Underground medium density inversion method and system based on high-precision microgravity
By using frequency domain analysis and mode decomposition to adjust the wavelet threshold, the problem of multi-factor superposition interference in gravity data was solved, and high-precision underground medium density inversion was achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- DAQING YILAI TESTING TECH SERVICE CO LTD
- Filing Date
- 2025-08-07
- Publication Date
- 2026-04-14
AI Technical Summary
Existing technologies cannot fully account for the superimposed interference of multiple factors in gravity data inversion, resulting in poor noise reduction and affecting the accuracy of underground medium density inversion.
By using frequency domain analysis and mode decomposition, the superposition outliers and superposition interference confidence of gravity subsequences are extracted, the wavelet threshold is adjusted for denoising, and the density of the subsurface medium is inverted by combining the Parker-Oldenburg density interface inversion algorithm.
This improved the denoising effect of gravity data, enhanced the accuracy of underground medium density inversion, and ensured the accuracy of the inversion results.
Smart Images
Figure CN120908875B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of underground medium density inversion technology, specifically to an underground medium density inversion method and system based on high-precision microgravity. Background Technology
[0002] With the rapid development of gravity measurement technology, satellite gravity detection technology, and satellite altimetry technology, people's understanding of the gravitational field has been continuously deepened. By using gravity anomaly monitoring instruments with an accuracy of microgal level, the monitoring needs of microgal level gravity anomalies can be met, thereby more accurately inverting the density of underground media, which helps to conduct refined monitoring of underground oil reservoirs.
[0003] In existing technologies, gravity data is used to invert the density of subsurface media. However, due to interference from factors such as elevation errors and zero-point drift, existing technologies typically employ Dual-tree Complex Wavelet Transform (DTCWT) to denoise the gravity data, aiming to improve the accuracy of gravity data and thus more accurately invert the density of subsurface media. However, because gravity data is subject to the superposition of interference from multiple factors, existing technologies cannot adequately account for this superposition interference in their denoising, resulting in poor denoising performance and affecting the accuracy of subsurface media density inversion. Summary of the Invention
[0004] To address the aforementioned technical problems, the purpose of this application is to provide a method and system for inverting the density of underground media based on high-precision microgravity. The specific technical solution adopted is as follows:
[0005] This application provides a method for inverting the density of underground media based on high-precision microgravity, including the following steps:
[0006] Obtain gravity data at each observation point within the surface measurement area for underground medium density inversion;
[0007] Gravity subsequences at each observation point are extracted and frequency domain analysis is performed. Based on the degree of difference between the instantaneous amplitude of the gravity subsequence in the frequency domain and the corresponding amplitude in the other frequency domain, as well as the discrepancy anomaly, the superposition anomaly value of each gravity subsequence at each observation point is obtained.
[0008] Mode decomposition is performed on each gravity subsequence, and the superposition feature value of each mode component is extracted by the random variation of each mode component. Based on the differences between the superposition feature values of each gravity subsequence at different observation points and the deviation of superposition outliers, the superposition interference confidence of each gravity subsequence at each observation point is obtained. By the change of the superposition interference confidence of all gravity subsequences at each observation point, the superposition influence duration at each observation point is obtained. This is used to adjust the wavelet threshold for wavelet transform of gravity data at each observation point to denoise the gravity data.
[0009] The gravity data of each observation point after denoising were corrected to extract the Bouguer gravity anomaly of each observation point, and the inversion result of the subsurface medium density was obtained by the Parker-Oldenburg density interface inversion algorithm.
[0010] Preferably, the gravity data of each observation point are arranged in chronological order to form a gravity data sequence for each observation point, and then evenly divided into multiple subsequences as gravity subsequences for each observation point.
[0011] Preferably, the method for obtaining the superposition outlier values of each gravity subsequence at each observation point is as follows:
[0012] In the formula, This represents the superposition anomaly value of the k-th gravity subsequence at the current observation point. This is the first mean of the k-th gravity subsequence at the current observation point. This is the first sum of the k-th gravity subsequence at the current observation point.
