Slope seepage analysis method under flood discharge atomization condition

Abstract

The invention provides a slope seepage analysis method under a flood discharge atomization condition. The slope seepage analysis method comprises the following steps: S1, acquiring slope pore water pressure data and matrix suction data through a distributed sensor array; s2, the data fusion unit compares the pore water pressure data with a dynamic critical pressure threshold value, compares the matric suction data with a dynamic critical suction threshold value, and generates a seepage analysis trigger instruction when any data exceeds the limit; s3, starting a transient seepage field calculation model by the seepage calculation engine based on the fracture network topology data; s4, generating a slope stability control parameter set according to a seepage pressure distribution diagram output by the transient seepage field calculation model; and S5, the execution system regulates and controls the operation state of the drainage facility according to the slope stability control parameter set. The slope seepage pressure can be reduced, the slope stability is improved, and safe operation of a water conservancy project is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of slope stability monitoring and control technology in water conservancy and hydropower projects, and more specifically, to a slope seepage analysis method under flood discharge atomization conditions. Background Technology

[0002] In the field of water conservancy and hydropower engineering, slope stability is one of the key factors affecting project safety. Under the influence of complex natural environments and engineering activities, slope stability is affected by a combination of factors, among which seepage is one of the main factors influencing slope stability. During flood discharge, the high-speed movement and atomization of water flow cause significant changes in seepage conditions on and inside the slope. Traditional slope seepage analysis methods are mainly based on seepage models under static conditions, which are difficult to accurately reflect the spatiotemporal variations of the seepage field inside the slope under dynamic conditions such as flood discharge and atomization. These methods typically assume that the fracture network and permeability coefficient inside the slope are uniformly distributed, ignoring the non-uniformity and anisotropy of fractures in actual slopes, leading to significant deviations between calculation results and actual conditions. Furthermore, existing technologies also have shortcomings in the acquisition and processing of monitoring data, failing to obtain real-time and accurate information on changes in key parameters such as pore water pressure and matrix suction, making it difficult to promptly detect potential seepage risks. In terms of the regulation of drainage facilities, existing technologies lack a dynamic feedback mechanism based on seepage analysis results, making it impossible to adjust the operation of drainage facilities in real time according to the actual seepage state of the slope. As a result, it is impossible to effectively reduce the seepage pressure on the slope and improve the stability of the slope.

[0003] In implementing the embodiments of the present invention, the prior art has at least the following problems or defects: First, it cannot accurately reflect the spatiotemporal variation characteristics of the seepage field inside the slope under dynamic conditions such as flood discharge atomization; second, the collection and processing of monitoring data are not real-time or accurate enough, making it difficult to detect potential seepage risks in a timely manner; third, the control of drainage facilities lacks a dynamic feedback mechanism with seepage analysis results, which cannot effectively reduce the seepage pressure of the slope and improve the stability of the slope. Summary of the Invention

[0004] This invention provides a method for slope seepage analysis under flood discharge atomization conditions, including: S1: Collect slope pore water pressure data and matrix suction data through a distributed sensor array; S2: The data fusion unit compares the pore water pressure data with the dynamic critical pressure threshold and the matrix suction data with the dynamic critical suction threshold. When any data exceeds the limit, a seepage analysis trigger command is generated. S3: The seepage calculation engine starts the transient seepage field calculation model based on the fractured network topology data; S4: Generate a set of slope stability control parameters based on the seepage pressure distribution map output by the transient seepage field calculation model; S5: The execution system regulates the operation status of drainage facilities based on the slope stability control parameter set.

[0005] Furthermore, S3 includes: S3.1: Parse the pre-stored dataset of spatial distribution of rock fractures; S3.2: Calculation of the three-dimensional permeability tensor field based on the fracture aperture attenuation law; S3.3: Construct the control differential equation for unsaturated-saturated coupled seepage flow; S3.4: Adaptive unstructured grid discrete slope geometry model is adopted; S3.5: Define the convergence criteria for Newton's iteration method; S3.6: Solve the discretized algebraic equations to obtain the spatiotemporal distribution of the seepage field.

[0006] Furthermore, the unsaturated-saturated coupled seepage control differential equation described in S3.3 is as follows:

[0007] Where: K is the permeability coefficient tensor; h is the head function; C is the capacity function; t is the time variable.

[0008] Furthermore, S2 includes: S2.1: Establish a mapping database between flood discharge atomization intensity levels and dynamic critical pressure thresholds; S2.2: Monitor the current flood discharge atomization intensity level using acoustic sensors; S2.3: Based on the current flood discharge atomization intensity level, retrieve the dynamic critical pressure threshold from the mapping database; S2.4: When the pore water pressure data exceeds the dynamic critical pressure threshold, a pore water pressure over-limit command is generated; S2.5: When the matrix suction data is lower than the dynamic critical suction threshold, a matrix suction insufficient command is generated; S2.6: Combines the pore water pressure over-limit command and the matrix suction insufficient command to generate the seepage analysis trigger command.

[0009] Furthermore, S5 includes: S5.1: Construct a functional relationship table between the flood discharge atomization intensity level and the adjustment coefficient of the drainage hole spacing; S5.2: Obtain the drainage hole spacing adjustment coefficient by querying the function relationship table based on the current flood discharge atomization intensity level; S5.3: Generate the opening and closing sequence of the drain valve based on the drain hole spacing adjustment coefficient; S5.4: Control the opening degree of the electric valve according to the opening and closing sequence of the drain valve; S5.5: Real-time collection of drainage flow data and feedback to the seepage calculation engine; S5.6: Correct the boundary conditions of the transient seepage field calculation model based on the drainage flow data.

[0010] Furthermore, S4 includes: S4.1: Identify areas of abnormal displacement rate in the seepage pressure distribution map; S4.2: Match the filter layer thickness correction factor according to the displacement rate level; S4.3: Generate a set of construction instructions for the modification of the filter layer structure; S4.4: Generate collaborative optimization instructions for parameters of adjacent drainage holes in areas with abnormal displacement rates; S4.5: Output reinforcement material space allocation scheme; S4.6: Verify the matching degree between the spatial distribution scheme of reinforcement materials and the displacement control target.

[0011] Furthermore, S4 also includes: S4.7: Monitor the trajectory of slope stability coefficient changes; S4.8: Activate the anchoring loading protocol when the slope stability coefficient enters the critical decrease range; S4.9: Determine the number of anchorage force loading stages based on the rate of decrease of the slope stability coefficient; S4.10: Apply anchoring force in stages according to time gradient; S4.11: Dynamically adjust the anchoring force loading rate based on displacement sensor feedback; S4.12: Establish the three-dimensional response surface of anchoring force-displacement-time.

[0012] Furthermore, S3.5 includes: S3.5.1: Record the residual rate of change data of Newton's iterative method; S3.5.2: Trigger the algorithm switching flag when the residuals continuously exceed the convergence criterion; S3.5.3: Reconstruct the slope stability equation using the simplified slice method; S3.5.4: Calculate the safety factor of the potential sliding surface based on the Mohr-Coulomb criterion; S3.5.5: Output the sliding surface space coordinates corresponding to the minimum safety factor; S3.5.6: Feeds back the spatial coordinates of the sliding surface to the adaptive unstructured mesh discretization module.

[0013] Furthermore, S1 includes: S1.1: Multiple sets of parallel monitoring profiles are set up along the main sliding direction of the slope; S1.2: Determine the spatial envelope of the potential slip surface through ground-penetrating radar scanning; S1.3: Deploy a high-density array of monitoring points within the spatial envelope of the potential sliding surface; S1.4: Set the maximum allowable point spacing constraint for the high-density monitoring point array; S1.5: Implement a real-time cross-validation process for monitoring data; S1.6: Establish an automatic mechanism for removing abnormal data.

