A Novel De-Noising Algorithm of Fission Neutrons Signals by Photoelectric Detection

Abstract

The empirical mode decomposition (EMD) is applied to analysis and processing of the fission neutrons events based on photoelectric detection. A new model on the de-noise of fission neutrons signals is established with self-adapting double-smoothness-detecting (DSD) algorithm on intrinsic mode function (IMF), and wavelet threshold filter. With the simulated and survey data, a series of comparisons with other signal de-noise methods are made. On an equal footing, the EMD-based-DSD has a better adaptability on various fission neutrons signals. The signal-to-noise ratio of the treated signal is higher while the waveform is more even. The loss of useful information caused by over-de-noise is avoided. Moreover, the de-noise degree can be customized quantifiably.