A method for coupling decomposition and quantification of multi-source noise in borehole transient electromagnetic based on multi-physical mechanism disturbance

Abstract

The application discloses a well transient electromagnetic multi-source noise coupling decomposition and quantification method based on multi-physical mechanism disturbance, which comprises the following steps: firstly, solving the well transient electromagnetic response signal under the undisturbed reference working condition; then, applying high-frequency pose vibration disturbance to the transmitting coil by using an adaptive dynamic grid, applying a layered temperature field in the surrounding rock stratum based on the heat conduction equation, coupling the temperature mapping and the electromagnetic diffusion equation, and superimposing an equivalent small loop alternating current coil as a narrowband harmonic additional source term to generate a noisy response signal; extracting a single-source noise component, a combined noise component and a multi-source mixed interaction component, and constructing noise intensity, energy proportion and time-varying non-stationarity indexes for the single-source noise component and the combined noise component, so as to express the multi-source noise as a decomposable, traceable and mapable data structure, and form a structured noise sample library. The application can be used for well transient electromagnetic noise mechanism research, noise removal or reconstruction algorithm evaluation and data set construction.

Description

Technical Field

[0001] This invention relates to the field of geophysical electromagnetic exploration technology, specifically a method for decomposition and quantification of multi-source transient electromagnetic noise in wells based on multi-physical mechanism perturbation. Background Technology

[0002] Borehole transient electromagnetic (BTEM) is a high-precision geophysical exploration technology based on electromagnetic induction and diffusion theory. By applying a time-domain excitation current through a transmitting coil and receiving the induced electromotive force response in the well, it can infer the distribution of formation conductivity or resistivity. It has important application value in scenarios such as mineral resource exploration, reservoir identification and stratification interpretation.

[0003] Downhole transient electromagnetic exploration can significantly shorten the propagation path and enhance near-field sensitivity to targets, but its measurement environment is more complex. The signal link is simultaneously superimposed with multi-source disturbances from the complex downhole environment, resulting in significant time-varying, non-stationary, and multi-source coupling characteristics in the observation data, which weakens the reliability of inversion and interpretation. Typical disturbance characteristics include: high-frequency micro-vibrations of the transmitting coil will cause time-varying spatial pose and current distribution of the transmitting coil, thereby changing the electromagnetic diffusion process and receiving response characteristics; the temperature field of the stratified formation changes with time and depth and affects electromagnetic diffusion through conductivity-temperature mapping; and narrowband harmonic interference introduced by cables and power supply links is superimposed on the observation signal. These interference factors can occur individually or often coexist in a mixed form, forming significant noise components in the time domain, frequency domain, and time-frequency domain.

[0004] In existing technologies, the processing of transient electromagnetic noise in wells mostly relies on empirical noise addition, single-source independent assumptions, or posterior filtering denoising strategies, which generally have the following shortcomings:

[0005] 1) Lack of parameterized and controllable generation capability for noise mechanism: Existing methods mostly approximate transient electromagnetic noise in wells as statistical noise, empirical disturbance terms, or simple linear superposition terms. They usually remain at the level of a posteriori noise addition or signal-level disturbance, and fail to map typical noise mechanisms such as mechanical vibration, temperature change, and cable harmonics into adjustable geometric pose parameters, medium parameters, and additional source term parameters, respectively. Therefore, it is difficult to systematically scan and combine the amplitude, frequency, phase, spatial location, action window, and coupling strength of noise under a unified physical model, and it is also difficult to form working condition labels that correspond one-to-one with the noise mechanism parameters.

[0006] 2) Lack of explicit coupling decomposition capability for multi-source coupled noise: Multi-source noise is not strictly linearly separable during electromagnetic diffusion and observation. If only single-source differential or mixed signal is filtered directly, it is difficult to explicitly identify and output single-source contribution, combined contribution and multi-source mixed interaction components, resulting in unclear noise attribution, inexplicable coupling path and untraceable error source.

[0007] 3) Lack of structured representation for multi-source noise coupling decomposition: Traditional methods mostly describe noise in an unstructured way based on a single statistic, a single energy value or overall waveform difference, treating multi-source noise as an indivisible overall disturbance. It is difficult to explicitly distinguish the contribution of single-source noise, the contribution of combined noise and the contribution of multi-source mixed interaction, and it is also difficult to establish the mapping relationship between noise components and the reference response, generation conditions and parameter combinations. Therefore, it is impossible to form a decomposable, traceable and mappable noise data structure.

[0008] In summary, there is an urgent need for a method for coupling decomposition and quantification of transient electromagnetic multi-source noise in wells based on multi-physics mechanism perturbation. This method would enable noise sources such as mechanical vibration, temperature change, and cable harmonics to achieve parameterized and controllable generation oriented towards noise mechanism, explicit decomposition oriented towards multi-source coupling, and structured quantitative expression of decomposition results within a unified multi-physics coupling framework. This would provide a reproducible technical benchmark for noise mechanism research, algorithm evaluation, and sample library construction. Summary of the Invention

[0009] The purpose of this invention is to provide a method for decomposition and quantization of coupled multi-source transient electromagnetic noise in wells based on multi-physics mechanism perturbation, comprising the following steps:

[0010] Step 1) Construct a three-dimensional geometric model of the wellbore-formation-transmitting coil, and define the physical property parameters of the three-dimensional geometric model of the wellbore-formation-transmitting coil and the time-domain excitation current of the transmitting coil. ;

[0011] Based on the time-domain Maxwell's equations, the electromagnetic diffusion control equations were derived, and the transient electromagnetic response signal in the well under undisturbed baseline conditions was calculated. ;

[0012] Step 2) Apply high-frequency pose vibration perturbation to the transmitting coil, update the coil pose using an adaptive moving mesh method, and modify the new electromagnetic diffusion control equations based on the coil displacement perturbation and the source terms of the transmitting coil, to calculate the response signal under mechanical vibration conditions. ;

[0013] A layered temperature field is applied to the surrounding rock strata, and the response signal under temperature variation conditions is calculated based on the coupling of the heat conduction equation, the electrical conductivity-temperature mapping relationship, and the electromagnetic diffusion control equation. ;

[0014] A small-loop AC coil is added as a narrowband harmonic additional source term. Based on the electromagnetic diffusion control equation of the superposition of the transmitting coil source term and the harmonic additional source term, the response signal under cable harmonic interference conditions is calculated. ;

[0015] Step 3) Based on the controllable parameterization strategy, the amplitude, frequency, phase, spatial location, action window and coupling strength of each noise source are scanned and combined to construct a set of transient electromagnetic noise response signals in the well under single source and multi-source superposition conditions, and form a condition label corresponding to the noise source parameters one by one.

[0016] Step 4) Align the noisy response signal with the baseline operating condition response signal, and extract the noise component using a differential method to obtain the noise sequence. Interaction components with multi-source mixtures This results in a traceable separation of noise components and residual signals;

[0017] Step 5) Based on the noise sequence and multi-source mixed interaction components, construct a structured expression and quantification system for coupling decomposition of transient electromagnetic multi-source noise in wells. Characterize multi-source noise as a decomposable, traceable, and mappable data structure composed of single-source noise components, combined noise components, and multi-source mixed interaction components, and quantify the noise intensity, energy ratio, and time-varying nonstationarity.

[0018] Step 6) The structured noise decomposition results and quantitative characterization results are associated, integrated, stored, and exported in batches to form a structured noise sample library.

[0019] Furthermore, the wellbore-formation-transmitting coil three-dimensional geometric model includes the wellbore space domain, the transmitting coil domain, the signal receiving array, and the surrounding rock formation domain;

[0020] The surrounding rock stratigraphic domain is composed of discretized media units and has parameterizable geological conditions.

[0021] The wellbore-formation-transmitting coil three-dimensional geometric model also includes an equivalent small loop AC coil domain for cable harmonic interference modeling, coil pose degrees of freedom for mechanical vibration modeling, and a layered temperature field domain for temperature change modeling.

[0022] Furthermore, the physical properties of the three-dimensional geometric model of the wellbore-formation-transmitting coil include wellbore diameter and wellbore trajectory parameters, and in-well electrical conductivity. Axial continuous stratification variation parameters, surrounding rock resistivity The assigned parameters, geometric and electrical parameters of the transmitting coil and signal receiving array, and controllable parameters of typical noise sources;

[0023] The coil geometry and electrical parameters include the coil radius. Number of turns Axial position Radial eccentricity The amplitude of the coil excitation current relative to the normal direction With shutdown time ;

[0024] Mechanical vibration parameters include vibration displacement. Vibration frequency With phase ;

[0025] Cable harmonic parameters include harmonic frequencies. Amplitude With phase ;

[0026] Temperature variation parameters include stratified temperature disturbances And the window of action.

[0027] Furthermore, in step 2), the steps for deriving the electromagnetic diffusion control equations include:

[0028] Step 2.1) Apply boundary conditions of continuous tangential electric field and continuous normal magnetic flux density at the multi-dielectric interface, and construct the time-domain Maxwell's equations and current constitutive relations, namely:

[0029] (1)

[0030] (2)

[0031] in, For electric field strength, It represents the magnetic flux density. The magnetic field strength, For current density, Permeability, For conduction current density, The source term current density of the transmitting coil, To account for the harmonic interference source current density corresponding to the equivalent current carrying capacity of the additional small-loop AC coil, Temperature-dependent conductivity;

[0032] Among them, electric field strength and magnetic induction intensity As shown below:

[0033] Magnetic vector potential With electric scalar potential :

[0034] (3)

[0035] In the formula, , For magnetic vector potential and electric scalar potential;

[0036] Step 2.2) Construct the electromagnetic diffusion control equations, namely:

[0037] (4)

[0038] Among them, the transmitting coil source term As shown below:

[0039] (5)

[0040] In the formula, This is the excitation current for the transmitting coil; The number of turns of the transmitting coil. The equivalent current-carrying cross-sectional area of ​​the transmitting coil. The unit tangent vector in the current-carrying direction. For the current-carrying domain of the transmitting coil Indicator functions;

[0041] Harmonic source term As shown below:

[0042] (6)

[0043] (7)

[0044] In the formula, This refers to the harmonic current of the small-loop AC coil. For the number of turns of the small-loop AC coil, Its equivalent current-carrying cross-sectional area, Let it be the unit tangent vector in its current-carrying direction. For harmonic current-carrying domain Indicator functions;

[0045] Temperature-dependent conductivity As shown below:

[0046] (8)

[0047] In the formula, Initial calibration temperature Lower reference conductivity is the temperature coefficient of electrical conductivity.