[0013] Preferably, for the current observation point, a frequency domain transformation is performed on each gravity subsequence to obtain the amplitude corresponding to all frequencies within the gravity subsequence, and a Hilbert-Huang transform is performed on each gravity subsequence to extract the instantaneous amplitude of the gravity subsequence. The mean of the absolute differences between the instantaneous amplitude of each gravity subsequence and the amplitude corresponding to all frequencies is taken as the first mean of each gravity subsequence.
[0014] Preferably, anomaly detection is performed on the absolute differences corresponding to each gravity subsequence, and the sum of the local outlier factors of all absolute differences is taken as the first sum of each gravity subsequence.
[0015] Preferably, the method for extracting the superposition feature values of each modal component is as follows:
[0016] Calculate the first-order difference vector of each modal component and the permutation entropy of each modal component. The normalized result of the product of the mean of the absolute values of all elements in the first-order difference vector and the permutation entropy is used as the superposition eigenvalue of each modal component.
[0017] Preferably, the method for obtaining the superposition interference confidence of each gravity subsequence at each observation point is as follows:
[0018] In the formula, The confidence level of the superposition interference of the k-th gravity subsequence at the current observation point. This represents the degree of difference between the superimposed eigenvectors of the current observation point and its j-th neighboring observation point in the k-th gravity subsequence. This is the absolute difference between the superposition anomaly of the current observation point and its j-th neighboring observation point in the k-th gravity subsequence. To avoid constants with a denominator of 0, the multiple observation points that are closest to the current observation point are all recorded as adjacent observation points of the current observation point.
[0019] Preferably, the method for obtaining the duration of the superimposed influence of each observation point is as follows:
[0020] The superposition interference confidence scores of all gravity subsequences at each observation point are arranged in time sequence to obtain the superposition confidence score sequence of each observation point. The ratio between the mean of all elements in the superposition confidence score sequence and the mean of the absolute values of all elements in the first difference sequence of the superposition confidence score sequence is taken as the superposition influence duration of each observation point.
[0021] Preferably, the method for adjusting the wavelet threshold for wavelet transform of gravity data at each observation point is as follows:
[0022] In the formula, The wavelet adjustment threshold for the gravity data at the current observation point. This represents the standard deviation of the noise estimated in the wavelet denoising algorithm. The length of the gravity data sequence at the current observation point. For adjustment coefficients, It is the hyperbolic tangent function. This represents the duration of the superimposed influence at the current observation point.
[0023] This application also provides a high-precision microgravity-based underground medium density inversion system, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described high-precision microgravity-based underground medium density inversion methods.
[0024] As can be seen from the above, the method and system for inverting the density of underground media based on high-precision microgravity provided in this application have at least the following beneficial effects:
[0025] This application employs frequency domain analysis combined with outlier detection algorithms to more accurately measure the abnormal characteristics of the instantaneous amplitude of gravity subsequences under the superposition of multiple frequencies. This clearly demonstrates the severity of the interference of external factors on the gravity data within the gravity subsequence, which is beneficial for more accurately determining the wavelet threshold and more accurately denoising the gravity data sequence.
[0026] Furthermore, this application extracts superimposed feature vectors by employing mode decomposition and considers the consistency of the superimposed influence of external noise between adjacent observation points. This allows for an accurate measurement of the confidence characteristics of the gravity subsequence at the observation point being affected by the superimposed interference of multiple external noise factors, thereby improving the accuracy of the analysis of multi-factor superimposed interference.
[0027] This application accurately measures the persistence of the impact of superimposed interference on gravity data by superimposing the confidence of interference over multiple time periods. It also accurately adjusts the wavelet threshold in the wavelet denoising algorithm by superimposing the persistence of the interference, thereby improving the denoising effect on gravity data and enabling more effective and accurate inversion of underground medium density using high-precision microgravity data. Attached Figure Description
[0028] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0029] Figure 1 A flowchart illustrating the steps of the high-precision microgravity-based underground medium density inversion method provided in this application. Detailed Implementation
[0030] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the underground medium density inversion method and system based on high-precision microgravity proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.
