Three-dimensional gravity gradient inversion method for oil and gas resource exploration, medium and equipment

Abstract

The invention relates to the technical field of oil and gas resource exploration data processing, in particular to a three-dimensional gravity gradient inversion method for oil and gas resource exploration, a medium and equipment. The method comprises the following steps: acquiring gravity gradient plane two-dimensional observation data of an exploration target area; setting an inversion constraint function; and performing inversion iterative operation to obtain an underground three-dimensional density structure, and delineating a space boundary of a salt hill, a rock cover or an oil storage structure according to the structure. According to the method, a regularization objective function is constructed based on the L1 norm of the model and the total variation TV1 of the model, an L1-TV1 weight factor is introduced and reasonably selected, and a density model vector after depth weighting is used in the total variation of the model. The method can effectively overcome the shielding effect of the shallow rock cover on the deep salt nucleus in salt hill detection, improve the depth resolution and the boundary depiction definition, reduce the inversion multiplicity of solutions, and provide accurate guidance for oil-gas exploration and drilling risk avoidance.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of oil and gas resource exploration data processing, and particularly relates to a three-dimensional gravity gradient inversion method for oil and gas resource exploration, a medium and equipment. BACKGROUND

[0002] Gravity gradient exploration, as an important part of geophysical exploration, has the characteristics of being sensitive to the boundary of shallow abnormal bodies and high resolution, and is widely used in the study of geological structure and oil and gas resource exploration in oil and gas basins. The basic principle is to infer the density distribution difference of underground matter by measuring the spatial variation of the earth's gravity gradient field, so as to reveal the geological structure. In the field of oil and gas exploration, salt dome structure is an important type of oil and gas trap. A typical salt dome structure is usually composed of a deep large-scale low-density salt core (mainly composed of salt rock) and a shallow high-density rock cover (mainly composed of limestone, gypsum, etc.) covering the top of the salt core. Due to the strong gravity gradient signal generated by the shallow high-density rock cover, it often has a "shielding effect" or "false extreme value interference" on the deep low-density salt core.

[0003] The existing three-dimensional gravity gradient inversion technology mainly focuses on solving the core problem of multi-solution inversion, and develops in several core directions such as regularization constraint, structural constraint, statistical learning and fast algorithm. These methods have different theoretical basis, implementation strategy and application scene, and together constitute the current gravity gradient inversion technology system, as follows: (1) Joint inversion method based on cross-gradient structural constraint Cross-gradient structural constraint joint inversion is an effective method developed in recent years, which does not depend on the rock physical property relationship, but uses the structural similarity between different physical models as a constraint condition, including: ① Smooth focusing structural constraint: a smooth focusing structural constraint joint inversion method for three-dimensional gravity, gravity gradient and magnetotelluric data is proposed, which combines the advantages of smooth and focusing regularization methods. It can not only ensure the smoothness of the abnormal body profile on a large scale, reduce local false anomalies, but also capture small-scale details of the inversion solution, enhance the sparsity of the inversion result, and obtain sharp abnormal body boundaries. By setting a focusing intensity threshold, this method avoids the problem of excessive focusing. In addition, a new strategy of discontinuous joint cross-gradient joint inversion is proposed, which alternates between separate inversion and joint inversion, ensuring the structural similarity constraint of cross-gradient method and allowing the iterative model response to fully fit the observation data.

[0004] ②, Joint inversion of cross-gradient and re-shock: Invention patent ZL202111234629.9 discloses a joint inversion method, system and electronic equipment based on cross-gradient function, including: based on different sources of geophysical data, the underground medium is meshed to obtain an initial physical property model; calculate the cross-gradient function between different physical property models; separately invert the geophysical data to obtain model updates; combine the separate inversion physical property model updates and the cross-gradient function and solve the linear inversion equation set; determine whether the solution meets the convergence criteria, if it meets, output the joint inversion physical property model updates, if it does not meet, update the physical property model using the joint inversion physical property model updates and repeat the above steps. By constraining different physical property models with structural similarity, the model updates from separate inversion are combined with the cross-gradient function, effectively reducing the multi-solution of joint inversion, improving the resolution of joint inversion, and obtaining more accurate underground medium physical property distribution results.