[0048] Furthermore, the displacement perturbation of the transmitting coil is used to characterize the time-varying change of the transmitting coil pose under mechanical vibration conditions, and is used to drive the time-varying update of the transmitting coil source terms, thereby participating in the calculation of the response signal under mechanical vibration conditions;

[0049] The displacement disturbance of the transmitting coil is shown below:

[0050] (9)

[0051] In the formula, The reference point for the geometric center of the equivalent current-carrying domain of the transmitting coil at time... Spatial position vector, This is the initial position vector of the transmitting coil under undisturbed reference conditions.

[0052] Furthermore, the heat conduction equation is as follows:

[0053] (10)

[0054] in For the density of the medium, For isobaric specific heat capacity, Thermal conductivity, The heat source intensity per unit volume.

[0055] Furthermore, response signals under mechanical vibration conditions Response signal under cable harmonic interference Response signal under temperature change conditions The solution process is as follows: Locally refined unstructured meshing and adaptive mesh adjustment are performed on the transmitting coil domain, the wellbore space domain, the additional interference AC coil domain, and the wellbore interface domain; under mechanical vibration conditions, a dynamic mesh update strategy is used to smooth and reconstruct the mesh, ensuring that the mesh maintains element mass constraints and avoids element distortion as the coil pose changes; an infinite element equivalent domain is used to process the outer boundary of the computational domain to reduce non-physical reflection errors caused by finite domain truncation; the response signal under mechanical vibration conditions is obtained by the finite element discretization method. Response signal under cable harmonic interference Response signal under temperature change conditions .

[0056] Furthermore, a single noise sequence As shown below:

[0057] (11)

[0058] Combined noise sequence As shown below:

[0059] (12)

[0060] Multi-source hybrid interaction components As shown below:

[0061] (13)

[0062] In the formula, It is a mixed response signal under the influence of multiple physical mechanisms of disturbance; This is the response signal under the perturbation of the k-th physical mechanism.

[0063] Furthermore, the noise intensity is shown below:

[0064] (14)

[0065] In the formula, the noise intensity index NI is an index characterizing the overall noise intensity level over the entire time period. The single-source and combined noise sequences to be evaluated are as follows: and These are the L2 norms of the noise sequence and the unperturbed reference operating condition response signal over the entire time period, respectively. To avoid numerically unstable small regularization values;

[0066] The energy breakdown is as follows:

[0067] (15)

[0068] In the formula, the energy percentage index (NEP) is an indicator that characterizes the overall level of noise energy percentage throughout the entire time period. and These are the L2 norms of the noise sequence and the unperturbed reference condition response signal over the entire time period, respectively.

[0069] The time-varying nonstationarity is shown below:

[0070] (16)

[0071] In the formula, the time-varying nonstationarity index TVNS is a comprehensive index characterizing the time-varying nonstationarity of noise over the entire time period. Let Variance be the operator for all sliding window root mean square statistic sequences. Noise sequence In the Root mean square statistic over a sliding window For the first Length of each sliding window For the first A sliding window.

[0072] A well transient electromagnetic multi-source noise coupling decomposition and quantization system based on the method described above includes a geometric and parametric modeling module, a multi-physics coupling forward modeling and condition generation module, a controllable parameter scanning and condition labeling module, a comparison difference and interaction component extraction module, a coupling decomposition structured expression and quantization module, and a structured noise sample library export module.

[0073] The geometry and parametric modeling module constructs a three-dimensional geometric model of the wellbore-formation-transmitting coil and completes the parametric definition of physical property parameters and excitation current;

[0074] The multiphysics coupling forward modeling and working condition generation module solves the electromagnetic diffusion control equation under quasi-static conditions, introduces typical transient electromagnetic noise sources in the well, applies high-frequency pose vibration disturbance to the transmitting coil using an adaptive dynamic mesh, applies a layered temperature field in the surrounding rock formation, and adds a small-loop AC coil as a narrowband harmonic additional source term. Based on the time-varying update of pose disturbance, temperature-conductivity mapping, and electromagnetic diffusion control equation coupled with harmonic additional source term, the module obtains the undisturbed reference working condition response signal and the response signals of each noisy working condition.

[0075] The response signals of the unperturbed reference working condition and each noisy working condition were solved using the finite element discretization method; during the solution process, an infinite element equivalent domain was applied to the outer boundary.

[0076] The controllable parameter scanning and operating condition labeling module scans and combines the amplitude, frequency, phase, spatial location, action window and coupling strength of each noise source to construct a set of noisy response signals under single-source and multi-source superposition operating conditions, and forms operating condition labels that correspond one-to-one with the noise source parameters.

[0077] The contrast difference and interaction component extraction module aligns the corresponding components of the response signal with a unified time sampling point and a unified depth sampling point as the alignment reference, and performs contrast difference extraction of the noise sequence based on the unperturbed reference working condition response and the noise working condition response. At the same time, it outputs multi-source mixed interaction components to achieve traceable separation of noise components and residual signals.

[0078] The coupled decomposition structured representation and quantization module constructs a structured noise representation based on single-source noise sequences, combined noise sequences, and multi-source mixed interaction components. It represents multi-source noise as a decomposable, traceable, and mappable data structure, and quantifies the noise intensity, energy ratio, and time-varying nonstationarity of single-source noise components and combined noise components.

[0079] The structured noise sample library export module associates and stores the structured noise decomposition results and their quantization characterization results, and exports them in batches to form a structured noise sample library that can be used for simulation verification, algorithm evaluation, and dataset construction.

[0080] The technical effects of this invention are undeniable, and its beneficial effects are as follows:

[0081] 1) Based on the quasi-static electromagnetic diffusion control equation, a unified parameterized model is constructed that couples dynamic mesh pose disturbance, AC harmonic additional source terms and temperature change mapping. This enables the injection of physical consistency working conditions and parameterized controllable generation of transient electromagnetic noise in the well, thereby reducing the risk of physical instability caused by empirical denoising from the source.

[0082] 2) Based on the comparison difference between the undisturbed baseline operating condition and the noisy operating condition, the single-source noise, combined noise and multi-source mixed interaction components are explicitly output, so as to realize the traceable separation of noise attribution and coupling contribution, and provide a calculable basis for noise coupling strength assessment and mechanism analysis.

[0083] 3) Unlike traditional unstructured noise description methods based on a single statistic, this invention does not simply describe noise by adding multiple quantitative indicators. Instead, it explicitly decomposes multi-source noise into single-source noise components, combined noise components, and multi-source mixed interaction components by using differential and multi-source coupling decomposition, while maintaining the correspondence between them and the baseline response and the generated operating conditions. This transforms noise from an overall disturbance into a decomposable, traceable, and mappable data structure.

[0084] 4) Achieve data-driven output of coupling decomposition results. By uniformly outputting single-source noise components, combined noise components, multi-source mixed interaction components and their quantization results, a structured noise sample library that can be directly retrieved, traced and reproduced, and batch expanded is formed, significantly improving experimental reproducibility, sample manageability and evaluation comparability.

[0085] 5) The structured noise sample library can be directly used for supervised training, noise injection enhancement and domain randomization training of signal denoising or reconstruction models. The coupling decomposition results can be used to achieve training sample difficulty classification and distribution alignment, thereby improving the model's generalization ability and robustness to real-world complex noise scenarios, and providing a unified quantitative indicator closed loop for model evaluation.

[0086] 6) By scanning and quantifying the results of noise source parameters, a sensitivity mapping of key parameters to signal distortion and noise energy ratio can be formed, providing a quantitative basis for downhole tool optimization design, temperature control strategy, anti-interference scheme and field test variable configuration, reducing test costs and improving data quality controllability.

[0087] 7) The framework is highly scalable and covers a wide range of operating conditions. It can remain modular and can be expanded to include more noise types and more complex coupling conditions without changing the master equation and data loop. It can also cover the operating conditions space through parameter sweeping, thereby enhancing the scenario coverage and continuous evolution capability of the sample library.

[0088] In summary, this invention addresses the challenges of parameterized and controllable generation of multi-source noise in well transient electromagnetic observations, the difficulty in explicitly decomposing multi-source coupling contributions, and the lack of structured and quantitative characterization of noise. It proposes a method for coupling decomposition and quantification of multi-source noise in well transient electromagnetic observations based on multi-physics mechanism perturbations. This method uses the electromagnetic diffusion control equation as its core framework, incorporating geometrically time-varying perturbations, superposition of additional AC source terms, and temperature-driven medium parameter mapping into a unified generation framework. It achieves integrated decomposition of single-source noise, combined noise, and coupling residues by combining differential and interaction component definitions. Furthermore, it provides a unified description of single-source and combined noise components through quantitative indices, transforming multi-source noise from a holistic perturbation into a decomposable, traceable, and mappable data structure. This provides traceable, reproducible, and scalable quantitative benchmarks and methodological support for evaluating well transient electromagnetic denoising algorithms, verifying robustness, and constructing datasets. Attached Figure Description

[0089] Figure 1 This is a schematic diagram of a three-dimensional geometric model of the wellbore-formation-transmitting coil-noise source, showing the spatial relationship between the wellbore space domain, the surrounding rock formation domain, the transmitting coil domain, and the signal receiving array domain, as well as the typical noise sources and their impact distribution in well exploration.

