Information fusion method and system for deep geological exploration

Abstract

The invention provides an information fusion method and system for deep geological exploration, and relates to the technical field of data processing. Comprising the steps that multiple sets of depth information of a to-be-explored area are acquired, and the depth information comprises shallow layer data and deep layer data; processing the shallow data by using an improved nonlinear conversion algorithm to obtain a first conversion information set; processing the deep data by using a deep learning algorithm to obtain a second conversion information set; performing data fusion on the first conversion information set and the second conversion information set to obtain a first fusion information subset and a second fusion information subset; and performing data optimization and fusion on the first fusion information subset and the second fusion information subset through a tensor decomposition algorithm and a joint inverse algorithm to obtain final fusion data. According to the method, the problems of insufficient baseline length and low information integration efficiency during processing of a complex geological environment in the prior art are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an information fusion method and system for deep geological exploration. Background Art

[0002] With the continuous development of deep geological exploration technology, traditional geological exploration methods are facing many challenges. Existing technologies usually rely on a single source of depth information, resulting in incomplete geological information and difficulty in meeting exploration needs in complex geological environments. Especially in the case of multiple depth parallaxes, the linear conversion of traditional methods will lead to a significant expansion of the number of bits of information, affecting the efficiency and accuracy of data processing. In addition, when processing deep geological information, the existing de-masking program has a short baseline length and cannot obtain high-precision depth information, which limits the reliability of exploration results.

[0003] The existing patent CN113537276B discloses a fusion method and system for multi-depth information. It uses the nonlinear conversion method adopted in the disclosed patent to process multi-depth information, avoiding the problem of bit expansion caused by traditional linear conversion, and fully utilizes high-precision depth information through optical center conversion technology, which significantly improves the accuracy of the fusion result. However, when dealing with complex geological environments, there are still problems such as insufficient baseline length and low information integration efficiency. Summary of the invention

[0004] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide an information fusion method and system for deep geological exploration.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] An information fusion method for deep geological exploration, comprising:

[0007] Acquire multiple sets of depth information of the area to be explored, wherein the depth information includes shallow layer data and deep layer data;

[0008] Processing the shallow data using an improved nonlinear conversion algorithm to obtain a first conversion information set;

[0009] Processing the deep data using a deep learning algorithm to obtain a second transformation information set;

[0010] Performing data fusion on the first conversion information set and the second conversion information set respectively to obtain a first fused information subset and a second fused information subset;

[0011] The first fused information subset and the second fused information subset are optimized and fused by using a tensor decomposition algorithm and a joint inverse algorithm to obtain final fused data.

[0012] Preferably, the obtaining of multiple sets of depth information of the area to be explored includes:

[0013] Establish a three-dimensional model corresponding to the area to be tested;

[0014] Obtain exploration routes with different baselines based on the 3D model;

[0015] Select the corresponding sensor equipment according to the exploration routes of different baselines to obtain depth information.

[0016] Preferably, the step of establishing a three-dimensional model corresponding to the area to be measured includes:

[0017] Obtain real-time and historical multi-source data;

[0018] Obtaining geological characteristics of the area to be measured according to the real-time multi-source data and the historical multi-source data;

[0019] Determining the geological unit for constructing the three-dimensional model according to the geological characteristics;

[0020] Based on geological modeling software, a three-dimensional model is constructed according to the geological units and the Kriging interpolation method.

[0021] Preferably, the multi-source data includes:

[0022] Seismic data, drilling data, electromagnetic data, and remote sensing imagery.

[0023] Preferably, obtaining the geological characteristics of the area to be measured based on the real-time multi-source data and the historical multi-source data includes:

[0024] Performing data cleaning, data standardization and noise processing on the real-time multi-source data to obtain a pre-processed real-time data set;

[0025] Normalizing the historical multi-source data and filling missing values ​​using an interpolation method to obtain a preprocessed historical data set;

[0026] Using an improved Kalman filter algorithm to fuse the preprocessed real-time data set to obtain real-time fused geological information;

[0027] Processing the preprocessed historical data set using a geostatistical method to obtain historical fused geological information;

[0028] Extracting geological features from the real-time fused geological information based on a machine learning algorithm to obtain a first initial geological feature;

[0029] Extracting geological features from the historical fused geological information using an improved time series analysis method to obtain a second initial geological feature;

[0030] The first initial geological feature and the second initial geological feature are subjected to feature fusion to obtain the geological feature of the area to be measured.

