Data denoising method for distributed artificial source electromagnetic exploration method
By combining step functions and short-time Fourier transforms, the problem of noise interference in electromagnetic exploration is solved, and the signal-to-noise ratio of data is improved and noise is automatically removed. This method is suitable for data processing in electromagnetic interference environments.
Patent Information
- Application Number
- CN202511722282.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-21
- Publication Date
- 2026-02-24
AI Technical Summary
In electromagnetic exploration, noise interference makes it impossible to obtain accurate resistivity information. Existing noise reduction methods are not effective in urban environments and cannot effectively handle complex noise.
By employing a combined approach of step function construction, data extension, short-time Fourier transform, and time-spectrum analysis, noise is eliminated and the signal-to-noise ratio is improved by detecting step position information in the data.
It effectively suppresses baseline noise, improves the signal-to-noise ratio of data, eliminates strong interference, reduces the difficulty of subsequent data processing, and is suitable for automated processing of large amounts of data.
Smart Images
Figure CN121559617A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of geophysical electromagnetic exploration technology, specifically relating to a data denoising method for distributed artificial source electromagnetic exploration methods. Background Technology
[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.
[0003] Artificial source electromagnetic methods are a type of geophysical exploration method that studies geological structures and solves geological problems related to engineering, environment, and disasters by observing the distribution of artificially established electric and electromagnetic fields based on the differences in electromagnetic properties between different underground media. It has advantages such as strong anti-interference ability, high resolution, and large exploration depth, and is widely used in oil and gas, mineral, and engineering fields. However, in electromagnetic exploration, noise interference with the effective signal is an unavoidable problem, making it impossible to obtain accurate resistivity information from the acquired data. Depending on the source, noise can be categorized as natural noise, human-induced noise, and internal instrument noise. In environments with strong interference, the sources of noise are often very complex. Faced with diverse noise types, different denoising methods each have their advantages and disadvantages. Using a single method to process data acquired in urban environments often encounters limitations. Summary of the Invention
[0004] To address the aforementioned problems, this invention proposes a data denoising method for distributed artificial source electromagnetic exploration methods. This invention processes artificial source electromagnetic data that is subject to strong interference, thereby achieving effective noise suppression and improving the data signal-to-noise ratio.
[0005] According to some embodiments, the present invention adopts the following technical solution: A data denoising method for distributed artificial source electromagnetic exploration methods includes the following steps: (1) Based on the data length of the target artificial source electromagnetic detection data, construct a step function H with a certain length; (2) Take the average value of the first and last parts of the target data by a certain length, and extend the data to both sides using this average value; (3) Convolve the step function H with the extended target data to obtain the detection sequence D that reflects the step position information in the data; (4) Divide the target data into segments at the step of the detection sequence D, fit each segment with a second-order polynomial, subtract the fitted binomial from the target data to obtain the (n-1)th processed data, with the initial value of n-1 being 1. (5) Using the period length of the signal used in the exploration as the time window length, and according to the preset time window overlap, perform short-time Fourier transform on the (n-1)th processed data to obtain the time spectrum matrix of the (n-1)th processed data. (6) Squaring all elements of the time-spectrum matrix, and then summing the elements corresponding to each time window to obtain the time-energy distribution of the (n-1)th processed data; (7) Calculate the relative standard deviation S of the energy of the (n-1)th processed data in all time windows. If the relative standard deviation S is greater than the quality index threshold, delete the time window with the largest energy in the first processed data to obtain the nth processed data; otherwise, end the processing. (8) For the nth data to be processed, n≥2, iteratively execute steps (5)-(7) until the set conditions are met.
[0006] As an alternative implementation method, the step function H is:
[0007] The length of the step function H is 1 / 20 of the data length, rounded to even numbers.
[0008] As an alternative implementation, when extending the data to both sides, the extension length is the length of the step function H.
[0009] As an alternative implementation, the time window overlap is set to 10% of the signal period length, and the calculation result is rounded down.
[0010] As an alternative implementation, for the nth processed data, n≥2, the process of iteratively executing steps (5)-(7) includes: Using the period length of the signal used in exploration as the time window length, and according to the preset time window overlap, a short-time Fourier transform is performed on the processed data n to obtain the time spectrum matrix of the processed data n; Squaring all elements of the time-spectrum matrix and then summing the elements corresponding to each time window yields the time-window-energy distribution of the processed data n. Calculate the relative standard deviation S of the energy of the processed data n across all time windows. If the relative standard deviation S is greater than the quality index threshold, delete the time window with the highest energy in the processed data n and obtain processed data n+1; otherwise, end the processing.
[0011] As a further implementation, the time window overlap is set to 10% of the signal period length, rounded to the nearest integer.
[0012] As an alternative implementation method, the iteration is performed at least once.