[0014] Furthermore, S3.1 includes: S3.1.1: Obtain rock mass fracture point cloud dataset through 3D laser scanning; S3.1.2: Reconstructing the spatial distribution function of fracture aperture based on rock mass fracture point cloud dataset; S3.1.3: Calculate the topological connectivity index matrix of the rock mass fracture network; S3.1.4: Generate the eigenvalue spectrum of the anisotropic permeation tensor; S3.1.5: Verify the reliability of the anisotropic permeability tensor eigenvalue spectrum through on-site water injection tests; S3.1.6: Optimize the fracture parameter inversion algorithm based on the verification results.

[0015] The embodiments of the present invention have at least the following beneficial effects: 1. This invention uses a distributed sensor array to collect real-time data on pore water pressure and matrix suction of slopes and compares it with dynamic critical thresholds. This enables timely and accurate detection of seepage anomalies on slopes under flood discharge atomization conditions. It solves the problems of non-real-time and inaccurate data collection in existing technologies, which makes it difficult to detect potential seepage risks in a timely manner. This effectively improves the reliability and timeliness of slope seepage monitoring and provides a solid data foundation for subsequent seepage analysis and slope stability control.

[0016] 2. This invention employs a transient seepage field calculation model based on fracture network topology data, which can fully consider the non-uniformity and anisotropy of internal fractures in slopes, and accurately calculate the spatiotemporal variation characteristics of the internal seepage field of slopes under dynamic conditions such as flood discharge atomization. This solves the problem that existing seepage analysis models cannot accurately reflect the changes in the internal seepage field of slopes under dynamic conditions, making the seepage analysis results closer to the actual situation, providing a more accurate basis for the generation of slope stability control parameters, and thus improving the scientificity and effectiveness of slope stability regulation.

[0017] 3. This invention regulates the operation of drainage facilities based on a set of slope stability control parameters, achieving intelligent dynamic regulation of drainage facilities. By collecting drainage flow data in real time and feeding it back to the seepage calculation engine, the boundary conditions of the transient seepage field calculation model are corrected, forming a dynamic feedback regulation mechanism. This solves the problems of lack of drainage facility regulation and dynamic feedback mechanism for seepage analysis results in the prior art. It can adjust the operation of drainage facilities in real time according to the actual seepage state of the slope, effectively reducing the seepage pressure of the slope, improving the stability of the slope, and ensuring the safe operation of water conservancy projects. Attached Figure Description

[0018] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings. Several embodiments of the invention are illustrated in the drawings by way of example and not limitation, wherein: Figure 1 This is a flowchart illustrating a slope seepage analysis method under flood discharge atomization conditions according to an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application. It should be noted that similar reference numerals and letters in the following drawings indicate similar items; therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. Furthermore, in the description of this application, the terms first, second, etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0020] like Figure 1As shown, this application proposes a slope seepage analysis method under flood discharge atomization conditions, comprising: S1: collecting slope pore water pressure data and matrix suction data through a distributed sensor array; S2: a data fusion unit compares the pore water pressure data with a dynamic critical pressure threshold and the matrix suction data with a dynamic critical suction threshold, generating a seepage analysis trigger command when any data exceeds the limit; S3: a seepage calculation engine starts a transient seepage field calculation model based on fracture network topology data; S4: a slope stability control parameter set is generated based on the seepage pressure distribution map output by the transient seepage field calculation model; S5: the execution system regulates the operation status of drainage facilities according to the slope stability control parameter set.

[0021] A distributed sensor array refers to a monitoring network composed of multiple sensors arranged at different spatial locations on a slope. Specifically, it can be implemented using fiber optic grating sensors or piezoelectric sensors. By deploying sensor nodes on the slope surface and at key internal nodes, it forms a continuous spatial monitoring capability, solving the problem of insufficient coverage of non-uniform seepage fields by traditional single-point monitoring. A dynamic critical pressure threshold refers to a pore water pressure warning value that is adjusted in real time according to the intensity of flood discharge atomization. Specifically, it can be obtained by monitoring the flood discharge atomization intensity level using acoustic sensors and then querying a mapping database, solving the problem that static thresholds cannot adapt to dynamic changes in flood discharge conditions. A transient seepage field calculation model refers to a seepage numerical model considering the anisotropic characteristics of the fracture network. Specifically, it can be constructed using a three-dimensional permeability tensor field to create unsaturated-saturated coupled differential equations and solved using an adaptive grid discretization method, overcoming the calculation bias caused by the traditional uniform permeability assumption. A slope stability control parameter set refers to a set of engineering control indicators that transforms the seepage pressure distribution. Specifically, it can include drainage hole spacing adjustment coefficients, filter layer thickness correction coefficients, and anchoring force loading gradient parameters, realizing a quantitative correlation between seepage field calculation results and engineering control measures. Drainage facility operation status control refers to adjusting the working mode of the drainage system based on real-time seepage analysis results. Specifically, it can be achieved by controlling the opening degree of electric valves and optimizing the opening and closing sequence of drainage holes to form a dynamic matching mechanism between seepage pressure and drainage capacity.

[0022] The core innovation of this application lies in constructing a closed-loop system for dynamic monitoring and feedback control. It collects multi-source data in real time through a distributed sensor array, and initiates transient seepage field calculations by combining a dynamic threshold triggering mechanism. The seepage pressure distribution is transformed into executable engineering control parameters, and a dynamic match between the seepage field and drainage capacity is formed based on real-time drainage control, thereby achieving real-time suppression and stability control of slope seepage risk under flood discharge atomization conditions.

[0023] The working process and principle of this application are as follows: A distributed sensor array collects pore water pressure and matrix suction data of the slope, enabling comprehensive monitoring of key areas of the slope. A data fusion unit compares the collected data with dynamic critical thresholds; when any data exceeds the limit, a seepage analysis trigger command is generated, ensuring the system can respond promptly to changes in the intensity of flood discharge atomization. The seepage calculation engine initiates a transient seepage field calculation model based on fracture network topology data, accurately modeling the anisotropic characteristics of the seepage field by introducing fracture structure parameters. Based on the seepage pressure distribution map output by the calculation model, a set of slope stability control parameters is generated, transforming the seepage field calculation results into executable engineering control indicators. The execution system dynamically adjusts the operating status of the drainage facilities based on these parameter sets, forming a real-time matching mechanism between seepage status and drainage capacity.

[0024] The collaborative process of each step is as follows: First, the distributed sensor array continuously collects data to provide real-time monitoring information for the system. The data fusion unit ensures that the generation of trigger commands is adapted to the current flood discharge atomization intensity through a dynamic threshold comparison mechanism. After receiving the trigger command, the seepage calculation engine immediately starts a transient model based on the fracture network topology to calculate the current seepage field state. The slope stability control parameter set is generated based on the seepage pressure distribution map, providing a decision-making basis for the execution system. Finally, the execution system adjusts the operating status of the drainage facilities according to these parameters to achieve dynamic control of the slope seepage pressure.

[0025] The rationale for selecting key technical features and their impact on performance include: The use of a distributed sensor array improves the spatial resolution of monitoring, overcoming the limitations of traditional single-point monitoring. The dynamic critical threshold comparison mechanism adapts to changes in flood discharge atomization intensity, improving the system's response sensitivity. The transient seepage field calculation model based on fracture network topology data overcomes the limitations of the traditional uniform seepage assumption, improving model accuracy. The conversion from seepage pressure distribution map to control parameter set effectively connects calculation results with engineering practice. Dynamic control of drainage facility operation forms a closed-loop feedback mechanism, effectively suppressing the risk of accumulated seepage pressure.

