Industrial Automatic Control Systems and Controllers Annotation << Back An Adaptive Algorithm for Nonlinear Filtering Determine the State Vector of a dynamic System Based on Conditional Cumulants Morozova T.Yu. Existing methods for determining the exact density of the posterior probability can be reduced to the solution of stochastic equations whose integration is not possible. An algorithm for solving systems of equations for the parameters that characterize this density. As these parameters are chosen cumulants. On the basis of the conditional cumulants developed an adaptive method for nonlinear filtering and recovery, which allows you to build non-linear filter for a finite number of ordinary differential equations. Algorithms filter settings to reduce the displacement estimates, they are simple, and their implementation is not too difficult. The simulation results of a number of problems give rise to the practical application of the algorithm by reducing the bias estimates in determining the coordinates of the state vector of a nonlinear dynamical system in the presence of additive white noise. Keywords: Kalman filter; cumulants; nonlinear filtering; adaptive method for nonlinear filtering; dynamic system. Contacts: E-mail: tmorozova200@rambler.ru Pp. 33-38.