Genesis and Migration Path Tracing Methods of Deep Geothermal Fluids

Abstract

This invention discloses a method for tracing the genesis and migration paths of deep geothermal fluids. This invention relates to the field of Earth science exploration technology and specifically includes the following steps: A1, collecting geothermal fluid samples; A2, fusing high-precision 3D seismic data, well logging data, and borehole data to construct a multi-scale fracture network model, quantifying fracture connectivity and permeability anisotropy; A3, developing a Long Short-Term Memory (LSTM) network model, inputting real-time monitored temperature, pressure, and chemical field data, dynamically predicting fluid migration paths, and introducing a Generative Adversarial Network (GAN) to optimize the model; A4, deploying a distributed fiber optic sensor network. This sensing method for high-speed moving targets in airborne applications, using core-shell structured tracers, balances high fluorescence efficiency and geological compatibility, forming a complete innovation chain in the material design, detection methods, and environmental adaptability of nano-tracers, significantly improving the accuracy and sustainability of tracing the migration paths of deep geothermal fluids.

Description

Technical Field

[0001] This invention relates to the field of Earth science exploration technology, specifically to a method for tracing the genesis and migration paths of deep geothermal fluids. Background Technology

[0002] As a clean and renewable energy source, deep geothermal resources hold significant strategic importance for alleviating the energy crisis and promoting green and sustainable development in the context of global energy structure transformation. Accurately revealing the genesis and migration pathways of deep geothermal fluids is a crucial prerequisite for the scientific development and rational utilization of geothermal resources.

[0003] The development of deep geothermal resources relies on the accurate identification of the origin and migration path of geothermal fluids. Existing technologies have the following shortcomings:

[0004] 1. Traditional methods rely on single isotopes or geochemical parameters, which are easily affected by the mixing of multiple fluid sources, leading to errors in gene determination:

[0005] 2. Numerical simulation based on two-dimensional geological profiles cannot reflect the fluid transport patterns in complex three-dimensional fracture networks, and lacks the ability to track the real-time changes in temperature-pressure-chemical field coupling during geothermal fluid transport. Therefore, this invention provides a method for tracing the genesis and transport path of deep geothermal fluids. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method for tracing the genesis and migration path of deep geothermal fluids, which solves the problem that traditional methods rely on single isotopes or geochemical parameters and are easily affected by the mixing of multiple fluid sources, leading to errors in gene determination.

[0007] To achieve the above objectives, the present invention provides a method for tracing the genesis and migration path of deep geothermal fluids, specifically comprising the following steps:

[0008] A1. Collect geothermal fluid samples, determine isotopic composition, rare gas composition, trace element fingerprint and water-rock reaction index, and establish a multidimensional feature database;

[0009] A2. By integrating high-precision 3D seismic data, well logging data and borehole data, a multi-scale fracture network model is constructed to quantify fracture connectivity and permeability anisotropy. Geochemical fingerprint data is coupled with fracture network parameters, and a fluid transport probability field is generated through Monte Carlo simulation.

[0010] A3. Develop a long short-term memory network model, input real-time monitored temperature, pressure, and chemical field data, dynamically predict fluid transport paths, introduce an adversarial generative network to optimize the model, and generate virtual path scenarios through training with historical data to improve prediction robustness.

[0011] A4. Deploy a distributed fiber optic sensor network to achieve synchronous monitoring of multiple parameters, including temperature, strain, and chemical field.

[0012] The determination of isotopic components includes: δ 18 O, δ 2 H, 3 He / 4 He 87 Sr / 86 Sr, rare gas components include: He, Ne, Ar, trace element fingerprints include: Li, B, As, Sb, water-rock reaction indicators include: SiO2, pH, Eh, the Al is introduced with an environmentally friendly nano-tracer, and high-sensitivity in-situ detection is achieved through laser-induced breakdown spectroscopy;

[0013] The specific method for achieving high-sensitivity in-situ detection by introducing an environmentally friendly nano-tracer and using laser-induced breakdown spectroscopy is as follows:

[0014] B1. CdSe / ZnS core-shell quantum dots are selected, emitting wavelengths covering the visible to near-infrared region. Multicolor encoding is achieved by controlling the particle size, and an improved method is employed. The sol-gel method deposits a dense SiO2 shell on the surface of quantum dots to form a QDs@SiO2 core-shell structure, which prevents the quantum dots from directly contacting the external environment.

