A method and system for evaluating the quality of semi-airborne transient electromagnetic data

Abstract

The application discloses a kind of semi-airborne transient electromagnetic data quality evaluation method and system, comprising: semi-airborne transient electromagnetic data acquisition is carried out to the region to be detected, and cruise trajectory error judgment and pretreatment are carried out;Semi-airborne transient electromagnetic data is carried out empirical mode decomposition, and low-frequency residual component is obtained;Semi-airborne transient electromagnetic data is subtracted low-frequency residual component, and first data are obtained;Noise removal is carried out to first data using singular value decomposition method, and equal logarithm time window is extracted, and multi-channel electromagnetic response data are obtained;According to the coil movement posture data, the multi-channel electromagnetic response data are corrected, and second data are obtained;The region to be detected is repeatedly flown, and third data are obtained;Correlation calculation is carried out to second data and third data, and correlation threshold is preset, and the data corresponding to less than correlation threshold is eliminated;Fourth data and fifth data are evaluated using mean square relative error, and the final transient electromagnetic data are obtained.

Description

Technical Field

[0001] This invention relates to the field of geophysical exploration technology, and in particular to a method and system for assessing the quality of semi-airborne transient electromagnetic data. Background Technology

[0002] Semi-airborne transient electromagnetic method (SEM) is an emerging geophysical exploration technology that combines the advantages of both ground-based and airborne SEM methods. This method typically uses an ungrounded loop source or a grounded long conductor source placed on the ground as the transmitter, and a hollow coil mounted on a UAV as the receiver. The method utilizes the ground transmitter to supply power to the underground, generating a primary magnetic field. The underground medium, influenced by the primary magnetic field, generates a secondary induced magnetic field. The receiver on the UAV then receives the induced electromotive force generated by the change in the secondary induced magnetic field, thus completing the data acquisition. In areas with complex terrain, SEM is widely used in mineral exploration, engineering geological exploration, and geological hazard investigation due to its high efficiency, flexibility, and low cost.

[0003] The ultimate goal of geophysical exploration technology is to obtain the structural characteristics of subsurface geological bodies using data collected in the field through inversion interpretation methods. Therefore, reliable data is the foundation of geophysical exploration work. Data acquired through the semi-airborne transient electromagnetic method consists of induced electromotive forces across multiple time windows. However, because data is collected during the flight of the UAV, the useful signals in semi-airborne transient electromagnetic data are relatively weak and susceptible to noise and other interference. Therefore, evaluating the quality of semi-airborne transient electromagnetic data acquired in the field and whether it meets the accuracy requirements for the final inversion interpretation is crucial.

[0004] Currently, existing technologies address issues such as atmospheric noise, coil motion attitude noise, and noise generated by speed changes and drone flight vibrations, primarily employing the following methods:

[0005] For example, the Chinese invention patent with publication number CN113341469A and title "A Method and System for Correcting Semi-Airborne Transient Electromagnetic Data" includes: performing wavelet transform decomposition on semi-airborne transient electromagnetic data, reconstructing the data based on the obtained decomposition parameters, and determining the start and end points of the effective data segments; denoising the semi-airborne transient electromagnetic data excluding the effective data segments for atmospheric noise; performing polynomial fitting on the single-cycle semi-airborne transient electromagnetic data excluding the effective data segments, resampling all data points according to the obtained fitting equation to denoise motion noise; screening semi-airborne transient electromagnetic data that meet quality standards based on the average energy ratio of the effective data segments and the remaining data segments after denoising; and performing attitude correction on the screened semi-airborne transient electromagnetic data based on the triaxial attitude angle of the receiving coil and the deflection of the reference coordinate system. Its disadvantages are: First, this method uses the fitting equation to resample all data points to eliminate motion noise; different fitting equations can lead to significant differences in the final fitting results, resulting in poor stability of the processing results and a significant impact on the quality of later data. Second, the correction method does not evaluate the quality of the corrected data, resulting in low final data reliability.

