Research On Time Domain Noise Reduction Method Of Magnetotelluric Sounding And Its Engineering Application

Audio magnetotelluric sounding uses the natural geomagnetic field as the field source,estimates the impedance using the observed electric field and magnetic field components,and then calculates the electrical structure of the underground rock and soil layer,so as to achieve the exploration purpose.This method is widely used in engineering exploration because it is not shielded by high resistance and the instrument is lightweight.However,the audio magnetotelluric sounding method is very susceptible to noise interference within the observation frequency band.If not processed,the interpretation results will be affected.Therefore,how to suppress noise in time-domain data effectively to improve the data quality of audio-magnetotelluric sounding is a problem worth studying.This paper takes improving the accuracy of audio magnetotelluric sounding as a starting point,and studies the noise removal of the collected data in time domain.On the basis of fully understanding the audio-magnetotelluric sounding method,the characteristics of natural magnetotelluric field sources are studied,the waveforms of common human noise disturbances in the measured data in time series and their possible effects are systematically analyzed.In view of the noise interference of audio magnetotelluric sounding data,the means of wavelet analysis is introduced,and the threshold denoising method of wavelet analysis is selected as the data processing method.The effect of different parameters of the method on noise suppression is analyzed in detail,and the signal-to-noise ratio and root mean square error are used as the objective evaluation index of the denoising effect.By processing the analog signal,the optimal combination parameters for pulse and square noise are selected.Through analysis,it is found that the wavelet threshold denoising method has a good suppressing effect on white Gaussian noise,but not on pulse and square noise.In this case,a combination of wavelet threshold denoising and variational mode decomposition is presented.The results of numerical simulation show that this method can suppress impulse and square noise very well.In order to further verify the validity of this method,it is applied to the processing and analysis of the measured data.By comparing the impedance curves before and after noise removal,it shows that this method can suppress noise.In order to validate the effectiveness of this method,the research results are applied to the exploration project of Dagou River water transmission tunnel,which greatly improves the exploration accuracy.