Data processing system and method for fault zone exploration

Abstract

The invention provides a data processing system and method for fault zone survey, and belongs to the technical field of seismic survey, and the system comprises a quantum enhanced sensing module, an edge intelligent calculation module and a multi-physics field modeling module. The quantum enhanced sensing module comprises a superconducting quantum interference magnetic gradiometer unit and a diamond NV color center strain sensor array unit; the edge intelligent calculation module comprises an FPGA self-adaptive noise reduction unit and a space-time data compression engine unit; the multi-physics field modeling module comprises a quantum-classical hybrid inversion engine unit and a dynamic fracture expansion simulator unit. According to the data processing system and method for fault zone exploration, the data resolution is improved, multi-source data fusion is simple, the real-time performance is high, the algorithm generalization ability is high, noise interference is reduced, the signal separation condition is relieved, the geologic model is complete, the cost is low, the automation degree is high, and the dynamic monitoring ability is high.

Description

Technical Field

[0001] The present invention relates to the technical field of seismic surveys, and in particular to a data processing system and method for fault zone surveys. Background Art

[0002] The existing data processing systems and methods for fault zone surveys have the following problems in terms of technology, efficiency, and application:

[0003] 1. Data acquisition and accuracy issues:

[0004] (1) Insufficient data resolution:

[0005] Traditional seismic surveys, gravity / magnetic exploration, etc. are limited by instrument accuracy or environmental noise (such as urban interference), making it difficult to capture the fine structures of micro-faults or deep faults.

[0006] (2) Difficulty in multi-source data fusion:

[0007] Data sources such as seismic, InSAR (Interferometric Synthetic Aperture Radar), GNSS (Global Navigation Satellite System), and low-quality surveys are heterogeneous, with inconsistent spatio-temporal scales and a lack of a unified data fusion framework.

[0008] (3) Poor real-time performance:

[0009] Most surveys rely on historical data or periodic acquisitions (such as satellite remote sensing), making it difficult to achieve real-time monitoring of fault zone activities.

[0010] 2. Data processing and analysis bottlenecks

[0011] (1) Insufficient algorithm generalization ability:

[0012] Existing algorithms (such as inversion algorithms, machine learning models) are mostly designed for specific regions or fault types, and the accuracy decreases when applied across regions.

[0013] (2) Challenges in noise interference and signal separation:

[0014] Fault zone signals are often masked by environmental noise (such as artificial vibrations, climate factors), and traditional filtering methods (such as wavelet transform) may lose effective signals.

[0015] (3) High computational complexity:

[0016] High-resolution 3D modeling or full-waveform inversion requires supercomputer support and is difficult to achieve fast analysis on ordinary hardware platforms.

[0017] 3. Limitations of models and interpretations

[0018] (1) Excessive simplification of geological models:

[0019] Most systems assume that the fault zone is a homogeneous medium or has a simple geometry, ignoring complex factors such as anisotropy and pore fluids, resulting in prediction deviations.

[0020] (2) Lack of uncertainty quantification:

[0021] There is a lack of systematic evaluation of data errors and model parameter uncertainties, affecting the reliability of risk assessment.

[0022] (3) Insufficient multidisciplinary collaboration:

[0023] Models in fields such as geophysics, geology, and engineering operate independently and lack cross-validation (e.g., there is no unified interpretation framework when seismic data conflicts with surface deformation data).

[0024] 4. Challenges in real-time application of technology

[0025] (1) Cost and accessibility:

[0026] High-precision exploration (such as UAV LiDAR and dense seismic networks) is costly and difficult to popularize.

[0027] (2) Low degree of automation:

[0028] Fault identification and parameter extraction still rely on manual intervention (such as seismic profile interpretation), which is highly subjective and inefficient.

[0029] (3) Weak dynamic monitoring ability:

[0030] Existing systems mostly focus on static structure analysis and have limited means for monitoring dynamic processes such as creep and stress accumulation in fault zones.

[0031] Therefore, there is an urgent need in this field for a technical solution that can solve the problems existing in the chain from data collection to decision support for fault zone exploration.

