Aerospace Instrument-Making Annotation << Back Modified algorithm for optimal signal filtering based on modeling special branching process Rybakov K.A. New approach for solving the optimal nonlinear filtering problem by statistical modeling is described. It is based on the reducing the filtration problem to the analysis of stochastic systems with terminating and branching paths by the interpretation of the term in Duncan–Mortensen–Zakai equation (nonlinear filtering equation for unnormalized conditional density of the state vector for the observation object) as an absorption and recovery function of sample paths for some auxiliary random process. The solution of analysis problem can be found approximately by using numerical methods for solving stochastic differential equations and methods for modeling inhomogeneous Poisson flows. The modeling algorithm for observation system and optimal estimation of its state based on the maximal section method is given in the paper. The main advantages of this algorithm are easy implementation and universality, namely the possibility of solving the optimal filtering problem for linear and non-linear models of the observation system, for one-dimensional and multidimensional case. Key words: branching processes, conditional density, Duncan–Mortensen–Zakai equation, Monte Carlo method, optimal filtering problem, stochastic system. Contacts: E-mail: rkoffice@mail.ru Pp. 15-21.