A targeted analysis method and device for landslide stability considering displacement traction

Abstract

The present invention provides a method for targeted analysis of landslide stability considering displacement traction, the method comprising: processing and analyzing GNSS monitoring point data, considering the influence of displacement traction on grid unit scale division, performing targeted analysis on displacement data, and determining the best evaluation unit; extracting evaluation factors under the condition of determining the best evaluation unit; extracting displacement feature data of each monitoring point, using a clustering algorithm to perform cluster analysis on the displacement feature data, constructing an initial evaluation sample set, and expanding it into training data. After training multiple machine learning models, the accuracy of the analysis is compared to determine the optimal model; the classification accuracy of the optimal model under conventional and displacement traction grid scales is compared horizontally, and the optimal evaluation model is determined to evaluate the stability of the landslide. The present invention realizes the accurate evaluation of landslide stability and provides a strong scientific basis for landslide prevention and control.

Description

Technical Field

[0001] The invention relates to the technical field of landslide stability evaluation and prediction, and in particular to a landslide stability targeted analysis method and equipment considering displacement traction. Background Art

[0002] Landslides, as a common geological disaster, pose a serious threat to human society and the natural environment. Therefore, accurate evaluation of landslide susceptibility and effective prediction of landslide stability are important research topics in the field of geological engineering. Landslide susceptibility evaluation aims to qualitatively or quantitatively evaluate the possibility of landslides in a specific area through the analysis of the disaster-prone conditions of the basic geological environment, and to draw a distribution map of potential landslides based on this. In the evaluation process, the selection of evaluation units is a crucial link, which not only determines the basis of the evaluation work, but also directly affects the accuracy and reliability of the evaluation results.

[0003] At present, there are various types of evaluation units used in landslide susceptibility assessment at home and abroad, mainly including regular grid, slope, region, landform and unique condition. Among them, the regular grid unit has become one of the widely used types of evaluation units due to its simplicity and ease of operation. However, the determination of the scale of the regular grid unit often relies on the knowledge and experience of experts, and lacks objective and unified standards. This limits the accuracy and reliability of landslide susceptibility assessment to a certain extent. Therefore, how to objectively and accurately determine the size of the regular grid unit has become an urgent problem to be solved in the current research on landslide stability evaluation.

[0004] And in the construction of landslide stability evaluation model, the quality of data sample set has a decisive influence on the evaluation accuracy of the model. However, in existing studies, the construction of sample set often ignores the consideration of landslide displacement characteristics in the region. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes a method and device for targeted analysis of landslide stability considering displacement traction to solve the technical problems existing in the prior art. The present invention uses regular grid units as evaluation units, processes and analyzes GNSS monitoring point data, and determines the best evaluation unit by considering the influence of displacement traction on grid unit scale division; and constructs a training marked sample set by clustering the displacement feature data of GNSS monitoring points; and obtains the best evaluation unit division method and the corresponding landslide stability evaluation results through different model training.

[0006] The technical solution adopted by the present invention is as follows:

[0007] In a first aspect, a method for targeted analysis of landslide stability considering displacement traction is provided, the method comprising the following steps:

[0008] Considering the impact of displacement traction on the grid unit scale, targeted analysis is performed on the displacement data to determine the best evaluation unit;

[0009] Extracting evaluation factors under the condition of considering the best evaluation unit;

[0010] Based on the global navigation satellite system GNSS, the displacement feature data of the GNSS monitoring points in the landslide area under the correlation analysis of rainfall data are extracted, and the clustering algorithm is used to perform clustering analysis on the displacement feature data to construct an initial evaluation sample set; the initial evaluation sample set is expanded to obtain training marked sample set data;

[0011] Inputting the evaluation factors and training sample set data into a machine learning model for training to obtain a landslide stability evaluation model;

[0012] The landslide stability evaluation model is used to evaluate the landslide stability.

[0013] Furthermore, the consideration of the influence of displacement traction on the grid unit scale, the targeted analysis of the displacement data, and the determination of the best evaluation unit specifically include:

[0014] The displacement size and displacement direction of the GNSS monitoring point are obtained, and the displacement vectors of each GNSS monitoring point along the slope are plotted according to the time series. A rectangular coordinate system is established with the monitoring point as the origin for targeted analysis. The displacement trajectory of the monitoring point is divided into four categories according to the four quadrants of the coordinate axis, Ⅰ, Ⅱ, Ⅲ, and Ⅳ. The classification operation is repeated to obtain the sample data of the GNSS monitoring point with displacement direction labels.

