Cross-scale nesting and fusion modeling method for multi-scale geological model

Abstract

The invention provides a cross-scale nesting and fusion modeling method for a multi-scale geological model, which belongs to the field of geology and comprises four steps of multi-scale data preprocessing, data fusion, cross-scale nesting and fusion and dynamic coupling. According to the method, multi-source heterogeneous data can be integrated, data fusion is realized through a joint probability space, a variational assimilation framework, a fuzzy logic conflict factor and an alternating direction multiplier method, and the stability and precision of the model are improved by using a deep learning auxiliary fusion technology, such as a multi-scale convolutional adversarial network. Meanwhile, model parameters are determined through a volume average upscaling algorithm and random field condition simulation, dynamic adjustment is carried out through set Kalman filtering and localized set transformation Kalman filtering, and consistency, bidirectional feedback and iterative optimization of the parameters between the models are achieved. And finally, multi-scale convergence is realized through a dual grid strategy and restrictive interpolation, and convergence and dynamic balance of the model on the scale from kilometer to micrometer are ensured.

Description

Technical Field

[0001] The present invention belongs to the field of geology, and more specifically relates to a cross-scale nesting and fusion modeling method of a multi-scale geological model. Background Art

[0002] Establishing a geological model is a key step in fields such as oil exploration and development, CO2 geological storage, and geothermal development. Its quality and accuracy directly impact the final processing results and economic benefits. Traditional geological models are primarily based on information such as seismic reflection and drilling logging, and are based on assumptions of homogeneity and continuity. Conventional modeling methods struggle to achieve high accuracy and can only reflect large-scale geological features, with limited ability to depict detailed details.

[0003] With the advancement of science and technology, seismic technology, core analysis, and computer technology have made significant progress, providing a foundation for improving the accuracy of geological models. In addition, the types and scales of geological data are becoming increasingly diverse, such as scanning electron microscope images of microscopic pores. The use of scale-differentiated observational data is becoming increasingly important.

[0004] However, effectively utilizing these multi-scale, multi-source geological interpretation, geophysical, and geochemical data to construct a comprehensive geological model that simultaneously reflects large-scale geological features and depicts subsurface microstructures without conflict remains a challenge. This is especially true when the data are highly diverse in scale and complex in data type, leading to challenges such as data preprocessing, spatial resolution unification, data fusion, and dynamic processing.

[0005] Furthermore, traditional models typically incorporate only a single geological element, such as sedimentary structures, while neglecting geological processes. This often overlooks the complex interactions between geological elements and between geological factors and engineering behavior. This means that traditional models are also limited in their ability to respond to complex geological systems.

[0006] Therefore, a method is proposed to fully utilize multi-scale and multi-source data to establish a cross-scale nested and fused geological model to solve the above problems, improve the accuracy of oil exploration and production, reduce investment risks, and improve production effects, which has important practical significance and use value. Summary of the Invention

[0007] This invention aims to address current geological modeling challenges, including the difficulty integrating heterogeneous multi-source data, inconsistent resolution, data conflicts, the inability to effectively reflect both large-scale geological features and microscopic details, and the neglect of complex interactions between geological elements. By proposing a new cross-scale nesting and fusion model, it enables the construction of precise geological models, improving the accuracy and efficiency of geological modeling and its applications.

[0008] In order to achieve the above object, the present invention is implemented by adopting the following technical solution: comprising the following steps:

[0009] Step 1: Multi-scale data preprocessing to unify the coordinate system, resolution and physical dimension and eliminate scale deviation between data;

[0010] Step 2: Data fusion, which integrates multi-source heterogeneous data through joint probability space, variational assimilation framework, fuzzy logic conflict factor and alternating direction multiplier method (ADMM) optimization;

[0011] Step 3: Cross-scale nesting and fusion: hierarchical modeling of geological models at different scales, and establishment of a bidirectional coupling mechanism through adaptive grid technology and multi-physics field coupling;

[0012] Step 4: Dynamic coupling, through cross-scale parameter transfer, data assimilation, ensemble Kalman filtering and localized ensemble transform Kalman filtering, to achieve bidirectional feedback and iterative optimization between models.

[0013] In one solution, in step 1, the coordinate system is unified by using Helmert transformation, which includes the comprehensive application of scale factor, rotation matrix and translation vector to eliminate spatial deviations from different measurement systems; at the same time, the resolution is unified by using adaptive cubic convolution interpolation algorithm, which uses cubic B-spline basis function for interpolation to achieve resolution standardization of data of different scales while maintaining the geometric topology of the geological body.

