Core hydration strain monitoring method and device based on multi-source fusion data

Abstract

The present specification relates to the technical field of oil and gas drilling and completion engineering, and particularly relates to a core hydration strain monitoring method and device based on multi-source fusion data, the method comprising: collecting first hydration strain data generated by a core in a hydration process through a core hydration strain monitoring system; inputting the first hydration strain data into a noise reduction model to obtain second hydration strain data after noise reduction; and converting the second hydration strain data into hydration stress data by using an evolution model to obtain distribution and evolution prediction results of the hydration stress in time and space, wherein the evolution model is trained based on strain sample sequence data and stress labels and is used to represent a mapping relationship between hydration strain and hydration stress. The present application accurately identifies core hydration expansion evolution characteristics and predicts wellbore stability, improves experimental analysis efficiency and result reliability, and has important value for studying wellbore stability monitoring and prevention and control.

Description

Technical Field

[0001] This specification relates to the field of oil and gas drilling and completion engineering technology, and in particular to a core hydration strain monitoring method and device based on multi-source fusion data. Background Technology

[0002] During oil and gas well drilling and completion operations, the surrounding rock of the wellbore is prone to hydration and expansion after contact with drilling fluid, leading to changes in formation stress and wellbore deformation. In severe cases, this can cause wellbore instability, affecting drilling safety and efficiency. Experimental research on the evolution of wellbore core hydration and expansion is of great significance for predicting and preventing wellbore instability.

[0003] Currently, core hydration swelling experiments typically employ conventional equipment combined with point sensors for monitoring, such as using resistance strain gauges to measure strain at a few locations on the core surface. However, such monitoring methods have limited data coverage, cannot continuously acquire the strain distribution throughout the entire core, and are difficult to accurately capture the dynamic evolution of hydration swelling.

[0004] Distributed fiber optic strain monitoring technology enables long-distance, high-resolution continuous strain measurement along a single optical fiber, making it highly suitable for monitoring the entire deformation process of soil and rock media. In the field of geotechnical engineering, there are reports of using distributed fiber optic sensing for deformation monitoring of structures such as slopes and tunnels. However, there are currently no reports of experimental methods using distributed fiber optic strain monitoring for the evolution of hydration and expansion in well core samples.

[0005] With the development of monitoring technology and the surge in data volume, artificial intelligence and big data technologies are playing an increasingly prominent role in monitoring data analysis. Currently, research and applications of artificial intelligence and big data technologies have shown initial success in fields such as intelligent drilling and completion, subsurface mechanics analysis, and structural health monitoring. For example, in the intelligent drilling and completion process of oil and gas wells, machine learning algorithms can automatically analyze the large amounts of data collected by downhole sensors to identify wellbore stability trends and predict wellbore instability risks. Similarly, in structural health monitoring, combining distributed fiber optic sensing with artificial intelligence data analysis can automatically identify abnormal deformation and damage signs in structures, enabling early warning. However, there is currently no technology for in-depth research on the evolution process of wellbore core hydration and expansion. Summary of the Invention

[0006] To address the problems in the prior art, this specification provides a method and apparatus for monitoring core hydration strain based on multi-source fusion data. The method includes: acquiring first hydration strain data generated during the hydration process of the core using a core hydration strain monitoring system; inputting the first hydration strain data into a denoising model to obtain denoised and restored second hydration strain data; and using an evolution model to convert the second hydration strain data into hydration stress data to obtain the temporal and spatial distribution and evolution prediction results of hydration stress. The evolution model is trained based on strain sample sequence data and stress labels and is used to characterize the mapping relationship between hydration strain and hydration stress.

[0007] According to one aspect of the embodiments of this specification, inputting the first hydration strain data into a denoising model to obtain denoised and restored second hydration strain data includes: denoising the first hydration strain data to obtain initial denoised hydration strain data; and inputting the initial denoised hydration strain data into a filling network to obtain second hydration strain data.

[0008] According to one aspect of the embodiments of this specification, the step of inputting initial hydration strain noise reduction data into a filling network to obtain second hydration strain data includes: performing temperature correction on the initial hydration strain noise reduction data to obtain hydration strain noise reduction fused data; and inputting the hydration strain noise reduction fused data into a filling network for completion to obtain second hydration strain data, wherein the filling network is constructed based on a convolutional neural network, LSTM, and a multi-head self-attention mechanism.

[0009] According to one aspect of the embodiments of this specification, the evolutionary model is constructed in the following manner: a training sample dataset is obtained, wherein each training sample data in the training sample dataset includes: strain sample sequence data of various locations in the core changing over historical time and corresponding stress labels; the strain sample sequence data is input into an initial evolutionary model to obtain an initial prediction result of hydration stress corresponding to the strain sample sequence data; a loss value is calculated based on the initial prediction result of hydration stress and the stress labels; the model parameters of the initial evolutionary model are iteratively trained until the model iteration termination condition is met, thereby constructing the evolutionary model.