[0013] As an alternative implementation method, the set condition is that the relative standard deviation S is less than or equal to a set threshold.
[0014] As an alternative implementation method, the threshold value is set to a range of 0.5% to 3%.
[0015] A data denoising system for distributed artificial source electromagnetic exploration methods includes: The step function construction module is configured to construct a step function H of a certain length based on the data length of the target artificial source electromagnetic detection data; The extension module is configured to take the average value of the first and last parts of the target data by a certain length, and then extend the data to both sides using this average value. The convolution module is configured to convolve the step function H with the extended target data to obtain the detection sequence D that reflects the step position information in the data. The fitting module is configured to segment the target data at the step of the detection sequence D, fit each segment with a second-order polynomial, subtract the fitting binomial from the target data to obtain the (n-1)th processed data, with the initial value of n-1 being 1. The transformation module is configured to use the period length of the signal used in exploration as the time window length, and perform a short-time Fourier transform on the (n-1)th processed data according to the preset time window overlap to obtain the time spectrum matrix of the (n-1)th processed data. The calculation module is configured to square all elements of the time-spectrum matrix and then sum the elements corresponding to each time window to obtain the time-window-energy distribution of the (n-1)th processed data. The comparison module is configured to calculate the relative standard deviation S of the energy of the (n-1)th processed data across all time windows. If the relative standard deviation S is greater than the quality index threshold, the time window with the highest energy in the first processed data is deleted to obtain the nth processed data; otherwise, the processing ends. The iteration module is configured to process the nth data (n≥2), iteratively calling the transformation module, calculation module, and comparison module until the set conditions are met.
[0016] An electronic device includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, wherein the computer instructions, when executed by the processor, perform the steps in the method described above.
[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention processes artificial source electromagnetic method (EMB) detection data affected by background noise, suppressing baseline noise and improving the signal-to-noise ratio. Simultaneously, it removes heavily interfered portions of the data, thereby increasing the utilization rate of EMB detection data under electromagnetic interference environments and reducing the difficulty of subsequent data processing. As a data processing method, it can be widely applied to artificial source EMB exploration in electromagnetic interference environments.
[0018] After the quality index threshold is preset in this invention, all subsequent operations have precise calculation formulas or restrictions, which can be completely executed by computer programs. This can free up manpower and is suitable for the rapid and automated processing of large amounts of artificial source electromagnetic detection data.
[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description
[0020] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.
[0021] Figure 1 This is a flowchart of a data denoising method for a distributed artificial source electromagnetic exploration method according to one or more embodiments of the present invention; Figure 2 The diagram shows the comparison between the extended target data (illustrated) and the detection sequence. It can be seen that the peak position of the detection sequence corresponds to the step position in the target data. Figure 3 A comparison chart showing the simulation target data before and after baseline removal; Figure 4 for Figure 3 The time spectrum and energy-time window distribution curve of the processed data 1 obtained in the process; Figure 5 To Figure 3 The processed data 1 obtained in step 7 is processed to obtain the time spectrum and energy-time window distribution curve of processed data 2; Figure 6 To Figure 3 A comparison between the simulated target data and the result after complete processing. Detailed Implementation
[0022] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0023] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.
[0024] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.
[0025] Where there is no conflict, the embodiments and features described in this application may be combined with each other.
[0026] Example 1 This embodiment discloses an automated denoising method for charging method data under strong interference conditions. It uses techniques such as periodic superposition, step detection, and potential difference extraction to process the noise-affected charging method data. Figure 1 As shown, the method includes the following steps: Step 1: Based on the data length (total number of data points) of the target artificial source electromagnetic detection data, construct the step function H according to the following formula:
[0027] In this step, the length of the step function H is 1 / 20 of the data length (rounded to even numbers).
[0028] Step 2: Take the average value of the first and last A data points of the target data, and extend the data on both sides using this average value. The extension length is the length of the step function H.
[0029] In this step, the value of A is the smaller of 1 / 100 of the data length (rounded to the nearest integer) and 10.
[0030] Step 3: Convolve H with the extended target data to obtain the detection sequence D, which reflects the step position information in the data. The correspondence between the detection sequence and the step in the target data is as follows: Figure 2 As shown.
[0031] Step 4: Divide the target data into segments at the peak of the detection sequence D, fit each segment with a second-order polynomial, and subtract the fitted binomial from the target data to obtain processed data 1.
[0032] Figure 3 The paper presents a comparison of steps 1 through 4 before and after performing them on a simulated target data.
[0033] Step 5: Using the period length of the signal used in the exploration as the time window length, set the time window overlap length to 10% of the signal period length (rounded to the nearest integer), perform a short-time Fourier transform on processed data 1 to obtain the time-spectrum matrix of processed data 1, as follows. Figure 4As shown in the upper part.
