.png){kind=icon}
Return
Compactly Restrictable Metric Policy Optimization Problems
.png){kind=icon}
OA
AI
DOI：10.1109/TAC.2022.3217269.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
We study policy optimization problems for deterministic Markov decision processes (MDPs) with metric state and action spaces, which we refer to as metric policy optimization problems (MPOPs). Our goal is to establish theoretical results on the well-posedness of MPOPs that can characterize practically relevant continuous control systems. To do so, we define a special class of MPOPs called compactly restrictable MPOPs (CR-MPOPs), which are flexible enough to capture the complex behavior of robotic systems but specific enough to admit solutions using dynamic programming methods such as value iteration. We show how to arrive at CR-MPOPs using forward-invariance. We further show that our theoretical results on CR-MPOPs can be used to characterize feedback linearizable control affine systems.
Keywords:
Optimization
Aerospace electronics
Extraterrestrial measurements
Control systems
Robots
Reinforcement learning
Markov processes
Continuous Markov decision processes (MDPs)
optimal control
physical systems
reinforcement learning
sampled-data
selection theorems
value iteration
Journal
IF:
7
Papers:
1.3W
Citations:
6.7W
Organization
Cited Papers
Preparations ofN, N?-ethylene-bridged dipeptides(eXX) constructed from (S)-methionine, -tryptophan, -tyrosine and-N(?)-benzyloxycarbonyllysine through acid-catalyzed cyclization.png)
Amino Acids
Amino AcidsIF0
PREAI
ShareSave
ShareSaveShareSaveShareSaveSynthesis of new tetrazolyldienylphenothiazines as potential multidrug resistance inhibitory compoundsArkivocOAAI
ArkivocIF0
OAAIShareSave
View more