[0010] According to one aspect of an embodiment of this specification, the evolution model characterizes the mapping relationship between hydration strain and hydration stress as follows:

[0011] ;

[0012] in: ; This represents the hydration stress in the i-th and j-th directions. Represents the hydration strain in the i-th and j-th directions; E represents volumetric strain; E represents the elastic modulus; ν represents Poisson's ratio. This represents the Kronecker symbol, which takes the value 1 when i=j and the value 0 when i≠j.

[0013] According to one aspect of an embodiment of this specification, the loss function is expressed by the following formula:

[0014] ;in, This represents the hydration stress predicted by the model; This represents the actual hydration stress; N represents the number of training samples.

[0015] This specification provides a core hydration strain monitoring system, comprising: a strain data detection module, which is arranged around the outer surface of the core and is used to collect first hydration strain data of various positions on the core surface over time; a core holder, which holds the core equipped with the strain data detection module; a pressurization module, which is used to pressurize the core to induce hydration expansion; and a control module, which is used to store the strain data, process the first hydration strain data into second hydration strain data, and convert the second hydration strain data into hydration stress data.

[0016] According to one aspect of the embodiments of this specification, the core hydration strain monitoring system further includes a temperature control module; the temperature control module is disposed on the outer surface of the core and inside the core holder, and is used to control the core to be within a preset temperature range.

[0017] This specification also provides a core hydration strain monitoring device based on multi-source fusion data. The device includes: a data acquisition unit for acquiring first hydration strain data generated during the hydration process of the core through a core hydration strain monitoring system; a noise reduction and restoration unit for inputting the first hydration strain data into a noise reduction model to obtain noise-reduced and restored second hydration strain data; and an evolution unit for using an evolution model to convert the second hydration strain data into hydration stress data, thereby obtaining the temporal and spatial distribution and evolution prediction results of hydration stress. The evolution model is trained based on strain sample sequence data and stress labels and is used to characterize the mapping relationship between hydration strain and hydration stress.

[0018] This specification also provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method.

[0019] This specification also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method.

[0020] This application introduces artificial intelligence algorithms and big data analysis technology into the monitoring system to extract features, identify change trends, and predict the hydration expansion state of core samples from strain data acquired by distributed optical fibers, thereby improving the intelligence level and predictive capability of hydration strain monitoring. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments or prior art of this specification, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 The diagram shown is a structural schematic of a core hydration strain monitoring system according to an embodiment of this specification.

[0023] Figure 2 The diagram shown is a flowchart of a core hydration strain monitoring method based on multi-source fusion data, as described in this specification.

[0024] Figure 3 The diagram shown is a flowchart of a method for obtaining second hydration strain data according to an embodiment of this specification.

[0025] Figure 4 The diagram shown is a flowchart of another method for determining second hydration strain data according to an embodiment of this specification.

[0026] Figure 5 The diagram shown is a flowchart of a method for constructing an evolutionary model according to an embodiment of this specification.

[0027] Figure 6 This is a schematic diagram of the structure of a distributed optical fiber hydration strain monitoring experimental device according to an embodiment of this specification;

[0028] Figure 7 The diagram shown is a schematic representation of a core hydration strain monitoring device based on multi-source fusion data, according to an embodiment of this specification.

[0029] Figure 8 The figure shown is a strain response curve of a core hydration process according to an embodiment of this specification.

[0030] Figure 9 The diagram shown is a schematic diagram of the simulation conditions for hydration and expansion of core samples from a wellbore in the laboratory, according to an embodiment of this specification.

[0031] Figure 10 The diagram shown is a structural schematic of a computer device according to an embodiment of this specification.

[0032] Explanation of symbols in the attached drawings:

[0033] 100. Core sample;

[0034] 110. Strain data detection module;

[0035] 120. Core holder;

[0036] 130. Pressure module;

[0037] 1301, Pressure Loading Module;

[0038] 1302, Liquid Loading Module;

[0039] 140. Control module;

[0040] 150. Temperature control module;

[0041] 701. Strain data detection module;

[0042] 702. Core holder;

[0043] 703. Pressurization module;

[0044] 704. Control Module;

[0045] 1002. Computer equipment;

[0046] 1004, Processor;

[0047] 1006. Memory;

[0048] 1008. Drive mechanism;

[0049] 1010. Input / Output Module;

[0050] 1012. Input devices;

[0051] 1014. Output devices;

[0052] 1016. Presentation device;

[0053] 1018. Graphical User Interface;

[0054] 1020. Network interface;

[0055] 1022. Communication link;

[0056] 1024. Communication bus. Detailed Implementation

[0057] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0058] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0059] This specification provides the operational steps of the methods described in the embodiments or flowcharts, but based on conventional or non-inventive labor, more or fewer operational steps may be included. The order of steps listed in the embodiments is merely one possible execution order among many and does not represent the only possible execution order. In actual system or device products, the methods shown in the embodiments or drawings can be executed sequentially or in parallel.

