Gravity anomaly body inversion method based on PMU-net deep learning network

Abstract

The application discloses a gravity anomaly body inversion method based on a PMU-Net deep learning network, belongs to the field of gravity inversion, and comprises the following steps: two kinds of different gravity anomaly data are inversely calculated in combination, the number of observation equation groups is increased, and underground geological structures are comprehensively presented. Data is processed by means of two independent channels, one channel uses a Mamba module to capture the dependence between observation points, and the other channel uses a spatial pre-training model to capture the position correlation between observation points and underground density. The processed features are fused and input into a U-shaped neural network constructed by one-dimensional convolution, so that the calculation amount is reduced and the efficiency is improved. The application further comprises a simulation data generation module, and the weak problem of weak interpretability of deep learning in gravity inversion is solved by embedding prior knowledge. The application effectively improves the accuracy and reliability of inversion under complex geological conditions, and provides a new efficient and accurate method for geophysical exploration.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of gravity inversion, and particularly relates to a gravity anomaly body inversion method based on a PMU-Net deep learning network. BACKGROUND

[0002] Gravity inversion is a key technology in geophysical exploration, which measures the gravity field on the ground and deduces the underground density distribution by means of algorithm model, and has the significant advantages of low cost and high efficiency. In recent years, with the development of artificial intelligence technology, people have begun to use deep learning to predict gravity data to quickly image the underground geological structure. The early method mainly uses single gravity field data as input to directly predict the underground density body, but this way can only reflect the gravity anomaly of the underground geological body in a particular aspect, the information is extremely one-sided, and the uncertainty of the inversion solution is high.

[0003] To improve this situation, some people have designed a GraInvNet model based on U-Net, which uses gravity anomaly data and gravity gradient anomaly data for joint inversion, and has achieved good prediction results. However, this method is relatively simple in processing different gravity anomaly data, and it is difficult to capture the complex features and mutual relationships between data. Later, some people predicted different features through different decoders to further improve the information capture ability, but there are still problems of insufficient capture of the dependence between observation points and the position correlation between observation points and underground density model. Some people have proposed the fusion of UNet++ network architecture with N×N convolution kernel and LeakyReLU activation function to improve the accuracy of the inversion result, but the high computational complexity of two-dimensional convolution limits the efficiency of the model.

[0004] In summary, the existing deep learning model still has deficiencies in comprehensively describing the underground geological structure, mining the correlation between different data, learning the relationship between observation points and density grid, and computational efficiency, and needs to be improved to meet the needs of high-precision and high-efficiency gravity inversion under complex geological conditions. SUMMARY

[0005] The application provides a gravity anomaly body inversion method based on a PMU-Net deep learning network to solve the problems in the prior art.

[0006] To achieve the above purpose, the application provides a gravity anomaly body inversion method based on a PMU-Net deep learning network, which comprises the following steps:

[0007] Obtain two different types of gravity anomaly data as input;

[0008] Two kinds of gravity anomaly data are processed through two independent channels to mine the correlation and complementarity between the data, one channel uses a Mamba module to capture the dependence between observation points, and the other channel captures the position correlation between observation points and underground density through a spatial pre-training model;

[0009] The processed features are fused, and the fused features are input into a U-shaped neural network constructed by one-dimensional convolution for inversion to obtain the underground density distribution result.

[0010] Optionally, the two different types of gravity anomaly data include gravity anomaly data and gravity gradient anomaly data.

[0011] Optionally, the Mamba module includes:

[0012] The relative coordinate coding is performed on each observation point of the gravity anomaly data to construct a spatial correlation matrix;

[0013] The spatial pattern features of the gravity anomaly data are extracted through a multi-scale convolution kernel;

[0014] Based on a dynamic state space model, a measurement point evolution sequence is constructed based on a plane coordinate grid index, and the physical state of each measurement point is evolved according to the law of conservation of mass;

[0015] The gravity anomaly signal is mapped through an underground physical property structure response matrix, and the mapping result is coupled with the initial geological state through a state transition matrix;

[0016] The updated geological state is projected to the detectable physical quantity space through an observation operator, and a lithology-density conversion matrix is introduced to retain the direct influence term of the original input signal;

[0017] Through a double-path fusion mechanism, the tectonic evolution dynamics process and the density physical static distribution are cooperatively constrained.

