Multi-objective Magnetotelluric Data Screening Method, Device, Equipment and Medium

Abstract

The present invention provides a multi-objective based magnetotelluric data screening method, device, equipment and medium, which relates to the technical field of magnetotelluric data processing, so as to alleviate problems existing in the prior art such as low efficiency, strong subjectivity, inability to optimize multiple electromagnetic objectives simultaneously, and poor screening effect when the noise ratio is high in the magnetotelluric power spectrum data screening method. The original power spectrum data is obtained by segmenting and windowing the magnetotelluric data and performing fast Fourier transform; the Mahalanobis distance is used to preliminarily screen the original power spectrum data to obtain the original transfer function estimation value and multiple objective functions; based on the non-dominated sorting genetic algorithm II and multiple objective functions, the original power spectrum data is screened again to obtain the target power spectrum data, so as to automatically and efficiently screen high-quality power spectrum data, thereby improving the transfer function estimation accuracy and the reliability of data analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of magnetotelluric data processing, and in particular, to a magnetotelluric data screening method, device, equipment, and medium based on multiple objectives. Background Technique

[0002] The Magnetotelluric (MT) method is an important geophysical exploration means for studying the electrical structure of the Earth's interior with the natural electromagnetic field as the field source. In the traditional magnetotelluric impedance tensor estimation method, before estimating the transfer function, the quality of the transfer function estimation can be improved by removing the power spectrum data of the noisy period, that is, power spectrum screening.

[0003] Common methods for screening magnetotelluric power spectrum data include manual screening methods and automatic screening methods based on statistical characteristics. Among them, the manual screening method has the disadvantages of strong subjectivity, inconsistent results, low efficiency, easy omission of important information, and inability to quantify and optimize multi-objective requirements. The data screening method based on Mahalanobis distance has technical problems such as being easily affected by data with a high noise ratio, single-objective requirements, need to meet statistical assumptions, and insufficient robustness. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a magnetotelluric data screening method, device, equipment, and medium based on multiple objectives to alleviate problems such as low efficiency, strong subjectivity, inability to optimize multiple electromagnetic objectives simultaneously, and poor screening effect when the noise ratio is high in the existing magnetotelluric power spectrum data screening method.

[0005] In the first aspect, the present invention provides a magnetotelluric data screening method based on multiple objectives, including:

[0006] Obtain the original power spectrum data;

[0007] Perform preliminary screening on the original power spectrum data using Mahalanobis distance to obtain the transfer function estimation value corresponding to the original power spectrum data;

[0008] According to the transfer function estimation value, determine multiple objective functions corresponding to the original power spectrum data; where the objective functions include the deviation between the transfer function and the expected value, the deviation between the apparent resistivity and the expected value, the dispersion degree of electromagnetic field parameters, the correlation degree of the electromagnetic field, and the deviation of the magnetotelluric impedance.

[0009] Optionally, screening the original power spectrum data based on the non-dominated sorting genetic algorithm II and multiple objective functions to obtain the target power spectrum data, including:

[0010] Randomly generate an initial population; where each individual in the population is a subset of the original power spectrum data;

[0011] Perform iterative screening operations based on the population until it is determined that the first iterative optimization termination condition is met. Based on the power spectrum data output during the last execution of the iterative screening operation, the target power spectrum data is obtained. Among them, the iterative screening operations include:

[0012] Use the fitness function to calculate the fitness value of each individual in the population.

[0013] Perform non-dominated sorting on the individuals in the population based on the values of multiple objective functions to obtain multiple levels, and calculate the crowding distance of each individual in each level based on the Mahalanobis distance.

[0014] Based on multiple levels and crowding distances, select parents, perform crossover, and mutate to update the population.

[0015] Optionally, screening the original power spectrum data based on the non-dominated sorting genetic algorithm II and multiple objective functions to obtain the target power spectrum data further includes:

[0016] Obtain multiple excellent individuals in the Pareto front solution set of the new population;

[0017] Based on the introduction mechanism, introduce multiple excellent individuals into the current population.

[0018] Optionally, obtaining multiple excellent individuals in the Pareto front solution set of the new population further includes:

[0019] If the number of multiple excellent individuals exceeds the threshold, screen the multiple excellent individuals according to non-dominated sorting and crowding distance.

[0020] Optionally, screening the original power spectrum data based on the non-dominated sorting genetic algorithm II and multiple objective functions to obtain the target power spectrum data further includes:

[0021] Obtain the frequency data set corresponding to the original power spectrum data; among them, the frequency data set includes multiple frequencies;

[0022] Based on the frequency data set, perform iterative single-frequency point screening operations on the original power spectrum until it is determined that the second iterative termination condition is met. Based on the power spectrum data output during the last execution of the iterative single-frequency point screening operation, the target power spectrum data is determined; among them, the iterative single-frequency point screening operations include:

[0023] Select the current frequency from the frequency data set;

[0024] Based on the current frequency and the original power spectrum data, use the non-dominated sorting genetic algorithm II to screen the original power spectrum data to obtain the target power spectrum data.