[0060] It should be noted that the core hydration strain monitoring method and device based on multi-source fusion data described in this specification can be used in the field of oil and gas drilling and completion engineering technology, as well as in the field of rock mechanics experimental technology. This specification does not limit the application field of the core hydration strain monitoring method and device based on multi-source fusion data.

[0061] Figure 1 The diagram shown is a schematic diagram of a core hydration strain monitoring system according to an embodiment of this specification, which specifically includes: core 100, strain data detection module 110, core holder 120, pressurization module 130, and control module 140.

[0062] First, a standard core with a smooth surface and no impurities is cut from the mudstone and shale sample. The core size can be 25 mm in diameter and 50 mm in length.

[0063] In this embodiment, the strain data detection module 110 is arranged around the outer surface of the core 100 to collect strain data generated at various locations and points during the hydration process of the core, comprehensively capturing the changes in the core. The strain data detection module can be a structure such as an optical fiber, an optical fiber network, an optical fiber array, or a distributed optical fiber strain sensor.

[0064] Specifically, the bare optical fiber is wound onto the surface of the core 100 using ascending and descending angle spirals. The winding angle and spacing are adjusted according to experimental requirements to ensure close contact between the fiber and the core surface. Adhesive dots can be used to fix the fiber to the core surface to prevent slippage or loosening during the experiment. This application utilizes distributed optical fiber strain sensors to achieve continuous, real-time monitoring of the entire core, overcoming the limitations of traditional point-based monitoring.

[0065] In the embodiments described in this specification, both ends of the optical fiber are further connected to an optical fiber demodulator (not shown in the figure) to check whether the connection between the optical fiber and the demodulator is normal and to perform preliminary signal testing to ensure that the optical fiber sensor can work properly. Liquid nitrogen is used to press on the start and end points of the optical fiber to ensure the accurate positioning of the sensing segment.

[0066] The core 100, with the strain data detection module 110 installed, is placed in the core holder 120, and a sealing system is used to ensure the isolation of the core from the external environment and the airtightness of the core holder to avoid seepage interference. The core is then fixed and a pressurization system is installed.

[0067] The pressurization module 130 includes a pressure loading module 1301 and a liquid loading module 1302. The liquid loading module 1302 injects a hydration fluid at a preset temperature and pressure into the core sample to simulate the hydration process of the wellbore core. With the injection of the hydration fluid, a hydration expansion reaction occurs on the core surface. The strain data detection module 110 records the strain changes of the core during the hydration process in real time, periodically recording strain data at different stages to analyze the spatiotemporal evolution of the hydration expansion. In some embodiments of this specification, the core hydration strain monitoring system also includes a temperature control module 150. The temperature control module 150 is installed on the outer surface of the core 100 and inside the core holder 120 to simulate the formation temperature environment and control the core within a set temperature range.

[0068] The pressure loading module 1301 applies confining pressure to the core 100 via a pressurization system to simulate the downhole formation stress environment. It ensures stable operation of the pressurization system, bringing the core surface stress to the set target and performing cyclic pressure loading. In this embodiment, a hydration fluid at a preset temperature and pressure is further injected into the core sample via a liquid injection system to simulate the hydration process of the wellbore core. It ensures precise adjustment of the temperature control system, keeping the core within the set temperature range to simulate the formation temperature environment.

[0069] This application enables intelligent real-time monitoring and quantitative evaluation of the core hydration and expansion process, providing a reliable experimental method for wellbore stability research.

[0070] In this manual, distributed optical fiber strain monitoring technology is used to monitor the strain evolution during the core hydration and expansion process in real time under the simulated drilling fluid annulus effect. This accurately captures the anisotropic strain characteristics of the core during the hydration process, providing a scientific basis for the study of the core hydration and expansion mechanism and the optimization of drilling technology.

[0071] Figure 2 The diagram shown is a flowchart of a core hydration strain monitoring method based on multi-source fusion data according to an embodiment of this specification, specifically including steps 201 to 203:

[0072] Step 201: Collect the first hydration strain data generated in the core during the hydration process using the core hydration strain monitoring system.

[0073] In this step, initial hydration strain data is collected using the strain data detection module in the core hydration strain monitoring system to obtain the first hydration strain data. During the core hydration expansion process, a distributed optical fiber strain monitoring system is used to monitor and record the strain and temperature changes on the core surface in real time.

[0074] This specification describes a dynamic distributed optical fiber sensing system (OSI-D) based on optical frequency domain reflectance (OFDR) technology to monitor temperature in real time during the hydration strain monitoring process. OFDR is a coherent detection technology based on continuous frequency modulated light waves. It locates the scattered signal by measuring the frequency of the Rayleigh scattering signal generated by the modulated probe light, and has millimeter-level spatial resolution and extremely high sensing accuracy.