[0005] (2) Joint inversion method based on geostatistics The geostatistics method uses statistical concepts and tools to integrate prior geological information and spatial variability into the inversion process, improving the reliability and accuracy of inversion. Mainly includes the following: ①, Collaborative Kriging technology: The gravity gradient data three-dimensional inversion method based on geostatistics uses collaborative Kriging technology to establish a collaborative Kriging equation for density value estimation with underground space density value as the main variable and gravity gradient data as the secondary variable. This method introduces a sensitivity matrix to offset the rapid decay of the kernel function with increasing depth, and sets a threshold to compress the density covariance matrix in the collaborative Kriging equation into a sparse matrix for storage and calculation, greatly reducing memory consumption and improving calculation speed.

[0006] ②, Random simulation technology: Based on the collaborative Kriging inversion model, random inversion of gravity gradient data can also be achieved. The random inversion method can obtain various equally possible realizations from the probability model, which can satisfy the known observation data and the geological understanding contained in the covariance matrix. The differences between a large number of realizations can reflect the heterogeneity and uncertainty of the spatial distribution of geological properties.

[0007] The main shortcomings of the prior art can be summarized as follows: First, the resolution of the inversion result is low, and the geological interpretation is poor, specifically: existing methods (such as smoothing constraint inversion based on L2 norm) tend to produce overly smoothed and fuzzy boundary models. The results cannot clearly depict the sharp boundaries of geological units such as ore bodies and faults, like "cloud clusters", which severely depend on the subjective experience of interpreters, reducing the geological reliability and practical value of the inversion results.

[0008] Second, the calculation efficiency is low, and it is difficult to deal with large-scale problems. Specifically, traditional algorithms (such as direct matrix inversion or conventional iterative method) have the bottleneck of long calculation time and huge memory demand when dealing with massive elements (several hundred thousand to several million) involved in three-dimensional inversion. This makes it difficult to carry out fast and large-scale fine inversion in actual production, limiting its application efficiency.

[0009] The ability to introduce prior information is weak, and the flexibility is insufficient. Specifically, the existing method system performs poorly in flexibly and effectively integrating diversified prior geological and geophysical constraints (such as known drilling information, seismic interpretation results, or sparse constraints pursuing block structure). The algorithm structure is rigid, and it is difficult to handle data fitting and complex constraints simultaneously, resulting in low consistency between the inversion results and known geological knowledge, and unable to effectively reduce the multi-solution. However, existing three-dimensional gravity gradient inversion methods (such as traditional L2 norm smoothing inversion) face serious technical bottlenecks in dealing with such complex structures, such as low depth resolution and unclear boundary. They often have difficulty in effectively distinguishing between shallow rock cover and deep salt core, resulting in blurred salt dome shape and inaccurate depth extension of the inversion, which cannot provide accurate guidance for drilling deployment of deep subsalt hydrocarbon reservoirs. Therefore, there is an urgent need for a three-dimensional gravity gradient inversion method that can balance depth resolution and boundary clarity, and effectively overcome the shielding interference of rock cover. SUMMARY

[0010] The present application provides a three-dimensional gravity gradient inversion method, medium and equipment for oil and gas resource exploration, aiming to solve the problem of low depth resolution and unclear boundary of existing inversion methods for deep salt core and subsalt structure due to the shielding effect of shallow high-density rock cover in oil and gas structure (especially salt dome structure) detection.

[0011] To achieve the above purpose, the implementation of the present application includes the following steps: Step 1, obtaining gravity gradient of the study area Planar two-dimensional data; Step 2, setting inversion constraint function; Step 3, based on the gravity gradient obtained in step 1 Planar two-dimensional data and the inversion constraint function obtained in step 2, inversion iterative operation is carried out to obtain the three-dimensional density structure of the inversion area; In step 2, setting the inversion constraint function specifically includes: Step 2.1.1, constructing the inversion objective function based on the L1 norm and total variation TV1 of the model As follows: ; Where: Indicates the two-norm; Denotes the norm 1; Represents the regularization factor; This represents the weighted kernel matrix. , This represents the unweighted kernel matrix. Represents the depth-weighted matrix; Represents the inverted density vector. , Represents the density vector that is not depth-weighted; , and They represent direction, direction and Gradient operator for direction; , and Is the total variation of the model in direction, direction and The weight of direction; These are the weights of the model's L1 norm; Step 2.1.2: Discretize the gradient operator of the objective function to obtain a new form of the objective function: ; in: Represents the difference matrix. , It is the identity matrix. for Gradient operator of direction discrete form, for Gradient operator of direction discrete form, for Gradient operator of direction Discrete form; This represents the transpose of a matrix.

