,建立CAV博弈决策模型收益矩阵,建立收益函数,构建换道博弈决策模型。采用PreScan和MATLAB/Simulink进行仿真验证,在不同工况下,对换道决策模型进行仿真实验。结果显示,初始速度分别设置为30 m/s和31 m/s时,两种换道场景下的博弈决策模型在换道成功率方面表现更为优异。 |
| 关键词: 高速公路 CAV专用车道 强制换道 博弈论 换道决策 |
| DOI:10.11918/202412043 |
| 分类号:U491 |
| 文献标识码:A |
| 基金项目:道路交通安全公安部重点实验室开放课题(2024ZDSYSKFKT01-1);中央高校基本科研业务费专项资金(2572023CT21) |
|
| Lane change intention recognition and game decision-making model for dedicated CAV lanes on freeways |
|
Cheng Guozhu1,Lu Chengcheng1,Fang Lizhuang2,You Zhidong2,Pan Yilei3
|
|
(1.School of Civil Engineering & Transportation, Northeast Forestry University, Harbin 150040, China; 2.Traffic Management Research Institute of the Ministry of Public Security, Wuxi 214151, Jiangsu, China; 3.Beidahuang Group Co., Ltd, Harbin 150036, China)
|
| Abstract: |
| In order to ensure that the connected and autonomous vehicle (CAV) can safely and efficiently change lanes many times, aiming at the specific scene of CAV on the expressway at the exit of the dedicated lane, the lane-changing behavior of CAV is analyzed. Taking the sum of lane utility and lane-changing urgency exceeding the threshold as the condition to trigger the lane-changing intention, a forced lane-changing intention recognition model for the exit of the CAV dedicated lane on the expressway is constructed. Based on the game theory, the process of multiple lane change of vehicles is divided into two stages according to the different positions of conflict points, and the action strategy of each stage is formed. The safety, speed and comfort are taken as the evaluation indexes of game decision-making. The income matrix of the CAV game decision-making model is established, the income function is established, and the lane change game decision-making model is constructed. PreScan and MATLAB/Simulink are used for simulation verification. Under different working conditions, the lane change decision model is simulated. The results show that when the initial speed is set to 30 m/s and 31 m/s respectively, the game decision models in the two lane-changing scenes perform better in terms of lane-changing success rate. |
| Key words: freeway CAV dedicated lanes mandatory lane change game theory lane change decision-making |