Intelligent regulation and control method and system for rock mass grouting process based on big data

Abstract

The invention relates to a rock mass grouting process intelligent regulation and control method and system based on big data. The method comprises the following steps: acquiring real-time image data, real-time slurry performance parameters and real-time rock mass response data of a grouting position; on the basis of the real-time image data, the real-time slurry performance parameters and the real-time rock mass response data, analyzing the characteristics of the crack at the grouting position to obtain a comprehensive crack characteristic vector; based on the comprehensive fracture feature vector and a preset threshold mapping rule set, performing state evaluation on the grouting position to obtain fracture early warning data; generating an intelligent regulation and control parameter vector based on the fracture early warning data, the comprehensive fracture feature vector and a preset historical case library; the intelligent regulation and control parameter vector is used for indicating adjustment of grouting parameters; the historical case library comprises the crack early warning data, the comprehensive crack feature vector, grouting adjustment parameters and effect scores of historical grouting cases. By adopting the method, the grouting safety and accuracy can be improved.

Description

Intelligent Control Method and System for Rock Mass Grouting Process Based on Big Data Technical Field

[0001] This invention belongs to the field of rock grouting engineering, and in particular relates to an intelligent control method and system for the rock grouting process based on big data. Background Technology

[0002] With the development of rock grouting engineering technology, grouting control technology based on sensor monitoring and empirical rules has emerged. This technology collects grouting process parameters in real time by deploying equipment such as pressure sensors and flow meters, and adjusts the parameters in combination with the empirical thresholds set by engineers in advance, thus forming a traditional grouting control method with "monitoring-threshold judgment-human intervention" as the core.

[0003] In traditional grouting control methods, engineers first pre-determine fixed grouting pressure ranges, flow rates, and grout mix ratios based on geological survey reports and past experience. During grouting construction, a sensor system continuously collects parameters such as grouting pressure, flow rate, and grout density. When the monitored data exceeds preset thresholds, the system issues an alarm, and on-site engineers determine whether parameter adjustments are necessary based on their experience.

[0004] However, current traditional grouting control methods suffer from a fundamental contradiction between the dynamic changes in rock mass fractures and the static setting of grouting parameters. Lacking real-time sensing and prediction capabilities for fracture propagation, the system can only passively respond after fractures change and affect grouting parameters, resulting in a significant lag in parameter adjustments. Particularly under complex geological conditions, when characteristic parameters such as the width, length, and distribution density of rock mass fractures exhibit nonlinear changes due to stress fields, permeability, and the external environment, traditional monitoring methods struggle to capture fracture characteristics in real-time and accurately. This prevents adjustments to grouting pressure and grout mix ratio from keeping pace with dynamic fracture changes, leading to a decline in the safety and accuracy of grouting. Summary of the Invention

[0005] Therefore, it is necessary to provide a big data-based intelligent control method and system for the rock mass grouting process that can improve the safety and accuracy of grouting, addressing the aforementioned technical problems.

[0006] Firstly, this application provides an intelligent control method for rock mass grouting process based on big data, including:

[0007] Acquire real-time image data, real-time grout performance parameters, and real-time rock mass response data of the grouting location;

[0008] Based on real-time image data, real-time grout performance parameters, and real-time rock mass response data, the characteristics of the fractures at the grouting location are analyzed to obtain a comprehensive fracture feature vector. The comprehensive fracture feature vector includes at least the fracture foundation geometric parameters, fracture morphological complexity, real-time grout performance parameters, and real-time rock mass response data.

[0009] Based on the comprehensive fracture feature vector and the preset threshold mapping rule set, the state of the grouting location is assessed to obtain fracture early warning data; the fracture early warning data includes early warning level value and state confidence level.

[0010] Based on crack early warning data, comprehensive crack feature vectors, and a pre-set historical case library, an intelligent control parameter vector is generated. The intelligent control parameter vector is used to indicate the adjustment of grouting parameters. The historical case library includes crack early warning data, comprehensive crack feature vectors, grouting adjustment parameters, and effect scores of historical grouting cases.

[0011] Furthermore, based on real-time image data, real-time grout performance parameters, and real-time rock mass response data, the characteristics of fractures at the grouting location are analyzed to obtain a comprehensive fracture feature vector, including:

[0012] Based on real-time image data, denoising processing is performed on the real-time image data to obtain denoised image data;

[0013] Based on real-time image data, edge enhancement processing is performed on the denoised image data to obtain preprocessed image data;

[0014] Based on a pre-defined crack segmentation model, the pre-processed image data is segmented to obtain a binary crack segmentation map; in the binary crack segmentation map, pixels with a value of 1 represent cracks, and pixels with a value of 0 represent the background.

[0015] Based on the binarized fracture segmentation map, the geometric features of the fractures are extracted to obtain the basic geometric parameters of the fractures; the basic geometric parameters of the fractures include the fracture width value sequence, fracture length, and fracture distribution density;

[0016] Based on the fundamental geometric parameters of the fracture, the morphological complexity of the fracture is quantified to obtain the fracture morphological complexity.

[0017] Based on the basic geometric parameters of the fracture, the complexity of the fracture morphology, the real-time slurry performance parameters, and the real-time rock mass response data, a comprehensive fracture feature vector is obtained.

[0018] Furthermore, based on the binarized fracture segmentation map, the geometric features of the fractures are extracted to obtain the basic geometric parameters of the fractures, including:

[0019] Based on the binary fracture segmentation map, the pixel coordinates of the fracture centerline are extracted to obtain the fracture centerline sequence;

[0020] Based on the fracture centerline sequence, the cross-sections perpendicular to the centerline are extracted to obtain the normal cross-section set;

[0021] Based on the set of normal profiles, the position of the edge points of each profile is identified to obtain a sequence of crack width values;

[0022] Based on the fracture centerline sequence, the fracture length is calculated using the following formula:

[0023]

[0024] in, It is the crack length. It is the total number of pixels in the crack centerline sequence. It is any pixel in the sequence of crack centerlines. It is a pixel. x-coordinate It is a pixel. x-coordinate It is a pixel. The ordinate, It is a pixel. The ordinate;

[0025] Based on the binary crack segmentation map and the preset spatial grid size, the distribution density of crack pixels is statistically analyzed to obtain the crack distribution density.

[0026] Based on the crack width value sequence, crack length, and crack distribution density, the basic geometric parameters of the crack are obtained.

[0027] Furthermore, based on the comprehensive fracture feature vector and a preset threshold mapping rule set, the state of the grouting location is assessed to obtain fracture early warning data, including:

[0028] Based on the threshold mapping rule set, the comprehensive crack feature vector is mapped to the corresponding matching threshold to obtain the matching threshold vector;

[0029] Based on the matching threshold vector and the preset state level rule set, the matching threshold is comprehensively evaluated to obtain the state level confidence distribution;

[0030] The warning level value is obtained based on the confidence distribution of the state level and the rule set of the state level;

[0031] Based on the warning level value, the confidence distribution of the state level is normalized to obtain the state confidence; and based on the warning level value and the state confidence, crack warning data is generated.

[0032] Furthermore, based on fracture early warning data, comprehensive fracture feature vectors, and a pre-set historical case library, an intelligent control parameter vector is generated, including:

[0033] For each historical grouting case in the historical case database, based on crack early warning data, comprehensive crack feature vectors, and effect scores, the similarity between historical grouting cases and grouting locations is quantified to obtain the similarity of each historical grouting case; and the corresponding historical grouting cases whose similarity meets the preset similarity threshold are selected to form a set of similar cases.

[0034] For each historical grouting case in the set of similar cases, the accurate similarity between the historical grouting case and the grouting location is calculated based on preset feature weight parameters, crack early warning data, comprehensive crack feature vector and effect score.