[0015] B2. The surface of SiO2 is modified with a silane coupling agent to introduce active functional groups, enhancing the dispersibility of nanoparticles in aqueous solution and their compatibility with geothermal fluid components. An outer layer of polyethylene glycol or zwitterionic polymer is then applied to reduce non-specific adsorption of nanoparticles in the geothermal system and extend the circulation time. Nano-tracer injection and migration monitoring are then performed. Based on the geothermal system volume and monitoring requirements, nano-tracers are injected at a concentration of 0.1-1 ppm, reducing the total amount used by more than 90% compared to traditional dyes. A dedicated injection pump is used to slowly inject the tracer solution into the target layer along the borehole wall, avoiding disturbance to the fracture network. Fiber optic cables are deployed along the geothermal wellbore and potential migration paths, with monitoring points spaced 10-50 m apart, enabling simultaneous acquisition of temperature, strain, and chemical fields. A portable LIBS system is installed at the key monitoring wellhead, with a laser wavelength of 1064 nm, pulse energy of 10 mJ, and a focused spot diameter of <100 μm, allowing it to penetrate the wellbore fluid and directly excite nanoparticles.

[0016] B3. Laser excitation: A high-energy pulsed laser is focused onto the fluid sample, instantly vaporizing and ionizing nanoparticles to generate plasma. During the plasma cooling process, characteristic spectra are emitted and dispersed by a spectrometer. A CCD detector records the spectral signals. The quantum dot emission peaks include Cd, Se, and Zn element lines, used to quantify tracer concentration. The SiO2 matrix peaks include Si and O lines, used to verify the integrity of the nanoparticles. Finally, a calibration curve of spectral intensity versus tracer concentration is established. Multivariate linear regression is used to eliminate matrix effect interference. Combined with temperature and strain data from fiber optic sensors, a particle filtering algorithm is used to reconstruct the nanoparticle migration trajectory, achieving a spatial resolution of 1m.

[0017] Preferably, the specific method for constructing the multi-scale fracture network model and quantifying fracture connectivity and permeability anisotropy is as follows:

[0018] C1. Macroscopic fracture distribution is constrained by 3D seismic data. Spatial locations of fracture zones are extracted through seismic attribute analysis. Sequence stratigraphy is used to delineate the top and bottom interfaces of thermal reservoirs, constructing a macroscopic fracture network framework. Then, well logging data is used to correct mesoscopic fracture density: imaging logging (FMI, UBI) is used to identify the wellbore fracture orientation and aperture (0.1-10 mm). Combined with sonic full-wavelength logging, fracture filling types are inverted, establishing a fracture density decay model with depth (exponential function: ρ(z)=ρ0e^(-z / λ), where λ is the decay coefficient). Finally, borehole data is used to calibrate microscopic fracture connectivity. Microscopic fracture porosity (0.1%-5%) is statistically analyzed through core thin section observations, and pore throat radius distribution (0.01-1 μm) is obtained through mercury intrusion porosimetry. A microscopic fracture connectivity probability function (P_c=1-e^(-k·r)) is constructed. 2 (where k is the connectivity coefficient);

[0019] C2. Discrete Fragment Network (DFN) Modeling and Anisotropy Quantization: Multi-scale fracture generation is performed based on fractal theory. Fragment networks are generated at macroscopic (length > 100m), mesoscopic (1-100m), and microscopic (< 1m) scales. A Poisson process is used to control the fracture spacing distribution. Monte Carlo sampling is used to determine the fracture azimuth (following a Fisher distribution with a concentration parameter κ = 5-20). Repermeability anisotropy is calculated, and a permeability tensor (k_ij = (t_ij)) is assigned to each fracture. 2 / 12)·δ_ij, where t is the aperture), calculate the equivalent permeability tensor of the region (K_eq=ΣV_i·k_i / V_total) using the volume averaging method, and quantify the main permeability direction (K_max / K_min ratio reaches 10). 2 -10 3 );

