Improved empirical mode decomposition method based on amplitude–frequency characteristic of vortex signals

Jie Chen,Fu-Rong Tian,Bin Li,Wei Zhang,Shu-Xu Wan,Xue-Qing Hou

DOI: https://doi.org/10.1063/5.0175496

IF: 1.6

2024-04-01

Review of Scientific Instruments

Abstract:The vortex flowmeter occupies a vital position in flow measurement with its unique advantages. It is essentially a fluid vibration instrument, and its measurement process is susceptible to interference, which seriously affects measurement accuracy. In particular, at low flow rates, it is an urgent problem to extract vortex signals from the complex noise. Among many signal processing methods, Empirical Mode Decomposition (EMD) is a time-frequency analysis method suitable for nonlinear, non-stationary signals. EMD can adaptively decompose noisy signals into noise and useful signal components arranged from high frequency to low frequency. For the above problems, an innovative, improved EMD method is proposed in this paper. The digital filter is designed according to the amplitude–frequency characteristic of vortex signals. After filtering, the vortex signal is adjusted to a fixed value, and high-frequency noise is filtered. According to the consistency of the filtered signal’s amplitude, we design a decomposition stop criterion for EMD to process the output signal of the vortex sensor. This method not only maintains the characteristic of adaptive decomposition in EMD but also completes the automatic extraction of the vortex signal under complex noise. It provides a new comprehensive method for realizing high-precision and anti-interference vortex flowmeters.

instruments & instrumentation,physics, applied