An electromagnetic interference feature learning separation method and device and medium

Abstract

This invention relates to the field of geophysical exploration technology, and particularly to an electromagnetic interference feature learning and separation method, apparatus, and medium. The method includes: collecting first electromagnetic data at each type of human electromagnetic interference source; extracting time-domain waveform features of known types of human electromagnetic interference sources based on the first electromagnetic data; constructing a basic electromagnetic interference feature library based on the time-domain waveform features; collecting second electromagnetic data at a distant reference point in the exploration area; and extracting clean electromagnetic signals; obtaining a signal-to-noise separation training dataset based on the basic electromagnetic interference feature library and the clean electromagnetic signals; training a signal-to-noise separation model based on the training dataset; collecting third electromagnetic data at a target acquisition area in the exploration area; and separating usable electromagnetic data based on the signal-to-noise separation model and the third electromagnetic data, thereby effectively improving the detection capability of frequency domain electromagnetic exploration in complex interference environments, especially in deep and peripheral resource exploration.

Description

Technical Field

[0001] This invention relates to the field of geophysical exploration technology, and in particular to a method, apparatus and medium for learning and separating electromagnetic interference characteristics. Background Technology

[0002] With the depletion of shallow mineral resources, detailed exploration of deep and concealed mineral deposits has become an important strategic direction for ensuring mineral resource security and promoting the high-quality development of green mining. Frequency domain electromagnetic exploration methods, such as magnetotelluric (MT), audio magnetotelluric (AMT), controlled-source audio magnetotelluric (CSAMT), and wide-field electromagnetic (WFEM), play an irreplaceable role in deep mineral exploration due to their advantages of large detection depth, wide frequency response, and flexible construction.

[0003] Existing research indicates that "mineral exploration in situ" is more successful and efficient in established mining areas with favorable mineralization conditions and clear geological backgrounds. However, in densely populated mining areas and surrounding regions, the electromagnetic interference generated by high-voltage power lines, transportation facilities, communication base stations, and mining equipment is often several orders of magnitude stronger than the effective signals from deep sources. This leads to a sharp deterioration in the signal-to-noise ratio of observational data, severely obscuring weak deep response signals and becoming a key bottleneck restricting the depth and resolution of exploration.

[0004] Currently, technologies for dealing with human-induced electromagnetic interference are mainly divided into frequency domain methods and time domain methods. Frequency domain methods (such as the far-reference method, Robust estimation, and inversion correction method) typically suppress noise in the frequency domain, but they suffer from difficulties in selecting far-reference points and unstable impedance estimation under strong interference conditions. Time domain methods (such as sparse representation, empirical mode decomposition, and neural network methods) operate directly on the original time series, preserving more effective information. Among these, artificial intelligence technologies such as deep learning have shown significant potential. For example, some studies have used deep neural networks (DNNs) to build interference knowledge bases to improve communication quality, or used deep temporal convolutional networks and dictionary learning to suppress strong cultural noise in controllable source electromagnetic data, maintaining high recognition accuracy even under low signal-to-noise ratio conditions.

[0005] However, these technologies still have obvious limitations: Staticization of characteristics: Most methods focus on identifying or removing noise from single observation data, and the interference signal discrimination criteria they rely on are often static or empirically dependent, lacking the ability to evolve. However, human electromagnetic interference in the real environment is multi-source, time-varying, and non-stationary, making it difficult for static discrimination criteria to adapt to its dynamic changes.

[0006] Insufficient generalization ability: Many deep learning-based methods rely on large amounts of labeled data for supervised training. However, acquiring a large number of high-quality labeled signals in real, complex electromagnetic environments is extremely difficult. Although some studies have attempted to improve adaptability and generalization ability in dynamic, complex environments and under small sample conditions by using unsupervised learning to eliminate the dependence on labeled data, how to effectively combine cutting-edge advancements with the specific interference suppression requirements in geophysical exploration to avoid overfitting of the sample library remains an unexplored problem.

[0007] Therefore, how to accurately separate and extract weak and effective signals from a strong interference background is a technical problem that urgently needs to be solved. Summary of the Invention

[0008] In view of the above problems, the present invention provides an electromagnetic interference feature learning and separation method, apparatus and medium to overcome the above problems or at least partially solve the above problems.