[0075] The specific measurement process is as follows: The linearly swept light emitted by the tunable laser is split into two paths by the coupler. One path enters the reference fiber and is reflected back to the coupler by the tail mirror as the local oscillator reference light. The other path enters the measurement fiber to sense the temperature / strain changes acting on the fiber. After measurement, Rayleigh scattering light is continuously generated within the fiber and returns along the original path. The two backscattered lights undergo beat frequency interference in the coupler and then enter the photodetector. The photodetector converts the optical signal into an electrical signal, thereby obtaining the Rayleigh scattering distribution information along the entire fiber. Demodulating the Rayleigh scattering spectrum signal yields the strain and temperature data, which constitute the first hydration strain data in this step.

[0076] In this manual, the dynamic distributed optical fiber sensing system has an OSI-D measurement point spatial resolution of up to 0.64 mm, a temperature measurement accuracy of ±0.1℃, a strain measurement accuracy of ±1με, and a sampling rate of 120Hz. Through optimized algorithms, it can achieve real-time dynamic demodulation of OFDR signals.

[0077] Step 202: Input the first hydration strain data into the denoising model to obtain the denoised and restored second hydration strain data.

[0078] In this step, for the temperature and strain data acquired by the optical fiber, multimodal signal processing methods such as wavelet transform and ensemble empirical mode decomposition are used for denoising and feature enhancement. This improves the signal quality of the distributed optical fiber optic hydration strain sensing system in complex environments and solves the problems of noise interference and data loss in the acquired first hydration strain data. For detailed processing procedures, please refer to [link to relevant documentation]. Figure 3 describe.

[0079] Step 203: Using the evolution model, the second hydration strain data is converted into hydration stress data to obtain the distribution and evolution prediction results of hydration stress in time and space.

[0080] The evolution model is trained based on strain sample sequence data and stress labels, and is used to characterize the mapping relationship between hydration strain and hydration stress.

[0081] In this step, an evolutionary model is used to analyze the second hydration strain data to identify the characteristic patterns of core hydration expansion evolution. For details on the training and construction process of the evolutionary model, please refer to [link to relevant documentation]. Figure 4 describe.

[0082] This manual can accurately capture the anisotropic strain characteristics of core samples during hydration, precisely identify the evolution characteristics of core hydration expansion, and predict wellbore stability, thereby improving experimental analysis efficiency and result reliability. It provides a scientific basis for the study of wellbore core hydration expansion mechanisms and the optimization of drilling techniques. It has significant application value in wellbore stability monitoring and control.

[0083] Figure 3 The diagram shown is a flowchart of a method for obtaining second hydration strain data according to an embodiment of this specification, specifically including steps 301 to 302:

[0084] Step 301: Denoise the first hydration strain data to obtain initial denoised hydration strain data.

[0085] In this step, the first hydration strain data is recorded as... Where x ∈ [0, L], x represents the spatial location of the optical fiber, and t represents the sampling time. The first hydration strain data is the initial data after acquisition but without processing, which contains noisy data, and some data is incomplete or missing. Therefore, in order to suppress multi-source noise in the first hydration strain data, a hierarchical processing strategy is adopted for the first hydration strain data, and wavelet transform, ensemble empirical mode decomposition (EEMD) and singular spectrum analysis (SSA) are performed separately.

[0086] Specifically, the first step is to perform frequency domain decomposition on the first hydration strain data using discrete wavelet transform to obtain the detailed coefficients of each frequency band. Frequency domain decomposition refines the frequency range step by step from high to low through coefficient layering (i.e., multi-resolution decomposition), forming a layered structure. Layering is essentially a step-by-step splitting of different frequency components, making the decomposition results more closely match the frequency distribution characteristics of the actual data.

[0087] Based on the soft thresholding strategy, the following transformation is performed on each coefficient layer:

[0088] ;in, Indicates the threshold value of the stratification coefficient; , This represents the standard deviation of the noise estimate. Indicates the adjustment parameter. This represents the detail coefficients for each frequency band.

[0089] To further decontaminate low-frequency trends and nonlinear noise, this step employs Ensemble Empirical Mode Decomposition (EEMD) to decompose the signal data: Gaussian white noise is introduced to adaptively decompose the non-stationary, nonlinear first hydration strain data into multiple intrinsic mode functions (IMFs) with distinct physical meanings, while preserving the original signal's time-domain characteristics and local frequency information. The specific processing procedure is represented by the following formula:

[0090] ;in, Let represent the i-th eigenfunction obtained after EEMD decomposition, x represent the spatial location of the fiber, t represent the sampling time, and N represent the total number of eigenmode functions obtained after EEMD decomposition. This represents the first hydration strain data.

[0091] The principal modal components are selected for reconstruction in order to preserve physically significant information.

[0092] Furthermore, in signal processing and data analysis, constructing a trajectory matrix and performing singular value decomposition (SVD) is a common method for extracting signal features or performing noise reduction. Combining it with empirical mode decomposition can effectively handle nonlinear and non-stationary signals.

[0093] Finally, the trajectory matrix X ∈ R is constructed for the signal. m×(N-m+1) Perform singular value decomposition:

[0094] ;in, denoted by ; u represents the left singular vector; v represents the right singular vector; by retaining the first r largest singular values ​​and their corresponding singular vectors, the reconstructed matrix can achieve signal denoising (removing noise components corresponding to small singular values) or feature extraction to remove periodic oscillation components.