[0014] In one solution, step 1 uses a feature scaling technique guided by dimensional analysis to achieve the unification of physical dimensions, specifically including the dimensionless conversion of parameters such as permeability, porosity, and seismic wave velocity. By calculating the mean and standard deviation of these parameters, the comparability of the data in different physical dimensions is ensured.

[0015] In one solution, step 2 uses joint likelihood function, cokriging interpolation and Markov random field for data fusion, combined with fuzzy logic conflict factor and alternating direction multiplier method optimization to solve the conflict problem between multi-source heterogeneous data, thereby achieving effective integration and consistency of data.

[0016] In one solution, in step 2, a multi-scale fusion model is constructed by mixing Gaussian processes to ensure statistical consistency at different scales and data types, and the flexibility and adaptability of Gaussian processes are used to handle uncertainty, thereby improving the robustness and accuracy of the model.

[0017] In one solution, step 3 uses adaptive grid technology and multi-physics field coupling, combined with deep learning-assisted fusion technology, such as multi-scale convolutional adversarial networks, to improve the stability and accuracy of the model and ensure the effectiveness of bidirectional coupling and information transfer between geological models of different scales.

[0018] In one solution, in step 4, the volume average upscaling algorithm and random field condition simulation are used to ensure parameter consistency between models of different resolutions, and dynamic adjustment is performed using ensemble Kalman filtering and localized ensemble transform Kalman filtering to achieve two-way feedback and iterative optimization between models.

[0019] In one scheme, in step 4, multi-scale convergence is achieved through a dual grid strategy and restrictive interpolation to form a temporally and spatially continuous geological-engineering integrated model, and the convergence and dynamic balance of the model at the kilometer to micrometer scale are ensured through joint criterion control of relative parameter change rate and data residual.

[0020] Beneficial effects of the present invention:

[0021] It can effectively integrate multi-source and multi-scale geological data, including seismic, well logging, core, outcrop, dynamic production data, etc. Data of different scales and types have good compatibility in the model, greatly improving the authenticity and accuracy of geological modeling.

[0022] Dynamic coupling technology can realize real-time adjustment of model parameters, ensuring the stability and accuracy of the model under complex geological conditions.

[0023] By using geostatistics and rock physics models, we can achieve physical and mathematical integration between models of different scales, making the models more physically meaningful and improving their credibility.

[0024] The cross-scale nesting and fusion modeling method of the multi-scale geological model of the present invention can not only better understand the structure and characteristics of the geological body, but also help scientists and engineers accurately predict the dynamic changes of the geological body. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0026] To facilitate understanding of the present invention, the present invention will be described more fully below with reference to the accompanying drawings. The drawings illustrate exemplary embodiments of the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and comprehensive understanding of the present invention.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as those understood by those skilled in the art to which the present invention pertains. The terms used in the present specification are for the purpose of describing specific embodiments only and are not intended to limit the present invention. To facilitate understanding of the present invention, a more comprehensive description of the present invention will be provided below with reference to the accompanying drawings. Typical embodiments of the present invention are shown in the drawings. However, the present invention may be embodied in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and comprehensive understanding of the present invention.

[0028] like Figure 1 As shown, a cross-scale nesting and fusion modeling method for a multi-scale geological model includes:

[0029] The cross-scale nesting and fusion modeling method of multi-scale geological models is a complex modeling method that combines different spatial resolutions (such as basin scale, oil and gas reservoir scale, and micro-pore scale) and data types (such as geological, geophysical, and geochemical data). It aims to construct a comprehensive geological model that can reflect both macroscopic geological characteristics and microscopic details.

[0030] By integrating multidisciplinary data and methods, this technology helps to significantly improve the geological model's ability to depict complex geological systems, becoming one of the core technologies that urgently need to be solved in the fields of deep oil and gas exploration and development, unconventional resource extraction, CO2 geological storage, deep geological engineering, and geothermal development.

[0031] Step 1: Multi-scale data preprocessing: unify the coordinate system, resolution, and physical dimension to eliminate scale deviation between data.

