DeepGait: Planning and Control of Quadrupedal Gaits Using Deep Reinforcement Learning.
Vassilios TsounisMitja AlgeJoonho LeeFarbod FarshidianMarco HutterPublished in: IEEE Robotics Autom. Lett. (2020)
Keyphrases
- legged robots
- reinforcement learning
- mechanical systems
- action selection
- inverted pendulum
- control problems
- control system
- optimal control
- robot control
- sagittal plane
- quadruped robot
- planning problems
- motion planning
- function approximation
- control policy
- stochastic domains
- domain independent
- markov decision processes
- control method
- control strategy
- reinforcement learning algorithms
- adaptive control
- nonlinear systems
- optimal policy
- control rules
- planning domains
- robotic control
- multi agent
- machine learning