A method for correcting power frequency interference of pre-stack data in transient electromagnetic exploration

Abstract

The application discloses a transient electromagnetic exploration pre-stack data power frequency interference correction method, which comprises the following steps: step one, observing to obtain no electromagnetic interference pre-stack full waveform data and electromagnetic interference area pre-stack full waveform data; step two, acquiring full waveform electromagnetic interference noise data set according to the no electromagnetic interference pre-stack full waveform data and the electromagnetic interference area pre-stack full waveform data; step three, fitting the noise data set by using the least square method, and iteratively fitting the value of the model parameter; step four, calculating the power frequency electromagnetic noise according to the model parameter; and step five, correcting the noise data set by using the power frequency electromagnetic noise to obtain the transient electromagnetic exploration data after the power frequency interference is corrected. The power frequency electromagnetic interference meets the characteristics of the harmonic equation, the power frequency electromagnetic interference in the transient electromagnetic pre-stack full waveform data is solved by using the optimization theory, the composition of the electromagnetic noise in the observation signal is effectively reduced, and the field data quality and the exploration precision of the transient electromagnetic method are improved.

Description

Technical Field

[0001] This invention relates to the field of geophysical exploration technology, specifically to a method for correcting power frequency interference in pre-stack data of transient electromagnetic exploration. Background Technology

[0002] Transient electromagnetic (TEM) methods are widely used in mineral exploration, coal mine hydrogeological surveys, and engineering geology. Higher quality raw TEM data from field observations leads to higher exploration accuracy. However, with the continuous development of national industrialization, most exploration areas now have high-voltage power lines, resulting in low signal-to-noise ratios in the raw TEM data and affecting actual exploration accuracy. To suppress or eliminate this power frequency electromagnetic interference, notch filters are often added to the instrument design to target the interfering frequency. However, this method is not entirely effective. To address this problem, geophysicists have incorporated various filtering algorithms into post-processing data to suppress interference.

[0003] Chinese invention patent CN108776357B discloses a method and apparatus for correcting electromagnetic interference in transient electromagnetic methods in sedimentary formations. It suppresses noise data by performing polynomial fitting on densely sampled transient electromagnetic data. Chinese invention patent CN116449438B discloses a reference denoising device and method for transient electromagnetic methods. It estimates the ambient electromagnetic noise collected by the signal coil by adding a set of reference coils to receive ambient electromagnetic noise, thereby obtaining the denoised transient electromagnetic signal.

[0004] The aforementioned existing technologies all start from transient electromagnetic superposition data, which leads to the periodic power frequency signal becoming complicated in the process of positive and negative square wave superposition, and thus does not satisfy the harmonic equation. Regardless of whether polynomial fitting or other filtering methods such as electromagnetic interference estimation are used, the effect of suppressing power frequency interference signals is limited, thereby affecting the exploration results. Summary of the Invention

[0005] The purpose of this invention is to provide a method for correcting power frequency interference in pre-stack data of transient electromagnetic exploration, so as to solve the technical problem of low exploration accuracy caused by power frequency interference affecting the raw transient electromagnetic data in the prior art.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0007] A method for correcting power frequency interference in pre-stack data from transient electromagnetic exploration includes the following steps:

[0008] Step 1: Observe and obtain the full waveform data U before stacking without electromagnetic interference. 无电磁干扰 Electromagnetic interference zone pre-stack full waveform data U 电磁干扰 ;

[0009] Step 2: Based on the electromagnetic interference-free pre-stack full waveform data U obtained in Step 1 无电磁干扰 Electromagnetic interference zone pre-stack full waveform data U 电磁干扰 Obtain the full waveform electromagnetic interference noise dataset;

[0010] Step 3: Fit the noise dataset U obtained in Step 2 using the least squares method. 噪声干扰 Iteratively fit the values ​​of the model parameters, including the maximum value of the transmission line electromotive force E. 工频干扰 and phase difference

[0011] Step 4: Calculate the power frequency electromagnetic noise based on the model parameters obtained in Step 3;

[0012] Step 5: Correct the noise dataset obtained in Step 2 using power frequency electromagnetic noise to obtain transient electromagnetic exploration data U after power frequency interference correction. 校 :

[0013] U 校 =U 电磁干扰 -U 工频干扰 .

