Method and system for risk assessment of moraine dam in multi-seismic region based on insar time series analysis and finite element method

Abstract

The application provides a kind of risk assessment method and system of moraine dam in multiple seismic regions based on InSAR time series analysis and finite element method, comprising: InSAR time series analysis and deformation feature extraction are carried out on multiple SAR images, and the freeze-thaw deformation index FDI of moraine dam is obtained;Using finite element simulation method, the internal mechanical response of dam body is simulated and calculated by simulating the three-dimensional finite element model of moraine dam, and the seismic response index SEI is calculated;Establish risk index RI model;Using the risk index RI model with the determined weight coefficient, the working condition of moraine dam in different monitoring intervals is evaluated, and the weak position of moraine dam and the comprehensive risk index RI of moraine dam are determined.The application comprehensively utilizes the surface deformation information of moraine dam obtained by InSAR and the internal mechanical response of dam body simulated by finite element method, realizes the quantitative evaluation of the risk of moraine dam in multiple seismic regions, and provides a scientific basis for the safety monitoring and disaster warning of moraine dam.

Description

Technical Field

[0001] This invention relates to the field of geological disaster monitoring and assessment technology, specifically to a method and system for risk assessment of glacial moraine dams in multi-seismic regions based on InSAR time series analysis and finite element method. Background Technology

[0002] Glacial moraine dams in earthquake-prone areas face severe challenges to their stability due to the combined effects of seismic activity, freeze-thaw cycles, glacial melt, and precipitation. A glacial moraine dam breach would trigger large-scale floods, mudslides, and other secondary disasters, causing immense damage to the lives and property of downstream residents and the ecological environment. Therefore, accurately assessing the risk status of glacial moraine dams and providing early warnings of dam breaches are of paramount importance for disaster prevention and mitigation.

[0003] Traditional methods for assessing the risk of glacial moraine dams primarily rely on field surveys, leveling, and GPS monitoring. Field surveys are limited by terrain conditions; in high mountain and canyon areas prone to earthquakes, transportation is inconvenient and safety risks exist, making it difficult to obtain comprehensive and timely overall information about glacial moraine dams. Leveling and GPS monitoring are point-based monitoring methods with limited coverage, unable to continuously monitor large areas of glacial moraine dams, and unable to capture subtle deformations and overall deformation trends of the dam body.

[0004] InSAR (Synthetic Aperture Radar Interferometry) technology, as an advanced remote sensing monitoring method, has the capability to monitor surface deformation over large areas with high precision, in all weather conditions, and at all times. It can acquire millimeter-level deformation data of glacial till dam surfaces, providing a new approach for monitoring glacial till dam deformation. However, InSAR monitoring alone can only obtain deformation information of the glacial till dam surface and cannot provide in-depth analysis of the mechanical response and failure mechanisms inside the dam body.

[0005] The finite element method (FEM) is a powerful numerical simulation technique capable of performing mechanical analysis on complex engineering structures. By establishing a three-dimensional finite element model of a glacial till dam, the stress and strain distribution of the dam body under loads such as earthquakes and freeze-thaw cycles can be simulated, revealing the dam's failure process and potential collapse risk. However, finite element simulation requires accurate boundary conditions and material parameters, which are complex and variable in practice, making precise acquisition difficult.

[0006] Therefore, there is an urgent need for a risk assessment method for glacial moraine dams in multi-seismic regions that combines InSAR time series analysis with the finite element method, so as to give full play to the advantages of both methods and achieve a comprehensive, accurate and real-time risk assessment of glacial moraine dams. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a method and system for risk assessment of glacial moraine dams in multi-seismic regions based on InSAR time series analysis and finite element method, which can effectively solve the above-mentioned problems.

[0008] The technical solution adopted in this invention is as follows:

[0009] The primary objective of this invention is to provide a method for risk assessment of glacial till dams in multi-seismic regions based on InSAR time series analysis and the finite element method, comprising:

[0010] Step S1: Construct a sample set; each sample in the sample set corresponds to a different monitoring interval, and the monitoring interval of each sample covers the seismic activity period and freeze-thaw cycle period of glacial till dams, including multiple SAR images, seismic monitoring data and the actual risk index of glacial till dams in the monitoring interval collected within the monitoring interval.

[0011] Step S2: Using the freeze-thaw deformation index calculation model, InSAR time series analysis and deformation feature extraction are performed on multiple SAR images in each sample to obtain the freeze-thaw deformation index (FDI) of the glacial till dam during the freeze-thaw cycle in the monitoring interval.

[0012] Step S3: Construct a three-dimensional finite element model of the glacial till dam; using the finite element simulation method, simulate the seismic and freeze-thaw characteristics of the three-dimensional finite element model of the glacial till dam within the monitoring interval and calculate the internal mechanical response of the dam body. The method is as follows:

[0013] Using the seismic monitoring data in each sample as input, and considering the effects of ground motion and freeze-thaw cycles, a seismic-freeze-thaw coupling simulation is performed to obtain the internal mechanical response of the glacial till dam under the combined action of seismic-freeze-thaw, thereby locating the weak points of the glacial till dam under the combined action of earthquake and freeze-thaw.

[0014] Step S4: Using the seismic response index calculation model, the seismic response index (SEI) is calculated based on the internal mechanical response of the dam body.

[0015] Step S5, establish the risk index RI model:

[0016] RI = w1 × FDI + w2 × SEI + w3 × FDI 2 +w4×SEI 2 +w5×(FDI×SEI)

[0017] Where: RI is the comprehensive risk index of glacial moraine dams; w1, w2, w3, w4, and w5 are weighting coefficients to be determined.

[0018] Through steps S1 to S4, the freeze-thaw deformation index FDI and the seismic response index SEI are obtained for each sample. Using the freeze-thaw deformation index FDI and the seismic response index SEI of each sample as input, and the actual risk index of glacial moraine dam as the target, the risk index RI model is trained using each sample to obtain the risk index RI model with determined weight coefficients.

[0019] Step S6: Using the risk index RI model with determined weight coefficients, conduct a risk assessment of the glacial moraine dam's operating conditions under different monitoring intervals, and determine the weak points of the glacial moraine dam and the comprehensive risk index RI of the glacial moraine dam.

[0020] The second objective of this invention is to provide a risk assessment system for glacial moraine dams in multi-seismic regions based on InSAR time series analysis and the finite element method, comprising:

[0021] A sample set construction module is used to construct a sample set; each sample in the sample set corresponds to a different monitoring interval, and the monitoring interval of each sample covers the seismic activity period and freeze-thaw cycle period of glacial till dams, including multiple SAR images, seismic monitoring data and the actual risk index of glacial till dams in the monitoring interval collected within the monitoring interval.

[0022] The freeze-thaw deformation index calculation model is used to perform InSAR time series analysis and deformation feature extraction on multiple SAR images in each sample to obtain the freeze-thaw deformation index (FDI) of the glacial till dam during the freeze-thaw cycle in the monitoring interval.

[0023] The finite element simulation module is used to construct a three-dimensional finite element model of the glacial till dam. Using the finite element simulation method, the seismic and freeze-thaw characteristics of the three-dimensional finite element model of the glacial till dam within the monitoring interval are simulated, and the internal mechanical response of the dam body is calculated. The method is as follows:

[0024] Using the seismic monitoring data in each sample as input, and considering the effects of ground motion and freeze-thaw cycles, a seismic-freeze-thaw coupling simulation is performed to obtain the internal mechanical response of the glacial till dam under the combined action of seismic-freeze-thaw, thereby locating the weak points of the glacial till dam under the combined action of earthquake and freeze-thaw.

[0025] A seismic response index calculation model is used to calculate the seismic response index (SEI) based on the internal mechanical response of the dam body.

[0026] Risk Index (RI) Model and Training Module: Used to build the Risk Index (RI) model.

[0027] RI = w1 × FDI + w2 × SEI + w3 × FDI 2 +w4×SEI 2 +w5×(FDI×SEI)

[0028] Where: RI is the comprehensive risk index of glacial moraine dams; w1, w2, w3, w4, and w5 are weighting coefficients to be determined.

[0029] For each sample, the freeze-thaw deformation index (FDI) and the seismic response index (SEI) are obtained. Using the freeze-thaw deformation index (FDI) and the seismic response index (SEI) of each sample as inputs, and the actual risk index of glacial moraine dam as the target, the risk index RI model is trained using each sample to obtain the risk index RI model with determined weight coefficients.

[0030] The risk assessment module is used to conduct risk assessments on the working conditions of glacial moraine dams under different monitoring intervals using a risk index (RI) model with predetermined weight coefficients, thereby identifying the weak points of the glacial moraine dams and the comprehensive risk index (RI) of the glacial moraine dams.

