BriefGPT.xyz
Apr, 2025
基于视觉的赛车模拟环境驾驶代理
Vision based driving agent for race car simulation environments
HTML
PDF
Gergely Bári, László Palkovics
TL;DR
本研究解决了在模拟环境中控制赛车轮胎抓地力极限的问题，填补了传统控制方法在紧急情况下的不足。论文采用深度强化学习 (DRL) 方法和近端策略优化 (PPO) 算法，成功训练出能够仅通过视觉输入实现专业人类圈速的驾驶代理。研究结果表明，该代理表现出人类般的学习和驾驶行为，充分利用轮胎的抓地力潜力。
Abstract
In recent years,
Autonomous Driving
has become a popular field of study. As control at tire grip limit is essential during emergency situations, algorithms developed for racecars are useful for road cars too. This paper examines the use of
→