Canyon deposit curtain grouting construction control method based on SVM and PID

Abstract

The present application relates to the field of geotechnical engineering construction technology, and in particular to a canyon deposit curtain grouting construction control method based on SVM and PID, which includes: obtaining construction data including formation permeability, grouting pressure, slurry consistency, grouting flow rate and slurry diffusion range; constructing a three-dimensional geological model of the canyon deposit, and associating the above parameters with corresponding positions in the model to generate a construction feature vector with geological attributes, and inputting it into a pre-trained SVM model to predict the geological response trend, and accordingly correct the control parameters of the PID control model, and then dynamically adjust the operating parameters of the grouting equipment, so that the PID control parameters and key grouting parameters can be continuously optimized. Thus, the problem that it is difficult to achieve adaptive control of complex and changeable geological conditions in the existing technology, resulting in inaccurate grouting parameter settings and unstable construction quality, is solved.

Description

Technical Field

[0001] The present application relates to the field of geotechnical engineering construction technology, and in particular to a canyon deposit curtain grouting construction control method based on SVM and PID. Background Art

[0002] Canyon deposit curtain grouting is a critical component of water conservancy and hydropower infrastructure construction. Its purpose is to form a continuous, impermeable curtain within the canyon deposit to reduce groundwater leakage and improve project stability and safety. However, canyon deposits are often characterized by extremely complex geological conditions, with numerous adverse geological phenomena such as faults, fissures, and caves, and formation parameters exhibit significant dynamic variations.

[0003] Traditional curtain grouting construction in canyon deposits typically uses empirical formulas or static control methods to determine key parameters such as grouting pressure, flow rate, and slurry diffusion range. These methods struggle to adapt to the complex and changing geological conditions and dynamic changes in parameters, and can easily lead to poor grouting results, such as uneven curtain formation and substandard anti-seepage performance. Furthermore, the lack of effective dynamic control methods can increase construction risks, such as slurry leakage and ground heave, while also wasting resources and increasing construction costs. Summary of the Invention

[0004] This application provides a canyon deposit curtain grouting construction control method based on SVM and PID to solve the problems in the existing technology that it is difficult to achieve adaptive control of complex and changeable geological conditions, resulting in inaccurate grouting parameter settings and unstable construction quality.

[0005] The first aspect of the present application provides a canyon deposit curtain grouting construction control method based on SVM and PID, comprising the following steps: acquiring construction data, wherein the construction data includes formation permeability coefficient, grouting pressure, slurry consistency, grouting flow rate and slurry diffusion range; constructing a three-dimensional geological model of the canyon deposit, associating the formation permeability coefficient, grouting pressure, slurry consistency, grouting flow rate and slurry diffusion range with corresponding positions in the three-dimensional geological model, and generating a construction feature vector with geological attributes; inputting the construction feature vector into a pre-trained SVM model to predict the geological response trend, and correcting the control parameters of the PID control model according to the geological response trend, wherein the control parameters of the PID control model include proportional coefficient, integral coefficient and differential coefficient; using the corrected PID control model to dynamically regulate the operating parameters of the grouting equipment, and inputting the grouting effect data after the dynamic regulation into the SVM model again to form a closed-loop feedback mechanism, and continuously optimizing the PID control parameters and the key grouting parameters.

[0006] Optionally, constructing a three-dimensional geological model of a canyon deposit includes: acquiring geological survey data, topographic data, and groundwater level data; dividing the strata of the canyon deposit based on the geological survey data, identifying fault structures in the canyon deposit, and establishing a three-dimensional model of the fault; constructing a topographic surface model of the canyon deposit based on the topographic data; and constructing a groundwater level model using an interpolation method based on the groundwater level data; integrating the three-dimensional fault model, the topographic surface model, and the groundwater level model to obtain a three-dimensional geological model.

[0007] Optionally, a pre-trained SVM model includes: selecting a radial basis kernel function according to the risk condition classification, setting the value range of the penalty factor and the kernel function parameters, and determining the target penalty factor and the kernel function parameters based on grid search and cross-validation; setting a dynamic weight coefficient to increase the weight of abnormal risk condition samples of drilling in pile foundation construction, including but not limited to hole collapse, slurry leakage and drill sticking; using the determined target penalty factor, kernel function parameters and dynamic weight coefficient to perform model training on the training data set to obtain a pre-trained SVM model.

[0008] Optionally, the control parameters of the PID control model are modified according to the geological response trend, wherein the modification formula is:

[0009] K p =K p0 ×(1+α×ΔP);

[0010] T i =T i0 / (1+β×ΔI);

[0011] T d =T d0 ×(1+γ×ΔD);

[0012] Among them, ΔP, ΔI, and ΔD are the trend changes of SVM output, α, β, and γ are adaptive adjustment coefficients, and K p is the proportional coefficient, T i is the integral coefficient, T d is the differential coefficient.

[0013] Optionally, the modified PID control model is used to dynamically regulate the operating parameters of the grouting equipment, including: obtaining the operating status parameters of the grouting equipment, such as the vibration signal, temperature data, pressure fluctuation curve and motor current change rate; performing feature matching on the operating status parameters and historical fault data, establishing an equipment fault prediction model based on the long-short-term memory network, and predicting the future failure probability of the equipment; when the predicted failure probability exceeds the warning threshold, automatically issuing a warning signal, and generating a corresponding emergency treatment plan based on the fault type and severity, and adjusting the control strategy of the PID control model.

[0014] Optionally, the grouting effect data after dynamic regulation is input into the SVM model again to form a closed-loop feedback mechanism, and continuously optimize the PID control parameters and key grouting parameters, including: setting a feedback cycle, and in each feedback cycle, inputting the grouting effect data after dynamic regulation into a pre-trained SVM model, and outputting an evaluation result of the current grouting status; comparing and analyzing the grouting effect data corresponding to the evaluation result with the preset grouting effect index to obtain a comparison result between the actual grouting effect and the expected one; adjusting the input parameters of the SVM model according to the comparison result, so that the model adapts to the ever-changing grouting conditions, and adjusting the control strategy of the PID control model at the same time.