[0034] Step 6: Square all elements of the time-spectrum matrix, then sum the elements corresponding to each time window to obtain the time-window-energy distribution of processed data 1, as shown below. Figure 4 The lower half is shown.
[0035] Step 7: Calculate the relative standard deviation S of the energy of processed data 1 across all time windows. If S is greater than the quality index threshold (which can be set according to actual needs, and the recommended value range is 0.5%~3%), then delete the time window with the highest energy in processed data 1 to obtain processed data 2; otherwise, end the processing.
[0036] Step 8: If the processing is not finished, repeat the following loop for the processed data n (n≥2): Using the period length of the signal used in exploration as the time window length, and setting the time window overlap to 10% of the signal period length (rounded to the nearest integer), a short-time Fourier transform is performed on the processed data n to obtain the time spectrum matrix of the processed data n.
[0037] Squaring all elements of the time-spectrum matrix and then summing the elements corresponding to each time window yields the energy-time-window distribution of the processed data n.
[0038] Calculate the relative standard deviation S of the energy of the processed data n across all time windows. If S is greater than the quality index threshold, delete the time window with the highest energy in the processed data n to obtain processed data n+1; otherwise, end the processing.
[0039] Figure 5 That is to Figure 3 , 4 The time-spectrum and energy-time window distribution of processed data 2 obtained by performing the above loop on processed data 1 are shown.
[0040] In this embodiment, the quality index threshold is set to 1%, and the processing result after a complete loop is shown below. Figure 6 The processing successfully removed strong noise and baseline noise from the original target data.
[0041] Example 2 A data denoising method for distributed artificial source electromagnetic exploration methods includes the following steps: Based on the data length of the electromagnetic detection data from the target artificial source, a step function H with a certain length is constructed; Take the average value of the first and last parts of the target data by a certain length, and extend the data to both sides using this average value. The extension length is the length of the step function H. Convolve H with the extended target data to obtain the detection sequence D, which reflects the step position information in the data; The target data is segmented at the step of the detection sequence D, and a second-order polynomial is used to fit each segment. The fitted binomial is subtracted from the target data to obtain the processed data 1. Using the period length of the signal used in the exploration as the time window length, and setting the time window overlap to 10% of the signal period length (rounded to the nearest integer), a short-time Fourier transform is performed on the processed data 1 to obtain the time spectrum matrix of processed data 1. Squaring all elements of the time-spectrum matrix and then summing the elements corresponding to each time window yields the time-window-energy distribution of processed data 1. Calculate the relative standard deviation S of the energy of processed data 1 across all time windows. If S is greater than the quality index threshold, delete the time window with the highest energy in processed data 1 to obtain processed data 2; otherwise, end the processing. Continue executing this loop for the data n (n≥2): Using the period length of the signal used in exploration as the time window length, and setting the time window overlap to 10% of the signal period length (rounded to the nearest integer), a short-time Fourier transform is performed on the processed data n to obtain the time spectrum matrix of the processed data n. Squaring all elements of the time-spectrum matrix and then summing the elements corresponding to each time window yields the time-window-energy distribution of the processed data n. Calculate the relative standard deviation S of the energy of the processed data n across all time windows. If S is greater than the quality index threshold, delete the time window with the highest energy in the processed data n to obtain processed data n+1; otherwise, end the processing.
[0042] Example 3 A data denoising system for distributed artificial source electromagnetic exploration methods includes: The step function construction module is configured to construct a step function H of a certain length based on the data length of the target artificial source electromagnetic detection data; The extension module is configured to take the average value of the first and last parts of the target data by a certain length, and then extend the data to both sides using this average value. The convolution module is configured to convolve the step function H with the extended target data to obtain the detection sequence D that reflects the step position information in the data. The fitting module is configured to segment the target data at the step of the detection sequence D, fit each segment with a second-order polynomial, subtract the fitting binomial from the target data to obtain the (n-1)th processed data, with the initial value of n-1 being 1. The transformation module is configured to use the period length of the signal used in exploration as the time window length, and perform a short-time Fourier transform on the (n-1)th processed data according to the preset time window overlap to obtain the time spectrum matrix of the (n-1)th processed data. The calculation module is configured to square all elements of the time-spectrum matrix and then sum the elements corresponding to each time window to obtain the time-window-energy distribution of the (n-1)th processed data. The comparison module is configured to calculate the relative standard deviation S of the energy of the (n-1)th processed data across all time windows. If the relative standard deviation S is greater than the quality index threshold, the time window with the highest energy in the first processed data is deleted to obtain the nth processed data; otherwise, the processing ends. The iteration module is configured to process the nth data (n≥2), iteratively calling the transformation module, calculation module, and comparison module until the set conditions are met.
[0043] Example 4 An electronic device includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, wherein the computer instructions, when executed by the processor, perform the steps in the method described above.