[0012] Preferably, the inversion iterative operation includes the following steps: Step 2.2.1: Introduce relaxation vectors and Lagrange vectors The augmented Lagrange function is obtained. ,as follows: ; Step 2.2.2: Solve iteratively using the alternating direction multiplier method; Step 2.2.3: Obtain the inversion density vector. relaxation vector and Lagrange vectors The updated formula is as follows: ; in: , , The first Density vector in the next iteration relaxation vector and Lagrange vectors ; ; This is a soft threshold operator.

[0013] Preferably, the formula for iterative solution using the alternating direction multiplier method in step 2.2.2 is as follows: ; in, Indicates the first Lagrange vector in the next iteration .

[0014] Preferably, it also includes convergence determination, specifically including: Define the original residual and dual residual ,as follows: ; The mean norm of the residuals is calculated using the following formula. : ; To make a judgment, specifically: if Then the iteration accuracy requirement is met; where: Indicates the preset iteration precision; This indicates the length of the vector.

[0015] The present invention also discloses a computer storage medium storing instructions, wherein when a computer reads the instructions, the computer executes the three-dimensional gravity gradient inversion method described above.

[0016] The present invention also discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it causes the computer to perform the three-dimensional gravity gradient inversion method as described above.

[0017] The effect of applying the technical solution of this invention is: 1. Effectively overcomes strong interference and accurately identifies deep subsalt structures: For typical salt dome structures in oil and gas exploration, this invention utilizes the joint constraints of the L1 norm and the model's total variation (TV1) to effectively distinguish between the strong gravity gradient signal generated by shallow high-density cap rock and the weak signal generated by deep low-density salt cores. Compared to traditional methods, this invention successfully eliminates the shielding effect of cap rock on deep structures and the interference of false extrema, thus clearly reproducing the spatial morphology and bottom depth of deep salt cores.

[0018] 2. Significantly improves the resolution of geological boundaries and reduces drilling risks: Traditional smooth inversion (L2 norm) yields anomalous body boundaries that are blurry, making it difficult to determine the exact location of reservoirs. This invention, through sparse constraints and total variational constraints, can recover the steep boundaries and blocky structure of subsurface geological bodies (such as the flanks of salt domes), which highly matches the actual geological characteristics of salt intrusions. This high-precision boundary characterization capability is expected to provide reliable geophysical evidence for the identification of oil and gas traps and the precise deployment of drilling well locations, thereby significantly reducing exploration and development risks. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the three-dimensional gravity gradient inversion method based on L1-TV1 in an embodiment of the present invention; Figure 2 The salt dome model and gravity gradient components in this embodiment of the invention. Contour map, where: (a) is a schematic diagram of the salt dome model; (b) is the gravity gradient component. Contour map, where dashed lines indicate the location of the profile; Figure 3 To compare the inversion results of the objective functions (a)-(c) of previous researchers with the objective functions (e)-(f) in this embodiment of the invention, where: (a) and (d) are three-dimensional plots of the inversion results; (b) and (e) are plots along AA' ( in (a) and (d)). x =-0.2 km) The results of the profile, (c) and (f) are respectively along BB' ( =-0.2 km) in (a) and (d). y =-0.2 km) results of the profile; the black line represents the actual model boundary.

[0021] Figure 4Gravity gradient data after correction for salt dome topography in a certain area Where: (a) is the gravity gradient data after salt dome topography correction in a certain area, and (b) is the gravity gradient of the rectangular area in (a) after curvature leveling (observation height 85m). (a) Data; (b) The black dashed lines AA' and BB' in the figure are cross-sectional lines, representing Figure 5 The location of the cross section.