[0015] Optionally, the key grouting parameters include slurry viscosity, gel time, grouting pressure gradient value and grouting rate fluctuation range.

[0016] The second embodiment of the present application provides a canyon deposit curtain grouting construction control system based on SVM and PID, including: an acquisition module for acquiring construction data, wherein the construction data includes formation permeability coefficient, grouting pressure, slurry consistency, grouting flow rate and slurry diffusion range; a construction module for constructing a three-dimensional geological model of the canyon deposit, associating the formation permeability coefficient, grouting pressure, slurry consistency, grouting flow rate and slurry diffusion range with corresponding positions in the three-dimensional geological model, and generating a construction feature vector with geological attributes; a correction module for inputting the construction feature vector into a pre-trained SVM model, predicting the geological response trend, and correcting the control parameters of the PID control model according to the geological response trend, wherein the control parameters of the PID control model include proportional coefficient, integral coefficient and differential coefficient; an optimization module for dynamically regulating the operating parameters of the grouting equipment using the corrected PID control model, and inputting the grouting effect data after the dynamic regulation into the SVM model again to form a closed-loop feedback mechanism, and continuously optimizing the PID control parameters and key grouting parameters.

[0017] The third aspect of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to execute the canyon deposit curtain grouting construction control method based on SVM and PID as described in the above embodiment.

[0018] The fourth embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to perform the canyon deposit curtain grouting construction control method based on SVM and PID as described in the above embodiment.

[0019] Therefore, this application has at least the following beneficial effects:

[0020] This embodiment of the present application correlates construction data with a three-dimensional geological model to generate geological characteristic vectors. The SVM model is then used to accurately predict geological response trends, providing a basis for dynamic modification of PID control model parameters and achieving intelligent adaptation of grouting equipment operating parameters. Grouting effect data is then fed back into the SVM model to form a closed-loop feedback loop, continuously optimizing control parameters and key process indicators. By using the SVM model to predict geological response trends, this embodiment of the present application can keenly capture potential changes in geological conditions during construction, proactively detect potential problems, and provide a scientific and reliable basis for timely and precise adjustment of construction parameters. This fundamentally changes the blindness and lag of traditional empirical judgments and greatly improves the accuracy and reliability of construction decisions. The PID control model dynamically regulates the operating parameters of the grouting equipment, and combined with a closed-loop feedback mechanism, continuously optimizes control parameters and key grouting parameters, achieving refined and intelligent control of the grouting construction process. Grouting parameters such as grouting pressure, flow rate, and slurry ratio can be precisely adjusted based on real-time changes in geological conditions, effectively improving grouting quality and ensuring that the curtain's impermeability and structural strength meet or even exceed design requirements. This reduces quality defects and resource waste caused by improper construction, reducing project costs and safety risks. The equipment failure prediction model, built using a long-short-term memory network, can accurately predict potential equipment failures in advance, issue early warning signals, and quickly generate emergency response plans based on the type and severity of the failure. It also flexibly adjusts PID control strategies. This model is not only applicable to canyon deposit curtain grouting projects, but can also be widely used in other complex deposit reinforcement projects, such as slopes. This solves the existing problems of achieving adaptive control for complex and changing geological conditions, which can lead to inaccurate grouting parameter settings and unstable construction quality.

[0021] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0023] Figure 1 Flowchart of a canyon deposit curtain grouting construction control method based on SVM and PID according to an embodiment of the present application;

[0024] Figure 2 A schematic diagram of the change in the training accuracy of the SVM model provided according to one embodiment of the present application;

[0025] Figure 3A schematic diagram of a PID control parameter adjustment trend provided according to an embodiment of the present application;

[0026] Figure 4 A schematic diagram showing how the probability of equipment failure varies over time according to one embodiment of the present application;

[0027] Figure 5 This is a block diagram of an example control system for curtain grouting of canyon deposits based on SVM and PID according to an embodiment of the present application;

[0028] Figure 6 A schematic diagram of the structure of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0029] The following describes in detail embodiments of the present application, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0030] The following describes the canyon deposit curtain grouting construction control method based on SVM and PID in the embodiment of the present application with reference to the accompanying drawings. In response to the problem of insufficient adaptability of PID control in nonlinear and time-varying geological scenarios mentioned in the above background technology, the present application provides a canyon deposit curtain grouting construction control method based on SVM and PID. In this method, by associating construction data with a three-dimensional geological model to generate a geological characteristic vector, the SVM model is used to accurately predict the geological response trend, providing a basis for the dynamic correction of the PID control model parameters, and realizing the intelligent adaptation of the grouting equipment operating parameters; the grouting effect data is then fed back to the SVM model to form a closed-loop feedback, and the control parameters and key process indicators are continuously optimized. The embodiment of the present application uses the SVM model to predict the geological response trend, which can keenly capture the potential changes in geological conditions during the construction process, detect possible problems in advance, and provide a scientific and reliable basis for the timely and accurate adjustment of construction parameters. It completely changes the blindness and lag of traditional judgment based on experience, and greatly improves the accuracy and reliability of construction decisions. A PID control model dynamically regulates the operating parameters of the grouting equipment, and a closed-loop feedback mechanism is used to continuously optimize control parameters and key grouting parameters, achieving refined and intelligent control of the grouting process. The system can precisely adjust grouting parameters such as grouting pressure, flow rate, and slurry ratio based on real-time changes in geological conditions, effectively improving grouting quality and ensuring that the curtain's impermeability and structural strength meet or exceed design requirements. This reduces quality defects and resource waste caused by improper construction, thereby reducing project costs and safety risks. A device failure prediction model, built using a long-short-term memory network, accurately predicts potential equipment failures in advance, issues early warning signals, and rapidly generates emergency response plans based on the type and severity of the failure. The model also flexibly adjusts the PID control strategy. This model is not only applicable to canyon deposit curtain grouting projects, but can also be widely applied to other complex deposit reinforcement projects, such as slope reinforcement projects. This solves the existing problems of achieving adaptive control for complex and changing geological conditions, which can lead to inaccurate grouting parameter settings and unstable construction quality.

