AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Reduced Policy Optimization for Continuous Control with Hard Constraints

Shutong Ding, Jingya Wang, Yali Du, Ye Shi

NeurIPS 2023 · Reinforcement learning › Policy gradient and actor-critic · approach: Constrained policy optimization · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add two-stage decision process (construction plus GRG projection) to DDPG with Lagrangian relaxation (ablation)DDPG · Safe CartPoleepisodic reward: 15.5000 -> 200.0000Improved
add two-stage decision process to SAC with Lagrangian relaxation (ablation)SAC · Safe CartPoleepisodic reward: 40.5000 -> 200.0000Improved
add two-stage decision process to DDPG (ablation)DDPG · Spring Pendulumepisodic reward: 76.4383 -> 175.1558Improved
add two-stage decision process to SAC (ablation)SAC · Spring Pendulumepisodic reward: 149.6762 -> 182.4221Improved
add two-stage decision process to DDPG (ablation)DDPG · OPF with Battery Energy Storageepisodic reward: -22.0649 -> 42.9710Improved
add two-stage decision process to SAC (ablation)SAC · OPF with Battery Energy Storageepisodic reward: -19.4738 -> 42.0764Improved
add two-stage decision process to DDPG (ablation)DDPG · Spring Pendulummax instantaneous equality violation: 0.3805 -> 0.0000Improved
RPO-SAC versus Safety LayerRPO-SAC · Safe CartPolemax instantaneous equality violation: 1.7099 -> 0.0000Improved
RPO versus Safe RL baseline CUPRPO-SAC · Spring Pendulumepisodic reward: 0.8599 -> 182.4221Improved
RPO versus CPORPO-DDPG · OPF with Battery Energy Storagemax instantaneous equality violation: 0.4413 -> 0.0001Improved
RPO versus DDPG-L on inequality violationRPO-DDPG · OPF with Battery Energy Storagemax instantaneous inequality violation: 0.0000 -> 0.0002Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy