Neural Q- Learning Based on Residual Gradient for Nonlinear Control Systems.
Yanna SiJiexin PuShaofei ZangPublished in: ICCAIS (2019)
Keyphrases
- control system
- recurrent neural network model
- cooperative
- neural model
- reinforcement learning
- network architecture
- function approximation
- nonlinear dynamics
- neural network
- multi agent
- state space
- learning algorithm
- biologically inspired
- closed loop
- bio inspired
- edge detection
- control scheme
- associative memory
- real time
- model free
- multiscale
- gradient information
- gradient direction
- potential field
- model predictive control
- artificial neural networks