Power transmission line tower slope instability monitoring and early warning method and system under heavy rainfall condition

Abstract

The invention discloses a power transmission line tower slope instability monitoring and early warning method and system under a heavy rainfall condition, and relates to the technical field of power transmission line safety monitoring, and the method comprises the steps: dividing DEM grids according to the coordinates of a power transmission line tower, deploying a multi-source sensor, calculating the catchment area and the soil layer thickness of each grid, and obtaining the rock-soil parameters; forecasting areas are divided according to the power transmission line, hourly rainfall intensity is calculated, rainfall intensity of a monitoring unit is obtained through interpolation, and further future effective rainfall is predicted; and performing depth discrete processing on the grids, establishing a one-dimensional Richards equation in the normal direction of the slope surface, performing implicit difference solution to output a pressure head profile, and calculating a physical instability index of the monitoring unit based on the pressure head profile and rock-soil parameters. According to the method, refined early warning of the tower slope scale is realized, the actual hydraulic load degree of different tower slopes is accurately reflected, the forecasting timeliness and physical consistency are improved, and the early warning reliability and interpretability are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power transmission line safety monitoring, in particular to a power transmission line tower slope instability monitoring and early warning method and system under heavy rainfall conditions. BACKGROUND

[0002] After the large-scale extension of the power transmission line to the mountainous, hilly and complex geological areas, the instability landslide of the tower foundation surrounding slope under the action of heavy rainfall has become one of the main disaster factors of power grid operation. In the existing engineering practice, on the one hand, rain gauges, inclinometers, GNSS displacement piles and other monitoring devices are widely deployed, and the landslide hidden danger investigation and qualitative zoning are carried out combined with digital elevation model (DEM), geological survey data; on the other hand, the rainfall duration-intensity (I-D) threshold based on empirical statistics, the effective rainfall amount index in the early stage and the simplified infinite slope stability calculation model have been used to construct a regional geological disaster warning system. At the same time, numerical weather prediction, three-dimensional GIS, Internet of Things communication and other technologies have also been gradually introduced into the field of power transmission line disaster prevention and reduction, which are used to provide short-term rainfall prediction and visual display of the tower environment. In recent years, some research has begun to explore the introduction of machine learning algorithms into landslide susceptibility evaluation and disaster prediction, but most of them stay at the regional grid or engineering example level, and a mature engineering application system has not yet been formed at the fine scale of tower slope. However, the existing power transmission line slope monitoring and early warning technology under heavy rainfall conditions still has some deficiencies. Most of the systems use meteorological station rainfall or coarse grid numerical prediction, and do not finely map the instantaneous precipitation flux and hourly rainfall intensity to specific tower monitoring units in high-resolution space, which is difficult to reflect the spatial heterogeneity of rainfall-concentration under complex terrain, and lacks the depth evolution of geological water infiltration, which is difficult to form a closed loop coupling with real-time monitoring data. SUMMARY

[0003] In view of the above existing problems, the present application is proposed.

[0004] Therefore, the present application provides a power transmission line tower slope instability monitoring and early warning method and system under heavy rainfall conditions, which solves the problem that the existing technology is difficult to reflect the spatial heterogeneity of rainfall-concentration under complex terrain, and lacks the depth evolution of geological water infiltration, which is difficult to form a closed loop coupling with real-time monitoring data.

[0005] To solve the above technical problems, the present application provides the following technical solutions: In a first aspect, the present application provides a power transmission line tower slope instability monitoring and early warning method under heavy rainfall conditions, which comprises, dividing the DEM grid according to the coordinates of the power transmission line tower to deploy multi-source sensors and define monitoring units, calculating the catchment area and soil thickness of each grid and obtaining rock and soil parameters for preprocessing; According to the power transmission line, a forecast area is divided, and the instantaneous precipitation flux of the forecast area is predicted to calculate the hourly rainfall intensity, the rainfall intensity of the monitoring unit is obtained by interpolation, and then the vertical pore pressure transmission time scale is calculated to obtain the attenuation coefficient to predict the future effective rainfall; The grid is deeply dispersed, a one-dimensional Richards equation along the slope normal is established, and an implicit difference solution is output to obtain the pressure head profile, and the physical instability index of the monitoring unit is calculated based on the pressure head profile and the rock-soil parameters; A feature vector is constructed for each monitoring unit, and an XGBoost-based landslide probability prediction model is constructed to output the landslide probability, and the comprehensive risk index is calculated by combining the physical instability index of the monitoring unit and the landslide probability to perform risk classification and three-dimensional visualization early warning.

