Sparse Fourier transform and amplitude–frequency characteristics analysis of vortex street signal

Abstract

Vortex flowmeter is a commonly used flow measurement device. It is almost not affected by the density and viscosity of the fluid, so the vortex flowmeter can be used for the detection of various medium, such as gas, liquid and steam. When dealing with vortex street signal, we usually use fast Fourier transform to calculate the signal frequency, but this traditional vortex street signal processing method is not only inefficient, but also difficult to filter out noise signals in the same frequency band as the vortex signal. The sparse Fourier transform utilizes the sparsity of the signal to efficiently calculate the signal spectrum, and the computational complexity is lower than that of the fast Fourier transform algorithm. In this paper, the amplitude and frequency of the vortex signal is analyzed by sparse Fourier transform and the noise signal is removed based on the amplitude–frequency characteristics of the vortex signal. Finally, by comparing with the other methods, we found that the time complexity of our algorithm is one-tenth of others’ methods. This means that our approach is 10 times faster than others.