PRM-RL: Long-range Robotic Navigation Tasks by Combining Reinforcement Learning and Sampling-Based Planning.
Aleksandra FaustKenneth OslundOscar RamirezAnthony G. FrancisLydia TapiaMarek FiserJames DavidsonPublished in: ICRA (2018)
Keyphrases
- long range
- navigation tasks
- reinforcement learning
- motion planning
- mobile robot
- path planning
- short range
- real robot
- collision free
- belief space
- function approximation
- indoor environments
- robotic systems
- reinforcement learning algorithms
- multi robot
- model free
- state space
- dynamic environments
- markov decision processes
- temporal difference
- learning algorithm
- action selection
- conditional random fields
- long range interactions
- optimal control
- degrees of freedom
- collision avoidance
- optimal policy
- long range correlations
- partially observable markov decision processes
- rl algorithms
- multi agent
- machine learning
- partially observable
- dynamic programming