基于博弈论的自动驾驶车辆协同换道分析  

Cooperative Lane Changing Analysis of Autonomous Vehicles Based on Game Theory

作　　者：陈华 CHEN Hua(College of Traffic&Transportation,Chongqing Jiaotong University,Chongqing 400074,China)

机构地区：[1]重庆交通大学交通运输学院,重庆400074

出　　处：《武汉理工大学学报(交通科学与工程版)》2025年第1期42-46,共5页Journal of Wuhan University of Technology(Transportation Science & Engineering)

摘　　要：文中基于两方博弈思想建立自动驾驶车辆换道决策模型.依据车辆行驶状态确定具备换道动机的博弈对象,选取安全距离、速度收益、行驶自由性作为博弈换道综合收益,寻求最优行驶条件的车道,并基于两方博弈思想建立自动驾驶车辆换道博弈决策模型并求解换道概率,基于SUMO软件并选取不同交通流密度下的车辆平均速度、车均损失时间以及车辆通过数作为评价指标将传统LC2013换道模型与博弈换道决策模型进行仿真对比,结果表明:换道博弈决策模型下整体车流运行效益优于传统换道模型.Based on the idea of two-party game,the decision model of lane change for autonomous vehicles was established.According to the driving state of the vehicle,the game object with lane-changing motivation was determined,and the safe distance,speed benefit and driving freedom were selected as the comprehensive benefits of the game lane-changing,so as to find the lane with the best driving conditions.Based on the two-party game idea,the game decision-making model of lane-changing for autonomous vehicles was established and the lane-changing probability was solved.Based on SUMO software,the average speed of vehicles,the average lost time of vehicles and the number of vehicles passing under different traffic flow densities were selected as evaluation indexes.The simulation comparison between the traditional LC2013 lane change model and the game lane change decision model shows that the overall traffic flow operation benefit under the lane change game decision model is better than the traditional lane change model.

关 键 词：自动驾驶 换道行为 协同换道 博弈论 SUMO 

分 类 号：U491.2[交通运输工程—交通运输规划与管理]