Analysis and Modeling of Lane-Changing Game Strategy for Autonomous Driving Vehicles

Abstract

Autonomous driving vehicles have some advantages, such as alleviating tasks of drivers and reducing carbon emissions. With the advancement of intelligent network connection technology, autonomous driving vehicles are showing a trend of practicality and popularization, so it is crucial to study the decision-making mechanism of autonomous driving vehicles. This paper focuses on the lane-changing decision-making behavior of autonomous driving vehicles. Firstly, the objective quantification of lane-changing intention is carried out to reasonably show the lane-changing intention of autonomous driving vehicles as a prerequisite for lane-changing decision-making. Besides, the lane-changing collision probability and the lane-changing dynamic risky coefficient are introduced, and explore the dynamic influencing factors of lane-changing process for autonomous driving vehicles. Based on the game theory, the decision-making behavior model of the lane-changing game for autonomous driving vehicles is established. Analyze the decision-making behavior mechanism of lane-changing, so that the autonomous driving vehicle can change lanes safely and reasonably. Finally, with SUMO software, the traditional LC2013 lane-changing model and the decision-making behavior model of the lane-changing game are used for simulation experiments and comparative analysis. The results show that under the decision-making behavior model of the lane-changing game, the average speed of vehicles increases by 3.6%, and the average number of passed vehicles increases by 10.3%, which has higher stability, safety, speed gains, and lane utilization. The modeling of lane-changing game strategy for autonomous driving vehicles comprehensively considers the dynamic factors in the traffic environment, and scientifically shows the lane-changing decision-making mechanism of autonomous driving vehicles.