An Intelligent Full-Wavenumber Fitting Algorithm for Proton Precession Magnetometer Frequency Measurement

For proton precession magnetometers, the influence of free induction decay (FID) signal noise on the frequency measurement procedure is a significant impediment to improving the precision of geomagnetic field measurements. However, the existing noise suppression methods treat the noise as a system, aiming to reduce its impact on frequency estimation without directly addressing the specific noise interference. To address this issue, an intelligent full-wavenumber fitting algorithm for measuring the FID signal frequency is proposed. It entails converting all the FID signal waves within the sampling cycle into a time series with the signal period as the independent variable. Effectively determining the frequency of the FID signal by fitting a linear regression on this time series of the full-wavenumber FID signal. Moreover, we introduce an intelligent noisy point detection method by analyzing the deviation between the time series and the fit curve. This enables targeted and localized noise elimination, resulting in improved noise suppression. To validate the effectiveness and character of the performance of the proposed algorithm, we conducted rigorous testing on a purpose-built experimental platform and the standard magnetism laboratory. Using a simulated FID signal, the results demonstrate that the proposed algorithm can obtain a precision greater than
${1.47} \times {10}^{-{5}}$
Hz and a resolution greater than
${3} \times {10}^{-{6}}$
Hz within a measurement time of 1000 ms. When aligned the sensor of the proposed prototype is with a standard artificial magnetic field, it demonstrates a 0.0667-nT measurement precision.