[0031] Preferably, the expression for real-time fusion of geological information is:

[0032] A= ;

[0033] Among them, A is the real-time fusion of geological information, is the state transfer matrix, is the state vector, is the Kalman gain matrix, is the observation vector.

[0034] Preferably, the calculation expression of the second initial geological characteristic is:

[0035] ;

[0036] in, is the second initial geological feature, is the variational mode decomposition of the mth mode, is the bidirectional long short-term memory network of the mth mode, is the spatiotemporal gating weight of the mth mode, is the spatial attention mechanism, is the historical fusion geological information, and M is the total number of modes.

[0037] Preferably, the process of processing the shallow data using an improved nonlinear conversion algorithm to obtain a first conversion information set includes:

[0038] De-noising the shallow data using an adaptive wavelet threshold, and segmenting and normalizing the de-noised data according to the baseline length to obtain pre-processed shallow data;

[0039] Using a polynomial fitting function to obtain the local nonlinear characteristics of the preprocessed shallow data;

[0040] Obtaining the global geological structure of the preprocessed shallow data using a Gaussian kernel function;

[0041] A first conversion information set is determined according to the local nonlinear characteristics and the global geological structure.

[0042] An information fusion system for deep geological exploration, comprising:

[0043] A data acquisition module, used to acquire multiple sets of depth information of the area to be explored, wherein the depth information includes shallow layer data and deep layer data;

[0044] A first conversion module, used for processing the shallow data by using an improved nonlinear conversion algorithm to obtain a first conversion information set;

[0045] A second conversion module, used for processing the deep data using a deep learning algorithm to obtain a second conversion information set;

[0046] A primary fusion module, used to perform data fusion on the first conversion information set and the second conversion information set respectively to obtain a first fused information subset and a second fused information subset;

[0047] The final fusion module is used to optimize and fuse the first fusion information subset and the second fusion information subset through a tensor decomposition algorithm and a joint inverse algorithm to obtain final fusion data.

[0048] The present invention discloses the following technical effects:

[0049] The present invention provides an information fusion method and system for deep geological exploration, the method comprising: obtaining multiple sets of depth information of the area to be explored, the depth information comprising shallow data and deep data; processing the shallow data using an improved nonlinear conversion algorithm to obtain a first conversion information set; processing the deep data using a deep learning algorithm to obtain a second conversion information set; fusing the first conversion information set and the second conversion information set respectively to obtain a first fused information subset and a second fused information subset; optimizing and fusing the first fused information subset and the second fused information subset by a tensor decomposition algorithm and a joint inversion algorithm to obtain the final fused data. The present invention avoids the loss of deep information caused by signal attenuation during traditional single baseline fusion by distinguishing between independent processing of shallow data (short baseline) and deep data (long baseline); compressing the shallow data dimension by 65% ​​by an improved nonlinear conversion algorithm to eliminate the expansion of redundant information bits; and using a "tensor decomposition + joint inversion" algorithm in the final fusion stage to align the shallow-deep data of different baselines across scales to solve the spatial misalignment problem of long baseline data and short baseline data. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0051] Figure 1 A flow chart of an information fusion method for deep geological exploration provided by an embodiment of the present invention;

[0052] Figure 2A schematic diagram of the structure of an information fusion system for deep geological exploration provided by an embodiment of the present invention.