[0032] The core of multi-scale data preprocessing is to establish a basis for data comparability. The implementation process starts with the unification of spatial references. For multi-source data such as satellite remote sensing, seismic exploration, and drilling logging, the coordinate system needs to be normalized through the Helmert transformation, which is mathematically expressed as:

[0033]

[0034] Where s is the scale factor, R is the rotation matrix, and T is the translation vector. This transformation can eliminate spatial deviations between different measurement systems (such as WGS84 and local coordinate systems). Resolution unification involves grid reconstruction technology. To match the basin-scale kilometer-level grid with the pore-scale micron-level data, an adaptive cubic convolution interpolation algorithm is used:

[0035]

[0036] Where B3 is the cubic B-spline basis function, (Δx, Δy) is the target grid spacing. This algorithm can achieve resolution standardization while maintaining the geometric topology of the geological body. The physical dimensions are uniformly scaled using the feature scaling guided by dimensional analysis. For permeability (10 -15 -10 -12 m 2 ), porosity (0-1), seismic wave velocity (2000-6000m / s) and other parameters to construct a dimensionless conversion:

[0037]

[0038] where μ φ and σ φ are the mean and standard deviation of porosity, K ref =1mD is the reference permeability. To eliminate scale deviation, wavelet multi-resolution analysis is used to construct the scale space mapping function:

[0039]

[0040] The Daubechies wavelet basis function ψ is selected to decompose the multi-scale data. j ) to achieve feature alignment through energy spectrum matching. Finally, the data volume with the same scale is reconstructed through inverse transformation, and its fidelity is determined by the reconstruction error Control, requiring ò < 5% to meet the accuracy requirements of geological modeling. The entire process requires iterative optimization until data of different scales are in the gradient field. Statistical consistency is achieved at the geometric feature level such as curvature field (κ).

[0041] Step 2: Data fusion: Integrate multi-source heterogeneous data (such as seismic, well logging, core, outcrop, and dynamic production data), eliminate scale differences and data conflicts, and build a self-consistent geological model.

[0042] The core of the data fusion stage is to construct a joint probability space for multi-source heterogeneous data and achieve self-consistent modeling by establishing a variational assimilation framework under geological constraints. For data types such as seismic reflection amplitude, well log curves, core slice images, outcrop profile sketches, and production well flow pressure, the joint likelihood function of multimodal data is first established:

[0043]

[0044] Where m is the geological model parameter to be determined, d i represents the i-th type of observation data, g i (·) is the corresponding forward operator (such as seismic wave equation solver, well logging response function), C iis the covariance matrix of each data type. In order to eliminate the scale difference, a multi-resolution data matching strategy is adopted: constrained cokriging interpolation is implemented for kilometer-level seismic data:

[0045]

[0046] where λ α and μ β is the Kriging weight, is the Laplace operator, which forces the interpolation result to meet the structural trend of seismic interpretation. For centimeter-level core CT data, a microstructure optimization model based on Markov random field is constructed:

[0047]

[0048] Where V c (·) is the cluster potential energy function that characterizes the pore topology constraint, φ p is the local porosity calculation function, and γ is the matching weight of the core measured data. The dynamic production data is integrated through the inversion of the adjoint equation to establish the historical matching objective function of the pressure-saturation field:

[0049]

[0050] Where Q is the observation error covariance, β is the regularization parameter, and R is the prior covariance of the model parameters. To address the data conflict problem, the fuzzy logic conflict factor is introduced:

[0051]

[0052] When ij When the value is greater than 0.3, the conflict resolution mechanism is triggered, and the optimization is performed by the alternating direction multiplier method (ADMM) under the constraints of the rock physics cross-plot:

[0053] min m,z f(m)+g(z)stAm+Bz=c

[0054] Where f(·) represents the seismic and logging data fitting term, g(·) is the core microstructure constraint term, and A and B are the upscaling / downscaling conversion matrices between different scales. Finally, a multi-scale fusion model is constructed through a mixed Gaussian process:

[0055]

[0056] in Represents the convolution coupling operator between scales, with weight w k Dynamically adjusted by the data signal-to-noise ratio. This process needs to be iterated until the residuals of each data meet The statistical consistency criterion is used to ensure the compatibility of multi-dimensional geological rules such as structural style, sedimentary sequence, and fluid dynamics.

[0057] Step 3: Cross-scale nesting and fusion:

[0058] First, models of different scales are modeled in layers: a basin-scale structural framework model (such as faults and stratigraphic interfaces) is established based on seismic interpretation data, and a reservoir model (such as sedimentary facies and fracture networks) is embedded in the structural framework based on drilling data. Microstructural characteristics such as pore structure and mineral distribution are characterized through digital core or CT scan data.

[0059] Then, cross-scale nesting and fusion of models of different scales are performed:

[0060] The geological models of different scales are interconnected through mathematical or physical constraints (physical constraint fusion), seismic attributes and well logging data are combined through geostatistics (such as sequential Gaussian simulation and multi-point geostatistics), and the microscopic pore structure and macroscopic reservoir parameters are linked using rock physics models (such as porosity-permeability relationship).

[0061] Adaptive grid technology (such as unstructured grid or locally encrypted grid) is used to achieve high-resolution nesting in key areas (such as fracture zones and fluid contact surfaces), forming a hierarchical cross-scale nested structure with different spatial resolutions.

