Sandy gravel stratum catastrophe grouting prevention and control virtual-real blending collaborative test method and system

Abstract

The invention relates to the technical field of grouting simulation tests, in particular to a sandy gravel stratum catastrophe grouting prevention and control virtual-real blending collaborative test method and system, and the method comprises the steps: carrying out a grouting test on a sandy gravel stratum simulation model, and obtaining grouting parameters and monitoring parameters in the grouting process; based on the resistivity and the radar data, further obtaining a slurry diffusion path and a diffusion range of the area with the higher surrounding rock grade; based on the resistivity and the optical fiber sensing data, the slurry diffusion path and the diffusion range of the area with the low surrounding rock grade are further obtained; inputting the initial grouting parameters into the trained neural network model to obtain simulation data; carrying out preprocessing and alignment operation on the previous data, and carrying out fusion analysis; and according to the fusion analysis result, grouting equipment parameters and a grouting scheme are pre-controlled and adjusted. And real-time regulation and control of a grouting scheme and equipment are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of grouting simulation test, and in particular to a virtual-real integrated collaborative test method and system for grouting prevention and control of disasters in sand and gravel strata. Background Art

[0002] Grouting is one of the important technologies for controlling water disasters and reinforcing loose and weak strata in underground engineering fields such as mines, tunnels, and water conservancy. When the grouting medium is a porous medium such as sand or gravel, since the particles of these media are rigid bodies, it is difficult for the slurry to compress the particles during the diffusion process, so it mainly diffuses through infiltration, accompanied by a certain particle pore compaction effect. The diffusion, migration and solidification process of the slurry in the sand and gravel medium is the key to determining the success or failure of the grouting disaster control engineering design and construction. However, due to the concealment and complexity of the sand and gravel medium strata, the grouting process is difficult to accurately control, and the slurry diffusion range is difficult to quantify and evaluate, resulting in the formulation of grouting treatment plans and the selection of construction processes relying more on subjective experience. The existing grouting scheme optimization research is mainly applicable to single or specific conditions, and it is difficult to meet the grouting design needs of highly complex and variable water-rich porous medium strata. Summary of the invention

[0003] In order to solve the deficiencies of the prior art, the present invention provides a virtual-real fusion collaborative test method and system for sand and gravel stratum disaster grouting prevention and control;

[0004] On the one hand, a virtual-real synergistic test method for disaster grouting prevention and control in sandy and gravel strata is provided, including:

[0005] (1) conducting a grouting test on a sand and gravel formation simulation model to obtain grouting parameters and monitoring parameters during the grouting process, wherein the grouting parameters include: grouting pressure, grouting flow rate and slurry ratio; and the monitoring parameters include: resistivity data, radar data and optical fiber sensor data;

[0006] (2) Based on the resistivity data and radar data, the slurry diffusion path and diffusion range in the area with higher surrounding rock grade are obtained; based on the resistivity data and optical fiber sensing data, the slurry diffusion path and diffusion range in the area with lower surrounding rock grade are obtained;

[0007] (3) Inputting the initial grouting parameters into the trained neural network model to obtain simulation data;

[0008] (4): Preprocess and align the data of (1) to (2) and (3) respectively, and perform fusion analysis;

[0009] (5) Based on the fusion analysis results, pre-control adjustments are made to the grouting equipment parameters and grouting scheme.

[0010] On the other hand, a virtual-real collaborative test system for disaster grouting prevention and control in sand and gravel formations is provided, including:

[0011] An acquisition module is configured to: conduct a grouting test on a sand-pebble formation simulation model to obtain grouting parameters and monitoring parameters during the grouting process, wherein the grouting parameters include: grouting pressure, grouting flow rate and slurry ratio, and the monitoring parameters include: resistivity data, radar data and optical fiber sensor data;

[0012] An inversion module is configured to: obtain the slurry diffusion path and diffusion range in the area with higher surrounding rock grade based on the resistivity data and the radar data; obtain the slurry diffusion path and diffusion range in the area with lower surrounding rock grade based on the resistivity data and the optical fiber sensing data;

[0013] The simulation module is configured to: input the initial grouting parameters into the trained neural network model to obtain simulation data;

[0014] A fusion analysis module is configured to: pre-process and align the data of the acquisition module, the inversion module and the simulation module respectively, and perform fusion analysis;

[0015] The pre-control adjustment module is configured to: perform pre-control adjustments on the grouting equipment parameters and the grouting scheme according to the fusion analysis results.

[0016] The above technical solution has the following advantages or beneficial effects:

[0017] Taking the disaster treatment process of gravel formation as the object, a test system for water-gushing reinforcement and disaster control in gravel formation was built. Based on electrical resistance tomography, fiber optic sensing and radar to capture the slurry diffusion process during the grouting process, the data of gravel grouting pressure, flow, velocity and slurry selection were integrated to realize the exploration of the grouting treatment mechanism and data fusion analysis of multiple types of disasters in gravel formations. A grouting intelligent simulation method combining numerical solution and neural network was proposed to realize the virtual model solution and analysis of equal-dimensional 1:1 model. According to the fusion analysis results of real-time monitoring data and the numerical solution analysis results, the actual grouting equipment parameters and grouting scheme were adjusted in real time to realize the visualization of the whole grouting treatment process and the real-time regulation of grouting scheme and equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0019] Figure 1 This is a flow chart of the method of embodiment 1. DETAILED DESCRIPTION

[0020] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0021] Embodiment 1

[0022] like Figure 1 As shown, this embodiment provides a virtual-real synergistic test method for disaster grouting prevention and control in sandy and gravel formations, including:

[0023] S101: conducting a grouting test on a sand and gravel formation simulation model to obtain grouting parameters and monitoring parameters during the grouting process, wherein the grouting parameters include: grouting pressure, grouting flow rate and slurry ratio, and the monitoring parameters include: resistivity data, radar data and optical fiber sensor data;

[0024] S102: based on the resistivity data and the radar data, obtaining the slurry diffusion path and diffusion range in the area with a higher surrounding rock grade; based on the resistivity data and the optical fiber sensing data, obtaining the slurry diffusion path and diffusion range in the area with a lower surrounding rock grade;

[0025] S103: inputting the initial grouting parameters into the trained neural network model to obtain simulation data;

[0026] S104: preprocessing and aligning the data in steps S101 to S102 and S103 respectively, and performing fusion analysis;

[0027] S105: Pre-control and adjust the grouting equipment parameters and grouting scheme according to the fusion analysis results.

