PBCS : Efficient Exploration and Exploitation Using a Synergy between Reinforcement Learning and Motion Planning.
Guillaume MatheronNicolas PerrinOlivier SigaudPublished in: CoRR (2020)
Keyphrases
- motion planning
- reinforcement learning
- degrees of freedom
- mobile robot
- robot arm
- trajectory planning
- path planning
- function approximation
- robotic tasks
- exploration exploitation tradeoff
- multi robot
- humanoid robot
- inverse kinematics
- obstacle avoidance
- belief space
- mechanical systems
- robotic arm
- multi agent
- markov decision processes
- learning algorithm
- dynamic programming
- model free
- robot control
- real robot
- autonomous mobile robot
- collision free
- optimal policy
- partially observable
- state space
- long run
- three dimensional
- control system
- object detection
- machine learning