[0009] In a first aspect, the present invention provides an electromagnetic interference feature learning and separation method, comprising: First electromagnetic data were collected at each known type of human electromagnetic interference source; Based on the first electromagnetic data, extract the time-domain waveform features of the known types of human electromagnetic interference sources; Based on the aforementioned time-domain waveform characteristics, a basic electromagnetic interference feature library is constructed; In the exploration area, second electromagnetic data are collected by selecting distant reference points; Based on the second electromagnetic data, extract the pure electromagnetic signal; Based on the aforementioned basic electromagnetic interference feature library and the aforementioned clean electromagnetic signal, a training dataset for signal-to-noise separation is obtained; Based on the training dataset, a signal-to-noise separation model is trained. In the exploration area, select the target acquisition area to collect third electromagnetic data; Based on the signal-to-noise separation model and the third electromagnetic data, usable electromagnetic data is separated out, which is used to indicate the deep ore body in the target acquisition area.

[0010] Preferably, after selecting the target exploration area and separating usable electromagnetic data based on the signal-to-noise separation model, and after the usable electromagnetic data is used to indicate the deep ore body in the target acquisition area, the method further includes: The quality of the available electromagnetic data is evaluated to determine whether it is acceptable. If not, it is determined that the available electromagnetic data contains new interference data; Based on the new interference data, new interference data features are extracted and added to the basic electromagnetic interference feature library.

[0011] Preferably, the quality of the available electromagnetic data is evaluated to determine whether it is acceptable, including: Based on the available electromagnetic data, the corresponding time series plot is obtained; Convert the available electromagnetic data into frequency domain data; Based on the frequency domain data, the corresponding apparent resistivity diagram, phase curve diagram, polarization direction diagram and Nyquist diagram are plotted. Determine whether the time series plot, apparent resistivity plot, phase curve plot, polarization direction plot and Nyquist plot satisfy their respective preset forms, and obtain the determination result; Based on the judgment results, an assessment is made as to whether the qualification is qualified.

[0012] Preferably, first electromagnetic data is collected at each known type of human electromagnetic interference source, including: For each known type of human electromagnetic interference source, the minimum safe acquisition distance is determined based on the signal-to-noise ratio model and terrain attenuation factor; Based on the minimum safe acquisition distance, survey lines are laid out according to a preset range to acquire the first electromagnetic data.

[0013] Preferably, at each known type of anthropogenic electromagnetic interference source, the minimum safe acquisition distance is determined based on the signal-to-noise ratio model and terrain attenuation factor, including: The minimum safe acquisition distance is obtained using the following formula:

[0014] in, To minimize the safe sampling distance, The minimum signal-to-noise ratio threshold. For terrain attenuation factor, The attenuation factor of the source signal is denoted as . The type factor of the field source signal. The attenuation factor of the interference source signal. This is the type factor of the interference source signal.

[0015] Preferably, based on the first electromagnetic data, the time-domain waveform features of the known types of anthropogenic electromagnetic interference sources are extracted, including: Based on the first electromagnetic data, the wavelet coefficients are determined using a soft threshold function, specifically according to the following calculation formula:

[0016] in, For symbolic functions, For the first The first wavelet decomposition scale Wavelet coefficients corresponding to each time position The threshold is used to dynamically adjust the energy amplitude of the primary and secondary fields of the noise signal from known types of human electromagnetic interference sources. Based on the wavelet coefficients, the time-domain waveform features are reconstructed using inverse wavelet transform.

[0017] Preferably, based on the basic electromagnetic interference feature library and the clean electromagnetic signal, a training dataset for signal-to-noise separation is obtained, including: Based on the aforementioned basic electromagnetic interference feature library, noise data is obtained; Based on the noise data and the clean electromagnetic signal, a training dataset for signal-to-noise separation is obtained.

[0018] Preferably, the known types of human electromagnetic interference sources include: Mining construction areas, high-power industrial equipment, high-voltage transmission lines, low-voltage distribution lines, highways, railways, communication base stations, and radar stations.