[0032] The information disclosed in this background section is only intended to enhance the overall understanding of the present invention and should not be regarded as an admission or any form of suggestion that this information constitutes prior art already known to those of ordinary skill in the art. Summary of the Invention

[0033] The object of the present invention is to provide a technical solution that can solve the problems existing in the chain from data collection to decision support for fault zone exploration.

[0034] To achieve the above object, the present invention provides the following solution:

[0035] A data processing system for fault zone exploration, comprising:

[0036] A quantum-enhanced sensing module, an edge intelligent computing module, and a multi-physics field modeling module;

[0037] The quantum enhanced sensing module includes:

[0038] A superconducting quantum interference magnetometer unit for detecting nano-Tesla level magnetic anomalies generated by microfractures. The quantum flux locking equation is:

[0039] where φ total is the total magnetic flux; n is an integer representing the magnetic flux quantum number; φ0 is the magnetic flux quantum; φ0 = h / (2e) = 2.07×10 -15 Wb; φ ext is the external magnetic flux; α is the geometric sensitivity factor, α = 1.7×10 -8 m 3 ; is the second derivative of the magnetic field, and ΔV is the volume change;

[0040] A diamond NV color center strain sensor array unit for measuring lattice strain with an accuracy of 10 -9 The formula is: the spin-strain coupling Hamiltonian:

[0041]

[0042] where λ⊥ = 3.7MH Z / strain: the transverse spin-strain coupling coefficient; λ|| is the longitudinal spin-strain coupling coefficient; ε xy is the shear strain component; ε zz is the normal strain component; S x , S y , S z are the spin operator components;

[0043] The edge intelligent computing module includes:

[0044] An FPGA adaptive noise reduction unit for suppressing environmental noise in real time. The core algorithm is: the dual-domain hybrid filtering algorithm:

[0045]

[0046] where, is the estimated signal spectrum after noise reduction; W(f,t) is the wavelet-Wiener hybrid weight matrix;

[0047] |X(f,t)| 2 is the power spectrum of the original signal; β(t) is the time-varying regularization factor, β(t) = 1 - e -t / τ ; τ = 0.83s is the time constant; N est (f,t) is the estimated noise power spectrum, with the unit: V 2 / Hz;

[0048] A spatio-temporal data compression engine unit for compressing data with a data compression ratio ≥ 100:1;

[0049] The core algorithm is: fractal dictionary learning algorithm:

[0050]

[0051] where Y is the observed data matrix, D is the fractal dictionary matrix, and X is the sparse coefficient matrix. is the L1 / 2 norm; ||xi|| 1 / 2 = ∑ j |x ij | 1 / 2 is the L1 / 2 non-convex sparse regularization term number; λ = 0.47 is the regularization parameter.

[0052] The multi-physics field modeling module includes:

[0053] A quantum-classical hybrid inversion engine unit for jointly inverting seismic wave velocity, resistivity, and gravity anomaly;

[0054] The core algorithm is:

[0055] Hybrid objective function:

[0056]

[0057] where J(m) is the inversion objective function; w k is the multi-physics field weight, ∑w k = 1; d k is the k-th geophysical observation data; G k (m) is the forward model of the k-th physical field; Total variation regularization; ||m|| QP = ∑ i,j |m i -m j | 0.7 : Quantum potential regularization; μ, v are regularization coefficients;

[0058] A dynamic fracture propagation simulator unit for predicting the evolution trend of fracture zones;

[0059] The core algorithm is: fractional order phase field model:

[0060]

[0061] where φ is the phase field variable, φ ∈ [0, 1]; M φ is the phase field mobility; ε is the gradient energy coefficient; is the fractional order gradient operator; ψ(φ) is the double well free energy function.

[0062] A data processing method for fault zone survey, characterized by comprising:

[0063] Data fusion and alignment; including: hyperdimensional spatio-temporal registration and multi-source uncertainty propagation;

[0064] Fault intelligent identification; including physical constraint and establishment of a multi-scale graph convolutional network;

[0065] Dynamic risk assessment; for dynamic risk assessment.

