Deep Reinforcement Learning with Long-Time Memory Capability for Robot Mapless Navigation.
Qinglin ZhouLei LyuHong LiuPublished in: CSCWD (2022)
Keyphrases
- reinforcement learning
- obstacle avoidance
- mobile robot
- robot navigation
- robot control
- real robot
- rough terrain
- service robots
- indoor environments
- navigation tasks
- autonomous navigation
- human robot interaction
- autonomous learning
- unknown environments
- collision free
- autonomous vehicles
- outdoor environments
- perceptual aliasing
- function approximation
- path planning
- memory usage
- learning capabilities
- visual navigation
- robot behavior
- robotic control
- learning algorithm
- optimal policy
- multi agent
- robotic systems
- state space
- potential field
- reinforcement learning algorithms
- robotic arm
- past experience
- memory space
- memory requirements
- navigation systems
- topological map
- mobile robotics
- markov decision processes
- humanoid robot
- real time
- human robot
- dynamic environments
- vision system
- robot manipulators
- multi robot
- motion planning
- model free
- dynamic programming
- website
- machine learning
- search and rescue
- reinforcement learning methods
- robot soccer
- optimal control
- human users
- temporal difference
- action selection
- information space
- goal directed