[0035] From the set of similar cases, select the historical grouting case with the maximum accurate similarity as the optimal similar case;

[0036] Based on the grouting adjustment parameters of the best similar case, real-time grout performance parameters, real-time rock mass response data, and preset operating constraints, a parameter optimization solution set is generated; the parameter optimization solution set includes the solution vectors for each grouting adjustment parameter.

[0037] Based on crack early warning data and comprehensive crack feature vector, the solution vector for grouting adjustment parameters is selected from the parameter optimization solution set to obtain the optimal combination of grouting parameters;

[0038] Based on the optimal combination of grouting parameters, an intelligent control parameter vector is generated.

[0039] Furthermore, based on crack early warning data and comprehensive crack feature vectors, the solution vector for grouting adjustment parameters is selected from the parameter optimization solution set to obtain the optimal combination of grouting parameters, including:

[0040] Based on the parameter optimization solution set and the comprehensive fracture feature vector, the solution vectors of each grouting adjustment parameter are evaluated to obtain the evaluation index values ​​of each grouting adjustment parameter; the evaluation index values ​​include the calculation quality index value, the time index value, and the cost index value.

[0041] Based on the evaluation index values ​​of each grouting adjustment parameter, the maximum and minimum values ​​of the quality index, the maximum and minimum values ​​of the time index, and the maximum and minimum values ​​of the cost index are extracted to form a set of parameter evaluation extreme values;

[0042] Based on the set of extreme values ​​of parameter evaluation, generate positive ideal adjustment parameter solution vectors and negative ideal parameter solution vectors;

[0043] For each grouting adjustment parameter solution vector in the parameter optimization solution set, based on crack early warning data and a preset parameter adjustment rule set, the difference between the evaluation index value and the positive ideal adjustment parameter solution vector is quantified to obtain the positive solution distance of each grouting adjustment parameter solution vector; the difference between the evaluation index value and the negative ideal parameter solution vector is quantified to obtain the negative solution distance of each grouting adjustment parameter solution vector.

[0044] For each solution vector of the grouting adjustment parameters in the parameter optimization solution set, the relative proximity of the solution vectors is calculated using the following formula, based on the positive and negative solution distances:

[0045]

[0046] in, It is the first The relative closeness of the solution vector for each grouting adjustment parameter. It is the first The positive solution distance of the vector obtained by solving for each grouting adjustment parameter. It is the first The distance between the negative solutions of the grouting adjustment parameters solution vector;

[0047] From the parameter optimization solution set, select the grouting adjustment parameter solution vector corresponding to the maximum relative closeness as the optimal grouting parameter combination.

[0048] Secondly, this application also provides an intelligent control system for rock mass grouting process based on big data, including:

[0049] The data acquisition module is used to acquire real-time image data, real-time grout performance parameters, and real-time rock mass response data of the grouting location;

[0050] The data analysis module is used to analyze the characteristics of the fractures at the grouting location based on real-time image data, real-time grout performance parameters, and real-time rock mass response data, and obtain a comprehensive fracture feature vector. The comprehensive fracture feature vector includes at least the fracture basic geometric parameters, fracture morphological complexity, real-time grout performance parameters, and real-time rock mass response data.

[0051] The early warning calculation module is used to assess the state of grouting locations based on comprehensive fracture feature vectors and a preset threshold mapping rule set, and obtain fracture early warning data; the fracture early warning data includes early warning level values ​​and state confidence levels;

[0052] The regulation generation module is used to generate intelligent regulation parameter vectors based on crack early warning data, comprehensive crack feature vectors, and a preset historical case library. The intelligent regulation parameter vectors are used to indicate the adjustment of grouting parameters. The historical case library includes crack early warning data, comprehensive crack feature vectors, grouting adjustment parameters, and effect scores of historical grouting cases.

[0053] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement any of the intelligent control methods for rock mass grouting processes based on big data described in the first aspect of this application.

[0054] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any of the big data-based intelligent control methods for rock mass grouting processes described in the first aspect of this application.

[0055] The aforementioned intelligent control method and system for rock grouting process based on big data acquires real-time image data, real-time grout performance parameters, and real-time rock mass response data at the grouting location. Based on these data, the system analyzes the characteristics of fractures at the grouting location to obtain a comprehensive fracture feature vector. This vector includes at least the fracture's basic geometric parameters, fracture morphological complexity, real-time grout performance parameters, and real-time rock mass response data. Based on the comprehensive fracture feature vector and a preset threshold mapping rule set, the system performs a state assessment of the grouting location to obtain fracture early warning data. This data includes an early warning level value and a state confidence level. Based on the early warning data, the comprehensive fracture feature vector, and a preset historical case library, an intelligent control parameter vector is generated. This vector indicates adjustments to the grouting parameters. The historical case library includes fracture early warning data, the comprehensive fracture feature vector, grouting adjustment parameters, and effect scores for historical grouting cases. It breaks through the contradiction between the dynamic changes of cracks and the static setting of parameters in traditional grouting. By replacing manual observation with real-time image recognition, replacing fixed thresholds with adaptive thresholds, and optimizing experience-based reliance through case reasoning, it improves the safety and accuracy of the grouting process. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 is a flowchart illustrating an intelligent control method for rock mass grouting process based on big data, provided in an embodiment of this application.

[0058] Figure 2 is a schematic diagram of the structure of an intelligent control system for rock grouting process based on big data provided in an embodiment of this application;

[0059] Figure 3 is a schematic diagram of the structure of a computer device for an intelligent control method of rock mass grouting process based on big data, provided in an embodiment of this application. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0061] In one embodiment, as shown in Figure 1, a method for intelligent control of rock grouting process based on big data is provided. This embodiment illustrates the application of this method to a terminal. It is understood that this method can also be applied to a server, and can also be applied to a system including a terminal and a server, and implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following S101-S104, wherein:

[0062] S101 acquires real-time image data, real-time grout performance parameters, and real-time rock mass response data of the grouting location.

[0063] Specifically, the terminal acquires real-time image data, real-time grout performance parameters, and real-time rock mass response data at the grouting location. Real-time image data characterizes the surface state of the rock mass in the grouting area, in pixel matrix form, and can be acquired using a borehole camera. Real-time grout performance parameters represent the delivery state and permeability of the grouting material, including at least the grouting pressure, and can be obtained in real-time through pressure sensors, flow meters, and density meters. Real-time rock mass response data reflects the mechanical response and structural changes of the rock mass during the grouting process, including at least acoustic emission signal data, and can be acquired using acoustic sensors and displacement gauges.

[0064] S102. Based on real-time image data, real-time grout performance parameters, and real-time rock mass response data, the characteristics of the fractures at the grouting location are analyzed to obtain a comprehensive fracture feature vector. The comprehensive fracture feature vector includes at least the fracture foundation geometric parameters, fracture morphological complexity, real-time grout performance parameters, and real-time rock mass response data.

[0065] Specifically, the terminal performs collaborative analysis on real-time image data, real-time grout performance parameters, and real-time rock mass response data to extract key features characterizing the fracture state, obtaining a comprehensive fracture feature vector. Among these, the basic geometric parameters of the fracture characterize the geometric features of the rock mass fractures at the grouting location, including the macroscopic size and distribution of the fractures. The fracture morphology complexity characterizes the irregularity, tortuosity, and morphological characteristics of the fracture edges at multiple scales; for example, a winding, rough-edged fracture has a much higher morphological complexity than a straight, smooth-edged fracture. The real-time grout performance parameters and real-time rock mass response data are those obtained in step S101. For example, the terminal can use image processing algorithms to identify the fracture geometry and obtain the basic geometric parameters of the fracture.