[0028] The surrounding rock grades are divided into the first, second, third, fourth, fifth and sixth grades. Among them, the fourth, fifth and sixth grades are areas with higher surrounding rock grades; among them, the first, second and third grades are areas with lower surrounding rock grades. Surrounding rock classification refers to dividing an infinite rock mass sequence into a finite number of categories with different degrees of stability based on indicators such as rock mass integrity and rock strength, that is, classifying some surrounding rocks with similar stability into one category, and dividing all surrounding rocks into several categories. On the basis of surrounding rock classification, the best construction method and support structure design are given according to the stability of each type of surrounding rock. Surrounding rock classification is the basis for selecting construction methods, conducting scientific management and correctly evaluating economic benefits, determining the load on the structure (loose load), determining the type and size of the lining structure, and formulating labor quotas and material consumption standards.

[0029] Furthermore, the sand and gravel stratum simulation model is a model obtained by 3D printing.

[0030] Furthermore, electrodes are arranged on the outer surface of the sand and gravel formation simulation model, a radar is arranged within a set distance range outside the sand and gravel formation simulation model, and optical fibers are arranged in the drilling holes and grouting holes of the sand and gravel formation simulation model.

[0031] Furthermore, the sand and gravel stratum simulation model includes: a model shell; the model shell is made of acrylic or glass material, and the microstructure image of the actual natural stratum is obtained by microscope and CT scanning technology, and it is digitized and three-dimensionally reconstructed to obtain a three-dimensional model of the sand and gravel stratum; settings include interface thickness, interface material, printing speed, etc., and the 3D printer is further prepared to ensure the cleanliness of the print bed, and level it as needed, load suitable printing materials, upload the sliced ​​file to the 3D printer, and complete the 3D printing of the three-dimensional model of the sand and gravel stratum.

[0032] By adjusting the water source, the water characteristics of different water-rich sand and gravel formations and the dynamic water flow rate environment simulation can be realized, including water pumps, flow meters and insulated water tanks.

[0033] The grouting system consists of a three-cylinder special pump, an intelligent control box, a flow sensor, a pressure sensor, etc. The three-cylinder special pump sucks and discharges cement slurry and water glass respectively, and the data detected by the sensor is transmitted to the intelligent control box after the data is calculated and fed back to adjust the displacement of cement slurry and water glass respectively, and keep the grouting speed stable.

[0034] Furthermore, electrodes are arranged on the outer surface of the sand and gravel formation simulation model, wherein the step of arranging the electrodes comprises:

[0035] Through the simulated annealing algorithm, the initialization parameters are first set, and an electrode arrangement scheme is randomly selected as the initial solution; an initial temperature (T0) is set; a cooling rate (α) is set; and a termination temperature (T min ); Maximum number of iterations (N max ), and define the objective function:

[0036]

[0037] Among them, d i is the distance between the ith electrode pair, d targe is the target electrode spacing; Variance(e j ) is the uniformity variance of the electrode arrangement; R k is the kth geological condition, R targe is the expected resistivity; P l is the porosity of the first geological condition, P targe is the expected porosity, w1, w2, w3, w4 are weighting coefficients.

[0038] Iterate the algorithm, select a random initial solution x0, set the initial temperature T0, and start the main loop until the temperature drops to T min Or the number of iterations exceeds N max , by adjusting the electrode position or spacing from the current solution x k Generate a neighborhood solution x new , calculate the objective function value f(x new ) and the objective function value f(x k ), calculate the energy difference:

[0039] ΔE=f(x new )-f(x k )

[0040] Determine whether to accept the new solution:

[0041] If ΔE≤0 (i.e. the new solution is better than the current solution), then accept the new solution:

[0042] x k+1 =x new

[0043] If ΔE>0 (i.e. the new solution is worse), keep the current solution:

[0044] x k+1 =x k

[0045] After each iteration, the temperature decays at the cooling rate:

[0046] T k+1 =αT k

[0047] When T k ≤T mib Or the maximum number of iterations N is reached max , the algorithm is terminated and the optimal solution is output to find the optimal electrode arrangement solution.

[0048] Based on the optimal electrode arrangement plan, the electrode position is determined by laser ranging technology, and the electrode is installed at the predetermined position. During the installation process, the contact condition between the electrode and the formation is monitored in real time by contact quality detectors and pressure sensors to ensure stable contact.

[0049] Furthermore, a radar is arranged within a set distance range outside the sand and gravel stratum simulation model, wherein the radar arrangement step comprises:

[0050] For the highly complex areas in the deployed strata, such as fracture zones, broken zones, and shallow areas of heterogeneous rock formations, radar sensors are deployed in the peripheral areas, and the simulated annealing algorithm is integrated to determine the optimal deployment location and number of radar sensors to ensure coverage of all areas of concern. The specific steps are as follows:

[0051] Assume that the radar is located at x=(x1,x2,...,x n ), where x i represents the position of the i-th radar, and the objective function is designed as:

[0052] f(x)=w1·Variance(x)+w2·Coverage(x)+w3·Signal Strength(x)

[0053] Among them, Variance(x) represents the variance of the radar position, Coverage(x) represents the coverage of the radar on the geological layer or target area, and Signal Strength(x) represents the radar signal strength.

[0054] Set the initialization parameters:

[0055] Randomly generate an initial radar layout coordinate x0, and bring it into the objective function to calculate the objective function value f(x0) of the initial radar layout; set a higher initial temperature (T0); set the cooling rate (α); set the termination temperature (T min ); Maximum number of iterations (N max );

[0056] Iterate the algorithm, select a random initial solution x0, set the initial temperature T0, and start the main loop until the temperature drops to T min Or the number of iterations exceeds N max , by adjusting the radar position or spacing from the current solution x k Generate a neighborhood solution x new , calculate the objective function value f(x new ) and the objective function value f(x k ), calculate the energy difference:

[0057] ΔE=f(x new )-f(x k )

[0058] Determine whether to accept the new solution:

[0059] If ΔE≤0 (i.e. the new solution is better than the current solution), then accept the new solution:

[0060] x k+1 =x new

[0061] If ΔE>0 (i.e. the new solution is worse), keep the current solution:

[0062] x k+1 =x k

[0063] After each iteration, the temperature decays at the cooling rate:

[0064] T k+1 =αT k

[0065] When T k ≤T min Or the maximum number of iterations N is reached max , the algorithm is terminated and the optimal solution is output to find the optimal radar layout plan.

[0066] Set the sensor spacing and direction according to the sensor's working range and detection angle; use a total station for positioning.

