Risk field enhanced game theoretic model for interpretable and consistent lane-changing decision makings

Author:

Xia Taokai1,Chen Hui1,Su Shaoka1

Affiliation:

1. Tongji University

Abstract

<div class="section abstract"><div class="htmlview paragraph">This paper presents an integrated modeling approach for real-time discretionary lane-changing decisions by autonomous vehicles, aiming to achieve human-like behavior. The approach incorporates a two-player normal-form game and a novel risk field method. The normal-form game represents the strategic interactions among traffic participants. It captures the trade-offs between lane-changing benefits and risks based on vehicle motion states during a lane change. By continuously determining the Nash equilibrium of the game at each time step, the model decides when it is appropriate to change the lane. A novel risk field method is integrated with the game to model risks in the game pay-offs. The risk field introduces regions along the desired target lane with different time headway ranges and risk weights, capturing traffic participants' complex risk perceptions and considerations in lane-changing scenarios. It goes beyond simple gap acceptance assumptions used in previous studies, providing more human-like risk estimations. Discretionary lane-changing data from human drivers extracted from the NGSIM I80 dataset were employed to calibrate the integrated model for human-like lane-change decisions. The calibration results demonstrate the high prediction accuracy of the proposed model compared to previous studies. The calibrated risk field parameters in the model provide interpretability and contribute to a deeper understanding of human lane-changing decisions. The proposed model also exhibits improved consistency in lane-changing decisions within a continuous time range around the lane-crossing moment. It outperforms previous game-theoretic models that rely on acceleration and time pay-offs with specific assumptions about future vehicle motions. Several case studies were carried out in the co-simulations of CARLA and SUMO software and based on the NGSIM dataset samples. The model's ability to produce reliable and interpretable lane-changing decisions enhances autonomous vehicles' overall safety and user experience.</div></div>

Publisher

SAE International

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3