[0026] Preferably, in the slope seepage analysis under flood discharge atomization conditions, a distributed sensor array is first deployed. This array includes pore water pressure sensors and matrix suction sensors, distributed in a grid pattern along the main sliding direction of the slope, covering the spatial envelope of the potential sliding surface. The spacing between the sensors is determined based on the slope geological conditions and the expected gradient of seepage field changes to ensure the representativeness and continuity of the monitoring data.

[0027] The data fusion unit receives data collected by the sensor array in real time. This unit has a built-in mapping relationship between the flood discharge atomization intensity level and the dynamic critical pressure threshold. It monitors the current flood discharge atomization intensity level through acoustic sensors and retrieves the corresponding dynamic critical pressure threshold and critical suction threshold from the mapping database. The data fusion unit compares the real-time monitoring data with these dynamic thresholds, and generates a seepage analysis trigger command when the pore water pressure exceeds the critical pressure threshold or the matrix suction falls below the critical suction threshold.

[0028] Upon receiving a trigger command, the seepage calculation engine initiates a transient seepage field calculation model. This model is constructed based on pre-obtained rock mass fracture network topology data acquired through 3D laser scanning. During the calculation, the fracture aperture attenuation law is considered, the 3D seepage tensor field is calculated, and unsaturated-saturated coupled seepage control differential equations are constructed. An adaptive unstructured mesh is used to discretize the slope geometry model, and a convergence criterion for Newton's iteration method is set. By solving the discretized algebraic equations, the spatiotemporal distribution of the seepage field is obtained.

[0029] Based on the seepage pressure distribution map output by the transient seepage field calculation model, the system generates a set of slope stability control parameters. This includes identifying areas of abnormal displacement rate in the seepage pressure distribution map, matching the filter layer thickness correction coefficient according to the displacement rate level, generating a set of construction instructions for filter layer structural modification, and generating collaborative optimization instructions by associating the parameters of adjacent drainage holes in areas of abnormal displacement rate.

[0030] The execution system regulates the operation of drainage facilities based on the slope stability control parameter set. First, a functional relationship table is constructed between the flood discharge atomization intensity level and the drainage hole spacing adjustment coefficient. The drainage hole spacing adjustment coefficient is retrieved based on the current flood discharge atomization intensity level. Based on this coefficient, a drainage valve opening and closing sequence is generated, and the electric valve opening is controlled according to the sequence. The system collects drainage flow data in real time and feeds it back to the seepage calculation engine to correct the boundary conditions of the transient seepage field calculation model, forming a closed-loop control.

[0031] This application further proposes to analyze the pre-stored spatial distribution dataset of rock mass fractures; calculate the three-dimensional permeability tensor field based on the fracture aperture decay law; construct the unsaturated-saturated coupled seepage control differential equation; adopt an adaptive unstructured grid to discretize the slope geometric model; set the convergence criterion of Newton's iteration method; and solve the discretized algebraic equation system to obtain the spatiotemporal distribution of the seepage field.

[0032] The spatial distribution dataset of rock fractures can be obtained through 3D laser scanning or acoustic tomography. The fracture aperture attenuation law can be represented by a mathematical model that varies exponentially with depth. The calculation of the 3D permeability tensor field can combine statistical parameters of fracture attitude with permeability anisotropy correction coefficients, for example, using the cosine of the angle between the fracture normal vector and the main permeability direction as a weighting factor. The construction of the unsaturated-saturated coupled seepage control differential equations requires the introduction of dynamic coupling terms between matrix suction and pore water pressure, specifically describing the permeability coefficient variation law in different saturation regions through piecewise continuous functions. The generation of adaptive unstructured meshes can be based on the Delaunay triangulation algorithm, automatically refining the mesh nodes in densely fractured areas, with the mesh size controllable within the range of 0.1-0.5 meters. The convergence criterion for the Newton iteration method can be set to trigger algorithm switching when the residual norm decreases to 1e-6 times the initial value or the number of iterations exceeds 50. The solution of the algebraic equation system can be achieved using the preprocessed conjugate gradient method, combined with multi-core parallel computing technology to accelerate the solution process.

[0033] Specifically, the spatial distribution dataset of rock fractures was obtained with millimeter-level precision using a 3D laser scanner to acquire fracture geometric parameters. The fracture aperture decay law was established by fitting water injection test data at different depths to establish an exponential decay model. In the calculation of the 3D permeability tensor field, fracture attitude parameters were converted into tensor components of the principal permeability direction, and the anisotropy correction coefficients were validated for their effectiveness through Monte Carlo simulation. The construction of the unsaturated-saturated coupled seepage control differential equation adopted a hybrid form of the Richards equation, introducing matrix suction as an independent variable into the Darcy's law correction term. During the adaptive unstructured mesh discretization process, the mesh generator dynamically adjusted the element size according to the fracture density gradient, generating tetrahedral and hexahedral hybrid elements in the fracture intersection region. The convergence determination of the Newton iteration method was achieved by real-time monitoring of the residual change rate; when the ratio of the residuals of two adjacent iterations exceeded a set threshold, it automatically switched to the simplified slice method. The solution process of the algebraic equation system adopted a sparse matrix storage format, combined with GPU acceleration technology to achieve fast solution of large-scale equation systems. Thus, the non-uniform distribution characteristics of rock mass fractures are accurately characterized by a three-dimensional permeability tensor field, the dynamic coupling effect between unsaturated and saturated zones is fully described by a hybrid differential equation, the complex fracture network morphology is modeled with high precision by adaptive grid discretization, and the stability of numerical solutions is guaranteed by convergence criteria and algorithm switching mechanism, ultimately obtaining spatiotemporal distribution data of seepage field that conforms to actual geological conditions.

[0034] Preferably, firstly, a pre-stored dataset of spatial distribution of rock mass fractures is parsed. This dataset contains information such as the three-dimensional coordinates, orientation, dip angle, and aperture of fractures within the slope rock mass. Topological features of the fracture network are then extracted using data processing algorithms.

[0035] Secondly, the three-dimensional permeability tensor field is calculated based on the fracture aperture decay law. An exponential decay model is used to describe the change of fracture aperture with depth, and an anisotropic permeability tensor is constructed by combining fracture orientation information. The slope is meshed, and a corresponding permeability tensor value is assigned to each mesh cell.

[0036] Furthermore, a differential equation governing unsaturated-saturated coupled seepage flow is constructed. This equation considers the dynamic coupling relationship between matrix suction and pore water pressure, and can describe the seepage characteristics of the saturated and unsaturated zones within the slope. The equation includes parameters such as the permeability tensor, hydraulic head function, and capacity function.

[0037] Next, an adaptive unstructured mesh discretization model of the slope geometry is adopted. Based on the fracture distribution characteristics, the mesh density is dynamically adjusted, with a finer mesh division used in areas with dense fractures. This method can better adapt to complex fracture network morphologies.

[0038] Then, the convergence criterion for Newton's iteration method is set. By monitoring the rate of change of the residuals, when the iteration process shows a divergent trend, it automatically switches to other numerical solution algorithms, such as the simplified slice method, to ensure the stability of the calculation.

[0039] Finally, the spatiotemporal distribution of the seepage field is obtained by solving the discretized algebraic equations. Efficient matrix solving techniques, such as sparse matrix storage and parallel computing, are employed to process large-scale equation systems and output spatiotemporal distribution data of pore water pressure and velocity fields within the slope.

[0040] This application further proposes a non-saturated-saturated coupled seepage control differential equation, specifically in the form of: Where: K is the permeability coefficient tensor; h is the head function; C is the capacity function; t is the time variable.

