Multi Pseudo Q-learning Based Deterministic Policy Gradient for Tracking Control of Autonomous Underwater Vehicles.
Wenjie ShiShiji SongCheng WuC. L. Philip ChenPublished in: CoRR (2019)
Keyphrases
- tracking control
- policy gradient
- function approximation
- reinforcement learning
- reinforcement learning algorithms
- nonlinear systems
- control law
- cooperative
- autonomous underwater vehicles
- learning algorithm
- reinforcement learning methods
- optimal control
- state space
- single agent
- learning rate
- model free
- markov decision processes
- multi agent
- state action
- optimal policy
- gradient method
- control scheme
- fuzzy controller
- closed loop
- variance reduction
- least squares
- machine learning