Data-Driven Solutions to Mixed $H_{2}/H_{\infty}$ Control: A Hamilton-Inequality-Driven Reinforcement Learning Approach.
Yongliang YangMajid MazouchiHamidreza ModaresPublished in: CCTA (2020)
Keyphrases
- data driven
- reinforcement learning
- control problems
- control strategies
- robot control
- optimal control
- control policy
- control system
- action selection
- solution space
- function approximation
- model driven
- control method
- benchmark problems
- data acquisition
- control algorithm
- real time
- supervised learning
- temporal difference
- optimal solution
- machine learning
- control theory
- neural network