Reinforcement Learning for Legged Robots: Motion Imitation from Model-Based Optimal Control.
AJ MillerShamel FahmiMatthew ChignoliSangbae KimPublished in: CoRR (2023)
Keyphrases
- optimal control
- reinforcement learning
- legged robots
- model free
- control problems
- dynamic programming
- quadruped robot
- feedback control
- image sequences
- function approximation
- infinite horizon
- state space
- optimal control problems
- inverted pendulum
- human motion
- optimal policy
- control law
- real robot
- machine learning
- moving objects
- control strategy
- markov decision processes
- motion planning
- inverse kinematics
- actor critic
- mobile robot
- autonomous navigation
- dynamical systems
- monte carlo