[0006] As a preferred scheme of the power transmission line tower slope instability monitoring and early warning method under the condition of heavy rainfall, wherein: the DEM grid is divided according to the coordinates of the power transmission line tower, the multi-source sensor is deployed to define the monitoring unit, the catchment area and soil thickness of each grid are calculated, and the rock-soil parameters are obtained for preprocessing, the coordinates of the power transmission line tower are obtained, the DEM grid is divided at a resolution of 10m*10m, the multi-source sensor is deployed at each power transmission tower, the planar distance from each center point of the DEM grid to the power transmission tower is calculated, the DEM grid within the control radius of the power transmission tower Is distributed to the nearest power transmission tower to form a monitoring unit and collect multi-source data, the terrain slope in the 3*3 field of each grid belonging to the monitoring unit is calculated using finite difference, and the slope angle And slope direction Are further calculated; The catchment area And soil thickness Of each grid are calculated; The geological survey data of the power transmission line are obtained through geological survey, the rock-soil parameters of each type of rock-soil in the power transmission line are obtained, including cohesion , internal friction angle , saturated wet soil bulk density , saturated hydraulic conductivity , and specific water storage , and the water diffusion coefficient Is calculated synchronously; After the collected multi-source data is filtered by moving average and outliers are removed, it is packaged and output.

[0007] As a preferred scheme of the tower slope instability monitoring and early warning method of the power transmission line under the strong rainfall condition, wherein: the power transmission line is taken as the skeleton to form a matrix prediction area by buffering 50km to both sides of the line, the rectangular prediction area is divided into a 1km*1km mode grid, and the center coordinates of each mode grid are recorded ; The 6-hour large-scale reanalysis field is obtained from the National Numerical Weather Prediction Center, and the reanalysis field is recorded as a background field ; The same meteorological data as the background field is collected by a multi-source sensor to form an observation vector y, and the observation vector is assimilated into an analysis field using the GSI 3DVar algorithm ; A precipitation prediction model is constructed using an LSTM network, and the model input is set to the analysis field and the output is set to the predicted precipitation flux The precipitation prediction model is trained and the predicted precipitation flux is output as the hourly rainfall intensity .

[0008] As a preferred scheme of the tower slope instability monitoring and early warning method of the power transmission line under the strong rainfall condition, wherein: the rainfall intensity of the monitoring unit is obtained by interpolation, and then the vertical pore pressure transmission time scale is calculated to obtain the decay coefficient to predict the future effective rainfall amount, and the area-weighted centroid coordinates of each monitoring unit are calculated ; The mode grid containing the centroid coordinates is searched according to the area-weighted centroid coordinates of the monitoring unit, the corner point coordinates are recorded, and the normalized coordinates of the centroid relative to the lower left corner point of the mode grid are calculated The rainfall intensity of the monitoring unit is obtained by using a bilinear interpolation formula ; The representative average soil thickness in the monitoring unit is calculated And the average catchment area ; The dominant rock type in the monitoring unit is taken as the dominant lithology to obtain the moisture diffusion coefficient , and then the vertical pore pressure transmission time scale of the monitoring unit is calculated ; According to the vertical pore pressure transmission time scale The decay coefficient coupled with the vertical time scale is defined ; The future 6-hour effective rainfall amount is calculated in an exponential decay manner according to the rainfall intensity of the monitoring unit and the decay coefficient.

[0009] As a preferred embodiment of the monitoring and early warning method for slope instability of transmission line towers under heavy rainfall conditions described in this invention, the following steps are performed: The grid is deeply discretized, a one-dimensional Richards equation along the slope normal is established, and an implicit difference solution is used to output the pressure head profile. Based on the pressure head profile and soil parameters, the physical instability index of the monitoring unit is calculated, which is determined by the water diffusion coefficient of the dominant lithology. and average catchment area Calculate the transverse pore pressure transmission time scale ; Dimensionless scale parameters are calculated by combining the vertical and transverse pore pressure transmission time scales. And define the credibility weights of the physical model. ; For each DEM grid g, from the surface Z=0 to the bottom of the soil Z= Establish a fixed number of depth layers Calculate the node position at depth m. ; The pressure head of the infiltration process is defined as follows: Establish a one-dimensional Richards equation along the slope normal; For grid g belonging to monitoring unit i at the depth node Calculate the initial pressure head profile ; Discretize the Richards equation for the depth node m; The complete indenter profile is obtained by iteratively linearizing and solving the one-dimensional Richards equation using backward Euler implicit difference in time and second-order central difference in space. ; For mesh g at depth Z= Calculate the safety factor ; For all DEM grids in monitoring unit i at time... Take the minimum safety factor and normalize it to the physical instability index .

[0010] As a preferred embodiment of the monitoring and early warning method for slope instability of transmission line towers under heavy rainfall conditions described in this invention, the following steps are taken: A feature vector is constructed for each monitoring unit, and a landslide probability prediction model is built based on XGBoost to output the landslide probability index. Multi-source data and geotechnical parameters are then spliced ​​together to form the feature vector of monitoring unit i. A landslide probability prediction model was constructed using XGBoost, employing a standard gradient boosting tree binary classification structure. The overall loss function was defined by combining logistic regression loss with a complexity regularization term to train the landslide probability prediction model. The feature vectors were then used to... The landslide probability prediction model outputs the landslide probability after being trained. .

