Least-squares design of digital differentiators using neural networks with closed-form derivations

Author

Jou, Yue-Dar

Author_Institution

Dept. of Comput. & Inf. Sci., Mil. Acad., Kaohsiung, Taiwan

Volume

12

Issue

11

fYear

2005

Firstpage

760

Lastpage

763

Abstract

In this letter, a neural network implementation for the least-squares design of digital differentiators with closed-form derivations for Hopfield-related parameters are proposed. Using this technique, the optimal filter coefficients are obtained by iteratively updating the dynamic nonlinear equations of the network. Simulation results indicate that the proposed technique has the advantage of effectiveness and is suitable for hardware implementation in real time.

Keywords

Hopfield neural nets; differentiation; digital filters; iterative methods; least squares approximations; nonlinear differential equations; real-time systems; signal processing; Hopfield-related parameter; closed-form derivation; digital differentiator; dynamic nonlinear equation; least-squares design; neural network implementation; optimal filter coefficient; real time system; Closed-form solution; Digital filters; Equations; Finite impulse response filter; Hardware; Hopfield neural networks; Military computing; Neural networks; Symmetric matrices; Transformers; Digital differentiators; least squares; neural network; real time;

fLanguage

English

Journal_Title

Signal Processing Letters, IEEE

Publisher

ieee

ISSN

1070-9908

Type

jour

DOI

10.1109/LSP.2005.856880

Filename

1518895