Continuous-time reinforcement learning for robust control under worst-case uncertainty.
Adolfo PerrusquíaWen YuPublished in: Int. J. Syst. Sci. (2021)
Keyphrases
- optimal control
- reinforcement learning
- worst case
- control problems
- state space
- robot control
- control system
- markov chain
- function approximation
- iterative learning control
- dynamic programming
- upper bound
- supervised learning
- robust optimization
- control strategy
- uncertain data
- action selection
- belief functions
- markov decision processes
- computationally efficient
- neural network
- model free
- adaptive control
- average case
- dynamical systems
- lower bound
- control theory
- control policy
- inverted pendulum
- robust stability
- learning algorithm