[0041] The permeability tensor K is generated using three-dimensional fracture network topology data. Its anisotropic characteristics are calculated through the fracture aperture attenuation law. For example, the spatial distribution function is reconstructed using fracture point cloud datasets obtained from three-dimensional laser scanning, thereby generating anisotropic permeability tensor eigenvalue spectra. The gradient term ∂h of the head function h and the time-varying term of the head are also considered. The combination of these methods achieves spatial discretization through an adaptive unstructured grid discretization model of the slope geometry, for example, using tetrahedral grids to subdivide complex slope morphologies. The nonlinear characteristics of the water content function C are determined by fitting experimental data on the relationship between soil water content and hydraulic head, for example, using the van Genuchten model to describe the water content variation in the unsaturated zone. The explicit expression of the time variable t is achieved through a transient seepage field calculation model, for example, using the implicit time integration method to solve the time step process.

[0042] Specifically, under flood discharge atomization conditions, the dynamic changes of the seepage field inside the slope are realized by solving this differential equation. The introduction of the permeability coefficient tensor K enables the characterization of the spatial anisotropy of the rock mass fracture network. For example, the permeability coefficient in the principal direction of the fracture can be 1-2 orders of magnitude higher than that in the secondary direction. Through the calculation of the three-dimensional permeability tensor field, the differences in permeability capacity in different directions are accurately reflected. The combination of the gradient term and the time-varying term of the head function h completely describes the dynamic relationship between pressure potential energy and time evolution in the seepage field. For example, Darcy's law is used in the saturated region, and the generalized Darcy's law is used for coupled solution in the unsaturated region. The dynamic adjustment of the capacity function C is driven by real-time monitoring of pore water pressure data. For example, when the soil water content reaches a critical value, the capacity function automatically switches to the value corresponding to the saturated state. The explicit expression of the time variable t enables the modeling of the transient seepage process. For example, a second-level time step is used to capture the rapid seepage response under the impact of flood discharge atomization. Through the synergistic effect of the above features, the model can accurately simulate the transition process of the slope from unsaturated to saturated state and reflect the influence of anisotropic permeability on the seepage path, thereby improving the spatiotemporal resolution of seepage field calculation.

[0043] Preferably, the control differential equation for unsaturated-saturated coupled seepage is:

[0044] K represents the permeability tensor, h represents the head function, C represents the capacity function, and t represents the time variable.

[0045] In practical applications, the permeability tensor K can be obtained through field and laboratory tests. For example, in-situ permeability tests can be conducted at different locations and directions on a slope to measure the permeability coefficients in the horizontal and vertical directions. For anisotropic media, K can be represented as a 3×3 matrix.

[0046] The head function h can be obtained in real time by deploying a network of pore water pressure sensors. The sensor network can be arranged in a layered manner, with measuring points set up in the surface, middle and deep layers of the slope to form a three-dimensional monitoring network.

[0047] The water-holding capacity function C can be determined through the soil-water characteristic curve (SWCC). SWCC can be measured in the laboratory using equipment such as a pressure plate apparatus, and the relationship between water-holding capacity and matrix suction can be obtained by fitting the curve.

[0048] The time variable t can be set as a discrete time step, for example, calculating the seepage field distribution every 10 minutes. By adjusting the time step, the calculation accuracy and efficiency can be balanced.

[0049] In the solution process, the finite element method can be used to discretize the governing equations. First, the slope is divided into several elements, and nodal displacement interpolation functions are established. Then, these are substituted into the governing equations to obtain the element stiffness matrix and load vector, which are then assembled into a global equation system. Finally, the iterative method is used to solve the equation system to obtain the head values ​​at each time point and at each node.

[0050] This application further proposes establishing a mapping database between flood discharge atomization intensity levels and dynamic critical pressure thresholds; monitoring the current flood discharge atomization intensity level using acoustic sensors; retrieving the dynamic critical pressure threshold from the mapping database based on the current flood discharge atomization intensity level; generating a pore water pressure over-limit command when the pore water pressure data exceeds the dynamic critical pressure threshold; generating a matrix suction insufficiency command when the matrix suction data is lower than the dynamic critical suction threshold; and merging the pore water pressure over-limit command and the matrix suction insufficiency command to generate a seepage analysis trigger command.

[0051] The mapping database between flood discharge atomization intensity levels and dynamic critical pressure thresholds was constructed through a combination of field water injection tests and numerical simulations. For example, atomization intensity was divided into three levels, with corresponding dynamic critical pressure thresholds of 40 kPa, 60 kPa, and 80 kPa, respectively. Acoustic sensors employ the principle of multi-band sound wave reflection, determining the current atomization intensity level by analyzing the attenuation characteristics of sound waves in atomized water vapor. For instance, three characteristic frequency detection windows were set within the 20 kHz to 50 kHz frequency band. The dynamic critical suction threshold was determined through indoor geotechnical tests based on soil type; for example, the critical suction threshold for cohesive soil was 25 kPa, and for sandy soil, it was 15 kPa. Pore water pressure exceeding limits and matrix suction insufficient limits were fused using logical AND operations. A seepage analysis trigger command was generated only when both commands were activated simultaneously; for example, the analysis process was triggered when pore water pressure exceeded 60 kPa and matrix suction was below 20 kPa.

[0052] Specifically, the flood discharge atomization intensity level is monitored in real time by acoustic sensors. The propagation speed and attenuation of sound waves in the atomized water vapor are positively correlated with the concentration of atomized particles. Atomization intensity grading standards are established by calibrating acoustic characteristic parameters corresponding to different concentrations. The dynamic critical pressure threshold is dynamically retrieved from a pre-established mapping database based on the atomization intensity level, ensuring that the threshold parameter accurately matches the current water vapor infiltration intensity. Pore water pressure data and matrix suction data are synchronously collected through a distributed sensor array. When the pore water pressure exceeds the dynamic critical pressure threshold, it indicates that the osmotic pressure has exceeded the drainage capacity of the slope; when the matrix suction is lower than the dynamic critical suction threshold, it indicates that the soil shear strength has decreased due to water intrusion. The fusion triggering mechanism of the two commands effectively distinguishes between two seepage failure modes—abrupt osmotic pressure changes and soil strength decay—through dual physical quantity constraints. For example, when the atomization intensity level is level two, the dynamic critical pressure threshold is set to 60 kPa, and the dynamic critical suction threshold is set to 20 kPa. The seepage analysis triggering command is activated only when the pore water pressure exceeds 60 kPa and the matrix suction is lower than 20 kPa. This mechanism reduces the probability of false triggering caused by fluctuations in a single parameter through dynamic correlation and logical judgment between physical quantities, while ensuring timely identification of complex seepage risks.

[0053] Preferably, a mapping database between flood discharge atomization intensity levels and dynamic critical pressure thresholds is established. This database contains multiple atomization intensity levels, each corresponding to a set of dynamic critical pressure thresholds. For example, the atomization intensity levels can be divided into three levels: low, medium, and high, with corresponding dynamic critical pressure thresholds of 100 kPa, 150 kPa, and 200 kPa, respectively.

[0054] The current flood discharge atomization intensity level is monitored using acoustic sensors. These sensors are deployed around the flood discharge area to collect acoustic signals in real time. Signal processing algorithms then convert the collected acoustic signals into atomization intensity levels.

[0055] The system retrieves the dynamic critical pressure threshold from the mapping database based on the current flood discharge atomization intensity level. For example, when the current atomization intensity is detected to be at the medium level, the system automatically extracts the corresponding dynamic critical pressure threshold of 150 kPa from the database.

[0056] A pore water pressure over-limit command is generated when the pore water pressure data exceeds the dynamic critical pressure threshold. The system continuously compares the real-time collected pore water pressure data with the current dynamic critical pressure threshold. Once the pore water pressure exceeds the threshold, an over-limit command is immediately generated.

[0057] A matrix suction insufficiency command is generated when the matrix suction data falls below the dynamic critical suction threshold. The system simultaneously monitors the matrix suction data and compares it to the preset dynamic critical suction threshold. When the matrix suction falls below the threshold, the system generates an insufficiency command.