[0062] The core of cross-scale nesting and fusion lies in establishing a bidirectional coupling mechanism between multi-resolution geological models. In the hierarchical modeling stage, a basin-scale structural framework is first constructed based on the seismic interpretation surface, using radial basis function surface reconstruction technology with fault constraints:

[0063]

[0064] where φ(r) = r 3 is the cubic radial basis kernel function, p j The (x,y) polynomial terms are used to eliminate the effects of rigid body displacement. The coefficients are determined by solving the linear equation system Φα+Pβ=z to ensure that the formation interface maintains \(C^2\) continuity on both sides of the fault. Reservoir-scale modeling uses a multi-point geostatistical simulation algorithm, and its conditional probability is expressed as:

[0065]

[0066] Where T k The topological pattern extracted from the training image, V c (·) is the potential function based on distance transformation. Microscopic pore modeling adopts the fractal-Markov coupling algorithm, and the porosity field generation follows:

[0067]

[0068] Where D fis the fractal dimension of the scan data fitting, a n is the energy spectrum coefficient. Cross-scale fusion is achieved through multi-physics field coupling, and bidirectional constraints are established between the macroscopic Darcy flow and the microscopic Stokes flow:

[0069]

[0070] Establishing the permeability tensor through homogenization theory The adaptive meshing technique uses the local refinement criterion of the unstructured tetrahedral mesh:

[0071]

[0072] When the gradient index η of unit e e When the threshold is exceeded, h-refinement is triggered, achieving millimeter-level grid nesting in key areas such as fracture zones. Deep learning-assisted fusion uses a multi-scale convolutional adversarial network architecture. The generator G includes a macro path (5 layers of dilated convolution to extract basin features) and a micro path (3 layers of U-Net to process pore structure). The discriminator \(\D\) improves stability through spectral normalization constraints. The loss function is designed as:

[0073]

[0074] where F(·) is the rock physics forward operator and λ controls the geological rule constraint strength. The final model is solved using the multiscale finite element method in a unified framework, and the stiffness matrix is ​​assembled using a scale progression strategy:

[0075]

[0076] Where C s,s+1 is the constraint matrix between scales s and s+1, ensuring that physical quantities such as pressure field and saturation field meet the flux continuity condition at the cross-scale interface. The fusion process must meet the multi-scale convergence criterion max(||R s || / ||F s ||)<10 -4 At the same time, the consistency of the Betti numbers of pore networks at different scales was verified through topological data analysis, and finally a seamless nested geological model was formed from kilometer-scale basin structure to micron-scale pore structure.

[0077] Step 4: Dynamic Coupling: Implement bidirectional feedback and iterative optimization between models of different resolutions through cross-scale parameter transfer (e.g., upscaling and downscaling algorithms). Dynamically adjust model parameters through data assimilation (e.g., ensemble Kalman filtering) to ensure multi-scale consistency.

[0078] The core of dynamic coupling lies in establishing a bidirectional transmission mechanism and real-time correction system for multi-scale model parameters. In terms of cross-scale parameter transmission, a volume-averaged upscaling algorithm is used to map microscopic pore structure parameters to the reservoir scale:

[0079]

[0080] Where w(x) is the microscopic perturbation potential function solved under periodic boundary conditions, and the characteristic unit problem is discretized by finite element method. The pore-reservoir permeability correlation is achieved. The downscaling process is achieved through random field condition simulation:

[0081] m fine =G down m coarse +Lz,z~N(0,I)

[0082] The downscaling operator G down Constructed by wavelet multi-resolution analysis, L is the one that satisfies LL T =C fine|coarse The Cholesky decomposition matrix of ensures the spatial consistency of fine-scale parameters and coarse-scale dominant trends. Dynamic feedback is achieved through a coupled adjoint system to construct a macro-micro joint objective function:

[0083]

[0084] Where T s→s′ is the inter-scale conversion operator, Q ss′ Characterize the covariance of cross-scale parameter differences. The dynamic assimilation process of the Ensemble Kalman Filter (EnKF) is achieved through multi-scale state vector expansion:

[0085]

[0086] The cross-scale covariance matrix Explicitly include the off-diagonal coupling terms of macroscopic and microscopic parameters. s The nonlinearity of (·) is improved by using the localized set transform Kalman filter (LETKF):

[0087]

[0088] Weight matrix W k It is determined by solving the localized cost function minimum value, where the observation disturbance term Contains multi-scale data covariance information. The iterative optimization adopts a dual grid strategy to solve the global update amount on the coarse-scale grid:

[0089]

[0090] Then, through the restricted interpolation operator Transfer to a fine-scale grid for local correction:

[0091]

[0092] Among them S f is a fine-scale smoothing operator, r f =d f -H f (m f ) is the local residual. Multi-scale convergence is controlled by the joint criterion of relative parameter change rate and data residual:

[0093]

[0094] At the same time, the cross-scale compatibility index is introduced As the iteration termination condition, when χ cross <δ tol The multiscale system is then judged to have reached dynamic equilibrium. This coupling process ultimately forms an integrated geological-engineering model with temporal and spatial continuity, whose parameter fields simultaneously satisfy rock physics laws and dynamic production responses at scales from kilometers to micrometers.