[0011] As a preferred embodiment of the monitoring and early warning method for slope instability of transmission line towers under heavy rainfall conditions described in this invention, the method involves: calculating a comprehensive risk index based on the physical instability index and landslide probability of the comprehensive monitoring unit, performing risk classification, and providing three-dimensional visualization for early warning. ; After obtaining the comprehensive risk index, a risk threshold is set to map the comprehensive risk index of the monitoring unit to the risk level output. In the 3D GIS platform, the DEM and transmission line data are loaded into a 3D scene and rendered in the 3D scene using color according to the risk level, and a visual early warning signal is output to the staff at the same time.

[0012] Secondly, this invention provides a monitoring and early warning system for slope instability of transmission line towers under heavy rainfall conditions, comprising, The data processing module is used to divide the DEM grid according to the coordinates of the transmission line towers, deploy multi-source sensors as monitoring units, calculate the catchment area and soil thickness of each grid, and obtain geotechnical parameters for preprocessing. The rainfall prediction module is used to divide the forecast area according to the transmission line, predict the instantaneous precipitation flux of the forecast area, calculate the hourly rainfall intensity, obtain the rainfall intensity of the monitoring unit through interpolation, and then calculate the attenuation coefficient to predict the future effective rainfall. The geological analysis module is used to perform in-depth discretization of the grid, establish a one-dimensional Richards equation along the slope normal, perform implicit difference solution to output the pressure head profile, and calculate the physical instability index of the monitoring unit based on the pressure head profile and soil parameters. The analysis and early warning module is used to construct feature vectors for each monitoring unit, and to build a landslide probability prediction model based on XGBoost to output the landslide probability. It also calculates a comprehensive risk index by combining the physical instability index of the monitoring unit and the landslide probability to perform risk classification and three-dimensional visualization early warning.

[0013] Thirdly, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the monitoring and early warning method for slope instability of transmission line towers under heavy rainfall conditions as described in the first aspect of the present invention.

[0014] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the method for monitoring and early warning of instability of transmission line tower slopes under heavy rainfall conditions as described in the first aspect of the present invention.

[0015] The beneficial effects of this invention are as follows: This invention constructs a monitoring unit by using DEM grid and multi-source sensors, calculates the physical instability index by combining instantaneous precipitation flux and one-dimensional Richards equation, and outputs a comprehensive risk index by combining XGBoost landslide probability, thereby achieving refined early warning at the tower slope scale, accurately reflecting the actual hydraulic load on different tower slopes, improving forecast timeliness and physical consistency, and significantly enhancing the reliability and interpretability of early warning. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the monitoring and early warning method for slope instability of transmission line towers under heavy rainfall conditions in Example 1.

[0018] Figure 2 This is a structural diagram of the monitoring and early warning system for slope instability of transmission line towers under heavy rainfall conditions in Example 1. Detailed Implementation

[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0020] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0021] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0022] Example 1, referring to Figure 1 and Figure 2 This is the first embodiment of the present invention, which provides a method for monitoring and early warning of slope instability of transmission line towers under heavy rainfall conditions, including the following steps: S1. Based on the coordinates of the transmission line towers, divide the DEM grid and deploy multi-source sensors as monitoring units. Calculate the catchment area and soil thickness of each grid and obtain geotechnical parameters for preprocessing. Specifically, based on the coordinates of the transmission line towers, a DEM grid is divided, and multi-source sensors are deployed as monitoring units. The catchment area and soil thickness of each grid are calculated, and geotechnical parameters are obtained for preprocessing. This involves acquiring the tower coordinates, dividing the DEM grid at a resolution of 10m x 10m, and deploying multi-source sensors on each tower, including rain gauges, triaxial inclinometers, tower base inclinometers, GNSS displacement piles, anemometers, and thermo-hygrometers. These sensors are deployed both on and around the tower. The planar distance from the center point of each DEM grid to the tower is calculated, and the area within the control radius of the tower is determined. The DEM grid within the monitoring unit is assigned to the transmission tower with the smallest distance to form a monitoring unit, and multi-source data is collected. For each grid belonging to the monitoring unit, the terrain slope is calculated using finite difference within a 3*3 neighborhood, and the slope angle is further calculated. and slope : in It is an east-west slope. It is a north-south slope; Calculate the catchment area for each grid. and soil layer thickness : in For grid area, For the set of upstream cells that converge to grid g, the water flow outlet direction of each grid is determined and the upstream cells are identified by performing the D8 unidirectional flow direction algorithm on the DEM grid. Let h be the catchment area of ​​the upstream grid cell. This refers to relative slope position factors (e.g., 0.8 at the top, 1.0 at the middle, and 1.2 at the angle). Typical soil layer thicknesses are obtained by querying the corresponding lithology-geomorphology combination in the preliminary survey report. Geological survey data of the transmission lines were obtained through geological exploration, and soil and rock parameters (indoor experiments) were acquired for each type of soil and rock in the transmission lines, including cohesion. (Unit: kPa), internal friction angle (Unit: °), saturated wet soil unit weight (Unit: kN / m) 3 ), saturated hydraulic conductivity (Unit: m / s) and specific water storage (Dimensionless, representing the amount of water stored per unit volume of soil per unit pressure head change), simultaneously calculate the moisture diffusion coefficient. (Indicates the diffusion rate of water in this rock and soil medium): Where j represents the type of soil or rock; The collected multi-source data is filtered by moving average and outlier removal before being packaged and output.