[0018] Optionally, the spatial pre-training model includes:

[0019] A physical correlation framework of gravity field features and inversion space is constructed to perform cross-domain feature coupling of observation signals and underground structures;

[0020] Multi-scale local features in the gravity field map are extracted through an F-Encoder to capture the spatial correlation of observation points, and a low-dimensional vector representing the gravity field features is generated;

[0021] The position coordinates of the inversion space are geometrically transformed and aligned through a P-Encoder to generate a high-dimensional vector containing the geometric structure information of the target body;

[0022] The low-dimensional vector and the high-dimensional vector are cooperated, the gradient change caused by the gravity anomaly density difference is processed synchronously, and a field-space feature coupling mechanism is formed by combining the geometric constraint of the inversion space;

[0023] The field-space feature coupling mechanism provides a regularization constraint for the inversion process.

[0024] Optionally, the U-shaped neural network adopts a one-dimensional convolution layer to replace a traditional two-dimensional convolution layer, and is used for processing one-dimensional sequence characteristics of the gravity anomaly data.

[0025] Optionally, the U-shaped neural network further comprises a full-scale skip connection mechanism, which is used for fusing features of different scales in the encoder and the decoder.

[0026] Optionally, the full-scale skip connection mechanism adjusts the features of different scales to the same scale through linear mapping and then fuses the features, and further integrates the features through one-dimensional convolution.

[0027] Optionally, the method further comprises generating simulation training data:

[0028] Initializing the inversion space parameters and the geological body feature parameters;

[0029] Generating a survey network, setting a grid parameter and constructing a structured sampling grid;

[0030] Constructing a density distribution model in the inversion space according to the geological body feature parameters;

[0031] Calculating gravity anomaly data corresponding to the density distribution model according to a forward formula;

[0032] Storing the gravity anomaly data and the corresponding density distribution model to form a training data set.

[0033] The application further provides a computer device, which comprises a memory, a processor and a computer program stored in the memory, and the processor executes the computer program to realize the steps of the method.

[0034] The application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the steps of the method.

[0035] Compared with the prior art, the application has the following advantages and technical effects:

[0036] The gravity anomaly body inversion method based on the space pre-training PMU-Net deep learning network of the present application is significantly superior to the prior art in many aspects: first, in terms of inversion accuracy, by jointly inverting two kinds of gravity anomaly data Gz and Gxz and processing these data through two independent channels respectively, the system can more comprehensively capture the characteristics of the underground geological structure, thereby significantly improving key indicators such as average gravity anomaly offset error and average gravity anomaly accuracy. Secondly, in terms of edge information restoration ability, PMU-Net performs well, effectively avoiding the common edge blur and position offset problems in the prior art. This is mainly due to the introduction of the pre-training model and the attention mechanism, as well as the design of the full-scale skip connection, which enables the system to better capture the dependency between the position of the density anomaly body and the position of the measuring point, while reducing the loss of features in the downsampling process. In addition, in terms of computational efficiency, the U-shaped neural network constructed by one-dimensional convolution significantly reduces the computational complexity and improves the computational efficiency of the model, providing a lightweight solution for fast gravity field inversion. At the same time, through full-scale skip connection, the system can more comprehensively integrate gravity features of different scales, effectively capturing semantic information from fine-grained to coarse-grained, further improving the accuracy of the inversion result. Finally, by embedding prior knowledge in the training data, the system can better combine geophysical constraints and gravity data-driven, further improving the prediction ability of deep learning, effectively solving the problem of weak interpretability of deep learning in gravity inversion. In summary, the PMU-Net deep learning network of the present application has shown significant superiority in gravity anomaly body inversion under complex geological conditions, not only improving the accuracy and reliability of the inversion, but also significantly improving the computational efficiency, providing an efficient and accurate new method for geophysical exploration. BRIEF DESCRIPTION OF DRAWINGS

[0037] The accompanying drawings, which form a part of this application, are included to provide a further understanding of the application and are incorporated in and constitute a part of this application. The embodiments of this application, and of the description of the application, are to explain the application and are not intended to limit the application. In the drawings:

[0038] Figure 1 A feature fusion schematic diagram of an embodiment of the present application;

[0039] Figure 2 A Mamba module schematic diagram of an embodiment of the present application;

[0040] Figure 3 A space pre-training model schematic diagram of an embodiment of the present application;

[0041] Figure 4 A full-scale skip connection schematic diagram of an embodiment of the present application;