[0020] C3. Monte Carlo simulation and transport probability field generation, parameter random sampling: for fracture aperture (t~N(μ_t,σ_t) 2)), permeability (K~LN(μ_K,σ_K) 2 Latin hypercube sampling (LHS) was performed on geochemical indicators (such as δ18O~U(-10‰, -5‰)) to generate 10 5 -10 6 The parameters are combined, and then a path tracing algorithm is used. An improved random walk particle tracing method (RWPT) is employed. Each particle carries geochemical fingerprint characteristics. When moving along the fracture network, the fingerprint value is updated according to the fracture attributes (aperture, permeability) and fluid mixing rules (complete mixing / laminar flow). The path frequency of all particles reaching the observation point is counted. The posterior probability is updated by combining Bayes' theorem (P(θ|D)=P(D|θ)·P(θ) / P(D), where θ is the fracture parameter and D is the geochemical observation data), generating a migration probability density cloud map (resolution 10m×10m×1m).

[0021] Preferably, the calculation formula for the development of the Long Short-Term Memory (LSTM) network model in A3 is as follows:

[0022] X t =α·f LSTM (T t ,P t C t )+(1-α)·g GAN (D hisory )

[0023] X t represents the predicted geothermal fluid migration path vector at time t. This vector contains key information such as migration direction and velocity, describing the migration state of the geothermal fluid at that moment. α is a weighting coefficient used to balance the contributions of the LSTM model prediction results and the GAN optimization results, and can be adjusted according to actual data and prediction requirements. f LSTM (T t ,P t C t ) is a prediction function of a Long Short-Term Memory (LSTM) network based on real-time monitoring data at time t, outputting a predicted value for the geothermal fluid migration path, T t ,P t C t Represents pressure, temperature, chemical field data, g GAN (D hisory ) is a Generative Adversarial Network (GAN) based on historical data D hisory The generated virtual path scenario is used to optimize the prediction results of the LSTM model and enhance the robustness of the prediction.

[0024] Preferably, the calculation formula for the Generative Adversarial Network (GAN) optimization model in A3 is as follows:

[0025]

[0026] x t The input vector (temperature, pressure, chemical field data) representing time t, h t-1 C represents the hidden state in the previous moment. t-1 The symbol represents the cell state at the previous time step, ⊙ represents element-wise multiplication, tanh represents the hyperbolic tangent function, and f represents the cell state at the previous time step. t This indicates the decision on how much historical information to retain. This indicates the decision of how much new information to add, o t This indicates which information to control the output.

[0027] Preferably, the algorithm formula for dynamically predicting fluid transport paths in A3 is as follows:

[0028]

[0029] This represents the objective function of the GAN, and D represents the discriminator, which determines whether the input data is real data. The generator generates virtual data, D(x) represents the discriminator's output on the real data x, and G represents the generator. This represents receiving random noise, and D(G(z)) represents generating a virtual geothermal fluid migration path scenario.

[0030] A4 deploys a distributed fiber optic sensor network to achieve synchronous monitoring of multiple parameters, including temperature, strain, and chemical field. An edge computing module enables real-time data processing and online model updates. The algorithm formula for the edge computing module is as follows: θ t+1 Let θ be the model parameters updated at the next time step t+1. t Here are the model parameters at the current time step, and α represents the learning rate. Let x′ be the gradient of the loss function calculated at the edge. t This represents the preprocessed sensor data, y t This represents the target for real-time path prediction.

[0031] Beneficial effects

[0032] This invention provides a method for tracing the genesis and migration paths of deep geothermal fluids. Compared with existing technologies, it has the following advantages:

[0033] By utilizing core-shell structured tracers, LIBS technology achieves in-situ, non-destructive, real-time downhole detection, balancing high fluorescence efficiency and geological compatibility. This replaces the traditional sampling-laboratory analysis process, reducing usage by over 90% and lowering ecological risks to 1 / 50th of traditional methods, aligning with the trend of green geothermal development. It forms a complete innovation chain in the material design, detection methods, and environmental adaptability of nano-tracers, significantly improving the accuracy and sustainability of deep geothermal fluid migration path tracing. Furthermore, it feeds back fracture parameters corresponding to low-probability paths in Monte Carlo simulations to the DFN model, optimizing fracture connectivity through simulated annealing algorithms (objective function: F = ω1·CCI + ω2·K_eq, ω1 + ω2 = 1), to achieve model self-calibration, compare the thermal breakthrough time predicted by the probability field with the measured data of the fiber optic sensor (error < 5%), verify the reliability of the model, achieve deep integration of geochemical information and three-dimensional fracture physical field, transform the traditional static geological model into a dynamic probabilistic prediction tool, significantly improve the analytical accuracy of deep geothermal fluid migration path, and achieve real-time prediction and error self-correction of migration path by combining the three-dimensional fracture network model with the LSTM-GAN algorithm, and improve the accuracy of causal discrimination by joint tracing of isotopes, rare gases and trace elements. Attached Figure Description