[0031] The present invention provides a method and system for risk assessment of glacial moraine dams in multi-seismic regions based on InSAR time series analysis and finite element method, which has the following advantages:

[0032] This invention can comprehensively utilize the surface deformation information of glacial till dams obtained by InSAR and the internal mechanical response of the dam body simulated by the finite element method to achieve a quantitative assessment of the risk of glacial till dams in seismically active areas, and provide a scientific basis for the safety monitoring and disaster early warning of glacial till dams. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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.

[0034] Figure 1 A flowchart of a risk assessment method for glacial till dams in multi-seismic regions based on InSAR time series analysis and finite element method provided by the present invention;

[0035] Figure 2 A flowchart for risk index assessment provided for this invention. Detailed Implementation

[0036] To make the technical problems solved, the technical solutions, and the beneficial effects of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and are not intended to limit the invention.

[0037] This invention provides a risk assessment method for glacial moraine dams in seismically active regions based on InSAR time series analysis and the finite element method. It can comprehensively utilize the surface deformation information of glacial moraine dams obtained by InSAR and the internal mechanical response of the dam body simulated by the finite element method to achieve a quantitative assessment of the risk of glacial moraine dams in seismically active regions, and provide a scientific basis for the safety monitoring and disaster early warning of glacial moraine dams.

[0038] This invention provides a method for risk assessment of glacial till dams in multi-seismic regions based on InSAR time series analysis and the finite element method, comprising:

[0039] Step S1: Construct a sample set; each sample in the sample set corresponds to a different monitoring interval, and the monitoring interval of each sample covers the seismic activity period and freeze-thaw cycle period of glacial till dams, including multiple SAR images, seismic monitoring data and the actual risk index of glacial till dams in the monitoring interval collected within the monitoring interval.

[0040] Step S2: Using the freeze-thaw deformation index calculation model, InSAR time series analysis and deformation feature extraction are performed on multiple SAR images in each sample to obtain the freeze-thaw deformation index (FDI) of the glacial till dam during the freeze-thaw cycle in the monitoring interval.

[0041] This step is specifically as follows:

[0042] Step S21: Perform data preprocessing on the multiple SAR images in the sample to obtain multiple preprocessed SAR images;

[0043] The data preprocessing includes radiometric calibration, geometric correction, and image registration.

[0044] The radiometric calibration is as follows: converting the gray values ​​of SAR images into physical scattering coefficients to eliminate the influence of sensor gain and system noise factors, so that images acquired at different times are comparable;

[0045] The geometric correction is as follows: using the terrain elevation data provided by the DEM, a geometric model of SAR image imaging is established to clarify the spatial mapping relationship between the radar beam and ground points; based on the three-dimensional coordinates of each grid point in the DEM, its theoretical pixel position on the SAR image is calculated; comparing the deviation between the actual pixel position and the theoretical pixel position in the SAR image, the coordinates of the SAR image are corrected by interpolation method to eliminate the geometric distortion caused by terrain undulations, so that the SAR image is accurately matched with the actual geographic coordinates, providing a reliable geometric benchmark for interferometric processing and deformation extraction;

[0046] The image registration involves accurately registering multiple SAR images to ensure that the same ground feature corresponds precisely in different SAR images, thus providing a basis for interferometric processing.

[0047] Step S22: Select image pairs with shorter spatial and temporal baselines from multiple pre-processed SAR images and combine them to construct an interferometric network; extract the deformation rate field and cumulative deformation of the glacial moraine surface in the monitoring range by performing phase unwrapping, atmospheric delay correction and orbital error correction on the interferometric network.

[0048] Step S23: Determine the freeze-thaw period based on the climate characteristics of the earthquake-prone areas; there are N freeze-thaw periods in the monitoring interval, therefore, the number of freeze-thaw cycles is N; in each freeze-thaw period, the average deformation D_T of the freeze-thaw period is obtained using time series analysis.

[0049] In this step, the average deformation D_T during the freeze-thaw period is obtained using the following method:

[0050] During each freeze-thaw period, the time-series shape variables of each monitoring point of the glacial till dam were obtained using InSAR time-series analysis, as shown in the following formula:

[0051]

[0052] in: For the first The monitoring point during the freeze-thaw period Deformation of a time period; For the first Effective deformation length of the dam body at the monitoring point; For the first Volumetric moisture content of the glacial till corresponding to the monitoring point; This refers to the dry density of glacial till. The freeze-thaw expansion coefficient of glacial till is a mechanical parameter related to water content and dry density. For the first The monitoring point during the freeze-thaw period The average temperature change over a period of time;

[0053] The average deformation of the same monitoring point is obtained by taking the arithmetic mean of the deformation at each time interval during the entire freeze-thaw period.

[0054] Spatial statistics were performed on the average deformation of all monitoring points in the key areas of the dam body, such as weighted average, with the weight being the area represented by each monitoring point, to obtain the average deformation D_T of the entire glacial moraine dam during the freeze-thaw period.

[0055] Step S24: The freeze-thaw deformation index FDI of the monitoring interval is obtained by using the formula FDI=|D_T|×N, which is used to quantify the degree of impact of freeze-thaw on the glacial moraine dam.

[0056] Step S3: Construct a three-dimensional finite element model of the glacial till dam; using the finite element simulation method, simulate the seismic and freeze-thaw characteristics of the three-dimensional finite element model of the glacial till dam within the monitoring interval and calculate the internal mechanical response of the dam body. The method is as follows:

[0057] Using the seismic monitoring data in each sample as input, and considering the effects of ground motion and freeze-thaw cycles, a seismic-freeze-thaw coupling simulation is performed to obtain the internal mechanical response of the glacial till dam under the combined action of seismic-freeze-thaw, thereby locating the weak points of the glacial till dam under the combined action of earthquake and freeze-thaw.

[0058] This step is specifically as follows:

[0059] Step S31, construct a three-dimensional finite element model of the glacial till dam:

[0060] Step S311, construct the three-dimensional geometric model of the glacial till dam:

[0061] Based on high-precision DEM data and geometric information of glacial till dam obtained from field surveys, a three-dimensional geometric model of the glacial till dam was established in finite element analysis software to simulate the shape, size, and upstream and downstream topography of the glacial till dam. For the dam body itself, especially in the height direction, the mesh size was controlled to not exceed 1 / 10 of the minimum characteristic size of the dam body in order to accurately capture the stress and strain distribution inside the dam body.

[0062] Step S312, Data Preprocessing and Coordinate System I:

[0063] DEM data adaptation: Import high-precision DEM data into the 3D geometric model of the glacial till, retain the terrain of the glacial till and the surrounding 1-2km range through the terrain data clipping module, and remove irrelevant areas; convert the DEM data into point cloud or mesh surface, and calibrate the elevation benchmark;

[0064] Step S313, import field survey data: import the dam crest boundary line, upstream and downstream slope toe lines and dam body centerline vector data into the 3D geometric model of the glacial moraine dam to generate the baseline of the 3D geometric model of the glacial moraine dam; draw the profile outline in the sketch function module for the cross and longitudinal profile data of the dam body obtained from the survey, as the cross section basis for dam body modeling.

[0065] Step S314, Basic terrain generation: Based on the DEM point cloud, generate a terrain surface to ensure that the terrain surface is smooth and fits the measured elevation.

[0066] Step S315, Material Parameter Determination: Indoor tests are conducted by sampling the glacial moraine dam on-site. Samples are taken from the centerline of the dam crest and 2-3 meters on both sides, covering the top 0-1 meter of the dam crest and a shallow layer of 1-3 meters. For the upstream and downstream slopes, a profile is taken every 10-15 meters along the elevation, with three points (toe, middle, and crest) for each profile. The material parameters of the glacial moraine are determined, including density, elastic modulus, Poisson's ratio, internal friction angle, and cohesion. Simultaneously, considering the influence of freeze-thaw cycles on the material properties of the glacial moraine, the freeze-thaw parameters of the glacial moraine are determined, including the coefficient of thermal expansion and frost heave stress.

[0067] Step S316, Boundary condition setting:

[0068] Based on the actual stress conditions of the glacial moraine dam within the monitoring section, the boundary conditions of the model were set as follows: the bottom of the glacial moraine dam was fixed and constrained to simulate the supporting effect of the bedrock on the dam body; water pressure was applied at the upstream and downstream faces to simulate the changes in water levels upstream and downstream of the dam; upstream water pressure... Downstream water pressure ,in The density of water, Let t be the upstream and downstream water level elevations. The vertical coordinates of any point on the dam surface are given; a temperature load is applied to the model surface to simulate the temperature change during the freeze-thaw cycle; based on the seismic activity characteristics of seismic regions, the seismic ground acceleration time history is input to simulate the effect of earthquakes on the glacial till dam.

[0069] Step S32: Using the finite element simulation method, the seismic and freeze-thaw characteristics of the three-dimensional finite element model of the glacial till dam in the monitoring interval are simulated and the internal mechanical response of the dam body is calculated, including the stress field and strain field distribution of the glacial till dam under the earthquake-freeze-thaw coupling action, as well as the displacement and acceleration response of key parts of the dam body.