[0031] The following describes the canyon deposit curtain grouting construction control method based on SVM and PID in an embodiment of the present application with reference to the accompanying drawings.

[0032] Specifically, Figure 1 A flow chart of a canyon deposit curtain grouting construction control method based on SVM and PID provided in an embodiment of the present application.

[0033] like Figure 1 As shown in FIG, the canyon accumulation curtain grouting construction control method based on SVM and PID includes the following steps:

[0034] In step S101 , construction data is acquired.

[0035] Among them, the construction data may include formation permeability, grouting pressure, slurry consistency, grouting flow rate and slurry diffusion range.

[0036] It can be understood that the embodiments of the present application provide data support for subsequent model construction and parameter optimization by acquiring construction data.

[0037] In step S102, a three-dimensional geological model of the canyon deposit is constructed, and the formation permeability coefficient, grouting pressure, slurry consistency, grouting flow rate and slurry diffusion range are associated with corresponding positions in the three-dimensional geological model to generate a construction feature vector with geological attributes.

[0038] Among them, the three-dimensional geological model of the canyon deposit is a model that can digitally present the geological information of the canyon deposit in a three-dimensional form.

[0039] It can be understood that in the embodiment of the present application, in the grouting construction of the canyon deposit, the geological inherent attribute data such as the stratum permeability coefficient, the operating parameter data such as the grouting pressure, and the effect index data such as the slurry diffusion range are associated with the three-dimensional geological model to construct a comprehensive cognitive system and clarify the spatial positioning, and then generate a construction feature vector covering geological and construction information to assist the model in learning and prediction.

[0040] In an embodiment of the present application, a three-dimensional geological model of a canyon deposit is constructed, including: obtaining geological survey data, topographic data, and groundwater level data; dividing the strata of the canyon deposit based on the geological survey data, identifying the fault structure in the canyon deposit, and establishing a three-dimensional model of the fault; constructing a topographic surface model of the canyon deposit based on the topographic data; and constructing a groundwater level model using an interpolation method based on the groundwater level data; integrating the three-dimensional fault model, the topographic surface model, and the groundwater level model to obtain a three-dimensional geological model.

[0041] It is understood that the embodiments of this application comprehensively depict the geological conditions of the canyon accumulation body in multiple dimensions by acquiring geological survey data covering basic geological characteristics such as stratum lithology, topographic data reflecting surface undulations, and groundwater level data reflecting the dynamic distribution of groundwater. This constructs a complete characterization system covering geological structure, surface morphology, and hydrological conditions, and reveals the inherent relationships between different geological elements. Layered modeling is then performed, with a three-dimensional fault model based on the geological survey data accurately presenting fault details. A terrain surface model is constructed using topographic data to intuitively display the surface morphology. A groundwater level model is constructed using interpolation to reflect groundwater dynamics, providing key information for the project.

[0042] Specifically, in the Yunling Gorge, a 3D model of the fault was constructed using specialized geological modeling software based on data such as the location, strike, and scale of the fault determined by geological surveys. The software inputs parameters such as the fault's spatial coordinates, dip, and inclination, and uses a 3D modeling algorithm to generate the fault's three-dimensional form. For example, the software inputs data for a detected north-south strike and a 3-meter-wide fault. The resulting 3D model intuitively demonstrates the fault's underground extension and its relationship to the surrounding strata.

[0043] The digital elevation model generated by drone oblique photogrammetry is imported into Geographic Information System software for further processing to generate a terrain surface model. In GIS software, the terrain model can be edited, analyzed, and visualized, and terrain features can be added to make the terrain surface model more accurate and detailed.

[0044] Kriging interpolation was used to process groundwater level monitoring data. Using the water level and coordinates of each monitoring well as input, the groundwater level distribution across the entire canyon was calculated through interpolation. This model then constructed a groundwater level surface model. This model can demonstrate spatial variations in groundwater levels, such as relatively high levels in the center of the canyon and lower levels on the sides.

[0045] Import the constructed 3D fault model, terrain surface model, and groundwater level model into a unified 3D geological modeling platform, such as 3DMine, and integrate them according to their actual spatial relationships. During the integration process, ensure that the various models are accurately connected to form a complete 3D geological model of the canyon accumulation.

[0046] In step S103, the construction feature vector is input into a pre-trained SVM model to predict the geological response trend, and the control parameters of the PID control model are modified according to the geological response trend.

[0047] The control parameters of the PID control model may include a proportional coefficient, an integral coefficient, and a differential coefficient.

[0048] It can be understood that the embodiment of the present application will integrate the characteristic vectors of multi-dimensional construction and geological information into the SVM model, mine the complex correlations of the data to accurately predict the geological response trend; establish a connection between the geological conditions and the PID control model parameters, realize dynamic adjustment of the parameters, and significantly enhance the adaptability and stability of the project.

[0049] In an embodiment of the present application, a pre-trained SVM model includes: selecting a radial basis kernel function according to the risk condition classification, setting the value range of the penalty factor and the kernel function parameters, and determining the target penalty factor and the kernel function parameters based on grid search and cross-validation; setting a dynamic weight coefficient to increase the weight of abnormal risk condition samples of drilling in pile foundation construction, including but not limited to hole collapse, slurry leakage and drill sticking; using the determined target penalty factor, kernel function parameters and dynamic weight coefficient to perform model training on the training data set to obtain a pre-trained SVM model.

[0050] Among them, the radial basis kernel function is a commonly used support vector machine kernel function, and its basic form is ,in is the input vector, γ is the kernel function parameter, Represents a vector The Euclidean distance between them; the penalty factor range is [0.1, 100], and the kernel function parameter range is [0.01, 1].

[0051] It is understandable that the embodiment of the present application optimizes the classification accuracy of the model. Based on the unique geological-construction characteristics and data distribution patterns of different risk conditions, the radial basis kernel function is selected in a targeted manner to give full play to its mapping ability for nonlinear data and accurately capture the boundary differences of working conditions to reduce classification errors. At the same time, grid search and cross-validation are used to determine the optimal combination of penalty factors and kernel function parameters, balance model complexity and classification accuracy, reduce the misclassification rate, and enhance the ability to identify risk conditions with complex geological conditions. Enhance the generalization ability of the model, use cross-validation to comprehensively evaluate the performance of the model on different data subsets, promptly discover and adjust overfitting or underfitting problems, and avoid excessive reliance on training data; set dynamic weight coefficients, increase the sample weights of key risk conditions such as collapse and slurry leakage, solve the problem of data imbalance, and improve the model's generalization and prediction capabilities for various risk conditions, ensuring stable and reliable practical applications.