[0006] Furthermore, in Yang Yang's paper "A Method for Removing Motion Noise from Semi-Airborne Transient Electromagnetic Pure Secondary Field Data Based on Least Squares Inversion," the non-full-time electromagnetic data is first extended to full-time. An overdetermined linear equation system for motion noise in the later stages of semi-airborne transient electromagnetic data is constructed based on Fourier series. The motion noise is then solved using least squares inversion, and the obtained motion noise is removed from the original data. This technique denoises both simulation and measured data. Its drawbacks are: First, this method only processes motion noise, resulting in a relatively limited scope. Second, it only utilizes later-stage data to construct the overdetermined equation system, neglecting earlier-stage data. Third, later-stage data is significantly affected by other noise components; when the signal-to-noise ratio of the later-stage data is too low, the calculated motion noise will be inaccurate.

[0007] For example, in Wang Yongxin's paper "Research on the Expansion of Semi-Airborne Transient Electromagnetic Datasets and Denoising of Fully Convolutional Networks Based on Generative Adversarial Networks," this technique uses real-world noise and forward modeling data to create a dataset for supervised training of the denoising network. This technique proposes a sample expansion method for real-world noise data using generative adversarial networks. Based on the sample distribution of the real-world noise data, it uses two network structures—a generator G and an evaluator C—to identify and simulate the sample distribution of the real-world noise data in an adversarial learning manner. By sampling extensively from this distribution, the dataset expansion is achieved. Its drawbacks are: First, the real-world noise in this method only represents a portion of random noise and does not involve non-random noise, resulting in an incomplete dataset. Second, the forward modeling data does not consider the impact of changes in the coordinates of measuring points on the forward modeling response under the same geological model conditions, leading to incomplete forward modeling data.

[0008] For example, in Wang Shixing's paper "Research on Semi-Airborne Transient Electromagnetic Signal Processing Based on Ensemble Empirical Mode Decomposition," he adds random Gaussian white noise to simulated semi-airborne transient electromagnetic signals. Utilizing the uniform distribution of the Gaussian white noise spectrum, and its characteristic of being distributed throughout the entire time-frequency space, the noise cancels out after multiple averagings, and the result of the ensemble average is taken as the final result. Its drawbacks are: First, this method is only applicable to processing secondary field data of semi-airborne transient electromagnetic signals. However, secondary field data is a logarithmically decaying curve, making it difficult to satisfy the basic condition in empirical mode decomposition that "the number of extrema and the number of zero-crossings differ by no more than 1." Second, the order of empirical mode decomposition is not accurately specified. Third, although the power spectral density of Gaussian white noise is uniformly distributed, its phase exhibits significant differences, making it difficult to achieve noise cancellation through multiple averagings.

[0009] Therefore, there is an urgent need for a simple and reliable method and system for assessing the quality of semi-airborne transient electromagnetic data. Summary of the Invention

[0010] To address the above problems, the present invention aims to provide a method and system for assessing the quality of semi-airborne transient electromagnetic data. The technical solution adopted by the present invention is as follows:

[0011] The first part of this technology provides a method for assessing the quality of semi-airborne transient electromagnetic data, which includes the following steps:

[0012] Step S01: Plan the patrol detection path for the area to be detected;

[0013] Step S02: Acquire semi-airborne transient electromagnetic data of the area to be detected according to the cruise detection path, and perform cruise trajectory error judgment and preprocessing; obtain coil motion attitude data during the semi-airborne transient electromagnetic data acquisition process.

[0014] Step S03: Perform empirical mode decomposition on the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing to obtain low-frequency residual components; subtract the low-frequency residual components from the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing to obtain the first data.

[0015] Step S04: The first data is noise-removed by singular value decomposition, and the first data after noise removal is sampled by an equal logarithmic time window to obtain multi-channel electromagnetic response data.

[0016] Step S05: Correct the multi-channel electromagnetic response data based on the coil motion attitude data to obtain the second data;

[0017] Step S06: Repeat the flight to the area to be detected according to the cruise detection path, repeating steps S02 to S05 to obtain the third data.

