THE USE OF NONLINEAR PROBABILISTIC
CRITERIA
TO SOLVE ADAPTIVE FILTERING PROBLEMS
S. V. Sokolov, I. V. Shcherban', V. A. Bertenev
Rostov Military Institute of Rocket Forces
The paper presents a general method for solution
of the adaptive parametric filtering problem requiring much less computational
effort at a potentially greater accuracy of estimation-identification as compared
with the traditional methods. The accuracy increase is obtained at the expense
of the use of more general probabilistic criteria instead of the mean -square
criterion. An example of identifying a stochastic nonlinear object is given.