[0070] Total Stress Tensor , For the three directions in space, the calculation formula is:

[0071]

[0072] in: This refers to the temperature stress tensor induced by freeze-thaw cycles. This refers to the dynamic stress tensor induced by earthquake action. The static stress tensor induced by the dam's own weight and water pressure;

[0073] The formula for calculating acceleration response is:

[0074]

[0075]

[0076]

[0077] in: For key parts exist The acceleration response at any given moment; This represents the freeze-thaw temperature acceleration component; These are the components of earthquake dynamic acceleration; For key parts The effective deformation length; For key parts The coefficient of freeze-thaw expansion of glacial till; For key parts exist The difference between the time and the initial temperature; For key parts Concentrated mass; The damping coefficient; This is the stiffness coefficient; Indicates the direction in space The length of the infinitesimal element; This represents the change in ground acceleration over time under earthquake action. For key parts exist Time along The dynamic velocity components of earthquakes in different directions; For key parts exist Time along Seismic dynamic displacement components in the direction.

[0078] Step S4: Using the seismic response index calculation model, the seismic response index (SEI) is calculated based on the internal mechanical response of the dam body.

[0079] This step is specifically as follows:

[0080] The seismic response index calculation model is SEI=PGA×D; where PGA is the peak ground acceleration, obtained through regional seismic monitoring data; D is the distance attenuation coefficient from the fault zone, quantitatively calculated based on the distance relationship between the glacial till dam and the main active fault zone in the region, D=1 / (r+1); r is the straight-line distance from the center of the glacial till dam to the main active fault zone; the seismic response index SEI is obtained based on the seismic response index calculation model.

[0081] Step S5, establish the risk index RI model:

[0082] RI = w1 × FDI + w2 × SEI + w3 × FDI 2 +w4×SEI 2 +w5×(FDI×SEI)

[0083] Where: RI is the comprehensive risk index of glacial moraine dams; w1, w2, w3, w4, and w5 are weighting coefficients to be determined.

[0084] Through steps S1 to S4, the freeze-thaw deformation index FDI and the seismic response index SEI are obtained for each sample. Using the freeze-thaw deformation index FDI and the seismic response index SEI of each sample as input, and the actual risk index of glacial moraine dam as the target, the risk index RI model is trained using each sample to obtain the risk index RI model with determined weight coefficients.

[0085] In this step, the risk index RI model is trained using various samples to obtain a risk index RI model with determined weight coefficients, specifically as follows:

[0086] Let the sample size in the sample set be... ;No. The actual risk index for each sample is: , Based on the first The FDI, SEI, FDI2, SEI2, and (FDI×SEI) obtained from each sample are represented as follows: Based on the first The model-predicted risk index obtained from each sample is expressed as follows: ;

[0087] When training the risk index RI model using individual samples, the loss function L is: L = ;

[0088] The matrix form of the loss function L is as follows: ;

[0089] in: For the coefficient vector, ;

[0090] The matrix of independent variables, ;

[0091] This represents the actual risk index vector. ;

[0092] Take the derivative of the loss function L with respect to the coefficient vector w, and set the derivative to 0:

[0093]

[0094] The normal equation is obtained by rearranging: ;

[0095] Solving this equation yields the optimal estimates of the coefficients. ; For matrix The inverse matrix;

[0096] The optimal estimate of the coefficients A significance test is performed on the entire model. If all tests pass, the weight coefficients are calculated to obtain the risk index RI model with the weight coefficients determined.

[0097] Step S6: Using the risk index RI model with determined weight coefficients, conduct a risk assessment of the glacial moraine dam's operating conditions under different monitoring intervals, and determine the weak points of the glacial moraine dam and the comprehensive risk index RI of the glacial moraine dam.

[0098] Step S7, Risk Classification and Early Warning:

[0099] Based on the Comprehensive Risk Index (RI) of glacial till dams, the risk levels of glacial till dams are classified into low risk, medium risk, and high risk. Corresponding risk thresholds are set: when RI < 0.4, it is judged as low risk, the dam structure is basically stable, and routine monitoring can be carried out; when 0.4 ≤ RI < 0.7, it is judged as medium risk, freeze-thaw cracks begin to develop, earthquakes may trigger local damage, and monitoring frequency needs to be increased; when RI ≥ 0.7, it is judged as high risk, the freeze-thaw-earthquake coupling effect is significant, the potential rupture surface is close to being connected, and early warning information needs to be issued immediately and corresponding emergency measures need to be taken.

[0100] Based on newly acquired multi-view SAR images and seismic monitoring data of glacial moraine dams, the three-dimensional finite element model of glacial moraine dams is updated in real time, and the latest freeze-thaw deformation index (FDI) and seismic response index (SEI) are calculated, thereby obtaining the latest comprehensive risk index (RI) of glacial moraine dams; based on the latest comprehensive risk index (RI) of glacial moraine dams, the risk status of glacial moraine dams is dynamically assessed.

[0101] This invention also provides a risk assessment system for glacial moraine dams in multi-seismic regions based on InSAR time series analysis and the finite element method, comprising:

[0102] A sample set construction module is used to construct a sample set; each sample in the sample set corresponds to a different monitoring interval, and the monitoring interval of each sample covers the seismic activity period and freeze-thaw cycle period of glacial till dams, including multiple SAR images, seismic monitoring data and the actual risk index of glacial till dams in the monitoring interval collected within the monitoring interval.

[0103] The freeze-thaw deformation index calculation model is used to perform InSAR time series analysis and deformation feature extraction on multiple SAR images in each sample to obtain the freeze-thaw deformation index (FDI) of the glacial till dam during the freeze-thaw cycle in the monitoring interval.

[0104] The finite element simulation module is used to construct a three-dimensional finite element model of the glacial till dam. Using the finite element simulation method, the seismic and freeze-thaw characteristics of the three-dimensional finite element model of the glacial till dam within the monitoring interval are simulated, and the internal mechanical response of the dam body is calculated. The method is as follows:

[0105] Using the seismic monitoring data in each sample as input, and considering the effects of ground motion and freeze-thaw cycles, a seismic-freeze-thaw coupling simulation is performed to obtain the internal mechanical response of the glacial till dam under the combined action of seismic-freeze-thaw, thereby locating the weak points of the glacial till dam under the combined action of earthquake and freeze-thaw.

[0106] A seismic response index calculation model is used to calculate the seismic response index (SEI) based on the internal mechanical response of the dam body.

[0107] Risk Index (RI) Model and Training Module: Used to build the Risk Index (RI) model.

[0108] RI = w1 × FDI + w2 × SEI + w3 × FDI 2 +w4×SEI 2 +w5×(FDI×SEI)

[0109] Where: RI is the comprehensive risk index of glacial moraine dams; w1, w2, w3, w4, and w5 are weighting coefficients to be determined.

[0110] For each sample, the freeze-thaw deformation index (FDI) and the seismic response index (SEI) are obtained. Using the freeze-thaw deformation index (FDI) and the seismic response index (SEI) of each sample as inputs, and the actual risk index of glacial moraine dam as the target, the risk index RI model is trained using each sample to obtain the risk index RI model with determined weight coefficients.

[0111] The risk assessment module is used to conduct risk assessments on the working conditions of glacial moraine dams under different monitoring intervals using a risk index (RI) model with predetermined weight coefficients, thereby identifying the weak points of the glacial moraine dams and the comprehensive risk index (RI) of the glacial moraine dams.

[0112] This invention comprehensively utilizes surface deformation information of glacial till dams obtained by InSAR and internal mechanical response of the dam body simulated by the finite element method to achieve quantitative assessment of the risk of glacial till dams in earthquake-prone areas, providing a scientific basis for safety monitoring and disaster early warning of glacial till dams.

[0113] Two examples are described below:

[0114] Example 1:

[0115] Step S1, InSAR data acquisition and preprocessing:

[0116] Step S11, InSAR data acquisition:

[0117] SAR satellite data with high resolution and short revisit periods, such as Sentinel-1 satellite C-band data (wavelength 5.4 cm, spatial resolution 5 m × 20 m), were selected. At least 20 SAR images with a time span of ≥1 year were collected in the study area to ensure coverage of different operating conditions of the glacial moraine dam, including periods of seismic activity and freeze-thaw cycles.

[0118] Step S2, Data Preprocessing:

[0119] Radiometric calibration: Converting the gray values ​​of SAR images into physical scattering coefficients eliminates the influence of factors such as sensor gain and system noise, making images acquired at different times comparable.

