arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2108.08448cs.LG
收藏

基于先验正则化的预适应Meta强化学习鲁棒性和安全性改进

Improved Robustness and Safety for Pre-Adaptation of Meta Reinforcement Learning with Prior Regularization

University of Michigan(密歇根大学);Ford Research and Advanced Engineering(福特研究与高级工程中心)

Lu Wen, Songan Zhang, H. Eric Tseng, Baljeet Singh, Dimitar Filev, Huei Peng

摘要:

Meta强化学习(Meta-RL)最近取得了重大进展。特别是,开发了off-policy方法来提高Meta-RL技术的数据效率。\textit{Probabilistic embeddings for actor-critic RL}(PEARL)是解决多MDP适应问题的领先方法。许多现有的Meta-RL方法(包括PEARL)的一个主要缺点是,当第一次暴露给新任务时,它们没有明确考虑先前策略的安全性。安全对于许多现实世界的应用至关重要,包括现场机器人和自动驾驶汽车(AV)。在本文中,我们开发了PEARL PLUS(PEARL $^+$)算法,该算法优化了先前(预适应)的安全性和之后(适应后)的性能。在PEARL算法的基础上,我们提出的PEARL $^+$算法在奖励函数中引入了一个先验正则化项,并引入了一个新的Q-网络,用于在先验上下文假设下恢复状态-动作值,以提高对任务分布偏移的鲁棒性和训练网络首次暴露于新任务时的安全性。PEARL $^+$的性能进行了验证,通过解决三个安全关键问题的机器人和自动驾驶汽车,包括两个MuJoCo基准问题。仿真实验表明,与PEARL相比,该策略的安全性得到了显著提高,并且对任务分布偏移具有更强的鲁棒性。

英文摘要:

Meta Reinforcement Learning (Meta-RL) has seen substantial advancements recently. In particular, off-policy methods were developed to improve the data efficiency of Meta-RL techniques. \textit{Probabilistic embeddings for actor-critic RL} (PEARL) is a leading approach for multi-MDP adaptation problems. A major drawback of many existing Meta-RL methods, including PEARL, is that they do not explicitly consider the safety of the prior policy when it is exposed to a new task for the first time. Safety is essential for many real-world applications, including field robots and Autonomous Vehicles (AVs). In this paper, we develop the PEARL PLUS (PEARL$^+$) algorithm, which optimizes the policy for both prior (pre-adaptation) safety and posterior (after-adaptation) performance. Building on top of PEARL, our proposed PEARL$^+$ algorithm introduces a prior regularization term in the reward function and a new Q-network for recovering the state-action value under prior context assumptions, to improve the robustness to task distribution shift and safety of the trained network exposed to a new task for the first time. The performance of PEARL$^+$ is validated by solving three safety-critical problems related to robots and AVs, including two MuJoCo benchmark problems. From the simulation experiments, we show that safety of the prior policy is significantly improved and more robust to task distribution shift compared to PEARL.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