شماره ركورد كنفرانس :
4891
عنوان مقاله :
Data filtering methods for collected data by Acoustic Doppler velocimeter
Author/Authors :
Hossein Mohajeri Dept. of Civil and environmental Engineering - Tarbiat Modares University , Akbar Safarzade Faculty of Engineering - Mohaghegh Ardebili University , Ali Akbar Salehi Neyshabouri Dept. of Civil and environmental Engineering - Tarbiat Modares University , Masoud Ghodsian Dept. of Civil and environmental Engineering - Tarbiat Modares University
كليدواژه :
Turbulence , Data filtering , Acoustic Doppler velocimeter , Velocity signal processing
عنوان كنفرانس :
نهمين كنگره بين المللي مهندسي عمران
چكيده لاتين :
Acoustic Doppler velocimeter (ADV) measurements are useful for measurements of mean and turbulent
components of fluids in both experimental and field studies. Although ADV has been used commonly for
velocity measurements, existence of spikes in the recorded velocity series leads to unreliability of these
series. In the present study, four famous despiking algorithms, consist of RC Filter method, Tukey 53H
method, acceleration thresholding method and Wavelet thresholding method, were applied to collected
data in an straight channel with smooth and rough bed condition. In addition, the wavelet thresholding
method as a common and high efficient method was modified. The results show that acceleration
thresholding method and modified wavelet method reject spike points and accept valid points. So these
methods purify the contaminated series appropriately.