[0042] Figure 5Fig. 1 is a comparison chart of inversion results of embodiments of the present application, wherein (a1) is a three-dimensional visual image of a real model, (a2) is an x-z section view of the real model along y=1800 meters, (a3) is a y-z section view of the real model along x=700 meters, (b1) is a three-dimensional visual image of the U-Net inversion result, (b2) is an x-z section view of the U-Net inversion result along y=1800, (b3) is a y-z section view of the U-Net inversion result along x=700, (c1) is a three-dimensional visual image of the TransUnet inversion result, (c2) is an x-z section view of the TransUnet inversion result along y=1800, (c3) is a y-z section view of the TransUnet inversion result along x=700, (d1) is a three-dimensional visual image of the PMU-Net inversion result, (d2) is an x-z section view of the PMU-Net inversion result along y=1800, and (d3) is a y-z section view of the PMU-Net inversion result along x=700;

[0043] Figure 6 Fig. 2 is a flowchart of generating a data set of an embodiment of the present application;

[0044] Figure 7 Fig. 3 is a corresponding schematic diagram of gravity forward modeling of an embodiment of the present application. DETAILED DESCRIPTION

[0045] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments.

[0046] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a group of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in an order different from that shown herein.

[0047] Embodiment one

[0048] In this embodiment, a gravity anomaly body inversion method based on a PMU-Net deep learning network is provided, comprising the following steps:

[0049] Two different types of gravity anomaly data are obtained as input;

[0050] The two types of gravity anomaly data are processed through two independent channels respectively to mine the correlation and complementarity between the data, wherein one channel uses a Mamba module to capture the dependence between observation points, and the other channel captures the positional correlation between observation points and underground density through a spatial pre-training model;

[0051] The processed features are fused, and the fused features are input into a U-shaped neural network constructed by one-dimensional convolution to obtain the underground density distribution result.

[0052] Specifically, the model uses gravity component Gz and gravity gradient component Gxz as the input of the model, and processes two different types of gravity anomaly data through two independent channels respectively. Among them, the gravity anomaly map Gz first solves the long distance dependence problem through relative position coding, and then captures the dependence relationship between each observation point through the Mamba module; the gravity anomaly data Gxz captures the position correlation between the observation points and each density grid of the underground density model by means of the spatial pre-training model. Then, the two types of features processed by the above processing are fused, and then input into a U-shaped neural network constructed by one-dimensional convolution, and full-scale skip connection is used to fuse features of different levels and different scales. As shown in Figure 1 .

[0053] Further, the Mamba module comprises:

[0054] The relative coordinate coding is performed on each measuring point of the gravity anomaly data to construct a spatial correlation matrix;

[0055] The spatial pattern features of the gravity anomaly data are extracted by a multi-scale convolution kernel;

[0056] Based on the dynamic state space model, the measuring point evolution sequence is constructed based on the plane coordinate grid index, and the physical state of each measuring point is evolved according to the law of conservation of mass;

[0057] The gravity anomaly signal is mapped through the underground physical property structure response matrix, and the mapping result is coupled with the initial geological state through the state transition matrix;

[0058] The updated geological state is projected into the detectable physical quantity space through the observation operator, and the lithology-density conversion matrix is introduced to retain the direct influence term of the original input signal;

[0059] Through the double-path fusion mechanism, the tectonic evolution dynamics process and the density physical static distribution are cooperatively constrained.

[0060] Specifically, the Mamba module is as shown in Figure 2 .

[0061] In Figure 2 , the mathematical expression is:

[0062]

[0063] In the formula, h represents the geological state, A represents the state transition matrix, t represents the measuring point, B represents the response matrix, X represents the gravity anomaly signal, Y represents the observation operator, Z represents the lithology-density conversion matrix, and y represents the observation output.

[0064] The embodiment constructs a gravity data feature extraction method based on dynamic state space modeling. A spatial correlation matrix is constructed by encoding the relative coordinates of the gravity data Gz measuring points. A multi-scale convolution kernel corresponding to a typical structure scale of 5-50 km is used to accurately extract the spatial pattern features of regional gravity anomalies. After processing, a feature tensor containing 512-dimensional density physical parameters (batch dimension 16 x measuring point sequence 1681 x feature dimension 512) is formed, which fully represents key geological information such as the development degree of salt structures in the study area.

[0065] In the dynamic modeling stage, the measuring point evolution sequence (t=0-1680) is constructed using a planar coordinate grid index. The physical state evolution of each measuring point strictly follows the law of conservation of mass. Taking measuring point t=1 as an example, its gravity anomaly signal x(1) is mapped through the underground physical structure response matrix B (dimension 1024x512). The matrix elements B ij The sensitivity of the jth density physical parameter to the ith geological structure deformation is quantitatively described. The mapping result is coupled with the initial geological state h(0) (containing regional tectonic stress field background information) through the state transition matrix A (dimension 1024x1024). Among them, the matrix A kl The mutual coupling relationship between the kth geological structure unit and the lth lithofacies density unit caused by geological processes such as fault activity and magma intrusion is described. The updated geological state h(1) is projected to the detectable physical quantity space through the observation operator Y (512x1024). The row vectors of this operator correspond to the gravity response kernel functions of different density interfaces. At the same time, the lithology-density conversion matrix Z (512x512) is introduced to retain the direct influence term of the original input signal x(1). Through the dual-path fusion mechanism, the tectonic evolution dynamics and the static distribution of density physical properties are cooperatively constrained.