[0090] Figure 2 This is a schematic diagram of vibration posture and current waveform, showing the high-frequency disturbance mode of mechanical vibration, the time-domain excitation current waveform of the transmitting coil, the harmonic current waveform of the additional small loop AC coil, and their input condition parameters.

[0091] Figure 3 This is a comparison chart of the transient electromagnetic response signals in the well under undisturbed reference conditions and single-source and multi-source superposition conditions, showing the time-domain differences in the response signals under undisturbed reference conditions, mechanical vibration conditions, cable harmonic interference conditions, temperature change conditions, and multi-source superposition conditions.

[0092] Figure 4 This is a flowchart for the controllable generation of typical multi-source noise in transient electromagnetic wells, showing the process from three-dimensional modeling of the wellbore-formation-transmitting coil, solving the baseline operating condition response, parameterizing the injection of three types of typical noise sources, combining noise parameters and operating conditions, generating single-source and multi-source noisy responses, to aligning differential extraction of noise components and outputting operating condition labels.

[0093] Figure 5 The diagram shows the extraction results of single-source noise components and multi-source combined noise components. It displays the single-source noise sequence and combined noise sequence obtained by comparing and differentiating the response signal of the unperturbed reference condition with the response signals of each noisy condition.

[0094] Figure 6This is a multi-source coupling interaction component analysis diagram, showing the interaction components and coupling contributions obtained by subtracting the independent contributions of each single source from the mixed response under the multi-source superposition condition, as well as their time-domain response characteristics.

[0095] Figure 7 The graph shows the results of multidimensional quantitative characterization of single-source and multi-source noise, displaying the calculation results and quantitative comparison differences of noise intensity index, energy ratio index and time-varying non-stationarity index. Detailed Implementation

[0096] The present invention will be further described below with reference to embodiments, but it should not be construed that the scope of the present invention is limited to the following embodiments. Various substitutions and modifications made based on ordinary technical knowledge and common practices in the art without departing from the above-described technical concept of the present invention should be included within the scope of protection of the present invention.

[0097] Example 1:

[0098] See Figures 1 to 7 A well-drilled transient electromagnetic multi-source noise coupling decomposition and quantization method based on multi-physics mechanism perturbation includes the following steps:

[0099] Step 1) Construct a three-dimensional geometric model of the wellbore-formation-transmitting coil, and define the physical property parameters of the three-dimensional geometric model of the wellbore-formation-transmitting coil and the time-domain excitation current of the transmitting coil. ;

[0100] Based on the time-domain Maxwell's equations, the electromagnetic diffusion control equations were derived, and the transient electromagnetic response signal in the well under undisturbed baseline conditions was calculated. ;

[0101] Step 2) Introduce typical transient electromagnetic noise sources in the well and calculate the response signals under different disturbance conditions: Apply high-frequency pose vibration disturbance to the transmitting coil, update the coil pose using an adaptive dynamic mesh method, and modify the new electromagnetic diffusion control equations based on the coil displacement disturbance and the source terms of the transmitting coil to calculate the response signal under mechanical vibration conditions. ;

[0102] A layered temperature field is applied to the surrounding rock strata, and the response signal under temperature variation conditions is calculated based on the coupling of the heat conduction equation, the electrical conductivity-temperature mapping relationship, and the electromagnetic diffusion control equation. ;

[0103] A small-loop AC coil is added as a narrowband harmonic additional source term. Based on the electromagnetic diffusion control equation of the superposition of the transmitting coil source term and the harmonic additional source term, the response signal under cable harmonic interference conditions is calculated. ;

[0104] Step 3) Based on the controllable parameterization strategy, the amplitude, frequency, phase, spatial location, action window and coupling strength of each noise source are scanned and combined to construct a set of transient electromagnetic noise response signals in the well under single source and multi-source superposition conditions, and form a condition label corresponding to the noise source parameters one by one.

[0105] Step 4) Align the noisy response signal with the baseline operating condition response signal, and extract the noise component using a differential method to obtain the noise sequence. Interaction components with multi-source mixtures This results in a traceable separation of noise components and residual signals;

[0106] Step 5) Based on the noise sequence and multi-source mixed interaction components, construct a structured expression and quantification system for coupling decomposition of transient electromagnetic multi-source noise in wells. Characterize multi-source noise as a decomposable, traceable, and mappable data structure composed of single-source noise components, combined noise components, and multi-source mixed interaction components, and quantify the noise intensity, energy ratio, and time-varying nonstationarity.

[0107] Step 6) The structured noise decomposition results and quantitative characterization results are associated, integrated, stored, and exported in batches to form a structured noise sample library.

[0108] Example 2:

[0109] A method for decomposition and quantization of transient electromagnetic multi-source noise in well based on multi-physics mechanism disturbance, with the same technical content as in Embodiment 1, further wherein the three-dimensional geometric model of wellbore-formation-transmitting coil includes a wellbore spatial domain, a transmitting coil domain, a signal receiving array, and a surrounding rock formation domain;

[0110] The surrounding rock stratigraphic domain is composed of discretized media units and has parameterizable geological conditions.

[0111] Example 3:

[0112] A method for coupling decomposition and quantization of transient electromagnetic multi-source noise in wells based on multi-physics mechanism perturbation, with the same technical content as any one of Embodiments 1-2. Furthermore, the three-dimensional geometric model of wellbore-formation-transmitting coil also includes an equivalent small loop AC coil domain for cable harmonic interference modeling, coil pose degrees of freedom for mechanical vibration modeling, and a layered temperature field domain for temperature change modeling.

[0113] Example 4:

[0114] A method for coupled decomposition and quantization of transient electromagnetic multi-source noise in wells based on multi-physics mechanism perturbation, with the same technical content as any one of embodiments 1-3, further comprising the following physical property parameters of the three-dimensional geometric model of wellbore-formation-transmitting coil: wellbore diameter and wellbore trajectory parameters, and in-well conductivity. Axial continuous stratification variation parameters, surrounding rock resistivity The assigned parameters, geometric and electrical parameters of the transmitting coil and signal receiving array, and controllable parameters of typical noise sources;

[0115] The coil geometry and electrical parameters include the coil radius. Number of turns Axial position Radial eccentricity The amplitude of the coil excitation current relative to the normal direction With shutdown time ;

[0116] Mechanical vibration parameters include vibration displacement. Vibration frequency With phase ;

[0117] Cable harmonic parameters include harmonic frequencies. Amplitude With phase ;

[0118] Temperature variation parameters include stratified temperature disturbances And the window of action.

[0119] Example 5:

[0120] A method for decomposition and quantization of coupled transient electromagnetic multi-source noise in wells based on multi-physics mechanism perturbation, with the same technical content as any one of embodiments 1-4, further comprising, in step 2), deriving the electromagnetic diffusion control equations, including:

[0121] Step 2.1) Apply boundary conditions of continuous tangential electric field and continuous normal magnetic flux density at the multi-dielectric interface, and construct the time-domain Maxwell's equations and current constitutive relations, namely:

[0122] (1)

[0123] (2)

[0124] in, For electric field strength, It represents the magnetic flux density. The magnetic field strength, For current density, Permeability, For conduction current density, The source term current density of the transmitting coil, To account for the harmonic interference source current density corresponding to the equivalent current carrying capacity of the additional small-loop AC coil, Temperature-dependent conductivity;

[0125] Among them, electric field strength and magnetic induction intensity As shown below:

[0126] Magnetic vector potential With electric scalar potential :

[0127] (3)

[0128] In the formula, , For magnetic vector potential and electric scalar potential;

[0129] Step 2.2) Construct the electromagnetic diffusion control equations, namely:

[0130] (4)

[0131] Among them, the transmitting coil source term As shown below:

[0132] (5)

[0133] In the formula, This is the excitation current for the transmitting coil; The number of turns of the transmitting coil. The equivalent current-carrying cross-sectional area of ​​the transmitting coil. The unit tangent vector in the current-carrying direction. For the current-carrying domain of the transmitting coil Indicator functions;

[0134] Harmonic source term As shown below:

[0135] (6)

[0136] (7)

[0137] In the formula, This refers to the harmonic current of the small-loop AC coil. For the number of turns of the small-loop AC coil, Its equivalent current-carrying cross-sectional area, Let it be the unit tangent vector in its current-carrying direction. For harmonic current-carrying domain Indicator functions;

[0138] Temperature-dependent conductivity As shown below:

[0139] (8)

[0140] In the formula, Initial calibration temperature Lower reference conductivity is the temperature coefficient of electrical conductivity.

[0141] Example 6:

[0142] A method for decomposition and quantization of coupled multi-source transient electromagnetic noise in wells based on multi-physics mechanism perturbation, with the same technical content as any one of embodiments 1-5, further wherein the coil displacement perturbation is as follows:

[0143] (9)

[0144] In the formula, The reference point for the geometric center of the equivalent current-carrying domain of the transmitting coil at time... Spatial position vector, This is the initial position vector under the unperturbed reference condition.

[0145] Example 7:

[0146] A method for decomposition and quantization of coupled transient electromagnetic multi-source noise in wells based on multi-physics mechanism perturbation, with the same technical content as any one of embodiments 1-6, further wherein the heat conduction equation is as follows:

[0147] (10)

[0148] in For the density of the medium, For isobaric specific heat capacity, Thermal conductivity, The heat source intensity per unit volume.

