Reinforcement learning in dynamic environment: abstraction of state-action space utilizing properties of the robot body and environment.
Kazuyuki ItoYutaka TakeuchiPublished in: Artif. Life Robotics (2016)
Keyphrases
- dynamic environments
- mobile robot
- reinforcement learning
- state action space
- path planning
- function approximation
- real environment
- real robot
- changing environment
- sensory information
- autonomous agents
- mobile robotics
- potential field
- reinforcement learning agents
- indoor environments
- autonomous robots
- reinforcement learning algorithms
- simultaneous localization and mapping
- multi robot
- collision free
- model free
- single agent
- path planner
- state space
- learning agent
- robotic systems
- temporal difference
- robot navigation
- optimal policy
- vision system
- learning algorithm
- adaptive control
- unknown environments
- state abstraction