[0031] Unless otherwise specified and limited, terms such as “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a circuit structure, article, or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the article or device that includes said element. Furthermore, the term “and / or” as used herein includes any and all combinations of one or more of the associated listed items. 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 application pertains.
[0032] The following description, in conjunction with the accompanying drawings, details the specific scheme of the high-precision microgravity-based underground medium density inversion method and system provided in this application.
[0033] Please see Figure 1 The document illustrates a flowchart of a high-precision microgravity-based method for inverting the density of underground media according to an embodiment of this application, including the following steps:
[0034] Step 1: Obtain gravity data for each observation point within the surface measurement area for underground medium density inversion.
[0035] To improve the accuracy of inverting the density of underground media, it is necessary to more accurately denoise the gravity data collected by quantum smart sensors, eliminating noise interference from external factors such as elevation errors and zero-point drift, so as to more effectively and accurately invert the density of underground media using high-precision microgravity signals.
[0036] Firstly, in this embodiment, preferably, by selecting a surface measurement area for underground medium density inversion and setting 10 observation points within the measurement area, the implementer can set the location of the measurement area and the number of observation points according to the actual situation.
[0037] Furthermore, micro-galaxy-level gravity data is collected at each observation point using a quantum intelligent sensor. In this embodiment, the gravity acquisition time interval is 1 second, and the acquisition time for a single observation point is 30 minutes. Implementers can adaptively set these parameters according to actual needs. The quantum intelligent sensor used in this embodiment is a quantum gravimeter, which possesses the core functions of an intelligent sensor, including gravity sensing, data acquisition, analysis and processing, storage, and communication. The quantum gravimeter uses sensor technology to sense the differences in gravity of atoms or particles in a gravitational field, thereby achieving the function of acquiring gravity data in the gravitational field.
[0038] Meanwhile, to facilitate subsequent noise reduction analysis of the micro-galaxy level gravity data, the analysis and processing functions of the intelligent sensor are preferably used in this embodiment to arrange the micro-galaxy level gravity data collected at a single observation point in chronological order to obtain the gravity data sequence of each observation point in the measurement area, thereby facilitating the subsequent storage and communication of gravity data at different observation points.
[0039] Step 2: Extract the gravity subsequences of each observation point and perform frequency domain analysis. Based on the degree of difference between the instantaneous amplitude of the gravity subsequence in the frequency domain and the corresponding amplitude in the other frequency domain, as well as the discrepancy anomaly, obtain the superposition anomaly value of each gravity subsequence at each observation point.
[0040] Existing technologies typically employ Dual-tree Complex Wavelet Transform (DTCWT) to denoise gravity data. However, the wavelet threshold used in these wavelet denoising algorithms is a general threshold that does not adequately account for the superimposed interference of multiple factors to accurately denoise gravity data. This can easily lead to the wavelet threshold being too large or too small.
[0041] However, an excessively large wavelet threshold can lead to an overly smooth denoised signal, while an excessively small threshold can fail to effectively suppress noise components in gravity data, ultimately affecting the accuracy of subsurface density inversion. Therefore, in practice, gravity data collected by quantum smart sensors is susceptible to interference from external factors. It is necessary to fully consider the superimposed interference of multiple factors on the gravity field signal and improve the wavelet threshold in the wavelet denoising algorithm. By enhancing the denoising effect of gravity data, the accuracy of subsurface density inversion can be improved.
[0042] In order to extract the interference characteristics of gravity data at different time periods, the gravity data sequence of each observation point is evenly divided into K subsequences, which are denoted as the K gravity subsequences of each observation point. In this embodiment, K is set to 30, so that the time length of each gravity subsequence is 1 minute.
[0043] Furthermore, for each gravity subsequence at each observation point, each gravity subsequence is used as input to the Discrete Fourier Transform (DFT), and the amplitude corresponding to all frequencies within each gravity subsequence is obtained through the DFT. Simultaneously, each gravity subsequence at each observation point is used as input to the Hilbert-Huang Transform (HFT), and the instantaneous amplitude of each gravity subsequence is obtained through the HFT. Both the DFT and the Hilbert-Huang Transform are well-known techniques and will not be elaborated further.