[0120] Geometric correction: First, using the terrain elevation data provided by the DEM, a geometric model for SAR image imaging, such as a range-Doppler model, is established to clarify the spatial mapping relationship between the radar beam and ground points. Then, based on the three-dimensional coordinates (longitude, latitude, and elevation) of each grid point in the DEM, its theoretical pixel position on the SAR image is calculated. Finally, by comparing the deviation between the actual pixel position and the theoretical pixel position in the SAR image, the coordinates of the image are corrected using interpolation and other methods to eliminate geometric distortion caused by terrain undulations, ensuring accurate matching between the SAR image and the actual geographic coordinates, and providing a reliable geometric benchmark for subsequent interferometric processing and deformation extraction.

[0121] Image registration: Accurate registration of multiple SAR images ensures that the same ground feature is precisely located in different SAR images, providing a basis for subsequent interferometric processing.

[0122] Step S2, InSAR time series analysis and deformation feature extraction:

[0123] The Small Baseline Set (SBAS) technique was employed to select image pairs with short spatial and temporal baselines from preprocessed SAR images and combine them to construct an interferometric network. By performing phase unwrapping, atmospheric delay correction, and orbital error correction on the interferometric network, the deformation rate field and cumulative deformation of the glacial till bar surface were extracted.

[0124] Identifying the periodic deformation characteristics during the freeze-thaw cycle: Based on the climatic characteristics of seismically active regions, the freeze-thaw period (generally November to March of the following year) is determined. Using time series analysis, the deformation rate field during the freeze-thaw period is analyzed to identify the periodic expansion and contraction deformation characteristics of the glacial till dam under freeze-thaw cycles. The average deformation D_T (mm) during the freeze-thaw period and the number of freeze-thaw cycles N are calculated, and then the freeze-thaw deformation index FDI = |D_T| × N is calculated. This index is used to quantify the impact of freeze-thaw cycles on the glacial till dam. The average deformation D_T (mm) during the freeze-thaw period is calculated as follows: Within the determined freeze-thaw period (e.g., November to March of the following year), InSAR time series analysis is used to obtain the time-specific deformation of each monitoring point of the glacial till dam (e.g., deformation values ​​every 12 days or months), as shown in the following formula: ,in: For the first The monitoring point during the freeze-thaw period Deformation of a time period; For the first Volumetric moisture content of the glacial till corresponding to the monitoring point; This refers to the dry density of glacial till. For the first The monitoring point during the freeze-thaw period Average temperature change over a period of time; The coefficient of thermal expansion of glacial till (a mechanical parameter related to water content and dry density). For the first The effective deformation length (mm) of the dam body at each monitoring point is calculated by taking the arithmetic mean of all deformation values ​​at the same monitoring point throughout the entire freeze-thaw period, thus obtaining the average deformation value at that point. Then, spatial statistics are performed on the average deformation values ​​of all monitoring points in key areas of the dam body (such as the dam crest and upstream and downstream slopes), such as a weighted average, with the weight being the area represented by each point. Finally, the average deformation value D_T of the entire glacial moraine dam during the freeze-thaw period is obtained. The core principle is to quantify the overall deformation amplitude of the dam body during the freeze-thaw period by integrating the temporal averaging and spatial integration of time-series data.

[0125] Step S2, Construction of the 3D finite element model of the glacial till dam:

[0126] Establishment of a 3D geometric model of the glacial till dam: Based on high-precision (±0.5~2m) DEM data and geometric information of the glacial till dam obtained from field surveys, a 3D geometric model of the glacial till dam is established in finite element analysis software (such as ANSYS, ABAQUS, etc.) to accurately simulate the shape, size, and upstream and downstream topography of the glacial till dam. For the dam body itself, especially in the height direction, the mesh size needs to be controlled to not exceed 1 / 10 of the minimum characteristic size of the dam body. For example, when the dam height is 50 meters, the mesh size should be ≤5 meters to accurately capture the stress and strain distribution inside the dam body.

[0127] Data Preprocessing and Coordinate System I:

[0128] DEM Data Adaptation: Import high-precision DEM data with a horizontal resolution of 5-30 meters. For detailed modeling, use a 5-10 meter resolution UAV LiDAR DEM or 10 meter resolution ALOS World 3D data. For regional terrain simulation, use a 10-30 meter resolution ASTER GDEM. Use the terrain data clipping module to retain the terrain of the glacial moraine and its surrounding 1-2 km range, removing irrelevant areas to reduce computation. In the ANSYS Space Claim or ABAQUS Part module, convert the DEM data into point clouds or mesh surfaces and calibrate the elevation datum, such as unifying it to the WGS84 coordinate system.

[0129] Importing Field Survey Data: Import vector data such as the dam crest boundary line (e.g., GPS-measured coordinates of the dam crest inflection point), upstream and downstream slope toe lines, and the dam's central axis, in DXF or CSV format, to generate the baseline for the 3D geometric model of the glacial moraine dam. For the dam's transverse and longitudinal profile data obtained from the survey, such as a dam height of 50m, top width of 10m, and bottom width of 150m, draw the profile outline in the sketching module as the cross-sectional basis for dam modeling.

[0130] Basic terrain generation: Based on the DEM point cloud, generate a "NURBS surface" in ANSYS using the "From Points" function, or generate a terrain surface in ABAQUS using "Create Surface from Point Cloud". Ensure that the surface is smooth and fits the measured elevation with an error ≤0.5m.

[0131] Material parameter determination: Indoor tests were conducted by sampling the glacial moraine dam on-site. Samples were taken from the centerline of the dam crest and 2-3 meters on both sides, covering the top 0-1 meter of the dam crest and a shallow layer of 1-3 meters. A profile was taken every 10-15 meters along the upstream and downstream slopes, with three points (toe, middle, and crest) on each profile. The material parameters of the glacial moraine, including density, elastic modulus, Poisson's ratio, angle of internal friction, and cohesion, were determined. Simultaneously, the influence of freeze-thaw cycles on the material properties of the glacial moraine was considered, and the coefficient of thermal expansion of the glacial moraine (e.g., 8 × 10⁻⁻⁶) was determined. 5 Freeze-thaw parameters such as / ℃ and freeze-thaw stress (e.g., 200MPa).

[0132] Boundary condition settings: Based on the actual stress conditions of the glacial moraine dam, the boundary conditions of the model are set. The bottom of the glacial moraine dam is fixed and constrained to simulate the supporting effect of the bedrock on the dam body; water pressure is applied at the upstream and downstream faces to simulate the changes in water level upstream and downstream of the dam; upstream water pressure... Downstream water pressure ,in The specific gravity of water, measured in N / m³, is taken at room temperature. N / m³; The upstream and downstream water level elevations at time t are in meters. The vertical coordinates of any point on the dam surface are given in meters. A temperature load is applied to the model surface to simulate the temperature change during the freeze-thaw cycle, such as from -15℃ to 10℃, with a cycle period of 24 hours. Based on the seismic activity characteristics of the seismic region, the ground motion acceleration time history is input, such as using the acceleration record of historical strong earthquakes in the region or artificial seismic waves generated based on the seismic zoning map, to simulate the effect of earthquakes on the glacial till dam.

[0133] Finite element simulation and seismic response index calculation:

[0134] Simulation of Seismic-Freeze-Thaw Coupling: A multi-physics coupled simulation is conducted in a three-dimensional finite element model, simultaneously considering the effects of seismic motion and freeze-thaw cycles. Calculations reveal the stress and strain field distributions of the glacial till dam under seismic-freeze-thaw coupling, as well as the displacement and acceleration responses of key dam components. This allows for precise location of weak points in the glacial till dam under the combined effects of seismic and freeze-thaw cycles, such as stress concentration zones and areas of excessive displacement. The simulation quantifies the degree of danger in the dam's mechanical response, reveals the failure mechanism under coupled effects, and provides direct evidence for assessing dam stability, classifying risk levels, and developing targeted reinforcement schemes.

[0135] Total Stress Tensor , The corresponding three directions in space are calculated using the following formula: ,in The temperature stress tensor induced by freeze-thaw cycles, The dynamic stress tensor induced by earthquake action. It is the static stress tensor induced by the dam's own weight, water pressure, etc.

[0136] The formula for calculating acceleration response is: , The acceleration component is the freeze-thaw temperature component, in m / s². The dynamic acceleration component of the earthquake is expressed in m / s².

[0137] in: Let k be the effective deformation length of the critical component, in meters. denoted as the freeze-thaw expansion coefficient of the glacial till at the critical location k, 1 / ℃; The difference (°C) between the critical component k at time t and the initial temperature \(T_0\) is given. Let K be the concentrated mass (kg) of the critical component k. is the damping coefficient (N·s / m). Stiffness coefficient (N / m); Indicates the direction in space The length of the infinitesimal element; This represents the change in ground acceleration over time under earthquake action. For key parts exist Time along The dynamic velocity components of earthquakes in different directions; For key parts exist Time along Seismic dynamic displacement components in the direction.

