Abstract: Aiming at driver’s lane selection behavior in a clustered situation, with a view to the background of Internet of Things, the factors which influence driving behavior, such as vehicle group situation and driving propensity, are given a comprehensive consideration, and the driver’s lane selection model based on multi-player dynamic game with complete information is built. The driver’s payoffs under different combination of lane selection strategies are analyzed, and on this basis, the subgame perfect Nash equilibrium solution is solved through backward induction, and the driver’s optimal lane selection strategy is obtained. Actual driving experiment and other means are applied for model validation. The results show that the model built in this paper can objectively reflect driver’s lane selection behavior and the operation characteristic of traffic flow.

Key words: intelligence transportation, lane selection, dynamic game, vehicle group situation, driving propensity

摘要： 针对集群车辆驾驶员的车道选择行为，着眼物联网背景，综合考虑车辆集群态势、驾驶倾向性等影响驾驶员行为的因素，建立基于完全信息多人动态博弈的车道选择模型。通过分析不同策略组合下的驾驶员收益，运用逆向归纳法，求解子博弈精炼纳什均衡，得到驾驶员的最优车道选择策略。应用实车实验等手段验证模型，结果表明，所建模型能够较为客观地反映驾驶员车道选择行为及交通流特性。

关键词: 智能交通, 车道选择, 动态博弈, 车辆集群态势, 驾驶倾向性