[0014] Further, the specific operation of step two: arbitrarily select a data point U free from power frequency electromagnetic interference. 无电磁干扰 As the background field, the pre-stack full waveform data U of the electromagnetic interference region obtained in step one is used. 电磁干扰 With pre-stack full waveform data U without electromagnetic interference 无电磁干扰 By subtracting the values, we obtain the dataset, which consists of the full-waveform electromagnetic interference noise. The noise dataset U is composed of 噪声干扰 i represents the sampling time t i The serial number.

[0015] Furthermore, in step three, the iterative formula is fitted:

[0016] (J T J+μI)ΔP k =J T Δd k

[0017] In the formula, J is the Jacobian matrix, k is the number of fitting iterations, k = 1, 2, ..., m, and m is the maximum number of fitting iterations; Δd k =U 噪声干扰 -U k For U 噪声干扰 The response U of the k-th fitted model k The residual; ΔP k The value is the correction factor for the k-th fitted model, obtained from the fitting iteration formula in each iteration; I is the unit vector; μ is the regularization factor, expressed as:

[0018]

[0019] In the formula, μ k Let P be the regularization factor corresponding to the k-th iteration. k The model parameters in the k-th iteration include the maximum value of the transmission line electromotive force E. 工频干扰 and phase difference

[0020] U 噪声干扰 =U 随机干扰 +U 地层感应 +U 工频干扰

[0021]

[0022] In the formula, U 噪声干扰 For a noisy dataset; U 随机干扰 This is random noise, measured in V; U 地层感应 Residual induction signal in the formation, unit: V; U 工频干扰 Power frequency electromagnetic noise, measured in V; E 工频干扰 t is the maximum electromotive force of the transmission line, in V; f is the frequency of the transmission line, which is 50Hz; t is the receiver observation time, in s. The phase difference is expressed in radians.

[0023] Finally, the model parameters of the transmission line are obtained after the m-th fitting iteration.

[0024] Further, the specific operation of step four: The model parameters calculated in step three... Substituting into the following formula, we can obtain the power frequency electromagnetic noise U generated by the transmission line. 工频干扰 ;

[0025]

[0026] In the formula, U 工频干扰 Power frequency electromagnetic noise is the power frequency noise data generated by power transmission lines, and its unit is V; E 工频干扰 t is the maximum electromotive force of the transmission line, in V; f is the frequency of the transmission line, which is 50Hz; t is the receiver observation time, in s. The phase difference is expressed in radians.

[0027] Compared to existing technologies, the method of this invention utilizes pre-stack data of transient electromagnetic full waveforms and achieves the separation of power frequency interference from effective signals through optimization theory. Specifically, during implementation, it leverages the characteristic that power frequency electromagnetic interference satisfies harmonic equations, and uses optimization theory to accurately solve for power frequency electromagnetic interference in the pre-stack full waveform data of transient electromagnetics. This effectively reduces the electromagnetic noise component in the observed signal, improves the signal-to-noise ratio of transient electromagnetic field data, and thus enhances the field data quality and exploration accuracy of the transient electromagnetic method. This method is applicable to transient electromagnetic exploration in areas with power frequency interference. Attached Figure Description

[0028] Figure 1 The general implementation steps of the method in this embodiment;

[0029] Figure 2 This refers to the full waveform data without power frequency noise received by the transient electromagnetic device in the embodiment;

[0030] Figure 3 The transient electromagnetic device in this embodiment receives full waveform data with power frequency noise.

[0031] Figure 4 The transmitted waveform of the transient electromagnetic device in the embodiment is shown.

[0032] Figure 5 The full waveform data of power frequency noise obtained in the example;

[0033] Figure 6 The transient electromagnetic full waveform data after power frequency electromagnetic interference was corrected for the example. Detailed Implementation

[0034] Unless otherwise specified, the technical terminology of this invention shall be understood by those skilled in the art.