[0067] The radar sensor emits electromagnetic wave signals, receives signals reflected from the formation, and converts them into digital signals to obtain radar signals. Signal processing is used to filter out noise and enhance signal quality, supporting formation structure analysis of the slurry diffusion path and real-time tracking of slurry.

[0068] Furthermore, optical fibers are arranged in the drill holes and grouting holes of the sand and gravel formation simulation model, including:

[0069] For water-rich and sand-rich strata, deformed strata, deep strata and multi-layer alternating geological structures, after drilling the grouting borehole, optical fiber is arranged on the borehole wall; for the borehole using catheter grouting, optical fiber is arranged on the catheter wall.

[0070] Further, the S101: conducting a grouting test on a sand-pebble formation simulation model to obtain grouting parameters and monitoring parameters during the grouting process, wherein the monitoring parameters include: resistivity data, radar data, and optical fiber sensing data;

[0071] Among them, resistivity data is collected by collecting current and voltage data through electrodes, and then the resistivity is determined by the collected current and voltage data; among them, radar data is collected by radar;

[0072] Among them, fiber optic sensing data is collected through optical fibers.

[0073] The electrodes, radar and optical fiber are all connected to a host computer.

[0074] The diffusion path and range of the slurry are preliminarily located by resistivity, and the diffusion path and diffusion boundary of shallow, non-deformed strata and shallow strata are constrained and corrected in combination with the radar signal of the radar sensor. For weak or deformed strata or deep multi-layer alternating geological structures, the deformation is obtained by using fiber optic monitoring data, and the position of the deformed boundary is projected onto the resistivity imaging, thereby achieving a refined characterization of the slurry diffusion front and the deformed boundary.

[0075] Further, determining the resistivity by collecting the current and voltage data includes:

[0076] Before grouting, resistivity data collection is performed in the initial stage to obtain the initial resistivity distribution.

[0077] First, a known current (I) is generated by a current source in the electrode network, and the resistivity of the formation is calculated by the voltage difference (ΔV).

[0078] During the grouting process, as the slurry diffuses, the resistivity of the formation will change. The injected slurry has a certain conductivity. When the slurry enters the formation, it will change the resistivity distribution of the formation. Usually, the resistivity of the slurry is low, so after the slurry is injected, the resistivity of the area will be reduced, making it easier for current to pass through the area, and the voltage difference will change accordingly.

[0079] Therefore, during the grouting process, the voltage difference (ΔV) will change dynamically as the slurry diffuses, reflecting the change in formation resistivity.

[0080] Then, the Gauss-Newton inversion algorithm is used to infer the initial resistivity distribution of the underground formation.

[0081] During the grouting process, resistivity data is collected in real time to record the resistivity changes during slurry diffusion.

[0082] Furthermore, the method further includes: based on the change of resistivity data, inverting the global diffusion path and diffusion range of the slurry, and the inversion process specifically includes:

[0083] (1) By continuously calling the Gauss-Newton inversion algorithm, the resistivity distribution map at each moment is generated, and then the resistivity change image of the slurry diffusion is obtained.

[0084] (2) Based on the resistivity change image, determine the area where the resistivity decreases by more than a set threshold, and the area where the resistivity decreases by more than the set threshold corresponds to the injection area of ​​the slurry and the range to which the slurry diffuses.

[0085] For example, in a specific area, the resistivity decreases significantly, and the resistivity value of the area is lower than a certain threshold (the threshold is based on previous experiments or previous data acquisition), and it is inferred that the area is the range of slurry diffusion.

[0086] During the slurry injection process, the real-time tracking of the slurry diffusion path can be achieved by collecting time-series data of resistivity images multiple times; each resistivity image inversion result provides a snapshot of the underground resistivity distribution at a specific moment; with the continuous injection of slurry, the resistivity of the diffusion area will change dynamically. By continuously updating the resistivity data, the slurry diffusion path can be tracked, the spatial distribution and change trend of the resistivity change can be analyzed, and the time evolution process of slurry diffusion can be obtained.

[0087] Furthermore, the S102: based on the resistivity data and the radar data, obtaining the slurry diffusion path and diffusion range in the area with a higher surrounding rock grade includes:

[0088] S102-1: For radar signal collection, a radar sensor is used to scan the tunnel face at multiple angles and orientations, and finally the location and structural characteristics of the water-conducting channels and cavities in the surrounding rock in front of the tunnel face based on the radar image in the stratum are obtained.

[0089] S102-2: Data synchronization and calibration to ensure that the radar signal and resistivity data are consistent in time and space. If they are inconsistent, calibration and alignment are performed;

[0090] S102-3: Through the joint tracking of resistivity and radar signals, the slurry diffusion path in the area with higher surrounding rock grade can be monitored.

[0091] Furthermore, the S102-2: data synchronization and calibration to ensure that the radar signal and resistivity data are consistent in time and space, and if they are inconsistent, necessary calibration and alignment are performed, including:

[0092] By preprocessing and standardizing the resistivity data and radar images, we ensure that the two are consistent in data format and resolution; using spatial registration technology, the data points of the resistivity data and radar data are mapped to the same spatial coordinate system for alignment to ensure that they correspond to the same spatial position such as the same depth, and at the same time, correct whether the time of the two data is the same.

[0093] The original format of resistivity data is in matrix form (containing spatial coordinate points and resistivity data of each coordinate point), and the original format of radar signal is in image format, which needs to be converted into matrix format; if the resolution of radar image is lower than that of resistivity data, the interpolation method is used to improve the resolution of image; if the resolution of radar image is higher than that of resistivity data, the pixel averaging method is used to lower the resolution; to further realize the format conversion, the gray value of each pixel represents the reflection intensity of a certain position underground, so the gray value matrix of the image can be directly used as the data matrix of radar image.

[0094] Furthermore, the S102-3: monitoring the slurry diffusion path in the area with higher surrounding rock grade by joint tracking of resistivity and radar signals includes:

[0095] S102-31: Based on the initial resistivity distribution and radar data, identify the boundary of the water channel in the area with higher surrounding rock grade; the area with higher surrounding rock grade includes: fracture zone, fracture zone, and heterogeneous rock layer area;

[0096] S102-32: During the grouting process, determine the area where the resistivity decreases by more than a set threshold, the determined area where the resistivity decreases by more than the set threshold corresponds to the slurry injection area and the range to which the slurry diffuses.