[0138] Seismic Response Index (SEI) Extraction: To quantify the impact of seismic action on glacial till dams, the SEI is defined as SEI = PGA × D, where PGA is the peak ground acceleration, which can be obtained from historical regional seismic monitoring data or based on seismic hazard analysis; D is the distance attenuation coefficient from the fault zone, quantitatively calculated based on the distance relationship between the glacial till dam and the main active fault zone in the region, D = 1 / (r+1), where r is the straight-line distance from the center of the glacial till dam to the main active fault zone (unit: km). This coefficient decreases as the distance from the fault zone increases, consistent with the law of seismic energy attenuation with propagation distance. Combining the SEI with the results of three-dimensional finite element coupled simulation, the weak points of the glacial till dam under the combined action of earthquake and freeze-thaw cycles can be accurately located, such as stress concentration zones and displacement exceeding limits. The damage risk of these weak points can be quantified by combining the SEI value (the higher the SEI, the stronger the seismic impact), revealing the failure mechanism under coupled action, and providing a basis for dam stability assessment, risk level classification, and reinforcement scheme formulation.

[0139] Risk index construction and assessment:

[0140] Establish a risk index (RI) model:

[0141] For the risk assessment of glacial till dams under the coupled effects of freeze-thaw cycles and earthquakes, based on the above-mentioned quantification of the impacts of freeze-thaw cycles and earthquakes on the dam body, an interaction term is introduced to capture the nonlinear synergistic effect of the two, and a risk index model is constructed: RI=w1×FDI+w2×SEI+w3×FDI 2 +w4×SEI 2 +w5×(FDI×SEI)

[0142] Wherein: RI is the comprehensive risk index of glacial moraine dams, with a value ranging from 0 to 1, and the higher the value, the higher the risk;

[0143] FDI is the freeze-thaw deformation index, in mm·times, which quantifies the cumulative damage to the dam body caused by freeze-thaw cycles.

[0144] SEI stands for Seismic Response Index, g・km⁻¹, which quantifies the dynamic impact of seismic motion on the dam body.

[0145] FDI×SEI is an interaction term that characterizes the synergistic / antagonistic effects of freeze-thaw cycles and earthquakes;

[0146] w1, w2, w3, w4, and w5 are weighting coefficients to be determined, reflecting FDI, SEI, and FDI respectively. 2 SEI 2 The strength of the contribution of the risk to the total risk, and the interaction between them.

[0147] The coefficients w1, w2, w3, w4, and w5 were determined through multiple linear regression driven by historical sample data. The core logic is "to fit the optimal coefficients using statistical methods based on measured risk data." The specific steps are as follows:

[0148] The coefficients are solved using the least squares method to calculate the sum of squared errors between predicted and actual risks. The loss function is: L =

[0149] Where: n is the sample size Let be the actual risk index of the i-th sample. Predict risk indices for the model;

[0150] To simplify calculations, variables and coefficients are represented in matrix form:

[0151] Coefficient vector:

[0152] Independent variable matrix (n×5):

[0153] Actual risk index vector:

[0154] Predicted value vector:

[0155] The loss function can be simplified as:

[0156] Differentiation and Optimal Solution: Differentiate the loss function L with respect to the coefficient vector w, and set the derivative to 0:

[0157] The normal equation is obtained by rearranging:

[0158] Solving this equation yields the optimal estimates of the coefficients: in, For matrix The inverse matrix.

[0159] Significance test of coefficients (t-test):

[0160] Test whether each coefficient is significantly different from 0 (null hypothesis H0: w j (=0), calculate the t-statistic:

[0161]

[0162] The standard error of the coefficient estimate is given by... The coefficient is obtained by taking the square root of the diagonal element. If the p-value is less than 0.05, the null hypothesis is rejected, and the coefficient is considered significant, meaning that the corresponding variable has a significant impact on risk.

[0163] Overall significance test of the model (F-test):

[0164] To test whether the model is superior to the constant term model (null hypothesis H0: w1=w2=w3=w4=w5=0), the F-statistic is: , (R 2 (where k=5 is the number of independent variables). If the p-value is <0.05, the model is significant overall.

[0165] Goodness-of-fit test:

[0166] Calculate the adjusted R 2 (Eliminating the influence of sample size), the adjusted R is required. 2 ≥0.8 (indicating that the model can explain more than 80% of the risk variance).

[0167] R 2 Definition and calculation:

[0168] R 2 The value of is between [0,1]. The closer the value is to 1, the better the model fits the data. Its calculation formula is:

[0169] Wherein: SST (Total Sum of Squares): measures the total variation of the dependent variable itself, reflecting the degree to which the data deviates from its mean.

[0170] Calculation formula: ,in It is the risk index observation value of the i-th sample. is the mean of the risk index, and n is the sample size.

[0171] SSE (Sum of Squared Residuals): Measures the deviation between the model-predicted risk index and the actual observed risk index, reflecting the unexplained variation in the model.

[0172] Calculation formula: ,in It is the predicted risk index value of the i-th sample, calculated by the regression model.

[0173] Through the above steps, the risk index model can finally be determined.

[0174] Risk Classification and Early Warning: Based on the Risk Index (RI), glacial till dams are classified into three risk levels: low, medium, and high. Corresponding risk thresholds are set: when RI < 0.4, it is considered low risk, the dam structure is basically stable, and routine monitoring is permissible; when 0.4 ≤ RI < 0.7, it is considered medium risk, freeze-thaw cracks are beginning to develop, and earthquakes may trigger local damage, requiring increased monitoring frequency; when RI ≥ 0.7, it is considered high risk, the freeze-thaw-earthquake coupling effect is significant, the potential rupture surface is close to completion, and an immediate early warning is required, along with corresponding emergency measures. Simultaneously, the risk index is updated in real time based on newly acquired InSAR data and earthquake monitoring data to dynamically assess the risk status of glacial till dams.

[0175] This invention provides a risk assessment method for glacial till dams in multi-seismic regions based on InSAR time series analysis and the finite element method, which has the following advantages:

[0176] (1) Comprehensive and accurate risk assessment: This invention combines the surface deformation information of glacial till dam obtained by InSAR time series analysis with the internal mechanical response of the dam body simulated by finite element method. It comprehensively considers the influence of multiple factors such as earthquake and freeze-thaw on the stability of glacial till dam, and can achieve a comprehensive and accurate assessment of the risk of glacial till dam, overcoming the limitations of traditional methods with single monitoring means.

[0177] (2) Quantitative assessment and dynamic early warning: By establishing a risk index model, the risk status of glacial moraine dams is quantified into specific values, and risk classification and dynamic early warning are carried out according to risk thresholds. This can provide relevant departments and the public with risk information on glacial moraine dams in a timely and accurate manner, provide a scientific basis for disaster prevention and mitigation decision-making, and improve the timeliness and effectiveness of disaster response.

[0178] (3) Improve monitoring efficiency and reduce costs: The large-area monitoring capability of InSAR technology can quickly obtain the overall deformation information of glacial moraine dams, reducing the workload and cost of on-site investigation. At the same time, finite element simulation can simulate and analyze various working conditions of glacial moraine dams on a computer without the need for a large number of field tests, further reducing monitoring and evaluation costs and improving work efficiency.

[0179] (4) Adapting to complex geological conditions: Geological conditions in earthquake-prone areas are complex, and the material properties and boundary conditions of glacial till dams are difficult to obtain accurately. This invention monitors the actual deformation of glacial till dams through InSAR time series analysis, providing realistic boundary conditions and deformation constraints for the finite element model, improving the accuracy and reliability of finite element simulation, and enabling it to better adapt to the risk assessment needs of glacial till dams under complex geological conditions.

[0180] Example 2: Risk Assessment of Glacial Tillage Dams in Earthquake-Prone Areas

[0181] Step S1, determining the weighting coefficients w1, w2, w3, w4, and w5:

[0182] Ten sets of glacial moraine dam data were obtained through historical surveys, including FDI, SEI, and the corresponding actual risk index y. i The values ​​range from 0 to 1, with 1 indicating high risk, as detailed below:

[0183] Sample number i FDI (mm・times) SEI (g·km⁻¹) <![CDATA[x i1 =FDI]]> <![CDATA[x i2 =BE]]> <![CDATA[x i3 =FDI 2 ]]> <![CDATA[x i4 =BE 2 (×10 -5 )]]> <![CDATA[x i5 =FDI×SEI]]> <![CDATA[Actual risk index y i > 1 280 0.0064 280 0.0064 78400 4.096 1.792 0.52 2 360 0.0048 360 0.0048 129600 2.304 1.728 0.65 3 220 0.0050 220 0.0050 48400 2.500 1.100 0.40 4 420 0.0070 420 0.0070 176400 4.900 2.940 0.78 5 180 0.0035 180 0.0035 32400 1.225 0.630 0.32 6 320 0.0062 320 0.0062 102400 3.844 1.984 0.59 7 480 0.0055 480 0.0055 230400 3.025 2.640 0.72 8 250 0.0042 250 0.0042 62500 1.764 1.050 0.45 9 520 0.0080 520 0.0080 270400 6.400 4.160 0.85 10 300 0.0058 300 0.0058 90000 3.364 1.740 0.55

[0184] Substitute the data into the optimal estimate equation get

[0185] That is RI=0.013×FDI+9.6×SEI−1.8×10 −5 ×FDI 2 -85×SEI 2 +0.082×(FDI×SEI)

[0186] Coefficient explanation:

[0187] w1=0.013: For every 1 mm·time increase in FDI, RI increases by an average of 0.013 (linear positive contribution).