[0025] Optionally, the first iteration termination condition is: reaching a preset number of iterations or the change range of the Pareto front solution set of the population is lower than a set threshold;

[0026] The second iteration termination condition is: reaching a preset number of iterations.

[0027] Optionally, obtaining the original power spectrum data further includes:

[0028] Obtaining the original magnetotelluric data;

[0029] Segmenting, windowing, performing fast Fourier transform and combining the original magnetotelluric data to obtain the original power spectrum data corresponding to the original magnetotelluric data at each frequency.

[0030] In a second aspect, the present invention provides a magnetotelluric data screening device based on multiple objectives, including:

[0031] A data acquisition unit for acquiring the original power spectrum data;

[0032] A preliminary screening unit for preliminarily screening the original power spectrum data using the Mahalanobis distance to obtain the estimated value of the transfer function corresponding to the original power spectrum data;

[0033] A target function determination unit for determining multiple target functions corresponding to the original power spectrum data according to the estimated value of the transfer function; wherein, the target functions include the deviation between the transfer function and the expected value, the deviation between the apparent resistivity and the expected value, the dispersion degree of electromagnetic field parameters, the correlation degree of the electromagnetic field, and the deviation degree of the magnetotelluric impedance;

[0034] A data screening unit for screening the original power spectrum data based on the non-dominated sorting genetic algorithm II and multiple target functions to obtain the target power spectrum data.

[0035] In a third aspect, the present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the magnetotelluric data screening method based on multiple objectives.

[0036] In a fourth aspect, the present invention further provides a computer-readable storage medium. The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the magnetotelluric data screening method based on multiple objectives is implemented.

[0037] A method, device, equipment and medium for screening magnetotelluric data based on multiple objectives provided by an embodiment of the present invention. The method includes obtaining original power spectrum data; preliminarily screening the original power spectrum data using Mahalanobis distance to obtain an estimated value of the transfer function; determining multiple objective functions based on the estimated value of the transfer function; and screening the original power spectrum data based on the non-dominated sorting genetic algorithm II and the multiple objective functions to obtain target power spectrum data. Compared with the magnetotelluric power spectrum data screening method in the prior art, through multi-objective optimization technology, high-quality power spectrum data can be automatically and efficiently screened, thereby improving the estimation accuracy of the transfer function and the reliability of data analysis.

[0038] To make the above objects, features and advantages of the present invention more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the accompanying drawings, provides a detailed description as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0040] Figure 1 Shows a schematic flowchart of a method for screening magnetotelluric data based on multiple objectives provided by an embodiment of the present invention;

[0041] Figure 2 Shows a flowchart of screening the original power spectrum data based on the non-dominated sorting genetic algorithm II (NSGA-II) provided by an embodiment of the present invention;

[0042] Figure 3 Shows a schematic flowchart of the operation of the non-dominated sorting genetic algorithm II (NSGA-II) provided by an embodiment of the present invention;

[0043] Figure 4 Shows the apparent resistivity and phase curves of the processing results of a certain measured data by the data screening method based on Mahalanobis distance provided by an embodiment of the present invention;

[0044] Figure 5 Shows the apparent resistivity and phase curves of the processing results of a certain measured data by the method for screening magnetotelluric data based on multiple objectives provided by an embodiment of the present invention;

[0045] Figure 6 Shows a schematic structural diagram of a device for screening magnetotelluric data based on multiple objectives provided by an embodiment of the present invention;

[0046] Figure 7The structural schematic diagram of an electronic device provided by an embodiment of the present invention is shown. Detailed implementation manners

[0047] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are only a part rather than all of the embodiments of the present invention. Components of the embodiments of the present invention described and illustrated herein generally may be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts fall within the protection scope of the present invention.

[0048] Figure 1 The flowchart of a multi-objective magnetotelluric data screening method provided by an embodiment of the present invention is shown. As Figure 1 shown, the method at least includes the following steps:

[0049] Step S110: Obtain original power spectrum data.

[0050] In the embodiments of the present application, the original power spectrum data is obtained through the following method:

[0051] Obtain original magnetotelluric data;

[0052] Perform segmentation, windowing, fast Fourier transform, and combination on the original magnetotelluric data to obtain the original power spectrum data corresponding to the original magnetotelluric data at each frequency.