[0097] For example, in a specific area, the resistivity is significantly reduced, and the resistivity value of the area is lower than a certain threshold (the threshold is based on previous experiments or previous data acquisition), and it is inferred that the area is the range of slurry diffusion. During the slurry injection process, the initial diffusion path of the slurry is captured by collecting time series data of resistivity images multiple times;

[0098] The water channels, cracks, and stratum spatial structure information captured by radar images are used as boundary constraints of the water channel. The resistivity inversion images at the same time and spatial position are superimposed. The slurry-water interface in the slurry diffusion front direction is based on the resistivity inversion result; the slurry diffusion boundary of the blocked water channel is based on the water channel boundary obtained by radar signals;

[0099] S102-33: Superimpose the continuous inversion results of radar and resistivity data and construct a complete diffusion path to track the slurry diffusion.

[0100] The inversion results at each moment are integrated using data fusion methods such as color mapping and transparent stacking to generate visual two-dimensional or three-dimensional images, providing a comprehensive display of the formation structure and slurry diffusion path.

[0101] Furthermore, the S102: based on the resistivity data and the optical fiber sensing data, obtaining the slurry diffusion path and diffusion range in the area with lower surrounding rock grade includes:

[0102] S102-1: Data preprocessing: Ensure the time synchronization of resistivity data and fiber optic sensor data, and ensure that the global resistivity and local strain, temperature, and pressure data are aligned on the time axis; filter and eliminate noise from the fiber optic sensor data, and resample the spatial resolution and time resolution of different data to ensure consistency in space and time.

[0103] S102-2: Data fusion: The prediction of the grouting diffusion range is achieved based on feature extraction and fusion, or the prediction of the grouting diffusion range is achieved based on data analysis.

[0104] Furthermore, the feature extraction and fusion-based method realizes the prediction of the grouting diffusion range, including:

[0105] S102-2a1: First, the resistivity data is preprocessed and converted into two-dimensional or three-dimensional grid data. The global spatial diffusion feature vector F is extracted through the convolutional neural network CNN. Resistivity ; The local strain, temperature, pressure and other data monitored by the optical fiber are processed in time series, and the local time series feature vector F is extracted using the long short-term memory network Fiber ;

[0106] S102-2a2: Use the attention mechanism to fuse the global and local feature vectors, and assign different weights to the global feature vector and the local feature vector (resistivity data weight w Resistivity , fiber data weight w Fiber ), and output the fused feature vector F Fusion .

[0107] w Resistivity +w Fiber =1;

[0108] F Fusion =w Resistivity ·F Resistivity +w Fiber ·F Fiber ;

[0109] S102-2a3: Add physical constraints to the fused features. For example, the resistivity diffusion feature must comply with the diffusion equation constraint, and the optical fiber strain, temperature, and pressure signals must satisfy Hooke's law. The constraints are implemented by adding regularization terms to the loss function.

[0110] L physics =λ1·L diffusion +λ2·L strain;

[0111] Among them, λ1 and λ2 are physical constraint weight coefficients, L diffusion With L strain Represents the loss term for the diffusion equation and strain constraint.

[0112] L diffusion It can be expressed as:

[0113]

[0114] in, Represents the resistivity value predicted by the model; N represents the number of sampling points; D represents the diffusion coefficient, which is summarized through a large number of experimental rules:

[0115] L strain It can be expressed as:

[0116]

[0117] in, Represents the strain value output by the model; σ i represents the stress obtained from the optical fiber monitoring data; E represents the Young's modulus of the material; M represents the number of sampling points;

[0118] Ultimately, by minimizing the total loss function, we ensure that the model not only accurately fits the data, but also follows the actual physical laws, thereby improving the accuracy and physical consistency of the prediction.

[0119] L total =L data (F Fusion )+L physics ;

[0120] Among them, L data is the loss of data in the fitting process, L total Represents the total loss.

[0121] S102-2a4: Input the fused feature vector into the fully connected layer to achieve continuous output of the slurry diffusion range.

[0122] Furthermore, the method based on data analysis to predict the grouting diffusion range includes:

[0123] S102-2b1: Based on the pre-processed global resistivity change data, determine the area where the resistivity decreases beyond the set threshold, corresponding to the injection area of ​​the slurry and the range to which the slurry diffuses, and obtain a rough diffusion range;

[0124] S102-2b2: Based on the pre-processed local strain, temperature and pressure data, the abnormal change points of strain, temperature and pressure are extracted, and the rough diffusion range is refined by combining the slurry front position and slurry diffusion boundary determined by the grouting theory.

[0125] Furthermore, the process of determining the slurry front position includes:

[0126] Based on the slurry diffusion range obtained from the resistivity data, for the refined characterization of the slurry diffusion front, the data of abnormal change points of temperature and pressure captured by the optical fiber sensing data are first cleaned and preprocessed, and the high-frequency noise is removed using a low-pass filtering algorithm. The data for a long period of time is divided into multiple time windows or space windows for local extraction and analysis, and the spatial coordinate positions of the data points are spatially calibrated with the resistivity data to ensure that the rate of change of the acquired data is calculated under the same spatial coordinates. When the rate of change exceeds a certain critical value, the point is considered to be an abnormal change point, the abnormal change point is extracted, and the abnormal change point of the optical fiber sensing path is projected onto the resistivity imaging to finally determine the position of the slurry front.

[0127]

[0128] in, is the rate of temperature change in the spatial direction; is the rate of change of temperature in the time direction.

[0129] Furthermore, the slurry diffusion boundary determination process includes:

[0130] For the refined characterization of the slurry diffusion boundary, the strain, temperature and pressure monitoring change data extracted along the grouting path and the extracted abnormal change points are analyzed, the strain gradient of the strain data is used to identify the area with large strain changes, and the data is normalized with the parameters such as the surrounding rock permeability and elastic modulus along the optical fiber arrangement, and each data is given a weight. The weight is obtained based on repeated experiments. For example, a high permeability area may have a greater impact on the slurry diffusion, so the permeability of this area may need to be given a higher weight; in a low permeability area, the strain change may be more sensitive, so a higher weight should be given; for rock formations with higher surrounding rock grades, the elastic modulus has a greater impact, so the weight of the elastic modulus can be appropriately increased. For example, the optical fiber monitoring strain data, surrounding rock permeability and elastic modulus are used to achieve the refinement of the boundary. The weighted calculation formula can be:

[0131] ΔX=w ∈ ·E ∈ +w K ·E K +w E ·E E

[0132] Among them, ΔX represents the deformation, E ∈ represents the normalized strain, E K represents the normalized permeability, E E represents the normalized elastic modulus, w ∈ 、w K 、w E denote the weights of strain, permeability and elastic modulus, respectively.

[0133] The boundary position after deformation is:

[0134] X new =X old +ΔX

[0135] Among them, X old Represented as the position of the original diffusion boundary, X new Indicates the position of the boundary after deformation.