[0034] Figure 1 This is a flowchart of the sensing method of the present invention. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] Please see Figure 1 The present invention provides a technical solution:

[0037] A method for tracing the genesis and migration paths of deep geothermal fluids, specifically including the following steps:

[0038] A1. Collect geothermal fluid samples, determine isotopic composition, rare gas composition, trace element fingerprint and water-rock reaction index, and establish a multidimensional feature database;

[0039] A2. By integrating high-precision 3D seismic data, well logging data and borehole data, a multi-scale fracture network model is constructed to quantify fracture connectivity and permeability anisotropy. Geochemical fingerprint data is coupled with fracture network parameters, and a fluid transport probability field is generated through Monte Carlo simulation.

[0040] A3. Develop a Long Short-Term Memory (LSTM) network model, input real-time monitored temperature, pressure, and chemical field data, dynamically predict fluid transport paths, introduce a Generative Adversarial Network (GAN) to optimize the model, and generate virtual path scenarios through training with historical data to improve prediction robustness.

[0041] A4. Deploy a distributed fiber optic sensor network to achieve synchronous monitoring of multiple parameters, including temperature, strain, and chemical field.

[0042] In this embodiment of the invention, the determination of isotopic components includes: δ 18 O, δ 2 H, 3 He / 4 He 87 Sr / 86 Sr, rare gas components including He, Ne, Ar, trace element fingerprints including Li, B, As, Sb, and water-rock reaction indicators including SiO2, pH, Eh, are introduced into Al. An environmentally friendly nano-tracer is introduced, and high-sensitivity in-situ detection is achieved through laser-induced breakdown spectroscopy (LIBS). Geothermal fluid samples at different depths (500m, 1000m, 1500m) are obtained through drilling, and their δ... 18 O (-8.5‰ to -6.2‰), 3 He / 4 He(1.2×10 -6 Up to 2.8×10 -6 Based on the Li / B ratio (0.5 to 1.2) and the regional geological background, the fluid was determined to be a mixture of deep-circulating metamorphic water and magmatic hydrothermal fluid.

[0043] In this embodiment of the invention, the specific method for achieving high-sensitivity in-situ detection by introducing an environmentally friendly nano-tracer and using laser-induced breakdown spectroscopy (LIBS) is as follows:

[0044] B1. CdSe / ZnS core-shell quantum dots are selected, with emission wavelengths covering the visible to near-infrared region (500-800nm). Multicolor encoding is achieved by controlling the particle size (2-10nm), and an improved method is used. The sol-gel method deposits a dense SiO2 shell (10-20 nm thick) on the surface of quantum dots, forming a QDs@SiO2 core-shell structure, which prevents the quantum dots from directly contacting the external environment.

[0045] B2. The surface of SiO2 is modified with a silane coupling agent to introduce active functional groups, enhancing the dispersibility of nanoparticles in aqueous solution and their compatibility with geothermal fluid components. An outer layer of polyethylene glycol or zwitterionic polymer is then applied to reduce non-specific adsorption of nanoparticles in the geothermal system and extend the circulation time. Nano-tracer injection and migration monitoring are then performed. Based on the geothermal system volume (V) and monitoring requirements, nano-tracers are injected at a concentration of 0.1-1 ppm, reducing the total amount used by more than 90% compared to traditional dyes. A dedicated injection pump is used to slowly inject the tracer solution into the target layer along the borehole wall, avoiding disturbance to the fracture network. Fiber optic cables are deployed along the geothermal wellbore and potential migration paths, with monitoring points spaced 10-50 m apart, enabling simultaneous acquisition of temperature, strain, and chemical fields. A portable LIBS system is installed at the key monitoring wellhead, with a laser wavelength of 1064 nm, pulse energy of 10 mJ, and a focused spot diameter of <100 μm, allowing direct excitation of nanoparticles through the wellbore fluid.

