Research On De-Noising Method Of Grounded Electrical Source Airborne Transient Electromagnetic Signals Based On Singular Spectrum Analysis

Grounded Electrical Source Airborne Transient Electromagnetic Method(GREATEM)has the characteristics of high detection efficiency,low cost,and high resolution.It has been widely used in deep mineral exploration,oil and gas resource evaluation,and hydrological environmental surveys.However,due to the characteristics of the frequency bandwidth of the ground-air transient electromagnetic signal,the actual collected data is easily mixed with a variety of noise interference,which seriously affects the signal quality and easily causes false anomalies in subsequent data processing.To process the signature,it is necessary to use an appropriate method to denoise the received data.Singular Spectrum Analysis(SSA)is developed from the Karhumen-Leove decomposition theory.It is a new time series analysis method in recent years,and it is also a semi-parametric spectral analysis method.The SSA algorithm realizes signal-tonoise separation of signals by constructing time-delay matrix,singular value decomposition,regrouping and diagonal averaging,and analyzes singular spectrum characteristics and signal attenuation characteristics.According to the characteristics of ground-air transient electromagnetic signals,this paper proposes to use SSA to denoise the received ground-space transient electromagnetic data.First,a denoising simulation experiment with Gaussian white noise,power frequency harmonic noise,sky-electric noise and mixed noise was carried out.By comparing the curves before and after denoising,it was proved that the SSA algorithm can eliminate the noise in the GREATEM signal;a quasi-two-dimensional The earth simulation model,by comparing the inversion results of the apparent resistivity profile,proves that the SSA algorithm can eliminate false anomalies caused by noise.Aiming at the problem of parameter selection of SSA algorithm,an iterative experiment is firstly carried out to simulate and analyze the effect of different window lengths on the signal-to-noise separation effect of GREATEM signals,and a particle swarm optimization algorithm(PSO)is proposed to use the permutation entropy as the fitness function for the SSA algorithm.Perform parameter optimization.The PSO-SSA algorithm was used to denoise the GREATEM signal containing Gaussian white noise,power frequency harmonic noise,sky-electric noise and mixed noise,and compared with the existing classical denoising algorithms,it proved the effectiveness of the PSOSSA algorithm.Finally,the experiments of eliminating measured noise and denoising of measured data were carried out.The algorithm proposed in this paper adaptively selects the SSA parameters,which can effectively remove the random environmental noise in the GREATEM signal.Noise processing,the inversion results of the processed data are closer to the actual geological situation,which further verifies the practicability of the algorithm in this paper.