[0019] Secondly, the present invention also provides an electromagnetic interference feature learning and separation device, comprising: The first acquisition module is used to acquire first electromagnetic data at each known type of human electromagnetic interference source; The first extraction module is used to extract the time-domain waveform features of the known type of human electromagnetic interference source based on the first electromagnetic data. The construction module is used to construct a basic electromagnetic interference feature library based on the time-domain waveform features; The second acquisition module is used to acquire second electromagnetic data by selecting a distant reference point in the exploration area; The second extraction module is used to extract pure electromagnetic signals based on the second electromagnetic data; The module is used to obtain a training dataset for signal-to-noise separation based on the basic electromagnetic interference feature library and the clean electromagnetic signal. The training module is used to train a signal-to-noise separation model based on the training dataset. The third acquisition module is used to select a target acquisition area in the exploration area to acquire third electromagnetic data; The separation module is used to separate usable electromagnetic data based on the signal-to-noise separation model and the third electromagnetic data. The usable electromagnetic data is used to indicate the deep ore body in the target acquisition area.

[0020] Thirdly, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method described in the first aspect.

[0021] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in the first aspect.

[0022] One or more technical solutions in the embodiments of the present invention have at least the following technical effects or advantages: This invention provides an electromagnetic interference feature learning and separation method, comprising: collecting first electromagnetic data at each known type of human electromagnetic interference source; extracting time-domain waveform features of the known types of human electromagnetic interference sources based on the first electromagnetic data; constructing a basic electromagnetic interference feature library based on the time-domain waveform features; collecting second electromagnetic data at a distant reference point in the exploration area; extracting clean electromagnetic signals based on the second electromagnetic data; obtaining a signal-to-noise separation training dataset based on the basic electromagnetic interference feature library and the clean electromagnetic signals; training a signal-to-noise separation model based on the training dataset; selecting a target exploration area and collecting third electromagnetic data at a target acquisition area within the exploration area; and separating usable electromagnetic data based on the signal-to-noise separation model and the third electromagnetic data. The usable electromagnetic data is used to indicate deep ore bodies in the target acquisition area, thereby effectively improving the detection capability of frequency domain electromagnetic exploration in complex interference environments, especially in deep and peripheral resource exploration. Attached Figure Description

[0023] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A schematic flowchart of the electromagnetic interference feature learning and separation method in an embodiment of the present invention is shown. Figure 2 This invention illustrates a schematic diagram of the layout of a four-component frequency domain electromagnetic data acquisition survey line in an embodiment of the invention. Figure 3 A schematic diagram showing the planar locations of human disturbance sources around the mine, as marked in an embodiment of the present invention, is provided. Figure 4This is a schematic diagram of the time series of electromagnetic data of interference in the mine construction area collected in an embodiment of the present invention; Figure 5 This diagram illustrates a time series of electromagnetic data including interference from high-power equipment areas in an embodiment of the present invention. Figure 6 This invention illustrates a time series diagram of electromagnetic data for transmission line interference in an embodiment of the present invention. Figure 7 This invention illustrates a time series diagram of electromagnetic data including interference near railways and highways in an embodiment of the present invention; Figure 8 This diagram illustrates a time series of electromagnetic data containing interference near a communication base station or radar station, as shown in an embodiment of the present invention. Figure 9 An anomaly-free time series plot of available electromagnetic data is shown in an embodiment of the present invention; Figure 10 This invention presents a distortion-free and continuous apparent resistivity map of electromagnetic data available in an embodiment of the invention. Figure 11 This invention presents a distortion-free and continuous phase curve of the electromagnetic data available in an embodiment of the invention. Figure 12 The diagram illustrates a random distribution of electromagnetic data available in an embodiment of the present invention; Figure 13 The Nyquist plot of the distribution of available electromagnetic data according to a preset morphological trend is shown in an embodiment of the present invention; Figure 14 A schematic diagram of the electromagnetic interference feature learning and separation device in an embodiment of the present invention is shown. Figure 15 A schematic diagram of the structure of an electronic device implementing the electromagnetic interference feature learning and separation method in an embodiment of the present invention is shown. Detailed Implementation

[0024] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the invention and to fully convey the scope of the invention to those skilled in the art.