[0046] B3. Laser excitation: A high-energy pulsed laser is focused onto the fluid sample, instantly vaporizing and ionizing nanoparticles to generate plasma. During the plasma cooling process, characteristic spectra are emitted and dispersed by a spectrometer (0.1 nm resolution). A CCD detector records the spectral signals. The quantum dot emission peaks include Cd (226.5 nm), Se (196.0 nm), and Zn (213.8 nm) elemental lines, used to quantify tracer concentration. The SiO2 matrix peaks include Si (288.2 nm) and O (777.4 nm) lines to verify the integrity of the nanoparticles. Finally, a calibration curve of spectral intensity (I) versus tracer concentration (C) is established (I = kC^n, n = 0.5-1.2). Multivariate linear regression is used to eliminate matrix effect interference. Combined with temperature (ΔT) and strain (Δε) data from fiber optic sensors, a particle filtering algorithm is used to reconstruct the nanoparticle migration trajectory, achieving a spatial resolution of 1 m.

[0047] The aforementioned QDs@SiO2 core-shell structure tracer combines high fluorescence efficiency with geological compatibility. LIBS technology enables in-situ, non-destructive, real-time detection downhole, replacing the traditional sampling-laboratory analysis process. It reduces the amount of tracer used by more than 90% and lowers the ecological risk to 1 / 50 of the traditional method, aligning with the trend of green geothermal development. It forms a complete innovation chain in terms of nano-tracer material design, detection methods, and environmental adaptability, significantly improving the accuracy and sustainability of deep geothermal fluid migration path tracing.

[0048] In this embodiment of the invention, the specific method for constructing a multi-scale fracture network model and quantifying fracture connectivity and permeability anisotropy is as follows:

[0049] C1. Macroscopic fracture distribution is constrained by 3D seismic data. Spatial locations of fracture zones are extracted through seismic attribute analysis. Sequence stratigraphy is used to delineate the top and bottom interfaces of thermal reservoirs, constructing a macroscopic fracture network framework. Then, well logging data is used to correct mesoscopic fracture density: imaging logging (FMI, UBI) is used to identify the wellbore fracture orientation and aperture (0.1-10 mm). Combined with sonic full-wavelength logging, fracture filling types are inverted, establishing a fracture density decay model with depth (exponential function: ρ(z)=ρ0e^(-z / λ), where λ is the decay coefficient). Finally, borehole data is used to calibrate microscopic fracture connectivity. Microscopic fracture porosity (0.1%-5%) is statistically analyzed through core thin section observations, and pore throat radius distribution (0.01-1 μm) is obtained through mercury intrusion porosimetry. A microscopic fracture connectivity probability function (P_c=1-e^(-k·r)) is constructed. 2 (where k is the connectivity coefficient);

[0050] C2. Discrete Fragment Network (DFN) Modeling and Anisotropy Quantization: Multi-scale fracture generation is performed based on fractal theory. Fragment networks are generated at macroscopic (length > 100m), mesoscopic (1-100m), and microscopic (< 1m) scales. A Poisson process is used to control the fracture spacing distribution. Monte Carlo sampling is used to determine the fracture azimuth (following a Fisher distribution with a concentration parameter κ = 5-20). Repermeability anisotropy is calculated, and a permeability tensor (k_ij = (t_ij)) is assigned to each fracture. 2 / 12)·δ_ij, where t is the aperture), calculate the equivalent permeability tensor of the region (K_eq=ΣV_i·k_i / V_total) using the volume averaging method, and quantify the main permeability direction (K_max / K_min ratio reaches 10). 2 -10 3 );

[0051] C3. Monte Carlo simulation and transport probability field generation, parameter random sampling: for fracture aperture (t~N(μ_t,σ_t) 2 )), permeability (K~LN(μ_K,σ_K) 2 Geochemical indicators (such as δ) 18 Latin hypercube sampling (LHS) was performed on samples O to U (-10‰, -5‰) to generate 10 5 -10 6The parameters are combined, and then a path tracing algorithm is used. An improved random walk particle tracing method (RWPT) is employed. Each particle carries geochemical fingerprint characteristics. When moving along the fracture network, the fingerprint value is updated according to the fracture attributes (aperture, permeability) and fluid mixing rules (complete mixing / laminar flow). The path frequency of all particles reaching the observation point is counted. The posterior probability is updated by combining Bayes' theorem (P(θ|D)=P(D|θ)·P(θ) / P(D), where θ is the fracture parameter and D is the geochemical observation data), generating a migration probability density cloud map (resolution 10m×10m×1m).