[0066] Optionally, the formula for the hyperdimensional spatio-temporal registration is:

[0067]

[0068] Wherein, is the spatio-temporal transformation function; P=(x, y, z) is the spatial coordinate; c k is the center point of the k-th Gaussian kernel; σ k =1.7·2 k-1 is the Gaussian kernel width; w k is the time frequency component; φ k is the phase offset.

[0069] Optionally,

[0070] . The formula for the multi-source uncertainty propagation is:

[0071]

[0072] Wherein, ρ ij is the correlation coefficient between parameters.

[0073] Optionally, the formula for the physical constraint is;

[0074]

[0075] Wherein, Q, K, V: query, key, value matrices; d k is the key vector dimension; physical constraint term; β is the constraint strength coefficient.

[0076] Optionally, the formula for establishing the multi-scale convolutional network is:

[0077]

[0078] Wherein, c ij =||r i -r j || 0.7 : geometric normalization factor;

[0079] γ kis the multi-scale fusion weight.

[0080] Optionally, the formula for the dynamic risk assessment is:

[0081]

[0082] where P failure is the fracture instability probability;

[0083] β = 1 / k B T eff is the inverse temperature parameter;

[0084] I fail is the failure indicator function;

[0085]

[0086] where LE max is the maximum probability exponent, unit: 1 / s;

[0087] J(X k ) is the probability matrix;

[0088] Δt is the time step, unit: s.

[0089] Compared with the prior art, the present invention has the following beneficial effects:

[0090] The data processing system and method for fracture zone survey provided by the present invention improve the data resolution, the multi-source data fusion is simple, has strong real-time performance, strong algorithm generalization ability, reduces noise interference, alleviates signal separation, has a complete geological model, low cost, high automation degree, and strong dynamic monitoring ability. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0092] Figure 1 is the schematic structural diagram of the system provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0093] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0094] The object of the present invention is to provide a technical solution that can solve the problems existing in the chain from data acquisition to decision support for fault zone survey.

[0095] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0096] Embodiment 1:

[0097] This embodiment provides a data processing system for fault zone survey, as Figure 1 shown, including:

[0098] A quantum enhanced sensing module, an edge intelligent computing module, and a multi-physical field modeling module;

[0099] The quantum enhanced sensing module includes:

[0100] A superconducting quantum interference magnetometer unit for detecting nano-Tesla-level magnetic anomalies generated by microfractures. The quantum flux-locked equation is:

[0101] where φ total is the total magnetic flux; n is an integer representing the flux quantum number; φ0 is the flux quantum; φ0 = h / (2e) = 2.07×10 -15 Wb; φ ext is the external magnetic flux; α is the geometric sensitivity factor, α = 1.7×10 -8 m 3 ; is the second derivative of the magnetic field, and ΔV is the volume change;

[0102] A diamond NV center strain sensor array unit for measuring lattice strain with an accuracy of 10 -9 ; The formula is: spin-strain coupling Hamiltonian:

[0103]

[0104] where λ⊥ = 3.7MH Z / strain: transverse spin-strain coupling coefficient; λ|| is the longitudinal spin-strain coupling coefficient; ε xy is the shear strain component; ε zzis the normal strain component; S x , S y , S z is the spin operator component;

[0105] The edge intelligent computing module includes:

[0106] The FPGA adaptive noise reduction unit is used to suppress environmental noise in real time, and the core algorithm is: the dual-domain hybrid filtering algorithm:

[0107]

[0108] Among them, is the signal spectrum estimation after noise reduction; W(f,t) is the wavelet-Wiener hybrid weight matrix;

[0109] |X(f,t)| 2 is the original signal power spectrum; β(t) is the time-varying regularization factor, β(t) = 1 - e -t / τ ; τ = 0.83s is the time constant; N est (f,t) is the noise power spectrum estimation, and the unit is: V 2 / Hz;

[0110] The spatio-temporal data compression engine unit is used to compress data, and the data compression ratio ≥ 100:1;

[0111] The core algorithm is: the fractal dictionary learning algorithm:

[0112]