[0095] After the above fusion processing, the first hydration strain data is initially denoised, as shown below:

[0096] ;in, Let represent the first hydration strain data, where x ∈ [0, L], x represents the spatial position of the optical fiber, and t represents the sampling time. This indicates the initial noise reduction signal after the first hydration strain signal has undergone fusion processing.

[0097] To further improve signal smoothness and structural fidelity, the initial hydration strain denoising signal is processed into a sliding window form before being input into the residual autoencoder network. In this step, the residual autoencoder network is a time-sliding window-based residual autoencoder network, including: an encoder, a bottleneck layer, a decoder, and residual connections. Specifically, the encoder is used for multi-layer dilated convolution to extract temporal features; the bottleneck layer is the attention mechanism module that compresses and enhances important channels; the decoder is used for deconvolution to reconstruct the signal; and the residual connections are used to enhance stability and generalization ability.

[0098] ;in, This represents the input sliding window signal, i.e., Represents the set of network parameters. The denoising result represents the initial denoised signal of the hydration strain. The input sliding window signal is the hydration strain data processed by wavelet transform, ensemble empirical mode decomposition, and singular spectrum analysis.

[0099] Step 302: Input the initial noise reduction signal of hydration strain into the filling network to obtain the second hydration strain data.

[0100] For details regarding the specific structure and data processing of the filling network in this step, please refer to [link / reference]. Figure 4 The description of this step will not be repeated here.

[0101] Figure 4 The diagram shown is a flowchart of another method for determining second hydration strain data according to an embodiment of this specification, specifically including steps 401 to 402:

[0102] Step 401: Temperature correction is applied to the initial hydration strain noise reduction data to obtain hydration strain noise reduction fused data.

[0103] In this step, considering the coupling effect of temperature and strain, temperature drift correction needs to be applied to the initial noise reduction data of hydration strain. Therefore, the calibrated thermo-stress coupling coefficient is introduced. The drift correction model is as follows:

[0104] ;in, This represents the temperature drift corrected hydration strain noise reduction fusion data. Indicates the current temperature. Indicates the reference temperature. This represents the thermal-stress coupling coefficient.

[0105] Step 402: Input the hydration strain noise reduction fusion data into the filling network for completion to obtain the second hydration strain data. The filling network is constructed based on convolutional neural network, LSTM and multi-head self-attention mechanism.

[0106] In this specification, the hydration-strain purification fusion data contains gaps. Therefore, a filling network is employed to recover missing, damaged, or incomplete areas in the data. The recovery of missing segments relies on joint modeling of spatial proximity channels and temporal context information. Specifically, the filling network learns the inherent patterns of spatial texture, temporal trends, and semantic relationships in the hydration-strain purification fusion data, generating content logically consistent with the original data to fill in the gaps. In this application, a missing mask is used to locate the area to be filled, guiding the model to focus on the target area. Utilizing the temporal and spatial information of the hydration-strain purification fusion data improves the accuracy and reliability of missing portion recovery.

[0107] In some embodiments of this specification, the missing regions are accurately located by using a mask, and the spatiotemporal correlation of the missing segments is focused to generate a completion result that more closely matches the actual distribution.

[0108] Define a mask for identifying missing signal segments. A file is considered missing if it meets one of the following conditions:

[0109] ;in, Indicates the noise amplitude threshold. Indicates the maximum allowed time interval; Indicates a noise identification mask. This represents the original hydration strain signal.

[0110] After determining the missing segment identification mask,

[0111] (1) The input contains incomplete hydration strain purification fusion data and missing mask. Only the missing positions of the mask are filled to avoid interfering with the complete data.

[0112] (2) A padding network is constructed based on the fusion of convolutional neural network (CNN), bidirectional LSTM and multi-head self-attention mechanism.

[0113] Specifically, the filling network is defined by the following formula:

[0114] ;

[0115] in, This represents the hydration strain signal at spatial location x and sampling time t after the filling network is used for completion; that is, the complete hydration strain data after the missing region is accurately filled in. This represents the process by which the network performs feature fusion and computational mapping on data. Represents network parameters, and These represent the widths of the spatial and temporal windows, respectively. Represents spatial gradient auxiliary features. The denoised multivariate fusion data representing hydration strain is, in other words, hydration strain denoised fusion data. This represents physical quantities such as temperature that are related to hydration strain. The spatial location of the optical fiber is represented by t, and the sampling time is represented by t.

[0116] (3) Generate fill content by learning the context information around the missing area through the fill network.

[0117] The incomplete hydration strain purification fusion data and missing mask are input into the above-mentioned filling network to obtain the output completion result. Thus, missing data are filled in for the hydration strain noise reduction fusion data.