[0066] S103, based on the comprehensive fracture feature vector and the preset threshold mapping rule set, performs a status assessment on the grouting location to obtain fracture early warning data; the fracture early warning data includes early warning level value and status confidence.

[0067] Specifically, a preset threshold mapping rule set is used to characterize the mapping relationship between the values ​​of each data item included in the comprehensive crack feature vector and the corresponding matching thresholds. This set can be constructed through historical data analysis and domain expert experience. The terminal performs a comprehensive analysis of the matching thresholds of each data item included in the comprehensive crack feature vector to obtain crack early warning data for each data item included in the comprehensive crack feature vector. The early warning level value, representing the degree of risk, is obtained by fusing the matching thresholds of each data item included in the comprehensive crack feature vector. The state confidence level reflects the reliability of the early warning level value. For example, the preset threshold mapping rule set can take the form of: ,in It is a threshold mapping rule set. It is any data item included in the comprehensive fracture feature vector. It is the index number of the crack morphology complexity. This is the matching threshold for the complexity of the crack morphology. The preset threshold mapping rule set can be set according to actual work, and this embodiment does not further limit the setting of this preset threshold mapping rule set.

[0068] S104 generates an intelligent control parameter vector based on crack early warning data, comprehensive crack feature vector, and a preset historical case library. The intelligent control parameter vector is used to indicate the adjustment of grouting parameters. The historical case library includes crack early warning data, comprehensive crack feature vector, grouting adjustment parameters, and effect scores of historical grouting cases.

[0069] Specifically, the pre-set historical case library includes fracture early warning data, comprehensive fracture feature vectors, grouting adjustment parameters, and effect scores for historical grouting cases. The terminal calculates the similarity between the fracture early warning data and comprehensive fracture feature vector of the current grouting location and each historical grouting case in the pre-set historical case library, finding the historical case most similar to the current grouting location. It then adjusts the grouting adjustment parameters of this case based on the real-time grout performance parameters and real-time rock mass response data of the current grouting location, obtaining an intelligent control parameter vector. This intelligent control parameter vector guides the real-time adjustment of grouting parameters. For example, the pre-set historical case library can take the form of… ,in It is a pre-set historical case library. This is any one of the historical grouting cases. . This is a historical grouting case. Crack early warning data. This is a historical grouting case. The comprehensive fracture feature vector. This is a historical grouting case. Grouting adjustment parameters, , This is a historical grouting case. Grouting pressure, This is a historical grouting case. Grouting flow rate, This is a historical grouting case. The density of the slurry. This is a historical grouting case. The effectiveness score. The preset historical case library can be obtained from actual work, and this embodiment does not further limit the setting of this preset historical case library.

[0070] This embodiment provides an intelligent control method for rock mass grouting based on big data. Through four steps—multi-source data fusion, fracture feature quantification, intelligent state assessment, and parameter optimization decision-making—it achieves precise perception and intelligent control of the grouting process. It overcomes the contradiction between dynamic fracture changes and static parameter settings in traditional grouting by replacing manual observation with real-time image recognition, replacing fixed thresholds with adaptive thresholds, and optimizing experience-based approaches through case-based reasoning, thereby improving the safety and accuracy of the grouting process.

[0071] In one embodiment, based on real-time image data, real-time grout performance parameters, and real-time rock mass response data, the characteristics of the fractures at the grouting location are analyzed to obtain a comprehensive fracture feature vector, including:

[0072] S201, based on real-time image data, performs denoising processing on the real-time image data to obtain denoised image data.

[0073] Specifically, the terminal employs an adaptive weighted median filtering algorithm to denoise real-time image data, obtaining denoised image data. The adaptive weighted median filtering algorithm is a non-linear filtering technique. The terminal uses a sensor with a size of [size missing]. The sliding window displays each pixel in the real-time image data. Aggregate the data. The set of all pixels within this window is... The terminal uses the following formula: Calculate any pixel point within the window The denoised values ​​of each pixel are then integrated according to the order in the real-time image data to obtain the denoised image data. It is a pixel. The noise reduction value. It is one of the sizes of the aforementioned sliding window. . This indicates the operation of finding the midpoint. It refers to any pixel within the sliding window. These are adaptive weighting coefficients. Weights It is not fixed, but dynamically adjusted based on local gradient information: in image edge regions (where the local gradient is large), lower weights are assigned to reduce the smoothing of edge structures; in flat regions (where the local gradient is small), higher weights are assigned to enhance the denoising effect.

[0074] S202, based on real-time image data, performs edge enhancement processing on the denoised image data to obtain preprocessed image data.

[0075] Specifically, edge enhancement processing uses image processing techniques to strengthen the features of the crack boundary, improving the contrast between the crack and the background. The terminal uses the Sobel operator to perform convolution operations with the denoised image data, calculating the horizontal gradient of each pixel in the denoised image data. and vertical gradient And based on the horizontal gradient of each pixel. and vertical gradient Calculate the gradient magnitude of each pixel. and gradient direction The terminal performs contrast stretching on the gradient magnitude of each pixel to obtain preprocessed image data.

[0076] S203, based on the preset crack segmentation model, the preprocessed image data is segmented to obtain a binary crack segmentation map; in the binary crack segmentation map, pixels with a value of 1 represent cracks and pixels with a value of 0 represent the background.

[0077] Specifically, the terminal inputs preprocessed image data into a preset crack segmentation model to obtain a binary crack segmentation map. In the binary crack segmentation map, pixels with a value of 1 represent cracks, and pixels with a value of 0 represent the background. The preset crack segmentation model is built based on the U-Net++ deep learning network and trained using a preset historical segmentation dataset. The preset historical segmentation dataset includes preprocessed images and corresponding expert-annotated binary segmentation labels (pixels in the crack region have a value of 1, and the background has a value of 0), which can be obtained from actual work. During training, the terminal uses the binary cross-entropy loss function to calculate the difference between the segmentation map predicted by the U-Net++ deep learning network and the true labels, and iteratively updates the network parameters through backpropagation and an optimizer (such as Adam) to obtain the preset crack segmentation model. The terminal inputs the preprocessed image data of the current grouting location into the preset crack segmentation model, outputs a crack probability map, and obtains a binary crack segmentation map through thresholding (e.g., setting a threshold of 0.5). Pixels with a value of 1 represent cracks, and pixels with a value of 0 represent the background.

[0078] S204, based on the binary fracture segmentation map, extracts the geometric features of the fracture to obtain the basic geometric parameters of the fracture; the basic geometric parameters of the fracture include the fracture width value sequence, fracture length and fracture distribution density.

[0079] Specifically, geometric feature extraction quantifies the morphological attributes of fractures using computer vision algorithms, including fracture width value sequences, fracture length, and fracture distribution density. These parameters collectively describe the spatial distribution and dimensional characteristics of fractures. Specifically, the fracture width value sequence characterizes the degree of fracture opening (width) and its variation at different locations along the extension path, reflecting the non-uniformity of fracture opening and directly affecting grout flow; fracture length characterizes the macroscopic scale of fracture distribution on a two-dimensional plane, reflecting the extent to which fractures cut into the rock mass structure; and fracture distribution density characterizes the density of fracture development within the observation area, quantifying the degree of rock mass fragmentation by statistically analyzing the proportion of fracture pixels per unit area.

[0080] S205, based on the basic geometric parameters of the fracture, quantifies the morphological complexity of the fracture to obtain the fracture morphological complexity.