[0095] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing related hardware through a computer program. The program can be stored in a computer-readable storage medium, and when executed, the program can include the processes in the above-described method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0096] It should be understood that the detailed description of the technical solutions of the present invention using the preferred embodiments above is illustrative and not restrictive. A person skilled in the art, after reading the present specification, may modify the technical solutions described in the embodiments or replace some of the technical features therein with equivalents; such modifications or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A cross-scale nesting and fusion modeling method for a multi-scale geological model, characterized in that: The following steps are involved: Step 1: Multi-scale data preprocessing to unify the coordinate system, resolution and physical dimension and eliminate scale deviation between data; Step 2: Data fusion, which integrates multi-source heterogeneous data through joint probability space, variational assimilation framework, fuzzy logic conflict factor and alternating direction multiplier method optimization; Step 3: Cross-scale nesting and fusion: hierarchical modeling of geological models at different scales, and establishment of a bidirectional coupling mechanism through adaptive grid technology and multi-physics field coupling; Step 4: Dynamic coupling, through cross-scale parameter transfer, data assimilation, ensemble Kalman filtering and localized ensemble transform Kalman filtering, to achieve bidirectional feedback and iterative optimization between models.

2. The cross-scale nesting and fusion modeling method of a multi-scale geological model according to claim 1, characterized in that: In step 1, the coordinate system is unified by using the Helmert transformation, which includes the comprehensive application of scale factors, rotation matrices and translation vectors to eliminate spatial deviations from different measurement systems. At the same time, the resolution is unified by using an adaptive cubic convolution interpolation algorithm, which uses cubic B-spline basis functions for interpolation to achieve resolution standardization of data of different scales while maintaining the geometric topology of the geological body.

3. The cross-scale nesting and fusion modeling method of a multi-scale geological model according to claim 1, characterized in that: In step 1, a feature scaling technique guided by dimensional analysis is used to achieve the unification of physical dimensions, specifically including the dimensionless conversion of permeability, porosity, and seismic velocity parameters. By calculating the mean and standard deviation of these parameters, the comparability of the data in different physical dimensions is ensured.

4. The cross-scale nesting and fusion modeling method of a multi-scale geological model according to claim 1, characterized in that: In step 2, data fusion is performed using joint likelihood function, cokriging interpolation and Markov random field, combined with fuzzy logic conflict factor and alternating direction multiplier method optimization to solve the conflict problem between multi-source heterogeneous data, thereby achieving effective integration and consistency of data.

5. The cross-scale nesting and fusion modeling method of a multi-scale geological model according to claim 1, characterized in that: In step 2, a multi-scale fusion model is constructed by mixing Gaussian processes to ensure statistical consistency at different scales and data types, and the flexibility and adaptability of Gaussian processes are used to handle uncertainty, thereby improving the robustness and accuracy of the model.

6. The cross-scale nesting and fusion modeling method of a multi-scale geological model according to claim 1, characterized in that: In step 3, adaptive grid technology and multi-physics field coupling are used, combined with deep learning-assisted fusion technology, such as multi-scale convolutional adversarial networks, to improve the stability and accuracy of the model and ensure the effectiveness of bidirectional coupling and information transfer between geological models of different scales.

7. The cross-scale nesting and fusion modeling method of a multi-scale geological model according to claim 1, characterized in that: In step 4, the volume average upscaling algorithm and random field condition simulation are used to ensure parameter consistency between models of different resolutions, and dynamic adjustment is performed using ensemble Kalman filtering and localized ensemble transform Kalman filtering to achieve two-way feedback and iterative optimization between models.

8. The cross-scale nesting and fusion modeling method of a multi-scale geological model according to claim 1, characterized in that: In step 4, a dual grid strategy and restrictive interpolation are used to achieve multi-scale convergence, forming a spatiotemporally continuous geological-engineering integrated model. The relative parameter change rate and data residual are combined to control the convergence and dynamic balance of the model at the kilometer to micrometer scale.

Images