[0188] w2=9.6: For every 1g・km⁻¹ increase in SEI, RI increases by an average of 9.6 (linear positive contribution).

[0189] w3=−1.8×10−5: The quadratic term of FDI is negative, indicating that the marginal effect of FDI on RI decreases as FDI increases (there is a trend of "slowing growth").

[0190] w4=−85: The quadratic term of SEI is negative, indicating that the marginal effect of SEI on RI decreases as SEI increases;

[0191] w5=0.082: The interaction term is positive, indicating that the synergistic effect of freeze-thaw cycles and earthquakes will amplify the risk.

[0192] Significance test of coefficients (t-test)

[0193] Perform a t-test on each coefficient (null hypothesis H0: w) k (where k=0, k=1,2,…,5), the results are as follows:

[0194] coefficient t-statistic p-value Significance <![CDATA[w1]]> 6.82 0.0002 Significant <![CDATA[w2]]> 3.25 0.011 Significant <![CDATA[w3]]> -2.18 0.065 Significant edges <![CDATA[w4]]> -1.97 0.087 Significant edges <![CDATA[w5]]> 4.03 0.003 Significant

[0195] To test whether the model is superior to the "null model with only constant terms" (null hypothesis H0: w1=w2=w3=w4=w5=0), calculate the F-statistic: Given: R 2 =0.93 (coefficient of determination), k=5 (number of independent variables, including 5 coefficients), n=10 (sample size).

[0196] Substitute into the calculation:

[0197] Looking up the F-distribution table, the degrees of freedom df1=5, df2=4, and the critical value F 0.05 (5,4)=6.26. Since the calculated F=10.63>6.26, the corresponding P value ≈0.018<0.05, the null hypothesis is rejected, indicating that the model is significantly better than the null model overall, that is, at least one coefficient is not 0, and the model's interpretation of risk is statistically significant.

[0198] Step S2, InSAR Data Acquisition and Processing:

[0199] Sentinel-1 satellite C-band SAR images of the area where the glacial moraine dam is located were collected from January 2018 to December 2023, resulting in a total of 30 images.

[0200] Following the data preprocessing steps described above, radiometric calibration, geometric correction, and image registration are performed on the SAR images.

[0201] Step S3: The image is interferometrically processed using SBAS technology. After phase unwrapping, atmospheric delay correction and orbital error correction, the deformation rate field of the glacial till dam surface from 2018 to 2023 is obtained.

[0202] The deformation rate field during the freeze-thaw period (November 2018 - March 2019, November 2019 - March 2020... November 2022 - March 2023) was analyzed. The average deformation D_T during the freeze-thaw period was calculated to be 6.5 mm, and the number of freeze-thaw cycles N was 5. The freeze-thaw deformation index FDI was then calculated as FDI = |D_T| × N = 6.5 × 5 = 32.5 mm·cycle.

[0203] Step S4, Construction of the 3D finite element model of the glacial till dam:

[0204] Based on the field survey data and high-precision DEM of the glacial moraine dam, a three-dimensional geometric model was established in ANSYS software.

[0205] The material parameters of the glacial moraine dam were determined through indoor tests: density 2.2 × 10³ kg / m³, elastic modulus 80 MPa, Poisson's ratio 0.32, internal friction angle 30°, and cohesion 15 kPa. Freeze-thaw parameters were also determined: coefficient of thermal expansion 8 × 10⁻⁻⁻⁻⁶. 5 / ℃, the frost heave stress is 200MPa.

[0206] Set boundary conditions: bottom fixed constraint, apply water pressure to upstream and downstream surfaces according to the measured water level, apply temperature load of -15℃ to 10℃ with a cycle period of 24h to the surface, and input the acceleration time history of the historical 5.5 magnitude earthquake in this area as the seismic load.

[0207] Step S5, Finite Element Simulation and Seismic Response Index Calculation:

[0208] A finite element simulation of earthquake-freeze-thaw coupling was conducted to obtain the stress and strain distribution of the glacial till dam under simulated conditions.

[0209] The peak ground acceleration (PGA) of the simulated earthquake in this region was obtained from earthquake monitoring data as 0.2g. The distance r between the glacial till and the main active fault zone in the region is 30km. The distance attenuation coefficient D = 1 / (r+1) = 1 / (30+1) ≈ 0.032 was calculated. Then, the seismic response index SEI = PGA × D = 0.2 × 0.032 = 0.0064g・km⁻¹ was calculated.

[0210] Calculate the interaction term: FDI×SEI=32.5×0.0064=0.208mm・times・g・km⁻¹.

[0211] Calculate the quadratic term of FDI: FDI × FDI = 32.5 × 32.5 = 1056.25

[0212] Calculate the quadratic term of SEI: SEI × SEI = 0.0064 × 0.0064 = 4.096 × 10 -5

[0213] Step S6, Risk Index Construction and Assessment:

[0214] The risk index model RI with interaction terms is adopted:

[0215] RI=0.013×FDI+9.6×SEI−1.8×10 −5 ×FDI 2 -85×SEI 2 +0.082×(FDI×SEI), where the coefficients were determined by multiple linear regression, and the adjusted R² = 0.93 has been verified, indicating that the model is significant.

[0216] Calculate the risk index of this glacial moraine dam:

[0217] RI=0.013×32.5+9.6×0.0064-1.8×10 −5 ×1056.25-85×4.096×10 -5 +0.082 × 0.208 = 0.4785

[0218] According to the risk classification criteria (verified based on regional historical data):

[0219] Low risk: RI < 0.4; Medium risk: 0.4 ≤ RI < 0.7; High risk: RI ≥ 0.7

[0220] The glacial moraine dam has an RI of 0.4785, classifying it as a medium-risk area. Monitoring frequency needs to be increased, with InSAR data updates conducted quarterly, focusing on the deformation differences between the upstream and downstream toes of the dam.

[0221] Step S7, update the risk assessment in real time:

[0222] Based on steps S1-S6, new Sentinel-1 satellite SAR images were acquired in June 2024, and the risk status of the glacial moraine dam was updated and assessed in real time according to the latest seismic monitoring data.

[0223] Step S71, InSAR data update and processing:

[0224] The newly acquired SAR images were preprocessed and interferometrically processed to obtain the deformation rate field of the glacial moraine dam from January to June 2024.

[0225] Analysis of the deformation during the freeze-thaw period of 2024 (November 2024 to March 2025) revealed that the average deformation D_T during the freeze-thaw period became 7.2 mm, and the cumulative number of freeze-thaw cycles N was 6. The freeze-thaw deformation index FDI was recalculated as FDI = |D_T| × N = 7.2 × 6 = 43.2 mm·cycle.

[0226] Step S72, Finite element model parameter update and simulation:

[0227] Based on newly acquired geological survey information around the glacial moraine dam, some boundary conditions of the finite element model were fine-tuned, such as appropriately increasing the pore water pressure inside the dam body to take into account the recent rise in groundwater level caused by precipitation.

[0228] Due to changes in seismic activity, new seismic monitoring data for the region was obtained. The latest seismic ground acceleration time history (acceleration record of a magnitude 4.8 earthquake) was input into the finite element model, and a new earthquake-freeze-thaw coupling simulation was performed.

[0229] Simulation results show that the stress concentration area inside the glacial moraine dam has expanded, and the displacement and strain responses of key components have also changed.

[0230] Earthquake Response Index Update Calculation:

[0231] The peak ground acceleration (PGA) of this earthquake was obtained from the new earthquake monitoring data as 0.15g. The distance r between the glacial till dam and the main active fault zone in the region is still 30km. The distance attenuation coefficient D is calculated as D=1 / (r+1)=1 / (30+1)≈0.032. The seismic response index SEI is recalculated as SEI=PGA×D=0.15×0.032=0.0048g・km⁻¹.

[0232] Calculate the interaction term: FDI×SEI=43.2×0.0048=0.20736mm・times・g・km⁻¹.

