Maximum Entropy Models, Dynamic Games, and Robust Output Feedback Control for Automata

In this paper, we develop a framework for designing controllers for automata which are robust with respect to uncertainties. A deterministic model for uncertainties is introduced, leading to a dynamic game formulation of the robust control problem. This problem is solved using an appropriate information state. We derive a Hidden Markov Model as the maximum entropy stochastic model for the automaton. A risk-sensitive stochastic control problem is formulated and solved for this Hidden Markov Model. The two problems are related using small noise limits.