[0081] Specifically, the terminal employs a multifractal spectrum algorithm to analyze the irregularity and complexity of the crack contour, obtaining the crack morphological complexity. The multifractal spectrum algorithm quantifies the heterogeneity and complexity of the morphology by analyzing the singular behavior of the image at different scales and in local regions. The terminal then segments the binary crack image. Consider it as a measure space. Use different side lengths. A square grid covers the image, for each scale Calculate each box The proportion of pixels with internal cracks is defined as a probability measure. Introducing the order of moments ( Using as a weighting factor, calculate the partition function. The partition function and the scale Satisfies the power law relationship: ,in It is a quality index, obtained by analyzing... and The results were obtained through linear regression. Finally, the multifractal spectrum was obtained through Legendre transformation. ,in It is a singularity index that describes the intensity of local singularities. The terminal defines the singularity index as the complexity of the fracture morphology.

[0082] S206, based on the basic geometric parameters of the fracture, the complexity of the fracture morphology, the real-time grout performance parameters, and the real-time rock mass response data, obtains a comprehensive fracture feature vector.

[0083] Specifically, the terminal splices together the basic geometric parameters of the fracture, the complexity of the fracture morphology, the real-time slurry performance parameters, and the real-time rock mass response data to obtain a comprehensive fracture feature vector.

[0084] This embodiment provides an intelligent control method for rock mass grouting process based on big data. Through image preprocessing, precise fracture segmentation, comprehensive feature extraction, and multi-source data fusion, a fracture feature quantification process is performed to obtain a comprehensive fracture feature vector characterizing the overall features of fractures at the irrigation location. This achieves the conversion from raw images to comprehensive feature vectors, providing reliable input data for intelligent control and improving the accuracy and efficiency of fracture identification.

[0085] In one embodiment, based on the binarized fracture segmentation map, the geometric features of the fracture are extracted to obtain the basic geometric parameters of the fracture, including:

[0086] S301. Based on the binary fracture segmentation map, the pixel coordinates of the fracture centerline are extracted to obtain the fracture centerline sequence.

[0087] Specifically, the terminal uses a skeletonization algorithm to process the binary crack segmentation image, extracting the pixel coordinates of the center lines belonging to the cracks in the binary crack segmentation image to form a crack center line sequence. Skeletonization is an image processing technique that, through iterative erosion operations, refines connected regions in a binary image to center lines of a single pixel width without altering the target topology. For the binary crack segmentation image... The terminal uses a structuring element to repeatedly remove boundary pixels of the target region while maintaining the connectivity of the cracks. This process is iterated until no pixels can be removed, ultimately generating a crack skeleton with a width of one pixel, i.e., a sequence of crack centerlines. This sequence precisely describes the continuous orientation and topology of the fractures.

[0088] S302, based on the fracture centerline sequence, extract the cross-sections perpendicular to the centerline to obtain the normal cross-section set.

[0089] Specifically, for each pixel in the crack centerline sequence The terminal calculates its tangent direction vector at that point. The tangent vector is approximated by the coordinate difference between adjacent points on the center line, and its form is: The terminal calculates the perpendicular vector, i.e., the normal vector, of the tangent vector based on the tangent vector. Extending a preset distance of pixel value along this normal direction on both sides of the center line will generate a line segment perpendicular to the crack direction, which is a normal profile corresponding to that pixel. The terminal integrates the normal profiles of each pixel in the crack centerline sequence to obtain a set of normal profiles. For example, the preset surface distance pixel value is set to 10 pixels by default, but it can also be set according to the actual work. This embodiment does not further limit the setting of this preset surface distance pixel value.

[0090] S303, based on the set of normal profiles, identifies the edge point positions of each profile to obtain a sequence of crack width values.

[0091] Specifically, the preset surface distance pixel value is used to identify the crack edge points on each normal profile using an extreme value detection algorithm, resulting in a crack width value sequence. For any normal profile... The terminal starts scanning from the pixel position on the center line and proceeds to both sides of the profile. Edge points are defined as locations where pixel grayscale values ​​change drastically, i.e., gradient extrema. When a pixel value drops sharply from high (inside the crack) to low (background), it is identified as an edge point on that side. The crack width of the profile is the Euclidean distance between the left and right edge points. The terminal integrates the crack widths of each profile to obtain a crack width value sequence.

[0092] S304, based on the fracture centerline sequence, the fracture length is calculated using the following formula:

[0093]

[0094] in, It is the crack length. It is the total number of pixels in the crack centerline sequence. It is any pixel in the sequence of crack centerlines. It is a pixel. x-coordinate It is a pixel. x-coordinate It is a pixel. The ordinate, It is a pixel. The ordinate.

[0095] Specifically, the terminal approximates the continuous fracture centerline as a broken line composed of a series of short line segments according to the formula, and approximates the true arc length of the fracture by summing the lengths of these line segments, thus obtaining the fracture length. Wherein, the fracture length... Used to characterize the length of the crack. Any pixel in the crack centerline sequence. Pixel x-coordinate Pixel x-coordinate Pixel ordinate and pixels ordinate It can be obtained from the sequence of fracture centerlines.

[0096] S305: Based on the binary crack segmentation map and the preset spatial grid size, the distribution density of crack pixels is statistically analyzed to obtain the crack distribution density.

[0097] Specifically, based on the binary crack segmentation image, the image is divided into regular grids according to a preset spatial grid size. The proportion of crack pixels within each grid is calculated to obtain the crack distribution density. Distribution density statistics are obtained by dividing the image into regular grids, calculating the proportion of crack pixels within each grid, and averaging the proportions of crack pixels across all grids. The preset spatial grid size is an analysis scale set according to image resolution and engineering requirements, typically 10×10 pixels or larger. It can also be set according to actual work needs; this embodiment does not further limit the setting of this preset spatial grid size.

[0098] S306. Based on the crack width value sequence, crack length and crack distribution density, the basic geometric parameters of the crack are obtained.

[0099] Specifically, the terminal splices together the sequence of crack width values, crack length, and crack distribution density to obtain the basic geometric parameters of the crack.

[0100] This embodiment provides an intelligent control method for rock mass grouting process based on big data. Through precise centerline extraction, profile analysis, width measurement, and density statistics, it obtains the basic geometric parameters of fractures, including fracture width value sequences, fracture lengths, and fracture distribution densities. This ensures the accuracy and reliability of geometric feature extraction, providing solid geometric input data for subsequent intelligent decision-making.

[0101] In one embodiment, based on a comprehensive fracture feature vector and a preset threshold mapping rule set, the state of the grouting location is assessed to obtain fracture early warning data, including:

[0102] S401, based on the threshold mapping rule set, maps the comprehensive crack feature vector to the corresponding matching threshold to obtain the matching threshold vector.

[0103] Specifically, the terminal maps each data item in the comprehensive fracture feature vector to a corresponding matching threshold according to the threshold mapping rule set. The terminal then integrates the matching thresholds of each data item to obtain a matching threshold vector. For example, the comprehensive fracture feature vector can take the form of a comprehensive fracture feature vector. For any one of the data The corresponding matching threshold calculation formula can be found in the threshold mapping rule set. ,in It is data It is the matching threshold. It is data The lower threshold, It is data The upper limit threshold can be obtained from the threshold mapping rule set, where the data... This can be any of the following: fracture foundation geometric parameters, fracture morphology complexity, real-time grout performance parameters, or real-time rock mass response data. The terminal concatenates the matching thresholds of each data point in the comprehensive fracture feature vector to obtain a matching threshold vector, which can be in the form of... .

[0104] S402, based on the matching threshold vector and the preset state level rule set, comprehensively evaluates the matching threshold to obtain the state level confidence distribution.