[0136] The deformed boundary position is projected onto the resistivity image to determine the precise slurry diffusion boundary, ultimately achieving accurate capture of the slurry diffusion process.

[0137] The above technical solution optimizes the layout of electrode arrays, radar positions, optical fiber layout positions and densities of formation resistivity monitoring equipment, geological radar and optical fiber according to the formation conditions.

[0138] The resistivity data is acquired in real time by using the electrode network evenly arranged around the sand and gravel formation model, and the global resistivity change data during the grouting process is obtained by using the electrical resistance tomography technology.

[0139] Arrange radar sensors in the outer area of ​​the sand-pebble formation model to collect radar signals during the grouting process and obtain radar images;

[0140] Through the optical fiber sensors arranged inside the formation, for homogeneous sand and gravel formations, the optical fiber is evenly arranged in the formation; for layered formations, the optical fiber is evenly arranged in each formation, and the optical fiber is densely arranged at the formation interface to obtain the optical fiber sensing data in real time, and then obtain the local strain, temperature and pressure data;

[0141] The dynamic resistivity data, radar scanning data, and fiber optic sensing data are fused and analyzed to obtain the real-time slurry diffusion path and the final slurry diffusion range.

[0142] Furthermore, in S103: inputting the initial grouting parameters into the trained neural network model to obtain simulation data, the training process includes:

[0143] Constructing a training set, wherein the training set is grouting parameters of known simulation data;

[0144] Input the training set into the neural network model, train the neural network model, and stop training when the loss function value of the neural network model no longer decreases, or the number of iterations exceeds the set number, to obtain the trained neural network model;

[0145] Among them, the loss function of the neural network model includes a data fitting loss term, a physical residual loss term and a boundary loss function term; the momentum control equation with the damping term added is used as the physical residual loss term.

[0146] Furthermore, the S103: inputting the initial grouting parameters into the trained neural network model to obtain simulation data includes:

[0147] The boundary conditions, grouting velocity, grouting pressure, initial flow field of dynamic water, gravity field, slurry diffusion morphology, temperature field, slurry viscosity, constant volume specific heat, thermal conductivity and density are taken as input data. The input data are numerically simulated in the initial stage through the numerical simulation method to obtain the calculation results of slurry velocity, grouting pressure, phase fraction and viscosity parameters. The parameters of the last time step obtained by the numerical simulation in the initial stage are taken as initial data and input into the trained neural network model to obtain the prediction results of slurry velocity, grouting pressure, phase fraction and viscosity data.

[0148] In this embodiment, the neural network model is constructed based on the nonlinear mode decomposition algorithm (EDMD).

[0149] In this embodiment, a parameter feature database is constructed by randomly sampling a numerical simulation case library, wherein the numerical simulation case library stores field data such as pressure field, velocity field and phase fraction field of grouting at different time and position points.

[0150] Data is extracted according to the set time step to obtain the data values ​​of flow field characteristic data (pressure, velocity, etc.) and spatiotemporal characteristic data (time span, spatial position), and the outliers in the data set are checked and corrected. Based on the EDMD algorithm, the nonlinear modal decomposition of the time-space analytical flow field of the sampled data is realized.

[0151] First, the Gaussian radial basis function (RBF) is selected as the dictionary function of EDMD, and the expression is:

[0152] φ(x)=exp(-||xc|| 2 / (2σ 2 ));

[0153] Where c is the center of the kernel and σ is the standard deviation.

[0154] Enter each data point into the RBF dictionary, for each point x in the datasetj , using all defined RBFs to calculate its eigenvectors:

[0155] ψ j (x j )=[φ1(x j ),φ2(x j ),…,φ k (x j )];

[0156] in, c k is the kth center point.

[0157] Combine the RBF eigenvectors of all points into a feature matrix Ψ(X):

[0158]

[0159] Perform singular value decomposition on Ψ(X):

[0160] Ψ(X)=U∑V * ;

[0161] Among them, U is the left singular vector, ∑ is the singular value, and ∑ is the right singular vector.

[0162] Construct an approximation A of the Koopman operator:

[0163] A=Ψ(X′)V∑ -1 U * ;

[0164] Among them, X′ is the data matrix of the next time step, U * is the conjugate transposed matrix of U, which describes the row space characteristics of the data matrix. The column vectors constitute the orthogonal basis of the data row space. V is the right singular vector matrix, and its column vectors constitute the orthogonal basis of the data column space.

[0165] Perform eigenvalue decomposition on A, use numerical linear algebra libraries (such as NumPy, MATLAB, etc.) to perform eigenvalue decomposition on Koopman operator A, extract eigenvalues ​​and corresponding eigenvectors, and for each eigenvalue, calculate the modulus |λ| of the eigenvalue. If |λ|>1, the corresponding mode is growing, and if it is less than 1, the mode is decaying, indicating that the fluid is dynamically unstable. The system is optimized by adjusting the damping coefficient. In the grouting process, the damping coefficient is mainly used to describe the resistance or dissipation effect of the slurry during the flow and diffusion process, especially when the slurry passes through complex pores or cracks, the damping coefficient determines the rate of slurry flow decay. The damping coefficient can be related to the viscosity of the fluid, the formation characteristics, and the interaction between the slurry and the medium.

[0166] The instantaneous dynamic changes of the slurry on the time scale and the acquisition of structural and periodic dynamic features on the spatial scale are realized. In the process of capturing the flow field characteristics, the high-dimensional dilemma caused by random noise or non-critical features in the non-overfitting data is eliminated, and the dimension reduction of the data is realized. The extracted data is converted into a format suitable for machine learning training, such as a CSV file, and the data is standardized and converted into data with a mean of 0 and a standard deviation of 1. Based on the above steps, the extracted data is divided into a training set, a test set, and a comparison set according to the time ratio, and the neural network model is trained.

[0167] Perform pre-training tests, train the same case with different time steps, adjust the network weights and bias parameters according to the pre-training test results, and continue training until the training is completed.

[0168] The input data is numerically simulated in the initial stage through the numerical simulation method. The time span of the initial stage is dynamically adjusted according to the actual complexity of the working conditions. The factors considered include: grouting flow rate, grouting pressure, grouting flow rate, slurry selection, and geological conditions.