[0149] Example 8:

[0150] A method for decomposition and quantization of coupled multi-source transient electromagnetic noise in wells based on multi-physics perturbation, with the same technical content as any one of embodiments 1-7, further includes locally refined unstructured meshing and adaptive mesh adjustment for the transmitting coil domain, the in-well space domain, the additional interference AC coil domain, and the well-ground interface domain; under mechanical vibration conditions, a dynamic mesh update strategy is used to smooth and reconstruct the mesh, so that the mesh maintains element mass constraints and avoids element distortion as the coil pose changes; an infinite element equivalent domain is used to process the outer boundary of the computational domain to reduce non-physical reflection errors caused by finite domain truncation; and the response signal under mechanical vibration conditions is obtained by solving the finite element discretization method. Response signal under cable harmonic interference Response signal under temperature change conditions .

[0151] Example 9:

[0152] A method for coupled decomposition and quantization of transient electromagnetic multi-source noise in wells based on multi-physics perturbation, with the same technical content as any one of embodiments 1-8, further comprising: a single noise sequence As shown below:

[0153] (11)

[0154] Combined noise sequence As shown below:

[0155] (12)

[0156] Multi-source hybrid interaction components As shown below:

[0157] (13)

[0158] In the formula, It is a mixed response signal under the influence of multiple physical mechanisms of disturbance; This is the response signal under the perturbation of the k-th physical mechanism.

[0159] Example 10:

[0160] A method for coupled decomposition and quantization of transient electromagnetic multi-source noise in wells based on multi-physics mechanism perturbation, with the same technical content as any one of embodiments 1-9, further wherein the noise intensity is as follows:

[0161] (14)

[0162] In the formula, the noise intensity index NI is an index characterizing the overall noise intensity level over the entire time period. The single-source and combined noise sequences to be evaluated are as follows: and These are the L2 norms of the noise sequence and the unperturbed reference operating condition response signal over the entire time period, respectively. To avoid numerically unstable small regularization values;

[0163] The energy breakdown is as follows:

[0164] (15)

[0165] In the formula, the energy percentage index (NEP) is an indicator that characterizes the overall level of noise energy percentage throughout the entire time period. and These are the L2 norms of the noise sequence and the unperturbed reference condition response signal over the entire time period, respectively.

[0166] The time-varying nonstationarity is shown below:

[0167] (16)

[0168] In the formula, the time-varying nonstationarity index TVNS is a comprehensive index characterizing the time-varying nonstationarity of noise over the entire time period. Let Variance be the operator for all sliding window root mean square statistic sequences. Noise sequence In the Root mean square statistic over a sliding window For the first Length of each sliding window For the first A sliding window.

[0169] Example 11:

[0170] A well transient electromagnetic multi-source noise coupling decomposition and quantization system based on any one of the methods described in Examples 1-10 includes a geometric and parametric modeling module, a multi-physics coupling forward modeling and condition generation module, a controllable parameter scanning and condition labeling module, a comparison difference and interaction component extraction module, a coupling decomposition structured expression and quantization module, and a structured noise sample library export module.

[0171] The geometry and parametric modeling module constructs a three-dimensional geometric model of the wellbore-formation-transmitting coil and completes the parametric definition of physical property parameters and excitation current;

[0172] The multiphysics coupling forward modeling and working condition generation module solves the electromagnetic diffusion control equation under quasi-static conditions, introduces typical transient electromagnetic noise sources in the well, applies high-frequency pose vibration disturbance to the transmitting coil using an adaptive dynamic mesh, applies a layered temperature field in the surrounding rock formation, and adds a small-loop AC coil as a narrowband harmonic additional source term. Based on the time-varying update of pose disturbance, temperature-conductivity mapping, and electromagnetic diffusion control equation coupled with harmonic additional source term, the module obtains the undisturbed reference working condition response signal and the response signals of each noisy working condition.

[0173] The response signals of the unperturbed reference working condition and each noisy working condition were solved using the finite element discretization method; during the solution process, an infinite element equivalent domain was applied to the outer boundary.

[0174] The controllable parameter scanning and operating condition labeling module scans and combines the amplitude, frequency, phase, spatial location, action window and coupling strength of each noise source to construct a set of noisy response signals under single-source and multi-source superposition operating conditions, and forms operating condition labels that correspond one-to-one with the noise source parameters.

[0175] The contrast difference and interaction component extraction module aligns the corresponding components of the response signal with a unified time sampling point and a unified depth sampling point as the alignment reference, and performs contrast difference extraction of the noise sequence based on the unperturbed reference working condition response and the noise working condition response. At the same time, it outputs multi-source mixed interaction components to achieve traceable separation of noise components and residual signals.

[0176] The coupled decomposition structured representation and quantization module constructs a structured noise representation based on single-source noise sequences, combined noise sequences, and multi-source mixed interaction components. It represents multi-source noise as a decomposable, traceable, and mappable data structure, and quantifies the noise intensity, energy ratio, and time-varying nonstationarity of single-source noise components and combined noise components.

[0177] The structured noise sample library export module associates and stores the structured noise decomposition results and their quantitative characterization results, and exports them in batches to form a structured noise sample library that can be used for simulation verification, algorithm evaluation and dataset construction.

[0178] Example 12:

[0179] See Figures 1 to 7 A well-drilled transient electromagnetic multi-source noise coupling decomposition and quantization method based on multi-physics mechanism perturbation includes the following steps:

[0180] 1) Construct a three-dimensional geometric model of the wellbore-formation-transmitting coil, and define the model's physical property parameters and the time-domain excitation current of the transmitting coil. Based on Maxwell's equations in the time domain, the electromagnetic diffusion control equations were derived, and the transient electromagnetic response signal in the well under undisturbed baseline conditions was calculated. ;

[0181] 2) Introducing typical transient electromagnetic noise sources in the well, a high-frequency pose vibration disturbance is applied to the transmitting coil using an adaptive dynamic mesh. A layered temperature field is applied in the surrounding rock formation, and a small-loop AC coil is added as a narrowband harmonic additional source term. Based on the time-varying update of the pose disturbance, temperature-conductivity mapping, and electromagnetic diffusion control equations coupled with harmonic additional source terms, the response signal under mechanical vibration conditions is calculated. Response signal under temperature change conditions Response signal under cable harmonic interference conditions ;

[0182] 3) Based on the controllable parameterization strategy, the amplitude, frequency, phase, spatial location, action window and coupling strength of each noise source are scanned and combined to construct a set of transient electromagnetic noise response signals in well under single-source and multi-source superposition conditions, and form a working condition label corresponding to the noise source parameters one by one.

[0183] 4) Align the noisy response signal with the baseline operating condition response signal using the same time sampling point and the same depth sampling point as the alignment reference, and extract the noise component using the comparison difference method to obtain the noise sequence. Interaction components with multi-source mixtures This enables traceable separation of noise components from residual signals;

[0184] 5) Based on the noise sequence and multi-source mixed interaction components, a structured expression and quantification system for coupling decomposition of transient electromagnetic multi-source noise in wells is constructed. The multi-source noise is characterized as a decomposable, traceable, and mappable data structure composed of single-source noise components, combined noise components, and multi-source mixed interaction components. The noise intensity, energy ratio, and time-varying nonstationarity are quantified.

[0185] 6) The structured noise decomposition results and their quantitative characterization results are associated, integrated, stored, and exported in batches to form a structured noise sample library that can be used for simulation verification, algorithm evaluation, and dataset construction.

[0186] Example 13:

[0187] A method for decomposition and quantization of transient electromagnetic multi-source noise coupling in wells based on multi-physics mechanism disturbance, with the same technical content as in Embodiment 12, wherein the wellbore-formation-transmitting coil three-dimensional geometric model includes a wellbore space domain, a surrounding rock formation domain, a transmitting coil domain, a signal receiving array domain, and an equivalent small loop AC coil domain, wherein the surrounding rock formation domain is composed of discretized medium units, and allows parameterization of resistivity, conductivity, layer thickness, number of layers, and interface position according to geological conditions.

[0188] The parameterized definition of the three-dimensional geometric model includes well diameter and wellbore trajectory parameters, and wellbore conductivity. Axial continuous stratification variation parameters, surrounding rock resistivity The parameters assigned to the transmitter coil, the geometric and electrical parameters of the signal receiving array, and the controllable parameters of typical noise sources are included. The geometric and electrical parameters of the coil include the coil radius. Number of turns Axial position Radial eccentricity The amplitude of the coil excitation current relative to the normal direction With shutdown time .

[0189] Mechanical vibration parameters include vibration displacement. Vibration frequency With phase Cable harmonic parameters include harmonic frequencies. Amplitude With phase Temperature variation parameters include stratified temperature disturbances. And its operating window. Through the above parameterization definition, geometric parameters, source term parameters, and noise parameters can be incorporated into a unified input interface, and... Figure 1 The three-dimensional geometric model shown and Figure 4 The noise controllable generation process shown corresponds to this.

