Reinforcement learning based robot navigation using illegal actions for autonomous docking of surface vehicles in unknown environments.
Maria Inês PereiraAndry Maykol PintoPublished in: Eng. Appl. Artif. Intell. (2024)
Keyphrases
- robot navigation
- unknown environments
- autonomous robots
- robotic systems
- reinforcement learning
- initially unknown
- continuous state
- mobile robot
- exploration strategy
- simultaneous localization and mapping
- autonomous vehicles
- obstacle avoidance
- action space
- map building
- path planning
- three dimensional
- real time
- free space
- autonomous navigation
- topological map
- partially observable
- reward function
- machine learning
- learning agent
- scene understanding
- indoor environments
- state space
- mobile robotics
- multi robot
- markov decision processes
- d objects
- human operators
- real environment
- motion planning
- single image
- object recognition