Research On Envelope Extraction Algorithm Of Magnetic Resonance Sounding Signal Based On Adaptive Local Iterative Filtering

Magnetic Resonance Sounding(MRS)is a geophysical method to detect groundwater.Its most outstanding advantage is that the content of groundwater can be directly and quantitatively determined.In practical applications,the background fields using the geomagnetic field,MRS signal itself are faint and experiment in the field environment cannot take effective shielding method,thus the environment random noise,power-line harmonic noise and spiky noise will affect the nanovolt-level MRS signal,which makes it difficult to effectively extract the MRS signal.Adaptive Local Iterative Filtering(ALIF)is a nonlinear and non-stationary signal processing method proposed by Antonio Cicone in 2016.Based on the idea of iterative decomposition,the Intrinsic Mode Function(IMF)is constructed by using the basic solution system of The Fokker-Planck partial differential equation.The Intrinsic Mode Function(IMF)is adaptive,and accurate Intrinsic Mode Function(IMF)can be obtained by selecting appropriate parameters.It has a solid mathematical foundation and has been applied in the fields of hydraulic power generation,wind power generation and rolling bearing fault detection.In view of the many advantages of this algorithm and the difficulty of effective extraction of MRS signals under the interference of complex noises,this paper presents the research of MRS signal envelope extraction algorithm based on ALIF.Firstly,according to the mathematical model of the MRS signal and the timefrequency characteristics of common noise,a data model of the noisy MRS signal is established.Based on the basic principle of ALIF algorithm,the envelope extraction of MRS signal is realized.On this basis,based on the uncertainty of the initial amplitude and relaxation time,the selection of two key parameters(decomposition order and mask coefficient)affecting the performance of the ALIF algorithm is realized by traversal.Through simulation experiments,it is proved that under the interference of Gaussian noise and power-line harmonic noise,the ALIF algorithm can effectively extract the envelope of the MRS signal,and the correlation coefficient with the noise-free envelope reaches 0.97.Secondly,according to the selected algorithm parameters,the stability analysis of the ALIF algorithm is carried out.When the white noise amplitude,power-line harmonic noise amplitude,average relaxation time and Larmor frequency and other factors change,the ALIF algorithm is used for processing,and the average signal-to-noise ratio is increased by 30?42 d B after processing.The fitting error is within ±5%,and the fitting error of the mean relaxation time is within ±5%.According to the analysis of the obtained statistical results,the stability of the algorithm is verified.At the same time,a comparative analysis experiment was carried out between the ALIF algorithm and the classical harmonic modeling method and ensemble empirical mode decomposition(EEMD)algorithm.The results show that the ALIF algorithm has more advantages in terms of processing efficiency and accuracy of results.In addition,in order to verify the universality of the ALIF algorithm,the processing and analysis of the multiexponential MRS signal data was carried out in this paper,which still achieved relatively good results.Finally,In this paper,the acquisition and processing of outdoor simulated test data and field measured data based on ALIF algorithm are carried out.The outdoor simulated test data and field measured data were collected in Changchun Culture Square and Changchun Shaoguo Town respectively.The MRS signal of outdoor simulated test was processed by ALIF algorithm,and the SNR was improved by about30 d B.In addition,compared with the processing results of harmonic modeling method and EEMD algorithm,the advantages of the proposed algorithm are further demonstrated,and the effectiveness and practicability of the proposed algorithm are verified.The paper's innovative work is as follows :1.To solve the problem of difficult extraction of MRS signal envelope,an adaptive local iterative filtering method was proposed to effectively extract MRS signal envelope.2.The best combination of mask coefficient and decomposition order is determined by traversal method,so that the algorithm can be optimized and the accuracy of envelope extraction can be improved.