[0190] Example 14:

[0191] A method for decomposing and quantizing coupled transient electromagnetic multi-source noise in wells based on multi-physics perturbation, with technical content identical to any one of Embodiments 12 to 13, wherein the numerical solution of the electromagnetic field employs the electromagnetic diffusion control equations of the time-domain Maxwell's equations under quasi-static conditions, and boundary conditions of continuous tangential electric field and continuous normal magnetic flux density are applied at the multi-medium interface. The time-domain Maxwell's equations and the current constitutive relation are constructed, namely:

[0192] （1） （2）

[0193] in For electric field strength, It represents the magnetic flux density. The magnetic field strength, For current density, Permeability, For conduction current density, The source term current density of the transmitting coil, To account for the harmonic interference source current density corresponding to the equivalent current carrying capacity of the additional small-loop AC coil, Temperature-dependent conductivity. Introducing magnetic vector potential. With electric scalar potential :

[0194] （3）

[0195] The electromagnetic diffusion control equations were then obtained:

[0196] （4）

[0197] The transmitting coil source term Excitation current from the transmitting coil Together with the spatial distribution of the current-carrying domain of the coil, it is determined, and determined The number of turns of the transmitting coil. The equivalent current-carrying cross-sectional area of ​​the transmitting coil. The unit tangent vector in the current-carrying direction. For the current-carrying domain of the transmitting coil If the indicator function is such that the source term of the transmitting coil satisfies:

[0198] （5）

[0199] Harmonic source term Harmonic current from small-circuit AC coil Together with its spatial distribution, it determines and determines For the number of turns of the small-loop AC coil, Its equivalent current-carrying cross-sectional area, Let it be the unit tangent vector in its current-carrying direction. For harmonic current-carrying domain The indicator function, then the additional small loop source term satisfies:

[0200] （6） （7）

[0201] This process allows for the simultaneous description of the reference emission source and harmonic additional source terms within a unified governing equation framework.

[0202] In the model, the interface boundary conditions For the interface normal vector, and The electric field strength on both sides of the interface, and Given the magnetic flux density on both sides of the interface, the boundary conditions for the multi-medium interface satisfy:

[0203] （8）

[0204] An infinite-element equivalent domain is applied outside the computational domain to ensure that the error of the diffusion field is controllable under finite-domain truncation.

[0205] Mechanical vibration is achieved using an arbitrary Lagrange-Euler moving mesh method, causing the transmitter coil pose to change over time and synchronously updating the current-carrying domain of the transmitter coil. and and determine The reference point for the geometric center of the equivalent current-carrying domain of the transmitting coil at time... Spatial position vector, Let be the initial position vector under undisturbed reference conditions. The coil displacement perturbation is used to characterize the time-varying change of the transmitting coil pose under mechanical vibration conditions. The coil displacement perturbation satisfies:

[0206] （9）

[0207] Temperature change is derived based on the heat conduction equation and coupled with the electromagnetic diffusion equation through conductivity-temperature mapping. The heat conduction equation is:

[0208] （10）

[0209] in For the density of the medium, For isobaric specific heat capacity, Thermal conductivity, The heat source intensity per unit volume.

[0210] Then the temperature field Temperature-dependent conductivity acting on the electromagnetic diffusion control equation and determine Initial calibration temperature Lower reference conductivity Let be the temperature coefficient of electrical conductivity. Then, the mapping coupling between the heat conduction equation and the electrical conductivity temperature is achieved, and the mapping relationship is:

[0211] （11）

[0212] Therefore, the three physical processes of mechanical vibration, temperature change, and cable harmonic interference are incorporated into a unified solution framework.

[0213] The three-dimensional geometric model is solved in the time domain using finite element discretization, including local densification of the unstructured mesh and adaptive mesh adjustment for the transmitting coil domain, the well space domain, the additional interference AC coil domain, and the well-ground interface domain; under mechanical vibration conditions, a dynamic mesh update strategy is used to smooth and reconstruct the mesh, so that the mesh maintains the element mass constraint and avoids element distortion as the coil pose changes; the outer boundary of the computational domain is treated with an infinite element equivalent domain to reduce non-physical reflection errors caused by finite domain truncation.

[0214] The electromagnetic diffusion control equations are solved using a time-domain finite element discretization algorithm based on a hybrid form of magnetic vector potential and electric scalar potential. For spatial discretization, the magnetic vector potential is discretized using tetrahedral edge vector elements, and the electric scalar potential is discretized using nodal scalar elements to ensure numerical consistency of the curl term, divergence constraint, and continuity condition of the multi-medium interface. For time discretization, an implicit backward difference scheme is used to advance step-by-step. Within each time step, the electromagnetic diffusion control equations are discretized into a sparse algebraic equation system concerning the magnetic vector potential and electric scalar potential at the current time. After solving, the electromagnetic field distribution at the current time is obtained, and the induced electromotive force signal corresponding to the received response in the well is calculated.

[0215] The dynamic mesh update under mechanical vibration conditions employs an algorithm combining boundary displacement driving, mesh smoothing, and local reconstruction under the arbitrary Lagrange-Euler method. First, the change in the transmitter's pose at the current moment is calculated based on the coil displacement perturbation equation, and this change is applied as the boundary displacement condition for the transmitter's current-carrying domain. Then, the boundary displacement is propagated to the surrounding mesh using the mesh smoothing control equation, allowing neighboring elements to deform continuously with pose changes while maintaining topological connectivity. When the mass of a local element drops below a preset threshold, local reconstruction is triggered, the distorted element region is re-divided, and the field variables obtained in the previous time step are interpolated and mapped to the new mesh before continuing time-domain propagation calculations.

[0216] To ensure the repeatability and controllable error of the numerical results, the unstructured mesh adopts a partitioned local refinement strategy. The transmitter loop current-carrying domain, the additional small loop AC coil domain, and their neighboring domains are set as the first-level refinement zone, with a minimum mesh cell size of 1×10⁻⁶. -3 The wellbore space domain and its vicinity are designated as a secondary refinement zone, while the far-field surrounding rock region uses a gradually coarsening mesh connected to the infinite element equivalent domain. Mesh convergence is tested using the response curve of the induced electromotive force received in the well as the criterion. When other parameters remain constant and the minimum element size of the primary refinement zone is reduced to half of the previous size, if the relative change rate of the L2 norm of the full-time window response curve is less than 1%, and the relative change rate of the amplitude of the mid-to-late stage key sampling points is less than 0.5%, then the current mesh is considered to have met the convergence requirements. When the local element quality factor falls below 0.20 after a dynamic mesh update, local reconstruction is performed to avoid significant impacts on the accuracy of the time-domain response calculation due to element distortion. The above mesh partitioning accuracy control indicators and convergence criteria are representative settings used in this embodiment to ensure solution stability and result repeatability. Similar adjustments can be made based on model scale, noise parameter range, and computational resource conditions without departing from the technical concept of this invention.

[0217] The transient electromagnetic response signal output in the well is an induced electromotive force. Based on the response of the unperturbed baseline operating condition Response to single-source noise conditions Calculate a single noise sequence:

[0218] （12）

[0219] Based on multi-source superimposed working condition response signals Calculate the combined noise sequence:

[0220] （13）

[0221] And the multi-source mixing interaction components are defined as:

[0222] （14）

[0223] This enables the differential extraction and traceable output of individual noise, combined noise, and interacting components.

[0224] The quantitative characterization index system includes noise intensity index, energy proportion index, and time-varying non-stationarity index, wherein the noise intensity index is:

[0225] （15）

[0226] The Noise Intensity Index (NI) is an indicator that characterizes the overall noise intensity level over all time periods. The single-source and combined noise sequences to be evaluated are as follows: and These are the L2 norms of the noise sequence and the unperturbed reference operating condition response signal over the entire time period, respectively. To avoid numerically unstable regularization variables, a value of 10 times the magnitude of the characteristic value of the denominator is typically taken. -6 Up to 10 -3 times.

[0227] The energy percentage index is defined as:

[0228] （16）

[0229] The energy percentage index (NEP) is a comprehensive indicator that represents the proportion of noise energy across all time periods. and These are the L2 norms of the noise sequence and the unperturbed reference condition response signal over the entire time period, respectively.

[0230] The time-varying nonstationarity index is defined based on the sliding window statistic as follows:

[0231] （17）

[0232] The Time-Variable Nonstationarity Index (TVNS) is a comprehensive index characterizing the time-varying nonstationarity of noise over the entire time period. Let Variance be the operator for all sliding window root mean square statistic sequences. Noise sequence In the Root mean square statistic over a sliding window For the first Length of each sliding window For the first A sliding window.

[0233] Example 15:

[0234] A method for coupled decomposition and quantification of transient electromagnetic multi-source noise in wellbore based on multi-physics mechanism perturbation is provided. The technical content is the same as any one of Embodiments 12 to 14. This embodiment provides a specific construction method for the single-source working condition of mechanical vibration noise to verify the perturbation law of the time-varying position of the transmit loop on the transient electromagnetic response in wellbore.

[0235] This embodiment employs the same unified three-dimensional geometric model and multiphysics coupled solution framework as described in Embodiments 13 and 14. The surrounding rock strata have a side length of 2×10⁻⁶. 3 m, conductivity taken as 2×10 -3 S / m, relative permittivity is taken as 1, relative permeability is taken as 1, and constant-pressure heat capacity is taken as 9 × 10⁻⁶. 2 J / (kg·K), density is taken as 2.6×10 3 kg / m 3The thermal conductivity is taken as 3 W / (m·K). The wellbore radius is 0.1 m, and the depth is 2×10⁻⁶ m. 3 m, the medium inside the well is air, and the conductivity is taken as 1×10⁻⁶. -5 S / m, relative permittivity is taken as 1, relative permeability is taken as 1. Emission loop side length is 2 × 10⁻⁶. 2 m, the cross-sectional area of ​​the conductor is 1×10 -6 m 2 The number of turns is 1, and the conductivity is taken as 6×10. 7 S / m, relative permittivity is taken as 1, relative permeability is taken as 1, and constant-pressure heat capacity is taken as 3.85×10 2 J / (kg·K), density is taken as 8.94×10 3 kg / m 3 The thermal conductivity is taken as 4×10 2 W / (m·K). An additional small loop is set on the ground surface and concentric with the wellhead, with a side length of 10 m, for subsequent construction of harmonic and coupling operating conditions.

