构造舒适：学习型驾驶策略的自适应、舒适有界动作空间
Comfort by Construction: Adaptive, Comfort-Bounded Action Spaces for Learned Driving Policies
- University of Freiburg（弗莱堡大学）
- Bosch Center for Artificial Intelligence（博世人工智能中心）
- Coburg University of Applied Sciences（科堡应用科学大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
针对固定网格动作空间导致强化学习策略产生不舒适机动的问题，提出自适应动作参数化方法，通过闭式求逆重新离散化网格，在Waymo数据集上实现舒适违规低于1%并提升可导航性。
AI中文摘要：
数据驱动的驾驶模拟器从固定网格中指令加速度和转向速率，而不约束实际产生的加速度和冲击度。因此，强化学习策略通过突然的最后一刻机动来夸大安全指标，这些机动远超出人类驾驶范围，且对真实车辆乘员不可接受，因此这些指标衡量的是模拟器的宽容度而非策略质量。简单地强制舒适界限是不够的：横向限制随速度二次方缩小，因此对静态网格进行截断会使其饱和并破坏精细控制（“网格崩溃”）。我们提出一种自适应动作参数化方法，通过横向冲击度约束的闭式求逆，在每一步重新离散化网格，以精确覆盖每步可行控制集。我们进一步提出PufferDrive-Editor，一个基于浏览器的工具，用于审计实际运动学并创作运动学上具有挑战性的场景。在Waymo开放运动数据集和手工编写的蛇形赛道上，我们的自适应模型将舒适违规保持在1%以下，同时在可导航性上优于截断网格和直接冲击度基线。
英文摘要：
Data-driven driving simulators command accelerations and steering rates from a fixed grid without constraining the realized accelerations and jerks. As a result, reinforcement-learning policies inflate safety metrics through abrupt, last-second maneuvers that lie far outside the range of human driving and would be unacceptable to occupants of a real vehicle, so the metrics measure simulator permissiveness rather than policy quality. Enforcing comfort bounds naively is not enough: lateral limits shrink quadratically with speed, so clamping a static grid saturates it and destroys fine-grained control ("grid collapse"). We propose an adaptive action parameterization that rediscretizes the grid at every step to span exactly the per-step feasible control set, via closed-form inversion of the lateral-jerk constraint. We further present PufferDrive-Editor, a browser-based tool to audit realized kinematics and author kinematically challenging scenes. On the Waymo Open Motion Dataset and a hand-authored slalom, our adaptive model holds comfort violations below 1% while outperforming clipped-grid and direct-jerk baselines in navigability.