[0023] By dividing the digital elevation model into grids centered on transmission towers and constructing monitoring units, topographic features, soil and rock parameters, and multi-source sensor observations are managed uniformly under the same spatial reference system. This avoids the misalignment of monitoring points with the actual affected slopes in traditional methods, improving the spatial accuracy of risk identification from the source. By calculating slope, angle, and aspect within a small neighborhood, local topographic noise is smoothed, making the geometric information relied upon for subsequent stability analysis and water catchment path analysis more reliable and precise, facilitating the accurate identification of the most unfavorable sliding direction for the towers. Using recursive water catchment area calculations and slope position factors to correct soil thickness, an approximate restoration of the actual soil cover differences at different locations such as the slope top, slope waist, and slope toe is achieved even in the absence of intensive drilling, allowing rainwater runoff capacity and potential sliding depth to be reasonably reflected at the grid scale. Combined with geological surveys, strength and hydraulic parameters of various soils and rocks are uniformly obtained, and the diffusion capacity of water in the medium is characterized accordingly, providing a solid physical basis for subsequent vertical pore pressure propagation timescales and infiltration evolution. Finally, the multi-source monitoring data were filtered and smoothed to remove outliers, which significantly reduced the interference of noise on sensitive features such as displacement velocity and stress change rate. This made the subsequent physical model calculations and machine learning predictions more stable and reliable, thereby improving the accuracy and engineering usability of early warning of tower slope instability under heavy rainfall conditions.

[0024] S2. Divide the forecast area according to the transmission line, predict the instantaneous precipitation flux of the forecast area, calculate the hourly rainfall intensity, obtain the rainfall intensity of the monitoring unit by interpolation, and then calculate the vertical pore pressure transmission time scale to obtain the attenuation coefficient and predict the future effective rainfall. Specifically, the forecast area is divided according to the transmission lines, and the hourly rainfall intensity is calculated by predicting the instantaneous precipitation flux of the forecast area. The matrix forecast area is formed by buffering 50km on both sides of the transmission lines as the framework, and the rectangular forecast area is divided into 1km*1km model grids (1km resolution can distinguish the key topographic features that control strong convection and orographic lifting). The center coordinates of each model grid are recorded. ; in and For pattern grid coordinates, and The number of vertical grid divisions, where p and q are the pattern grid labels; Large-scale reanalysis fields (e.g., global grid data of 0.25°*0.25°) are obtained from the National Numerical Weather Prediction Center every 6 hours, and these reanalysis fields are denoted as the background field. It includes three-dimensional meteorological elements such as wind field, temperature, and humidity; Meteorological data identical to the background field are collected from multiple sensors to form an observation vector y. The GSI 3DVar algorithm is then used to assimilate the observation vector into an analysis field. : in The observation operator is used to project model variables into the observation space (e.g., station data interpolated from grid points, wind direction and speed obtained from a 3D wind field). It defines a nonlinear observation operator for each variable in the model state, mapping it to the observation space. The nonlinear observation operator is then linearized and approximated, and the data is extracted from the background field. Obtaining the Jacobian matrix As the gain matrix, the background covariance matrix B is extracted using the NMC method, and the observation error covariance matrix R is extracted for the observed variables. , For transpose; A precipitation prediction model is constructed using an LSTM network. The model input is the analysis field, and the output is the predicted precipitation flux. (Unit: mm / h) The precipitation prediction model is trained and the predicted precipitation flux is output as the hourly rainfall intensity. .

[0025] A rectangular forecasting area of ​​50 kilometers on each side of the transmission line is constructed, and a model grid with a resolution of 1 kilometer is used. This enables the analysis of key topographic units that control strong convection and orographic uplift at a scale directly covering the transmission corridor, ensuring that the spatial distribution of precipitation forecasts is highly consistent with the actual rainfall-receiving area of ​​the towers. By introducing the reanalysis field from the National Numerical Weather Prediction Center as the background field, and fusing observations of wind, temperature, and humidity collected by multi-source sensors along the line into the analysis field using a three-dimensional variational assimilation algorithm, the systematic bias of large-scale forecasts under complex local terrain and underlying surface conditions is corrected. Furthermore, the initial field of the model is ensured to more closely reflect local weather conditions near the transmission line. Based on this, the analysis field sequence is used as input to a long short-term memory network to directly learn the nonlinear mapping relationship between analysis field elements and subsequent precipitation fluxes. This allows the precipitation forecast to inherit the dynamic consistency of the physical assimilation field while using a data-driven approach to compensate for the shortcomings of traditional parameterization schemes in characterizing small- and medium-scale precipitation structures. Finally, hourly rainfall intensity fields were obtained at the model grid scale, providing high spatiotemporal resolution and physically consistent rainfall input for subsequent accurate projection of rainfall load onto tower monitoring units, driving infiltration-stability evolution and risk assessment, thereby significantly improving the accuracy and reliability of early warning of instability of transmission line tower slopes under heavy rainfall conditions.