[0105] Specifically, the terminal employs an information function theory algorithm for multi-source evidence fusion. This algorithm can handle uncertain information and match the threshold vector. Each matching threshold in It is considered an independent source of evidence. Firstly, it is based on a pre-defined set of state level rules. ,in These represent three status levels: "Normal," "Warning," and "Action." For the... Each matching threshold is used to calculate its basic probability allocation using the following formula. : This indicates that the evidence supports the "normal" state to a certain degree. The remaining uncertainty The alert and action states were evenly distributed. For any state... The confidence value after state fusion is calculated using the following formula. , , ,in, Yes, it is possible. In any of the states, It is the matching threshold vector The total number of matching thresholds. The terminal integrates the confidence values ​​of each state to obtain the state level confidence distribution. For example, the preset state level rule set can be set according to actual work, and this embodiment does not further limit the setting of this preset state level rule set.

[0106] S403, based on the confidence distribution of state level and the rule set of state level, the warning level value is obtained.

[0107] Specifically, the terminal determines the maximum confidence value from the state level confidence distribution, and sets the state corresponding to the maximum confidence value as the warning level value according to the state level rule set.

[0108] S404: Based on the warning level value, the confidence distribution of the state level is normalized to obtain the state confidence; and based on the warning level value and the state confidence, crack warning data is generated.

[0109] Specifically, the terminal uses the following formula based on the warning level value and the confidence distribution of the status level: The state confidence level is calculated, and the state confidence level and the early warning level value are concatenated to obtain the crack early warning data. Among them, It is the state confidence. It is the confidence level value corresponding to the warning level value. It is any confidence value in the state level confidence distribution.

[0110] This embodiment provides an intelligent control method for rock mass grouting process based on big data. Through threshold mapping, evidence fusion, graded decision-making, and confidence quantification, it obtains fracture early warning data including early warning level values ​​and state confidence levels. This achieves the conversion from feature vectors to early warning data, providing accurate risk assessment and reliable state judgment, thus improving the safety and accuracy of the grouting process.

[0111] In one embodiment, an intelligent control parameter vector is generated based on fracture early warning data, a comprehensive fracture feature vector, and a preset historical case library, including:

[0112] S501. For each historical grouting case in the historical case database, based on crack early warning data, comprehensive crack feature vector and effect score, quantify the similarity between historical grouting cases and grouting locations to obtain the similarity of each historical grouting case; and select the corresponding historical grouting cases whose similarity meets the preset similarity threshold to form a set of similar cases.

[0113] Specifically, for each historical grouting case in the historical case database, the terminal uses the following formula: The similarity of each historical grouting case is calculated, and historical grouting cases whose similarity meets a preset similarity threshold are selected to form a set of similar cases. Wherein, is... Any historical grouting case in the historical case library. This is a historical grouting case. The similarity. It is either the fracture early warning data or any one of the fracture early warning data. It can be any one of the following: fracture foundation geometric parameters, fracture morphology complexity, real-time grout performance parameters, real-time rock mass response data, early warning level value, or state confidence level. It is the total number of data included in the fracture early warning data and the comprehensive fracture feature vector. This is the data for the current grouting location. . This is a historical grouting case. Data . Data from the historical case library The standard deviation. This is a historical grouting case. Performance rating. This is the maximum performance score in the historical case library. The preset similarity threshold can be set according to actual work needs, with a default setting of 0.6. This embodiment does not further limit the setting of this preset similarity threshold.

[0114] S502, for each historical grouting case in the similar case set, calculates the accurate similarity between the historical grouting case and the grouting location based on preset feature weight parameters, crack early warning data, comprehensive crack feature vector and effect score.

[0115] Specifically, the preset feature weight parameters include a weighted combination of the parameters in the fracture early warning data and the comprehensive fracture feature vector. These weights are obtained by analyzing the influence of each parameter on the grouting effect in the historical case library using a gradient boosting decision tree algorithm. The terminal uses the following formula: The accurate similarity of each historical grouting case in the similar case set was calculated. This is a historical grouting case. The similarity. It is either the fracture early warning data or any one of the fracture early warning data. It can be any one of the following: fracture foundation geometric parameters, fracture morphology complexity, real-time grout performance parameters, real-time rock mass response data, early warning level value, or state confidence level. It is the total number of data included in the fracture early warning data and the comprehensive fracture feature vector. This is the data for the current grouting location. . This is a historical grouting case. Data . It is data The weighted combination weights can be obtained based on preset feature weight parameters. Data from the historical case library The standard deviation. This is a historical grouting case. Performance rating. It is the maximum performance score in the historical case library.

[0116] S503: From the set of similar cases, select the historical grouting case with the highest accurate similarity as the optimal similar case.

[0117] Specifically, the terminal selects the historical grouting case with the highest accurate similarity from the set of similar cases and determines it as the optimal similar case for the current grouting location.

[0118] S504 generates a parameter optimization solution set based on the grouting adjustment parameters, real-time grout performance parameters, real-time rock mass response data, and preset operating constraints of the best similar case; the parameter optimization solution set includes the solution vectors of each grouting adjustment parameter.

[0119] Specifically, the terminal employs a non-dominated sorting genetic algorithm with an elitist strategy to perform multi-objective optimization of the grouting adjustment parameters for the optimal similar cases. The optimization model focuses on grouting pressure among the grouting adjustment parameters. ,flow slurry density Decision variables Based on grouting quality ,time and cost The objective function must satisfy operational constraints determined by real-time slurry performance parameters and real-time rock mass response data, including at least the following: ,in, and It is based on the slurry density in the real-time slurry performance parameters. The lower and upper limits of safe pressure are obtained from dynamic calculations of pipeline friction characteristics; , It is the acoustic emission signal in the real-time rock mass response data. This is the critical threshold for the acoustic emission signal. The algorithm first uses the parameters of the optimal similar case. An initial population is generated based on this, and offspring are produced through genetic operations such as selection, crossover, and mutation. Non-dominated sorting and crowding calculation are used to maintain the diversity and convergence of the solution set, ultimately outputting a Pareto optimal solution set. Each of them Each of these is a solution vector for grouting adjustment parameters, representing the optimal trade-offs between the three objectives of quality, time, and cost in different senses.

[0120] S505, based on crack early warning data and comprehensive crack feature vector, selects the grouting adjustment parameter solution vector from the parameter optimization solution set to obtain the optimal grouting parameter combination.

[0121] Specifically, the terminal selects the grouting adjustment parameter solution vector from the parameter optimization solution set based on crack early warning data and comprehensive crack feature vectors, thus obtaining the optimal grouting parameter combination. Parameter selection is achieved by determining the final parameter combination from the optimization solution set through a decision algorithm, taking into account the impact of early warning level on parameter preferences. The optimal grouting parameter combination balances multiple objectives such as quality, efficiency, and economy.

[0122] S506 generates an intelligent control parameter vector based on the optimal combination of grouting parameters.

[0123] Specifically, the terminal standardizes and formats the parameters included in the optimal grouting parameter combination to generate an intelligent control parameter vector that can be directly sent to the grouting execution machine.

[0124] This embodiment provides an intelligent control method for rock mass grouting process based on big data. By introducing a weighted case similarity retrieval mechanism, it ensures the accurate reuse of historical experience. Furthermore, it employs a multi-objective optimization algorithm to systematically generate an optimal parameter set that balances quality, efficiency, and economy while meeting real-time engineering constraints. Combining this with fracture early warning levels, a decision-making method is used to select the grouting parameters that best meet the current working conditions from the optimal parameter set, resulting in the optimal grouting parameter combination. This reduces reliance on single expert experience, enhances the systematic, scientific, and adaptable nature of grouting control, and improves the safety and accuracy of the grouting process.

[0125] In one embodiment, based on crack early warning data and comprehensive crack feature vectors, a solution vector for grouting adjustment parameters is selected from the parameter optimization solution set to obtain the optimal combination of grouting parameters, including:

[0126] S601, based on the parameter optimization solution set and the comprehensive fracture feature vector, evaluates the solution vector of each grouting adjustment parameter to obtain the evaluation index value of each grouting adjustment parameter; the evaluation index value includes the calculation quality index value, time index value and cost index value.