[0044] Generally, the instantaneous amplitude is the amplitude under the superposition of different frequencies. The greater the difference between the instantaneous amplitude and the amplitude corresponding to all frequencies, and the more prominent the abnormality of the difference in amplitude corresponding to all frequencies, the more it reflects the abnormal characteristics of the instantaneous amplitude under the superposition of multiple frequencies. At this time, the gravity data is more severely affected by the superposition of multiple factors, indicating that the gravity data collected by the quantum intelligent sensor is more complexly affected by interference. Therefore, it is more necessary to set a larger wavelet threshold to more effectively suppress the superposition noise in the gravity data.
[0045] Therefore, for the k-th gravity subsequence at the current observation point, the absolute difference between the instantaneous amplitude of the k-th gravity subsequence and the amplitude corresponding to all frequencies within the gravity subsequence is calculated. The mean of all absolute differences is recorded as the first mean of the k-th gravity subsequence at the current observation point. The larger the first mean, the deeper the instantaneous amplitude is affected by the superposition of different frequencies. At this time, the gravity data is more easily affected by the superposition interference of external noise components.
[0046] Meanwhile, all absolute differences corresponding to the kth gravity subsequence at the current observation point are used as input to the LOF outlier detection algorithm. The preset neighborhood parameter in the algorithm is 10. The LOF outlier detection algorithm obtains the local outlier factors of all absolute differences. The sum of the local outlier factors of all absolute differences is recorded as the first sum of the kth gravity subsequence at the current observation point. The larger the first sum, the more prominent the abnormality of the amplitude difference corresponding to all frequencies as a whole, and the more it can reflect the abnormal characteristics of the instantaneous amplitude under the superposition of multiple frequencies.
[0047] Based on the above analysis, in this embodiment, taking the current observation point as an example, the superposition anomaly value of the k-th gravity subsequence at the current observation point is calculated:
[0048] In the formula, This represents the superposition anomaly value of the k-th gravity subsequence at the current observation point. This is the first mean of the k-th gravity subsequence at the current observation point. This is the first sum of the k-th gravity subsequence at the current observation point.
[0049] Among them, superimposed outliers reflect the abnormal characteristics of instantaneous amplitude under the superposition of multiple frequencies. The larger the superimposed outlier, the more prominent the abnormal characteristics of instantaneous amplitude under the superposition of multiple frequencies of the gravity subsequence. At this time, the gravity data is more severely affected by superimposed noise interference from external factors such as elevation error and zero drift. The gravity data sensed by the quantum intelligent sensor is less accurate, and a larger wavelet threshold needs to be set to more effectively suppress the superposition of noise in the gravity data.
[0050] Step 3: Perform mode decomposition on each gravity subsequence and extract the superposition feature value of each mode component based on the random variation of each mode component. Based on the differences between the superposition feature values of each gravity subsequence at different observation points and the deviation of superposition outliers, obtain the superposition interference confidence of each gravity subsequence at each observation point. By observing the changes in the superposition interference confidence of all gravity subsequences at each observation point, obtain the superposition influence duration at each observation point. This is used to adjust the wavelet threshold for wavelet transform of gravity data at each observation point to denoise the gravity data.
[0051] Furthermore, the k-th gravity subsequence at each observation point is used as the input to the EMD empirical mode decomposition algorithm. The EMD empirical mode decomposition algorithm decomposes the gravity subsequence into various modal components through an iterative process. If the volatility of the element changes within the modal component is higher and the uncertainty of the element changes within the modal component is greater, then the modal component is more likely to cause superimposed interference to the gravity data, thereby affecting the authenticity of the gravity data.
[0052] Therefore, the first-order difference vector of each modal component is calculated to reflect the fluctuation changes between adjacent elements within the modal component. At the same time, the permutation entropy of each modal component is calculated to reflect the uncertainty of element changes within the modal component. Furthermore, the normalized result of the product between the mean of the absolute values of all elements in the first-order difference vector and the permutation entropy is denoted as the superposition eigenvalue of each modal component. The larger the superposition eigenvalue, the more likely the modal component is to have a superposition interference effect on the gravity data.