[0022] Figure 5 The image shows a reconstructed underground density model of a salt dome in a certain area, where: (a) is a 3D model of the underground density of the area, and (b) is a model of the underground density along AA' in (a). x =442.5 km) The result diagram of the profile, (c) is along BB' ( in (a) y = 3334.3 km) Result diagram of the profile.

[0023] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0025] A three-dimensional gravity gradient inversion method based on L1-TV1 is detailed in [link to documentation]. Figure 1 This includes: obtaining gravity observation data of the study area based on numerical simulation / field measurement; constructing an L1-TV1 regularized objective function, solving its optimization problem, and obtaining the inversion formula; and obtaining the inversion results of the gravity gradient observation data based on the inversion formula. This embodiment uses variables... ( Taking the density of underground structures as an example, the specific steps include: Step 1: Obtain the gravity gradient of the study area Two-dimensional planar data. This data can be obtained through numerical simulation or through field measurements.

[0026] Step 2: Define the inversion constraint function; Step 3: Based on the gravity gradient obtained in Step 1 The planar two-dimensional data and the inversion constraint function obtained in step two are used for inversion iterative calculation to obtain the three-dimensional density structure of the inversion region.

[0027] In this preferred embodiment, the setting of the inversion constraint function specifically includes: Step 2.1.1: Construct the inversion objective function based on the L1 norm and total variation TV1 of the model. ,as follows: ; in: Represents the L2 norm; Denotes the norm 1; Represents the regularization factor; This represents the weighted kernel matrix. , This represents the unweighted kernel matrix. Represents the depth-weighted matrix; This represents the inverted density vector, which indicates the density distribution underground. It is also a diagonal matrix, with the diagonal elements being depth-weighted functions. , Represents the density vector that is not depth-weighted; , and They represent direction, direction and Gradient operator for direction; , and Is the total variation of the model in direction, direction and The weight of direction; These are the weights of the model's L1 norm; Step 2.1.2: Discretize the gradient operator of the objective function to obtain a new form of the objective function: ; in: Represents the difference matrix. , It is the identity matrix. for Gradient operator of direction discrete form, for Gradient operator of direction discrete form, for Gradient operator of direction Discrete form; This represents the transpose of a matrix.

[0028] Preferably, to solve the optimization problem of the objective function, this invention introduces the Alternating Direction Multiplier Method (ADMM) to solve the optimization problem of gravity gradient inversion. The characteristic of the Alternating Direction Multiplier Method (ADMM) is that in each step, only one variable is updated while the other two variables are fixed, and this alternating updating is repeated until convergence is achieved. The following is a brief explanation of the solution process of the Alternating Direction Multiplier Method (ADMM). The inversion iterative calculation includes the following steps: Step 2.2.1: Introduce relaxation vectors and Lagrange vectors The augmented Lagrange function is obtained. ,as follows: ; Step 2.2.2: Solve iteratively using the alternating direction multiplier method. The formula for solving iteratively using the alternating direction multiplier method is as follows: ; in, Indicates the first Lagrange vector in the next iteration .

[0029] Step 2.2.3: Obtain the inversion density vector. relaxation vector and Lagrange vectors The updated formula is as follows: ; Among them: Among them: , , The first Density vector in the next iteration relaxation vector and Lagrange vectors ; ; This is a soft threshold operator.

[0030] In this embodiment, to determine the convergence status during the ADMM iteration process, a convergence determination is also included, specifically comprising the following steps: Define the original residual and dual residual ,as follows: ; The mean norm of the residuals is calculated using the following formula. : ; Using the mean norm of the residuals ( To determine the convergence of the ADMM iteration, specifically: if Then the iteration accuracy requirement is met; where: Indicates the preset iteration precision; This indicates the length of the vector.

[0031] The key aspects of this embodiment are: ① Constructing a Tikhonov regularized objective function based on the L1 norm and total variation (TV1) of the model for inverting gravity gradient data: the L1 norm ensures that the inversion results have high resolution, and the total variation (TV1) of the model can obtain results with clear boundaries, thereby achieving high-resolution and clear-boundary inversion imaging of underground density structures; ② Further improving the depth resolution of the inversion by introducing and reasonably selecting L1-TV1 weighting factors; ③ Proposing to use the depth-weighted density model vector in the total variation (TV1) instead of the real density model to alleviate the over-focusing problem in L1-TV1 inversion.