[0026] Furthermore, the rainfall intensity of the monitoring unit is obtained through interpolation, and then the attenuation coefficient is calculated based on the vertical pore pressure transmission time scale to predict the future effective rainfall. This is achieved by calculating the area-weighted centroid coordinates of each monitoring unit. : in The coordinates of the grid center in the monitoring unit. For the set of coordinates of the monitoring unit, The coordinates of the centroid of monitoring unit i in the plane; Based on the area-weighted centroid coordinates of the monitoring unit, find the corner coordinates of the model grid containing the centroid coordinates, and calculate the normalized coordinates of the centroid relative to the lower left corner of the model grid. The rainfall intensity of the monitoring unit was obtained using a bilinear interpolation formula. ; For calculating the representative average soil layer thickness in the monitoring unit (unit: m) and average catchment area : The water diffusion coefficient was obtained by using the dominant soil and rock type in the monitoring unit as the dominant lithology. (unit: m) 2 / s), and then calculate the vertical pore pressure transmission timescale of the monitoring unit. : Based on the vertical pore pressure transmission timescale Define the attenuation coefficient coupled with the vertical time scale. : in The adjustment factor is dimensionless; in this invention, it is set to 0.5, representing approximately one to several times the previous value. Rainfall over the current timescale still has an impact on current pore pressure; The effective rainfall for the next 6 hours is calculated using an exponential decay method based on the rainfall intensity and attenuation coefficient of the monitoring unit. in This represents the estimated effective rainfall over the next 6 hours at the current moment.

[0027] Using the area-weighted centroid of the center point of each grid within the monitoring unit as the representative location of rainfall, the rainfall intensity of the model grid is accurately mapped to the monitoring unit through bilinear interpolation. This achieves spatial downsampling from one-kilometer numerical forecast grid points to slope units at the tens-meter level, avoiding abrupt changes and biases caused by simply taking the nearest grid point or a simple average. This ensures that each pole slope receives spatiotemporal rainfall input more consistent with its actual rainfall conditions. Based on this, by calculating representative average values ​​of soil thickness and catchment area within the monitoring unit and combining this with the water diffusion coefficient of the dominant lithology, a vertical pore pressure transfer timescale is constructed. This solves the technical problem of traditional effective rainfall models not distinguishing between the "fast-slow response" of different slopes, providing a clear physical basis for the length of the slope's memory of rainfall. Furthermore, this timescale is converted into an attenuation coefficient directly coupled to the pore pressure response, rather than using a uniform empirical constant. This allows the same rainfall process to exhibit differentiated effective rainfall evolution characteristics across different monitoring units, highlighting the high-risk cumulative effect of thick, low-permeability, and unfavorable slopes after continuous rainfall. Finally, by using the predicted rainfall intensity for the next six hours to calculate the forward-looking effective rainfall in an exponential decay manner, the "future rainfall" is linked with the "current hydrological-soil condition," enabling the early warning system to identify tower slopes that are close to the instability threshold under long-term cumulative effects, even if the future rainfall intensity is not extreme. This reduces false alarms based solely on rainfall thresholds and improves the ability to identify slowly changing and delayed instability disasters in a targeted manner.

[0028] S3. Perform deep discretization on the grid, establish a one-dimensional Richards equation along the slope normal, perform implicit difference solution to output the pressure head profile, and calculate the physical instability index of the monitoring unit based on the pressure head profile and soil parameters. Specifically, the grid is deeply discretized, and a one-dimensional Richards equation along the slope normal is established. An implicit difference solution is then performed to output the pressure head profile. Based on the pressure head profile and soil parameters, the physical instability index of the monitoring unit is calculated, which is determined by the water diffusion coefficient of the dominant lithology. and average catchment area Calculate the transverse pore pressure transmission time scale : Dimensionless scale parameters are calculated by combining the vertical and transverse pore pressure transmission time scales. And define the credibility weights of the physical model. : For each DEM grid g, from the surface Z=0 to the bottom of the soil Z= Establish a fixed number of depth layers Calculate the node position at depth m. : in For depth step size; The pressure head of the infiltration process is defined as follows: Establish the one-dimensional Richards equation along the slope normal: in The hydraulic conductivity coefficient, , Let g be the saturated hydraulic conductivity of the grid. These are empirical shape parameters, calibrated experimentally, used to describe the variation of unsaturated hydraulic conductivity with pressure head. For specific water capacity, , For moisture content, For the slope angle, This represents the hydraulic gradient along the slope direction; For grid g belonging to monitoring unit i at the depth node Calculate the initial pressure head profile : in This represents the typical pressure head of lithology under relatively dry conditions (generally a large negative value, in meters). The pressure head is in saturation state (close to 0m). The effective rainfall in the preceding period for monitoring unit i is calculated using historical data. For reference to effective rainfall, when The soil was nearly saturated at that time. This is an adjustment coefficient used to control the rate of change of the exponential function; Discretize the Richards equation for depth node m: in and For node m in time and The pressure head, and Let m be the water flow rate between m and its adjacent nodes; The and Defined as: in The hydraulic conductivity between m and m+1 is taken as the average hydraulic conductivity between the two depth nodes; The complete indenter profile is obtained by iteratively linearizing and solving the one-dimensional Richards equation using backward Euler implicit difference in time and second-order central difference in space. ; For mesh g at depth Z= Calculate the safety factor : in For friction, For water pressure, This is the cohesive term; The friction item Calculated using the internal friction angle and the slope angle: The water pressure item Calculations based on the indenter profile: in For DEM grid g wet soil unit weight, It is water-weighted. For depth Z= The indenter profile; The cohesion term Calculation based on soil cohesion: in The soil cohesion of the DEM mesh g; For all DEM grids in monitoring unit i at time... Take the minimum safety factor and normalize it to the physical instability index : in This is the minimum safety factor for monitoring unit i.

