Crowd Navigation in an Unknown and Dynamic Environment Based on Deep Reinforcement Learning.
Libo SunJinfeng ZhaiWenhu QinPublished in: IEEE Access (2019)
Keyphrases
- dynamic environments
- reinforcement learning
- potential field
- autonomous navigation
- reinforcement learning algorithms
- path planning
- mobile robot
- reinforcement learning agents
- obstacle avoidance
- autonomous agents
- collision free
- function approximation
- indoor environments
- adaptive control
- collision avoidance
- real environment
- model free
- temporal difference
- state space
- markov decision processes
- environmental conditions
- learning algorithm
- agent systems
- visual slam
- initially unknown
- simultaneous localization and mapping
- real time
- single agent
- dynamic programming
- machine learning
- robot control
- changing environment
- optimal policy