[0127] Specifically, the terminal uses a multi-objective evaluation algorithm to optimize the parameter solution set. Each solution vector in the algorithm is evaluated. For any grouting adjustment parameter, the solution vector is calculated. Use the following formulas respectively: , , Calculate its evaluation value on the three objective functions. Among them, It is the solution vector for any grouting adjustment parameter. It is the solution vector for grouting adjustment parameters. The quality index values. It is the weight of the crack closure rate. This is the predicted fracture closure rate, calculated using a regression model built from historical data. The model is in the form of... ,in These are the regression coefficients learned from historical grouting data. It is a stability weight. It is the predicted value of rock mass stability enhancement, and the formula is: ,in These are the rock mass mechanical parameters determined based on the current rock mass type and geological conditions. It is the solution vector for grouting adjustment parameters. The time index value. This is the predicted grouting time value, obtained through... Received, among which The total volume of the fracture is estimated based on the basic geometric parameters of the fracture in the comprehensive fracture feature vector. This is the preset equipment preparation time. These are preset cleaning and maintenance times, which can be set according to actual work requirements. It is the solution vector for grouting adjustment parameters. Cost indicator values. It is the cost of materials, expressed by the formula. Calculation, where The preset unit price is the price per unit mass of slurry. It is the cost of labor, expressed by the formula. Calculation, where The preset number of construction workers, The preset hourly wage rate can be set according to the actual work situation. It is the equipment cost, expressed by the formula. Calculation, where The preset hourly rate for equipment usage can be set according to actual work requirements.

[0128] S602, based on the evaluation index values ​​of each grouting adjustment parameter, extract the maximum and minimum values ​​of the quality index, the maximum and minimum values ​​of the time index, and the maximum and minimum values ​​of the cost index to form a set of parameter evaluation extreme values.

[0129] Specifically, the terminal selects the maximum and minimum values ​​of the quality index, the maximum and minimum values ​​of the time index, and the maximum and minimum values ​​of the cost index from the evaluation index values ​​of each grouting adjustment parameter to form a set of parameter evaluation extreme values.

[0130] S603 generates positive ideal adjustment parameter solution vectors and negative ideal parameter solution vectors based on the parameter evaluation extremum set.

[0131] Specifically, the terminal parameters are evaluated to determine the set of maximum and minimum values, generating positive ideal adjustment parameter solution vectors and negative ideal parameter solution vectors. The forms of the positive ideal adjustment parameter solution vectors and negative ideal parameter solution vectors are as follows: , .in, It is the solution vector of the positive ideal adjustment parameters. It is the maximum value of the quality indicator. It is the minimum value of the time index. It is the minimum value of the cost indicator. It is the solution vector of the negative ideal parameters. It is the minimum value of the quality indicator. It is the maximum value of the time index. It is the maximum value of the cost indicator.

[0132] S604: For each grouting adjustment parameter solution vector in the parameter optimization solution set, based on the crack early warning data and the preset parameter adjustment rule set, the difference between the evaluation index value and the positive ideal adjustment parameter solution vector is quantified to obtain the positive solution distance of each grouting adjustment parameter solution vector; the difference between the evaluation index value and the negative ideal parameter solution vector is quantified to obtain the negative solution distance of each grouting adjustment parameter solution vector.

[0133] Specifically, the preset parameter adjustment rule set is used to characterize the mapping relationship between the weighted weights of each parameter in the evaluation index value and the evaluation index value, in the form of: ,in It is a preset set of parameter adjustment rules. This is the warning level value for the current grouting location, which can correspond to three states—"Normal," "Warning," and "Action"—from a preset set of status level rules. . It is the weight used to calculate the quality index value in the evaluation index value. It is the weight of the time index value in the evaluation index value. This refers to the weight of the cost index value among the evaluation index values. Based on this weight, the terminal calculates the weighted Euclidean distance between the evaluation index value of each grouting adjustment parameter solution vector and the positive and negative ideal parameter solution vectors, thus obtaining the positive and negative solution distances for each grouting adjustment parameter solution vector.

[0134] S605, for each grouting adjustment parameter solution vector in the parameter optimization solution set, the relative proximity of the grouting adjustment parameter solution vector is calculated using the following formula based on the positive and negative solution distances:

[0135]

[0136] in, It is the first The relative closeness of the solution vector for each grouting adjustment parameter. It is the first The positive solution distance of the vector obtained by solving for each grouting adjustment parameter. It is the first The distance of the negative solution vector for each grouting adjustment parameter solution vector.

[0137] Specifically, the terminal calculates the relative closeness of the solution vectors for each grouting adjustment parameter according to the formula. A larger relative closeness value indicates that the solution is closer to the positive ideal solution and farther from the negative ideal solution. The positive solution distance of the grouting adjustment parameter solution vector and the Negative solution distance of the solution vector for each grouting adjustment parameter It can be obtained from S604.

[0138] S606. From the parameter optimization solution set, select the grouting adjustment parameter solution vector corresponding to the maximum relative proximity as the optimal grouting parameter combination.

[0139] Specifically, the terminal selects the grouting adjustment parameter solution vector corresponding to the maximum relative proximity from the parameter optimization solution set as the optimal grouting parameter combination.

[0140] This embodiment provides an intelligent control method for rock mass grouting process based on big data. It evaluates the quality, time, and cost performance of each solution in the Pareto solution set; determines the extreme value range of each index as a decision benchmark; dynamically constructs an ideal reference solution based on real-time early warning levels; quantifies the relative closeness of each solution to the ideal solution; and selects the solution with the best overall performance as the implementation plan. This effectively balances the three major objectives of grouting quality, engineering efficiency, and economic cost, ensuring the scientific nature and adaptability of parameter selection under complex working conditions, and improving the overall benefits and decision reliability of grouting projects.

[0141] The aforementioned intelligent control method for rock grouting based on big data acquires real-time image data, real-time grout performance parameters, and real-time rock mass response data at the grouting location. Based on these data, the characteristics of the fractures at the grouting location are analyzed to obtain a comprehensive fracture feature vector. This vector includes at least the fracture's basic geometric parameters, fracture morphological complexity, real-time grout performance parameters, and real-time rock mass response data. The grouting location is then assessed based on the comprehensive fracture feature vector and a preset threshold mapping rule set to obtain fracture early warning data. This data includes an early warning level and a state confidence level. An intelligent control parameter vector is generated based on the early warning data, the comprehensive fracture feature vector, and a preset historical case library. This vector indicates the adjustment of grouting parameters. The historical case library includes fracture early warning data, the comprehensive fracture feature vector, grouting adjustment parameters, and effect scores for historical grouting cases. It breaks through the contradiction between the dynamic changes of cracks and the static setting of parameters in traditional grouting. By replacing manual observation with real-time image recognition, replacing fixed thresholds with adaptive thresholds, and optimizing experience-based reliance through case reasoning, it improves the safety and accuracy of the grouting process.

[0142] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0143] Based on the same inventive concept, this application also provides a big data-based intelligent control system for rock grouting process, used to implement the aforementioned big data-based intelligent control method for rock grouting process. The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations of one or more big data-based intelligent control system embodiments provided below can be found in the limitations of the big data-based intelligent control method for rock grouting process described above, and will not be repeated here.