[0025] Example 1: Embodiments of the present invention provide an electromagnetic interference feature learning and separation method, such as... Figure 1 As shown, it includes: S101, Collect first electromagnetic data at each known type of human electromagnetic interference source; S102, Based on the first electromagnetic data, extract the time-domain waveform features of known types of human electromagnetic interference sources; S103, Based on time-domain waveform characteristics, a basic electromagnetic interference feature library is constructed; S104, in the exploration area, select a distant reference point to collect second electromagnetic data; S105, based on the second electromagnetic data, extracts the pure electromagnetic signal; S106, based on the basic electromagnetic interference feature library and clean electromagnetic signals, obtains the training dataset for signal-to-noise separation; S107, a signal-to-noise separation model is trained based on the training dataset; S108, In the exploration area, select the target acquisition area to collect third electromagnetic data; S109, based on the signal-to-noise separation model and third electromagnetic data, separates usable electromagnetic data, which is used to indicate the deep ore body in the target acquisition area.

[0026] The overall approach involves first constructing a basic electromagnetic interference feature library. Then, based on this library and second electromagnetic data collected at a distant reference point, a training dataset for signal-to-noise separation is built to train a signal-to-noise separation model. This model is designed to separate usable electromagnetic data from the third electromagnetic data collected in the target acquisition area as early as possible.

[0027] First, execute S101 to collect first electromagnetic data at each known type of human electromagnetic interference source.

[0028] Specifically, for each known type of human electromagnetic interference source, the minimum safe acquisition distance is determined based on the signal-to-noise ratio model and terrain attenuation factor; Based on the minimum safe acquisition distance, the survey lines are laid out according to the preset range to collect the first electromagnetic data.

[0029] like Figure 2 As shown, a geophysical instrument was used to construct a four-component frequency domain electromagnetic data acquisition line for collecting the first electromagnetic data. This four-component frequency domain electromagnetic data acquisition line was deployed in the east, south, west, and north directions.

[0030] Known types of human-mediated electromagnetic interference sources include: Mining construction areas, high-power industrial equipment, high-voltage power transmission lines, low-voltage power distribution cables, highways, railways, communication base stations, and radar stations.

[0031] Specifically, such as Figure 3 The diagram shown is a plan view of the marked locations of human disturbance sources around the mine. These include residential areas, high-power equipment areas, and other surrounding human disturbance sources.

[0032] The following is a detailed description of determining the minimum safe acquisition distance: The minimum safe acquisition distance is obtained using the following formula:

[0033] in, To minimize the safe sampling distance, The minimum signal-to-noise ratio threshold. For terrain attenuation factor, The attenuation factor of the source signal is denoted as . The type factor of the field source signal. The attenuation factor of the interference source signal. This is the type factor of the interference source signal.

[0034] By setting up survey lines within a preset range and adhering to this minimum safe acquisition distance, the first electromagnetic data can be acquired. The preset range for setting up the survey lines is as follows: .

[0035] Taking mine construction areas, high-power equipment areas, power transmission lines, railways and highways, communication base stations, or radar as examples, this paper presents a time series diagram of electromagnetic data collected from mine construction areas, such as... Figure 4 As shown in the diagram. This illustrates a time series of electromagnetic data containing interference from high-power equipment areas, such as... Figure 5 As shown in the diagram. A time series diagram of electromagnetic data for transmission line interference is presented, as follows. Figure 6 As shown in the diagram. This illustrates a time series of electromagnetic data including interference from railways and highways, as shown in the diagram. Figure 7 As shown. This is a schematic diagram illustrating the time series of electromagnetic data containing interference from near communication base stations or radar stations, such as... Figure 8 As shown.

[0036] Next, step S102 is executed to extract the time-domain waveform features of known types of anthropogenic electromagnetic interference sources based on the first electromagnetic data. Specifically: Based on the first electromagnetic data, the wavelet coefficients are determined using a soft threshold function, specifically according to the following calculation formula:

[0037] in, For symbolic functions, For the first The first wavelet decomposition scale Wavelet coefficients corresponding to each time position The threshold is used to dynamically adjust the energy amplitude of the primary and secondary fields of the noise signal from known types of human electromagnetic interference sources. Based on wavelet coefficients, inverse wavelet transform is used to reconstruct the time-domain waveform features.

[0038] Among them, wavelet coefficients contain local feature information of the signal at different time points and frequency components. By weighting and superimposing these wavelet coefficients, the time domain waveform features of the original signal are recombined.

[0039] After obtaining the time-domain waveform features, execute S103 to construct a basic electromagnetic interference feature library based on the time-domain waveform features.