[0053] Furthermore, for the k-th gravity subsequence at the current observation point, the superposition eigenvalues of all modal components corresponding to the k-th gravity subsequence are arranged in the order of modal decomposition to obtain the superposition eigenvector of the k-th gravity subsequence at each observation point. If the superposition eigenvectors of each observation point within the measurement area are more similar to those of its neighboring observation points, and the difference in superposition anomalies between each observation point and its neighboring observation points is smaller, it indicates that the superposition influence of multiple external noise factors on the observation points in the local area is more consistent. This more fully and accurately demonstrates that the central observation point is subject to the superposition interference of multiple external factors. In this case, the reliability of the gravity data collected by the quantum intelligent sensor is lower.
[0054] Therefore, the M observation points with the closest Euclidean distance to each observation point are denoted as the M neighboring observation points of each observation point, where M is 4, representing the nearest neighboring observation points in the four directions of the central observation point. Based on the neighboring observation points of each observation point, the superposition interference confidence of the k-th gravity subsequence of each observation point is calculated:
[0055] ;
[0056] In the formula, The confidence level of the superposition interference of the k-th gravity subsequence at the current observation point. To measure the degree of difference between the superimposed feature vector of the current observation point and its j-th neighboring observation point in the k-th gravity subsequence, this embodiment uses DTW dynamic programming distance to measure the degree of difference. This is the absolute difference between the superposition anomaly of the current observation point and its j-th neighboring observation point in the k-th gravity subsequence. To avoid constants with a denominator of 0, the value range is (0.01, 0.05), and in this embodiment, the value is 0.02.
[0057] Among them, the superposition interference confidence reflects the confidence characteristics of the observation point being affected by the superposition interference of multiple external noise factors. The higher the superposition interference confidence, the higher the credibility of the superposition interference of noise factors within the corresponding time period of the gravity subsequence. If the superposition interference confidence of multiple time periods is at a high level, it means that the duration of the superposition interference of noise factors on the gravity data is longer. In this case, a larger wavelet threshold should be set to more effectively suppress the superposition of noise in the gravity data. Conversely, the shorter the duration of the superposition interference of noise factors on the gravity data, the smaller the wavelet threshold should be set to avoid the denoised signal being too smooth, thereby improving the accuracy of the gravity data collected by the quantum intelligent sensor.
[0058] Therefore, the superposition interference confidence scores of all gravity subsequences at each observation point are arranged in chronological order to obtain a superposition confidence score sequence for each observation point. The mean of the absolute values of all elements in the first-order difference sequence of the superposition confidence score sequence is then used. The smaller this mean, the better the continuity of the superposition interference confidence score within the superposition confidence score sequence. Furthermore, the ratio between the mean of all elements in the superposition confidence score sequence and the mean of the absolute values of all elements in the first-order difference sequence of the superposition confidence score sequence is used as the superposition influence duration at each observation point. The larger the superposition influence duration, the longer the duration of the superposition interference from noise factors on the gravity data, and therefore a larger wavelet threshold should be set.
[0059] Therefore, based on the duration of the superposition effect at each observation point, the wavelet adjustment threshold of the gravity data sequence at each observation point is calculated:
[0060] ;
[0061] In the formula, The wavelet adjustment threshold for the gravity data at the current observation point. This represents the standard deviation of the noise estimated in the wavelet denoising algorithm. The length of the gravity data sequence at the current observation point. To adjust the coefficient and avoid the wavelet threshold being adjusted too high or too low, the adjustment coefficient in this embodiment is set to 0.5. It is the hyperbolic tangent function. The duration of the superimposed influence at the current observation point, where, The calculation is a well-known technique and will not be elaborated further.
[0062] The longer the duration of the superimposed noise interference on the gravity data, the larger the wavelet threshold should be set to more effectively suppress the superimposed noise in the gravity data; conversely, the shorter the duration of the superimposed noise interference on the gravity data, the smaller the wavelet threshold should be set to avoid the denoised signal being too smooth, thereby improving the denoising effect of the gravity data collected by the quantum intelligent sensor and avoiding affecting the accuracy of the inversion of the underground medium density.