[0029] By uniformly converting the soil layer thickness, catchment area, and soil moisture diffusion capacity of each monitoring unit into vertical and lateral pore pressure transfer time scales, and constructing dimensionless scale parameters and model reliability weights, a quantitative characterization of whether a one-dimensional infiltration model is applicable to the slope is achieved. This solves the problem of applying physical models indiscriminately and failing to differentiate the reliability of results in traditional methods. Based on this, the grid is discretized at a fixed number of layers from the surface to the bottom of the soil layer, and a one-dimensional infiltration control equation is established along the slope normal. An implicit iterative solution strategy using backward Euler time discretization and second-order spatial difference is employed, which maintains numerical stability under conditions of heavy rainfall and large pore pressure gradients. This yields a complete pore pressure profile that evolves with time and depth, rather than simply using simplified empirical pore pressure increments, significantly improving the physical accuracy of pore pressure-stability analysis. Furthermore, at potential sliding surface locations, the friction, water pressure, and cohesion terms are decomposed using pore pressure profiles and parameters such as cohesion, internal friction angle, soil weight, and slope angle. Safety factors for each grid are calculated, and the most unfavorable safety factor is extracted at the monitoring unit scale and normalized into a physical instability index. This achieves a transformation from "local profile calculation" to "overall risk characterization of the slope controlled by the tower base." This approach not only identifies the delayed and cumulative instability trends caused by rainfall infiltration but also provides rigid physical quantity support for subsequent probabilistic models and risk classification. Overall, it constructs a self-verifying, applicable, infiltration-pore pressure-shear integrated framework for characterizing slope physical risks.

[0030] S4. Construct a feature vector for each monitoring unit, and build a landslide probability prediction model based on XGBoost to output the landslide probability. Calculate a comprehensive risk index by combining the physical instability index of the monitoring unit and the landslide probability to perform risk classification and three-dimensional visualization early warning. Specifically, a feature vector is constructed for each monitoring unit, and a landslide probability prediction model is built based on XGBoost to output the landslide probability index. Multi-source data and geotechnical parameters are spliced ​​together to form the feature vector of monitoring unit i. A landslide probability prediction model was constructed using XGBoost, employing a standard gradient boosting tree binary classification structure. The overall loss function was defined by combining logistic regression loss with a complexity regularization term to train the landslide probability prediction model. The feature vectors were then used to... The landslide probability prediction model outputs the landslide probability after being trained. .

[0031] Furthermore, by comprehensively calculating a comprehensive risk index based on the physical instability index and landslide probability of the monitoring unit, risk classification and three-dimensional visualization early warning are achieved. : in These are weighting coefficients used to assign weights to the landslide probability; After obtaining the comprehensive risk index, a risk threshold is set to map the comprehensive risk index of the monitoring unit to the risk level output. In the 3D GIS platform, the DEM and transmission line data are loaded into a 3D scene and rendered in the 3D scene using color according to the risk level, and a visual early warning signal is output to the staff at the same time.

[0032] This embodiment also provides a monitoring and early warning system for slope instability of transmission line towers under heavy rainfall conditions, including: The data processing module is used to divide the DEM grid according to the coordinates of the transmission line towers, deploy multi-source sensors as monitoring units, calculate the catchment area and soil thickness of each grid, and obtain geotechnical parameters for preprocessing. The rainfall prediction module is used to divide the forecast area according to the transmission line, predict the instantaneous precipitation flux of the forecast area, calculate the hourly rainfall intensity, obtain the rainfall intensity of the monitoring unit through interpolation, and then calculate the attenuation coefficient to predict the future effective rainfall. The geological analysis module is used to perform in-depth discretization of the grid, establish a one-dimensional Richards equation along the slope normal, perform implicit difference solution to output the pressure head profile, and calculate the physical instability index of the monitoring unit based on the pressure head profile and soil parameters. The analysis and early warning module is used to construct feature vectors for each monitoring unit, and to build a landslide probability prediction model based on XGBoost to output the landslide probability. It also calculates a comprehensive risk index by combining the physical instability index of the monitoring unit and the landslide probability to perform risk classification and three-dimensional visualization early warning.