[0236] The settings for the aforementioned geometric dimensions, material parameters, and excitation parameters adhere to the principles of consistency in magnitude for typical detection conditions, feasibility of noise disturbance mechanisms, and stability and identifiability of numerical solutions. Specifically, a kilometer-scale computational domain is used for the surrounding rock strata to fully cover the propagation and diffusion range of the detection response at different depths and to reduce the impact of lateral boundary truncation on the electromagnetic diffusion field distribution and response signal calculation results within the study window. The electrical conductivity of the surrounding rock is taken as 10. -3 The S / m scale is used to characterize the background of low- to medium-conductivity surrounding rock. The wellbore uses a decimeter-scale radius and a kilometer-scale depth to characterize the downhole exploration spatial scale. The transmit loop uses a hundred-meter-scale scale and tens of ampere-level excitation current, and is equipped with millisecond-level current-on and turn-off processes to form a wellbore transient electromagnetic reference response with identifiable attenuation characteristics. The conductivity of the transmit loop conductor is taken as 10. 7 The S / m order of magnitude corresponds to the order of magnitude of engineering materials for high-conductivity metallic conductors. Therefore, the background model parameters and source parameters in this embodiment are representative parameter combinations selected under the premise of physical rationality, conformity to actual working conditions, engineering feasibility, and ease of parametric scanning verification.

[0237] In this embodiment, the excitation current of the transmitting coil is adopted. Figure 2 The waveform shown in (b) indicates a current amplitude of 30 A for 20 ms, which is then switched off to 0 A within 0.5 ms starting from 20 ms, and then held at 0 A for 19.5 ms. The mechanical vibration displacement load is... Figure 2 The function shown in (a) is applied to the pose degree of the launch loop according to equation (9), and its time-domain expression corresponds to a vibration frequency of 1×10. 3The Hz frequency ranges from 0 mm to 2 mm, with an average vibration displacement of 1 mm. During the solution process, an arbitrary Lagrange-Euler moving grid is used to synchronously update the position of the transmitting coil and couple it with the spatial distribution of the source terms of the transmitting coil, thereby obtaining the response signal under mechanical vibration conditions.

[0238] The aforementioned mechanical vibration parameters are used to describe the high-frequency, minute pose shifts of the excitation device under the coupling effect of mechanical disturbances. The set displacement amplitude is on the order of millimeters, significantly smaller than the wellbore radius and the characteristic scale of the transmission loop, falling within the range of small disturbances that do not alter the device's topology. The set vibration frequency is taken as a representative value among high-frequency micro-vibrations in the range of hundreds of hertz to kilohertz, used to characterize dynamic disturbance conditions that have a significant modulating effect on the late-stage local waveform structure in the transient response. The average vibration displacement is used to ensure that the coil pose still has an observable geometric offset based on the periodic oscillation. Therefore, this set of parameters not only satisfies the physical feasibility of downhole disturbances but also maintains consistency in order of magnitude with the common small-amplitude, high-frequency, time-varying disturbance characteristics under actual downhole conditions, while ensuring that the mechanical vibration noise has identifiable and quantifiable characterization features in the response signal.

[0239] By applying Figure 2 The mechanical vibration load in (a) is obtained Figure 3 (b) Time-domain response signal under mechanical vibration conditions and Figure 5 (a) The controllable generation results of mechanical vibration noise and Figure 7 The mechanical vibration noise quantification characterization results of n1; and can be obtained through Figure 4 The controllable generation process generates mechanical vibration condition labels under different parameter combinations. These condition labels include at least the mechanical vibration type identifier, displacement amplitude identifier, frequency identifier, phase identifier, and duration identifier.

[0240] Figure 3 (b) The overall decay trend of the response curve under mechanical vibration is consistent with that under undisturbed conditions. However, in the middle and late time windows, local oscillations are enhanced and curve fluctuations are aggravated, indicating that mechanical vibration first changes the spatial coupling relationship between the transmitting loop and the medium and receiving system, which then manifests as dynamic disturbance at the response end. Figure 3 (a) Using the baseline, perform comparative difference analysis according to equation (12) to obtain Figure 5 (a) shows the time-domain diagram of mechanical vibration noise. Figure 5 The mechanical vibration noise shown in (a) is not the strongest in terms of overall energy level, but it has significant local high-frequency fluctuations, indicating that mechanical vibration noise is more sensitive to short-time dynamic waveform structures. Then, calculate according to equations (15) to (17). Figure 7The quantization result corresponding to n1 yields the characterization values ​​of mechanical vibration noise in three dimensions: noise intensity, energy proportion, and time-varying non-stationarity. Thus, this embodiment completes a full closed loop from input load to response output, component extraction, and index output.

[0241] The above parameter values ​​are representative example values ​​used to verify the effectiveness of the method of the present invention. Without departing from the technical concept of the present invention, they can be adjusted within the same order of magnitude according to different well depths, well diameters, surrounding rock electrical conditions, launch device sizes and downhole disturbance intensity, and still fall within the protection scope of the controllable generation and quantitative characterization method of mechanical vibration noise described in the present invention.

[0242] Example 16:

[0243] A method for coupling decomposition and quantification of transient electromagnetic multi-source noise in wells based on multi-physics mechanism perturbation is provided. The technical content is the same as any one of Examples 12 to 15. This example provides a specific construction method for the single-source working condition of temperature change noise to verify the systematic interference of the layered thermal field on the transient electromagnetic diffusion process in the well after conductivity temperature mapping.

[0244] This embodiment follows 2×10 3 The surrounding rock strata are divided into four temperature zones along the depth direction, each with a thickness of 5×10. 2 m, with temperatures of 303.15 K, 313.15 K, 323.15 K, and 333.15 K from top to bottom. Except for the temperature parameters, the other electromagnetic, thermal, and geometric parameters are the same as in Example 15. The temperature field is solved according to Equation (10), the temperature-conductivity mapping is performed according to Equation (11), and the temperature-related conductivity is fed back to the electromagnetic diffusion control equation to obtain the response signal under temperature change conditions.

[0245] Figure 3 (d) shows the time-domain plot of the response signal under temperature variation conditions. This response signal does not exhibit high-frequency fluctuations or narrow-band oscillations, but rather shows an overall shift in the amplitude trajectory throughout the entire time period, a change in the late-stage attenuation slope, and a rearrangement of the relative amplitude relationships in different time segments. This result indicates that temperature variation noise is a typical slow-varying, diffusion-path-modulated noise. Its mechanism is neither a geometric instantaneous disturbance nor an external periodic source, but rather it exerts a systematic influence on the electromagnetic diffusion propagation process by altering the formation's conductive structure.

[0246] by Figure 3 (a) Using the base equation (12) as the difference, we can obtain Figure 5(c) shows the time-domain plot of the temperature change noise. Temperature change noise exhibits a continuous, smooth, low-oscillation, and overall offset morphological characteristic, indicating that it primarily alters the macroscopic attenuation structure of the response rather than local high-frequency components. Then, calculate according to equations (15) to (17). Figure 7 The quantization result corresponding to n3 provides a quantitative characterization of temperature change noise in three dimensions: overall intensity, energy contribution, and non-stationarity. Thus, this embodiment completes a full closed loop from stratified temperature setting to response formation, component extraction, and index output.

[0247] Example 17:

[0248] A method for decomposition and quantization of transient electromagnetic multi-source noise coupling in wells based on multi-physics mechanism perturbation is provided. The technical content is the same as any one of Embodiments 12 to 16. This embodiment provides the construction method of single-source and multi-source coupled cable harmonic noise to verify the coupling generation capability of the additional small loop AC source term and other noise sources under a unified forward modeling framework.

[0249] In this embodiment, the additional small loop excitation current is adopted. Figure 2 The waveform function shown in (c) has a current amplitude of 0.5 A and a frequency of 500 Hz. The additional small loop is located on the ground surface, with its center coinciding with the center of the wellhead. The loop has a side length of 10 m, and other material parameters are the same as in Example 15. The AC source term of the additional small loop participates in the electromagnetic diffusion solution through the harmonic source term. Its physical significance lies in equating the periodic interference introduced by the power supply link or cable laying to a narrowband additional source with a definite amplitude, frequency, and spatial location.

[0250] Figure 3 (c) shows the time-domain response signal of the cable under single-source harmonic operation. The late time window of this response exhibits significant periodic amplitude oscillations and local zero-crossing structures, indicating that the harmonic noise is a typical narrowband periodic disturbance. Its time pattern is clearly distinguishable from the local high-frequency fluctuations of mechanical vibration noise and the smooth shift characteristics of temperature change noise. Figure 3 (a) Using the baseline, perform comparative difference analysis according to equation (12) to obtain Figure 5 (b) shows the time-domain diagram of cable harmonic noise; then calculate according to equations (15) to (17). Figure 7 The quantization result corresponding to n2 can be used to form an independent spectral label for cable harmonic noise.

[0251] Based on the single-source harmonic operating condition, this embodiment constructs three types of coupled operating conditions: mechanical vibration coupled with cable harmonics, mechanical vibration coupled with temperature change, cable harmonics coupled with temperature change, and a three-source coupled operating condition where all three coexist. These correspond to... Figure 3 (e) Figure 3 (f) Figure 3 (g) and Figure 3 (h). These coupled operating conditions are not simply intended to increase the number of samples, but rather to verify the combined generation capability of multi-source disturbances within the framework of unified governing equations. The time-domain curves under coupled operating conditions no longer maintain the pure form of a single noise source, but instead exhibit a composite distortion mode with oscillations, shifts, local enhancements, and attenuation structural rearrangements. Figure 3 (a) Based on the standard, the combined noise is extracted according to equation (13), and the result is obtained. Figure 5 (d) to Figure 5 (g) shows the time-domain plot of the coupled noise, and respectively in Figure 7 The quantization results correspond to n4, n5, n6, and n7. This embodiment demonstrates that the present invention can not only stably generate three types of single-source noise, but also generate multi-source coupled noise within a physically consistent framework, and form operating condition labels and index labels that correspond one-to-one with the parameter configuration.