[0052] The above method feeds back fracture parameters corresponding to low-probability paths in Monte Carlo simulations to the DFN model, optimizes fracture connectivity using simulated annealing (objective function: F = ω1·CCI + ω2·K_eq, ω1 + ω2 = 1), achieves model self-calibration, and verifies model reliability by comparing the thermal breakthrough time predicted by the probability field with the measured data from fiber optic sensors (error < 5%). This method achieves deep fusion of geochemical information and three-dimensional fracture physical field, transforming the traditional static geological model into a dynamic probabilistic prediction tool and significantly improving the analytical accuracy of deep geothermal fluid migration paths.

[0053] In this embodiment of the invention, the calculation formula for the development of the Long Short-Term Memory (LSTM) network model in A3 is as follows:

[0054] X t =α·f LSTM (T t ,P t C t )+(1-α)·g GAN (D hisory )

[0055] X t represents the predicted geothermal fluid migration path vector at time t. This vector contains key information such as migration direction and velocity, describing the migration state of the geothermal fluid at that moment. α is a weighting coefficient used to balance the contributions of the LSTM model prediction results and the GAN optimization results, and can be adjusted according to actual data and prediction requirements. f LSTM (T t ,P t C t ) is a prediction function of a Long Short-Term Memory (LSTM) network based on real-time monitoring data at time t, outputting a predicted value for the geothermal fluid migration path, T t ,P t C t Represents pressure, temperature, chemical field data, g GAN (D hisory ) is a Generative Adversarial Network (GAN) based on historical data D hisoryThe generated virtual path scenario is used to optimize the prediction results of the LSTM model and enhance the robustness of the prediction.

[0056] In this embodiment of the invention, the calculation formula for the optimized Generative Adversarial Network (GAN) model in A3 is as follows:

[0057]

[0058] x t The input vector (temperature, pressure, chemical field data) representing time t, h t-1 C represents the hidden state at the previous moment. t-1 The symbol represents the cell state at the previous time step, ⊙ represents element-wise multiplication, tanh represents the hyperbolic tangent function, and f represents the cell state at the previous time step. t This indicates the decision on how much historical information to retain. This indicates the decision of how much new information to add, o t This indicates which information to control the output.

[0059] In this embodiment of the invention, the algorithm formula for dynamically predicting fluid transport paths in A3 is as follows:

[0060]

[0061] This represents the objective function of the GAN, and D represents the discriminator, which determines whether the input data is real data. The generator generates virtual data, D(x) represents the discriminator's output on the real data x, and G represents the generator. This represents receiving random noise, and D(G(z)) represents generating a virtual geothermal fluid migration path scenario.

[0062] In this embodiment of the invention, a distributed fiber optic sensor network is deployed in A4 to achieve synchronous monitoring of multiple parameters including temperature, strain, and chemical field. Real-time data processing and online model updates are achieved through an edge computing module. The algorithm formula for the edge computing module is as follows:

[0063] θ t+1 Let θ be the model parameters updated at the next time step t+1. t Here are the model parameters at the current time step, and α represents the learning rate. Let x′ be the gradient of the loss function calculated at the edge. t This represents the preprocessed sensor data, y t By representing the real-time path prediction target and through an edge-cloud collaborative computing architecture, this invention realizes a closed loop of "perception-decision-control" for geothermal fluid migration monitoring, significantly improving the safety and economy of deep geothermal resource development.

[0064] In summary, by collecting data from high-speed moving targets using acquisition devices, and then integrating the preprocessed data to establish a spatial model, the position and angle changes of high-speed moving targets can be identified. After learning and optimization, this model can be applied to the identification and acquisition devices, enabling the identification and acquisition of high-speed moving targets, the calculation of the position of high-speed moving targets based on data conversion, real-time updates of data based on the movement of the moving targets, and prediction of the future trajectory of the targets. This also facilitates the perception and real-time adjustment of the devices for high-speed moving targets, improving the accuracy of the data.