[0033] This embodiment also provides a computer device applicable to the monitoring and early warning method for slope instability of transmission line towers under heavy rainfall conditions, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the monitoring and early warning method for slope instability of transmission line towers under heavy rainfall conditions as proposed in the above embodiment.

[0034] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0035] This embodiment also provides a storage medium storing a computer program. When executed by a processor, the program implements the method for monitoring and early warning of slope instability of transmission line towers under heavy rainfall conditions as proposed in the above embodiment. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0036] In summary, this invention constructs monitoring units using DEM grids and multi-source sensors, calculates the physical instability index by combining instantaneous precipitation flux with the one-dimensional Richards equation, and outputs a comprehensive risk index using XGBoost landslide probability. This enables refined early warning at the tower slope scale, accurately reflects the actual hydraulic load on different tower slopes, improves forecast timeliness and physical consistency, and significantly enhances the reliability and interpretability of early warning.

[0037] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for monitoring and early warning of slope instability of a power transmission line tower under heavy rainfall conditions, characterized in that: The method comprises the following steps: According to the coordinates of the transmission line tower, the DEM grid is divided, the multi-source sensor is deployed, and the monitoring unit is defined. The water catchment area and soil thickness of each grid are calculated, and the geotechnical parameters are obtained for preprocessing; According to the transmission line, the forecast area is divided, and the instantaneous precipitation flux of the forecast area is predicted to calculate the hourly rainfall intensity. The rainfall intensity of the monitoring unit is obtained by interpolation, and then the vertical pore pressure transmission time scale is calculated to obtain the attenuation coefficient to predict the future effective rainfall. The grid is deeply discretized, a one-dimensional Richards equation along the normal direction of the slope surface is established, and an implicit difference solution is obtained to output the pressure head profile. Based on the pressure head profile and the geotechnical parameters, the physical instability index of the monitoring unit is calculated. For each monitoring unit, a feature vector is constructed, and a landslide probability prediction model is built based on XGBoost to output the landslide probability. The comprehensive risk index is calculated by comprehensively considering the physical instability index of the monitoring unit and the landslide probability to perform risk classification and three-dimensional visualization early warning.

2. The method for monitoring and early warning of slope instability of a tower of a power transmission line under heavy rainfall conditions according to claim 1, characterized in that: The definition of the deployment of multi-source sensors according to the DEM grid division of the transmission line tower coordinates is defined as a monitoring unit, the catchment area and soil thickness of each grid are calculated and the rock-soil parameters are obtained for preprocessing, the transmission line tower coordinates are obtained, the DEM grid is divided at a resolution of 10m*10m, multi-source sensors are deployed at each transmission tower, the planar distance from each DEM grid center point to the transmission tower is calculated, the DEM grids within the control radius of the transmission tower are assigned to the transmission tower with the smallest distance to form a monitoring unit and collect multi-source data, the terrain slope in the 3*3 field of each grid belonging to the monitoring unit is calculated using finite difference, and the slope angle and slope direction are further calculated; Calculate catchment area for each grid and soil thickness ; Obtaining geological survey data of the transmission line through geological survey, obtaining rock-soil parameters of each type of rock-soil in the transmission line, including cohesion , internal friction angle , saturated wet soil bulk density , saturated water permeability and specific water storage , synchronously calculating the moisture diffusion coefficient ; After the collected multi-source data is filtered by moving average and outliers are removed, it is packaged and output.

3. The method for monitoring and early warning of slope instability of transmission line tower under heavy rainfall condition according to claim 2, characterized in that: The method comprises the following steps: dividing a forecast area according to a power transmission line, and forecasting an instantaneous precipitation flux of the forecast area to calculate a per-hour rainfall intensity index; taking the power transmission line as a skeleton to form a matrix forecast area by buffering 50 km on both sides of the line; dividing the rectangular forecast area into a 1 km*1 km model grid; and recording the center coordinates of each model grid. ; Large-scale reanalysis fields, obtained from the National Centers for Environmental Prediction, are used as background fields, which are updated every 6 hours ; The observation vector y is formed by collecting the same meteorological data as the background field from multi-source sensors, and the observation vector is assimilated into the analysis field using the GSI 3DVar algorithm ; The precipitation prediction model is trained and the predicted precipitation flux is output as the hourly rainfall intensity , the precipitation prediction model is trained and the predicted precipitation flux is output as the hourly rainfall intensity .