[0035] like Figure 1 As shown, the power frequency interference correction method for pre-stack data of transient electromagnetic exploration provided by this invention is implemented using a transient electromagnetic exploration device. First, the pre-stack full waveform data U of the electromagnetic interference zone observed by the transient electromagnetic exploration device is processed. 电磁干扰 With pre-stack full waveform data U without electromagnetic interference 无电磁干扰 The difference is used to obtain the full waveform pure electromagnetic interference noise dataset U. 噪声干扰 Then, the least squares algorithm is used to fit the harmonic equations satisfied by the full waveform electromagnetic interference noise data to obtain the model parameters, thus obtaining the power frequency signal U in the electromagnetic interference zone. 工频干扰 Finally, the measured pre-stack full waveform data U in the electromagnetic interference region was used. 电磁干扰 The power frequency electromagnetic noise U in the electromagnetic interference region is solved. 工频干扰 By performing subtraction, transient electromagnetic exploration data U after correcting for power frequency electromagnetic interference is obtained. 校 .

[0036] Step 1: Obtain electromagnetic interference-free pre-stack full waveform data U using a transient electromagnetic exploration device. 无电磁干扰 Electromagnetic interference zone pre-stack full waveform data U 电磁干扰 ,See Figure 2 and Figure 3 .

[0037] Common transient electromagnetic exploration devices include a ground transmitter, a transmitting loop, a receiver, and a receiving probe. The ground transmitter, connected to the transmitting loop, is used to send trapezoidal magnetic field signals underground, with a transmitted waveform as shown in the image. Figure 4 As shown; the receiver is connected to the receiving probe and is used to receive the pre-stack full-waveform induced electromotive force signal.

[0038] Step 2: Obtain the full waveform electromagnetic interference noise dataset; the specific steps are as follows:

[0039] Arbitrarily select a data point U free from power frequency electromagnetic interference 无电磁干扰 As the background field, the pre-stack full waveform data U of the electromagnetic interference region obtained in step one is used. 电磁干扰 With pre-stack full waveform data U without electromagnetic interference 无电磁干扰 By subtracting the values, we obtain the dataset, which consists of the full-waveform electromagnetic interference noise. The noise dataset U is composed of 噪声干扰 i represents the sampling time t i The serial number.

[0040] Noise dataset U 噪声干扰 Includes random noise U 随机干扰 Residual induction signal U in the formation 地层感应 and power frequency electromagnetic noise U 工频干扰 As shown in equation (1):

[0041] U 噪声干扰 =U 随机干扰 +U 地层感应 +U 工频干扰 (1)

[0042] In the full waveform electromagnetic interference noise dataset U 噪声干扰 Only power frequency electromagnetic noise U 工频干扰 The power frequency interference harmonic equation is satisfied, as shown in equation (2):

[0043]

[0044] In the formula, U 工频干扰 Power frequency electromagnetic noise is the power frequency noise data generated by power transmission lines, and its unit is V; E 工频干扰 t is the maximum electromotive force of the transmission line, in V; f is the frequency of the transmission line, which is 50Hz; t is the receiver observation time, in s. The phase difference is expressed in radians.

[0045] Step 3: Fit the noise dataset U obtained in Step 2 using the least squares method. 噪声干扰 Iteratively fit the values ​​of the model parameters, including the maximum value of the transmission line electromotive force E. 工频干扰 and phase difference

[0046] In the power frequency interference harmonic equation (i.e., formula (2)), the power frequency electromagnetic noise U generated by the transmission line 工频干扰 The maximum electromotive force E of the transmission line 工频干扰 and phase difference Since the unknown variable is U, the least squares method can be used to fit the full waveform electromagnetic interference noise dataset U. 噪声干扰 Solve the problem.

[0047] Specifically, the objective function is shown in equation (3):

[0048]

[0049] In the formula, n is the number of sampling time channels of the receiver; For the i-th sampling time channel t i The corresponding electromagnetic interference noise (obtained from step two); U i To use model parameters The theoretical value of the power frequency electromagnetic noise on the i-th time channel is calculated. The objective function Φ(P) is expanded using Taylor, and higher-order terms are neglected. After transformation, the fitting iterative formula (4) is obtained:

[0050] (J T J+μI)ΔP k =J T Δd k (4)

[0051] In the formula, J is the Jacobian matrix, and its elements are: k is the number of fitting iterations, typically taken as k = 1, 2, ..., m, where m is the maximum number of fitting iterations; Δd k =U 噪声干扰 -U k For U 噪声干扰 The response U of the k-th fitted model k The residual (i.e., power frequency electromagnetic noise); ΔP k The value of the correction to the fitted model in the kth iteration is obtained from formula (4) in each iteration; I is the unit vector; μ is the regularization factor, and its expression is:

[0052]

[0053] In the formula, μ k Let P be the regularization factor corresponding to the k-th iteration. k The model parameters in the k-th iteration include the maximum value of the transmission line electromotive force E. 工频干扰 and phase difference

[0054] By performing fitting iterations according to equations (4) and (5), the model parameters of the transmission line after the m-th fitting iteration are obtained.

