Research On The Human Noise De-Noising For Frequency-Domain Electromagnetic Data Based On Wavelet Transform Algorithm

Nowadays,the increase of population and the rapid development of industrialization are the main reasons for the increase of the consumption of mineral resources.However,the shallow area of mineral resources after decades of exploitation,has gradually dried up.Therefore,in order to alleviate the energy crisis,it is urgent to improve the prospecting technology to detect the second mining space(underground 500m~1500m depth).Magnetotelluric sounding method(MT)has a high application value in deep exploration due to its advantages of large depth of detection,high resolution in horizontal direction,convenience in field experiment and low cost.Because of the weak energy and wide frequency band,MT data is easy to be disturbed by the external disturbance.Generally,there are five main types human interference,such as the impulse noise,square wave noise,step noise,triangle wave noise and periodic noise,may be affected during the field experiment.These noises are relatively large in energy,which seriously affect the quality of the observed data,and even drown the signal,which lead to the deviation of the inversion results,and the error of the resource exploration results.Therefore,it is necessary to introduce a reasonable noise reduction method to deal with the MT signal,so as to improve the accuracy of electromagnetic data.Not only for high frequency electromagnetic signal is susceptible to noise interference,and the low frequency part also easily affected by the long cycle square wave noise,combining the characteristics of multi-resolution analysis algorithm and wavelet threshold algorithm,a combined wavelet de-noising method is proposed based on wavelet transform.The optimal wavelet base and threshold rule are selected by the combination of the simulated signal de-noising experiment and the measured signal experiment.The long-period noise was removed using multi-resolution analysis algorithm;using wavelet threshold algorithm with db6 wavelet-base based on Bayes estimation with improved threshold function to remove noise and short cycle.The processing results of measured data show that the quality of time series waveforms and the Cagniard apparent resistivity curves and the phase curves have been improved,and the near-source effect has been obviously suppressed.Finally,a software based on MATLAB GUI is designed using a comprehensive wavelet noise reduction method in the article,it can process analog signals and measured signals.You can start by observation and analysis of the pollution type measurement of noise signal by the noise,and then use the analog signal generated sine wave superposition and adding related species,and then through the wavelet parameter modification interface to adjust the optimal wavelet parameters,which is applied to the noise reduction was measured by similar data interference conditions,and,a better noise reduction effect is obtained.The software not only to simplify the process of noise reduction and improve the efficiency of parameter adjustment,but also to make the signal processing work more intuitively and to shorten the data processing timeIn summary,the results show that the comprehensive wavelet de-noising method can suppress the human disturbance noise in the geomagnetic data,improve the quality of field MT data,and provide the more accurate geomagnetic data for further geological exploration.In addition,with the noise reduction software for signal processing,it is able to achieve a great effect and has a high value of application.