Optimal control for unknown mean-field discrete-time system based on Q-Learning.
Yingying GeXikui LiuYan LiPublished in: Int. J. Syst. Sci. (2021)
Keyphrases
- optimal control
- optimal control problems
- reinforcement learning
- linear quadratic
- dynamic programming
- control problems
- policy iteration
- feedback control
- state space
- actor critic
- function approximation
- brownian motion
- class of nonlinear systems
- risk sensitive
- infinite horizon
- control strategy
- average cost
- rl algorithms
- optimal policy
- reinforcement learning algorithms
- production planning
- learning algorithm
- markov chain
- real time
- machine learning
- continuous stirred tank reactor
- markov random field
- solving nonlinear
- model free
- lyapunov function
- control law
- differential equations
- closed form
- neural network