[0113] Among them, Y is the observed data matrix, D is the fractal dictionary matrix, X is the sparse coefficient matrix, is the L1 / 2 norm; ||xi|| 1 / 2 = ∑ j |x ij | 1 / 2 is the L1 / 2 non-convex sparse regularization term number; λ = 0.47 is the regularization parameter;

[0114] The multi-physical field modeling module includes:

[0115] The quantum-classical hybrid inversion engine unit is used to jointly invert seismic wave velocity, resistivity, and gravity anomaly;

[0116] The core algorithm is:

[0117] The hybrid objective function:

[0118]

[0119] Among them, J(m) is the inversion objective function; w k is the multi-physical field weight, ∑wk = 1; d k is the k-th kind of geophysical observation data; G k (m) is the forward model of the k-th physical field; Total variation regularization; ||m|| QP = ∑ i,j |m i -m j | 0.7 : Quantum potential regularization; μ, v are regularization coefficients;

[0120] Dynamic fracture propagation simulator unit for predicting the evolution trend of the fracture zone;

[0121] The core algorithm is: Fractional phase field model:

[0122]

[0123] where φ is the phase field variable, φ ∈ [0, 1]; M φ is the phase field mobility; ε is the gradient energy coefficient; is the fractional gradient operator; ψ(φ) is the double-well free energy function.

[0124] A data processing method for fracture zone survey, including:

[0125] Data fusion and alignment; including: Hyperdimensional spatio-temporal registration and multi-source uncertainty propagation;

[0126] Fracture intelligent identification; including physical constraints and establishing a multi-scale graph convolutional network;

[0127] Dynamic risk assessment; for dynamic risk assessment.

[0128] In one embodiment, the formula for the hyperdimensional spatio-temporal registration is:

[0129]

[0130] where, is the spatio-temporal transformation function; P = (x, y, z) is the spatial coordinate; c k is the center point of the k-th Gaussian kernel; σ k = 1.7·2 k-1 is the Gaussian kernel width; w k is the time frequency component; φ k is the phase shift.

[0131] In one embodiment, the formula for the multi-source uncertainty propagation is:

[0132]

[0133] where ρij is the correlation coefficient between parameters.

[0134] In one embodiment, the formula for physical constraint is:

[0135]

[0136] where Q, K, V are query, key, and value matrices; d k is the dimension of the key vector; is the physical constraint term; β is the constraint strength coefficient.

[0137] In one embodiment, the formula for establishing a multi-scale convolutional network is:

[0138]

[0139] where c ij = ||r i - r j || 0.7 : geometric normalization factor;

[0140] γ k is the multi-scale fusion weight.

[0141] In one embodiment, the formula for dynamic risk assessment is:

[0142]

[0143] where P failure is the fracture instability probability;

[0144] β = 1 / k B T eff is the inverse temperature parameter;

[0145] I fail is the failure indicator function;

[0146]

[0147] where LE max is the maximum probability exponent, unit: 1 / s;

[0148] J(X k ) is the probability matrix;

[0149] Δt is the time step, unit: s.

[0150] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple. For the relevant parts, reference can be made to the description in the method section.

[0151] In this article, specific examples are used to elaborate on the principles and implementation manners of the present invention. The descriptions of the above embodiments are only used to help understand the method of the present invention and its core idea. At the same time, for those of ordinary skill in the art, based on the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation on the present invention.