[0053] Specifically, obtain the original magnetotelluric data, and divide the original magnetotelluric data into A segments according to time periods to facilitate subsequent analysis of the magnetotelluric data; for each segmented time series data (i.e., the original magnetotelluric data), apply the fast Fourier transform (FFT) to convert the original magnetotelluric data into frequency domain data, thereby decomposing the original magnetotelluric data into different frequency components; calculate the original auto / cross power spectrum data corresponding to the original magnetotelluric data at each frequency.

[0054] In the embodiments of the present application, the FFT conversion is respectively performed on the original magnetotelluric data of 5 channels (Ex, Ey, Hx, Hy, Hz), so 5 groups of original spectrum data can be obtained. 25 groups of different auto / cross power spectra can be calculated to form a 5×5 power spectrum matrix, and the calculation can be performed using the following formula:

[0055]

[0056] In the formula, is the FFT result of the th channel, is the complex conjugate of the FFT result of the th channel. If i ≠ j, is the cross-power spectrum. If i = j, then is the auto-power spectrum.

[0057] Furthermore, after obtaining A power spectrum matrices, divide them into B groups and superimpose them to obtain B superimposed power spectrum matrices as the original power spectrum data.

[0058] Step S120: Use the Mahalanobis distance to preliminarily screen the original power spectrum data to obtain the estimated value of the transfer function corresponding to the original power spectrum data.

[0059] To improve the robustness of the screening, the original power spectrum data is preliminarily screened using the Mahalanobis distance data with the minimum determinant to obtain the estimated value of the transfer function. The method can be as follows:

[0060] Based on the B superimposed power spectrum matrices, calculate the real and imaginary parts of the B groups of transfer functions, calculate their Mahalanobis distance covariance matrix and Mahalanobis distance, retain some power spectra with smaller Mahalanobis distances, update the covariance matrix, and repeat this step until the determinant of the covariance matrix no longer decreases. Superimpose the finally retained power spectrum matrices and calculate to obtain the estimated value of the transfer function.

[0061] Step S130: According to the estimated value of the transfer function, determine multiple objective functions corresponding to the original power spectrum data; among them, the objective functions include the deviation between the transfer function and the expected value, the deviation between the apparent resistivity and the expected value, the dispersion degree of electromagnetic field parameters, the correlation degree of the electromagnetic field, and the deviation degree of the magnetotelluric impedance.

[0062] In the embodiments of the present application, based on the original power spectrum data, electromagnetic field parameters are calculated, and the optimization objectives for participating in subsequent data screening are determined. The optimization objectives include the real and imaginary parts of the transfer function, the auto-power spectral density of the electromagnetic field, and the phase tensor.

[0063] Furthermore, the real and imaginary parts of the transfer function: , , , , , , , , , , and ;

[0064] The auto-power spectral density of the electromagnetic field: , , , ;

[0065] The phase tensor is a derived parameter of the impedance tensor. The impedance tensor is divided into a real part and an imaginary part, and the impedance tensor is: , where is the real part of the impedance tensor, is the imaginary part of the impedance tensor. Therefore, the phase tensor can be defined as: , where is four elements in

[0066] In the embodiments of the present application, the setting of the objective function can be increased or adjusted according to actual application requirements. In the present application, the objective function includes the deviation between the transfer function and the expected value, the deviation between the apparent resistivity and the expected value, the dispersion degree of electromagnetic field parameters, the coherence degree of the electromagnetic field, and the deviation degree of magnetotelluric impedance. Among them, the deviation between the transfer function and the expected value is used to measure the difference between the estimated value of the impedance tensor and the expected value, and is used to evaluate the reliability of the power spectrum data;

[0067] The expected value of the transfer function is predicted through the continuity of a single transfer function component and the correlation between each transfer function component. The prediction method can be an artificial intelligence method or a curve fitting method;

[0068] The expected values of the apparent resistivity and the phase can be calculated through the expected value of the transfer function, or can be obtained using the one-dimensional inversion method Rhoplus;

[0069] The dispersion degree of electromagnetic field parameters is used to measure the consistency of the power spectrum data distribution to reflect the coordination between data points. The dispersion degree can be one or more of the range, variance, standard deviation, interquartile range, coefficient of variation, or mean absolute deviation according to different electromagnetic parameters and application conditions. The data median used to calculate the dispersion degree can be replaced by the expected value of the corresponding parameter to make the objective function more robust. Among them, the electromagnetic field parameters include the phase tensor, the transfer function, and the electromagnetic field self-power spectral density;

[0070] The coherence degree is used for the correlation between the corresponding electric field and magnetic field signals of the power spectrum data, and is one of the indication standards for the strength of the signal and noise;

[0071] The magnetotelluric impedance deviation degree includes a one-dimensional deviation degree and a two-dimensional deviation degree, and is one of the indication standards for the strength of the data noise.