[0066] Further, the spatial pre-training model includes:

[0067] A physical association framework of gravity field features and inversion space is constructed to couple the cross-domain features of observation signals and underground structures;

[0068] Multi-scale local features in the gravity field map are extracted by F-Encoder and the spatial correlation of observation points is captured to generate a low-dimensional vector representing the features of the gravity field;

[0069] The P-Encoder performs geometric transformation alignment on the position coordinates of the inversion space to generate a high-dimensional vector containing the geometric structure information of the target body;

[0070] The gradient change caused by the gravity anomaly density difference and the geometric constraint of the inversion space are processed synchronously in cooperation with the low-dimensional vector and the high-dimensional vector, to form a field-space feature coupling mechanism;

[0071] The field-space feature coupling mechanism provides regularization constraints for the inversion process.

[0072] Specifically, the traditional CNN, U-Net and other network models often fail to recognize the influence of the inversion space and the position information of the observation network on the gravity inversion accuracy in the inversion process. Therefore, the research goal of the present application is to enable the model to more accurately capture the spatial position features of each gravity anomaly data in the observation network and each density unit in the underground density inversion space, and what dependency exists between these features, so as to improve the accuracy and reliability of the inversion.

[0073] As shown in Figure 3 The spatial pre-training network model proposed in the present embodiment realizes cross-domain feature coupling of observation signals and underground structures by constructing a physical correlation framework of gravity field features and inversion space. The model designs a double-encoder mechanism: F-Encoder based on convolution and self-attention module extracts multi-scale local features in the gravity field map and captures the spatial correlation of observation points, which is essentially a "physical signal coding" of the field response of geological bodies such as density interface fluctuation and structure boundary, generating a low-dimensional vector F containing implicit gravity field features; however, P-Encoder generates a high-dimensional vector P containing geometric structure information of target bodies such as salt domes by introducing the T-Net module of PointNet to perform geometric transformation alignment on the position coordinates of the inversion space, realizing "spatial prior coding" of the morphological features of underground structures. The physical meaning of the model lies in breaking the separation processing mode of "gravity field features" and "density anomaly body spatial position" in traditional inversion: that is, through the cooperation of F-Encoder and P-Encoder, the gradient change caused by the density difference of gravity anomaly and the geometric constraint of the inversion space are considered synchronously, to form a "field-space feature coupling" mechanism; the structured representation of the position coordinates by the T-Net module is equivalent to introducing the morphological prior of geological bodies, and the modeling of the observation point correlation by the self-attention layer implicitly learns the spatial covariance of the gravity field; the cross-domain mapping relationship established by pre-training provides regularization constraints for the inversion process, effectively suppressing the non-uniqueness of the solution, so that the inversion result is closer to the real geological structure.

[0074] The PointNet architecture T-Net module is integrated into the design of the P-Encoder, and the core function thereof is to convert the position coordinate information of the inversion space in the simulation data into a high-dimensional feature vector P after a series of transformation and alignment operations, thereby enhancing the geometric invariance of the model. The module comprises an input transformation network and a feature transformation network: first, the spatial position features of the input are extracted through a multi-layer perceptron (MLP), and are aggregated into a global feature vector through maximum pooling, and then a transformation matrix of a corresponding dimension is output through a fully connected layer; finally, the input data is spatially aligned by using the transformation matrix, so that the transformation of the feature space is realized.

[0075] Further, the multiple gravity anomaly data feature fusion is as follows:

[0076] In the field of geophysical gravity anomaly body inversion, gravity anomaly data Gz and gravity gradient component Gxz are two common types of observation data, which can reflect different characteristics of underground abnormal bodies. The gravity anomaly data Gz usually contains more low-frequency information, and the low-frequency part corresponds to large-scale geological body structures. Analysis and processing of these low-frequency information can effectively extract the general shape and distribution range of large-scale deep underground geological bodies, providing macroscopic constraints for inversion. Gxz contains more high-frequency information, and the high-frequency information corresponds to the edges and details of the image or signal. In geophysical inversion, these high-frequency information helps to identify the boundaries, faults and other fine structures of underground geological bodies. Through extraction and analysis of high-frequency information, the fine structure of the geological body can be described. The joint inversion method using the two types of gravity anomaly data can comprehensively depict the underground geological structure.