[0065] Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0066] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0067] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for tracing the genesis and migration path of deep geothermal fluids, characterized in that: Specifically, the following steps are included: A1. Collect geothermal fluid samples, determine isotopic composition, rare gas composition, trace element fingerprint and water-rock reaction index, and establish a multidimensional feature database; A2. By integrating high-precision 3D seismic data, well logging data and borehole data, a multi-scale fracture network model is constructed to quantify fracture connectivity and permeability anisotropy. Geochemical fingerprint data is coupled with fracture network parameters, and a fluid transport probability field is generated through Monte Carlo simulation. A3. Develop a long short-term memory network model, input real-time monitored temperature, pressure, and chemical field data, dynamically predict fluid transport paths, introduce an adversarial generative network to optimize the model, and generate virtual path scenarios through training with historical data to improve prediction robustness. A4. Deploy a distributed fiber optic sensor network to achieve synchronous monitoring of multiple parameters, including temperature, strain, and chemical field. The determination of isotopic components includes: δ 18 O, δ 2 H, 3 He / 4 He 87 Sr / 86 Sr, rare gas components include: He, Ne, Ar, trace element fingerprints include: Li, B, As, Sb, water-rock reaction indicators include: SiO2, pH, Eh, the Al is introduced with an environmentally friendly nano-tracer, and high-sensitivity in-situ detection is achieved through laser-induced breakdown spectroscopy; The specific method for achieving high-sensitivity in-situ detection by introducing an environmentally friendly nano-tracer and using laser-induced breakdown spectroscopy is as follows: B1. CdSe / ZnS core-shell quantum dots are selected, emitting wavelengths covering the visible to near-infrared region. Multicolor encoding is achieved by controlling the particle size, and an improved method is employed. The sol-gel method deposits a dense SiO2 shell on the surface of quantum dots to form a QDs@SiO2 core-shell structure, which prevents the quantum dots from directly contacting the external environment. B2. The surface of SiO2 is modified with a silane coupling agent to introduce active functional groups, enhancing the dispersibility of nanoparticles in aqueous solution and their compatibility with geothermal fluid components. An outer layer of polyethylene glycol or zwitterionic polymer is then applied to reduce non-specific adsorption of nanoparticles in the geothermal system and extend the circulation time. Nano-tracer injection and migration monitoring are then performed. Based on the geothermal system volume and monitoring requirements, nano-tracers are injected at a concentration of 0.1-1 ppm, reducing the total amount used by more than 90% compared to traditional dyes. A dedicated injection pump is used to slowly inject the tracer solution into the target layer along the borehole wall, avoiding disturbance to the fracture network. Fiber optic cables are deployed along the geothermal wellbore and potential migration paths, with monitoring points spaced 10-50 m apart, enabling simultaneous acquisition of temperature, strain, and chemical fields. A portable LIBS system is installed at the key monitoring wellhead, with a laser wavelength of 1064 nm, pulse energy of 10 mJ, and a focused spot diameter of <100 μm, allowing it to penetrate the wellbore fluid and directly excite nanoparticles. B3. Laser excitation: A high-energy pulsed laser is focused onto the fluid sample, instantly vaporizing and ionizing nanoparticles to generate plasma. During the plasma cooling process, characteristic spectra are emitted and dispersed by a spectrometer. A CCD detector records the spectral signals. The quantum dot emission peaks include Cd, Se, and Zn element lines, used to quantify tracer concentration. The SiO2 matrix peaks include Si and O lines, used to verify the integrity of the nanoparticles. Finally, a calibration curve of spectral intensity versus tracer concentration is established. Multivariate linear regression is used to eliminate matrix effect interference. Combined with temperature and strain data from fiber optic sensors, a particle filtering algorithm is used to reconstruct the nanoparticle migration trajectory, achieving a spatial resolution of 1m.