[0072] Step S140: Screen the original power spectrum data based on the non-dominated sorting genetic algorithm II and multiple objective functions to obtain the target power spectrum data.

[0073] In the embodiments of the present application, a multi-objective optimization screening method based on the Non-dominated Sorting Genetic Algorithm II (NSGA-II) is used to encode the original magnetotelluric power spectrum data into binary individuals, and the population is iteratively optimized through the genetic algorithm. By comprehensively considering various objective functions such as the deviation between the transfer function and the expected value, the deviation between the apparent resistivity and phase and the expected value, the dispersion degree of electromagnetic field parameters, and the electromagnetic field coherence degree, the intelligent screening of the original power spectrum data is realized, and the target power spectrum data is obtained, thereby improving the estimation accuracy of the magnetotelluric transfer function and the reliability of subsequent geophysical analysis.

[0074] The following details the screening process of the original power spectrum data using the Non-dominated Sorting Genetic Algorithm II (NSGA-II) provided in the embodiments of the present application, as Figure 2 shown, the screening process of the original power spectrum data using the Non-dominated Sorting Genetic Algorithm II (NSGA-II) provided in the embodiments of the present application is as follows:

[0075] Step S210: Obtain the frequency data set corresponding to the original power spectrum data; wherein, the frequency data set includes multiple frequencies.

[0076] In the embodiments of the present application, when obtaining the frequency data set, the following methods can be used but are not limited to:

[0077] First, obtain the original magnetotelluric data and divide the original magnetotelluric data according to time periods;

[0078] Then, for each divided segment of the original magnetotelluric data, apply the Fast Fourier Transform (FFT) to convert the original magnetotelluric data into frequency-domain data, and calculate the original power spectrum data corresponding to the original geophysical data at each frequency;

[0079] Finally, based on the original power spectrum data and the frequency-domain data, obtain the frequency data set.

[0080] Step S220: Select the current frequency from the frequency data set.

[0081] Step S230: Based on the current frequency and the original power spectrum data, use the Non-dominated Sorting Genetic Algorithm II to screen the original power spectrum data to obtain the target power spectrum data.

[0082] Step S240: Determine whether the second iteration termination condition is satisfied; if so, execute step S250; if not, return to step S220; wherein, the second iteration termination condition is that the number of iterations is not less than the predicted number of iterations.

[0083] Step S250: Determine the power spectrum data output when the last single-frequency point screening operation of the iteration is performed as the target power spectrum data.

[0084] In this application, by using the Non-dominated Sorting Genetic Algorithm II (NSGA-II) at different frequencies to screen the original power spectrum data, it is possible to effectively denoise, improve the signal-to-noise ratio, enable the target power spectrum data to better reflect the actual signal characteristics, and have high accuracy at multiple key frequency points.

[0085] The following will detail the operation process of the Non-dominated Sorting Genetic Algorithm II (NSGA-II) provided in the embodiments of this application. As Figure 3 shown, the operation process of the Non-dominated Sorting Genetic Algorithm II (NSGA-II) provided in the embodiments of this application is as follows:

[0086] Step S310: Randomly generate an initial population; where each individual in the population is a subset of the original power spectrum data.

[0087] In the embodiments of this application, whether each original power spectrum data point participates in the calculation is encoded as a binary variable, where 0 indicates that the original power spectrum data point does not participate in the calculation, and 1 indicates that the original power spectrum data point participates in the calculation;

[0088] Randomly generate an initial population, where each individual in the population consists of a set of binary codes, the coding length is equal to the number of original power spectrum data points, and the population size N is a preset parameter that controls the number of individuals participating in the screening.

[0089] Step S320: Use a fitness function to calculate the fitness value of each individual in the population; where the fitness function can be, but is not limited to, the sum of squared prediction errors, etc.

[0090] Step S330: Perform non-dominated sorting on the individuals in the population based on the values of multiple objective functions to obtain multiple levels, and calculate the crowding distance of each individual in each level based on the Mahalanobis distance.

[0091] In the embodiments of this application, non-dominated sorting is performed on the individuals in the population according to the values of multiple objective functions. If an individual is not worse than another individual in all objective functions and is better in at least one objective function, then this individual is said to "dominate" the other individual; all individuals that are not dominated by other individuals are divided into the first Pareto front; after removing the first layer from the population, repeat the above process to determine the second, third, and fourth layers; and mark the non-dominated sorting level (Rank) for each individual. The lower the level value, the better the individual.

