arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2402.05149cs.LGcs.AI
收藏

FlowPG:基于归一化流的动作约束策略梯度

FlowPG: Action-constrained Policy Gradient with Normalizing Flows

School of Computing and Information Systems(计算与信息系统学院);Singapore Management University(新加坡管理大学)

Janaka Chathuranga Brahmanage, Jiajing Ling, Akshat Kumar

AI总结:

本文提出 FlowPG,利用归一化流学习可行动作空间与简单分布间的可逆映射,并与 DDPG 集成,避免投影层优化,显著减少约束违反并加速连续控制任务训练。

摘要:

动作约束强化学习(ACRL)是解决安全关键和资源分配相关决策问题的常用方法。ACRL 的一个主要挑战是确保智能体在每一步强化学习中采取满足约束的有效动作。常用的方法是在策略网络之上使用投影层,这需要求解一个优化程序,可能导致训练时间更长、收敛速度慢以及零梯度问题。为解决这些问题,首先,我们使用归一化流模型来学习可行动作空间与潜在变量(如高斯分布)上简单分布支撑之间的可逆、可微映射。其次,学习流模型需要从可行动作空间采样,这同样具有挑战性。我们基于哈密顿蒙特卡洛和概率命题决策图开发了多种方法,用于凸约束和非凸约束下的动作采样。第三,我们将学习到的归一化流与 DDPG 算法集成。通过设计,训练良好的归一化流可以将策略输出转换为有效动作,而无需优化求解器。实验上,我们的方法显著减少了约束违反次数(在多个实例中减少了一个数量级),并且在多种连续控制任务上速度提升了数倍。

英文摘要:

Action-constrained reinforcement learning (ACRL) is a popular approach for solving safety-critical and resource-allocation related decision making problems. A major challenge in ACRL is to ensure agent taking a valid action satisfying constraints in each RL step. Commonly used approach of using a projection layer on top of the policy network requires solving an optimization program which can result in longer training time, slow convergence, and zero gradient problem. To address this, first we use a normalizing flow model to learn an invertible, differentiable mapping between the feasible action space and the support of a simple distribution on a latent variable, such as Gaussian. Second, learning the flow model requires sampling from the feasible action space, which is also challenging. We develop multiple methods, based on Hamiltonian Monte-Carlo and probabilistic sentential decision diagrams for such action sampling for convex and non-convex constraints. Third, we integrate the learned normalizing flow with the DDPG algorithm. By design, a well-trained normalizing flow will transform policy output into a valid action without requiring an optimization solver. Empirically, our approach results in significantly fewer constraint violations (upto an order-of-magnitude for several instances) and is multiple times faster on a variety of continuous control tasks.

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。

↑