[0040] Therefore, this basic electromagnetic interference feature library contains time-domain waveform features corresponding to various human-caused electromagnetic interference sources.

[0041] Next, step S104 is executed to acquire second electromagnetic data from the depths of the target acquisition area. Since this second electromagnetic data contains low-interference signals, interference signals need to be filtered out. Therefore, step S105 is executed to extract clean electromagnetic signals based on the second electromagnetic data.

[0042] Specifically, in the exploration area, four-component frequency domain electromagnetic data are collected from sources of interference, and the collection time needs to cover the duration of future data collection.

[0043] A clean electromagnetic signal is extracted from the second electromagnetic data; this clean signal is the valid signal. Specifically, data with good signal integrity within a certain period are selected as the clean electromagnetic signal.

[0044] Next, S106 is executed to obtain a training dataset for signal-to-noise separation based on the basic electromagnetic interference feature library and clean electromagnetic signals.

[0045] In this training dataset, the time-domain waveform features corresponding to various human interference sources are combined with the pure electromagnetic signal in multiple ways, and the interference signal and the pure electromagnetic signal are labeled to obtain the signal-to-noise separation training.

[0046] Next, execute S107 to train the signal-to-noise separation model based on the training dataset.

[0047] By constructing a deep neural network model, this training dataset is used to train the deep neural network model.

[0048] Specifically, the training dataset is input into the deep neural network model according to the labeled signal types to train the model to distinguish signal types, i.e., to separate interference signals from useful signals. During training, an L1 norm loss function is used to optimize network parameters, and then the Adam algorithm is used to iteratively find the optimal value along the negative gradient direction based on the objective function.

[0049] Execute S108 to select the target acquisition area in the exploration area and collect third electromagnetic data; The center of the target acquisition area is 50-100 kilometers away from the distant reference point. The electromagnetic environment of the distant reference point has a low degree of electromagnetic coupling with human-caused interference sources. However, the electromagnetic environment of the target acquisition area has a high degree of electromagnetic coupling with human-caused interference sources. Therefore, a third electromagnetic data is acquired in this target acquisition area, which includes both human-caused electromagnetic interference and usable electromagnetic data.

[0050] Next, S109 is executed, based on the signal-to-noise separation model and the third electromagnetic data, to separate the usable electromagnetic data, which is used to indicate the deep ore body in the target acquisition area.

[0051] Specifically, a third electromagnetic signal is acquired in the target acquisition area, and the third electromagnetic signal is input into the signal-to-noise separation model to separate usable electromagnetic data.

[0052] To verify that the available electromagnetic data is an electromagnetic signal free from human interference, the following steps are included after S108: The quality of available electromagnetic data is assessed to determine whether it is acceptable. If not, it is determined that the available electromagnetic data contains new interference data; Based on the new interference data, new interference data features are extracted and added to the basic electromagnetic interference feature library.

[0053] If the quality of the separated usable electromagnetic data is acceptable, and the usable electromagnetic data is determined to be pure and free of unidentified interference signals, then no further processing is required.

[0054] The following describes the assessment of the quality of available electromagnetic data: First, based on the available electromagnetic data, the corresponding time series plot is obtained; Convert available electromagnetic data into frequency domain data; Based on frequency domain data, plot the corresponding apparent resistivity diagram, phase curve diagram, polarization direction diagram and Nyquist diagram; Determine whether the time series plot, apparent resistivity plot, phase curve plot, polarization direction plot, and Nyquist plot satisfy their respective preset forms, and obtain the determination results; Based on this judgment result, an assessment will be made as to whether the qualification is satisfactory.

[0055] The preset formats for the time series plot, apparent resistivity plot, phase curve plot, polarization direction plot, and Nyquist plot are as follows: Figures 9-13 As shown.

[0056] Among them, such as Figure 9 As shown, the default format of this time series plot is to prevent any abnormal situations from occurring; as... Figure 10As shown, the preset format of this apparent resistivity map is distortion-free and has good continuity; as Figure 11 As shown, the preset form of this phase curve is distortion-free and has good continuity; as Figure 12 As shown, the preset form of this polarization pattern is random distribution; as Figure 13 As shown, the default form of this Nyquist plot is a distribution according to a default trend.

