Reinforcement Learning for Versatile, Dynamic, and Robust Bipedal Locomotion Control.
Zhongyu LiXue Bin PengPieter AbbeelSergey LevineGlen BersethKoushil SreenathPublished in: CoRR (2024)
Keyphrases
- reinforcement learning
- robot control
- control problems
- autonomous robots
- mobile robot
- dynamic environments
- robotic systems
- machine learning
- optimal control
- control system
- action selection
- closed loop
- computationally efficient
- adaptive control
- legged robots
- robust stability
- function approximation
- multi modal
- parameter tuning
- robust estimation
- process control
- learning algorithm
- neural network
- inverted pendulum
- policy search