基于ARMA车速预测的智能车交叉口强化学习决策研究

打开文本图片集

YUZhicheng'，ZHAO Junpeng²，LIUYonggang'，XIA Pugeng，YE Ming4 （1. State Key Laboratory of Mechanical Transmission for Advanced Equipment， Chongqing University， Chongqing 40044，P.R.China;2.Beijing AerospaceLaunch Technology Research Institute，Beijing 100076， P.R.China;3.Chengdu Yiwei New Energy Vehicle Co.，Ltd.，Chengdu 611730，P.R.China; 4.Vehicle Engineering Institute， Chongqing University of Technology， Chongqing 40oo54， P.R. China）

Abstract：To address the chalenge of autonomous vehicle decision-making and control at unsignalized intersections，this study investigates the merging behaviorof twovehicles atatwo-way single-lane intersection. Reinforcement learning is used to establish a mapping between the vehicle state space and action space for autonomous decision-making.To overcome the limitations of overly simplified speed setings in existing studies， real-world trajectory data of surrounding vehiclesareused to construct an environmental traffic model.The autoregressive moving average （ARMA） model is applied to predict the speeds of surrounding vehicles. By integrating the predicted speed profiles with the autonomous vehicle's motion parameters，a forward decisionmaking model is established to calculate reference speeds.These reference speedsare incorporated into the reinforcement learning reward function to accelerate training convergence.Experimental results show that the proposed model achieves rapid convergence，and the trained agent can safely navigate the intersection while interacting with surrounding vehicles exhibiting diverse driving behaviors.This work provides a reference framework for improving the safety and effciency of autonomous vehicle decision-making at unsignalized intersections.

Keywords： intersections; autonomous vehicles; autoregressive moving average model （ARMA）; reinforcement learning

随着自动驾驶科技的持续进步，大量具备初级自动驾驶辅助系统（advanced driver asistance system，ADAS）的车辆已逐步投入使用。（剩余11670字）