Mapless Motion Planning System for an Autonomous Underwater Vehicle Using Policy Gradient-based Deep Reinforcement Learning.
Yushan SunJunhan ChengGuocheng ZhangHao XuPublished in: J. Intell. Robotic Syst. (2019)
Keyphrases
- autonomous underwater vehicle
- reinforcement learning
- optimal policy
- policy search
- markov decision process
- motion model
- model free
- image sequences
- motion estimation
- motion analysis
- policy evaluation
- motion field
- partially observable
- action selection
- space time
- state and action spaces
- motion segmentation
- motion patterns
- policy iteration
- moving objects
- action space
- optical flow
- control policy
- actor critic
- reinforcement learning problems
- machine learning
- camera motion
- reinforcement learning algorithms
- markov decision processes
- markov decision problems
- policy gradient
- approximate dynamic programming
- function approximation
- human motion
- partially observable environments
- infinite horizon
- state action
- partially observable domains
- long run
- transition model
- multi agent
- agent learns
- decision problems