[0055] Step 4: Calculate the power frequency electromagnetic noise based on the model parameters obtained in Step 3, and use the power frequency electromagnetic noise to correct the noise dataset obtained in Step 2.

[0056] Specifically, the model parameters calculated in step three... Substituting into equation (2), the power frequency electromagnetic noise U generated by the transmission line is obtained. 工频干扰 ,like Figure 5 As shown.

[0057] Step 5: Using formula (6), obtain the transient electromagnetic exploration data U after correcting for power frequency interference. 校 ,like Figure 6 As shown.

[0058] U 校 =U 电磁干扰 -U 工频干扰 (6).

[0059] contrast Figure 4 and Figure 6 As can be seen, power frequency interference noise has been effectively corrected, data quality has been significantly improved, and the exploration accuracy of transient electromagnetic interference has been guaranteed.

Claims

1. A method for correcting power frequency interference in pre-stack data from transient electromagnetic exploration, characterized in that, The method comprises the following steps: Step 1: Observe and obtain pre-stack full waveform data without electromagnetic interference. Electromagnetic interference zone pre-stack full waveform data ; Step 2: Based on the electromagnetic interference-free pre-stack full waveform data obtained in Step 1 Electromagnetic interference zone pre-stack full waveform data Obtain the full waveform electromagnetic interference noise dataset; the specific steps are as follows: arbitrarily select a point of pre-stack full waveform data without electromagnetic interference. As a background field, the pre-stack full waveform data of the electromagnetic interference region obtained in step one is used. With pre-stack full waveform data without electromagnetic interference By subtracting the values, we obtain the dataset, which consists of the full-waveform electromagnetic interference noise. Noise dataset , Indicates sampling time channel The serial number; Step 3: Fit the noise dataset obtained in Step 2 using the least squares method. Iteratively fit the values ​​of the model parameters, including the maximum value of the transmission line electromotive force. and phase difference The fitting iterative formula is as follows: In the formula, Let T be the Jacobian matrix, and T be the transpose of the matrix. To fit the number of iterations, , This represents the maximum number of fitting iterations. for With the The subfit model response U k The residual; For the first The correction amount of the fitted model is obtained from the fitting iteration formula in each iteration; It is a unit vector; The regularization factor is expressed as follows: In the formula, For the regularization factor corresponding to the k-th iteration, These are the model parameters in the k-th iteration, including the maximum value of the transmission line electromotive force. and phase difference ; In the formula, For noisy datasets; For random noise, the unit is ; Residual induction signal in the formation, unit: ; Power frequency electromagnetic noise, unit: ; This represents the maximum electromotive force of the transmission line, in units of... ; For the frequency of the transmission line, ; The receiver observation time is expressed in units of... ; The phase difference is expressed in radians. Finally, the fitting iterations are used to obtain the first... Model parameters of the transmission line after the second fitting iteration ; Step four: calculating the power frequency electromagnetic noise according to the model parameters obtained in step three; Step 5: Correct the noise dataset obtained in Step 2 using power frequency electromagnetic noise to obtain transient electromagnetic exploration data corrected for power frequency interference. : 。 2. The transient electromagnetic exploration pre-stack data power frequency interference correction method as described in claim 1, characterized in that, Step four involves taking the model parameters calculated in step three. Substituting into the following formula, the power frequency electromagnetic noise generated by the transmission line can be obtained. ; In the formula, Power frequency electromagnetic noise is the power frequency noise data generated by power transmission lines, and the unit is 1000 ppm. ; This represents the maximum electromotive force of the transmission line, in units of... ; The frequency of the transmission line is... ; The receiver observation time is expressed in units of... ; The phase difference is expressed in radians.

Images