[0053] Reference numerals:

[0054] 1-data acquisition module, 2-first conversion module, 3-second conversion module, 4-primary fusion module, 5-final fusion module. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0057] like Figure 1 As shown, the present invention provides an information fusion method for deep geological exploration, comprising:

[0058] Step 100: Acquire multiple sets of depth information of the area to be explored, wherein the depth information includes shallow layer data and deep layer data;

[0059] Step 200: Processing the shallow data using an improved nonlinear conversion algorithm to obtain a first conversion information set;

[0060] Step 300: Process the deep data using a deep learning algorithm to obtain a second transformation information set;

[0061] Step 400: performing data fusion on the first conversion information set and the second conversion information set respectively to obtain a first fused information subset and a second fused information subset;

[0062] Step 500: Optimizing and fusing the first fused information subset and the second fused information subset through a tensor decomposition algorithm and a joint inverse algorithm to obtain final fused data.

[0063] Specifically, baseline length refers to the spacing between sensor arrays (such as seismic wave transmitters and receivers) or the coverage of the detection path.

[0064] Short baseline (<500 m): High-frequency signals decay quickly, suitable for capturing shallow (0-1000 m) high-resolution data, but with limited penetration depth;

[0065] Long baseline (>2000 meters): Low-frequency signals have strong penetration and are suitable for detecting large-scale structures in deep layers (1000-5000 meters), but the resolution is relatively low.

[0066] More specifically, shallow data: short baseline (200 meters) seismic wave data, used to identify shallow rock interfaces (such as the boundary between sandstone and shale at 0-800 meters); sensor selection: high-resolution seismic detectors (frequency range 50-200 Hz) to match the high-frequency signal requirements of short baselines.

[0067] Deep data: long baseline (3000 meters) electromagnetic sounding data, used to detect deep fault extension (such as basement faults at 2000-4500 meters); sensor selection: low-frequency electromagnetic transmitter (frequency range 0.1-10 Hz), adapted to the low-frequency penetration requirements of long baselines.

[0068] Furthermore, the step of obtaining multiple sets of depth information of the area to be explored includes:

[0069] Establish a three-dimensional model corresponding to the area to be tested;

[0070] Obtain exploration routes with different baselines based on the 3D model;

[0071] Select the corresponding sensor equipment according to the exploration routes of different baselines to obtain depth information.

[0072] Specifically, short baseline route generation (200-500m spacing):

[0073] Extract the three-dimensional coordinate points of the shallow target area;

[0074] Perform Delaunay triangulation to generate the initial triangle mesh;

[0075] Insert new nodes in the area where the triangle side length is >200m, and iterate to refine the mesh until the side length is ≤200m;

[0076] Generate parallel survey lines along the triangle edges to ensure coverage density ≥ 2 lines / km 2 .

[0077] Long baseline route generation (1-3km spacing):

[0078] Calculate the structural continuity of each voxel in the 3D model (based on density gradient consistency);

[0079] Define pheromone update rules;

[0080] Ant colony iteratively searches for the optimal path (iteration times ≥ 100 times);

[0081] The path is smoothed using Bezier curves to avoid terrain obstacles.

[0082] Table 1 is a table of information of baselines of different lengths, and Table 1 is as follows:

[0083] Table 1

[0084] Baseline Type Coordinate range Coverage Target Line density Short baseline (1000m,2000m)-(1500m,2500m) Shallow caves and fissure zones 5 / / km² Long baseline (3000m,500m)→(3000m,3500m) Basement faults, deep magma chambers 1 / 2 / km² Hybrid Baseline Along the short baseline grid diagonal Cross-scale construct validation 2 / km²

[0085] Furthermore, the establishing of the three-dimensional model corresponding to the area to be measured includes:

[0086] Obtain real-time and historical multi-source data;

[0087] Obtaining geological characteristics of the area to be measured according to the real-time multi-source data and the historical multi-source data;

[0088] Determining the geological unit for constructing the three-dimensional model according to the geological characteristics;

[0089] Based on geological modeling software, a three-dimensional model is constructed according to the geological units and the Kriging interpolation method.

[0090] Furthermore, the multi-source data includes:

[0091] Seismic data, drilling data, electromagnetic data, and remote sensing imagery.

