Extraction of useful data from noise-contaminated PDFs

✍ Scribed by Warren C. Strahle; Wilhelmus A. Degroot

• Publisher: Elsevier Science
• Year: 1988
• Tongue: English
• Weight: 838 KB
• Volume: 74
• Category: Article
• ISSN: 0010-2180
• DOI: 10.1016/0010-2180(88)90017-x

✦ Synopsis

The problem treated in this paper is the extraction of desired turbulence data from experimentally determined probability density functions (PDFs) that are contaminated by undesired noise. The mathematical and numerical problem trealed is the inversion of an integral equation of the first kind. Starting with an exact solution to the problem, several variants are presented through Fourier transform techniques, based upon orthogonal-function expansions and delta-function approximations to continuous distributions. The applicability of this method is shown in the deconvolution of some joint PDFs of velocity and species concentration using laser velocimetry and Rayleigh molecular scattering. An analytical example is given that illustrates statistical sampling error with the given techniques.