[0018] Step S07: Calculate the correlation between the second data and the third data, preset the correlation threshold, and remove the data that is less than the correlation threshold to obtain the fourth data corresponding to the second data and the fifth data corresponding to the third data.

[0019] Step S08: The fourth and fifth data are evaluated using mean square relative error, and the final transient electromagnetic data are obtained.

[0020] The second part of this technology provides a system for assessing the quality of semi-airborne transient electromagnetic data, which includes:

[0021] The path planning module plans the patrol detection path for the area to be detected;

[0022] The cruise detection and acquisition module is connected to the path planning module. It performs semi-airborne transient electromagnetic data acquisition on the area to be detected according to the cruise detection path, and performs cruise trajectory error judgment and preprocessing; it acquires coil motion attitude data during the semi-airborne transient electromagnetic data acquisition process; and it repeatedly flies on the area to be detected according to the cruise detection path planning and acquires semi-airborne transient electromagnetic data from the repeated flights.

[0023] The low-frequency residual component processing module is connected to the cruise detection and acquisition module. It performs empirical mode decomposition on the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing to obtain the low-frequency residual component. The first data is obtained by subtracting the low-frequency residual component from the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing.

[0024] The noise removal module is connected to the low-frequency residual component processing module. It uses the singular value decomposition method to remove noise from the first data. The first data after noise removal is sampled by an equal logarithmic time window to obtain multi-channel electromagnetic response data.

[0025] The correction module, connected to the noise removal module, corrects the multi-channel electromagnetic response data based on the coil motion attitude data to obtain the second data and the third data corresponding to the repeated flight.

[0026] The correlation calculation module, connected to the correction module, performs correlation calculation on the second and third data, presets a correlation threshold, removes data that is less than the correlation threshold, and obtains the fourth data corresponding to the second data and the fifth data corresponding to the third data respectively.

[0027] In addition, the evaluation module, connected to the correlation calculation module, evaluates the fourth and fifth data using mean square relative error and obtains the final transient electromagnetic data.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] (1) Based on the characteristic that the semi-airborne transient electromagnetic induction electromotive force changes with altitude and track, and with the preset cruise track error, this invention provides a standard for judging whether the track altitude and track deviation in actual measurement are qualified, so as to solve the problem of judging the track quality of the survey line flight and avoid the problem of inaccurate data acquisition.

[0030] (2) This invention ingeniously performs empirical mode decomposition on the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing to obtain low-frequency residual components; by subtracting the low-frequency residual components from the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing, the first data is obtained. This invention provides a judgment criterion for the residuals obtained from empirical mode decomposition, performs power spectrum analysis on each residual, and finds the residual with the first maximum energy frequency less than 20Hz, so as to solve the problem of difficulty in judging suitable residuals and ensure the correctness of baseline drift correction.

[0031] (3) This invention employs singular value decomposition to remove noise from the first data, obtaining multi-channel electromagnetic response data; and corrects the multi-channel electromagnetic response data based on the coil motion attitude data to obtain the second data. It simplifies the displacement coefficient matrix in the Z-component coil motion state correction formula, and combines it with the attitude acquisition electronic equipment system to collect coil motion angles, thus solving the problem of complex attitude correction processes. This invention simplifies attitude correction while improving processing efficiency and ensuring data accuracy.

[0032] (4) This invention obtains the data corresponding to the go-around by taking the flight again, and then uses the Pearson correlation coefficient to calculate the correlation coefficient of the induced electromotive force curve of each time window track, and judges the degree of noise influence on the data of each time window track. The judgment is based on the correlation coefficient being greater than or less than 0.8, and a clear judgment standard is given to avoid the problem of difficult semi-aerial data quality assessment, so as to solve the problem of output data meeting the accuracy required for inversion interpretation.

[0033] (5) The present invention evaluates the fourth and fifth data by mean square relative error and removes the data of the measurement points with relative error greater than 10%, which effectively ensures the accuracy and reliability of the data.