[0032] The present invention also discloses a computer storage medium storing instructions, wherein when a computer reads the instructions, the computer executes the three-dimensional gravity gradient inversion method described above.

[0033] The present invention also discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it causes the computer to perform the three-dimensional gravity gradient inversion method as described above.

[0034] This embodiment employs an objective function constructed based on the L1 norm and the total variation (TV1) of the model (this objective function was proposed by Utsugi; see Mitsuru Utsugi, Magnetic inversion to recover the subsurface block structures based on...). L 1 norm and total variation regularization, Geophysical Journal International (2022, 228(1): 510–537) Applied to magnetic field data inversion.

[0035] ; In the formula, For observation data, Represents the inverted susceptibility vector ( ), representing the underground magnetic susceptibility distribution, This represents a depth-weighted matrix, and it is also a diagonal matrix, with the diagonal elements being the depth-weighting function; Inversion vector The forward modeling operation is performed to obtain the forward modeling response of the inverted region.

[0036] Despite obtaining inversion results with clear boundaries and high resolution, this method still suffers from insufficient depth resolution for geological bodies at greater depths.

[0037] To compare this technique with the present invention, the method in this embodiment is improved based on the above objective function to make it applicable to the inversion of gravity gradient data, resulting in the following equation: .

[0038] The rewritten inversion objective function is then inverted and compared with the technical solution of this embodiment, as follows: Design as Figure 2 The salt dome model shown in (a) is an example. Salt domes are favorable geological structures associated with oil and gas reservoirs, typically containing key mineral resources such as potash, gypsum, and sulfur. Furthermore, salt domes can also serve as underground storage tanks for liquefied gas and radioactive materials. Therefore, probing salt dome structures is of great significance. The model contains a shallow, high-density rock cap with its geometric center located at (-0.2 km, -0.2 km, 0.8 km) and dimensions of 1.2 × 1.2 × 0.4 m. 3 The remaining density is 0.5 g / cm³. 3 In addition, it includes a deep, low-density salt body, geometrically centered at (-0.2 km, 0.0 km, 5.1 km), with dimensions of 4.4 × 6.4 × 3.0 km. 3 The remaining density is -0.5 g / cm³. 3 The extent of the underground source area ( , , The range is defined as (-10.0km, -10.0km, 0km) to (10.0km, 10.0km, 8.0km), and the source region is divided into 50×50×20 regular prism elements, with each element having a size of 0.4×0.4×0.4km. 3 The observation surface is located on the Earth's surface (at an altitude of 0m) and consists of 50×50 observation points. To simulate instrument random errors and environmental disturbances, the gravity gradient obtained through forward modeling is used... The observed data were obtained by adding Gaussian noise with a standard deviation of 2% of the absolute value of the data, such as... Figure 2 As shown in (b).

[0039] In the inversion, this embodiment sets the maximum number of ADMM inversion iterations to 1000, and the accuracy of ADMM is... 1.0×10 -5 The accuracy during the CG iterative update process is 1.0 × 10⁻⁶. -3 .

[0040] First, the salt dome model is inverted using the objective function proposed by previous researchers. The inversion results are as follows: Figure 3 Images (a)-(c) show that the inverted shallow sluice block exhibits over-focusing, with a residual density of 2.31 g / cm³. 3 It deviates significantly from the true amplitude (0.5 g / cm). 3 For inverting deep anomalies, overstretching occurs, sometimes even extending to the bottom of the inversion region. Overall, this inversion result indicates that the objective function cannot effectively invert the salt dome model, and the inversion result suffers from low depth resolution.