[0063] Further, preferably, in this embodiment, the gravity data sequence of each observation point within the measurement area is used as the input of the DTCWT dual-tree complex wavelet transform to decompose and obtain the wavelet bottom-level low-frequency coefficients and wavelet high-frequency coefficients. The wavelet high-frequency coefficients are then used as the input of the wavelet denoising algorithm. Specifically, the wavelet adjustment threshold is selected as the wavelet threshold for the wavelet denoising algorithm, and a soft thresholding method is used to obtain the threshold-processed wavelet high-frequency coefficients. The threshold-processed wavelet high-frequency coefficients and the wavelet bottom-level low-frequency coefficients are used as the input of the dual-tree complex wavelet inverse transform. The signal is reconstructed through the dual-tree complex wavelet inverse transform to obtain the denoised gravity data sequence of each observation point. The DTCWT dual-tree complex wavelet transform and the wavelet denoising algorithm are both well-known technologies and will not be described in detail.
[0064] Step 4: Correct the gravity data of each observation point after denoising to extract the Bouguer gravity anomaly of each observation point, and obtain the inversion result of the subsurface medium density through the Parker-Oldenburg density interface inversion algorithm.
[0065] Furthermore, the density of the underground medium in the measurement area is inverted using the gravity data sequence of each observation point after noise reduction processing. Preferably, in this embodiment, the specific process is as follows:
[0066] By performing dimensional correction, topographic correction, and Bouguer correction on the gravity data within the gravity data sequence, the Bouguer gravity anomaly of all observation points in the surface measurement area is obtained. This Bouguer gravity anomaly is then used as input to the Parker-Oldenburg density interface inversion algorithm. The algorithm uses the Parker-Oldenburg formula to invert the subsurface medium density at all observation points. The algorithm outputs the inversion result of the subsurface medium density within the surface measurement area. Dimensional correction, topographic correction, Bouguer correction, and the Parker-Oldenburg density interface inversion algorithm are all well-known techniques and will not be elaborated upon further.
[0067] Based on the same inventive concept as the above method, this application also provides a high-precision microgravity-based underground medium density inversion system, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described high-precision microgravity-based underground medium density inversion methods.
[0068] Preferably, in this embodiment, the subsurface medium density inversion system includes a sensor acquisition module, a superposition anomaly analysis module, a wavelet denoising module, and a density inversion module. The sensor acquisition module collects gravity data of the measurement area using a quantum intelligent sensor and preprocesses the collected gravity data using the intelligent sensor's processing function to obtain a gravity data sequence. The superposition anomaly analysis module extracts superposition anomalies of the gravity subsequence at different time periods for each observation point using frequency domain analysis. The wavelet denoising module improves the wavelet threshold to obtain a wavelet adjustment threshold and performs wavelet denoising processing on the gravity data using DTCWT dual-tree complex wavelet transform. The density inversion module inverts the subsurface medium density of the measurement area using the Parker-Oldenburg density interface inversion algorithm. It should be noted that each module in the subsurface medium density inversion system is implemented through the steps and processes of the high-precision microgravity-based subsurface medium density inversion method.
[0069] It is understood that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.
[0070] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
[0071] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the protection scope of this application.
Claims
1. A method for inverting the density of underground media based on high-precision microgravity, characterized in that, Includes the following steps: Obtain gravity data at each observation point within the surface measurement area for underground medium density inversion; Gravity subsequences at each observation point are extracted and frequency domain analysis is performed. Based on the degree of difference between the instantaneous amplitude of the gravity subsequence in the frequency domain and the corresponding amplitude in the other frequency domain, as well as the discrepancy anomaly, the superposition anomaly value of each gravity subsequence at each observation point is obtained. Mode decomposition is performed on each gravity subsequence, and the superposition feature value of each mode component is extracted by the random variation of each mode component. Based on the differences between the superposition feature values of each gravity subsequence at different observation points and the deviation of superposition outliers, the superposition interference confidence of each gravity subsequence at each observation point is obtained. By the change of the superposition interference confidence of all gravity subsequences at each observation point, the superposition influence duration at each observation point is obtained. This is used to adjust the wavelet threshold for wavelet transform of gravity data at each observation point to denoise the gravity data. The gravity data of each observation point after denoising were corrected to extract the Bouguer gravity anomaly of each observation point, and the inversion result of the subsurface medium density was obtained by the Parker-Oldenburg density interface inversion algorithm.
