Construction of embedded Markov decision processes for optimal control of non-linear systems with continuous state spaces.
Daniel NikovskiAlan EsentherPublished in: CDC/ECC (2011)
Keyphrases
- markov decision processes
- optimal control
- continuous state spaces
- risk sensitive
- control problems
- dynamic programming
- reinforcement learning
- infinite horizon
- state space
- policy iteration
- average cost
- finite state
- action space
- optimal policy
- decision processes
- finite horizon
- partially observable
- control law
- control strategy
- continuous state
- markov decision process
- average reward
- model free
- finite number