Finite-horizon LQR controller for partially-observed Boolean dynamical systems.
Mahdi ImaniUlisses M. Braga-NetoPublished in: Autom. (2018)
Keyphrases
- dynamical systems
- partially observed
- finite horizon
- linear quadratic
- infinite horizon
- optimal control
- optimal policy
- expected reward
- state space
- markov decision processes
- differential equations
- multistage
- dynamic systems
- markov decision process
- control theory
- real valued
- control law
- nonlinear dynamical systems
- average cost
- partially observable
- reinforcement learning
- dynamic programming
- non stationary
- closed loop
- decision problems
- finite state
- lost sales
- control strategy
- predictive state representations
- initial state
- objective function