SARSA-based reinforcement learning for motion planning in serial manipulators.
Ignazio AleoPaolo ArenaLuca PatanéPublished in: IJCNN (2010)
Keyphrases
- motion planning
- reinforcement learning
- degrees of freedom
- reinforcement learning algorithms
- function approximation
- path planning
- mobile robot
- trajectory planning
- robot arm
- function approximators
- humanoid robot
- state space
- control law
- temporal difference learning
- temporal difference
- model free
- learning algorithm
- multi robot
- belief space
- robotic arm
- adaptive control
- action selection
- inverse kinematics
- markov decision processes
- autonomous mobile robot
- dynamic programming
- obstacle avoidance
- rl algorithms
- robotic tasks
- multi agent
- configuration space
- collision free
- optimal policy
- pose estimation
- three dimensional
- high dimensional
- partially observable
- machine learning
- mechanical systems