A Practical Method for Determining Coefficient Word Length in Digital Filters

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

This correspondence presents a new method for assessing the sensitivity of recursive digital filters to changes in their coefficients, and is particularly aimed at real-time control applications where continuous filters are being emulated in digital form. The method described enables sensitivity factors to be derived which directly relate the accuracy of the digital filter coefficients to the parameters of the continuous filter. By this means, for a particular percentage accuracy requirement in the parameters of the continuous filter, it is possible to determine directly the percentage accuracy (and hence word length) for the coefficients of the digital filter. Application of the method to typical examples is given, and the sensitivity predictions are confirmed by simulations of the digital filters in response to step inputs. An overview of its use in a design sense is also included.

LanguageEnglish
Pages981-985
Number of pages5
JournalIEEE Transactions on Signal Processing
Volume40
Issue number4
DOIs
Publication statusPublished - 1 Apr 1992
Externally publishedYes

Fingerprint

Digital filters
Real time control

Cite this

@article{7ddc1fb9f3e44e1fb3ca8a52ee097f30,
title = "A Practical Method for Determining Coefficient Word Length in Digital Filters",
abstract = "This correspondence presents a new method for assessing the sensitivity of recursive digital filters to changes in their coefficients, and is particularly aimed at real-time control applications where continuous filters are being emulated in digital form. The method described enables sensitivity factors to be derived which directly relate the accuracy of the digital filter coefficients to the parameters of the continuous filter. By this means, for a particular percentage accuracy requirement in the parameters of the continuous filter, it is possible to determine directly the percentage accuracy (and hence word length) for the coefficients of the digital filter. Application of the method to typical examples is given, and the sensitivity predictions are confirmed by simulations of the digital filters in response to step inputs. An overview of its use in a design sense is also included.",
keywords = "digital filters, Transfer function, Nonlinear Filter, predictive model, low pass filters, laplace equations, difference equations, sampling methods, accuracy, vents, Sensitivity analysis",
author = "Goodall, {R. M.}",
year = "1992",
month = "4",
day = "1",
doi = "10.1109/78.127972",
language = "English",
volume = "40",
pages = "981--985",
journal = "IEEE Transactions on Signal Processing",
issn = "1053-587X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

A Practical Method for Determining Coefficient Word Length in Digital Filters. / Goodall, R. M.

In: IEEE Transactions on Signal Processing, Vol. 40, No. 4, 01.04.1992, p. 981-985.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A Practical Method for Determining Coefficient Word Length in Digital Filters

AU - Goodall, R. M.

PY - 1992/4/1

Y1 - 1992/4/1

N2 - This correspondence presents a new method for assessing the sensitivity of recursive digital filters to changes in their coefficients, and is particularly aimed at real-time control applications where continuous filters are being emulated in digital form. The method described enables sensitivity factors to be derived which directly relate the accuracy of the digital filter coefficients to the parameters of the continuous filter. By this means, for a particular percentage accuracy requirement in the parameters of the continuous filter, it is possible to determine directly the percentage accuracy (and hence word length) for the coefficients of the digital filter. Application of the method to typical examples is given, and the sensitivity predictions are confirmed by simulations of the digital filters in response to step inputs. An overview of its use in a design sense is also included.

AB - This correspondence presents a new method for assessing the sensitivity of recursive digital filters to changes in their coefficients, and is particularly aimed at real-time control applications where continuous filters are being emulated in digital form. The method described enables sensitivity factors to be derived which directly relate the accuracy of the digital filter coefficients to the parameters of the continuous filter. By this means, for a particular percentage accuracy requirement in the parameters of the continuous filter, it is possible to determine directly the percentage accuracy (and hence word length) for the coefficients of the digital filter. Application of the method to typical examples is given, and the sensitivity predictions are confirmed by simulations of the digital filters in response to step inputs. An overview of its use in a design sense is also included.

KW - digital filters

KW - Transfer function

KW - Nonlinear Filter

KW - predictive model

KW - low pass filters

KW - laplace equations

KW - difference equations

KW - sampling methods

KW - accuracy

KW - vents

KW - Sensitivity analysis

UR - http://www.scopus.com/inward/record.url?scp=0026840866&partnerID=8YFLogxK

U2 - 10.1109/78.127972

DO - 10.1109/78.127972

M3 - Article

VL - 40

SP - 981

EP - 985

JO - IEEE Transactions on Signal Processing

T2 - IEEE Transactions on Signal Processing

JF - IEEE Transactions on Signal Processing

SN - 1053-587X

IS - 4

ER -