A Q-learning predictive control scheme with guaranteed stability.
Lukas BeckenbachPavel OsinenkoStefan StreifPublished in: Eur. J. Control (2020)
Keyphrases
- control scheme
- feedback controller
- fault tolerant control
- closed loop
- control system
- dynamic model
- reinforcement learning
- external disturbances
- control strategy
- limit cycle
- robotic manipulator
- control law
- controller design
- function approximation
- robot manipulators
- predictive control
- state space
- control loop
- model free
- neural network controller
- learning rate
- nonlinear systems
- neural model
- pid controller
- learning algorithm
- fuzzy controller
- artificial neural networks
- induction motor
- sliding mode
- tracking error
- adaptive fuzzy
- stability analysis
- force control