[0092] Further, for each individual in each layer, according to each objective function value, the crowding distance is calculated based on the Mahalanobis distance, and this crowding distance is used to measure the density of solutions around the individual; for each objective function, the individuals are sorted in ascending order of the crowding distance value; and the distances between adjacent individuals within the range of the objective function values are calculated. Among them, the crowding distance is the comprehensive distance on all objective functions, and the larger the crowding distance value, the more dispersed the solutions around the individual.

[0093] Step S340: Update the population by selecting parents, performing crossover, and mutating based on multiple levels and the crowding distance.

[0094] In the embodiments of the present application, based on non-dominated sorting and the crowding distance, N individuals are selected from the merged population of the current population and the previous generation population. Among them, individuals with lower level values are preferentially selected; in the same level, individuals with larger crowding distances are preferentially selected to maintain the diversity of the population; using the simulated binary crossover (SBX) method, two individuals are randomly selected from the selected individuals, and their partial codes are exchanged to generate two new individuals; among them, the crossover probability is a preset parameter, usually set to 0.8 - 0.9; each newly generated individual is flipped by a preset mutation probability (0 becomes 1, 1 becomes 0). Among them, the mutation probability is usually set to 3 / the coding length to increase the diversity of the population.

[0095] Step S350: Obtain multiple excellent individuals in the Pareto front solution set in the new population; based on the introduction mechanism, introduce the multiple excellent individuals into the current population.

[0096] In the embodiments of the present application, after each generation of iteration, a part of the individuals located on the Pareto front in the new generation population (i.e., excellent individuals) are randomly selected and added to the external archive; based on the introduction mechanism, individuals in the external archive are introduced into the current population with a certain probability to enhance the diversity of the population and prevent the solution from falling into local optimality. Among them, if the number of individuals in the external archive exceeds the set threshold, the individuals are screened according to non-dominated sorting and the crowding distance, and the inferior solutions are deleted.

[0097] Step S360: Determine whether the first iteration termination condition is satisfied; if so, execute step S370; if not, return to step S320; where the first iteration termination condition is that the number of iterations is not less than the predicted number of iterations or the change range of the Pareto front solution set of the population is lower than the set threshold, whichever is satisfied.

[0098] Step S370: Based on the Pareto front of the last generation population, calculate the Mahalanobis distance determinant of each individual, and take the individual with the smallest determinant as the target power spectrum data.

[0099] In the embodiments of the present application, through multi-objective optimization screening based on the Non-dominated Sorting Genetic Algorithm II (NSGA-II), multiple excellent solution sets are screened from the original power spectrum data to form a Pareto front, enhancing the diversity and optimization of the screening results; by introducing excellent solution sets and supplementing the external archive, convergence to local optima is avoided, and the optimal solution can be found within a limited number of iterations.

[0100] The following is a comparison of the processing results of measured data and the data processing results obtained by using a multi-objective magnetotelluric data screening method provided by the present invention: The measured data is from a magnetotelluric measurement point in a certain place. After segmenting the original time series, 1000 power spectrum data are calculated. The power spectrum screening is performed using the Mahalanobis distance method and the multi-objective magnetotelluric data screening method of the present invention respectively. Among them, the deviation between the apparent resistivity phase curves at multiple angles of the objective function selected in the method of the present invention and the expected values is used. The apparent resistivity phase curves are rotated to 6 different angles (0°, 15°, 30°, 45°, 60° and 75°), and the one-dimensional inversion method Rhoplus is used for inversion fitting to obtain the corresponding fitting apparent resistivity and phase as the expected values. During the operation of the NSGA-II algorithm provided by the present application, the objective fitness of each individual is the apparent resistivity and phase difference at 6 angles and 2 polarization directions (TE mode and TM mode) (a total of 24 objectives). After performing NSGA-II data screening on each frequency point in turn, the final processing results are obtained.

[0101] Figure 4 Apparent resistivity of the processing result of a certain measured data by the data screening method based on the Mahalanobis distance and phase curve; Figure 5 Apparent resistivity of the processing result of a certain measured data by the multi-objective magnetotelluric data screening method of the present invention and phase curve. It can be seen from this that the quality of the processing result curve of the method of the present invention is better than that of the data screening method based on the Mahalanobis distance.

[0102] A multi-objective magnetotelluric data screening method provided by the present invention has the following beneficial effects:

[0103] 1. By using the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to automatically screen magnetotelluric power spectrum data, replacing the traditional manual screening and single-objective screening methods; through multi-objective optimization and intelligent iteration, the efficiency of data processing is significantly improved, which is especially suitable for large-scale and multi-frequency magnetotelluric data sets.

