Robust Dynamic Locomotion via Reinforcement Learning and Novel Whole Body Controller.
Donghyun KimJaemin LeeLuis SentisPublished in: CoRR (2017)
Keyphrases
- reinforcement learning
- optimal control
- dynamic environments
- quadruped robot
- neural network
- robot control
- learning algorithm
- strong robustness
- robust stability
- computationally efficient
- markov decision processes
- closed loop
- function approximation
- dynamic programming
- fuzzy controller
- parameter tuning
- model free
- feedback control
- function approximators
- control policy
- legged robots
- real time