[0034] In summary, this invention has the advantages of simple logic and reliable extraction, and has high practical and promotional value in the field of geophysical exploration technology. Attached Figure Description

[0035] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope of protection. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a schematic diagram of the process of the present invention.

[0037] Figure 2 A comparison chart (altitude) of the designed flight path and the actual flight path for this invention.

[0038] Figure 3 A comparison chart of the designed flight path and the actual flight path (track deviation) for this invention.

[0039] Figure 4 This is the empirical mode decomposition diagram of the present invention.

[0040] Figure 5 This is the baseline drift correction diagram of the present invention.

[0041] Figure 6 This is a singular value decomposition denoising diagram from the present invention.

[0042] Figure 7 This is a schematic diagram of the coil motion state correction of the present invention.

[0043] Figure 8 This is a comparison chart of the correlation between data curves for different time windows in this invention.

[0044] Figure 9 This is a graph showing the relative mean square errors of each part of the invention. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this application clearer, the present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following embodiments. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0046] In this embodiment, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0047] The terms "first" and "second," etc., used in the specification and claims of this embodiment are used to distinguish different objects, not to describe a specific order of objects. For example, "first target object" and "second target object," etc., are used to distinguish different target objects, not to describe a specific order of target objects.

[0048] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0049] In the description of the embodiments in this application, unless otherwise stated, "multiple" means two or more. For example, multiple processing units means two or more processing units; multiple systems means two or more systems.

[0050] like Figures 1 to 9 As shown, this embodiment provides a method for assessing the quality of semi-airborne transient electromagnetic data, which mainly includes the following steps:

[0051] The first step is to plan a patrol and detection route for the area to be detected.

[0052] The second step involves acquiring semi-airborne transient electromagnetic data of the area to be tested according to the cruise detection path, and performing cruise trajectory error judgment and preprocessing; obtaining coil motion attitude data during the semi-airborne transient electromagnetic data acquisition process. In this step, the quality of the semi-airborne transient electromagnetic survey line flight track is evaluated. Based on the difference between the designed track and the actual track, the maximum error in flight altitude deviation and the maximum error in flight track deviation are calculated to determine whether the actual track meets the specified errors. Specifically: the maximum permissible error in altitude of the cruise trajectory error is set to 5% of the flight altitude of the planned cruise detection path; the maximum permissible error in track deviation of the cruise trajectory error is set to 10m.

[0053] The third step involves performing empirical mode decomposition on the semi-airborne transient electromagnetic data after trajectory error assessment and preprocessing to obtain low-frequency residual components. The first data is obtained by subtracting these low-frequency residual components from the preprocessed semi-airborne transient electromagnetic data. Specifically:

[0054] The raw semi-airborne transient electromagnetic data is continuously subjected to empirical mode decomposition until the frequency of the maximum energy distribution of the decomposition residual is less than the threshold of 20Hz. Decomposition is then stopped, yielding the low-frequency residual component. The original data is subtracted from the low-frequency residual component to complete baseline drift correction. The calculation formula is as follows:

[0055]

[0056]

[0057]

[0058]

[0059] MaxFre(PSD(r n ))≤20

[0060] Where x(t) represents the original semi-airborne transient electromagnetic data, n represents the total number of decompositions, and imf i (t) represents the decomposition modulus obtained from the i-th decomposition, r n (t) represents the residual, e max (t) denotes the maximum envelope, e min (t) represents the local minimum envelope. The new signal r1(t) with high-frequency components removed, ε is the judgment condition, maxFre is the frequency of the maximum energy distribution, and PSD is the power spectrum transform.

[0061] The fourth step involves using singular value decomposition (SVD) to remove noise from the first data. The denoised first data is then subjected to equal logarithmic time window sampling to obtain multi-channel electromagnetic response data. Noise removal includes: decomposing the first data, which contains atmospheric noise and white noise, using SVD to obtain several singular values; and reconstructing the first data using these singular values ​​to complete the noise removal process.

