Linear-quadratic stochastic control problem. III. Nonlinear optimal controllers (Q1324063)

The problem described by a linear difference equation with additive stochastic noise and a mean quadratic performance functional is considered. The noise is Markov with known transition density. It is shown that in the case of nonnormal noise, the optimal control strategy may be nonlinear. Under the assumption that the state of the system is known at each instant of time, the ``explicit'' formulas for the feedback controls which minimize the quadratic performance functional are derived. Examples are given demonstrating the efficiency of the proposed method for the synthesis of linear and nonliner optimal feedbacks.