De-noising of Gamma-ray Pulse Signal Based on HILBERT-Huang Transform

Abstract

Gamma-ray energy spectrum is obtained by statistical calculation of gamma-ray pulse peak data and based to do the qualitative and quantitative analysis of radioactivity. In practice, the acquisition of gamma-ray pulse data will inevitably be affected by noise, which is challenging to determine the peak value accurately. Traditional noise suppression methods mainly use hardware methods, threshold methods, etc. These methods prevent noise pulses from being wrongly judged as characteristic pulses. However, the superimposed effect of noise signals on real characteristic pulse data is usually ignored. According to the nonlinear and non-stationary characteristics of the gamma-ray pulse signal. This work uses the Hilbert-Huang Transform (HHT) to denoise the gamma pulse signal. The purpose of noise suppression is achieved by Empirical mode decomposition (EMD) and Hilbert spectrum analysis (HSA) of the original data. This work uses the experimental data and simulated data in the experiment. The experimental results under different noise intensities show that the method can effectively suppress the influence of noise on gammaray pulse signal processing. The classical algorithm wavelet transform was conducted as a comparative experiment. By comparison, it is found that HHT has a better noise suppression ability under strong noise background. Meanwhile, HHT has a better adaptability and shows good noise suppression ability under different noise intensities. This work verifies the feasibility of HHT to denoise gamma pulses and provides a new option for gamma-ray pulse signal analysis.