Action Mapping for Reinforcement Learning in Continuous Environments with Constraints

Mirco Theile, Lukas Dirnberger, Raphael Trumpp, Marco Caccamo, Alberto Sangiovanni-Vincentelli; 6:343−363, 2025.

Presented at the Reinforcement Learning Conference (RLC), Edmonton, Alberta, Canada, August 5–9, 2025.

Abstract

Deep reinforcement learning (DRL) has had success across various domains, but applying it to environments with constraints remains challenging due to poor sample efficiency and slow convergence. Recent literature explored incorporating model knowledge to mitigate these problems, particularly using models that assess the feasibility of proposed actions. However, integrating feasibility models efficiently into DRL pipelines in environments with continuous action spaces is non-trivial. We propose a novel DRL training strategy utilizing action mapping that leverages feasibility models to streamline the learning process. By decoupling the learning of feasible actions from policy optimization, action mapping allows DRL agents to focus on selecting the optimal action from a reduced feasible action set. We demonstrate that action mapping significantly improves training performance in two constrained environments with continuous action spaces, especially with imperfect feasibility models.

[abs][pdf]

BibTeX

@article{theile2025action,
    title={Action Mapping for Reinforcement Learning in Continuous Environments with Constraints},
    author={Theile, Mirco and Dirnberger, Lukas and Trumpp, Raphael and Caccamo, Marco and Sangiovanni-Vincentelli, Alberto},
    journal={Reinforcement Learning Journal},
    volume={6},
    pages={343--363},
    year={2025}
}