2. The method for tracing the genesis and migration path of deep geothermal fluids according to claim 1, characterized in that: The specific method for constructing a multi-scale fracture network model and quantifying fracture connectivity and permeability anisotropy is as follows: C1. Macroscopic fracture distribution is constrained by 3D seismic data. Spatial location of fracture zones is extracted through seismic attribute analysis. Sequence stratigraphy is used to delineate the top and bottom interfaces of thermal reservoirs and construct a macroscopic fracture network framework. Then, well logging data is used to correct the mesoscopic fracture density: imaging logging is used to identify the direction and aperture of fractures around the well. Combined with sonic full-wavelength logging, fracture filling material type is inverted and a fracture density decay model with depth is established. Finally, borehole data is used to calibrate microscopic fracture connectivity. Microscopic fracture porosity is statistically analyzed through core thin section observation. The pore throat radius distribution is obtained by combining mercury intrusion porosimetry and a microscopic fracture connectivity probability function is constructed. C2. Discrete fracture network modeling and anisotropy quantization: Multi-scale fracture generation. Based on fractal theory, fracture networks are generated at the macroscopic scale. Poisson process is used to control fracture spacing distribution. Monte Carlo sampling is used to determine fracture azimuth angles. Permeability anisotropy is then calculated. A permeability tensor is assigned to each fracture. The equivalent permeability tensor of the region is calculated using the volume averaging method to quantify the main permeability direction. C3. Monte Carlo simulation and migration probability field generation, parameter random sampling: Latin hypercube sampling is performed on fracture aperture, permeability, and geochemical indicators to generate 10... 5 -10 6 The parameters are combined, and then a path tracing algorithm is used. An improved random walk particle tracking method is adopted. Each particle carries geochemical fingerprint characteristics. When moving along the fracture network, the fingerprint value is updated according to the fracture properties and fluid mixing rules. The path frequency of all particles reaching the observation point is counted. The posterior probability is updated by combining Bayes' theorem to generate a migration probability density cloud map.

3. The method for tracing the genesis and migration path of deep geothermal fluids according to claim 1, characterized in that: The calculation formula for the Long Short-Term Memory (LSTM) network model in A3 is as follows: X t =α·f LSTM (T t ,P t ,C t )+(1-α)·g GAN (D hisory ) Where X t Let t represent the geothermal fluid migration path vector predicted by t, α be a weighting coefficient used to balance the contributions of the LSTM model prediction results and the GAN optimization results, and f be the weighting coefficient. LSTM (T t ,P t C t ) is a prediction function of a Long Short-Term Memory (LSTM) network based on real-time monitoring data at time t, outputting a predicted value for the geothermal fluid migration path, T t ,P t C t Represents pressure, temperature, chemical field data, g GAN (D hisory ) is a Generative Adversarial Network (GAN) based on historical data D hisory The generated virtual path scenario.

4. The method for tracing the genesis and migration path of deep geothermal fluids according to claim 1, characterized in that: The calculation formula for the adversarial generative network optimization model in A3 is as follows: Where x t h represents the input vector at time t. t-1 C represents the hidden state in the previous moment. t-1 The symbol represents the cell state at the previous time step, ⊙ represents element-wise multiplication, tanh represents the hyperbolic tangent function, and f represents the cell state at the previous time step. t This indicates the decision on how much historical information to retain. This indicates the decision of how much new information to add, o t This indicates which information to control the output.

5. The method for tracing the genesis and migration path of deep geothermal fluids according to claim 1, characterized in that: The formula for the algorithm of dynamically predicting fluid migration paths in A3 is as follows: in This represents the objective function of the GAN, and D represents the discriminator, which determines whether the input data is real data. The generator generates virtual data, D(x) represents the discriminator's output on the real data x, and G represents the generator. This represents receiving random noise, and D(G(z)) represents generating a virtual geothermal fluid migration path scenario.

6. The method for tracing the genesis and migration path of deep geothermal fluids according to claim 1, characterized in that: A4 deploys a distributed fiber optic sensor network to achieve synchronous monitoring of multiple parameters, including temperature, strain, and chemical field. An edge computing module enables real-time data processing and online model updates. The algorithm formula for the edge computing module is as follows: θ t+1 Let θ be the model parameters updated at the next time step t+1. t Here are the model parameters at the current time step, and α represents the learning rate. Let x′ be the gradient of the loss function calculated at the edge. t This represents the preprocessed sensor data, y t This represents the target for real-time path prediction.

Images