[0252] Example 18:

[0253] A method for decomposition and quantization of transient electromagnetic multi-source noise in wells based on multi-physics perturbation is provided. The technical content is the same as any one of Embodiments 12 to 17. This embodiment describes the decomposition results of single-source noise components, combined noise components and multi-source mixed interaction components, as well as the quantization characterization results of single-source noise components and combined noise components, in order to verify the complete output capability of the present invention from noise generation to noise analysis to noise evaluation.

[0254] By using differential processing, the independent contribution of a single source, the overall contribution of combined noise, and the multi-source mixed interaction components can be explicitly separated. Based on the single-source operating condition and coupled operating condition response signals generated in Examples 15 to 17 and the unperturbed reference operating condition response shown, the single-source noise component is extracted according to Equation (12), the combined noise component is extracted according to Equation (13), and the multi-source mixed interaction components are calculated according to Equation (14). Figure 5 (a) to Figure 5 (g) Corresponds in sequence to mechanical vibration noise, cable harmonic noise, temperature change noise, coupling noise between mechanical vibration and cable harmonics, coupling noise between mechanical vibration and temperature change, coupling noise between cable harmonics and temperature change, and simultaneous coupling noise from three sources. Mechanical vibration noise is mainly characterized by local high-frequency fluctuations, cable harmonic noise is mainly characterized by narrow-band periodic oscillations, temperature change noise is mainly characterized by smooth shifts and trend reshaping, while coupling noise exhibits time window rearrangement and composite modulation characteristics.

[0255] After obtaining the combined noise components, the interaction contribution between the multiple sources is calculated. Figure 6This is a multi-source coupling interaction component analysis diagram, used to characterize the interaction components and their time-domain characteristics obtained by subtracting the independent contributions of each individual source from the mixed response under multi-source superposition conditions. The physical reason for the generation of these interaction components is that the various disturbances are not independent additive variables acting on the same level, but rather act on three different physical levels: geometric pose, medium parameters, and additional source terms. They propagate and couple together under a unified electromagnetic diffusion control equation, ultimately mapping to the receiver. Therefore, the mixed response is no longer strictly equal to the simple algebraic sum of the individual source responses.

[0256] When mechanical vibration couples with cable harmonics, the time-varying pose of the transmitting loop alters the relative geometric relationships and mutual inductance coupling strength between the transmitting source, the additional harmonic source, and the receiving system. This causes the original narrowband periodic interference to be modulated by time-varying amplitude and phase, resulting in a coupling residue term that cannot be directly obtained by adding the contributions from the single source of mechanical vibration and the single source of harmonics. When mechanical vibration couples with temperature changes, the response sensitivity of the medium to the excitation field at different times and poses varies after temperature changes. Therefore, the spatiotemporal distribution changes of the source field caused by mechanical vibration will exhibit different propagation and response results in the medium after temperature changes, resulting in a coupling residue term under the combined effect of the geometric disturbance of the transmitting source and the change in underground medium parameters. When cable harmonics couple with temperature changes, the propagation attenuation, phase lag, and induced current closed path of the additional harmonic source change under different conductivity distributions, making the harmonic contribution dependent on the temperature field distribution. This results in a coupling residue term under the combined effect of the narrowband additional source term and the medium parameter mapping. When all three are coupled simultaneously, the aforementioned geometric pose modulation, additional source term modulation, and medium parameter modulation coexist, resulting in a more complex composite distortion structure in the interaction components. The multi-source interaction components essentially characterize the additional coupling contribution exceeding the sum of the independent contributions of a single source under the combined action of multiple physical mechanisms. They can be used to reveal the coupling strength, action path, and non-independent separability characteristics between noises. This result substantially distinguishes this invention from traditional empirical noise addition and simple linear superposition models.

[0257] After obtaining the noise components, the quantification characterization does not involve calculating several statistics on the overall noise in isolation, but rather performs unified quantification on the single-source noise components and combined noise components obtained through differential and coupled decomposition. The quantification index system includes a noise intensity index, an energy proportion index, and a time-varying nonstationarity index. The noise intensity index is used to characterize the overall intensity level of noise throughout the entire time period, the energy proportion index is used to characterize the energy contribution of noise relative to the reference response, and the time-varying nonstationarity index is used to characterize the time-varying fluctuation characteristics of noise under the statistical significance of a sliding time window.

[0258] The structured representation of this invention does not originate from the parallel summarization of multiple indicators, but rather from the internal hierarchical structure of multi-source noise after coupling decomposition. By comparing the difference and multi-source coupling decomposition, the multi-source noise, which originally existed as a whole disturbance, is represented as a structured noise object composed of single-source noise components, combined noise components, and multi-source mixed interaction components, thereby realizing noise decomposition, parameter backscaling, sample mapping, and structured sample management.

[0259] Figure 7 It can be used to characterize the differences in three categories of indicators: noise intensity, energy ratio, and time-varying nonstationarity under different noise types and coupling conditions. The seven noise conditions, n1 to n7, are uniformly quantified according to the quantification index formula. Figure 7 In the quantization results, n1 corresponds to mechanical vibration noise, n2 to cable harmonic noise, n3 to temperature change noise, n4 to the coupled noise of mechanical vibration and cable harmonics, n5 to the coupled noise of mechanical vibration and temperature change, n6 to the coupled noise of cable harmonics and temperature change, and n7 to the coupled noise of mechanical vibration, cable harmonics, and temperature change. The quantization results are used for component comparison between single-source noise components and combined noise components, parameter back-scaling, and sample mapping.

[0260] Example 19:

[0261] A well transient electromagnetic multi-source noise coupling decomposition and quantization system based on the method described in any one of Examples 12 to 18 includes a geometric and parametric modeling module, a multi-physics coupling forward modeling and condition generation module, a controllable parameter scanning and condition labeling module, a controllable difference and interaction component extraction module, a coupling decomposition structured expression and quantization module, and a structured noise sample library export module.

[0262] The geometry and parametric modeling module is used to construct a three-dimensional geometric model of the wellbore-formation-transmitting coil, and to complete the parametric definition of physical property parameters, boundary conditions, excitation current, and parameters of each noise source.

[0263] The multiphysics coupling forward modeling and operating condition generation module is used to solve the electromagnetic diffusion control equation under quasi-static conditions, and to generate unperturbed reference operating conditions and various noisy operating condition response signals through dynamic mesh, heat conduction equations and additional harmonic source terms.

[0264] The controllable parameter scanning and operating condition labeling module is used to scan and combine the amplitude, frequency, phase, spatial location, action window and coupling strength of each noise source to construct a set of noisy response signals under single-source and multi-source superposition conditions, and form operating condition labels that correspond one-to-one with the noise source parameters.

[0265] The control difference and interaction component extraction module is used to align the corresponding components of the response signal with a unified time sampling point and a unified depth sampling point as the alignment reference, and extract the noise sequence and multi-source mixed interaction components based on the control difference to achieve traceable separation of noise components and residual signals.

[0266] The coupled decomposition structured representation and quantization module is used to construct a structured noise representation based on single-source noise sequences, combined noise sequences, and multi-source mixed interaction components. It represents multi-source noise as a decomposable, traceable, and mappable data structure, and quantifies the noise intensity, energy ratio, and time-varying nonstationarity of single-source noise components and combined noise components.

[0267] The structured noise sample library export module is used to associate and store the structured noise decomposition results and their quantitative characterization results, and export them in batches to form a structured noise sample library that can be used for simulation verification, algorithm evaluation and dataset construction.

[0268] The system can be deployed on a software platform or integrated data processing platform that supports multiphysics forward modeling. It is used for simulation verification of transient electromagnetic multi-source noise in wells, noise mechanism research, evaluation of denoising or reconstruction algorithms, and construction of training datasets. Through this system, a closed-loop process can be achieved from parameter input, condition generation, difference extraction, quantitative evaluation to sample export.

Claims

1. A method for coupled decomposition and quantization of transient electromagnetic multi-source noise in wells based on multi-physics mechanism perturbation, characterized in that, Includes the following steps: Step 1) Construct a three-dimensional geometric model of the wellbore-formation-transmitting coil, and define the physical property parameters of the three-dimensional geometric model of the wellbore-formation-transmitting coil and the time-domain excitation current of the transmitting coil. ; Based on the time-domain Maxwell's equations, the electromagnetic diffusion control equations were derived, and the transient electromagnetic response signal in the well under undisturbed baseline conditions was calculated. ; Step 2) Apply high-frequency pose vibration perturbation to the transmitting coil, update the coil pose using an adaptive moving mesh method, and modify the new electromagnetic diffusion control equations based on the coil displacement perturbation and the source terms of the transmitting coil, to calculate the response signal under mechanical vibration conditions. ; A layered temperature field is applied to the surrounding rock strata, and the response signal under temperature variation conditions is calculated based on the coupling of the heat conduction equation, the electrical conductivity-temperature mapping relationship, and the electromagnetic diffusion control equation. ; A small-loop AC coil is added as a narrowband harmonic additional source term. Based on the electromagnetic diffusion control equation of the superposition of the transmitting coil source term and the harmonic additional source term, the response signal under cable harmonic interference conditions is calculated. ; Step 3) Based on the controllable parameterization strategy, the amplitude, frequency, phase, spatial location, action window and coupling strength of each noise source are scanned and combined to construct a set of transient electromagnetic noise response signals in the well under single source and multi-source superposition conditions, and form a condition label corresponding to the noise source parameters one by one. Step 4) Align the noisy response signal with the baseline operating condition response signal, and extract the noise component using a differential method to obtain the noise sequence. Interaction components with multi-source mixtures This results in a traceable separation of noise components and residual signals; Step 5) Based on the noise sequence and multi-source mixed interaction components, construct a structured expression and quantification system for coupling decomposition of transient electromagnetic multi-source noise in wells. Characterize multi-source noise as a decomposable, traceable, and mappable data structure composed of single-source noise components, combined noise components, and multi-source mixed interaction components, and quantify the noise intensity, energy ratio, and time-varying nonstationarity. Step 6) The structured noise decomposition results and quantitative characterization results are associated, integrated, stored, and exported in batches to form a structured noise sample library.