[0233] Calculate the quadratic term of FDI: FDI × FDI = 43.2 × 43.2 = 1866.24

[0234] Calculate the quadratic term of SEI: SEI × SEI = 0.0048 × 0.0048 = 2.304 × 10 -5

[0235] Step S73, real-time updating and assessment of the risk index:

[0236] Using the same risk index model RI:

[0237] RI=0.013×FDI+9.6×SEI−1.8×10 −5 ×FDI 2 -85×SEI 2 +0.082×(FDI×SEI), with the weighting coefficients remaining unchanged. Calculate the updated risk index:

[0238] RI=0.013×43.2+9.6×0.0048−1.8×10 −5 ×1866.24−85×2.304×10 -5 +0.082 × 0.20736 = 0.5891

[0239] The risk level remains medium, but the risk index has increased by approximately 24% compared to the previous reading (0.4785), indicating a worsening of the glacial moraine dam's risk. Based on the assessment results, monitoring of the dam will be further strengthened, with increased deployment density and frequency of monitoring equipment, such as the addition of tiltmeters and piezometers, to monitor dam tilt changes and seepage in real time. Simultaneously, experts will conduct an in-depth analysis of the dam's stability and develop corresponding emergency response plans to ensure timely and effective countermeasures should the risk of further escalation, safeguarding the lives and property of downstream residents and protecting the ecological environment.

[0240] As can be seen from the above embodiments, the risk assessment method for glacial moraine dams in multi-seismic regions based on InSAR time series analysis and finite element method of the present invention can effectively assess the risk of glacial moraine dams and update the assessment results in real time according to new data, providing strong technical support for the safety monitoring and disaster early warning of glacial moraine dams.

[0241] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A risk assessment method for glacial moraine dams in multi-seismic regions based on InSAR time series analysis and finite element method, characterized in that, include: Step S1: Construct a sample set; Each sample in the sample set corresponds to a different monitoring interval. The monitoring interval of each sample covers the seismic activity period and freeze-thaw cycle period of glacial till dams, including multiple SAR images, seismic monitoring data, and the actual risk index of glacial till dams in the monitoring interval. Step S2: Using the freeze-thaw deformation index calculation model, InSAR time series analysis and deformation feature extraction are performed on multiple SAR images in each sample to obtain the freeze-thaw deformation index (FDI) of the glacial till dam during the freeze-thaw cycle in the monitoring interval. Step S3: Construct a three-dimensional finite element model of the glacial till dam; using the finite element simulation method, simulate the seismic and freeze-thaw characteristics of the three-dimensional finite element model of the glacial till dam within the monitoring interval and calculate the internal mechanical response of the dam body. The method is as follows: Using the seismic monitoring data in each sample as input, and considering the effects of ground motion and freeze-thaw cycles, a seismic-freeze-thaw coupling simulation is performed to obtain the internal mechanical response of the glacial till dam under the combined action of seismic-freeze-thaw, thereby locating the weak points of the glacial till dam under the combined action of earthquake and freeze-thaw. Step S4: Using the seismic response index calculation model, the seismic response index (SEI) is calculated based on the internal mechanical response of the dam body. Step S5, establish the risk index RI model: RI=w1×FDI+w2×SEI+w3×FDI 2 +w4×SEI 2 +w5×(FDI×SEI) Where: RI is the comprehensive risk index of glacial moraine dams; w1, w2, w3, w4, and w5 are weighting coefficients to be determined. Through steps S1 to S4, the freeze-thaw deformation index FDI and the seismic response index SEI are obtained for each sample. Using the freeze-thaw deformation index FDI and the seismic response index SEI of each sample as input, and the actual risk index of glacial moraine dam as the target, the risk index RI model is trained using each sample to obtain the risk index RI model with determined weight coefficients. Step S6: Using the risk index RI model with determined weight coefficients, conduct a risk assessment of the working conditions of the glacial moraine dam under different monitoring intervals, and determine the weak points of the glacial moraine dam and the comprehensive risk index RI of the glacial moraine dam. The risk index RI model is trained using various samples to obtain a risk index RI model with determined weight coefficients, specifically as follows: Let the sample size in the sample set be... ;No. The actual risk index for each sample is: , Based on the first The FDI, SEI, FDI2, SEI2, and (FDI×SEI) obtained from each sample are represented as follows: Based on the first The model-predicted risk index obtained from each sample is expressed as follows: ; When training the risk index RI model using individual samples, the loss function L is: L = ; The matrix form of the loss function L is as follows: ; in: For the coefficient vector, ; The matrix of independent variables, ; This represents the actual risk index vector. ; Take the derivative of the loss function L with respect to the coefficient vector w, and set the derivative to 0: ； The normal equation is obtained by rearranging: ; Solving this equation yields the optimal estimates of the coefficients. ; For matrix The inverse matrix; The optimal estimate of the coefficients A significance test is performed on the entire model. If all tests pass, the weight coefficients are calculated to obtain the risk index RI model with the weight coefficients determined.

2. The method for risk assessment of glacial moraine dams in multi-seismic areas based on InSAR time series analysis and finite element method according to claim 1, characterized in that, Step S2 is as follows: Step S21: Perform data preprocessing on the multiple SAR images in the sample to obtain multiple preprocessed SAR images; Step S22: Select image pairs with shorter spatial and temporal baselines from multiple pre-processed SAR images and combine them to construct an interferometric network; extract the deformation rate field and cumulative deformation of the glacial moraine surface in the monitoring range by performing phase unwrapping, atmospheric delay correction and orbital error correction on the interferometric network. Step S23: Determine the freeze-thaw period based on the climate characteristics of the earthquake-prone areas; there are N freeze-thaw periods in the monitoring interval, therefore, the number of freeze-thaw cycles is N; in each freeze-thaw period, the average deformation D_T of the freeze-thaw period is obtained using time series analysis. Step S24: The freeze-thaw deformation index FDI of the monitoring interval is obtained by using the formula FDI=|D_T|×N, which is used to quantify the degree of impact of freeze-thaw on the glacial moraine dam.

3. The method for risk assessment of glacial moraine dams in multi-seismic areas based on InSAR time series analysis and finite element method according to claim 2, characterized in that, The data preprocessing includes radiometric calibration, geometric correction, and image registration. The radiometric calibration is as follows: converting the gray values ​​of SAR images into physical scattering coefficients to eliminate the influence of sensor gain and system noise factors, so that images acquired at different times are comparable; The geometric correction is as follows: using the terrain elevation data provided by the DEM, a geometric model of SAR image imaging is established to clarify the spatial mapping relationship between the radar beam and ground points; based on the three-dimensional coordinates of each grid point in the DEM, its theoretical pixel position on the SAR image is calculated; comparing the deviation between the actual pixel position and the theoretical pixel position in the SAR image, the coordinates of the SAR image are corrected by interpolation method to eliminate the geometric distortion caused by terrain undulations, so that the SAR image is accurately matched with the actual geographic coordinates, providing a reliable geometric benchmark for interferometric processing and deformation extraction; The image registration involves accurately registering multiple SAR images to ensure that the same ground feature corresponds precisely in different SAR images, thus providing a basis for interferometric processing.

4. The method for risk assessment of glacial moraine dams in multi-seismic areas based on InSAR time series analysis and finite element method according to claim 2, characterized in that, In each freeze-thaw period, the average deformation D_T for the freeze-thaw period is obtained using time series analysis, specifically: During each freeze-thaw period, the time-series shape variables of each monitoring point of the glacial till dam were obtained using InSAR time-series analysis, as shown in the following formula: ； in: For the first The monitoring point during the freeze-thaw period Deformation of a time period; For the first Effective deformation length of the dam body at the monitoring point; For the first Volumetric moisture content of the glacial till corresponding to the monitoring point; This refers to the dry density of glacial till. The freeze-thaw expansion coefficient of glacial till is a mechanical parameter related to water content and dry density. For the first The monitoring point during the freeze-thaw period The average temperature change over a period of time; The average deformation of the same monitoring point is obtained by taking the arithmetic mean of the deformation at each time interval during the entire freeze-thaw period. Spatial statistics were performed on the average deformation of all monitoring points in the key areas of the dam body, such as weighted average, with the weight being the area represented by each monitoring point, to obtain the average deformation D_T of the entire glacial moraine dam during the freeze-thaw period.