4. The method for monitoring and early warning of slope instability of a tower of a power transmission line under heavy rainfall conditions according to claim 3, characterized in that: The rainfall intensity of each monitoring unit is obtained by interpolation, and then the attenuation coefficient is obtained by calculating the vertical pore pressure transmission time scale to predict the future effective rainfall, and the area-weighted centroid coordinates of each monitoring unit are calculated ; According to the area-weighted barycentric coordinates of the monitoring unit, the mode grid record corner point coordinates containing the barycentric coordinates are found, and the normalized coordinates of the barycenter relative to the lower left corner point of the mode grid are calculated , the rainfall intensity of the monitoring unit is obtained by using a bilinear interpolation formula ; For monitoring the representative average soil layer thickness and the average catchment area ; Obtaining water diffusion coefficient with dominant lithology in monitoring unit as dominant lithology , and further calculating vertical pore pressure transmission time scale of the monitoring unit ; According to the vertical borehole pressure transmission time scale Defining the attenuation coefficient coupled with the vertical time scale ; According to the rainfall intensity of the monitoring unit and the attenuation coefficient, the future 6-hour effective rainfall is calculated in an exponential attenuation manner.

5. The method for monitoring and early warning of slope instability of transmission line tower under heavy rainfall condition according to claim 4, characterized in that: The grid is discretely processed in depth, a one-dimensional Richards equation along the slope normal is established, implicit difference is solved to output the pressure head profile, the physical instability index of the monitoring unit is calculated based on the pressure head profile and the rock and soil parameters, and the water diffusion coefficient of the dominant rock nature is obtained and the average catchment area The lateral hole pressure transfer time scale is calculated ; Combining the vertical pore pressure transmission time scale and the lateral pore pressure transmission time scale to calculate a dimensionless scale parameter and defining a physical model credibility weight ; For each DEM grid g, from ground surface Z = 0 to soil layer bottom Z = Zmax Establish fixed depth layer number , calculate the node position of the mth layer depth ; The pressure head defining the infiltration process is , and the one-dimensional Richards equation along the normal of the slope is established. For the grid g belonging to the monitoring unit i in the depth node Calculating the initial pressure head profile ; The Richards equation of the depth node m is discretized; The complete pressure head profile is obtained by solving the one-dimensional Richards equation through iterative linearization with backward Euler implicit difference in time and second-order central difference in space ; For a grid g at depth Z Computing the safety factor ; For all DEM grids in the monitoring unit i at time Take the minimum safety factor and normalize to the physical instability index .

6. The method for monitoring and early warning of slope instability of transmission line tower under heavy rainfall condition according to claim 5, characterized in that: The characteristic vector of each monitoring unit is constructed, and a landslide probability prediction model is constructed based on XGBoost to output landslide probability , and the landslide probability prediction model is constructed by adopting XGBoost, adopting a standard gradient boosting tree binary classification structure, and defining an overall loss function by combining a logistic regression loss function with a complexity regularization term to train the landslide probability prediction model, and the characteristic vector is input into the trained landslide probability prediction model to output landslide probability .

7. The method for monitoring and early warning of slope instability of transmission line tower under heavy rainfall condition according to claim 6, characterized in that: The comprehensive monitoring unit physical instability index and landslide probability calculate the comprehensive risk index to carry out risk classification and three-dimensional visualization early warning ; After obtaining the comprehensive risk index, the risk threshold is set to map the comprehensive risk index of the monitoring unit to the risk level for output. In the three-dimensional GIS platform, the DEM and transmission line data are loaded as a three-dimensional scene, and are rendered and displayed in the three-dimensional scene according to the risk classification and color. Visual early warning signals are also output to the staff.

8. A monitoring and early warning system for slope instability of a power transmission line tower under heavy rainfall conditions, based on the monitoring and early warning method for slope instability of a power transmission line tower under heavy rainfall conditions according to any one of claims 1-7, characterized in that: The method comprises the following steps: The data processing module is used to divide the DEM grid according to the coordinates of the transmission line tower, deploy the multi-source sensor, define the monitoring unit, calculate the water catchment area and soil thickness of each grid, and obtain the geotechnical parameters for preprocessing; The rainfall prediction module is used to divide the forecast area according to the transmission line, predict the instantaneous precipitation flux of the forecast area, calculate the hourly rainfall intensity, obtain the rainfall intensity of the monitoring unit by interpolation, and then calculate the vertical pore pressure transmission time scale to obtain the attenuation coefficient to predict the future effective rainfall. The geological analysis module is used to deeply discretize the grid, establish a one-dimensional Richards equation along the normal direction of the slope surface, perform implicit difference solution to output the pressure head profile, and calculate the physical instability index of the monitoring unit based on the pressure head profile and the geotechnical parameters. The analysis and early warning module is used to construct a feature vector for each monitoring unit, build a landslide probability prediction model based on XGBoost to output the landslide probability, and calculate the comprehensive risk index by comprehensively considering the physical instability index of the monitoring unit and the landslide probability to perform risk classification and three-dimensional visualization early warning. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to realize the steps of the transmission line tower slope instability monitoring and early warning method under strong rainfall conditions according to any one of claims 1-7.

10. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the transmission line tower slope instability monitoring and early warning method under strong rainfall conditions according to any one of claims 1-7.

Images