[0144] In an exemplary embodiment, as shown in Figure 2, a big data-based intelligent control system 200 for rock mass grouting process is provided, comprising:

[0145] The data acquisition module 201 is used to acquire real-time image data, real-time grout performance parameters, and real-time rock mass response data of the grouting location;

[0146] Data analysis module 202 is used to analyze the characteristics of fractures at the grouting location based on real-time image data, real-time grout performance parameters, and real-time rock mass response data, and obtain a comprehensive fracture feature vector; the comprehensive fracture feature vector includes at least the fracture basic geometric parameters, fracture morphological complexity, real-time grout performance parameters, and real-time rock mass response data;

[0147] The early warning calculation module 203 is used to perform a status assessment of the grouting location based on the comprehensive fracture feature vector and a preset threshold mapping rule set to obtain fracture early warning data; the fracture early warning data includes early warning level value and status confidence level;

[0148] The regulation generation module 204 is used to generate an intelligent regulation parameter vector based on crack early warning data, comprehensive crack feature vector, and a preset historical case library. The intelligent regulation parameter vector is used to indicate the adjustment of grouting parameters. The historical case library includes crack early warning data, comprehensive crack feature vector, grouting adjustment parameters, and effect scores of historical grouting cases.

[0149] Furthermore, the data analysis module includes:

[0150] The denoising processing unit is used to denoise the real-time image data based on the real-time image data to obtain denoised image data.

[0151] The edge enhancement unit is used to perform edge enhancement processing on the denoised image data based on real-time image data to obtain preprocessed image data;

[0152] The background segmentation unit is used to segment the preprocessed image data based on a preset crack segmentation model to obtain a binary crack segmentation map; in the binary crack segmentation map, pixels with a value of 1 represent cracks and pixels with a value of 0 represent the background.

[0153] The fracture feature calculation unit is used to extract the geometric features of fractures based on the binary fracture segmentation map, and obtain the basic geometric parameters of fractures. The basic geometric parameters of fractures include the fracture width value sequence, fracture length, and fracture distribution density.

[0154] The complexity calculation unit is used to quantify the morphological complexity of a fracture based on its basic geometric parameters, and to obtain the fracture morphological complexity.

[0155] The vector generation unit is used to obtain a comprehensive fracture feature vector based on the fracture basic geometric parameters, fracture morphological complexity, real-time slurry performance parameters, and real-time rock mass response data.

[0156] Furthermore, the fracture feature calculation unit is also used for:

[0157] Based on the binary fracture segmentation map, the pixel coordinates of the fracture centerline are extracted to obtain the fracture centerline sequence;

[0158] Based on the fracture centerline sequence, the cross-sections perpendicular to the centerline are extracted to obtain the normal cross-section set;

[0159] Based on the set of normal profiles, the position of the edge points of each profile is identified to obtain a sequence of crack width values;

[0160] Based on the fracture centerline sequence, the fracture length is calculated using the following formula:

[0161]

[0162] in, It is the crack length. It is the total number of pixels in the crack centerline sequence. It is any pixel in the sequence of crack centerlines. It is a pixel. x-coordinate It is a pixel. x-coordinate It is a pixel. The ordinate, It is a pixel. The ordinate;

[0163] Based on the binary crack segmentation map and the preset spatial grid size, the distribution density of crack pixels is statistically analyzed to obtain the crack distribution density.

[0164] Based on the crack width value sequence, crack length, and crack distribution density, the basic geometric parameters of the crack are obtained.

[0165] Furthermore, the early warning calculation module is also used for:

[0166] Based on the threshold mapping rule set, the comprehensive crack feature vector is mapped to the corresponding matching threshold to obtain the matching threshold vector;

[0167] Based on the matching threshold vector and the preset state level rule set, the matching threshold is comprehensively evaluated to obtain the state level confidence distribution;

[0168] The warning level value is obtained based on the confidence distribution of the state level and the rule set of the state level;

[0169] Based on the warning level value, the confidence distribution of the state level is normalized to obtain the state confidence; and based on the warning level value and the state confidence, crack warning data is generated.

[0170] Furthermore, the regulation generation module includes:

[0171] The first similar case determination unit is used to quantify the similarity between historical grouting cases and grouting locations for each historical grouting case in the historical case database, based on crack early warning data, comprehensive crack feature vectors, and effect scores, to obtain the similarity of each historical grouting case; and select the corresponding historical grouting cases whose similarity meets the preset similarity threshold to form a similar case set.

[0172] The second similarity case determination unit is used to calculate the accurate similarity between the historical grouting case and the grouting location for each historical grouting case in the similarity case set, based on preset feature weight parameters, crack early warning data, comprehensive crack feature vector, and effect score.

[0173] The optimal similar case determination unit is used to select the corresponding historical grouting case with the maximum accurate similarity from the set of similar cases as the optimal similar case;

[0174] The parameter optimization solution set determination unit is used to generate a parameter optimization solution set based on the grouting adjustment parameters of the best similar case, real-time grout performance parameters, real-time rock mass response data, and preset operating constraints; the parameter optimization solution set includes the solution vectors of each grouting adjustment parameter;

[0175] The optimal grouting parameter combination determination unit is used to select the grouting adjustment parameter solution vector from the parameter optimization solution set based on crack early warning data and comprehensive crack feature vector to obtain the optimal grouting parameter combination;

[0176] The intelligent control parameter vector generation unit is used to generate an intelligent control parameter vector based on the optimal combination of grouting parameters.

[0177] Furthermore, the optimal grouting parameter combination determination unit is also used for:

[0178] Based on the parameter optimization solution set and the comprehensive fracture feature vector, the solution vectors of each grouting adjustment parameter are evaluated to obtain the evaluation index values ​​of each grouting adjustment parameter; the evaluation index values ​​include the calculation quality index value, the time index value, and the cost index value.

[0179] Based on the evaluation index values ​​of each grouting adjustment parameter, the maximum and minimum values ​​of the quality index, the maximum and minimum values ​​of the time index, and the maximum and minimum values ​​of the cost index are extracted to form a set of parameter evaluation extreme values;

[0180] Based on the set of extreme values ​​of parameter evaluation, generate positive ideal adjustment parameter solution vectors and negative ideal parameter solution vectors;

[0181] For each grouting adjustment parameter solution vector in the parameter optimization solution set, based on crack early warning data and a preset parameter adjustment rule set, the difference between the evaluation index value and the positive ideal adjustment parameter solution vector is quantified to obtain the positive solution distance of each grouting adjustment parameter solution vector; the difference between the evaluation index value and the negative ideal parameter solution vector is quantified to obtain the negative solution distance of each grouting adjustment parameter solution vector.

[0182] For each solution vector of the grouting adjustment parameters in the parameter optimization solution set, the relative proximity of the solution vectors is calculated using the following formula, based on the positive and negative solution distances:

[0183]

[0184] in, It is the first The relative closeness of the solution vector for each grouting adjustment parameter. It is the first The positive solution distance of the vector obtained by solving for each grouting adjustment parameter. It is the first The distance between the negative solutions of the grouting adjustment parameters solution vector;

[0185] From the parameter optimization solution set, select the grouting adjustment parameter solution vector corresponding to the maximum relative closeness as the optimal grouting parameter combination.

[0186] In one embodiment, as shown in FIG3, a computer device is provided, including:

[0187] At least one processor 301, and a memory 302 communicatively connected to at least one of the processors 301: the memory stores application code that can be executed by at least one of the processors, the application code being executed by at least one of the processors to enable at least one of the processors to execute the intelligent control method for rock mass grouting process based on big data as described above.

[0188] Computer equipment may also include: sensor 303.

[0189] The processor 301, memory 302 and sensor 303 can be connected via a bus or other means, with the bus being an example in the figure.

