Normalization of systems with stochastically unsolved derivatives

For a system with stochastically unsolved derivatives, two approaches for reduction of such systems to deterministic systems are developed. The first approach is based on equations for mathematical expectations and covariance characteristics. The second approach considers equations for mathematical expectations and coordinate functions for canonical expansions. The theory of normal stochastic systems is the basis of the developed approaches. An illustrative example is given. Applications to estimation, identification, and calibration problems are considered. Some generalizations are mentioned.