[0118] In this step, to ensure the accuracy, smoothness, physical consistency, and spectral preservation of the network output, an end-to-end multi-objective composite loss function is constructed using the following formula:

[0119] ;

[0120] in:

[0121] ;

[0122] ;

[0123] ;

[0124] ;

[0125] ;

[0126] in: Represents the time-domain sub-loss function. This represents the i-th hydration strain value predicted by the model. Represents the true value of the hydration strain of the i-th element, and represents... Represents the frequency domain sub-loss function; This represents the sub-loss function of the total variation regularization term; This represents the loss function for the missing segment compensator. The physical consistency constraint sub-loss function is represented; α represents the coefficient of thermal expansion. Indicates reference temperature. This indicates the current actual temperature. This represents the strain response in theoretical mechanics.

[0127] Ultimately, by jointly optimizing the aforementioned loss terms, we can preserve the spatial structural features and physical consistency while ensuring the accuracy of signal reconstruction, thereby achieving stable monitoring and feature extraction of the entire hydration process.

[0128] Figure 5 The diagram shown is a flowchart of a method for constructing an evolutionary model according to an embodiment of this specification, specifically including steps 501 to 503:

[0129] Step 501: Obtain the training sample dataset. Each training sample in the training sample dataset includes: strain sample sequence data of different locations of the rock core changing over historical time and the corresponding stress label.

[0130] In this specification, the core sample undergoes expansion strain during hydration, leading to changes in the internal stress field. Following the method described above, distributed fiber optic sensing technology is used to acquire strain data of different locations within the core sample over historical time. This strain data is used as training sample data, denoted as: .

[0131] Furthermore, using time windows or spatial segment The strain sample sequence data is generated in units of axial strain and shear strain. The strain sample sequence data includes axial strain and shear strain components. In some embodiments of this specification, environmental parameters such as temperature and moisture content may be included as inputs to the strain sample sequence data, if necessary.

[0132] Step 502: Input the strain sample sequence data into the initial evolution model to obtain the initial prediction result of hydration stress corresponding to the strain sample sequence data.

[0133] Using strain sample sequence data as input parameters for the initial evolution model, and combining linear elastic constitutive relations, a function mapping model between hydration strain and hydration stress is established.

[0134] In some embodiments of this specification, the initial prediction result is: the change in hydration stress at the corresponding location of the core over time. .

[0135] Step 503: Calculate the loss value based on the initial prediction results, iteratively train the model parameters of the initial evolutionary model until the model iteration termination condition is met, and construct the evolutionary model.

[0136] This step uses supervised learning to train the initial evolution model. Based on the prediction results after each iteration of the initial evolution model, the loss value between the prediction results and the label (true hydration stress value) is calculated, thereby optimizing the model parameter θ to minimize the difference between the prediction results and the true hydration stress value.

[0137] The loss function formula is shown below:

[0138] ;in; This represents the hydration stress predicted by the model; The label represents the actual hydration stress; N represents the number of training samples.

[0139] The initial evolutionary model is trained and optimized until the loss function converges to a preset threshold, or the number of model iterations reaches a preset threshold. This completes the training of the initial evolutionary model, thus constructing the final evolutionary model. This allows for the prediction of the temporal and spatial distribution and evolution of hydration stress using fiber optic measured strain.

[0140] In the embodiments of this specification, the constructed evolution model characterizes the relationship between hydration strain and hydration stress as follows: Figure 8 As shown: Figure 8 The figure shown is a strain response curve of the core hydration process according to an embodiment of this specification.

[0141] The figure shows the fiber optic strain response curves of the core at different moments during the hydration expansion process, demonstrating the dynamic changes in the core surface strain over time during the hydration reaction. The figure uses a dynamic distributed optical fiber sensing system (OSI-D) based on optical frequency domain reflectance (OFDR) to monitor strain data in real time during the hydration strain monitoring process. It can be seen that at different locations in the core (sensing distance), the strain magnitude gradually increases with increasing hydration time. The degree of strain change also varies at different locations in the core.

[0142] The relationship between hydration strain and hydraulic stress can be represented by the following mapping: ;

[0143] in: This represents the hydration stress in the i-th and j-th directions; ; This represents the hydration strain in the x-direction. This represents the hydration strain in the y-direction; This represents the hydration strain in the z-direction; Represents the hydration strain in the i-th and j-th directions; E represents volumetric strain; E represents the elastic modulus; ν represents Poisson's ratio. This represents the Kronecker symbol, which takes the value 1 when i=j and the value 0 when i≠j.

[0144] Figure 6 This is a schematic diagram of the structure of a distributed optical fiber hydration strain monitoring experimental device according to an embodiment of this specification.

[0145] The experimental setup shown in the figure includes: core sample, pressurization system, liquid injection system, temperature control system, fiber optic demodulator, and computer.

[0146] Standard core samples were fixed on the experimental platform to ensure their position and orientation stability. The wellbore core physical simulation system includes an inner casing, a core simulator, and an outer casing. The casing uses API standard casing; the inner casing has an outer diameter of 139.7 mm, a wall thickness of 7.72 mm, and a steel grade of P110; the outer casing has an outer diameter of 177.8 mm, a wall thickness of 8.05 mm, and a steel grade of N80. Both the inner and outer casings are 1000 mm long.

