WIT Press


Constrained, Weighted, Least Square Technique For Correcting Experimental Data

Price

Free (open access)

Paper DOI

10.2495/CMEM930322

Volume

5

Pages

14

Published

1993

Size

871 kb

Author(s)

S.A. Lukasiewicz & M. Stanuszek

Abstract

Constrained, weighted, least square technique for correcting experimental data S.A. Lukasiewicz^ M. Stanuszek^ of Mechanical of Calgary, Canada ^Computer ABSTRACT The paper presents a simple filtering routine for the elimination of the noise from the experimental data. The method consists in the application of the con- strained, weighted least square technique and the finite difference representation of the constraints. A simple formula for the filter matrix was found which makes this approach very efficient in filtering of the noisy data. INTRODUCTION Very often, the experimental data obtained to identify or verify the phenom- ena in science and engineering are affected by errors introduced by measuring instruments and methods. Sometimes, the noise is at such a level that the data are difficult to interpret and some techniques are required to remove the noise from the data without altering the true character of the signal. A typical way of smoo

Keywords