Control of Markovian jumps linear systems with cost and information associated to jump times

Deals with a stochastic optimal control problem involving discrete-time jump Markov linear systems. The jumps or changes between the system operation modes evolve according to an underlying Markov chain, but the Markov states may be not completely available to the controller. The control problem horizon is defined by the occurrence of a finite number of jumps, and the information available allows the control action reconfiguration at each jump time, in the form of a linear feedback gain. An optimal solution for the problem with complete Markov states observation, and a sub-optimal solution for the problem with incomplete state observation are presented. These solutions are based on linear matrix inequalities.