[0147] A high-precision dynamic distributed optical fiber temperature / strain monitoring system was installed on the outer surface of the core sample, and the optical fiber was wound and connected.

[0148] The pressurization system includes a pressurization pump and a pressurization steel pipeline. The pressurization pump injects fluid into the core and applies confining pressure to ensure that the confining pressure reaches the set target pressure, and performs pressure cyclic loading. Another pressurization pump injects one end of the fluid into the core holder through a pipeline connected to the wellbore core physical simulation clamping system. Pressure data is measured by pressure sensors connected to the pipeline and transmitted in real time to a computer for data analysis.

[0149] The liquid injection system is used to inject simulated hydration fluid into the core, ensuring precise control of the injection rate and pressure. The temperature control system (not shown in the figure) is integrated into the experimental setup to simulate the formation temperature environment, ensuring that the temperature during the core hydration and expansion process matches actual environmental conditions.

[0150] The two ends of the optical fiber are further connected to an optical fiber demodulator to check the connection between the fiber and the demodulator and to perform preliminary signal testing to ensure that the optical fiber sensor is working properly. Liquid nitrogen is used to press down on the start and end points of the optical fiber to ensure the accurate positioning of the sensing segment.

[0151] In addition, the experimental setup also includes a fixed base (not shown in the figure), which supports the entire experimental setup and ensures the stability of each component, forming a complete indoor well core hydration simulation experimental setup.

[0152] Figure 7 The diagram shown is a schematic representation of a core hydration strain monitoring device based on multi-source fusion data, according to an embodiment of this specification. The basic structure of the device is illustrated in the diagram. The functional units and modules can be implemented using software, or using general-purpose or specific chips to monitor core hydration strain based on multi-source fusion data. The device specifically includes:

[0153] A strain data detection module 701 is arranged around the outer surface of the core to collect first hydration strain data of each position on the core surface as a function of time.

[0154] Core holder 702, wherein a core having the strain data detection module is placed in the core holder;

[0155] Pressurization module 703, the pressurization module is used to pressurize the core for hydration expansion;

[0156] The control module 704 is used to store the strain data, process the first hydration strain data into second hydration strain data, and convert the second hydration strain data into hydration stress data.

[0157] Figure 9The diagram shown illustrates a laboratory simulation of hydration expansion of wellbore core samples according to an embodiment of this specification. The experimental apparatus in this embodiment simulates the hydration behavior of the core sample under stress conditions, including hydraulic column pressure and surrounding rock pressure, within the drilling formation. Traditional hydration experimental apparatuses do not consider the influence of stress and rely on spontaneous absorption under no confining pressure or stress conditions, thus failing to simulate the hydration evolution process of wellbore core samples under stress conditions.

[0158] like Figure 10 The diagram shown is a schematic of a computer device provided in an embodiment of this specification. The core hydration strain monitoring method based on multi-source fusion data described in this application can be applied to the computer device. The computer device 1002 may include one or more processors 1004, such as one or more central processing units (CPUs), each of which can implement one or more hardware threads. The computer device 1002 may also include any memory 1006 for storing any kind of information such as code, settings, data, etc. Without limitation, for example, the memory 1006 may include any one or more combinations of the following: any type of RAM, any type of ROM, flash memory device, hard disk, optical disk, etc. More generally, any memory can use any technology to store information. Further, any memory can provide volatile or non-volatile retention of information. Further, any memory can represent a fixed or removable component of the computer device 1002. In one case, when the processor 1004 executes associated instructions stored in any memory or combination of memories, the computer device 1002 can perform any operation of the associated instructions. The computer device 1002 also includes one or more drive mechanisms 1008 for interacting with any memory, such as a hard disk drive mechanism, an optical disk drive mechanism, etc.

[0159] Computer device 1002 may further include an input / output module 1010 (I / O) for receiving various inputs (via input device 1012) and providing various outputs (via output device 1014). A specific output mechanism may include a presentation device 1016 and an associated graphical user interface (GUI) 1018. In other embodiments, the input / output module 1010 (I / O), input device 1012, and output device 1014 may be omitted, and the device may function solely as a computer device within a network. Computer device 1002 may also include one or more network interfaces 1020 for exchanging data with other devices via one or more communication links 1022. One or more communication buses 1024 couple the components described above together.

[0160] The communication link 1022 can be implemented in any way, such as via a local area network, a wide area network (e.g., the Internet), a point-to-point connection, or any combination thereof. The communication link 1022 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc., governed by any protocol or combination of protocols.

[0161] Corresponding to Figures 1 to 5 In addition to the methods described above, embodiments of this specification also provide a computer-readable storage medium storing a computer program that, when executed by a processor, performs the steps of the methods described above.

[0162] This specification also provides computer-readable instructions, wherein when a processor executes the instructions, the program therein causes the processor to perform the following... Figures 1 to 5 The method shown.

