REVISTA DYNA NEW TECHNOLOGIES

Key words:

estimador, secuencia de matrices, funcional del error, gradiente estocástico, seudoinversa, segundo momento de probabilidad, Estimator, the sequence of matrices, functional error, stochastic gradient, pseudoinverse, the second moment

Article type:

ARTICULO DE INVESTIGACION / RESEARCH ARTICLE

Section:

RESEARCH ARTICLES

ABSTRACT:
The internal parameters are not observable in a black box system, only its input-output ratio. The estimate is one of the areas of the filter that aims to describe the parameters system. In this paper, is used an estimator based on a technique of sequence of matrices to approximate the model reference to the answer given, considered: a) the recursive functional error, b) the filters in tremendous differences. Independent of the estimation structure, its results identify states known as the adaptive filtering process. Both estimators have a degree of convergence towards optimal probability, as shown in the simulations.

Keywords: Estimator, the sequence of matrices, functional error, stochastic gradient, pseudoinverse, the second moment