[0169]

[0170] χ=α v m+α τ τ+α S S;

[0171] k=α u u+α p p+α f f+α d d;

[0172] Among them, T0 is the basic time span, α, β, γ, δ, α v , α τ , α S , α u , α p , α f , α d are weight coefficients, is the inverse of the grouting rate, is the inverse of the grouting pressure, χ is the slurry selection parameter, k is the geological condition complexity coefficient, m is the slurry viscosity, τ is the curing time, S is the chemical stability coefficient, u is the stability of the formation mechanical properties, p is the formation permeability, f is the degree of crack development, and d is the disaster-prone structure coefficient.

[0173] After determining the time span of the initial stage, set the time of the final stage to be consistent with the time span of the initial stage. Set the initial conditions, and the time step is the time span of the initial stage. After completing the numerical simulation, input the parameters of the last time step of the initial stage as the initial data into the trained neural network model to obtain the prediction results of slurry velocity, grouting pressure, phase fraction and viscosity data.

[0174] For the output results of the neural network model, 10% of the total time steps are set for numerical solution verification. A three-stage step-by-step trust verification mechanism is established to ensure the reliability and accuracy of the model; the details are as follows:

[0175] Assuming the total time steps are 100, the first 10 time steps are used for validation.

[0176] In the first stage, a fixed number a1 of time step nodes is used for preliminary verification. If the error of the preliminary verification result is controlled within 5%, the second stage is entered.

[0177] In the second stage, the trust in the model is enhanced by gradually reducing the verification frequency until the model output is fully trusted. Specifically, if the error between the model output and the numerical calculation is less than 5% within the set time step a2, the model output is considered to be credible and enters a fully trusted state.

[0178] In the state of full trust, no numerical verification is performed and subsequent simulation prediction is carried out directly. After the prediction is completed, the third stage is entered to conduct strict numerical verification of the prediction results of the last a3 time steps. If the error of the verification result is still within 5%, the grouting simulation result can be output; if it is inconsistent, the initial grouting conditions are re-adjusted and calculated.

[0179] The sum of the time steps of the three stages is 10% of the total time step, that is: a1+a2+a3=10.

[0180] Of course, for the specific error range requirement, this embodiment is set to 5%. In other implementations, it can also be set according to actual needs.

[0181] The numerical simulation results and neural network simulation results are collated, and the neural network parameters are tuned and optimized according to the verification results and actual engineering excavation findings.

[0182] Furthermore, the S103: simulation data includes: prediction results of grouting speed, grouting pressure, phase fraction and viscosity parameters.

[0183] Furthermore, the S104: preprocessing and aligning the data of steps S101 to S102 and S103 respectively, and performing fusion analysis; wherein the alignment operation refers to alignment in the time dimension.

[0184] Furthermore, the step S104 preprocesses and aligns the data of steps S101 to S102 and S103, and performs fusion analysis; wherein the fusion analysis includes:

[0185] Preprocessing of physical model test data includes data cleaning, interpolation, and denoising of grouting pressure, flow rate, flow velocity, slurry ratio, etc.

[0186] The pre-processed physical model test data and numerical simulation results are unified in format and aligned in time and space coordinate systems and mapped into the same spatial domain;

[0187] The optimal grouting parameters for the next step are regarded as predicted values, and the physical test and numerical simulation data are regarded as observed values. The optimal prediction results are adjusted through Kalman filter iterative optimization. The specific steps are as follows:

[0188] Construct the grouting parameter equation:

[0189] x k =Ax k-1 +Bu k-1

[0190] Among them, x k represents the optimal grouting parameters to be estimated, such as pressure, flow rate, and slurry ratio; A represents the state transfer matrix, which is used to describe how the state at the previous moment is transferred to the current moment; B represents the control input matrix, which is used to describe the impact of external control input; x k-1 Indicates the grouting parameters at the last moment; u k-1 Indicates the control behavior of the system at the last moment.

[0191] Construct the grouting monitoring data equation:

[0192] z k =Hx k +v k

[0193] Among them, z k represents observation data; H represents the observation matrix; v k represents the monitoring data noise;

[0194] Based on the state estimate at the previous moment, predict the state and covariance at the current moment:

[0195]

[0196] in, Indicates the predicted state at the current moment; Represents the updated state at time k-1.

[0197] P k|k-1 =AP k-1|k-1 A T +Q

[0198] Where Q represents the process noise covariance; P k|k-1 represents the prediction covariance, which indicates the uncertainty of the current state estimate; P k-1|k-1 represents the updated covariance at time k-1.

[0199] According to monitoring data k Correct the predicted state and calculate the Kalman gain:

[0200] K k =P k|k-1 H T (HP k|k-1 H T +R) -1

[0201] Among them, K k represents the Kalman gain, which balances the weight between the predicted state and the observed data. The larger the Kalman gain, the more dependent it is on the observed data, and the smaller the gain, the more dependent it is on the predicted state. R represents the observation noise covariance, which indicates the noise level of the observed data.

[0202] Update grouting parameter estimates:

[0203]

[0204] in, Represents the residual between the predicted value and the actual observed value, and the Kalman filter updates the state estimate based on the residual.

[0205] Update the covariance:

[0206] P k|k =(IK k H)P k|k-1

[0207] Finally, the optimal grouting parameters and covariance estimation at the next moment are predicted through the grouting parameter data at the current moment (such as pressure, flow, slurry selection, etc.) and the monitoring data of the current physical test data (grouting pressure, flow, slurry diffusion range, etc.) and numerical simulation data (grouting pressure, flow, phase fraction, etc.), thereby realizing the fusion analysis of physical test data and numerical simulation data.

[0208] Furthermore, in S105: pre-control and adjust the grouting equipment parameters and the grouting scheme according to the fusion analysis results.

[0209] For the grouting scheme, according to the fusion analysis results, if the results show that the slurry spreads too fast or too slow, adjust the grouting pressure accordingly to ensure that the slurry can evenly cover the area to be reinforced;

[0210] Adjust the slurry flow rate and flow rate. If the results show that the slurry cannot completely penetrate the target area, increase the flow rate, flow rate or optimize the slurry ratio and adjust the location or angle of the grouting holes based on the analysis results.

[0211] For the grouting equipment parameters, according to the fusion analysis results, if the numerical simulation and test data show that the grouting pressure is unstable or the flow is insufficient, the pump power and speed are adjusted; if the monitored flow rate and pressure changes are abnormal, the valve settings or pipeline paths of the pipeline are adjusted.