2. The method for inverting the density of underground media based on high-precision microgravity as described in claim 1, characterized in that, The gravity data of each observation point are arranged in chronological order to form a gravity data sequence for each observation point, and then evenly divided into multiple subsequences as gravity subsequences for each observation point.
3. The method for inverting the density of underground media based on high-precision microgravity as described in claim 1, characterized in that, The method for obtaining the superposition outlier values of each gravity subsequence at each observation point is as follows: In the formula, This represents the superposition anomaly value of the k-th gravity subsequence at the current observation point. This is the first mean of the k-th gravity subsequence at the current observation point. This is the first sum of the k-th gravity subsequence at the current observation point.
4. The method for inverting the density of underground media based on high-precision microgravity as described in claim 3, characterized in that, For the current observation point, frequency domain transformation is performed on each gravity subsequence to obtain the amplitude corresponding to all frequencies within the gravity subsequence. Hilbert-Huang transform is then performed on each gravity subsequence to extract the instantaneous amplitude. The mean of the absolute differences between the instantaneous amplitude of each gravity subsequence and the amplitude corresponding to all frequencies is taken as the first mean of each gravity subsequence.
5. The method for inverting the density of underground media based on high-precision microgravity as described in claim 4, characterized in that, Anomaly detection is performed on the absolute differences corresponding to each gravity subsequence, and the sum of the local outlier factors of all absolute differences is taken as the first sum of each gravity subsequence.
6. The method for inverting the density of underground media based on high-precision microgravity as described in claim 1, characterized in that, The method for extracting the superposition feature values of each modal component is as follows: Calculate the first-order difference vector of each modal component and the permutation entropy of each modal component. The normalized result of the product of the mean of the absolute values of all elements in the first-order difference vector and the permutation entropy is used as the superposition eigenvalue of each modal component.
7. The method for inverting the density of underground media based on high-precision microgravity as described in claim 3, characterized in that, The method for obtaining the superposition interference confidence of each gravity subsequence at each observation point is as follows: In the formula, The confidence level of the superposition interference of the k-th gravity subsequence at the current observation point. This represents the degree of difference between the superimposed eigenvectors of the current observation point and its j-th neighboring observation point in the k-th gravity subsequence. This is the absolute difference between the superposition anomaly of the current observation point and its j-th neighboring observation point in the k-th gravity subsequence. To avoid constants with a denominator of 0, the multiple observation points that are closest to the current observation point are all recorded as adjacent observation points of the current observation point.
8. The method for inverting the density of underground media based on high-precision microgravity as described in claim 1, characterized in that, The method for obtaining the duration of the superimposed influence of each observation point is as follows: The superposition interference confidence scores of all gravity subsequences at each observation point are arranged in time sequence to obtain the superposition confidence score sequence of each observation point. The ratio between the mean of all elements in the superposition confidence score sequence and the mean of the absolute values of all elements in the first difference sequence of the superposition confidence score sequence is taken as the superposition influence duration of each observation point.
9. The method for inverting the density of underground media based on high-precision microgravity as described in claim 2, characterized in that, The method for adjusting the wavelet threshold for wavelet transform of gravity data at each observation point is as follows: In the formula, The wavelet adjustment threshold for the gravity data at the current observation point. This represents the standard deviation of the noise estimated in the wavelet denoising algorithm. The length of the gravity data sequence at the current observation point. For adjustment coefficients, It is the hyperbolic tangent function. This represents the duration of the superimposed influence at the current observation point.
10. A high-precision microgravity-based underground medium density inversion system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the high-precision microgravity-based underground medium density inversion method as described in any one of claims 1-9.
Citation Information
Patent Citations
Joint noise reduction method based on variational mode decomposition and permutation entropy
CN110659621A
Residual gravity anomalous field separation method
CN113158830A