[0104] 2. Comprehensively optimize multiple electromagnetic targets to enhance the objectivity and consistency of screening: By introducing multi-objective optimization, the screening process is based on scientifically defined objective functions such as transfer function deviation, apparent resistivity phase deviation, and electromagnetic field parameter dispersion, etc., to achieve an all-round balanced optimization of complex data requirements. It is completely driven by algorithms, avoiding the subjectivity and inconsistency of manual judgment, and ensuring the reliability and repeatability of screening results.

[0105] 3. In the case of high noise ratio or poor signal quality, through the population iteration of the genetic algorithm, external archive maintenance, and Pareto front solution set update, gradually eliminate noise points and retain high-quality data, significantly improving the robustness and anti-interference ability of the screening process.

[0106] 4. The objective function can be flexibly defined according to actual data and application requirements, such as the deviation between the transfer function and the expected value, the stability of electromagnetic field parameters, the physical rationality of apparent resistivity and phase, etc. The optimization objectives can be adjusted for different scenarios to ensure the adaptability and generality of the algorithm.

[0107] 5. Use the method of the minimum determinant of the Mahalanobis distance covariance matrix to select solutions from the Pareto front, avoiding the subjectivity of manually selecting the optimal solution from the Pareto front. The determinant of the Mahalanobis distance covariance matrix characterizes the consistency of data in each parameter dimension. Selecting the solution with the minimum determinant ensures the quality of the solution.

[0108] 6. The magnetotelluric power spectrum data screening is a process of gradually improving the quality of transfer function estimation and gradually approaching the true solution. Set reasonable initial iteration values through the Mahalanobis distance method, and then gradually optimize the expected values of the transfer function and electromagnetic field parameters through multiple rounds of iteration, and further optimize the objective function, so that the processing results of each round are improved compared with the previous ones.

[0109] Figure 6 The present invention provides a structural schematic diagram of a magnetotelluric data screening device based on multiple objectives, as Figure 6 shown. The device includes:

[0110] A data acquisition unit 410 for acquiring original power spectrum data;

[0111] A preliminary screening unit 420 for preliminarily screening the original power spectrum data using the Mahalanobis distance to obtain the estimated value of the transfer function corresponding to the original power spectrum data;

[0112] An objective function determination unit 430 for determining multiple objective functions corresponding to the original power spectrum data according to the estimated value of the transfer function; wherein, the objective functions include the deviation between the transfer function and the expected value, the deviation between the apparent resistivity and the expected value, the dispersion of electromagnetic field parameters, the correlation of the electromagnetic field, and the magnetotelluric impedance deviation.

[0113] A data screening unit 440 is configured to screen the original power spectrum data based on the non-dominated sorting genetic algorithm II and multiple objective functions to obtain the target power spectrum data.

[0114] In an alternative embodiment, the data screening unit 440 is further configured to:

[0115] Randomly generate an initial population; where each individual in the population is a subset of the original power spectrum data;

[0116] Perform iterative screening operations on the population until it is determined that the first iterative optimization termination condition is met, and based on the power spectrum data combination output during the last execution of the iterative screening operation as the target power spectrum data; wherein, the iterative screening operations include:

[0117] Adopt a fitness function to calculate the fitness value of each individual in the population;

[0118] Perform non-dominated sorting on the individuals in the population based on the values of multiple objective functions to obtain multiple levels, and calculate the crowding distance of each individual in each level based on the Mahalanobis distance;

[0119] Based on multiple levels and crowding distances, select parents, perform crossover and mutation to update the population.

[0120] In an alternative embodiment, the data screening unit 440 is further configured to:

[0121] Obtain multiple excellent individuals in the Pareto front solution set of the new population;

[0122] Based on an introduction mechanism, introduce multiple excellent individuals into the current population.

[0123] Optionally, obtaining multiple excellent individuals in the Pareto front solution set of the new population further includes:

[0124] If the number of multiple excellent individuals exceeds a threshold, screen the multiple excellent individuals according to non-dominated sorting and crowding distance.

[0125] In an alternative embodiment, the data screening unit 440 is further configured to:

[0126] Obtain a frequency data set corresponding to the original power spectrum data; wherein, the frequency data set includes multiple frequencies;

[0127] Based on the frequency data set, perform iterative single-frequency point screening operations on the original power spectrum until it is determined that the second iterative termination condition is met, and determine the power spectrum data output during the last execution of the iterative single-frequency point screening operation as the target power spectrum data; wherein, the iterative single-frequency point screening operations include:

[0128] Select the current frequency from the frequency data set;

[0129] Based on the current frequency and the original power spectrum data, use the non-dominated sorting genetic algorithm II to screen the original power spectrum data to obtain the target power spectrum data.