[0092] Furthermore, obtaining the geological characteristics of the area to be measured based on the real-time multi-source data and the historical multi-source data includes:

[0093] Performing data cleaning, data standardization and noise processing on the real-time multi-source data to obtain a pre-processed real-time data set;

[0094] Normalizing the historical multi-source data and filling missing values ​​using an interpolation method to obtain a preprocessed historical data set;

[0095] Using an improved Kalman filter algorithm to fuse the preprocessed real-time data set to obtain real-time fused geological information;

[0096] Processing the preprocessed historical data set using a geostatistical method to obtain historical fused geological information;

[0097] Extracting geological features from the real-time fused geological information based on a machine learning algorithm to obtain a first initial geological feature;

[0098] Extracting geological features from the historical fused geological information using an improved time series analysis method to obtain a second initial geological feature;

[0099] Performing feature fusion on the first initial geological feature and the second initial geological feature to obtain geological features of the area to be measured;

[0100] The expression of real-time fusion of geological information is:

[0101] A= ;

[0102] Among them, A is the real-time fusion of geological information, is the state transfer matrix, is the state vector, is the Kalman gain matrix, is the observation vector; the state vector includes the rock layer thickness, fault position and physical properties; the state vector and the observation vector are obtained by preprocessing real-time multi-source data through the first mapping function and the second mapping function.

[0103] The first mapping function:

[0104] Extract geological structure parameters:

[0105] Rock thickness: cross-validated by seismic wave reflection time series and drilling data;

[0106] Fault location: integrating electromagnetic anomaly gradient and remote sensing image fault line characteristics;

[0107] Physical properties: Inversion of rock density and porosity based on seismic wave velocity (Vp / Vs);

[0108] Data dimensionality reduction: compressing multi-dimensional sensor data (such as 128-dimensional electromagnetic spectrum features) into a low-dimensional state space (such as an 8-dimensional vector);

[0109] Timing alignment: Compensating for time delays between different sensors (such as the difference in propagation speed between seismic waves and electromagnetic signals).

[0110] The second mapping function:

[0111] Map the preprocessed real-time multi-source data into the observation vector Y(t);

[0112] Specific functions:

[0113] Generate direct observations:

[0114] Seismic wave amplitude (0.5-80Hz frequency band);

[0115] Electromagnetic field strength (10^ -6 ~10^ -3 V / m);

[0116] Drill core density (2.5-3.2 g / cm³);

[0117] Dynamic Noise Suppression:

[0118] Eliminate the interference of surface mechanical vibration on seismic data (>20Hz high-frequency noise);

[0119] Suppression of atmospheric ionospheric interference in electromagnetic data (day / night mode adaptation).

[0120] The state transfer matrix is ​​composed of the three-dimensional model information Determine; the three-dimensional model information here is determined by constructing an initial three-dimensional model using historical data.

[0121] ; in, is the state transfer matrix;

[0122] The state vector,is determined by preprocessing the real-time multi-source dataset;

[0123] The Kalman gain matrix is ​​given by the error covariance matrix and the weight matrix Sure;

[0124] ;

[0125] ;

[0126] in, is the Kalman gain matrix, is the state transfer matrix, represents the prediction error covariance matrix, is the forecast error covariance matrix of the previous moment, for The transpose of is the process noise covariance matrix; is the observation vector, which is determined by preprocessing the real-time multi-source data set; after introducing the weight matrix, the Kalman gain matrix can dynamically adjust the contribution of different data sources (seismic, drilling, electromagnetic, remote sensing) to improve the accuracy of the fusion result. Quantifies the uncertainty of the state vector, guides the calculation of the Kalman gain matrix and updates The algorithm can dynamically adjust the trust in the predicted values ​​and observed values ​​to improve the robustness of the fusion results.

[0127] =diag(σ S 2 (t),σ D 2 (t),σ E 2 (t),σ R 2 (t)); σ S 2 (t),σ D 2 (t),σ E2 (t),σ R 2 (t) are the error variances of seismic data, drilling data, electromagnetic data and remote sensing image data respectively.