[0058] The system integrates pore water pressure exceeding limits and matrix suction insufficient commands to generate seepage analysis trigger commands. The system comprehensively considers both types of commands, triggering the seepage analysis process when both exceeding and insufficient conditions are simultaneously met.

[0059] This application further proposes to construct a functional relationship table between the flood discharge atomization intensity level and the drainage hole spacing adjustment coefficient; obtain the drainage hole spacing adjustment coefficient by querying the functional relationship table according to the current flood discharge atomization intensity level; generate a drainage valve opening and closing sequence based on the drainage hole spacing adjustment coefficient; control the opening degree of the electric valve according to the drainage valve opening and closing sequence; collect drainage flow data in real time and feed it back to the seepage calculation engine; and correct the boundary conditions of the transient seepage field calculation model according to the drainage flow data.

[0060] The functional relationship between the flood discharge atomization intensity level and the drainage hole spacing adjustment coefficient can be established through experimental data fitting or numerical simulation. For example, the adjustment coefficient is 1.2 for level 1, 1.0 for level 2, and 0.8 for level 3. The drainage hole spacing adjustment coefficient is used to generate the valve opening and closing sequence, which can be achieved by controlling the opening time interval of the electric valves of adjacent drainage holes. For example, when the adjustment coefficient is 1.2, the opening interval of adjacent valves is shortened by 20%. The electric valve opening degree control adopts a PID algorithm, and the valve opening degree error is adjusted in real time through the flow sensor feedback. The drainage flow rate data is transmitted to the seepage calculation engine through the IoT communication module, and the data update frequency can be set to once per second. The boundary condition correction adopts the finite difference method, substituting the measured drainage flow rate into the Darcy's law expression of the seepage field calculation model, and iteratively updating the hydraulic gradient parameters.

[0061] Specifically, the intensity level of the flood discharge atomization is monitored in real time by acoustic sensors and matched with a function relationship table to obtain the corresponding adjustment coefficient for the spacing of the drainage holes. When generating the opening and closing sequence of the drainage valves based on the adjustment coefficients, timing control logic is used to determine the opening order and time interval of each drainage hole valve. The opening degree of the electric valves is adjusted in stages according to the opening and closing sequence, and flow data acquisition is triggered after each adjustment stage is completed. The real-time collected drainage flow data is transmitted to the seepage calculation engine through the data bus, and the drainage flow is used as a dynamic boundary condition input in the seepage field calculation model. The boundary condition correction process adopts the Newton-Raphson iteration method, adjusting the permeability coefficient tensor by comparing the residual between the calculated flow and the measured flow. A closed-loop feedback mechanism is used to achieve dynamic coupling between the seepage field calculation model and the control parameters of the drainage facilities, so that the transient seepage field calculation results are updated synchronously with the actual seepage state of the slope.

[0062] Preferably, firstly, a functional relationship table is constructed between the flood discharge atomization intensity level and the drainage hole spacing adjustment coefficient. This table can include multiple flood discharge atomization intensity levels, for example, it can be divided into 5 levels: light, moderate, strong, strong, and extremely strong. Each level corresponds to a drainage hole spacing adjustment coefficient, which can be a decimal between 0 and 1.

[0063] Next, the adjustment coefficient for the spacing between drainage holes is obtained by referring to the function relationship table based on the current flood discharge atomization intensity level. For example, if the current flood discharge atomization intensity level is relatively strong, the corresponding adjustment coefficient can be found to be 0.7.

[0064] Then, a sequence for opening and closing the drainage valves is generated based on the adjustment coefficient for the spacing of the drainage holes. Specifically, the drainage holes on the slope can be numbered, the number of drainage holes to be opened can be determined according to the adjustment coefficient, and a sequence containing the opening order can be generated.

[0065] Then, the opening degree of the electric valve is controlled according to the opening and closing sequence of the drain valves. The valve opening degree of each drain hole can be adjusted sequentially by the electric actuator to achieve the staged opening of the drain holes.

[0066] Simultaneously, real-time drainage flow data is collected and fed back to the seepage calculation engine. Flow sensors can be installed at each drainage hole to continuously monitor drainage flow and transmit the data to the central control system.

[0067] Finally, the boundary conditions of the transient seepage field calculation model are corrected based on the drainage flow data. The seepage calculation engine can dynamically adjust the boundary condition parameters in the seepage field model based on real-time feedback of drainage flow data to improve calculation accuracy.

[0068] This application further proposes to identify abnormal displacement rate regions in the seepage pressure distribution map; match the filter layer thickness correction coefficient according to the displacement rate level; generate a set of construction instructions for the structural modification of the filter layer; generate collaborative optimization instructions by associating the parameters of adjacent drainage holes in the abnormal displacement rate regions; output a spatial allocation scheme for reinforcement materials; and verify the matching degree between the spatial allocation scheme for reinforcement materials and the displacement control target.

[0069] Anomaly regions in the seepage pressure distribution map are located using an image recognition algorithm. This algorithm establishes a region division standard based on the coupling relationship between seepage pressure gradient and displacement rate. Displacement rate levels are divided into three levels, corresponding to discrete values ​​of the filter layer thickness correction coefficient ranging from 0.5 to 1.5. For example, when the displacement rate exceeds a threshold of 20%, a correction coefficient of 1.2 is applied. The filter layer structure modification construction instruction set includes parameters for drilling depth, material filling density, and construction sequence. Drilling depth and the correction coefficient are linearly related; for example, a correction coefficient of 1.2 corresponds to a 15% increase in drilling depth. Collaborative optimization instructions are generated using a spatial topology analysis algorithm, correlating and optimizing the drainage hole flow parameters within a 50-meter radius of the anomaly region with the filter layer modification parameters. For example, when the drainage hole spacing is reduced by 10%, the filter layer thickness correction coefficient is simultaneously adjusted to 0.1. The reinforcement material spatial allocation scheme is generated using a resource optimization algorithm, performing non-uniform allocation based on the area proportion of the anomaly region and the material shear strength parameters. For example, when the area proportion exceeds 30%, 60% of the total material is allocated. The matching degree verification is performed through the finite element simulation model. The deviation between the predicted displacement value after reinforcement and the control target value is analyzed. When the deviation exceeds 5%, the scheme iterative optimization is triggered.

[0070] Specifically, after the seepage pressure distribution map is generated by the transient seepage field calculation model, abnormal displacement rate areas are located using a pressure gradient abrupt change point identification algorithm. For example, when the pressure difference between adjacent grid points exceeds a set threshold, it is marked as an abnormal area. The displacement rate level is determined by comparing and analyzing historical monitoring data with real-time data, for example, using the ratio of the current rate to the historical average rate as the classification basis. The filter layer thickness correction coefficient is obtained by querying a preset displacement rate-correction coefficient mapping table, for example, the coefficient for rate level 2 is 1.0. The construction instruction set is generated by an engineering parameter conversion algorithm, converting the correction coefficient into specific borehole diameter and filling material ratio parameters. The collaborative optimization instructions are generated by calculating the correlation matrix between drainage hole flow parameters and filter layer modification parameters, for example, establishing a linear relationship between the drainage hole spacing adjustment and the filter layer thickness correction. The reinforcement material allocation scheme is generated by a spatial weight allocation algorithm, calculating the material allocation ratio of each sub-region based on the spatial distribution density of the abnormal area. The matching degree verification is achieved by establishing a post-reinforcement displacement prediction model, comparing the predicted displacement with the control target value in real time, and automatically triggering the parameter adjustment instruction set when the deviation exceeds the allowable range. Therefore, a closed-loop verification mechanism is used to ensure the dynamic consistency between the reinforcement scheme and the engineering safety objectives. At the same time, spatial correlation optimization is used to achieve the synergistic effect between the drainage system and the filter layer structure, thereby improving the accuracy of slope stability control measures and the efficiency of resource utilization.

