Title: Safe Reinforcement Learning with Probabilistic Control Barrier Functions for Ramp Merging Show Chinese title

Title: 带有概率控制屏障函数的匝道合并安全强化学习

Abstract: Prior work has looked at applying reinforcement learning and imitation learning approaches to autonomous driving scenarios, but either the safety or the efficiency of the algorithm is compromised. With the use of control barrier functions embedded into the reinforcement learning policy, we arrive at safe policies to optimize the performance of the autonomous driving vehicle. However, control barrier functions need a good approximation of the model of the car. We use probabilistic control barrier functions as an estimate of the model uncertainty. The algorithm is implemented as an online version in the CARLA (Dosovitskiy et al., 2017) Simulator and as an offline version on a dataset extracted from the NGSIM Database. The proposed algorithm is not just a safe ramp merging algorithm but a safe autonomous driving algorithm applied to address ramp merging on highways. Show Chinese abstract

Abstract: 先前的工作已经研究了将强化学习和模仿学习方法应用于自动驾驶场景，但要么算法的安全性受到影响，要么效率受到影响。 通过将控制屏障函数嵌入到强化学习策略中，我们得出了安全策略，以优化自动驾驶车辆的性能。 然而，控制屏障函数需要对车辆模型有良好的近似。 我们使用概率控制屏障函数作为模型不确定性的估计。 该算法在CARLA（Dosovitskiy等，2017）模拟器中实现为在线版本，并在从NGSIM数据库提取的数据集上实现为离线版本。 所提出的算法不仅仅是一个安全的匝道合并算法，而是一个应用于解决高速公路上匝道合并的安全自动驾驶算法。