Estimation of errors of quasilinear filtering method (Q800880)

An error analysis of the linear filtering method is presented that uses statistical linearization of the nonlinear functions in the right-hand sides of the differential equations describing the signal. The numerical investigation results show that if Kalman filters are applied to a statistically linearized nonlinear system (a linear model of the signal), this may cause biased estimates and a strong deviation of the model signal filtering error variance from the true signal filtering error variance.