Identification of parameters of discrete stochastic systems with unknown inputs based on an information filtering algorithm

The paper considers the problem of parameter identification of discrete linear stochastic systems in the state space. An identification criterion is proposed for systems with unknown inputs based on the information version of the Gillijns and De Moor algorithm. We apply this criterion to identify the diffusion coefficient of a one-dimensional diffusion model with unknown boundary conditions of the first kind. The results of computer modeling validate the presented approach.