[0077] Therefore, after using Mamba and space and training two independent channels to extract features, the model uses a concatenation operation to fuse the multi-scale information extracted by the two channels to realize the collaborative optimization of geophysical constraints and gravity data driving. In addition, the Mamba module extracts the low-frequency components of the gravity anomaly data Gz to reflect the macroscopic shape and distribution range of deep large-scale geological bodies, providing regional background constraints for inversion; the spatial pre-training module extracts the high-frequency components of the gravity gradient component Gxz, which is sensitive to fine structures such as shallow faults and boundaries, and enhances the boundary resolution capability. Through feature vector concatenation, the two types of field information form a multi-scale fusion mechanism of “low-frequency control background and high-frequency detail”, which provides complete physical constraints covering different geological depths and structural scales for inversion, and also provides more abundant input representations for subsequent network layers.

[0078] Further, the U-shaped neural network uses a one-dimensional convolutional layer to replace the traditional two-dimensional convolutional layer for processing one-dimensional sequence characteristics of the gravity anomaly data.

[0079] Specifically, the U-shaped network structure is originally derived from the classic literature "U-Net: Convolutional Networks for Biomedical Image Segmentation" in the field of biomedical image segmentation. Through the U-shaped topology structure and the cross-layer skip connection mechanism, the U-shaped network structure realizes efficient fusion of multi-scale features of images and exhibits excellent performance in medical image segmentation tasks. When migrated to the field of geophysical model inversion, the two-dimensional convolution architecture adopted by the traditional U-shaped network faces significant challenges: when inverting gravity anomaly data, whether it is a 1xN row direction convolution or an N*N two-dimensional region convolution, the model inversion precision is limited and the computational complexity is high due to the introduction of redundant feature coupling calculation.

[0080] In the geophysical gravity inversion scenario, gravity anomaly data essentially constitute a one-dimensional ordered sequence, and the gravity anomaly value of each measurement point corresponds to an element in the sequence, and its physical meaning mainly reflects the spatial correlation of the measurement point. In view of this, the embodiment proposes an improved scheme suitable for the characteristics of geophysical data: replacing the two-dimensional convolution layer in the traditional U-shaped network with a one-dimensional convolution, while maintaining the multi-scale feature fusion capability, the computational efficiency is significantly improved.

[0081] From the perspective of parameter quantity optimization, assuming that the input feature dimension is I (such as including gravity anomaly, terrain correction value and other multi-source data channels), the output feature dimension is O, and the one-dimensional convolution kernel length is K, the learnable parameter quantity of the one-dimensional convolution layer is:

[0082]

[0083] In comparison, the parameter quantity of the two-dimensional convolution layer (the convolution kernel size is K_h*K_w) is:

[0084]

[0085] When processing one-dimensional sequence data of the same scale, the parameter quantity of the two-dimensional convolution increases in a quadratic order with the kernel size, while the one-dimensional convolution only increases linearly. This feature makes the one-dimensional convolution have a significant advantage in geophysical long sequence data processing, which can effectively capture the spatial correlation of gravity anomaly data through convolution kernel sliding operation, and avoid the redundant calculation introduced by two-dimensional convolution, thereby greatly reducing the model parameter quantity and computational complexity while ensuring the inversion precision, providing a lightweight solution for rapid gravity field inversion.

[0086] Further, the U-shaped neural network also includes a full-scale skip connection mechanism for fusing features of different scales in the encoder and the decoder. The full-scale skip connection adjusts the features of different scales to the same scale through linear mapping and then fuses them, and further integrates them through one-dimensional convolution.

[0087] Specifically, in the process of geophysical gravity density anomaly body inversion using the traditional U-Net, different scale features can provide different information. Low-level features have small receptive fields and can capture the boundary texture information in the field map; high-level features have large receptive fields and can capture the semantic information in the field map. When the traditional U-Net performs layer-by-layer downsampling, part of the edge information is lost, and this lost feature cannot be recovered by upsampling. Although the U-Net uses a skip connection to fuse multi-scale information, the decoder still has difficulty in fully utilizing the full-scale feature information extracted by the encoder in the downsampling process. To solve this problem, the full-scale skip connection is used in the embodiment.