[0062] Specifically, here, one-and-a-half-dimensional airborne transient electromagnetic data is reconstructed into R... m×n For a two-dimensional matrix, perform singular value decomposition on the matrix, and then reconstruct the data using the multiple principal singular values ​​obtained from the decomposition. The calculation formula is as follows:

[0063]

[0064] Where, σ i Let u represent the i-th singular value. i Let v' represent the vector in the i-th row. i Let i represent the i-th column vector.

[0065] The fifth step involves correcting the multi-channel electromagnetic response data based on the coil motion attitude data to obtain the second set of data, which includes:

[0066] Define the coil's forward direction as the x-direction, the direction perpendicular to the x-axis in the horizontal plane as the y-direction, and the direction pointing towards the ground as the z-direction. The coil's motion state changes mainly involve three angles: spin angle α, roll angle β, and pitch angle γ. Based on these three motion angles recorded by the coil in the attitude acquisition system, geometric correction is used to correct the coil's motion state. The calculation formula is as follows:

[0067]

[0068] V=R α R β R γ V′

[0069] Where α is the coil spin angle, β is the coil roll angle, γ is the coil pitch angle, and R is the coil pitch angle. α R is the spin displacement coefficient matrix of the coil. β R is the spin displacement coefficient matrix of the coil. γ This is the spin displacement coefficient matrix of the coil.

[0070] For the Z-component coil, the spin angle change has no effect on the semi-airborne transient electromagnetic data. In this case, the correction formula can be simplified as follows:

[0071] V = V′cos(β)cos(γ)

[0072] Where V′ represents the signal before calibration, i.e., the multi-channel electromagnetic response data; V represents the signal after calibration, i.e., the second data; β represents the coil roll angle; and γ represents the coil pitch angle.

[0073] Step 6: For the same design survey line, it is assumed that all actual survey line data completed in the above steps can reflect the structural characteristics of the underground geological body at the design survey line location. Therefore, theoretically, the induced electromotive force (EMF) data of these actual survey lines are equal. However, due to complex noise pollution during actual measurement, the induced EMF data will be distorted. To intuitively represent the difference in induced EMF between different survey lines in each time window, the correlation coefficient of the induced EMF curves of each time window is calculated using the Pearson correlation coefficient. It is considered that the data of time window tracks with a correlation coefficient lower than 0.8 are too affected by noise and cannot be used for inversion interpretation. If the actual survey line has N measurement points, the curve correlation ρ of the k-th time window track of the two flight measurements is calculated according to the following formula. k The calculation formula is as follows:

[0074]

[0075]

[0076]

[0077] Where k represents the number of survey tracks; A ik μ represents the induced electromotive force at the i-th measuring point within the k-th time window of the second data; Ak B represents the mean of the k-th time window in the second data set; ik μ represents the induced electromotive force at the i-th measuring point within the k-th time window of the third data; Bk σ represents the mean of the third data point in the k-th time window; Ak σ represents the standard deviation of the k-th time window in the second data; Bk The standard deviation of the third data point in the k-th time window is represented by σ; μ represents the mean; x ik The induced electromotive force (EMF) at the i-th measuring point within the k-th time window of a certain measured data is A. ik Or B ik Corresponding data; ρ k (A,B) represents the curve correlation between the second and third data points in the k-th time window.

[0078] After calculating the correlation of the curves for all time windows, plot the Pearson correlation coefficient for each time window, as shown in the attached figure. Figure 5 As shown. All time windows with a correlation coefficient higher than 0.8 are statistically analyzed, along with the center time for each window. These parameters are then output, yielding the fourth data point corresponding to the second data point and the fifth data point corresponding to the third data point.

