Please use this identifier to cite or link to this item:
|Title:||COMPRESSIVE SAMPLING FOR ENERGY SPECTRUM ESTIMATION OF TURBULENT FLOWS|
|Authors:||Kevlahan, Nicholas K.-R.|
Adalsteinsson, Gudmundur F.
|Keywords:||compressive sampling;turbulence;energy spectrum;wavelets;optimization|
|Citation:||Adalsteinsson, G. and Kevlahan, N.K.-R. 2015 Compressive sampling for energy spectrum estimation of turbulent flows. SIAM J. Sci. Comput. 37, B452-B472|
|Series/Report no.:||SIAM J. SCI. COMPUT.;|
|Abstract:||Recent results from compressive sampling (CS) have demonstrated that accurate reconstruction of sparse signals often requires far fewer samples than suggested by the classical Nyquist–Shannon sampling theorem. Typically, signal reconstruction errors are measured in the 2 norm and the signal is assumed to be sparse or compressible. Our spectrum estimation by sparse optimization (SpESO) method uses a priori information about isotropic homogeneous turbulent flows with power law energy spectra and applies the methods of CS to one- and two-dimensional turbulence signals to estimate their energy spectra with small logarithmic errors. SpESO is distinct from existing energy spectrum estimation methods which are based on sparse support of the signal in Fourier space. SpESO approximates energy spectra with an order of magnitude fewer samples than needed with Shannon sampling. Our results demonstrate that SpESO performs much better than lumped orthogonal matching pursuit, and as well as or better than wavelet-based best M-term or M/2-term methods, even though these methods require complete sampling of the signal before compression.|
|Appears in Collections:||Faculty Publications|
Files in This Item:
|SIAM_J_Sci_Comput_v37_2015.pdf||Main article||1.04 MB||Adobe PDF||View/Open|
Items in MacSphere are protected by copyright, with all rights reserved, unless otherwise indicated.