Prefiltration in Kalman filter with a priori indeterminacy

A stable nonlinear discrete Kalman filter is synthesized for measurement objects with a priori indeterminate, and variable, signal and noise distribution. A robust algorithm of prefiltration is used for obtaining stable maximum-likelihood estimates of the state of a dynamic system in the case of non-Gaussian noise. The estimation problem is expressed in the form of a recurrence relation containing a positive-definite symmetric matrix, whereupon the synthesis problem is formulated in terms of a system of corresponding difference equations in which the measurement errors have been converted to Gaussian ones.