Claims

1. A data processing system for fracture zone survey, characterized in that, Including: A quantum-enhanced sensing module, an edge intelligent computing module, and a multi-physical field modeling module; The quantum-enhanced sensing module includes: A superconducting quantum interference magnetometer unit for detecting nano-Tesla magnetic anomalies generated by microfractures. The quantum flux-locked equation is: where φ total is the total magnetic flux; n is an integer representing the magnetic flux quantum number; φ0 is the magnetic flux quantum; φ0 = h / (2e) = 2.07×10 -15 Wb; φ ext is the external magnetic flux; α is the geometric sensitivity factor, α = 1.7×10 -8 m 3 ; is the second derivative of the magnetic field, and ΔV is the volume change; A diamond NV color center strain sensor array unit for measuring lattice strain up to 10 -9 precision; the formula is: spin-strain coupling Hamiltonian: where λ⊥ = 3.7 MH Z / strain: transverse spin-strain coupling coefficient; λ|| is the longitudinal spin-strain coupling coefficient; ε xy is the shear strain component; ε zz is the normal strain component; S x , S y , S z are the spin operator components; The edge intelligent computing module includes: An FPGA adaptive noise reduction unit for suppressing environmental noise in real time, with the core algorithm being: a dual-domain hybrid filtering algorithm: Among them, is the estimated signal spectrum after noise reduction; W(f, t) is the wavelet-Wiener hybrid weight matrix; |X(f,t)| 2 is the original signal power spectrum; β(t) is the time-varying regularization factor, β(t) = 1 - e -t / τ ; τ = 0.83 s is the time constant; N est (f,t) is the noise power spectrum estimate, with the unit: V 2 / Hz; A spatio-temporal data compression engine unit for compressing data, with a data compression ratio ≥ 100:1; The core algorithm is: a fractal dictionary learning algorithm: Among them, Y is the observed data matrix, D is the fractal dictionary matrix, and X is the sparse coefficient matrix. is the L1 / 2 norm; ||xi|| 1 / 2 = ∑ j |x ij | 1 / 2 is the L1 / 2 non-convex sparse regularization term number; λ = 0.47 is the regularization parameter. The multi-physical field modeling module includes: A quantum-classical hybrid inversion engine unit for jointly inverting seismic wave velocity, resistivity, and gravity anomaly; The core algorithm is: A hybrid objective function: Among them, J(m) is the inversion objective function; w k is the multi-physical field weight, ∑w k = 1; d k is the k-th geophysical observation data; G k (m) is the forward model of the k-th physical field; Total variation regularization; ||m|| QP = ∑ i,j |m i - m j | 0.7 : Quantum potential regularization; μ, v are regularization coefficients; A dynamic fracture propagation simulator unit for predicting the evolution trend of the fracture zone; The core algorithm is: a fractional-order phase field model: where φ is the phase-field variable, φ ∈ [0, 1]; M φ is the phase-field mobility; ε is the gradient energy coefficient; is the fractional gradient operator; ψ(φ) is the double-well free energy function.

2. A data processing method for fracture zone survey, characterized in that, Including: Data fusion and alignment; including: hyper-dimensional spatio-temporal registration and multi-source uncertainty propagation; Intelligent fracture identification; including imposing physical constraints and establishing a multi-scale graph convolutional network; Dynamic risk assessment; for dynamic risk assessment.

3. The data processing method for fracture zone survey according to claim 2, wherein The formula for the hyper-dimensional spatio-temporal registration is: Among them, is a spatio-temporal transformation function; P = (x, y, z) is a spatial coordinate; c k is the center point of the k-th Gaussian kernel; σ k = 1.7·2 k-1 is the Gaussian kernel width; w k is the time frequency component; φ k is the phase shift.

4. The data processing method for fracture zone survey according to claim 2, wherein The formula for the multi-source uncertainty propagation is: Among them, ρ ij is the correlation coefficient between parameters.

5. The data processing method for fracture zone survey according to claim 2, wherein The formula for imposing the physical constraints is; where Q, K, V: query, key, and value matrices; d k is the key vector dimension; is the physical constraint term; β is the constraint strength coefficient.

6. The data processing method for fracture zone survey according to claim 2, wherein, The formula for establishing the multi-scale convolutional network is: where c ij = ||r i - r j || 0.7 : geometric normalization factor; γ k is the multi-scale fusion weight.

7. The data processing method for fracture zone survey according to claim 2, characterized in that, The formula for the dynamic risk assessment is: Among them, P failure is the fracture instability probability; β = 1 / k B T eff is the inverse temperature parameter; I fail is a failure indication function; Among them, LE max is the maximum probability exponent, unit: 1 / s; J(X k ) is a probability matrix; Δt is the time step, unit: s.

Images