[0057] If any of the above diagrams does not meet the preset format, it is deemed unqualified. Conversely, if it does meet the preset format, it is considered qualified. Therefore, as shown... Figures 9-13 All of these are cases where each figure meets the corresponding preset format.

[0058] For cases that do not meet the requirements, based on the time series characteristics of the measurement points, we search for unknown interference sources in the vicinity, extract the corresponding interference data features, and add the new interference data features to the basic electromagnetic interference feature library to enrich the basic electromagnetic interference feature library and facilitate subsequent separation and identification.

[0059] One or more technical solutions in the embodiments of the present invention have at least the following technical effects or advantages: This invention provides an electromagnetic interference feature learning and separation method, comprising: collecting first electromagnetic data at each known type of human electromagnetic interference source; extracting time-domain waveform features of the known types of human electromagnetic interference sources based on the first electromagnetic data; constructing a basic electromagnetic interference feature library based on the time-domain waveform features; collecting second electromagnetic data at a distant reference point in the exploration area; extracting clean electromagnetic signals based on the second electromagnetic data; obtaining a signal-to-noise separation training dataset based on the basic electromagnetic interference feature library and the clean electromagnetic signals; training a signal-to-noise separation model based on the training dataset; collecting third electromagnetic data at a target acquisition area in the exploration area; and separating usable electromagnetic data based on the signal-to-noise separation model and the third electromagnetic data. The usable electromagnetic data is used to indicate deep ore bodies in the target acquisition area, thereby effectively improving the detection capability of frequency domain electromagnetic exploration in complex interference environments, especially in deep and peripheral resource exploration.

[0060] Example 2: Based on the same inventive concept, embodiments of the present invention also provide an electromagnetic interference feature learning and separation device, such as... Figure 14 As shown, it includes: The first acquisition module 1401 is used to acquire first electromagnetic data at each known type of human electromagnetic interference source. The first extraction module 1402 is used to extract the time-domain waveform features of the known type of human electromagnetic interference source based on the first electromagnetic data. Construction module 1403 is used to construct a basic electromagnetic interference feature library based on the time-domain waveform features; The second acquisition module 1404 is used to acquire second electromagnetic data by selecting a distant reference point in the exploration area; The second extraction module 1405 is used to extract pure electromagnetic signals based on the second electromagnetic data; Module 1406 is used to obtain a training dataset for signal-to-noise separation based on the basic electromagnetic interference feature library and the clean electromagnetic signal. Training module 1407 is used to train a signal-to-noise separation model based on the training dataset; The third acquisition module 1408 is used to select a target acquisition area in the exploration area to acquire third electromagnetic data. The separation module 1409 is used to separate usable electromagnetic data based on the signal-to-noise separation model and the third electromagnetic data. The usable electromagnetic data is used to indicate the deep ore body in the target acquisition area.

[0061] In one alternative implementation, it further includes: an evaluation module, used for: The quality of the available electromagnetic data is evaluated to determine whether it is acceptable. If not, it is determined that the available electromagnetic data contains new interference data; Based on the new interference data, new interference data features are extracted and added to the basic electromagnetic interference feature library.

[0062] In one alternative implementation, the evaluation module is used for: Based on the available electromagnetic data, the corresponding time series plot is obtained; Convert the available electromagnetic data into frequency domain data; Based on the frequency domain data, the corresponding apparent resistivity diagram, phase curve diagram, polarization direction diagram and Nyquist diagram are plotted. Determine whether the time series plot, apparent resistivity plot, phase curve plot, polarization direction plot and Nyquist plot satisfy their respective preset forms, and obtain the determination result; Based on the judgment results, an assessment is made as to whether the qualification is qualified.

[0063] In one optional implementation, the first acquisition module 1401 is used for: For each known type of human electromagnetic interference source, the minimum safe acquisition distance is determined based on the signal-to-noise ratio model and terrain attenuation factor; Based on the minimum safe acquisition distance, survey lines are laid out according to a preset range to acquire the first electromagnetic data.

[0064] In one optional implementation, the first acquisition module 1401 is specifically used for: The minimum safe acquisition distance is obtained using the following formula:

[0065] in, To minimize the safe sampling distance, The minimum signal-to-noise ratio threshold. For terrain attenuation factor, The attenuation factor of the source signal is denoted as . The type factor of the field source signal. The attenuation factor of the interference source signal. This is the type factor of the interference source signal.