[0041] The objective function of this invention is used to invert the intrusion model. Through extensive testing, we found that adjusting the L1-TV1 weights in the objective function (…) When the value is set to (1.0, 3.0, 3.0, 0.6), by appropriately increasing... , The weight value of the directional variation penalty is reduced, and the vertical smoothing penalty is reduced relatively. The contribution of this method allows for the preservation of sharper physical property boundaries in deep anomalies, ultimately resulting in inversion results with high depth resolution. Furthermore, the improvement by using a weighted density model vector instead of the true density model in the total variation (TV1) effectively alleviates the over-focusing of shallow anomalies, and their density is reasonably recovered (0.34 g / cm³). 3 (It is closer to the true value). Figure 3 As shown in (d)-(f), the location and depth of the recovered magnetization intensity distribution are consistent with the real model, and the boundaries are clear. This model experiment demonstrates that the method of this invention can completely invert the magma intrusion model. The location and amplitude of the shallow caprock match the real model, and the top surface burial depth (3.6 km) and bottom interface depth (6.6 km) of the deep low-density salt body are accurately delineated.

[0042] The method described in this embodiment can be used to invert salt dome areas using actual geological data, thereby enabling the exploration of underground structures in these areas.

[0043] Salt domes are diapiric structures widely distributed in major saline basins worldwide. These structures are not only favorable locations for oil and gas accumulation, but also often contain key mineral resources such as potash, gypsum, and sulfur. They can even be used as underground storage depots for liquefied gas and radioactive waste. Therefore, accurate exploration of their underground structures has extremely high economic value and engineering significance.

[0044] However, gravity exploration of salt dome structures faces significant challenges. Typically, salt diapirs develop within sedimentary strata and exhibit typical vertical density variations: the top cap rock, due to dehydration, forms high-density gypsum or anhydrite, showing a "positive density contrast" relative to the surrounding rock; while the underlying deep salt core usually maintains a constant density and is lower than the compacted surrounding rock, exhibiting a "negative density contrast." This complex vertical density structure, with "positive above and negative below," coupled with the gravity response generated by fluctuations in the background sedimentary strata, makes it difficult for conventional inversion methods to effectively separate strong shallow interferences from weak deep signals, easily leading to misjudgments of deep subsalt structures.

[0045] To verify the effectiveness of the method of the present invention in solving the aforementioned complex salt dome detection problem, this embodiment selects a salt dome located in a certain area as the research object. Based on known geological drilling and geophysical data, this salt dome fully conforms to the aforementioned typical characteristics: it consists of a deep, large-scale, low-density salt core (salt rock) and a shallow, high-density rock cap covering it. The gravity gradient data used in this embodiment has undergone denoising and topographic correction processing, and the gravity gradient g containing the core region of the salt dome was selected. zz Inversion is performed on the component data.

[0046] Inversion parameter settings: The underground space is divided into grid cells, and the inversion depth is set to 8000 meters below the surface. The L1-TV1-based regularized inversion method proposed in this invention is adopted, the L1-TV1 weight factor is set, and the ADMM algorithm is used for solving.

[0047] Geological interpretation and exploration value: such as Figure 4 As shown, the inversion results clearly reveal the geological bodies closely related to hydrocarbon accumulation: (1) Shallow rock cap identification: At a depth of approximately 300 to 700 meters, a high-density positive anomaly was identified, corresponding to a rock cap composed of limestone / gypsum, with a shape that is thicker in the west and thinner in the east. (2) Deep salt core characterization: Below the rock cap, at a depth of approximately 2300 to 4900 meters, a low-density salt core was clearly identified, exhibiting an asymmetrical columnar shape. (3) Technical effect: Compared with existing technologies, the method in this embodiment successfully distinguishes between the overlapping high-density rock cap and the low-density salt core, eliminates the interference of "false extreme values," and accurately delineates the steep boundaries of the salt dome's flanks. This provides accurate geophysical evidence for deploying oil and gas drilling wells in this area and avoiding drilling risks, solving the technical problem of difficult imaging of complex salt dome structures.

