让模仿学习在赛车任务中更安全可靠
A Simple Approach to Constraint-Aware Imitation Learning with Application to Autonomous Racing
- 在模仿学习目标中直接加入约束条件
- 仿真测试显示约束满足率和性能更稳定
- 适合高精度、极限操作场景的自动驾驶研究
在需要接近系统操控极限的任务中，保证约束满足是模仿学习（IL）的难点。传统方法如行为克隆（BC）难以有效执行约束，导致高精度任务表现不佳。本文提出一种简单方法，在模仿学习目标中引入安全约束。通过模拟实验，在具有全状态反馈和图像反馈的自动驾驶赛车任务上验证了该方法，结果表明其在约束满足度和任务性能一致性方面优于行为克隆。
原文摘要 · Abstract (English)
Guaranteeing constraint satisfaction is challenging in imitation learning (IL), particularly in tasks that require operating near a system's handling limits. Traditional IL methods, such as Behavior Cloning (BC), often struggle to enforce constraints, leading to suboptimal performance in high-precision tasks. In this paper, we present a simple approach to incorporating safety into the IL objective. Through simulations, we empirically validate our approach on an autonomous racing task with both full-state and image feedback, demonstrating improved constraint satisfaction and greater consistency in task performance compared to BC.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。