On-Policy Data-Driven Linear Quadratic Regulator via Combined Policy Iteration and Recursive Least Squares.
Lorenzo SforniGuido CarnevaleIvano NotarnicolaGiuseppe NotarstefanoPublished in: CDC (2023)
Keyphrases
- linear quadratic
- optimal control
- policy iteration
- actor critic
- recursive least squares
- infinite horizon
- markov decision processes
- reinforcement learning
- dynamical systems
- optimal policy
- dynamic programming
- closed loop
- policy gradient
- average reward
- fixed point
- neuro fuzzy
- model free
- vector valued
- control strategy
- markov decision process
- average cost
- temporal difference
- gaussian model
- multiscale
- least squares
- state space
- finite state
- neural network
- particle swarm optimization