[0052] Specifically, based on a preliminary analysis of construction data and past experience, the radial basis kernel (RBF) function was selected as the kernel function for the SVM model. The penalty factor C was set to a value range of [0.1, 10], with a step size of 0.1; the kernel function parameter γ was set to a value range of [0.01, 1], with a step size of 0.01. If there are 100 candidate values ​​for C and 100 candidate values ​​for γ, a total of 100 × 100 = 10,000 parameter combinations will be generated.

[0053] A grid search algorithm combined with 5-fold cross validation is used to find the optimal penalty factor and kernel function parameters. The training data set is randomly divided into 5 parts, 4 parts are selected as training sets each time, and 1 part is used as a validation set. For each set of (C, γ) parameter combinations, the SVM model is trained on the training set, and the model performance is evaluated on the validation set, using accuracy as the evaluation indicator. For example, when C=0.5 and γ=0.05, the average accuracy in a 5-fold cross validation is 85%. By traversing all parameter combinations, it is found that when C=1 and γ=0.1, the average accuracy of the model in cross validation is the highest, reaching 90%. Therefore, the target penalty factor C is determined. opt =1, target kernel function parameter γ opt =0.1.

[0054] Considering the serious impact of risk conditions such as hole collapse, slurry leakage, and slurry leakage on construction, we increase the weight of these risk condition samples. We set the weight of normal condition samples to 1, the weight of hole collapse samples to 2, the weight of slurry leakage samples to 2.5, and the weight of slurry leakage samples to 3. This enhances the influence of risk condition samples during model training, allowing the model to better learn the characteristics of risk conditions.

[0055] Using the determined target penalty factor C opt =1, target kernel function parameter γ opt = 0.1 and a dynamic weight coefficient were used to train the SVM model on the entire training dataset. After multiple rounds of iterative training, the model gradually learned the relationship between construction parameters and risk conditions, ultimately resulting in a pre-trained SVM model. During subsequent construction, real-time construction data was fed into this model, enabling it to predict the likelihood of risk conditions such as hole collapse, slurry leakage, and slurry leakage under the current construction conditions, providing a basis for construction personnel to take preventive measures in advance.

[0056] In the embodiment of the present application, the control parameters of the PID control model are modified according to the geological response trend, wherein the modification formula is:

[0057] K p =K p0 ×(1+α×ΔP);

[0058] T i =T i0 / (1+β×ΔI);

[0059] T d =T d0 ×(1+γ×ΔD);

[0060] Among them, ΔP, ΔI, and ΔD are the trend changes of SVM output, α, β, and γ are adaptive adjustment coefficients, and K p is the proportional coefficient, Ti is the integral coefficient, T d is the differential coefficient.

[0061] In step S104, the operating parameters of the grouting equipment are dynamically regulated using the modified PID control model, and the grouting effect data after dynamic regulation is input into the SVM model again to form a closed-loop feedback mechanism to continuously optimize the PID control parameters and key grouting parameters.

[0062] Among them, the key grouting parameters may include slurry viscosity, gel time, grouting pressure gradient value and grouting rate fluctuation range.

[0063] It can be understood that the embodiment of the present application improves the grouting quality and engineering stability by constructing a closed-loop feedback mechanism through the modified PID control model and the SVM model: on the one hand, the modified PID model dynamically controls the grouting pressure, flow rate and other parameters based on real-time geological parameters (such as rock formation permeability, etc.), adapts to complex geological differences, and reduces abnormal slurry loss or penetration problems. The SVM model then optimizes the parameters based on the grouting effect data feedback to improve the grouting uniformity; on the other hand, the modified PID model combines the real-time monitoring data (such as the grouting pipe pressure, etc.) to timely detect and control grouting defects. The SVM model analyzes the parameter correlation according to the defect situation, optimizes the parameters and slurry ratio, and reduces the probability of defects.

[0064] In an embodiment of the present application, a modified PID control model is used to dynamically regulate the operating parameters of the grouting equipment, including: obtaining the operating status parameters of the grouting equipment, such as the vibration signal, temperature data, pressure fluctuation curve, and motor current change rate; performing feature matching on the operating status parameters and historical fault data, establishing an equipment fault prediction model based on a long-short-term memory network, and predicting the future failure probability of the equipment; when the predicted failure probability exceeds the warning threshold, a warning signal is automatically issued, and at the same time, a corresponding emergency treatment plan is generated according to the type and severity of the fault, and the control strategy of the PID control model is adjusted.

[0065] Among them, the warning threshold can be determined according to actual conditions, such as 70%.

[0066] It is understandable that the embodiments of this application integrate multi-dimensional operating status parameters such as grouting equipment vibration, temperature, pressure fluctuations, and motor current change rate, and establish a fault prediction model with the help of long-short-term memory networks and historical fault data feature matching to achieve accurate prediction of equipment failures. Once the predicted failure probability exceeds the warning threshold, an automatic warning is issued and a personalized emergency plan is generated, and the PID control strategy is adjusted simultaneously. This not only ensures equipment operation and maintenance, reducing sudden failure downtime and repair costs, but also stabilizes equipment operating parameters, ensures grouting quality and progress, optimizes operation and maintenance decisions, and reduces full-cycle costs.