[0212] Embodiment 2

[0213] This embodiment provides a virtual-real synergistic test system for disaster grouting prevention and control in sandy and gravel strata, including:

[0214] An acquisition module is configured to: conduct a grouting test on a sand-pebble formation simulation model to obtain grouting parameters and monitoring parameters during the grouting process, wherein the grouting parameters include: grouting pressure, grouting flow rate and slurry ratio, and the monitoring parameters include: resistivity data, radar data and optical fiber sensor data;

[0215] An inversion module is configured to: obtain the slurry diffusion path and diffusion range in the area with higher surrounding rock grade based on the resistivity data and the radar data; obtain the slurry diffusion path and diffusion range in the area with lower surrounding rock grade based on the resistivity data and the optical fiber sensing data;

[0216] The simulation module is configured to: input the initial grouting parameters into the trained neural network model to obtain simulation data;

[0217] A fusion analysis module is configured to: pre-process and align the data of the acquisition module, the inversion module and the simulation module respectively, and perform fusion analysis;

[0218] The pre-control adjustment module is configured to: perform pre-control adjustments on the grouting equipment parameters and the grouting scheme according to the fusion analysis results.

[0219] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A virtual-real synergistic experimental method for disaster grouting prevention and control in sandy and gravel strata, characterized by: include: (1) conducting a grouting test on a sand and gravel formation simulation model to obtain grouting parameters and monitoring parameters during the grouting process, wherein the grouting parameters include: grouting pressure, grouting flow rate and slurry ratio; and the monitoring parameters include: resistivity data, radar data and optical fiber sensor data; (2) Based on the resistivity data and radar data, the slurry diffusion path and diffusion range in the area with higher surrounding rock grade are obtained; based on the resistivity data and optical fiber sensing data, the slurry diffusion path and diffusion range in the area with lower surrounding rock grade are obtained; (3) Inputting the initial grouting parameters into the trained neural network model to obtain simulation data; (4): Preprocess and align the data of (1) to (2) and (3) respectively, and perform fusion analysis; (5) Based on the fusion analysis results, pre-control adjustments are made to the grouting equipment parameters and grouting scheme.

2. The virtual-real fusion collaborative test method for disaster grouting prevention and control in sand and gravel strata as claimed in claim 1 is characterized by: The outer surface of the sand and gravel formation simulation model is arranged with electrodes, a radar is arranged within a set distance range outside the sand and gravel formation simulation model, and optical fibers are arranged in the drilling holes and grouting holes of the sand and gravel formation simulation model; Electrodes are arranged on the outer surface of the sand and gravel formation simulation model, wherein the electrode arrangement step includes: establishing a mathematical model based on electrode positions and slurry diffusion range according to the conductivity, porosity, fracture distribution formation characteristics of the formation and the preset number of electrodes, accuracy requirements, monitoring requirements of monitoring coverage, and the expected diffusion radius of the slurry through a simulated annealing algorithm, and obtaining the electrode arrangement positions by solving the mathematical model; A radar is arranged within a set distance range outside the sand and gravel formation simulation model, wherein the radar arrangement step includes: arranging radar sensors in the peripheral area where there are fracture zones, broken zones, and heterogeneous rock formation areas in the arranged formation, and integrating a simulated annealing algorithm to determine the optimal arrangement position and number of the radar sensors; Optical fibers are arranged in the boreholes and grouting holes of the sand-pebble stratum simulation model, including: for water-rich and sand-rich strata, deformed strata, deep strata and multi-layer alternating geological structures, after the grouting borehole is drilled, optical fibers are arranged on the borehole wall; for the borehole using catheter grouting, optical fibers are arranged on the catheter wall.

3. The virtual-real fusion collaborative test method for disaster grouting prevention and control in sand and gravel strata as claimed in claim 1 is characterized in that: Based on the resistivity data and radar data, the slurry diffusion path and diffusion range in the area with higher surrounding rock grade are obtained, including: for radar signal acquisition, the radar sensor is used to scan the tunnel face at multiple angles and orientations, and finally the location and structural characteristics of the surrounding rock water-conducting channels and cavities in front of the tunnel face based on the radar image in the formation are obtained; data synchronization and calibration are performed to ensure that the radar signal and resistivity data have consistent benchmarks in time and space, and calibration and alignment are performed if they are inconsistent; through the joint tracking of resistivity and radar signals, the slurry diffusion path in the area with higher surrounding rock grade is monitored; Based on the resistivity data and fiber optic sensing data, the slurry diffusion path and diffusion range in the area with lower surrounding rock grade are obtained, including: data preprocessing: ensuring the time synchronization of resistivity data and fiber optic sensing data, ensuring that the global resistivity and local strain, temperature, and pressure data are aligned on the time axis; filtering and noise elimination of fiber optic sensing data, and resampling the spatial resolution and time resolution of different data to ensure consistency in space and time; data fusion: based on feature extraction and fusion, the prediction of grouting diffusion range is realized, or, based on data analysis, the prediction of grouting diffusion range is realized.

4. The virtual-real fusion collaborative test method for disaster grouting prevention and control in sand and gravel strata as claimed in claim 3 is characterized by: The method based on feature extraction and fusion to realize the prediction of grouting diffusion range includes: First, the resistivity data is preprocessed and converted into two-dimensional or three-dimensional grid data. The global spatial diffusion feature vector F is extracted through the convolutional neural network (CNN). Resistivity ; The data monitored by the optical fiber is processed into time series data, and the local time series feature vector F is extracted using the long short-term memory network Fiber ; The attention mechanism is used to fuse the global and local feature vectors, giving different weights to the global feature vector and the local feature vector, and outputting the fused feature vector F Fusion ; In Resistivity +in Fiber =1; F Fusion =w Resistivity ·F Resistivity +w Fiber ·F Fiber ; Among them, w Resistivity Represents the resistivity data weight; w Fiber Indicates the fiber data weight; Constraints are implemented by adding regularization terms to the loss function; L physics =λ1·L diffusion +λ2·L strain ; Among them, λ1 and λ2 are physical constraint weight coefficients, L diffusion With L strain represents the loss term for the diffusion equation and strain constraints; L diffusion It is expressed as: in, Represents the resistivity value predicted by the model; N represents the number of sampling points; D represents the diffusion coefficient, which is summarized through a large number of experimental rules: L strain It is expressed as: in, Represents the strain value output by the model; σ i represents the stress obtained from the optical fiber monitoring data; E represents the Young's modulus of the material; M represents the number of sampling points; L total =L data (F Fusion )+L physics ; Among them, L data is the loss of data in the fitting process, L total represents the total loss; The fused feature vector is input into the fully connected layer to achieve continuous output of the slurry diffusion range.

