Study On Transient Electromagnetic Method Of Denoising Based On Hilbert-Huang Transform

Transient electromagnetic method(TEM)as an important branch of electrical exploration,because of the small influence of topography,high working efficiency,good resolution for low resistivity bodies and pure anomaly measurement.the transient electromagnetic method has been widely used in deep metal ore exploration,geothermal exploitation,oil,gas and water resources exploration and so on.However,the data of transient electromagnetic is liable to be disturbed by peripheral noise,especially for the latest data,during acquisition in the field,because the power is low,also the signal is weak in the latest time,and the existence of noise seriously pollutes the useful signal.It is easy to cause false exception in the process of data processing.Therefore,that is of great significance to find a suitable denoising method to deal with the measured data and improve the analysis quality of the data,which is of great significance for acquiring more accurate underground geological information.Through a large number of reading documents,the common noise interference in the field of transient electromagnetic method is summed up.The noise sources are divided into geological noise,natural noise and human noise,and the noise generation mechanism and noise characteristics are analyzed.EMD decomposition,as a proven signal processing method,has been widely used in data processing.The mechanism of EMD decomposition is understood by deducing the principle of HHT transformation.In view of the influence of EMD decomposition boundary effect,polynomial fitting method is adopted to suppress the endpoint effect by interpolating the data beyond the endpoints at the endpoints.According to the characteristics of the space-electric noise,this paper uses the Alpha stable distribution to simulate the space-time noise,changes the Alpha value to obtain the pulse sequence at different time points,and then simulates the theoretical simulation data by changing the amplitude.The threshold value is set by analyzing the curve after being interfered with by the weather electricity,and according to the law of TEM attenuation approximate exponent,however,in a certain window,the data can be fitted by polynomial sum function.The interpolation method is used to replace the data after the interference,and the denoising of the interference is realized.The results are compared with the results of MATLAB smoothing filter.The former has better fitting effect and low efficiency,while the latter can suppress noise,but the effect is not as good as the former.When gaussian noise with different intensities was added to the theoretical data,the shape of the curve became more obvious with the increase of the intensity of the noise.Finally,the EMD decomposition method is used to decompose,through time-frequency analysis,the empirical mode function of noise is set zero,and the useful signals are synthesized to achieve the purpose of de-noising.In order to find out the influence of harmonics on transient electromagnetic data,the paper simulates 50 Hz harmonics with different intensities,mixed frequency signals,and adds gaussian noise and sinusoidal wave to the theoretical data respectively.The law of influence is analyzed.At the same time,with the help of EMD decomposition,according to the multi-scale and multi-resolution of EMD decomposition,it can deal with the characteristics of nonlinearity and nonstationarity.By comparing the data before and after denoising,the data processed by EMD are in good agreement with the original data,which shows that Hilbert-Huang transform has a good effect in dealing with noise.Finally,through an example,give a transient electromagnetic data processing flow,meanwhile through the above method,data preprocessing is carried out by using the EMD method and compared with the original data,which indicates the practicability of the method.