5. The method for risk assessment of glacial moraine dams in multi-seismic areas based on InSAR time series analysis and finite element method according to claim 1, characterized in that, Step S3 is as follows: Step S31, construct a three-dimensional finite element model of the glacial till dam: Step S311, construct the three-dimensional geometric model of the glacial till dam: Based on high-precision DEM data and geometric information of glacial till dam obtained from field surveys, a three-dimensional geometric model of the glacial till dam was established in finite element analysis software to simulate the shape, size, and upstream and downstream topography of the glacial till dam. For the dam body itself, especially in the height direction, the mesh size was controlled to not exceed 1 / 10 of the minimum characteristic size of the dam body in order to accurately capture the stress and strain distribution inside the dam body. Step S312, Data Preprocessing and Coordinate System I: DEM data adaptation: Import high-precision DEM data into the 3D geometric model of the glacial moraine dam, and retain the terrain of the glacial moraine dam and the surrounding 1-2km range through the terrain data clipping module, while removing irrelevant areas; Convert DEM data into point clouds or mesh surfaces and calibrate elevation datums; Step S313, import field survey data: import the dam crest boundary line, upstream and downstream slope toe lines and dam body centerline vector data into the 3D geometric model of the glacial moraine dam to generate the baseline of the 3D geometric model of the glacial moraine dam; draw the profile outline in the sketch function module for the cross and longitudinal profile data of the dam body obtained from the survey, as the cross section basis for dam body modeling. Step S314, Basic terrain generation: Based on the DEM point cloud, generate a terrain surface to ensure that the terrain surface is smooth and fits the measured elevation. Step S315, Material Parameter Determination: Indoor tests are conducted by sampling the glacial moraine dam on-site. Samples are taken from the centerline of the dam crest and 2-3 meters on both sides, covering the top 0-1 meter of the dam crest and a shallow layer of 1-3 meters. For the upstream and downstream slopes, a profile is taken every 10-15 meters along the elevation, with three points (toe, middle, and crest) for each profile. The material parameters of the glacial moraine are determined, including density, elastic modulus, Poisson's ratio, internal friction angle, and cohesion. Simultaneously, considering the influence of freeze-thaw cycles on the material properties of the glacial moraine, the freeze-thaw parameters of the glacial moraine are determined, including the coefficient of thermal expansion and frost heave stress. Step S316, Boundary condition setting: Based on the actual stress conditions of the glacial moraine dam within the monitoring section, the boundary conditions of the model were set as follows: the bottom of the glacial moraine dam was fixed and constrained to simulate the supporting effect of the bedrock on the dam body; water pressure was applied at the upstream and downstream faces to simulate the changes in water levels upstream and downstream of the dam; upstream water pressure... Downstream water pressure ,in The density of water, Let t be the upstream and downstream water level elevations. The vertical coordinates of any point on the dam surface are given; a temperature load is applied to the model surface to simulate the temperature change during the freeze-thaw cycle; based on the seismic activity characteristics of seismic regions, the seismic ground acceleration time history is input to simulate the effect of earthquakes on the glacial till dam. Step S32: Using the finite element simulation method, the seismic and freeze-thaw characteristics of the three-dimensional finite element model of the glacial till dam in the monitoring interval are simulated and the internal mechanical response of the dam body is calculated, including the stress field and strain field distribution of the glacial till dam under the earthquake-freeze-thaw coupling action, as well as the displacement and acceleration response of key parts of the dam body. Total Stress Tensor , For the three directions in space, the calculation formula is: ； in: This refers to the temperature stress tensor induced by freeze-thaw cycles. This refers to the dynamic stress tensor induced by earthquake action. The static stress tensor induced by the dam's own weight and water pressure; The formula for calculating acceleration response is: ； ； ； in: For key parts exist The acceleration response at any given moment; This represents the freeze-thaw temperature acceleration component; These are the components of earthquake dynamic acceleration; For key parts The effective deformation length; For key parts The coefficient of freeze-thaw expansion of glacial till; For key parts exist The difference between the time and the initial temperature; For key parts Concentrated mass; The damping coefficient; This is the stiffness coefficient; Indicates the direction in space The length of the infinitesimal element; This represents the change in ground acceleration over time under earthquake action. For key parts exist Time along The dynamic velocity components of earthquakes in different directions; For key parts exist Time along Seismic dynamic displacement components in the direction.

6. The method for risk assessment of glacial moraine dams in multi-seismic areas based on InSAR time series analysis and finite element method according to claim 1, characterized in that, Step S4 is as follows: The seismic response index calculation model is SEI=PGA×D; where PGA is the peak ground acceleration, obtained through regional seismic monitoring data; D is the distance attenuation coefficient from the fault zone, which is quantitatively calculated based on the distance relationship between the glacial till dam and the main active fault zone in the region, D=1 / (r+1); r is the straight-line distance from the center of the glacial till dam to the main active fault zone; The Seismic Response Index (SEI) is obtained based on the seismic response index calculation model.

7. The method for risk assessment of glacial moraine dams in multi-seismic areas based on InSAR time series analysis and finite element method according to claim 1, characterized in that, After identifying the weak points of the glacial moraine dam and its comprehensive risk index (RI), the following also applies: Step S7, Risk Classification and Early Warning: Based on the Comprehensive Risk Index (RI) of glacial till dams, the risk levels of glacial till dams are classified into low risk, medium risk, and high risk. Corresponding risk thresholds are set: when RI < 0.4, it is judged as low risk, the dam structure is basically stable, and routine monitoring can be carried out; when 0.4 ≤ RI < 0.7, it is judged as medium risk, freeze-thaw cracks begin to develop, earthquakes may trigger local damage, and monitoring frequency needs to be increased; when RI ≥ 0.7, it is judged as high risk, the freeze-thaw-earthquake coupling effect is significant, the potential rupture surface is close to being connected, and early warning information needs to be issued immediately and corresponding emergency measures need to be taken.

8. The method for risk assessment of glacial moraine dams in multi-seismic regions based on InSAR time series analysis and finite element method according to claim 1 further includes: Based on newly acquired SAR images of glacial till and seismic monitoring data, the three-dimensional finite element model of the glacial till is updated in real time, and the latest freeze-thaw deformation index FDI and seismic response index SEI are calculated, thereby obtaining the latest comprehensive risk index RI of the glacial till. The risk status of glacial moraine dams is dynamically assessed based on the latest Integrated Risk Index (RI).

9. A risk assessment system for glacial moraine dams in multi-seismic regions based on InSAR time series analysis and finite element method, characterized in that, include: A sample set construction module is used to construct a sample set; each sample in the sample set corresponds to a different monitoring interval, and the monitoring interval of each sample covers the seismic activity period and freeze-thaw cycle period of glacial till dams, including multiple SAR images, seismic monitoring data and the actual risk index of glacial till dams in the monitoring interval collected within the monitoring interval. The freeze-thaw deformation index calculation model is used to perform InSAR time series analysis and deformation feature extraction on multiple SAR images in each sample to obtain the freeze-thaw deformation index (FDI) of the glacial till dam during the freeze-thaw cycle in the monitoring interval. The finite element simulation module is used to construct a three-dimensional finite element model of the glacial till dam. Using the finite element simulation method, the seismic and freeze-thaw characteristics of the three-dimensional finite element model of the glacial till dam within the monitoring interval are simulated, and the internal mechanical response of the dam body is calculated. The method is as follows: Using the seismic monitoring data in each sample as input, and considering the effects of ground motion and freeze-thaw cycles, a seismic-freeze-thaw coupling simulation is performed to obtain the internal mechanical response of the glacial till dam under the combined action of seismic-freeze-thaw, thereby locating the weak points of the glacial till dam under the combined action of earthquake and freeze-thaw. A seismic response index calculation model is used to calculate the seismic response index (SEI) based on the internal mechanical response of the dam body. Risk Index (RI) Model and Training Module: Used to build the Risk Index (RI) model. RI=w1×FDI+w2×SEI+w3×FDI 2 +w4×SEI 2 +w5×(FDI×SEI) Where: RI is the comprehensive risk index of glacial moraine dams; w1, w2, w3, w4, and w5 are weighting coefficients to be determined. For each sample, the freeze-thaw deformation index (FDI) and the seismic response index (SEI) are obtained. Using the freeze-thaw deformation index (FDI) and the seismic response index (SEI) of each sample as inputs, and the actual risk index of glacial moraine dam as the target, the risk index RI model is trained using each sample to obtain the risk index RI model with determined weight coefficients. The risk assessment module is used to conduct risk assessments on the working conditions of glacial moraine dams under different monitoring intervals using a risk index RI model with predetermined weight coefficients, thereby identifying the weak points of the glacial moraine dams and the comprehensive risk index RI of the glacial moraine dams. The risk index RI model is trained using various samples to obtain a risk index RI model with determined weight coefficients, specifically as follows: Let the sample size in the sample set be... ;No. The actual risk index for each sample is: , Based on the first The FDI, SEI, FDI2, SEI2, and (FDI×SEI) obtained from each sample are represented as follows: Based on the first The model-predicted risk index obtained from each sample is expressed as follows: ; When training the risk index RI model using individual samples, the loss function L is: L = ; The matrix form of the loss function L is as follows: ; in: For the coefficient vector, ; The matrix of independent variables, ; This represents the actual risk index vector. ; Take the derivative of the loss function L with respect to the coefficient vector w, and set the derivative to 0: ； The normal equation is obtained by rearranging: ; Solving this equation yields the optimal estimates of the coefficients. ; For matrix The inverse matrix; The optimal estimate of the coefficients A significance test is performed on the entire model. If all tests pass, the weight coefficients are calculated to obtain the risk index RI model with the weight coefficients determined.

Images