Undiscounted control policy generation for continuous-valued optimal control by approximate dynamic programming.
Jonathan LockTomas McKelveyPublished in: Int. J. Control (2022)
Keyphrases
- approximate dynamic programming
- optimal control
- continuous valued
- policy iteration
- control policy
- average cost
- infinite horizon
- dynamic programming
- reinforcement learning
- markov decision processes
- markov decision problems
- finite horizon
- control policies
- markov decision process
- long run
- average reward
- control strategy
- optimal control problems
- state space
- image processing
- step size
- linear program
- autoregressive model
- actor critic
- initial state
- multistage
- optimal policy
- linear programming
- least squares
- multiresolution
- machine learning
- real time