[0067] Specifically, various sensors, including acceleration, temperature, and pressure, were used to collect operational parameters of the grouting equipment, including vibration signals, temperature data, pressure fluctuation curves, and motor current change rates. These parameters were aligned in time series and features extracted, which were then used to train a long short-term memory (LSTM) network to construct an equipment failure prediction model. Three warning thresholds were set based on historical equipment failure data and expert experience: yellow (failure probability ≥ 50%), orange (failure probability ≥ 70%), and red (failure probability ≥ 90%). When the predicted failure probability exceeded the threshold, the system automatically issued a corresponding warning signal and implemented emergency response measures based on the warning level, such as reducing grouting pressure, switching to a backup pump, or ceasing operations. Furthermore, the system dynamically adjusted the control strategy of the PID control model, optimizing PID parameters and switching control targets based on the fault type and severity. Through the application of this system, equipment failure prediction accuracy reached 92%, with advance warnings up to 8 hours. Downtime caused by equipment failures was reduced by 65%, repair costs were reduced by 40%, grouting pressure stability was improved by 30%, and the curtain body's anti-seepage effectiveness was enhanced by 25%.

[0068] In an embodiment of the present application, the grouting effect data after dynamic regulation is input into the SVM model again to form a closed-loop feedback mechanism, and the PID control parameters and key grouting parameters are continuously optimized, including: setting a feedback cycle, and in each feedback cycle, inputting the grouting effect data after dynamic regulation into a pre-trained SVM model, and outputting an evaluation result of the current grouting status; comparing and analyzing the grouting effect data corresponding to the evaluation result with the preset grouting effect index to obtain a comparison result between the actual grouting effect and the expected one; adjusting the input parameters of the SVM model according to the comparison result, so that the model adapts to the ever-changing grouting conditions, and adjusting the control strategy of the PID control model at the same time.

[0069] The feedback cycle refers to the time interval for complete information feedback and adjustment, such as 1 hour.

[0070] As can be understood, the present embodiment utilizes a feedback cycle mechanism to regularly input dynamically regulated grouting effect data into a pre-trained SVM model, enabling a comprehensive and accurate assessment of the grouting status and timely identification of subtle issues. By comparing actual results with pre-set indicators, the gap between actual and expected results can be clearly identified, allowing timely measures to ensure quality. Furthermore, this solution can reduce resource waste, lower maintenance costs, improve construction efficiency, and optimize resources and costs. It also enables data-driven decision-making, enhancing the system's adaptive adjustment and intelligent early warning and prevention capabilities.

[0071] According to the canyon deposit curtain grouting construction control method based on SVM and PID proposed in the embodiment of the present application, by associating construction data with a three-dimensional geological model to generate a geological characteristic vector, the SVM model is used to accurately predict the geological response trend, providing a basis for the dynamic correction of the PID control model parameters, and realizing the intelligent adaptation of the grouting equipment operating parameters; the grouting effect data is then fed back to the SVM model to form a closed-loop feedback, and the control parameters and key process indicators are continuously optimized. The embodiment of the present application uses the SVM model to predict the geological response trend, which can keenly capture the potential changes in geological conditions during the construction process, detect possible problems in advance, and provide a scientific and reliable basis for the timely and accurate adjustment of construction parameters. It completely changes the blindness and lag of traditional judgment based on experience, and greatly improves the accuracy and reliability of construction decisions. The PID control model is used to dynamically adjust the operating parameters of the grouting equipment, and the closed-loop feedback mechanism is combined to continuously optimize the control parameters and key grouting parameters, thereby realizing refined and intelligent control of the grouting construction process. It can accurately adjust parameters such as grouting pressure, flow rate, and slurry ratio according to real-time changes in geological conditions, effectively improving the grouting quality and ensuring that the curtain body's impermeability and structural strength meet or even exceed design requirements, reducing quality defects and resource waste caused by improper construction, and reducing project costs and safety risks. The equipment failure prediction model established using the long short-term memory network can accurately predict potential equipment failures in advance, issue early warning signals in a timely manner, and quickly generate corresponding emergency response plans based on the type and severity of the failure. At the same time, it can flexibly adjust the PID control strategy. It is not only suitable for canyon deposit curtain grouting projects, but can also be widely used in other complex deposit reinforcement projects such as slopes. This solves the problem that it is difficult to achieve adaptive control of complex and changeable geological conditions in existing technologies, resulting in inaccurate grouting parameter settings and unstable construction quality.

[0072] The following will use a specific example to illustrate the curtain grouting control method for canyon deposits based on SVM and PID. In the canyon area of ​​a large hydropower project, the topography is complex and the geological conditions are even more challenging. There are multiple fault fracture zones in the area, the strata are diverse in lithology and have extremely different permeability coefficients, and the groundwater level fluctuates frequently, which brings great challenges to the curtain grouting construction. The content is as follows:

[0073] In the early stages of construction, the geological survey team used advanced geological radar, sonic detectors, and other equipment to conduct a detailed survey of the canyon deposits. They obtained a wealth of geological survey data, detailing the lithologic distribution of the strata, the location and orientation of faults, and the development of joints and fissures in the rock mass. At the same time, using high-precision total stations and drone-oblique photogrammetry, the surveying and mapping team obtained precise topographic data for the area, covering detailed features such as the canyon's topographical undulations, slope changes, and gully distribution. The hydrogeological team, through the deployment of multiple groundwater level monitoring wells and the use of advanced water level sensors and data acquisition systems, conducted long-term, continuous monitoring of groundwater level data, gaining an understanding of the dynamic changes in groundwater levels and their relationship to factors such as rainfall and seasons.

[0074] Based on this detailed data, technicians used specialized geological modeling software to meticulously construct a three-dimensional geological model of the canyon's accumulations. Strata were meticulously divided into distinct units based on their lithologic characteristics, age, and mechanical properties. Fault identification involved a comprehensive analysis of geological radar data, core samples, and the regional geological and tectonic context, enabling precise location of multiple faults. Their occurrence, fracture zone width, and impact on surrounding strata were then thoroughly investigated. Using geological modeling algorithms, the resulting three-dimensional fault model clearly depicted the fault's subsurface trajectory, its relationship to surrounding strata, and the internal structure of the fracture zone. To construct the terrain surface model, massive data acquired through drone-generated oblique photogrammetry was imported into geographic information system software. Through a series of steps, including data processing, terrain feature extraction, and model optimization, a high-resolution terrain surface model was generated, faithfully recreating every detail of the canyon's topography. For the groundwater level model, Kriging interpolation, combined with long-term monitoring data from groundwater level monitoring wells, accurately captured the dynamics of the groundwater level, visually displaying its distribution over time and space. Subsequently, the three models were integrated in a unified 3D geological modeling platform. Through precise coordinate matching and spatial position adjustment, seamless connection between the models was ensured, and finally a comprehensive, accurate and intuitive 3D geological model of the canyon deposit was formed.