[0128] diag is used to convert a vector into a diagonal matrix, define the error covariance matrix and the weight matrix, and ensure the independence between different data sources.

[0129] Furthermore, the calculation expression of the second initial geological characteristic is:

[0130] ;

[0131] in, is the second initial geological feature, is the variational mode decomposition of the mth mode, is the bidirectional long short-term memory network of the mth mode, is the spatiotemporal gating weight of the mth mode, is the spatial attention mechanism, is the historical fusion geological information, and M is the total number of modes.

[0132] Specifically, multimodal adaptive decomposition (VMD) replaces the preset wavelet basis to improve the signal decomposition accuracy; bidirectional time series modeling (BiLSTM) captures a more complete time dependency; spatiotemporal gating jointly optimizes time and space attention and dynamically adjusts modal weights; end-to-end spatial attention directly extracts geological spatial features to avoid information loss.

[0133] The process of processing the shallow data using an improved nonlinear conversion algorithm to obtain a first conversion information set includes:

[0134] De-noising the shallow data using an adaptive wavelet threshold, and segmenting and normalizing the de-noised data according to the baseline length to obtain pre-processed shallow data;

[0135] Specifically, adaptive wavelet denoising: targeting the high-frequency noise characteristics of short-baseline shallow data (such as surface vibration interference), the threshold is dynamically adjusted to retain details;

[0136] Segment normalization: Split the data by baseline length (e.g., every 200 meters) to eliminate local deviations caused by differences in sensor sensitivity.

[0137] Using a polynomial fitting function to obtain the local nonlinear characteristics of the preprocessed shallow data;

[0138] Specifically, a matrix of preprocessed shallow data (n data points, d-dimensional features) is obtained, the data is divided into a preset number of grids according to spatial coordinates, and each sub-block is processed independently, a quadratic polynomial is used to fit the sub-blocks to obtain a fitting coefficient matrix for each sub-block, and the local fitting quality is evaluated based on the fitting coefficient matrix; the local curvature and gradient direction are calculated through the quadratic term coefficients, and local nonlinear features are extracted based on the local curvature and gradient direction and their fitting coefficient matrix.

[0139] Obtaining the global geological structure of the preprocessed shallow data using a Gaussian kernel function;

[0140] Specifically, a Gaussian radial basis function (RBF) is used to calculate the similarity between data points, a global kernel matrix is ​​constructed based on the similarity between the data points, the global kernel matrix is ​​eigendecomposed, and the eigenvector corresponding to the maximum eigenvalue is used as the global structural basis vector to determine the global geological structure, wherein the global geological structure includes branch faults and anticline axes.

[0141] A first conversion information set is determined according to the local nonlinear characteristics and the global geological structure.

[0142] Specifically, the local nonlinear characteristics and the global geological structure are mapped to a unified coordinate system to be fused to obtain a first transformed information set.

[0143] like Figure 2 As shown, this embodiment also discloses an information fusion system for deep geological exploration, including:

[0144] A data acquisition module 1 is used to acquire multiple sets of depth information of the area to be explored, wherein the depth information includes shallow layer data and deep layer data;

[0145] A first conversion module 2, used for processing the shallow data using an improved nonlinear conversion algorithm to obtain a first conversion information set;

[0146] A second conversion module 3, used to process the deep data using a deep learning algorithm to obtain a second conversion information set;

[0147] A primary fusion module 4 is used to perform data fusion on the first conversion information set and the second conversion information set respectively to obtain a first fused information subset and a second fused information subset;

[0148] The final fusion module 5 is used to optimize and fuse the first fusion information subset and the second fusion information subset through a tensor decomposition algorithm and a joint inverse algorithm to obtain final fusion data.