[0088] As shown in Figure 4 , the gravity field data will pass through a U-shaped architecture composed of an encoder and a decoder. The encoder in the first half focuses on feature extraction, while the decoder in the second half is responsible for feature fusion and mapping. The full-scale skip connection strategy is used in the embodiment, aiming to fuse gravity features from different scales. Take the upsampling layer D2 as an example for detailed description. As shown in part t of the figure, D2 is different from the traditional skip connection. Instead of relying only on D3 to directly fuse features with E2 after deconvolution, it is composed of four parts. D2 first adjusts all the scale features extracted by each layer of the neural network to the same scale through linear mapping, and then fuses them. Finally, one-dimensional convolution operation is used to further integrate these fused features.

[0089] The mathematical expression formula is as follows:

[0090]

[0091] Where, the function C represents the convolution operation, the function H represents the feature aggregation mechanism (a convolution layer + a BN + a ReLU), the functions D and u represent the upsampling and downsampling operations respectively, “[·]” represents the channel dimension splicing fusion, E is the feature of each stage of the encoder, i marks the “current encoder stage of interest”, k is a temporary variable for traversing “earlier encoder stages” to aggregate multi-scale features.

[0092] In this way, each decoder module fuses feature maps from small and same scales in the encoder, as well as from large scales in the previous decoder. These feature maps capture fine-grained and coarse-grained semantics at full scale, and have fewer parameters. In geophysics, the use of full-scale skip connection can effectively reduce the loss of fine-grained spatial information, which leads to the inability to capture the boundary information of different geology and the decline of the accuracy of anomaly body edge positioning.

[0093] To illustrate the effectiveness of the network model, this embodiment designs a relatively simple simulation model ( Figure 7 Tests and analyses were conducted, and comparative experiments were performed using other deep learning inversion methods. The results are as follows: Figure 5 As shown, Figure 5 (a1) in the image represents a 3D visual image of the real model. Figure 5 (a2) in the diagram is the xz cross-section of the actual model along y=1800 meters. Figure 5 (a3) is a cross-sectional view of the actual model along the yz plane at x=700 meters. Figure 5 (b1) in the image is a three-dimensional visual image of the U-Net inversion result. Figure 5 (b2) in the image is the xz cross-section of the U-Net inversion result along y=180°. Figure 5 (b3) in the figure is the yz cross-section of the U-Net inversion result along x=700. Figure 5 (c1) in the image represents a 3D visual image of the TransUnet inversion result. Figure 5 (c2) in the figure is the xz cross-section of the TransUnet inversion result along y=180°. Figure 5 (c3) in the figure is the yz cross-section of the TransUnet inversion result along x=700. Figure 5 (d1) in the image represents a 3D visual image of the PMU-Net inversion result. Figure 5 (d2) in the figure is the xz cross-section of the PMU-Net inversion result along y=180°. Figure 6 (d3) in the image is the yz section plot of the PMU-Net inversion result along x=700. Figure 7 It is clearly observed that U-Net did not achieve ideal accuracy in both the 3D inversion space and the 2D field map. Although the 3D visualization model predicted by U-Net largely matches the simulation model, the accuracy of the 2D field map was significantly affected due to blurred edges. This phenomenon may be attributed to U-Net's shortcomings in extracting global information.

[0094] On the other hand, TransUNet has a lower accuracy in the three-dimensional inversion space, but its accuracy in the two-dimensional field map exceeds that of UNet. This is because TransUNet introduces a self-attention mechanism, which enables it to capture global information and more accurately depict edge details. Therefore, the density contour of the field map generated by TransUNet is smoother. However, due to the lack of position information learning of the model, the predicted density anomaly body position is shifted, which affects the accuracy of the three-dimensional inversion space. It is worth noting that the above-mentioned U-Net also has limitations in learning position information, but its deviation in anomaly body position prediction is not common. The case of TransUNet in the figure is a special case and is extremely rare.

[0095] The PMU-Net network proposed in this embodiment exhibits excellent edge information restoration capability in the inversion result. It effectively makes up for the information loss and position deviation problems often encountered by traditional U-Net and TransUNet in restoring edge information. This significant advantage is mainly due to the pre-trained model and attention mechanism adopted by the PMU-Net network, which enables it to better capture the internal dependency between the density anomaly body position and the station position, as well as the mutual dependency between the two. At the same time, the full-scale skip connection adopted by the PMU-Net network can more comprehensively capture coarse-grained semantic information and fine-grained semantic information, effectively reducing the feature loss in the downsampling process. Therefore, the inversion method of the network predicts a density structure that has a higher degree of consistency with the simulation model.