[0075] To collect construction data, technicians deployed a large number of sensors at the grouting site, including pressure sensors, flow sensors, and density sensors, to obtain real-time data on construction parameters such as grouting pressure, grouting flow rate, and slurry consistency. Advanced geological radar and borehole peepholes were also used to precisely measure the slurry diffusion range. In one construction area, geological radar detected that, under the current grouting pressure and flow rate conditions, the slurry diffusion range in the sandy rock formation was elliptical, with a major axis of approximately 8 meters and a minor axis of approximately 5 meters. In another area, the slurry diffusion range in the clay layer was relatively small due to the low permeability of the formation, forming a nearly circular shape with a radius of approximately 3 meters. To correlate these construction data with corresponding locations in the 3D geological model, technicians established a comprehensive data mapping mechanism. For example, for each grouting hole, the construction parameters and slurry diffusion range data were mapped to geological information such as the lithology, fault distance, and groundwater level at the location in the 3D geological model, generating a construction feature vector with rich geological attributes. These characteristic vectors not only include key parameters in the construction process, but also integrate geological environmental factors, providing comprehensive data support for subsequent model analysis and prediction.

[0076] In order to train the SVM model, the technicians classified and sorted the data according to the common risk conditions in the project, such as hole collapse, slurry leakage, and slurry leakage. Through in-depth analysis of construction data and summary of previous engineering experience, the radial basis kernel function was selected as the kernel function of the SVM model. When setting the value range of the penalty factor C and the kernel function parameter γ, after multiple preliminary tests and theoretical analysis, it was determined that the value range of the penalty factor C is [0.1, 10] and the value range of the kernel function parameter γ is [0.01, 1]. The grid search algorithm combined with the 5-fold cross-validation method was used to optimize the parameters. In the grid search process, the penalty factor C was traversed from 0.1 to 10 with a step size of 0.1; the kernel function parameter γ was traversed from 0.01 to 1 with a step size of 0.01. For each set of (C, γ) parameter combinations, the training data set was randomly divided into 5 parts, 4 parts were selected each time as training sets, and 1 part was selected as a validation set. The SVM model was trained on the training set, and the model performance was evaluated on the validation set, using accuracy as the evaluation indicator. As Figure 2 As shown in the figure, after extensive calculations and analysis, we found that when C = 1 and γ = 0.1, the model achieved the highest average accuracy in cross-validation, reaching 92%. Therefore, we determined the target penalty factor Copt = 1 and the target kernel function parameter γopt = 0.1.

[0077] Taking into account the serious impact of risk conditions such as hole collapse and slurry leakage on construction, technicians increased the weight of these risk condition samples. After a comprehensive assessment of the probability of occurrence and the degree of impact of risk conditions, the weight of normal condition samples was set to 1, the weight of hole collapse samples to 2, the weight of slurry leakage samples to 2.5, and the weight of slurry leakage samples to 3. Using the determined target penalty factor, kernel function parameters and dynamic weight coefficient, the SVM model was trained on the entire training data set. During the training process, the stochastic gradient descent algorithm was used to optimize the model. After multiple rounds of iterative training, the model gradually learned the complex nonlinear relationship between construction parameters and risk conditions, and finally obtained a pre-trained SVM model. In the subsequent construction process, the construction data collected in real time was input into this model. The model can accurately predict the possibility of risk conditions such as hole collapse, slurry leakage, and slurry leakage under the current construction status. For example, in a certain construction period, the model predicts that under the current construction parameters, the probability of slurry leakage in a certain grouting area is 75%, which exceeds the warning threshold. Based on this prediction result, the construction workers took preventive measures in advance, such as adjusting the grouting sequence and reducing the grouting pressure, which effectively avoided the occurrence of grouting accidents.

[0078] During construction, technicians input construction feature vectors into a pre-trained SVM model to predict geological response trends. For example, the SVM model predicts that in a specific area, due to the presence of a small fault in the stratum and the current high grouting pressure, grouting may be rapidly lost along the fault, affecting the grouting effect. Based on this prediction, the control parameters of the PID control model are modified according to the correction formula.

[0079] In practical applications, when the ΔP output by the SVM model indicates that the pressure change trend is a rapid increase, according to the above formula, the proportional coefficient Kp will increase accordingly, making the PID control model more sensitive to pressure changes and able to adjust the grouting pressure in time to avoid problems caused by excessive pressure.

[0080] The modified PID control model is used to dynamically control the operating parameters of the grouting equipment. Vibration signals, temperature data, pressure fluctuation curves, motor current change rates and other operating status parameters are obtained in real time through sensors installed on equipment such as grouting pumps, mixers, and slurry pipelines. These parameters are matched with historical fault data and a long short-term memory network is used to establish an equipment fault prediction model. Figure 3 As shown in the figure, in the process of building the model, the historical fault data was deeply analyzed to extract the characteristic change patterns of the equipment operating parameters before the fault occurred, such as the frequency and amplitude changes of the vibration signal, the abnormal temperature increase, the abnormal pattern of pressure fluctuation and the sudden change of the motor current. Figure 4As shown, the model can accurately predict the future probability of equipment failure. Assume that during construction, the equipment failure prediction model predicts that the probability of a grouting pump failing within the next three hours reaches 85%, exceeding the red warning threshold of 90%. A red warning signal is automatically issued, and construction personnel immediately stop the grouting pump and start the backup pump. At the same time, based on the type and severity of the fault, the system quickly adjusts the control strategy of the PID control model, such as reducing the startup pressure rise rate of the backup pump and optimizing the PID parameters to ensure the continuity and stability of the grouting construction. During the subsequent equipment maintenance process, through the disassembly and inspection of the faulty grouting pump, it was found that the fault type predicted by the model was completely consistent with the actual fault type, verifying the accuracy and reliability of the equipment failure prediction model.