2. The wellbore transient electromagnetic multi-source noise coupling decomposition and quantization method based on multi-physics mechanism perturbation according to claim 1, characterized in that, The wellbore-formation-transmitting coil three-dimensional geometric model includes the wellbore space domain, the transmitting coil domain, the signal receiving array, and the surrounding rock formation domain; The surrounding rock stratigraphic domain is composed of discretized media units and has parameterizable geological conditions; The wellbore-formation-transmitting coil three-dimensional geometric model also includes an equivalent small loop AC coil domain for cable harmonic interference modeling, coil pose degrees of freedom for mechanical vibration modeling, and a layered temperature field domain for temperature change modeling.

3. The wellbore transient electromagnetic multi-source noise coupling decomposition and quantization method based on multi-physics mechanism perturbation according to claim 2, characterized in that, The physical properties of the three-dimensional geometric model of the wellbore-formation-transmitting coil include wellbore diameter and wellbore trajectory parameters, and in-well conductivity. Axial continuous stratification variation parameters, surrounding rock resistivity The assigned parameters, geometric and electrical parameters of the transmitting coil and signal receiving array, and controllable parameters of typical noise sources; The coil geometry and electrical parameters include the coil radius. Number of turns Axial position Radial eccentricity The amplitude of the coil excitation current relative to the normal direction With shutdown time ; Mechanical vibration parameters include vibration displacement. Vibration frequency With phase ; Cable harmonic parameters include harmonic frequencies. Amplitude With phase ; Temperature variation parameters include stratified temperature disturbances And the window of action.

4. The wellbore transient electromagnetic multi-source noise coupling decomposition and quantization method based on multi-physics mechanism perturbation according to claim 1, characterized in that, Step 2) involves deriving the electromagnetic diffusion control equations, including: Step 2.1) Apply boundary conditions of continuous tangential electric field and continuous normal magnetic flux density at the multi-dielectric interface, and construct the time-domain Maxwell's equations and current constitutive relations, namely: （1） （2） in, For electric field strength, It represents the magnetic flux density. The magnetic field strength, For current density, Permeability, For conduction current density, The source term current density of the transmitting coil, To account for the harmonic interference source current density corresponding to the equivalent current carrying capacity of the additional small-loop AC coil, Temperature-dependent conductivity; Among them, electric field strength and magnetic induction intensity As shown below: Magnetic vector potential With electric scalar potential : （3） In the formula, , For magnetic vector potential and electric scalar potential; Step 2.2) Construct the electromagnetic diffusion control equations, namely: （4） Among them, the transmitting coil source term As shown below: （5） In the formula, This is the excitation current for the transmitting coil; The number of turns of the transmitting coil. The equivalent current-carrying cross-sectional area of ​​the transmitting coil. The unit tangent vector in the current-carrying direction. For the current-carrying domain of the transmitting coil Indicator functions; Harmonic source term As shown below: （6） （7） In the formula, This refers to the harmonic current of the small-loop AC coil. For the number of turns of the small-loop AC coil, Its equivalent current-carrying cross-sectional area, Let it be the unit tangent vector in its current-carrying direction. For harmonic current-carrying domain Indicator functions; Temperature-dependent conductivity As shown below: （8） In the formula, Initial calibration temperature Lower reference conductivity is the temperature coefficient of electrical conductivity.

5. The wellbore transient electromagnetic multi-source noise coupling decomposition and quantization method based on multi-physics mechanism perturbation according to claim 1, characterized in that, The displacement perturbation of the transmitting coil is used to characterize the time-varying change of the transmitting coil pose under mechanical vibration conditions, and to drive the time-varying update of the source terms of the transmitting coil, thereby participating in the calculation of the response signal under mechanical vibration conditions; The displacement disturbance of the transmitting coil is shown below: （9） In the formula, The reference point for the geometric center of the equivalent current-carrying domain of the transmitting coil at time... Spatial position vector, This is the initial position vector of the transmitting coil under undisturbed reference conditions.

6. The wellbore transient electromagnetic multi-source noise coupling decomposition and quantization method based on multi-physics mechanism perturbation according to claim 1, characterized in that, The heat conduction equation is as follows: （10） in For the density of the medium, For isobaric specific heat capacity, Thermal conductivity, The heat source intensity per unit volume.

7. The wellbore transient electromagnetic multi-source noise coupling decomposition and quantization method based on multi-physics mechanism perturbation according to claim 1, characterized in that, Response signal under mechanical vibration conditions Response signal under cable harmonic interference Response signal under temperature change conditions The solution process is as follows: local densification of unstructured mesh and adaptive mesh adjustment are performed on the transmitting coil domain, the well space domain, the additional interference AC coil domain, and the well interface domain; under mechanical vibration conditions, a dynamic mesh update strategy is used to smooth and reconstruct the mesh so that the mesh maintains element mass constraints and avoids element distortion as the coil pose changes. An infinite element equivalent domain is used to process the outer boundary of the computational domain to reduce non-physical reflection errors caused by finite domain truncation; the response signal under mechanical vibration conditions is obtained by solving the finite element discretization method. Response signal under cable harmonic interference Response signal under temperature change conditions .

8. The wellbore transient electromagnetic multi-source noise coupling decomposition and quantization method based on multi-physics mechanism perturbation according to claim 1, characterized in that, Single noise sequence As shown below: （11） Combined noise sequence As shown below: （12） Multi-source hybrid interaction components As shown below: （13） In the formula, It is a mixed response signal under the influence of multiple physical mechanisms of disturbance; This is the response signal under the perturbation of the k-th physical mechanism.

9. The wellbore transient electromagnetic multi-source noise coupling decomposition and quantization method based on multi-physics mechanism perturbation according to claim 1, characterized in that, The noise levels are shown below: （14） In the formula, the noise intensity index NI is an index characterizing the overall noise intensity level over the entire time period. The single-source and combined noise sequences to be evaluated are as follows: and These are the L2 norms of the noise sequence and the unperturbed reference operating condition response signal over the entire time period, respectively. To avoid numerically unstable small regularization values; The energy breakdown is as follows: （15） In the formula, the energy percentage index (NEP) is an indicator that characterizes the overall level of noise energy percentage throughout the entire time period. and These are the L2 square norms of the noise sequence and the unperturbed reference operating condition response signal over the entire time period, respectively. The time-varying nonstationarity is shown below: （16） In the formula, the time-varying nonstationarity index TVNS is a comprehensive index characterizing the time-varying nonstationarity of noise over the entire time period. Let Variance be the operator for all sliding window root mean square statistic sequences. Noise sequence In the Root mean square statistic over a sliding window For the first Length of each sliding window For the first A sliding window.

10. A wellbore transient electromagnetic multi-source noise coupling decomposition and quantization system based on the method of any one of claims 1-9, characterized in that, It includes modules for geometric and parametric modeling, multiphysics coupling forward modeling and condition generation, controllable parameter scanning and condition labeling, differential and interaction component extraction, coupled decomposition structured expression and quantization, and structured noise sample library export. The geometry and parametric modeling module constructs a three-dimensional geometric model of the wellbore-formation-transmitting coil and completes the parametric definition of physical property parameters and excitation current; The multiphysics coupling forward modeling and working condition generation module solves the electromagnetic diffusion control equation under quasi-static conditions, introduces typical transient electromagnetic noise sources in the well, applies high-frequency pose vibration disturbance to the transmitting coil using an adaptive dynamic mesh, applies a layered temperature field in the surrounding rock formation, and adds a small-loop AC coil as a narrowband harmonic additional source term. Based on the time-varying update of pose disturbance, temperature-conductivity mapping, and electromagnetic diffusion control equation coupled with harmonic additional source term, the module obtains the undisturbed reference working condition response signal and the response signals of each noisy working condition. The response signals of the unperturbed reference working condition and each noisy working condition were solved using the finite element discretization method; during the solution process, an infinite element equivalent domain was applied to the outer boundary. The controllable parameter scanning and operating condition labeling module scans and combines the amplitude, frequency, phase, spatial location, action window and coupling strength of each noise source to construct a set of noisy response signals under single-source and multi-source superposition operating conditions, and forms operating condition labels that correspond one-to-one with the noise source parameters. The contrast difference and interaction component extraction module aligns the corresponding components of the response signal with a unified time sampling point and a unified depth sampling point as the alignment reference, and performs contrast difference extraction of the noise sequence based on the unperturbed reference working condition response and the noise working condition response. At the same time, it outputs multi-source mixed interaction components to achieve traceable separation of noise components and residual signals. The coupled decomposition structured representation and quantization module constructs a structured noise representation based on single-source noise sequences, combined noise sequences, and multi-source mixed interaction components. It represents multi-source noise as a decomposable, traceable, and mappable data structure, and quantifies the noise intensity, energy ratio, and time-varying nonstationarity of single-source noise components and combined noise components. The structured noise sample library export module associates and stores the structured noise decomposition results and their quantitative characterization results, and exports them in batches to form a structured noise sample library that can be used for simulation verification, algorithm evaluation and dataset construction.

Images