[0048] Consistency verification with previous studies: The density structure characteristics obtained by inversion are in high agreement with existing geological and geophysical research results in this region, further verifying the reliability of this method. Specifically: 1) Consistent with the structural trend of the rock cap: The inversion results show that the rock cap has lateral heterogeneity, which is thicker on the west and north sides and gradually thins towards the east and south sides. This geometric feature is consistent with the structural trend map drawn by previous researchers (reference: C. Ennen, “Mapping gas-charged fault blocks around the Vinton salt dome, Louisiana using gravitygradiometry data,” MS thesis, Univ. Houston, Houston, TX, USA, 2012). 2) Good agreement on physical properties: The density difference of the rock cap estimated by inversion is about 0.5~0.73 g / cm3, which is basically consistent with the rock physical property values ​​in the geological report of this area (reference: C. Ennen, “Mapping gas-charged faultblocks around the Vinton salt dome, Louisiana using gravity gradiometry data,” MS thesis, Univ. Houston, Houston, TX, USA, 2012). 3) Deep structural compatibility: The depth of the deep salt core bottom boundary determined by the inversion (approximately 4.9 km) is compatible with the local seismic velocity analysis results (inferred salt bottom depth of approximately 4 km) (reference: C. Ennen, “Mapping gas-charged faultblocks around the Vinton salt dome, Louisiana using gravity gradiometry data,” MS thesis, Univ. Houston, Houston, TX, USA, 2012) and the range of the inversion model of Chen et al. (3.5-5 km) (reference: B. Chen, G. Qi, J. Du, S. Li and Y. Sun, "3-D Gravity Anomaly Inversion for Imaging Salt Structures, With Application to Vinton Salt Dome, Gulf of Mexico," in IEEE Transactions on Geoscience and Remote Sensing, 2023, 61(1): 1-9).

[0049] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A three-dimensional gravity gradient inversion method for oil and gas resource exploration, characterized in that, The method comprises the following steps: Step one, obtaining the gravity gradient of the geological area to be surveyed planar two-dimensional data; Step two, setting the inversion constraint function; Step three, the gravity gradient based on the gravity gradient obtained in step one The plane two-dimensional data and the inversion constraint function obtained in step two are subjected to iterative inversion operation to obtain the three-dimensional density structure of the inversion region. In step two, setting the inversion constraint function specifically comprises: Step 2.1.

1. Constructing the inversion objective function based on the model-based L1 norm and total variation TV1 As follows: ； where: denotes the two-norm; denotes the one-norm; denotes the regularization factor; denotes the weighted kernel matrix, , denotes the unweighted kernel matrix, denotes the depth weighting matrix; denotes the inverted density vector, , denotes the un-depth weighted density vector; , and denote the gradient operators in direction, direction and direction, respectively; , and are the weights of the model total variation in direction, direction and direction, respectively; is the weight of the L1 norm of the model; Step 2.1.2, discretize the gradient operator of the objective function to obtain a new form of the objective function: ； wherein: represents a difference matrix, , is an identity matrix, is a gradient operator in the direction a discrete form of is a gradient operator in the direction a discrete form of is a gradient operator in the direction a discrete form of denotes the transpose of a matrix.

2. The method for 3D gravity gradient inversion for oil and gas resource exploration according to claim 1, characterized in that, The inversion iterative operation comprises the following steps: Step 2.2.1, Introducing the relaxation vector and the Lagrangian vector to obtain the augmented Lagrangian function as follows: ； Step 2.2.2, iterative solution by using the alternating direction multiplier method; Step 2.2.3, obtain updates for the density vector , the relaxation vector , and the Lagrange vector , as follows: ； in: , , The first Density vector in the next iteration relaxation vector and Lagrange vectors ; ; This is a soft threshold operator.

3. The three-dimensional gravity gradient inversion method of claim 2, wherein, The formula for iterative solution by using the alternating direction multiplier method in step 2.2.2 is as follows: ； wherein, represents the Lagrange vector in the th iteration.

4. The three-dimensional gravity gradient inversion method of claim 3, wherein, Further comprising convergence condition determination, specifically comprising: defining original residuals and dual residuals as follows: ； The average norm of the residuals is calculated using the following formula : ； The judgment is specifically: if the iteration accuracy requirement is met; wherein: represents a preset iteration accuracy; represents the length of the vector.

5. A computer storage medium, characterized in that, The storage medium has instructions stored therein, and when the computer reads the instructions, the computer executes the three-dimensional gravity gradient inversion method according to any one of claims 1 to 4.

6. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program, and the computer executes the three-dimensional gravity gradient inversion method according to any one of claims 1 to 4.

Images