[0071] Preferably, firstly, abnormal displacement rate regions in the seepage pressure distribution map are identified. By performing image analysis on the seepage pressure distribution map, regions where the displacement rate exceeds a preset threshold are identified. For example, a displacement rate threshold of 5 mm / day can be set; when the displacement rate of a certain region exceeds this threshold, it is marked as an abnormal region.

[0072] Secondly, the thickness correction factor of the filter layer is matched according to the displacement rate level. A correspondence table between displacement rate and filter layer thickness correction factor is established. For example, when the displacement rate is in the range of 5-10 mm / day, the correction factor is 1.2; when the displacement rate is in the range of 10-20 mm / day, the correction factor is 1.5; and when the displacement rate exceeds 20 mm / day, the correction factor is 2.0.

[0073] Next, a set of construction instructions for the retrofitting of the filter layer structure is generated. Based on the filter layer thickness correction coefficient, the required increase in filter layer thickness is calculated, and specific construction instructions are generated. For example, adding a 300mm thick filter layer in area A, and adding a 450mm thick filter layer in area B, etc.

[0074] Then, collaborative optimization instructions are generated by associating the parameters of adjacent drainage holes in the displacement rate anomaly region. The distribution and parameters of drainage holes around the displacement rate anomaly region are analyzed to generate optimization instructions. For example, increasing the diameter of drainage hole C1 to 100mm, and adding two drainage holes with a spacing of 3m in region D, etc.

[0075] Next, a spatial distribution scheme for reinforcement materials is generated. Based on the distribution characteristics of areas with abnormal displacement rates, a spatial distribution scheme for reinforcement materials is formulated. For example, high-strength geogrids are laid in area E with a spacing of 0.5m, and shotcrete with a thickness of 100mm is sprayed in area F.

[0076] Finally, the matching degree between the spatial distribution scheme of the reinforcement materials and the displacement control target is verified. The effectiveness of the reinforcement scheme is verified through numerical simulation or field tests. For example, the stability coefficient of the reinforced slope is simulated, or the actual displacement change in the reinforced area is monitored to ensure that the expected displacement control target is achieved.

[0077] This application further proposes to monitor the change trajectory of the slope stability coefficient; activate the anchoring loading protocol when the slope stability coefficient enters the critical decline range; determine the number of anchoring force loading stages according to the decline rate of the slope stability coefficient; increase the anchoring force intensity in stages according to the time gradient; dynamically adjust the anchoring force loading rate based on displacement sensor feedback; and establish a three-dimensional response surface of anchoring force-displacement-time.

[0078] The critical descent range can be set to a stability coefficient between 1.05 and 1.15, and the trend of stability coefficient change is identified through a sliding window algorithm. The number of anchoring force loading stages is dynamically divided according to the descent rate; for example, when the descent rate reaches 0.05 / min, it is divided into three stages. The time gradient can be configured to increase the initial load by 30% every 10 minutes, forming a stepped loading curve. When the displacement sensor feedback adjusts the loading rate, a displacement rate threshold of 0.2 mm / s can be set as the adjustment trigger condition. The three-dimensional response surface is constructed using a cubic spline interpolation algorithm, fitting the anchoring force application, displacement increment, and time variable onto the surface.

[0079] Specifically, the anchoring loading protocol is automatically activated when the slope stability coefficient enters the preset critical descent range. The current descent rate is determined by real-time calculation of the first derivative of the stability coefficient, and a corresponding loading stage division strategy is matched. For example, at a descent rate of 0.08 / min, the system automatically generates a four-stage loading scheme, with each stage spaced 5 minutes apart and a load increase of 25%. During load application, displacement sensors monitor slope deformation in real time. When the displacement increment after a single loading exceeds a preset threshold, the system automatically extends the time interval of the next stage to 8 minutes. Load parameters, time parameters, and displacement data for all loading stages are recorded synchronously, and a three-dimensional response model is constructed using a surface generation algorithm. This model can display the displacement response surface characteristics corresponding to different anchoring force loading paths, providing a visual analysis basis for subsequent optimization. The anchoring force loading rate adjustment employs a closed-loop control algorithm, comparing real-time data from the displacement sensors with a preset target curve to dynamically adjust the output pressure of the hydraulic system.

[0080] Preferably, the trajectory of slope stability coefficient changes is monitored. Multiple inclinometers and displacement sensors are deployed on the slope surface to collect slope deformation data in real time. The data acquisition system continuously records the sensor readings, and slope stability analysis software is used to calculate the time series of the slope stability coefficient.

[0081] The anchoring loading protocol is activated when the slope stability coefficient enters the critical decrease range. The critical value for the stability coefficient is set to 1.2. When the calculated stability coefficient falls below this critical value three times consecutively, the anchoring loading protocol is automatically triggered.

[0082] The number of anchorage force loading stages is determined based on the rate of decrease of the slope stability coefficient. The rate of decrease of the stability coefficient is calculated. If the rate of decrease is greater than 0.01 / hour, the anchorage force loading is divided into 5 stages; if the rate of decrease is between 0.005 and 0.01 / hour, it is divided into 3 stages; and if the rate of decrease is less than 0.005 / hour, it is divided into 2 stages.

[0083] The anchoring force is applied in stages with a time gradient. For example, for 5 loading stages, each lasting 2 hours, the anchoring force is 20%, 40%, 60%, 80%, and 100% of the design maximum value, respectively.

[0084] The anchoring force loading rate is dynamically adjusted based on displacement sensor feedback. During each loading phase, the displacement sensor readings are monitored in real time. If the displacement increment exceeds a preset threshold, the anchoring force loading rate is reduced; if the displacement increment is below the preset threshold, the loading rate is appropriately increased.

[0085] A three-dimensional response surface based on anchoring force, displacement, and time is constructed. Anchoring force, displacement, and time data are input into 3D data visualization software to generate the 3D response surface. By analyzing the morphological characteristics of the surface, the optimal anchoring force loading path is identified.

[0086] This application further proposes to record the residual change rate data of the Newton iteration method. When the residual continuously exceeds the convergence criterion, the algorithm switching flag is triggered, the simplified slice method is called to reconstruct the slope stability equation, the safety factor of the potential sliding surface is calculated based on the Mohr-Coulomb criterion, the spatial coordinates of the sliding surface corresponding to the minimum safety factor are output, and the coordinates are fed back to the adaptive unstructured grid discretization module.

[0087] The residual change rate data can be obtained through the difference operation between two adjacent residual values ​​during the iteration process, for example, by using a second-order backward difference scheme to calculate the residual change rate. The trigger condition for the algorithm switching flag can be set to a preset value when the residual exceeds the convergence threshold for three consecutive iterations, for example, triggering the flag when the residual exceeds the threshold for three consecutive iterations. The reconstruction process of the simplified strip method can include the discretization of the slip surface geometry, for example, discretizing the slip surface into multiple strips and establishing force balance equations. The application of the Mohr-Coulomb criterion can be combined with the internal friction angle and cohesion parameters of the soil and rock mass, for example, using the strength index obtained from triaxial tests to calculate the ratio of anti-slip force to sliding force. The feedback of the spatial coordinates of the slip surface can be achieved through a coordinate mapping algorithm, for example, converting the slip surface coordinates into the displacement correction of the grid nodes.