5. The virtual-real fusion collaborative test method for disaster grouting prevention and control in sand and gravel strata as claimed in claim 3 is characterized by: The method based on data analysis realizes the prediction of the grouting diffusion range, including: based on the pre-processed global resistivity change data, determining the area where the resistivity decreases beyond the set threshold, corresponding to the injection area of ​​the slurry and the range to which the slurry diffuses, and obtaining a rough diffusion range; based on the pre-processed local strain, temperature and pressure data, extracting the abnormal change points of strain, temperature and pressure, and combining the slurry front position and the slurry diffusion boundary determined by the grouting theory to refine the rough diffusion range; The process of determining the slurry front position includes: firstly, cleaning and preprocessing the abnormal change point data of temperature and pressure captured by the optical fiber sensing data, using a low-pass filtering algorithm to remove high-frequency noise, and dividing the data of a long period of time into multiple time windows or space windows for local extraction and analysis, and spatially calibrating the spatial coordinate position of the data point with the resistivity data to ensure that the change rate of the acquired data is calculated under the same spatial coordinate, and when the change rate exceeds a certain critical value, the point is considered to be an abnormal change point, the abnormal change point is extracted, and the abnormal change point of the optical fiber sensing path is projected onto the resistivity imaging, and finally the slurry front position is determined; in, is the rate of temperature change in the spatial direction; is the rate of change of temperature in the time direction.

6. The virtual-real fusion collaborative test method for disaster grouting prevention and control in sand and gravel strata as claimed in claim 5 is characterized by: The slurry diffusion boundary determination process includes: The optical fiber monitoring strain data, surrounding rock permeability and elastic modulus are used to achieve boundary refinement. The weighted calculation formula is: ΔX=w ∈ ·E ∈ +w K ·E K +w E ·E E ; Among them, ΔX represents the deformation, E ∈ represents the normalized strain, E K represents the normalized permeability, E E represents the normalized elastic modulus, w ∈ 、w K 、w E denote the weights of strain, permeability, and elastic modulus, respectively; The boundary position after deformation is: X new =X old +ΔX; Among them, X old Represented as the position of the original diffusion boundary, X new Indicates the position of the boundary after deformation.

7. The virtual-real fusion collaborative test method for disaster grouting prevention and control in sand and gravel strata as claimed in claim 1 is characterized by: The initial grouting parameters are input into the trained neural network model to obtain simulation data. The training process includes: Constructing a training set, wherein the training set is grouting parameters of known simulation data; The training set is input into the neural network model to train the neural network model. When the loss function value of the neural network model no longer decreases, or the number of iterations exceeds the set number, the training is stopped to obtain the trained neural network model.

8. The virtual-real fusion collaborative test method for disaster grouting prevention and control in sand and gravel strata as claimed in claim 1 is characterized by: Fusion analysis, including: The preprocessing of physical model test data includes data cleaning, interpolation and denoising of grouting pressure, flow rate, flow velocity and slurry ratio; The pre-processed physical model test data and numerical simulation results are unified in format and aligned in time and space coordinate systems and mapped into the same spatial domain; The optimal grouting parameters for the next step are regarded as predicted values, and the physical test and numerical simulation data are regarded as observed values. The optimal prediction results are adjusted through Kalman filter iterative optimization. The specific steps are as follows: Construct the grouting parameter equation: x k =Ax k-1 +Bu k-1 4 Among them, x k represents the optimal grouting parameters to be estimated; A represents the state transfer matrix, which is used to describe the state transfer from the previous moment to the current moment; B represents the control input matrix, which is used to describe the influence of external control input; x k-1 Indicates the grouting parameters at the last moment; u k-1 Indicates the control behavior of the system at the last moment; Construct the grouting monitoring data equation: z k =Hx k +v k ; Among them, z k represents observation data; H represents the observation matrix; v k represents the monitoring data noise; Based on the state estimate at the previous moment, predict the state and covariance at the current moment: in, Indicates the predicted state at the current moment; Represents the updated state at time k-1; P k|k-1 =AP k-1|k-1 From T +Q; Where Q represents the process noise covariance; P k|k-1 represents the prediction covariance; P k-1|k-1 represents the updated covariance at time k-1; According to monitoring data k Correct the predicted state and calculate the Kalman gain: Among them, K k represents the Kalman gain, R represents the observation noise covariance; Update grouting parameter estimates: in, Represents the residual between the predicted value and the actual observed value. The Kalman filter updates the state estimate based on the residual. Update the covariance: P k|k =(I-K k H)P k|k-1 ; Finally, the optimal grouting parameters and covariance estimation at the next moment are predicted through the grouting parameter data at the current moment and the monitoring data of the physical test data and numerical simulation data at the current moment, realizing the fusion analysis of physical test data and numerical simulation data.

9. The virtual-real fusion collaborative test method for disaster grouting prevention and control in sand and gravel strata as claimed in claim 1 is characterized by: According to the fusion analysis results, the grouting equipment parameters and grouting scheme are adjusted in advance, including: For the grouting scheme, according to the fusion analysis results, if the results show that the slurry spreads too fast or too slow, adjust the grouting pressure accordingly to ensure that the slurry can evenly cover the area to be reinforced; Adjust the slurry flow rate and flow rate. If the results show that the slurry cannot fully penetrate the target area, increase the flow rate, flow rate or optimize the slurry ratio and adjust the layout or angle of the grouting holes according to the analysis results; For the parameters of grouting equipment, according to the fusion analysis results, if the numerical simulation and test data show that the grouting pressure is unstable or the flow rate is insufficient, the power and speed of the pump are adjusted; If the monitored flow rate and pressure changes are abnormal, adjust the valve settings or pipeline path of the pipeline.

10. A virtual-real synergistic test system for disaster grouting prevention and control in sand and gravel strata, which is characterized by including: An acquisition module is configured to: conduct a grouting test on a sand-pebble formation simulation model to obtain grouting parameters and monitoring parameters during the grouting process, wherein the grouting parameters include: grouting pressure, grouting flow rate and slurry ratio, and the monitoring parameters include: resistivity data, radar data and optical fiber sensor data; An inversion module is configured to: obtain the slurry diffusion path and diffusion range in the area with higher surrounding rock grade based on the resistivity data and the radar data; obtain the slurry diffusion path and diffusion range in the area with lower surrounding rock grade based on the resistivity data and the optical fiber sensing data; The simulation module is configured to: input the initial grouting parameters into the trained neural network model to obtain simulation data; A fusion analysis module is configured to: pre-process and align the data of the acquisition module, the inversion module and the simulation module respectively, and perform fusion analysis; The pre-control adjustment module is configured to: perform pre-control adjustments on the grouting equipment parameters and the grouting scheme according to the fusion analysis results.

Images