[0163] It should be understood that in the various embodiments of this specification, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this specification.

[0164] It should also be understood that, in the embodiments of this specification, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this specification generally indicates that the preceding and following related objects have an "or" relationship.

[0165] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this specification can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this specification.

[0166] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0167] In the several embodiments provided in this specification, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the couplings or direct couplings or communication connections shown or discussed may be indirect couplings or communication connections through some interfaces, devices, or units, or they may be electrical, mechanical, or other forms of connection.

[0168] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments described in this specification, depending on actual needs.

[0169] Furthermore, the functional units in the various embodiments of this specification can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0170] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this specification, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this specification. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0171] This specification uses specific embodiments to illustrate the principles and implementation methods of this specification. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this specification. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this specification. Therefore, the content of this specification should not be construed as a limitation of this specification.

Claims

1. A method for monitoring core hydration strain based on multi-source fusion data, characterized in that, The method includes: The strain data detection module, which is deployed around the outer surface of the core in the core hydration strain monitoring system, collects the initial hydration strain data and obtains the first hydration strain data. The first hydration strain data is the strain data generated at each position and point on the outer surface of the core. The first hydration strain data is input into the noise reduction model to obtain the noise-reduced and restored second hydration strain data; Using an evolutionary model, the second hydration strain data is converted into hydration stress data to obtain the temporal and spatial distribution and evolution prediction results of hydration stress. The evolutionary model is trained based on strain sample sequence data and stress labels and is used to characterize the mapping relationship between hydration strain and hydration stress.

2. The method according to claim 1, characterized in that, The first hydration strain data is input into the denoising model to obtain the denoised and restored second hydration strain data, including: The first hydration strain data is denoised to obtain the initial denoised hydration strain data; The initial hydration strain noise reduction data is input into the filling network to obtain the second hydration strain data.

3. The method according to claim 2, characterized in that, The step of inputting the initial hydration strain noise reduction data into the filling network to obtain the second hydration strain data includes: Temperature correction is applied to the initial hydration strain noise reduction data to obtain hydration strain noise reduction fused data; The hydration strain noise reduction fusion data is input into the filling network for completion to obtain the second hydration strain data. The filling network is constructed based on convolutional neural network, LSTM and multi-head self-attention mechanism.

4. The method according to claim 1, characterized in that, The evolutionary model is constructed in the following manner: Obtain a training sample dataset, wherein each training sample data in the training sample dataset includes: strain sample sequence data of each location of the rock core changing with historical time and the corresponding stress label; The strain sample sequence data is input into the initial evolution model to obtain the initial prediction result of the hydration stress corresponding to the strain sample sequence data; Based on the initial prediction results of the hydration stress and the loss value calculated by the stress label, the model parameters of the initial evolution model are iteratively trained until the model iteration termination condition is met, and the evolution model is constructed.

5. The method according to claim 4, characterized in that, The evolution model characterizes the mapping relationship between hydration strain and hydration stress as shown below: ； in: ; This represents the hydration stress in the i-th and j-th directions. Represents the hydration strain in the i-th and j-th directions; E represents volumetric strain; E represents the elastic modulus; ν represents Poisson's ratio. This represents the Kronecker symbol, which takes the value 1 when i=j and the value 0 when i≠j.

6. The method according to claim 4, characterized in that, The loss value is determined by the following formula: ;in, This represents the hydration stress predicted by the model; This represents the actual hydration stress; N represents the number of training samples.

7. A core hydration strain monitoring system, characterized in that, The system includes: The strain data detection module is arranged around the outer surface of the core to collect the first hydration strain data of the core surface at various locations over time. The first hydration strain data is the strain data generated at various locations and points on the outer surface of the core. A core holder, wherein a core equipped with the strain data detection module is placed in the core holder; A pressurization module, used to pressurize the core for hydration expansion; The control module is used to store the strain data, process the first hydration strain data into second hydration strain data, and convert the second hydration strain data into hydration stress data.

8. The system according to claim 7, characterized in that, The core hydration strain monitoring system also includes a temperature control module; The temperature control module is installed on the outer surface of the core and inside the core holder to control the core within a preset temperature range.

9. A core hydration strain monitoring device based on multi-source fusion data, characterized in that, The device includes: The acquisition unit is used to acquire initial hydration strain data through the strain data detection module deployed around the outer surface of the core in the core hydration strain monitoring system, and to obtain the first hydration strain data, which is the strain data generated at various locations and points on the outer surface of the core. The noise reduction and restoration unit is used to input the first hydration strain data into the noise reduction model to obtain the noise-reduced and restored second hydration strain data; An evolutionary unit is used to convert the second hydration strain data into hydration stress data using an evolutionary model, thereby obtaining the temporal and spatial distribution and evolution prediction results of hydration stress. The evolutionary model is trained based on strain sample sequence data and stress labels and is used to characterize the mapping relationship between hydration strain and hydration stress.

10. A computer device, characterized in that, The computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method according to any one of claims 1 to 6.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 6.

Images