[0149] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0150] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. An information fusion method for deep geological exploration, characterized in that: include: Acquire multiple sets of depth information of the area to be explored, wherein the depth information includes shallow layer data and deep layer data; Processing the shallow data using an improved nonlinear conversion algorithm to obtain a first conversion information set; Processing the deep data using a deep learning algorithm to obtain a second transformation information set; Performing data fusion on the first conversion information set and the second conversion information set respectively to obtain a first fused information subset and a second fused information subset; Performing data optimization and fusion on the first fusion information subset and the second fusion information subset through a tensor decomposition algorithm and a joint inverse algorithm to obtain final fusion data; The process of processing the shallow data using an improved nonlinear conversion algorithm to obtain a first conversion information set includes: De-noising the shallow data using an adaptive wavelet threshold, and segmenting and normalizing the de-noised data according to baseline length to obtain pre-processed shallow data; Using a polynomial fitting function to obtain the local nonlinear characteristics of the preprocessed shallow data; Obtaining the global geological structure of the preprocessed shallow data using a Gaussian kernel function; A first conversion information set is determined according to the local nonlinear characteristics and the global geological structure.

2. The information fusion method for deep geological exploration according to claim 1, characterized in that: The step of obtaining multiple sets of depth information of the area to be explored includes: Establish a three-dimensional model corresponding to the area to be tested; Obtain exploration routes with different baselines based on the 3D model; Select the corresponding sensor equipment according to the exploration routes of different baselines to obtain depth information.

3. The information fusion method for deep geological exploration according to claim 2, characterized in that: The establishing of the three-dimensional model corresponding to the area to be measured comprises: Obtain real-time and historical multi-source data; Obtaining geological characteristics of the area to be measured according to the real-time multi-source data and the historical multi-source data; Determining the geological unit for constructing the three-dimensional model according to the geological characteristics; Based on geological modeling software, a three-dimensional model is constructed according to the geological units and the Kriging interpolation method.

4. The information fusion method for deep geological exploration according to claim 3 is characterized in that: Multi-source data includes: Seismic data, drilling data, electromagnetic data, and remote sensing imagery.

5. The information fusion method for deep geological exploration according to claim 4, characterized in that: The step of obtaining geological characteristics of the area to be measured based on the real-time multi-source data and the historical multi-source data includes: Performing data cleaning, data standardization and noise processing on the real-time multi-source data to obtain a pre-processed real-time data set; Normalizing the historical multi-source data and filling missing values ​​using an interpolation method to obtain a preprocessed historical data set; Using an improved Kalman filter algorithm to fuse the preprocessed real-time data set to obtain real-time fused geological information; Processing the preprocessed historical data set using a geostatistical method to obtain historical fused geological information; Extracting geological features from the real-time fused geological information based on a machine learning algorithm to obtain a first initial geological feature; Extracting geological features from the historical fused geological information using an improved time series analysis method to obtain a second initial geological feature; The first initial geological feature and the second initial geological feature are subjected to feature fusion to obtain the geological feature of the area to be measured.

6. The information fusion method for deep geological exploration according to claim 5, characterized in that: The expression of real-time fusion of geological information is: A= ； Among them, A is the real-time fusion of geological information, is the state transfer matrix, is the state vector, is the Kalman gain matrix, is the observation vector.

7. The information fusion method for deep geological exploration according to claim 5, characterized in that: The calculation expression of the second initial geological characteristic is: ； in, is the second initial geological feature, is the variational mode decomposition of the mth mode, is the bidirectional long short-term memory network of the mth mode, is the spatiotemporal gating weight of the mth mode, is the spatial attention mechanism, is the historical fusion geological information, and M is the total number of modes.

8. An information fusion system for deep geological exploration, characterized in that: include: A data acquisition module, used to acquire multiple sets of depth information of the area to be explored, wherein the depth information includes shallow layer data and deep layer data; A first conversion module, used for processing the shallow data by using an improved nonlinear conversion algorithm to obtain a first conversion information set; A second conversion module, used for processing the deep data using a deep learning algorithm to obtain a second conversion information set; A primary fusion module, used to perform data fusion on the first conversion information set and the second conversion information set respectively to obtain a first fused information subset and a second fused information subset; The final fusion module is used to optimize and fuse the first fusion information subset and the second fusion information subset through a tensor decomposition algorithm and a joint inverse algorithm to obtain final fusion data.

Images