Research On An Envelope Extraction Algorithm Of Magnetic Resonance Sounding Signal Based On Spectrum Analysis

In the field of geophysical detection,the most advanced method of groundwater detection in the world is to use Magnetic Resonance Sounding(MRS)technology to quantitatively measure underground aquifers.Because the magnetic resonance sounding technology is non-invasive,does not require drilling,and is only sensitive to hydrogen protons in groundwater,this technology has been widely used in exploration work such as groundwater resources surveys and early warning of water-induced disasters in underground engineering,The content involves water resources survey,evaluation,leakage of dams,water inrush from mine tunnels and other fields.However,in the actual work process of this method,because the detection work is carried out in the natural geomagnetic field,shielding measures cannot be taken.In addition,the MRS signal amplitude is extremely small(nanovolt level),which makes it easy to collect MRS signals.Affected by interference such as spike noise,Gaussian white noise and power frequency harmonics in the external environment,the extraction of MRS signal envelope will be affected,and the difficulty and accuracy of inversion and interpretation of groundwater-related information will be increased.Therefore,research on the envelope extraction algorithm of the MRS signal under the interference of complex noise will help to solve the current technical barriers of NMR sounding,and is of great significance for improving the signal-to-noise ratio and improving the accuracy of the inversion results.Spectrum Analysis(Spectrum Analysis,SA)is a method used to study the characteristics of a signal or system in the frequency domain.The signal is sampled in the time domain,and the collected signal is Fourier transformed to obtain a spectrum signal,and then the frequency domain is sampled to obtain the frequency.Signals in the frequency domain can achieve high signal-to-noise ratio envelope extraction in the frequency domain.After comparing and analyzing several commonly used spectrum analysis methods,this paper proposes the use of sliding window spectrum analysis method to extract MRS signal envelope.By discussing the optimal window length and window shape,the optimal parameters suitable for MRS signal processing are determined,and the algorithm simulation analysis and measured data processing under different environmental conditions have further verified the effectiveness of the algorithm.The main research contents are as follows:(1)By analyzing the characteristics of the MRS signal and related noise,it is concluded that the characteristics of the MRS signal are different from the noise.It is a single-frequency oscillating signal that decays exponentially according to e.This feature lays the foundation for the subsequent development of a sliding window-based spectral analysis envelope extraction method basis.(2)Window length and window shape are two important parameters that affect whether the sliding window spectrum analysis method can obtain the best signal-to-noise ratio.Through theoretical analysis,this paper has determined that the optimal window length for MRS signal is 0.5s;at the same time,considering the exponential attenuation of MRS signal e,it is proposed to use exponential window as the window function,which is equivalent to imposing different signal and noise on the signal and noise.At the same time,when the attenuation rate of the window function is equal to and obtained through discussion,the signal-to-noise ratio of the extracted envelope can be maximized.(3)Based on the sliding window spectrum analysis method,the research on the envelope extraction algorithm of noiseless and noisy MRS data is carried out,and the feasibility of the algorithm is verified.Simultaneously,simulation experiments with different frequency deviations and ? were carried out to verify the universality of the algorithm,and compared experiments with classical harmonic modeling and empirical mode decomposition(EMD)algorithms were carried out to verify the performance of the method proposed in this paper.Superiority.(4)MRS data collection and processing experiments were carried out in the field.The experimental results show that the proposed algorithm can effectively extract the envelope with high signal-to-noise ratio in complex environmental noise,which fully demonstrates the effectiveness and practicability of this algorithm.