[0081] During each feedback cycle (set to 1 hour), technicians input dynamically controlled grouting performance data, such as the actual grouting range, grouting pressure stability, and grouting stone strength, into a pre-trained SVM model. The SVM model outputs an assessment of the current grouting status, which is then compared and analyzed against preset grouting performance indicators. These indicators include that the grouting range should reach at least 90% of the design requirements, that the grouting pressure fluctuation range should be controlled within ±0.5 MPa, and that the grouting stone strength should reach at least 95% of the design strength. If the actual grouting pressure fluctuations are found to be large and exceed the preset ranges, technicians adjust the SVM model's input parameters based on the comparison results. For example, they increase the weight of pressure fluctuation-related features and adjust the control strategy of the PID control model by increasing the differential coefficient Td to enhance the ability to suppress pressure fluctuations. During a certain feedback cycle, analysis of the grouting performance data revealed that the actual grouting range only reached 80% of the design requirements, failing to meet the preset indicators. After in-depth analysis, technicians adjusted the weights of input parameters related to formation permeability, grouting pressure, and flow rate in the SVM model. They also appropriately increased the grouting pressure setpoint in the PID control model and extended the grouting time. Subsequent construction verification confirmed that the slurry diffusion range gradually met the design requirements, demonstrating that the feedback adjustment mechanism can effectively optimize the grouting process and improve grouting quality.

[0082] In summary, the embodiment of the present application significantly improved the grouting construction quality of the canyon deposit curtain grouting project through the SVM and PID-based curtain grouting construction control method. The curtain body's impermeability and structural strength far exceeded the design requirements. Testing showed that the leakage rate was only 30% of the design allowable leakage rate, and the curtain body's compressive strength reached 120% of the design strength. No serious quality defects or safety accidents occurred during the construction process. Compared with traditional construction methods, resource waste was reduced by 40%, project costs were reduced by 30%, and construction efficiency was increased by 35%.

[0083] Next, the canyon deposit curtain grouting construction control system based on SVM and PID proposed in accordance with an embodiment of the present application will be described with reference to the accompanying drawings.

[0084] Figure 5 It is a block diagram of a canyon deposit curtain grouting construction control system based on SVM and PID in an embodiment of the present application.

[0085] like Figure 5 As shown, the canyon deposit curtain grouting construction control system 10 based on SVM and PID includes: an acquisition module 100, a construction module 200, a correction module 300 and an optimization module 400.

[0086] Among them, the acquisition module 100 is used to obtain construction data, wherein the construction data includes formation permeability coefficient, grouting pressure, slurry consistency, grouting flow rate and slurry diffusion range; the construction module 200 is used to construct a three-dimensional geological model of the canyon accumulation body, and associate the formation permeability coefficient, grouting pressure, slurry consistency, grouting flow rate and slurry diffusion range with the corresponding positions in the three-dimensional geological model to generate a construction feature vector with geological attributes; the correction module 300 is used to input the construction feature vector into a pre-trained SVM model, predict the geological response trend, and correct the control parameters of the PID control model according to the geological response trend, wherein the control parameters of the PID control model include proportional coefficient, integral coefficient and differential coefficient; the optimization module 400 is used to use the corrected PID control model to dynamically control the operating parameters of the grouting equipment, and input the grouting effect data after dynamic control into the SVM model again to form a closed-loop feedback mechanism, and continuously optimize the PID control parameters and key grouting parameters.

[0087] It should be noted that the above explanation of the embodiment of the canyon deposit curtain grouting construction control method based on SVM and PID is also applicable to the canyon deposit curtain grouting construction control system based on SVM and PID in this embodiment, and will not be repeated here.

[0088] According to the canyon deposit curtain grouting control system based on SVM and PID proposed in the embodiment of the present application, by associating construction data with a three-dimensional geological model to generate a geological characteristic vector, the SVM model is used to accurately predict the geological response trend, providing a basis for the dynamic correction of the PID control model parameters, and realizing the intelligent adaptation of the grouting equipment operating parameters; the grouting effect data is then fed back to the SVM model to form a closed-loop feedback, and the control parameters and key process indicators are continuously optimized. The embodiment of the present application uses the SVM model to predict the geological response trend, which can keenly capture the potential changes in geological conditions during the construction process, detect possible problems in advance, and provide a scientific and reliable basis for the timely and accurate adjustment of construction parameters. It completely changes the blindness and lag of traditional judgment based on experience, and greatly improves the accuracy and reliability of construction decisions. The PID control model is used to dynamically adjust the operating parameters of the grouting equipment, and the closed-loop feedback mechanism is combined to continuously optimize the control parameters and key grouting parameters, thereby realizing refined and intelligent control of the grouting construction process. It can accurately adjust parameters such as grouting pressure, flow rate, and slurry ratio according to real-time changes in geological conditions, effectively improving the grouting quality and ensuring that the curtain body's impermeability and structural strength meet or even exceed design requirements, reducing quality defects and resource waste caused by improper construction, and reducing project costs and safety risks. The equipment failure prediction model established using the long short-term memory network can accurately predict potential equipment failures in advance, issue early warning signals in a timely manner, and quickly generate corresponding emergency response plans based on the type and severity of the failure. At the same time, it can flexibly adjust the PID control strategy. It is not only suitable for canyon deposit curtain grouting projects, but can also be widely used in other complex deposit reinforcement projects such as slopes. This solves the problem that it is difficult to achieve adaptive control of complex and changeable geological conditions in existing technologies, resulting in inaccurate grouting parameter settings and unstable construction quality.

[0089] FIG6 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may include:

[0090] A memory 601 , a processor 602 , and a computer program stored in the memory 601 and executable on the processor 602 .

[0091] When the processor 602 executes the program, the canyon deposit curtain grouting construction control method based on SVM and PID provided in the above embodiment is implemented.

[0092] Furthermore, the electronic device further includes:

[0093] The communication interface 603 is used for communication between the memory 601 and the processor 602 .

[0094] The memory 601 is used to store computer programs that can be run on the processor 602 .

