Numerical methods of synthesis of an optimal control for stochastic dynamical systems of diffusion type

Several outlines of numerical methods of synthesis of optimal trajectories in the problem of control of a stochastic dynamic system of diffusion type are proposed. Both the systems with complete information on the state and the systems with information constraints, which mean that each component of the control strategy may depend only on the set of the components of the state vector measured precisely stipulated in advance, are considered. It is shown that in the problem with complete information on the state, one of the methods converges for just one iteration. For linear and quasilnear systems with quadratic cost function, the proposed methods are brought out to the regular computation procedures.