[0044] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of one or more computer-usable storage media (including, but not limited to, disk storage, etc.) containing computer-usable program code. CD - ROM It takes the form of a computer program product implemented on (such as optical memory, etc.).
[0045] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0046] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0047] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0048] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made by those skilled in the art without creative effort within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A data denoising method for distributed artificial source electromagnetic exploration methods, characterized in that, Includes the following steps: (1) Based on the data length of the target artificial source electromagnetic detection data, construct a step function H with a certain length; (2) Take the average value of the first and last parts of the target data by a certain length, and extend the data to both sides using this average value; (3) Convolve the step function H with the extended target data to obtain the detection sequence D that reflects the step position information in the data; (4) Divide the target data into segments at the step of the detection sequence D, fit each segment with a second-order polynomial, subtract the fitted binomial from the target data to obtain the (n-1)th processed data, with the initial value of n-1 being 1. (5) Using the period length of the signal used in the exploration as the time window length, and according to the preset time window overlap, perform short-time Fourier transform on the (n-1)th processed data to obtain the time spectrum matrix of the (n-1)th processed data. (6) Squaring all elements of the time-spectrum matrix, and then summing the elements corresponding to each time window to obtain the time-energy distribution of the (n-1)th processed data; (7) Calculate the relative standard deviation S of the energy of the (n-1)th processed data in all time windows. If the relative standard deviation S is greater than the quality index threshold, delete the time window with the largest energy in the first processed data to obtain the nth processed data; otherwise, end the processing. (8) For the nth data to be processed, n≥2, iteratively execute steps (5)-(7) until the set conditions are met.
2. The data denoising method for distributed artificial source electromagnetic exploration as described in claim 1, characterized in that, The step function H is: The length of the step function H is 1 / 20 of the data length, rounded to even numbers.
3. The data denoising method for distributed artificial source electromagnetic exploration as described in claim 1, characterized in that, When extending the data to both sides, the extension length is the length of the step function H.
4. The data denoising method for distributed artificial source electromagnetic exploration as described in claim 1, characterized in that, The time window overlap is set to 10% of the signal period length, and the calculation result is rounded down.
5. The data denoising method for distributed artificial source electromagnetic exploration as described in claim 1, characterized in that, For the nth data to be processed, n≥2, the iterative execution of steps (5)-(7) includes: Using the period length of the signal used in exploration as the time window length, and according to the preset time window overlap, a short-time Fourier transform is performed on the processed data n to obtain the time spectrum matrix of the processed data n; Squaring all elements of the time-spectrum matrix and then summing the elements corresponding to each time window yields the time-window-energy distribution of the processed data n. Calculate the relative standard deviation S of the energy of the processed data n across all time windows. If the relative standard deviation S is greater than the quality index threshold, delete the time window with the highest energy in the processed data n and obtain processed data n+1; otherwise, end the processing.
6. The data denoising method for distributed artificial source electromagnetic exploration as described in claim 5, characterized in that, The iteration must be executed at least once.
7. The data denoising method for distributed artificial source electromagnetic exploration as described in claim 1, characterized in that, The condition is met when the relative standard deviation S is less than or equal to the set threshold.
8. The data denoising method for distributed artificial source electromagnetic exploration as described in claim 1, characterized in that, The threshold value is set to range from 0.5% to 3%.
9. A data denoising system for distributed artificial source electromagnetic exploration methods, characterized in that, include: The step function construction module is configured to construct a step function H of a certain length based on the data length of the target artificial source electromagnetic detection data; The extension module is configured to take the average value of the first and last parts of the target data by a certain length, and then extend the data to both sides using this average value. The convolution module is configured to convolve the step function H with the extended target data to obtain the detection sequence D that reflects the step position information in the data. The fitting module is configured to segment the target data at the step of the detection sequence D, fit each segment with a second-order polynomial, subtract the fitting binomial from the target data to obtain the (n-1)th processed data, with the initial value of n-1 being 1. The transformation module is configured to use the period length of the signal used in exploration as the time window length, and perform a short-time Fourier transform on the (n-1)th processed data according to the preset time window overlap to obtain the time spectrum matrix of the (n-1)th processed data. The calculation module is configured to square all elements of the time-spectrum matrix and then sum the elements corresponding to each time window to obtain the time-window-energy distribution of the (n-1)th processed data. The comparison module is configured to calculate the relative standard deviation S of the energy of the (n-1)th processed data across all time windows. If the relative standard deviation S is greater than the quality index threshold, the time window with the highest energy in the first processed data is deleted to obtain the nth processed data; otherwise, the processing ends. The iteration module is configured to process the nth data (n≥2), iteratively calling the transformation module, calculation module, and comparison module until the set conditions are met.
10. An electronic device, characterized in that, It includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, which, when executed by the processor, perform the steps of the method according to any one of claims 1-8.