[0079] Step 7: The mean square relative error is used to evaluate the fourth and fifth data points, and the final transient electromagnetic data is obtained. In this embodiment, the effective measurement data obtained in step 6 is used to evaluate the data quality of each measurement point using the mean square relative error calculation formula. If the relative error of a measurement point is less than 10%, the measurement point is considered qualified and is saved. Conversely, if the relative error of a measurement point is greater than 10%, the measurement point is considered unqualified and is discarded. The mean square relative error calculation formula for the i-th measurement point in two flight measurements is as follows:

[0080]

[0081] Where Rms represents the mean square relative error, and M represents the total number of valid traces involved in the calculation.

[0082] Specific examples are as follows:

[0083] The first step is to assess the flight track quality of the semi-airborne transient electromagnetic survey line. If the maximum error between the actual track altitude and the designed track altitude is less than 5% of the designed flight altitude, and the maximum error between the actual track deviation and the designed track deviation is less than 10m, then the flight track quality of the survey line is qualified; otherwise, the flight track quality is unqualified and needs to be remeasured.

[0084] As attached Figures 2 to 3 As shown, the maximum error between the actual flight altitude and the design altitude is about 2m, which is less than the judgment error. The maximum error in the flight track deviation distance is about 2m, which is less than the judgment error of 10m, thus meeting the requirements for subsequent work.

[0085] The second step is to continuously perform adaptive empirical mode decomposition on the original semi-airborne transient electromagnetic data until the frequency of the maximum energy distribution of the decomposition residual is less than the threshold of 20Hz. Then, the decomposition is stopped, and the low-frequency residual component is obtained. The original data is subtracted from the low-frequency residual component to complete the baseline drift correction.

[0086] As attached Figure 4 As shown, the original semi-airborne transient electromagnetic data underwent a total of 10 empirical mode decompositions. (a) to (j) represent the 10th decomposition moduli, (k) represents the residual, whose maximum energy distribution frequency is less than 20 Hz, and (l) is a graph of the original semi-airborne transient electromagnetic data. (See attached diagram.) Figure 5 As shown, (c) is the data after baseline drift correction.

[0087] The third step is to reconstruct the one-and-a-half-dimensional airborne transient electromagnetic data into R... m×n A two-dimensional matrix is ​​subjected to singular value decomposition (SVD). SVD is used to decompose the signal, including atmospheric noise and white noise, and the data is reconstructed using the multiple principal singular values ​​after decomposition, thus completing conventional noise removal. (See attached diagram.) Figure 6 As shown.

[0088] The fourth step is to correct the coil motion state based on the coil motion angle in the attitude acquisition system and the rotation relationship between the motion angle and the Cartesian coordinate system.

[0089] As attached Figure 6 As shown, (a) is a schematic diagram of a three-component receiving device, (b) is a schematic diagram of the spin motion angle, (c) is a schematic diagram of the roll motion angle, and (d) is a schematic diagram of the pitch angle.

[0090] Fifth, for the same design survey line, it is assumed that all actual survey line data completed in the above steps can reflect the structural characteristics of the underground geological body at the design survey line location. Therefore, theoretically, the induced electromotive force (EMF) data of these actual survey lines are all equal. However, due to complex noise pollution during actual measurement, the induced EMF data will be distorted. To intuitively represent the differences in induced EMF of each time window track for different survey lines, this paper selects the induced EMF data from two flight measurements and uses the Pearson correlation coefficient to evaluate the data quality of the induced EMF curves of each time window track, judging the degree of noise influence on the data of each time window track. If the correlation coefficient is greater than 0.8, it is considered that the track is less affected by noise, the data quality is better, and it can be used for inversion interpretation, and should be regarded as a useful track and saved; if the correlation coefficient is greater than 0.8, it is considered that the track is greatly affected by noise, the data quality is poor, it cannot be used for inversion interpretation, and should be regarded as a noisy track and discarded.

[0091] As attached Figure 8 As shown, after calculating the curve correlation for all time windows, Pearson correlation coefficient plots for each time window are generated. All time windows with correlation coefficients higher than 0.8 and their corresponding center times are statistically analyzed, and these parameters are output. The curve correlation coefficients differ for different time windows. Time windows 1-23 have correlation coefficients greater than 0.8 and are considered valid data; time windows 24-30 have correlation coefficients less than 0.8 and are considered invalid data, discarded, and not used for subsequent inversion interpretation.