[0066] In one optional implementation, the first extraction module 1402 is used for: Based on the first electromagnetic data, the wavelet coefficients are determined using a soft threshold function, specifically according to the following calculation formula:

[0067] in, For symbolic functions, For the first The first wavelet decomposition scale Wavelet coefficients corresponding to each time position The threshold is used to dynamically adjust the energy amplitude of the primary and secondary fields of the noise signal from known types of human electromagnetic interference sources. Based on the wavelet coefficients, the time-domain waveform features are reconstructed using inverse wavelet transform.

[0068] In one alternative implementation, module 1406 is obtained, which is used for: Based on the aforementioned basic electromagnetic interference feature library, noise data is obtained; Based on the noise data and the clean electromagnetic signal, a training dataset for signal-to-noise separation is obtained.

[0069] In one alternative implementation, the known type of anthropomorphic electromagnetic interference source includes: Mining construction areas, high-power industrial equipment, high-voltage transmission lines, low-voltage distribution lines, highways, railways, communication base stations, and radar stations.

[0070] Example 3: Based on the same inventive concept, embodiments of the present invention provide a computer device, such as... Figure 15 As shown, it includes a memory 1504, a processor 1502, and a computer program stored in the memory 1504 and executable on the processor 1502. When the processor 1502 executes the program, it implements the steps of the above-described electromagnetic interference feature learning and separation method.

[0071] Among them, Figure 15 In this document, a bus architecture (represented by bus 1500) is used. Bus 1500 may include any number of interconnected buses and bridges, linking various circuits including one or more processors represented by processor 1502 and memory represented by memory 1504. Bus 1500 may also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. Bus interface 1506 provides an interface between bus 1500 and receiver 1501 and transmitter 1503. Receiver 1501 and transmitter 1503 may be the same element, i.e., a transceiver, providing a unit for communicating with various other devices over a transmission medium. Processor 1502 is responsible for managing bus 1500 and general processing, while memory 1504 may be used to store data used by processor 1502 during operation.

[0072] Example 4: Based on the same inventive concept, embodiments of the present invention provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described electromagnetic interference feature learning and separation method.

[0073] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0074] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0075] Similarly, it should be understood that, in order to simplify the invention and aid in understanding one or more of the various inventive aspects, in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof. However, this disclosure should not be construed as reflecting an intention that the claimed invention requires more features than are explicitly recited in each embodiment. Rather, as reflected in each embodiment, inventive aspects lie in fewer than all features of the single foregoing disclosed embodiment. Therefore, the claims, following the detailed description, are hereby expressly incorporated into this detailed description, wherein each claim itself is a separate embodiment of the invention.

[0076] Those skilled in the art will understand that modules in the device of the embodiments can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiments can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components. Except where at least some of such features and / or processes or units are mutually exclusive, any combination can be used to combine all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or units of any method or device so disclosed. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0077] Furthermore, those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the specific implementation, any of the claimed embodiments can be used in any combination.

[0078] The various component embodiments of the present invention can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) can be used in practice to implement some or all of the functions of some or all of the components in the electromagnetic interference feature learning and separation device or computer device according to embodiments of the present invention. The present invention can also be implemented as a device or apparatus program (e.g., a computer program and computer program product) for performing part or all of the methods described herein. Such programs implementing the present invention can be stored on a computer-readable medium or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0079] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names.

Claims

1. A method for learning and separating electromagnetic interference features, characterized in that, include: First electromagnetic data were collected at each known type of human electromagnetic interference source; Based on the first electromagnetic data, extract the time-domain waveform features of the known types of human electromagnetic interference sources; Based on the aforementioned time-domain waveform characteristics, a basic electromagnetic interference feature library is constructed; In the exploration area, second electromagnetic data are collected by selecting distant reference points; Based on the second electromagnetic data, extract the pure electromagnetic signal; Based on the aforementioned basic electromagnetic interference feature library and the aforementioned clean electromagnetic signal, a training dataset for signal-to-noise separation is obtained; Based on the training dataset, a signal-to-noise separation model is trained. In the exploration area, select the target acquisition area to collect third electromagnetic data; Based on the signal-to-noise separation model and the third electromagnetic data, usable electromagnetic data is separated out, which is used to indicate the deep ore body in the target acquisition area.