[0088] Specifically, during the solution of the unsaturated-saturated coupled seepage control differential equations, residual change rate data is recorded and analyzed in real time. When the residual continuously exceeds a preset threshold, the calculation process automatically switches to the simplified slice method for slope stability analysis. By establishing a discretized slice mechanical equilibrium model, the safety factor of each potential sliding surface is calculated in conjunction with the strength parameters of the soil and rock mass. After the coordinates of the sliding surface corresponding to the minimum safety factor are extracted, displacement correction instructions for the grid nodes are generated through a coordinate transformation algorithm. These correction instructions are input into the adaptive unstructured grid discretization module, driving the grid nodes to be denser in the dangerous area of ​​the sliding surface. For example, when stress concentration is detected in the sliding surface area, the grid node density can be increased to 1.5 times the original density. The model after grid optimization re-enters the seepage field calculation process, forming a closed-loop feedback mechanism. This process effectively overcomes the convergence defect of the Newton iteration method under complex fracture conditions, and at the same time, through the dynamic coupling of mechanical constitutive relations and numerical calculations, it realizes the collaborative analysis of the seepage field and stress field.

[0089] As a preferred embodiment, the specific implementation of this application's solution is as follows: During the transient seepage field calculation, when the residual change rate of the Newton iteration method exceeds a preset threshold, an algorithm switching mechanism is triggered. Specifically, the residual change rate is calculated by the relative change in the absolute value of the residuals in three consecutive iterations. If this change exceeds the set value of the convergence criterion, it is determined to be an abnormal convergence. At this time, the simplified slice method is automatically activated to reconstruct the slope stability equation. The seepage pressure distribution data and the shear strength parameters of the soil and rock mass are substituted into the reconstruction equation, and the safety factor of the potential sliding surface is calculated based on the Mohr-Coulomb criterion. The spatial coordinates of the sliding surface are mapped to the adaptive grid discretization module through a three-dimensional spatial coordinate transformation algorithm, triggering a local grid refinement operation. The grid refinement area expands along the normal direction with the sliding surface coordinates as the center to form an ellipsoidal refinement area, and the grid size decreases exponentially. The reconstructed grid model re-enters the seepage field calculation process, forming a dynamic balance between calculation accuracy and convergence speed.

[0090] This application further proposes to deploy multiple sets of parallel monitoring profiles along the main sliding direction of the slope, determine the spatial envelope of the potential sliding surface through ground-penetrating radar scanning, deploy a high-density monitoring point array within the spatial envelope of the potential sliding surface, set the maximum allowable point spacing constraint for the high-density monitoring point array, implement a real-time cross-validation process for monitoring data, and establish an automatic abnormal data removal mechanism.

[0091] The spacing between parallel monitoring profiles can be set to 5-10 meters based on the slope geometry, with each monitoring profile containing 3-5 vertical boreholes. Ground-penetrating radar (GPR) scanning uses an 800MHz antenna for three-dimensional data acquisition, identifying soil-rock structural interfaces through abrupt changes in reflected wave amplitude. The deployment density of the high-density monitoring point array is set to 2-4 monitoring points per square meter, with the maximum allowable point spacing constrained to 0.3-0.8 meters. A real-time cross-validation process involves synchronously acquiring data from three adjacent sets of sensors; a data verification procedure is triggered when the deviation between any two sets of data exceeds 5%. An automatic outlier removal mechanism employs the three-sigma criterion, marking and removing data points that exceed the mean ± 3 standard deviations within three consecutive sampling periods.

[0092] Specifically, parallel monitoring profiles are arranged at equal intervals along the main slip direction, forming an observation network orthogonal to the potential slip direction, ensuring complete coverage of seepage parameters along the mechanically sensitive axis. The spatial envelope of the slip surface generated by ground-penetrating radar scans identifies abrupt changes in the rock and soil structure through electromagnetic wave reflection characteristics, providing a precise target area for subsequent monitoring point deployment. A high-density monitoring point array is deployed within the slip envelope area according to the maximum point spacing constraint, satisfying the need to capture changes in the seepage field gradient while avoiding data redundancy caused by excessive density. The real-time cross-validation process eliminates drift errors from single devices through synchronous comparison of multiple sensors, and the automatic outlier removal mechanism dynamically filters noisy data using a statistical outlier detection algorithm. For example, when the atomization intensity reaches level IV, the monitoring point array automatically increases the sampling frequency to 10Hz while tightening the maximum allowable point spacing constraint to 0.5 meters, ensuring accurate capture of transient seepage processes. This closed-loop control mechanism for spatial deployment and data processing improves the acquisition accuracy of key seepage parameters by more than 40% and reduces the false alarm rate of abnormal data to below 3%.

[0093] As a preferred embodiment, the specific implementation of this application is as follows: In determining the main sliding direction of the slope, firstly, the sliding vector field is fitted using historical displacement monitoring data, and the maximum displacement direction is extracted as the reference axis for the main sliding direction based on principal component analysis. Three sets of parallel monitoring profiles are arranged at equal intervals along this reference axis, with the profile spacing controlled within the range of one-fifth to one-quarter of the slope height. When scanning with ground-penetrating radar, the transmitting antenna frequency is set to 100MHz, and reflected wave signals are collected at 0.5-meter intervals along the profile direction. The dielectric constant abrupt change interface is obtained through inversion using the finite-difference time-domain method, generating a three-dimensional envelope surface of the potential sliding surface. Within the coverage area of ​​this envelope surface, a pore water pressure sensor array is arranged according to a triangular mesh topology, with the distance between adjacent nodes not exceeding 1.5 meters. The monitoring data is transmitted to the central processing unit in real time. The Pearson correlation coefficient method is used to perform data consistency checks on three adjacent monitoring points. When the correlation coefficient is lower than 0.85, abnormal data points are automatically marked, and the data is cleaned using a median filtering algorithm.

[0094] This application further proposes to obtain a rock mass fracture point cloud dataset through three-dimensional laser scanning; reconstruct the fracture aperture spatial distribution function based on the rock mass fracture point cloud dataset; calculate the topological connectivity index matrix of the rock mass fracture network; generate anisotropic permeability tensor eigenvalue spectrum; verify the reliability of the anisotropic permeability tensor eigenvalue spectrum through field water injection tests; and optimize the fracture parameter inversion algorithm based on the verification results.

[0095] Three-dimensional laser scanning can be performed using a phase-detection scanner, with scanning accuracy controlled to the 0.1 mm level and point cloud density exceeding 5000 points per square meter. The spatial distribution function of fracture aperture is constructed using a non-uniform rational B-spline surface fitting method, with a fitting error not exceeding 5% of the actual aperture value. The topological connectivity index matrix of the rock mass fracture network is calculated using graph theory algorithms, where nodes represent fracture intersections and edges represent fracture branches, and connectivity indices include node degree distribution and path length distribution. The anisotropic permeability tensor eigenvalue spectrum is generated using a tensor decomposition algorithm, with the deviation angle of the main permeability direction controlled within 3 degrees. The field water injection test adopts a segmented pressurization method, with the water injection pressure gradient set to 0.1 MPa per stage, and the flow rate monitoring accuracy reaching 0.01 liters per minute. The fracture parameter inversion algorithm optimization uses a Bayesian regularization method, with the number of iterations limited to within 100.

[0096] Specifically, the dense point cloud data acquired by 3D laser scanning provides the geometric basis for subsequent analysis. The point cloud dataset, after denoising, is input into the fracture aperture reconstruction module. The spatial distribution function of fracture aperture is generated using a surface interpolation algorithm, reflecting the microscopic undulations of the fracture wall. The calculation of the topological connectivity index matrix employs an adjacency matrix storage method, combined with a depth-first search algorithm to traverse the fracture network path. The permeability tensor eigenvalue spectrum is extracted using eigenvalue decomposition technology, automatically aligning the main permeation direction with the dominant fracture orientation. Field water injection tests are conducted at selected monitoring well locations, with injection pressure and seepage velocity data transmitted to the verification module in real time. During the inversion algorithm optimization process, the difference between the measured and calculated values ​​of the permeability tensor eigenvalues ​​is quantified as an objective function, and the fracture aperture parameters are adjusted using the gradient descent method. Thus, the spatial connectivity state of the fracture network is accurately quantified, the anisotropic characteristics of the permeability tensor are verified through physical experiments, and the parameter inversion process forms a closed-loop correction mechanism, ultimately ensuring that the boundary conditions of the seepage model match real geological conditions, improving the convergence and reliability of the seepage field calculation.