[0095] The memory 601 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0096] If the memory 601, processor 602, and communication interface 603 are implemented independently, the communication interface 603, memory 601, and processor 602 can be interconnected via a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be divided into address buses, data buses, control buses, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0097] Optionally, in a specific implementation, if the memory 601, the processor 602 and the communication interface 603 are integrated on a chip, the memory 601, the processor 602 and the communication interface 603 can communicate with each other through an internal interface.

[0098] The processor 602 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0099] An embodiment of the present application also provides a computer-readable storage medium having a computer program or instruction stored thereon. When the computer program or instruction is executed by a processor, the above-mentioned canyon deposit curtain grouting construction control method based on SVM and PID is implemented.

[0100] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0101] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, a feature specified as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of this application, "N" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0102] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing a custom logical function or process step, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed in a different order than shown or discussed, including performing functions in a substantially simultaneous manner or in a reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application pertain.

[0103] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiment, the N steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, it can be implemented using any one or a combination of the following technologies known in the art: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having suitable combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0104] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

Claims

1. A canyon deposit curtain grouting construction control method based on SVM and PID, characterized in that: The following steps are involved: Acquiring construction data, wherein the construction data includes formation permeability, grouting pressure, slurry consistency, grouting flow rate, and slurry diffusion range; Constructing a three-dimensional geological model of the canyon deposit, correlating the formation permeability, grouting pressure, slurry consistency, grouting flow rate, and slurry diffusion range with corresponding positions in the three-dimensional geological model, and generating a construction feature vector with geological attributes; The construction feature vector is input into a pre-trained SVM model, wherein the radial basis kernel function is selected as the kernel function of the SVM model through in-depth analysis of construction data and summary of previous engineering experience; when setting the value range of the penalty factor C and the kernel function parameter γ, after multiple preliminary tests and theoretical analysis, it is determined that the value range of the penalty factor C is [0.1, 10] and the value range of the kernel function parameter γ is [0.01, 1]; a grid search algorithm combined with a 5-fold cross-validation method is used to optimize the parameters; in the grid search process, the penalty factor C is traversed from 0.1 to 10 with a step size of 0.1; the kernel function parameter γ is traversed from 0.01 to 1 with a step size of 0.01; for each set of (C, γ) parameter combinations, the training data set is randomly divided into 5 parts, 4 parts are selected as training sets each time, and 1 part is selected as a validation set, and the SVM model is trained on the training set. A VM model was developed and the model performance was evaluated on the validation set, with accuracy used as the evaluation indicator. After a comprehensive evaluation of the probability of occurrence and impact of risk conditions, the weight of the normal condition sample was set to 1, the weight of the hole collapse sample was set to 2, the weight of the slurry leakage sample was set to 2.5, and the weight of the slurry leakage sample was set to 3. The SVM model was trained on the entire training data set using the determined target penalty factor, kernel function parameters, and dynamic weight coefficient. During the training process, the stochastic gradient descent algorithm was used to optimize the model. After multiple rounds of iterative training, the model gradually learned the complex nonlinear relationship between construction parameters and risk conditions, and finally obtained a pre-trained SVM model to predict the geological response trend. According to the geological response trend, the control parameters of the PID control model were corrected, wherein the control parameters of the PID control model include proportional coefficient, integral coefficient, and differential coefficient. The correction formula is: K p =K p0 ×(1+α×ΔP); T i =T i0 / (1+β×ΔI); T d =T d0 ×(1+γ×ΔD); Among them, ΔP, ΔI, and ΔD are the trend changes of SVM output, α, β, and γ are adaptive adjustment coefficients, and K p is the proportional coefficient, T i is the integral coefficient, T d is the differential coefficient; The modified PID control model is used to dynamically control the operating parameters of the grouting equipment. The operating state parameters of the grouting equipment, including vibration signals, temperature data, pressure fluctuation curves, and motor current change rates, are obtained. Feature matching is performed between the operating state parameters and historical fault data, and an equipment fault prediction model is established based on a long-short-term memory network to predict the future failure probability of the equipment. When the predicted failure probability exceeds a warning threshold, a warning signal is automatically issued. At the same time, a corresponding emergency response plan is generated based on the fault type and severity, and the control strategy of the PID control model is adjusted. The dynamically controlled grouting effect data is again input into the SVM model to form a closed-loop feedback mechanism to continuously optimize the PID control parameters and key grouting parameters. A feedback cycle is set. Within each feedback cycle, the dynamically controlled grouting effect data is input into a pre-trained SVM model, and an evaluation result of the current grouting state is output. The grouting effect data corresponding to the evaluation result is compared and analyzed with preset grouting effect indicators to obtain a comparison result between the actual grouting effect and the expected grouting effect. The input parameters of the SVM model are adjusted based on the comparison result to adapt the model to the changing grouting conditions, and the control strategy of the PID control model is adjusted at the same time.

2. The canyon deposit curtain grouting construction control method based on SVM and PID according to claim 1 is characterized in that: Construct a 3D geological model of the canyon deposits, including: Obtain geological survey data, topographic data, and groundwater level data; Based on the geological survey data, the strata of the canyon deposits are divided, the fault structures in the canyon deposits are identified, and a three-dimensional fault model is established; based on the topographic data, a topographic surface model of the canyon deposits is constructed; based on the groundwater level data, a groundwater level model is constructed using an interpolation method; The three-dimensional fault model, the terrain surface model and the groundwater level model are integrated to obtain a three-dimensional geological model.

3. The canyon deposit curtain grouting construction control method based on SVM and PID according to claim 1 is characterized in that: Pre-trained SVM models, including: Select the radial basis kernel function according to the risk condition classification, set the value range of the penalty factor and kernel function parameters, and determine the target penalty factor and kernel function parameters through grid search and cross-validation; Set dynamic weight coefficients to increase weights for risky working condition samples such as hole collapse and slurry leakage; The determined target penalty factor, kernel function parameters and dynamic weight coefficient are used to train the model on the training data set to obtain a pre-trained SVM model.

4. The canyon deposit curtain grouting construction control method based on SVM and PID according to claim 1 is characterized in that: The key grouting parameters include slurry viscosity, gel time, grouting pressure gradient value and grouting rate fluctuation range.

Images