Convergence of Dynamic Programming on the Semidefinite Cone for Discrete-Time Infinite-Horizon LQR.
Donghwan LeePublished in: IEEE Trans. Autom. Control. (2022)
Keyphrases
- semidefinite
- infinite horizon
- dynamic programming
- optimal control
- semidefinite programming
- stationary policies
- finite horizon
- optimal policy
- linear programming
- markov decision processes
- production planning
- multistage
- sufficient conditions
- average cost
- long run
- dec pomdps
- interior point methods
- convex relaxation
- finite state
- higher dimensional
- markov decision process
- state space
- complementarity problems
- control strategy
- markov chain
- convex sets
- reinforcement learning
- positive semidefinite
- markov decision problems
- convergence rate
- linear program
- primal dual
- finite dimensional