[0097] As a preferred embodiment, the specific implementation of this application is as follows: During the acquisition of slope rock mass fracture parameters, a ground-based three-dimensional laser scanning system is used to scan the target slope at multiple stations, forming a complete fracture point cloud dataset through target stitching. An improved moving cube algorithm is used to reconstruct the three-dimensional geometric model of the fracture surface. Based on curvature variation characteristics, the spatial distribution law of fracture aperture is identified, and a functional relationship between fracture aperture and spatial coordinates is constructed. A graph theory algorithm is used to abstract the fracture network into a node-edge topology, establishing an adjacency matrix containing the number of connected paths and the degree of branch nodes. The principal direction of the permeability tensor is calculated using tensor decomposition, and the anisotropic characteristic spectrum of the permeability coefficient is generated by combining the statistical results of fracture attitude. Water injection test well groups are arranged at different elevations of the slope, and the measured values ​​of the permeability coefficient are obtained using a stepped flow injection method. Residual analysis is performed between the measured values ​​and the numerical calculation results. The exponential correction term in the fracture aperture attenuation model is adjusted according to the residual distribution pattern, optimizing the weight allocation mechanism of the inversion algorithm.

[0098] Through the above technical solutions, this application achieves accurate digital reconstruction of the spatial distribution characteristics of rock mass fractures, overcoming the data deficiency problem of traditional geological sketching methods; it accurately characterizes the spatial distribution law of the dominant seepage path through topological connectivity quantitative analysis, eliminating the simplification error of the equivalent continuous medium model; the verification mechanism based on field tests effectively corrects the boundary condition deviation of the numerical model, ensuring the engineering applicability of the permeability tensor field; the closed-loop optimization algorithm significantly improves the computational efficiency and convergence accuracy of fracture parameter inversion, providing a reliable parameter basis for the stable solution of the seepage control equation.

[0099] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for analyzing slope seepage under flood discharge atomization conditions, characterized in that, include: S1: Collect slope pore water pressure data and matrix suction data through a distributed sensor array; S2: The data fusion unit compares the pore water pressure data with the dynamic critical pressure threshold and the matrix suction data with the dynamic critical suction threshold. When any data exceeds the limit, a seepage analysis trigger command is generated. S3: The seepage calculation engine starts the transient seepage field calculation model based on the fractured network topology data; S4: Generate a set of slope stability control parameters based on the seepage pressure distribution map output by the transient seepage field calculation model; S5: The execution system regulates the operation status of drainage facilities based on the slope stability control parameter set.

2. The method according to claim 1, characterized in that, S3 include: S3.1: Parse the pre-stored dataset of spatial distribution of rock fractures; S3.2: Calculation of the three-dimensional permeability tensor field based on the fracture aperture attenuation law; S3.3: Construct the control differential equation for unsaturated-saturated coupled seepage flow; S3.4: Adaptive unstructured grid discrete slope geometry model is adopted; S3.5: Define the convergence criteria for Newton's iteration method; S3.6: Solve the discretized algebraic equations to obtain the spatiotemporal distribution of the seepage field.

3. The method according to claim 2, characterized in that: The unsaturated-saturated coupled seepage control differential equation described in S3.3 is as follows: Where: K is the permeability coefficient tensor; h is the head function; C is the capacity function; t is the time variable.

4. The method according to claim 3, characterized in that, S2 include: S2.1: Establish a mapping database between flood discharge atomization intensity levels and dynamic critical pressure thresholds; S2.2: Monitor the current flood discharge atomization intensity level using acoustic sensors; S2.3: Based on the current flood discharge atomization intensity level, retrieve the dynamic critical pressure threshold from the mapping database; S2.4: When the pore water pressure data exceeds the dynamic critical pressure threshold, a pore water pressure over-limit command is generated; S2.5: When the matrix suction data is lower than the dynamic critical suction threshold, a matrix suction insufficient command is generated; S2.6: Combines the pore water pressure over-limit command and the matrix suction insufficient command to generate the seepage analysis trigger command.

5. The method according to claim 4, characterized in that, S5 include: S5.1: Construct a functional relationship table between the flood discharge atomization intensity level and the adjustment coefficient of the drainage hole spacing; S5.2: Obtain the drainage hole spacing adjustment coefficient by querying the function relationship table based on the current flood discharge atomization intensity level; S5.3: Generate the opening and closing sequence of the drain valve based on the drain hole spacing adjustment coefficient; S5.4: Control the opening degree of the electric valve according to the opening and closing sequence of the drain valve; S5.5: Real-time collection of drainage flow data and feedback to the seepage calculation engine; S5.6: Correct the boundary conditions of the transient seepage field calculation model based on the drainage flow data.

6. The method according to claim 3, characterized in that, S4 include: S4.1: Identify areas of abnormal displacement rate in the seepage pressure distribution map; S4.2: Match the filter layer thickness correction factor according to the displacement rate level; S4.3: Generate a set of construction instructions for the modification of the filter layer structure; S4.4: Generate collaborative optimization instructions for parameters of adjacent drainage holes in areas with abnormal displacement rates; S4.5: Output reinforcement material space allocation scheme; S4.6: Verify the matching degree between the spatial distribution scheme of reinforcement materials and the displacement control target.

7. The method according to claim 6, characterized in that, S4 also includes: S4.7: Monitor the trajectory of slope stability coefficient changes; S4.8: Activate the anchoring loading protocol when the slope stability coefficient enters the critical decrease range; S4.9: Determine the number of anchorage force loading stages based on the rate of decrease of the slope stability coefficient; S4.10: Apply anchoring force in stages according to time gradient; S4.11: Dynamically adjust the anchoring force loading rate based on displacement sensor feedback; S4.12: Establish the three-dimensional response surface of anchoring force-displacement-time.

8. The method according to claim 3, characterized in that, S3.5 includes: S3.5.1: Record the residual rate of change data of Newton's iterative method; S3.5.2: Trigger the algorithm switching flag when the residuals continuously exceed the convergence criterion; S3.5.3: Reconstruct the slope stability equation using the simplified slice method; S3.5.4: Calculate the safety factor of the potential sliding surface based on the Mohr-Coulomb criterion; S3.5.5: Output the sliding surface space coordinates corresponding to the minimum safety factor; S3.5.6: Feeds back the spatial coordinates of the sliding surface to the adaptive unstructured mesh discretization module.

9. The method according to claim 1, characterized in that, S1 includes: S1.1: Multiple sets of parallel monitoring profiles are set up along the main sliding direction of the slope; S1.2: Determine the spatial envelope of the potential slip surface through ground-penetrating radar scanning; S1.3: Deploy a high-density array of monitoring points within the spatial envelope of the potential sliding surface; S1.4: Set the maximum allowable point spacing constraint for the high-density monitoring point array; S1.5: Implement a real-time cross-validation process for monitoring data; S1.6: Establish an automatic mechanism for removing abnormal data.

10. The method according to claim 2, characterized in that, S3.1 includes: S3.1.1: Obtain rock mass fracture point cloud dataset through 3D laser scanning; S3.1.2: Reconstructing the spatial distribution function of fracture aperture based on rock mass fracture point cloud dataset; S3.1.3: Calculate the topological connectivity index matrix of the rock mass fracture network; S3.1.4: Generate the eigenvalue spectrum of the anisotropic permeation tensor; S3.1.5: Verify the reliability of the anisotropic permeability tensor eigenvalue spectrum through on-site water injection tests; S3.1.6: Optimize the fracture parameter inversion algorithm based on the verification results.

Images