[0130] In an alternative embodiment, the first iteration termination condition is: reaching a preset number of iterations or the change range of the Pareto front solution set of the population is lower than a set threshold;

[0131] The second iteration termination condition is: reaching a preset number of iterations.

[0132] In an alternative embodiment, the data acquisition unit 410 is further configured to:

[0133] Obtain the original power spectrum data, further including:

[0134] Obtain the original magnetotelluric data;

[0135] Segment, window, perform fast Fourier transform and combination on the original magnetotelluric data to obtain the original power spectrum data corresponding to the original magnetotelluric data at each frequency.

[0136] The device provided by the embodiments of the present application has the same implementation principle and the same technical effects as the foregoing method embodiments. For the sake of brief description, for the parts not mentioned in the device embodiments, reference may be made to the corresponding contents in the foregoing method embodiments.

[0137] As Figure 7 As shown, an electronic device 600 provided by the embodiments of the present application includes: a processor 601, a memory 602, and a bus. The memory 602 stores machine-readable instructions executable by the processor 601. When the electronic device runs, the processor 601 communicates with the memory 602 through the bus, and the processor 601 executes the machine-readable instructions to perform the steps of the above-mentioned multi-objective magnetotelluric data screening method.

[0138] Specifically, the above-mentioned memory 602 and processor 601 can be general-purpose memory and processor, which are not specifically limited here. When the processor 601 runs the computer program stored in the memory 602, it can execute the above-mentioned multi-objective magnetotelluric data screening method.

[0139] The processor 601 may be an integrated circuit chip with the ability to process signals. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in the processor 601 or the instructions in the form of software. The above-mentioned processor 601 may be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it may also be a digital signal processor (DSP for short), an application specific integrated circuit (ASIC for short), a field-programmable gate array (FPGA for short) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as being executed and completed by the hardware decoding processor, or executed and completed by a combination of the hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory 602, and the processor 601 reads the information in the memory 602 and combines its hardware to complete the steps of the above method.

[0140] Corresponding to the above multi-objective magnetotelluric data screening method, an embodiment of the present application also provides a computer-readable storage medium. The computer-readable storage medium stores machine-executable instructions. When the machine-executable instructions are called and run by a processor, the machine-executable instructions cause the processor to run the steps of the above multi-objective magnetotelluric data screening method.

[0141] The multi-objective magnetotelluric data screening device provided by the embodiments of the present application may be specific hardware on the device or software or firmware installed on the device, etc. For the device provided by the embodiments of the present application, the implementation principle and the technical effects produced are the same as those of the foregoing method embodiments. For the sake of brief description, for the parts not mentioned in the device embodiments, reference may be made to the corresponding content in the foregoing method embodiments. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can all refer to the corresponding processes in the above method embodiments, and will not be repeated here.

[0142] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods. For another example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some communication interfaces. The indirect couplings or communication connections of devices or units can be in electrical, mechanical or other forms.

[0143] For another example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions, and operations of devices, methods, and computer program products according to multiple embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code. A module, a program segment, or a part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks can actually be executed substantially in parallel, and they can sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as the combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0144] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0145] In addition, the various functional units in the embodiments provided in the present application can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0146] If a function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing an electronic device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of various embodiments of this application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs.

[0147] It should be noted that: similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. In addition, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0148] Finally, it should be noted that: the above embodiments are only specific implementation manners of this application, used to illustrate the technical solutions of this application, rather than limiting it. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed in this application can still modify the technical solutions recorded in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes, or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A multi-objective magnetotelluric data screening method, characterized in that: include: Get the original power spectrum data; The original power spectrum data is preliminarily screened using Mahalanobis distance to obtain a transfer function estimation value corresponding to the original power spectrum data; Determine multiple objective functions corresponding to the original power spectrum data according to the estimated value of the transmission function; wherein the objective functions include the deviation of the transmission function from the expected value, the deviation of the apparent resistivity from the expected value, the dispersion of electromagnetic field parameters, the correlation of the electromagnetic field and the deviation of the magnetotelluric impedance; The original power spectrum data is screened based on the non-dominated sorting genetic algorithm II and multiple objective functions to obtain target power spectrum data; wherein, the original power spectrum data is screened based on the non-dominated sorting genetic algorithm II and multiple objective functions to obtain target power spectrum data, including: randomly generating an initialized population; wherein each individual in the population is a subset of the original power spectrum data; performing an iterative screening operation based on the population until it is determined that the first iterative optimization termination condition is met, and the power spectrum data output when the iterative screening operation is last performed is combined into the target power spectrum data; wherein, the iterative screening operation includes: using a fitness function to calculate the fitness value of each individual in the population; performing a non-dominated sorting of the individuals in the population based on the values ​​of multiple objective functions to obtain multiple levels, and calculating the crowding distance of each individual in each level based on the Mahalanobis distance; based on multiple levels and the crowding distance, selecting a parent, crossover, or similar method is used. and mutation to update the population; obtain multiple excellent individuals in the new population located in the Pareto front solution set; based on the introduction mechanism, introduce multiple excellent individuals into the current population; wherein, based on the non-dominated sorting genetic algorithm II and multiple objective functions, the original power spectrum data is screened to obtain the target power spectrum data, and also includes: obtaining the frequency data set corresponding to the original power spectrum data; wherein the frequency data set includes multiple frequencies; based on the frequency data set, performing an iterative single frequency point screening operation on the original power spectrum until it is determined that the second iteration termination condition is met, and the power spectrum data output when the iterative single frequency point screening operation is performed for the last time is determined as the target power spectrum data; wherein the iterative single frequency point screening operation includes: selecting the current frequency from the frequency data set; based on the current frequency and the original power spectrum data, using the non-dominated sorting genetic algorithm II to screen the original power spectrum data to obtain the target power spectrum data.

