Obtaining Robust Control and Navigation Policies for Multi-robot Navigation via Deep Reinforcement Learning.
Christian JestelHartmut SurmannJonas StenzelOliver UrbannMarius BrehlerPublished in: ICARA (2021)
Keyphrases
- robot navigation
- autonomous robots
- continuous state
- reinforcement learning
- autonomous mobile robot
- policy search
- robot control
- control policy
- optimal policy
- control policies
- mobile robot
- optimal control
- scene understanding
- map building
- control system
- robot localization
- landmark recognition
- topological map
- dynamic programming
- state space
- fitted q iteration
- machine learning
- unknown environments
- markov decision process
- function approximation
- robotic systems
- object recognition
- initially unknown
- object detection