[0092] Step 6: The survey line data obtained from the two flight measurements are ultimately represented by the induced electromotive force at each measuring point. However, external factors may cause inconsistencies in the number of measuring points between the two measurements. In this case, interpolation needs to be performed on one of the survey lines to ensure consistency in the measuring point data between the two measurements. Then, using only the valid survey line data obtained in step 6, the data quality of each measuring point is evaluated using the mean square relative error calculation formula. If the relative error of a measuring point is less than 10%, the measuring point is considered qualified and is saved. Conversely, if the relative error of a measuring point is greater than 10%, the measuring point is considered unqualified and is discarded.

[0093] Using only the valid measurement data from channels 1 to 24 obtained in step 6, the relative error of a certain measurement point is evaluated. For example... Figure 9 As shown, the mean square relative error varies at different measurement points. The mean square relative error of most measurement points in the figure is less than 10%. However, the mean square relative error of the measurement points at mileage -200m and mileage 100m is significantly greater than 10%. The data quality of the measurement points at these locations is unqualified and needs to be removed.

[0094] The above embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any changes made based on the design principles of the present invention, or any non-creative modifications made thereon, shall fall within the scope of protection of the present invention.

Claims

1. A method for assessing the quality of semi-airborne transient electromagnetic data, characterized in that, Includes the following steps: Step S01: Plan the patrol detection path for the area to be detected; Step S02: Acquire semi-airborne transient electromagnetic data of the area to be detected according to the cruise detection path, and perform cruise trajectory error judgment and preprocessing; obtain coil motion attitude data during the semi-airborne transient electromagnetic data acquisition process. Step S03: Perform empirical mode decomposition on the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing to obtain low-frequency residual components; subtract the low-frequency residual components from the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing to obtain the first data. Step S04: The first data is noise-removed by singular value decomposition, and the first data after noise removal is sampled by an equal logarithmic time window to obtain multi-channel electromagnetic response data. Step S05: Correct the multi-channel electromagnetic response data based on the coil motion attitude data to obtain the second data; Step S06: Repeat the flight to the area to be detected according to the cruise detection path, repeating steps S02 to S05 to obtain the third data. Step S07: Calculate the correlation between the second data and the third data, preset the correlation threshold, and remove the data that is less than the correlation threshold to obtain the fourth data corresponding to the second data and the fifth data corresponding to the third data. Step S08: The fourth and fifth data are evaluated using mean square relative error, and the final transient electromagnetic data are obtained.

2. The method for assessing the quality of semi-airborne transient electromagnetic data according to claim 1, characterized in that, According to the cruise detection path, semi-airborne transient electromagnetic data is acquired for the area to be detected, and cruise trajectory error judgment and preprocessing are performed, including: The actual cruise trajectory is obtained and compared with the planned cruise detection path to obtain the flight altitude deviation error value and the flight trajectory deviation error value respectively. Preset flight altitude deviation error threshold and flight track deviation error threshold; Semi-airborne transient electromagnetic data corresponding to flight altitude deviation error values ​​greater than the flight altitude deviation error threshold and / or flight track deviation error values ​​greater than the flight track deviation error threshold are excluded.

3. A method for assessing the quality of semi-airborne transient electromagnetic data according to claim 1 or 2, characterized in that, The noise removal of the first data using singular value decomposition includes: The singular value decomposition method was used to decompose the first data containing atmospheric noise and white noise, resulting in several singular values. The first data is reconstructed using the several singular values ​​after decomposition in order to remove noise from the first data.