[0190] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0191] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and 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 modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0192] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A method for intelligent control of rock mass grouting process based on big data, characterized in that, The method includes: acquiring real-time image data, real-time grout performance parameters, and real-time rock mass response data of the grouting location; analyzing the characteristics of the fractures at the grouting location based on the real-time image data, the real-time grout performance parameters, and the real-time rock mass response data to obtain a comprehensive fracture feature vector; the comprehensive fracture feature vector includes at least the fracture basic geometric parameters, fracture morphological complexity, the real-time grout performance parameters, and the real-time rock mass response data; performing a state assessment of the grouting location based on the comprehensive fracture feature vector and a preset threshold mapping rule set to obtain fracture early warning data; the fracture early warning data includes an early warning level value and a state confidence level; generating an intelligent control parameter vector based on the fracture early warning data, the comprehensive fracture feature vector, and a preset historical case library; the intelligent control parameter vector is used to indicate the adjustment of grouting parameters; the historical case library includes the fracture early warning data, the comprehensive fracture feature vector, grouting adjustment parameters, and effect scores of historical grouting cases.

2. The method according to claim 1, characterized in that, The method of analyzing the characteristics of fractures at the grouting location based on the real-time image data, the real-time grout performance parameters, and the real-time rock mass response data to obtain a comprehensive fracture feature vector includes: denoising the real-time image data to obtain denoised image data; performing edge enhancement on the denoised image data to obtain preprocessed image data; segmenting the preprocessed image data based on a preset fracture segmentation model to obtain a binary fracture segmentation map; in the binary fracture segmentation map, pixels with a value of 1 represent fractures, and pixels with a value of 0 represent the background; extracting the geometric features of the fractures based on the binary fracture segmentation map to obtain the basic geometric parameters of the fractures; the basic geometric parameters of the fractures include a fracture width value sequence, fracture length, and fracture distribution density; quantifying the morphological complexity of the fractures based on the basic geometric parameters of the fractures to obtain the fracture morphological complexity; and obtaining the comprehensive fracture feature vector based on the basic geometric parameters of the fractures, the fracture morphological complexity, the real-time grout performance parameters, and the real-time rock mass response data.

3. The method according to claim 2, characterized in that, The process of extracting geometric features of the crack based on the binarized crack segmentation image to obtain the basic geometric parameters of the crack includes: extracting the pixel coordinates of the centerline of the crack based on the binary crack segmentation image to obtain a crack centerline sequence; extracting cross-sections perpendicular to the centerline based on the crack centerline sequence to obtain a set of normal cross-sections; identifying the edge point positions of each cross-section based on the set of normal cross-sections to obtain a crack width value sequence; and calculating the crack length based on the crack centerline sequence using the following formula: in, It is the crack length. It is the total number of pixels in the crack centerline sequence. It is any pixel in the sequence of crack centerlines. It is a pixel. x-coordinate It is a pixel. x-coordinate It is a pixel. The ordinate, It is a pixel. The vertical coordinate is calculated; based on the binary crack segmentation map and the preset spatial grid size, the distribution density of the pixels of the crack is calculated to obtain the crack distribution density; based on the crack width value sequence, the crack length and the crack distribution density, the basic geometric parameters of the crack are obtained.

4. The method according to claim 1, characterized in that, The step of assessing the state of the grouting location based on the comprehensive fracture feature vector and a preset threshold mapping rule set to obtain fracture early warning data includes: mapping the comprehensive fracture feature vector to a corresponding matching threshold based on the threshold mapping rule set to obtain a matching threshold vector; comprehensively evaluating the matching threshold based on the matching threshold vector and a preset state level rule set to obtain a state level confidence distribution; obtaining the early warning level value based on the state level confidence distribution and the state level rule set; normalizing the state level confidence distribution based on the early warning level value to obtain the state confidence; and generating fracture early warning data based on the early warning level value and the state confidence.

5. The method according to claim 1, characterized in that, The step of generating an intelligent control parameter vector based on the crack early warning data, the comprehensive crack feature vector, and a preset historical case library includes: for each historical grouting case in the historical case library, quantifying the similarity between the historical grouting case and the grouting location based on the crack early warning data, the comprehensive crack feature vector, and the effect score to obtain the similarity of each historical grouting case; selecting the historical grouting cases whose similarity meets a preset similarity threshold to form a similar case set; and for each historical grouting case in the similar case set, based on preset feature weight parameters, the crack early warning data, the comprehensive crack feature vector, and the effect score... Calculate the accurate similarity between the historical grouting cases and the grouting locations; select the historical grouting case with the highest accurate similarity from the set of similar cases as the optimal similar case; generate a parameter optimization solution set based on the grouting adjustment parameters, real-time grout performance parameters, real-time rock mass response data, and preset operating constraints of the optimal similar case; the parameter optimization solution set includes solution vectors for each grouting adjustment parameter; select the solution vectors for the grouting adjustment parameters from the parameter optimization solution set based on the fracture early warning data and the comprehensive fracture feature vector to obtain the optimal grouting parameter combination; generate the intelligent control parameter vector based on the optimal grouting parameter combination.

6. The method according to claim 5, characterized in that, The step of selecting the grouting adjustment parameter solution vector from the parameter optimization solution set based on the crack early warning data and the comprehensive crack feature vector to obtain the optimal grouting parameter combination includes: evaluating each grouting adjustment parameter solution vector based on the parameter optimization solution set and the comprehensive crack feature vector to obtain evaluation index values ​​for each grouting adjustment parameter; the evaluation index values ​​include calculated quality index values, time index values, and cost index values; based on the evaluation index values ​​of each grouting adjustment parameter, extracting the maximum and minimum values ​​of the quality index values, the maximum and minimum values ​​of the time index values, and the maximum and minimum values ​​of the cost index values ​​to form a parameter evaluation extreme value set; and based on the parameter evaluation... The set of extreme values ​​is used to generate positive and negative ideal parameter solution vectors. For each grouting adjustment parameter solution vector in the parameter optimization solution set, based on the crack early warning data and the preset parameter adjustment rule set, the difference between the evaluation index value and the positive ideal parameter solution vector is quantified to obtain the positive solution distance of each grouting adjustment parameter solution vector. The difference between the evaluation index value and the negative ideal parameter solution vector is also quantified to obtain the negative solution distance of each grouting adjustment parameter solution vector. For each grouting adjustment parameter solution vector in the parameter optimization solution set, based on the positive and negative solution distances, the relative closeness of the grouting adjustment parameter solution vector is calculated using the following formula: in, It is the first The relative closeness of the solution vector for each grouting adjustment parameter. It is the first The positive solution distance of the vector obtained by solving for each grouting adjustment parameter. It is the first The negative solution distance of each grouting adjustment parameter solution vector; from the parameter optimization solution set, the grouting adjustment parameter solution vector corresponding to the maximum relative proximity is selected as the optimal grouting parameter combination.

7. A smart control system for rock mass grouting process based on big data, characterized in that, The system includes: a data acquisition module for acquiring real-time image data, real-time grout performance parameters, and real-time rock mass response data of the grouting location; a data analysis module for analyzing the characteristics of the fractures at the grouting location based on the real-time image data, the real-time grout performance parameters, and the real-time rock mass response data, to obtain a comprehensive fracture feature vector; the comprehensive fracture feature vector includes at least the fracture basic geometric parameters, fracture morphological complexity, the real-time grout performance parameters, and the real-time rock mass response data; an early warning calculation module for performing a state assessment of the grouting location based on the comprehensive fracture feature vector and a preset threshold mapping rule set, to obtain fracture early warning data; the fracture early warning data includes an early warning level value and a state confidence level; and a control generation module for generating an intelligent control parameter vector based on the fracture early warning data, the comprehensive fracture feature vector, and a preset historical case library; the intelligent control parameter vector is used to indicate the adjustment of grouting parameters; the historical case library includes the fracture early warning data, the comprehensive fracture feature vector, grouting adjustment parameters, and effect scores of historical grouting cases.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images