[0096] Further, the method further comprises generating simulation training data:

[0097] Initializing the inversion space parameters and the geological body feature parameters;

[0098] Generating a survey network, setting grid parameters and constructing a structured sampling grid;

[0099] Constructing a density distribution model in the inversion space according to the geological body feature parameters;

[0100] Calculating the gravity anomaly data corresponding to the density distribution model according to the forward formula;

[0101] Storing the gravity anomaly data and the corresponding density distribution model to form a training data set.

[0102] Specifically, integrating prior knowledge into the training data and input data of the neural network is an effective method to impose constraints on data-driven deep learning, which can effectively solve the weak interpretability of deep learning in gravity inversion and further improve the prediction ability of deep learning. Generating simulation data is an important way to embed prior knowledge in training data, and deep learning has always faced the challenge of lacking data in the field of gravity inversion. Therefore, it is very important to generate good simulation data.

[0103] In theory, the gravity anomaly observed by the measurement network is caused by the density anomaly body in the inversion space, which is irrelevant to the background wall rock, and the gravity anomaly is a positive anomaly. And the real geological body is usually irregular, but the influence of simple geological body on gravity anomaly can be simulated by generating some regular density anomaly bodies, and finally the superposition of some regular anomaly bodies can simulate complex geological bodies.

[0104] ​ A flowchart is generated for the data set. The flowchart shows the data set generation process using deep learning method for geophysical inversion research. The flow starts from "initial configuration", which defines the inversion space parameters (including x-axis, y-axis, z-axis range, where mp_inter is the division interval of the inversion space, and num_block is the number of inversion space division blocks) and geological body characteristics (x1, x2, y1, y2, z1, z2 are the starting and ending coordinates of x, y, z axes, and q represents the filling density of the geological body), laying the foundation for subsequent modeling. Next is "measurement network generation", which sets the grid parameters (the measurement network is divided by ob_inter as the interval distance), and assigns values to obx, obz by traversing the X / Y direction, to construct a structured sampling grid. Then, "inversion space generation" resets the inversion space parameters (the geological body is divided by mp_inter=200 as the interval distance), and traverses all blocks: when the block space position satisfies x1≤i≤x2, y1≤j≤y2, z1≤k≤z2, its density is set to q, thereby constructing a density distribution model. The "forward calculation" step calculates the gravity anomaly data using the forward formula, simulates the observation results based on the defined density model. Finally, "data storage and file management" creates a save directory, generates a file name and writes the gravity anomaly data, ensuring the systematic saving of synthetic data. These data are crucial for training deep learning models and can be used for subsequent inversion and estimation of actual density distribution from observed anomalies, bridging the forward modeling and inversion tasks in the field of geophysics.

[0105] where the gravity anomaly is caused by the change of the gravitational field. For example ​As shown, when the density of the inversion space is uniform, point A is only subject to the action of the earth's gravity, and the direction of the force is vertically downward; however, when there is a density anomaly body X in the inversion space, which is different from the surrounding environment, X will generate a new gravity, thereby causing point A to be affected by the gravity of the earth's center and the gravity of the density anomaly body X. The superposition of the two kinds of gravity causes the gravity field of point A to change, thereby generating a gravity anomaly. The gravity anomaly values are obtained by the forward formula respectively:

[0106]

[0107]

[0108] In the formula, V represents the gravity potential, x, y, z represent the spatial coordinates of the observation point, which are used to determine the position of the observation point in space, x', z' represent the source point (anomaly body) coordinate variables, which are used to determine the position of the mass source (such as underground density anomaly body, etc.) generating the gravity field in space. r represents the position vector of the field point, which represents the spatial position of the observation point and can be composed of coordinates (x, y, z), r' represents the position vector of the source point, which represents the spatial vector of the position of the mass source, |r'-r| represents the distance between the field point and the source point, that is, the modulus of the difference between the position vectors r' and r, reflecting the spatial distance between the two. ρ(r') represents the mass density at the source point, which describes the mass distribution of the unit volume of the mass source. γ represents the proportionality constant related to gravity, which is related to the physical law of the gravity field, the unit system, etc. (for example, the gravitational constant G is related). represents a triple volume integral, which is integrated with respect to the volume element dv' of the mass source, and the purpose is to accumulate the contribution of all mass source elements to the second-order partial derivative of the gravity potential of the observation point (gravity gradient component). σ represents the mass surface density (here the integral is a volume integral, more accurately, it can be understood as a mass density, which represents the mass in a unit volume and describes the density characteristics of the mass distribution generating the gravity field). ξ, η, ζ represent the spatial coordinates of the source point, that is, the spatial coordinates of the position of the mass element generating the gravity field. dξ, dη, dζ represent the volume element, which is a small volume part in the integral region, used to differentially accumulate the mass distribution.