2. The method as described in claim 1, characterized in that, After selecting the target exploration area, based on the signal-to-noise separation model, usable electromagnetic data is separated. This usable electromagnetic data is used to indicate deep ore bodies within the target acquisition area. The process further includes: The quality of the available electromagnetic data is evaluated to determine whether it is acceptable. If not, it is determined that the available electromagnetic data contains new interference data; Based on the new interference data, new interference data features are extracted and added to the basic electromagnetic interference feature library.

3. The method as described in claim 2, characterized in that, The quality of the available electromagnetic data is evaluated to determine whether it is acceptable, including: Based on the available electromagnetic data, the corresponding time series plot is obtained; Convert the available electromagnetic data into frequency domain data; Based on the frequency domain data, the corresponding apparent resistivity diagram, phase curve diagram, polarization direction diagram and Nyquist diagram are plotted. Determine whether the time series plot, apparent resistivity plot, phase curve plot, polarization direction plot and Nyquist plot satisfy their respective preset forms, and obtain the determination result; Based on the judgment results, an assessment is made as to whether the qualification is qualified.

4. The method as described in claim 1, characterized in that, First electromagnetic data were collected at each known type of human electromagnetic interference source, including: For each known type of human electromagnetic interference source, the minimum safe acquisition distance is determined based on the signal-to-noise ratio model and terrain attenuation factor; Based on the minimum safe acquisition distance, survey lines are laid out according to a preset range to acquire the first electromagnetic data.

5. The method as described in claim 4, characterized in that, For each known type of anthropogenic electromagnetic interference source, the minimum safe acquisition distance is determined based on the signal-to-noise ratio model and terrain attenuation factor, including: The minimum safe acquisition distance is obtained using the following formula: ； in, To minimize the safe sampling distance, The minimum signal-to-noise ratio threshold. For terrain attenuation factor, Here, is the attenuation factor of the source signal, and is the type factor of the source signal. The attenuation factor of the interference source signal. This is the type factor of the interference source signal.

6. The method as described in claim 1, characterized in that, Based on the first electromagnetic data, the time-domain waveform features of the known types of human electromagnetic interference sources are extracted, including: Based on the first electromagnetic data, the wavelet coefficients are determined using a soft threshold function, specifically according to the following calculation formula: ； in, For symbolic functions, For the first The first wavelet decomposition scale Wavelet coefficients corresponding to each time position The threshold is used to dynamically adjust the energy amplitude of the primary and secondary fields of the noise signal from known types of human electromagnetic interference sources. Based on the wavelet coefficients, the time-domain waveform features are reconstructed using inverse wavelet transform.

7. The method as described in claim 1, characterized in that, Based on the aforementioned basic electromagnetic interference feature library and the aforementioned clean electromagnetic signal, a training dataset for signal-to-noise separation is obtained, including: Based on the aforementioned basic electromagnetic interference feature library, noise data is obtained; Based on the noise data and the clean electromagnetic signal, a training dataset for signal-to-noise separation is obtained.

8. The method as described in claim 1, characterized in that, The known types of human-caused electromagnetic interference sources include: Mining construction areas, high-power industrial equipment, high-voltage transmission lines, low-voltage distribution lines, highways, railways, communication base stations, and radar stations.

9. An electromagnetic interference feature learning and separation device, characterized in that, include: The first acquisition module is used to acquire first electromagnetic data at each known type of human electromagnetic interference source; The first extraction module is used to extract the time-domain waveform features of the known type of human electromagnetic interference source based on the first electromagnetic data. The construction module is used to construct a basic electromagnetic interference feature library based on the time-domain waveform features; The second acquisition module is used to acquire second electromagnetic data by selecting a distant reference point in the exploration area; The second extraction module is used to extract pure electromagnetic signals based on the second electromagnetic data; The module is used to obtain a training dataset for signal-to-noise separation based on the basic electromagnetic interference feature library and the clean electromagnetic signal. The training module is used to train a signal-to-noise separation model based on the training dataset. The third acquisition module is used to select a target acquisition area in the exploration area to acquire third electromagnetic data; The separation module is used to separate usable electromagnetic data based on the signal-to-noise separation model and the third electromagnetic data. The usable electromagnetic data is used to indicate the deep ore body in the target acquisition area.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 8.

Images