2. The multi-objective magnetotelluric data screening method according to claim 1 is characterized in that: Obtain multiple excellent individuals in the Pareto front solution set in the new population, including: If the number of multiple excellent individuals exceeds a threshold, the multiple excellent individuals are screened according to non-dominated sorting and crowding distance.

3. The multi-objective magnetotelluric data screening method according to claim 1, characterized in that: include: The first iteration termination condition is: reaching a preset number of iterations or the change range of the Pareto front solution set of the population is lower than a set threshold; The second iteration termination condition is: reaching a preset number of iterations.

4. The multi-objective magnetotelluric data screening method according to claim 1, characterized in that: Before obtaining the original power spectrum data, it also includes: Obtain raw magnetotelluric data; The original magnetotelluric data are segmented, windowed, fast Fourier transformed and combined to obtain original power spectrum data corresponding to the original magnetotelluric data at each frequency.

5. A multi-objective magnetotelluric data screening device, characterized in that: include: A data acquisition unit, used for acquiring original power spectrum data; A preliminary screening unit, used for performing preliminary screening on the original power spectrum data by using Mahalanobis distance to obtain a transfer function estimation value corresponding to the original power spectrum data; A target function determination unit is used to determine a plurality of target functions corresponding to the original power spectrum data according to the estimated value of the transmission function; wherein the target functions include the deviation of the transmission function from the expected value, the deviation of the apparent resistivity from the expected value, the dispersion of the electromagnetic field parameters, the correlation of the electromagnetic field and the deviation of the magnetotelluric impedance; A data screening unit is used to screen the original power spectrum data based on a non-dominated sorting genetic algorithm II and a plurality of the objective functions to obtain target power spectrum data; wherein, the original power spectrum data is screened based on a non-dominated sorting genetic algorithm II and a plurality of the objective functions to obtain target power spectrum data, including: randomly generating an initialized population; wherein each individual in the population is a subset of the original power spectrum data; performing an iterative screening operation based on the population until it is determined that the first iterative optimization termination condition is met, and combining the power spectrum data output when the iterative screening operation is performed for the last time as the target power spectrum data; wherein the iterative screening operation includes: using a fitness function to calculate the fitness value of each individual in the population; performing a non-dominated sorting of the individuals in the population based on the values ​​of the plurality of the objective functions to obtain a plurality of levels, and calculating the crowding distance of each individual in each level based on the Mahalanobis distance; based on the plurality of the levels and the crowding distance, using a selection The population is updated by parent generation, crossover and mutation; a plurality of excellent individuals in the new population located in the Pareto front solution set are obtained; based on the introduction mechanism, a plurality of excellent individuals are introduced into the current population; wherein, the original power spectrum data is screened based on the non-dominated sorting genetic algorithm II and a plurality of the objective functions to obtain the target power spectrum data, and further comprises: obtaining a frequency data set corresponding to the original power spectrum data; wherein, the frequency data set comprises a plurality of frequencies; based on the frequency data set, an iterative single frequency point screening operation is performed on the original power spectrum until it is determined that the second iteration termination condition is met, and the power spectrum data outputted when the iterative single frequency point screening operation is performed for the last time is determined as the target power spectrum data; wherein, the iterative single frequency point screening operation comprises: selecting the current frequency from the frequency data set; based on the current frequency and the original power spectrum data, the original power spectrum data is screened using the non-dominated sorting genetic algorithm II to obtain the target power spectrum data.

6. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the multi-objective magnetotelluric data screening method according to any one of claims 1 to 4 is implemented.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the multi-objective-based magnetotelluric data screening method according to any one of claims 1 to 4 is implemented.

Images