Variable Impedance Control in End-Effector Space: An Action Space for Reinforcement Learning in Contact-Rich Tasks.
Roberto Martín-MartínMichelle A. LeeRachel GardnerSilvio SavareseJeannette BohgAnimesh GargPublished in: CoRR (2019)
Keyphrases
- action space
- impedance control
- reinforcement learning
- model free
- end effector
- manipulation tasks
- state space
- markov decision processes
- real valued
- optimal control
- force control
- force feedback
- degrees of freedom
- robot manipulators
- robot arm
- inverse kinematics
- position control
- action selection
- stochastic processes
- function approximation
- vision system
- transfer learning
- dynamic programming
- real robot
- motion planning
- optimal policy
- visual servoing
- multi agent
- robot navigation
- single agent
- markov decision process
- machine learning
- robotic arm
- control law
- robotic systems
- mobile robot
- configuration space
- motor control
- search space
- learning algorithm