[0109] This embodiment adopts a survey network with a size of 4000m*4000m, and sets the grid spacing to 200m, so that it can be accurately divided into 400 observation points. For the convenience of further research and discussion, it is assumed that the survey network is a plane, that is, the z coordinates of all observation points are equal to 0. Gz and Gxz obtained by the forward formula are used as the input of the deep neural network, so that the geological information in the model can be more accurately extracted. Through this method, the causes of gravity anomalies and the relationship between gravity anomaly values and the density of the inversion space can be more deeply understood, and a solid theoretical basis can be provided for subsequent inversion work.

[0110] The embodiment also provides a computer device, comprising a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method.

[0111] The embodiment also provides a computer readable storage medium, wherein a computer program is stored in the computer readable storage medium, and the computer program is executed by a processor to implement the steps of the method.

[0112] The above merely describes the preferred embodiments of the present application, but the protection scope of the present application is not limited to this. Any changes or replacements within the technical scope disclosed by the present application can be easily conceived by those skilled in the art, and should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A gravity anomaly body inversion method based on a PMU-Net deep learning network, characterized in that, The method comprises the following steps: obtaining two different types of gravity anomaly data as input; processing the two types of gravity anomaly data through two independent channels to mine the correlation and complementarity between the data, wherein one channel uses a Mamba module to capture the dependence between observation points, and the other channel uses a spatial pre-training model to capture the position correlation between observation points and underground density; the Mamba module comprises: performing relative coordinate coding on each observation point of the gravity anomaly data to construct a spatial correlation matrix; extracting spatial pattern features of the gravity anomaly data through a multi-scale convolution kernel; based on a dynamic state space model, constructing an observation point evolution sequence using a planar coordinate grid index, and evolving the physical state of each observation point according to the law of conservation of mass; mapping the gravity anomaly signal through an underground physical property structure response matrix, and tensor coupling the mapping result and the initial geological state through a state transition matrix; projecting the updated geological state to the detectable physical quantity space through an observation operator, and introducing a lithology-density conversion matrix to retain the direct influence term of the original input signal; through a double-path fusion mechanism, the structural evolution dynamics process and the static distribution of density and physical properties are cooperatively constrained; the spatial pre-training model comprises: constructing a physical correlation framework of gravity field features and inversion space for cross-domain feature coupling of observation signals and underground structures; extracting multi-scale local features in the gravity field map through an F-Encoder and capturing the spatial correlation of observation points to generate a low-dimensional vector representing the gravity field features; performing geometric transformation alignment on the position coordinates of the inversion space through a P-Encoder to generate a high-dimensional vector containing the geometric structure information of the target body; synchronously processing the gradient changes caused by gravity anomaly density differences and the geometric constraints of the inversion space by cooperating the low-dimensional vector and the high-dimensional vector to form a field-space feature coupling mechanism; providing regularization constraints for the inversion process through the field-space feature coupling mechanism; fusing the processed features, inputting the fused features into a U-shaped neural network constructed by one-dimensional convolution to perform inversion, and obtaining the underground density distribution result.

2. The method of claim 1, wherein, The two different types of gravity anomaly data include gravity anomaly data and gravity gradient anomaly data.

3. The method of claim 1, wherein, The U-shaped neural network uses a one-dimensional convolution layer to replace a traditional two-dimensional convolution layer to process one-dimensional sequence characteristics of the gravity anomaly data.

4. The method of claim 1, wherein, The U-shaped neural network further comprises a full-scale skip connection mechanism for fusing features of different scales in the encoder and the decoder.

5. The method of claim 4, wherein, The full-scale skip connection mechanism adjusts the features of different scales to the same scale through linear mapping and then fuses them, and further integrates them through one-dimensional convolution.

6. The method of claim 1, wherein, It also includes generating simulated training data: initializing inversion space parameters and geological body feature parameters; generating a survey network, setting grid parameters and constructing a structured sampling grid; constructing a density distribution model in the inversion space according to the geological body feature parameters; calculating the gravity anomaly data corresponding to the density distribution model according to the forward formula; store the gravity anomaly data and the corresponding density distribution model to form a training data set.

7. A computer apparatus comprising a memory, a processor, and a computer program stored on the memory, wherein the computer program, when executed by the processor, causes the processor to perform the method of any one of claims 1 to 6. The processor executes the computer program to implement the steps of the method of any one of claims 1-6.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by the processor, implements the steps of the method of any one of claims 1-6.

Images