4. A method for assessing the quality of semi-airborne transient electromagnetic data according to claim 1 or 2, characterized in that, The step of correcting the multi-channel electromagnetic response data based on the coil motion attitude data to obtain the second data includes the following steps: Acquire coil motion attitude data; the coil motion attitude data includes coil motion angles; the coil motion angles include coil roll motion angles. β and coil pitch angle γ; The formula for correcting the motion state of the coil is derived from the relationship between the motion angle and the rotation of the Cartesian coordinate system, and is expressed as follows: ； in, This represents the signal before calibration, i.e., the multi-channel electromagnetic response data; This represents the corrected signal, i.e., the second data. Indicates the angle of the coil's lateral movement; This indicates the pitch angle of the coil.

5. A method for assessing the quality of semi-airborne transient electromagnetic data according to claim 1 or 2, characterized in that, In step S03, empirical mode decomposition is performed on the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing. The decomposition is stopped when the frequency of the maximum energy distribution of the decomposition residual is less than the threshold of 20Hz, and low-frequency residual components are obtained.

6. A method for assessing the quality of semi-airborne transient electromagnetic data according to claim 1 or 2, characterized in that, In step S07, the expression for the correlation is: ； ； ； in, Indicates the number of survey tracks; Indicates the second data. The first time window Induced electromotive force at the measuring point; Indicates the second data. The mean of the time window; Indicates the third data The first time window Induced electromotive force at the measuring point; Indicates the third data The mean of the time window; Indicates the second data. The standard deviation of the time window; Indicates the third data The standard deviation of the time window; Indicates standard deviation; This represents the mean; For a certain measured data point The first time window Induced electromotive force at the measuring point; Indicates the first The correlation between the curves corresponding to the second and third data points within the time window; N represents the number of measurement points on the measurement line.

7. The method for assessing the quality of semi-airborne transient electromagnetic data according to claim 6, characterized in that, In step S08, the mean square relative error is used to evaluate the measurement points corresponding to the fourth and fifth data points. If the relative error of the measurement point is greater than 10%, it is discarded. Otherwise, the data is retained, and the final transient electromagnetic data is obtained; the mean square relative error The expression is: ； Where M represents the total number of valid channels involved in the calculation, that is, the total number of channels for the fourth or fifth data.

8. The method for assessing the quality of semi-airborne transient electromagnetic data according to claim 1, characterized in that, In step S02, the maximum permissible error in altitude of the cruise trajectory error is 5% of the flight altitude of the planned cruise detection path; the maximum permissible error in track deviation of the cruise trajectory error is 10m.

9. A system employing the semi-airborne transient electromagnetic data quality assessment method according to any one of claims 1 to 8, characterized in that, include: The path planning module plans the patrol detection path for the area to be detected; The cruise detection and acquisition module is connected to the path planning module. It performs semi-airborne transient electromagnetic data acquisition on the area to be detected according to the cruise detection path, and performs cruise trajectory error judgment and preprocessing; it acquires coil motion attitude data during the semi-airborne transient electromagnetic data acquisition process; and it repeatedly flies on the area to be detected according to the cruise detection path planning and acquires semi-airborne transient electromagnetic data from the repeated flights. The low-frequency residual component processing module is connected to the cruise detection and acquisition module. It performs empirical mode decomposition on the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing to obtain low-frequency residual components. The first data is obtained by subtracting the low-frequency residual component from the semi-airborne transient electromagnetic data after trajectory error judgment and preprocessing; The noise removal module is connected to the low-frequency residual component processing module. It uses the singular value decomposition method to remove noise from the first data. The first data after noise removal is sampled by an equal logarithmic time window to obtain multi-channel electromagnetic response data. The correction module, connected to the noise removal module, corrects the multi-channel electromagnetic response data based on the coil motion attitude data to obtain the second data and the third data corresponding to the repeated flight. The correlation calculation module, connected to the correction module, performs correlation calculation on the second and third data, presets a correlation threshold, removes data that is less than the correlation threshold, and obtains the fourth data corresponding to the second data and the fifth data corresponding to the third data respectively. In addition, the evaluation module, connected to the correlation calculation module, evaluates the fourth and fifth data using mean square relative error and obtains the final transient electromagnetic data.

Images