[0015] Generate a number of random points as unlabeled sample data, use the detection point data with displacement direction labels as labeled data, perform semi-supervised learning on the two samples at the same time, transfer the displacement direction labels to the random point data, and obtain a number of GNSS monitoring point sample data with displacement direction labels, wherein the GNSS monitoring point sample data with displacement direction labels are landslide displacement monitoring point data with displacement direction labels;

[0016] Based on the landslide displacement monitoring point data with displacement direction labels, the consistency of the displacement direction of each landslide displacement monitoring point data under different size evaluation units is analyzed, the evaluation unit size is determined, and the best evaluation unit is determined; the size and shape of the best evaluation unit are not completely the same.

[0017] Furthermore, the extraction of evaluation factors includes: extracting data in each evaluation unit from the aspects of topography, geological conditions and hydrology.

[0018] Furthermore, the extracted evaluation factors include: surface roughness, surface undulation, terrain moisture index, elevation, slope, slope aspect, distance to water system, distance to road, distance to building, distance to crack and stratum lithology.

[0019] Furthermore, the displacement characteristic data include: cumulative displacement during the rainfall period, average daily displacement, maximum daily displacement, and total displacement.

[0020] Furthermore, the ISSA-K-means K-means clustering algorithm based on the improved sparrow search algorithm was used to perform cluster analysis on the displacement feature data, and the displacement feature data were aggregated into four different risk categories. The evaluation units where the monitoring points were located were marked as four different categories: extremely low risk, low risk, medium risk, and high risk, completing the construction of the initial evaluation sample set.

[0021] Furthermore, the initial evaluation sample set is expanded through a semi-supervised learning method to obtain training labeled sample set data.

[0022] Furthermore, the machine learning models include: Stacking model, SVC model, Gaussian process model, random forest model, XGB model and logistic regression model; during training, the accuracy of each model is compared and analyzed to determine the optimal accuracy model, and the classification accuracy of the optimal accuracy model under conventional and displacement traction grid unit scale division methods is horizontally compared. According to the comparison results, the optimal machine learning model is determined as the landslide stability evaluation model.

[0023] In a second aspect, an electronic device is provided, which, in a second possible implementation manner, includes:

[0024] one or more processors;

[0025] A storage device for storing one or more programs;

[0026] When one or more programs are executed by one or more processors, the one or more processors implement the targeted landslide stability analysis method considering displacement traction described in the first aspect.

[0027] It can be seen from the above technical solution that the beneficial technical effects of the present invention are as follows:

[0028] 1. The present invention determines the best evaluation unit by considering the influence of displacement traction on the grid unit scale. The evaluation units obtained by this division method are not completely the same in size and shape. It can more scientifically and accurately reflect the actual situation of the landslide area and improve the evaluation accuracy.

[0029] 2. The present invention extracts evaluation factors from multiple aspects, constructs a comprehensive evaluation factor system, and provides more comprehensive data support for landslide stability evaluation.

[0030] 3. The present invention divides GNSS monitoring points into different risk categories through cluster analysis of displacement feature data and constructs an initial evaluation sample set. Subsequently, the sample set is expanded using a semi-supervised learning method, further improving the diversity and richness of the sample data and enhancing the generalization ability of the model.

[0031] 4. By training multiple machine learning models and comparing and analyzing their accuracy, the present invention can select the optimal landslide stability evaluation model according to actual conditions, thereby ensuring the flexibility and accuracy of the evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the prior art description. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0033] Figure 1 This is a flow chart of a method according to an embodiment of the present invention;

[0034] Figure 2 This is a displacement trajectory diagram of GNSS monitoring points 1-6 according to an embodiment of the present invention; Figure 2 (a) is the displacement trajectory of GNSS monitoring point 1; Figure 2 (b) is the displacement trajectory diagram of GNSS monitoring point 2; Figure 2 (c) is the displacement trajectory diagram of GNSS monitoring point No. 3; Figure 2 (d) is the displacement trajectory diagram of GNSS monitoring point 4; Figure 2 (e) is the displacement trajectory diagram of GNSS monitoring point No. 5; Figure 2 (f) is the displacement trajectory diagram of GNSS monitoring point No. 6;

[0035] Figure 3 A distribution diagram of displacement trajectories of all GNSS monitoring points in an embodiment of the present invention;

[0036] Figure 4 The sample distribution diagram in the 5m×5m grid where monitoring point No. 1 is located in the embodiment of the present invention;

[0037] Figure 5 The sample distribution diagram in the 7m×7m grid where monitoring point No. 1 is located in the embodiment of the present invention;

[0038] Figure 6This is a sample distribution diagram within a 9m×9m grid where monitoring point No. 1 is located in an embodiment of the present invention;

[0039] Figure 7 This is a sample distribution diagram within the 11m×11m grid where monitoring point No. 1 is located in the embodiment of the present invention;

[0040] Figure 8 This is a sample distribution diagram within a 13m×13m grid where monitoring point No. 1 is located in an embodiment of the present invention;

[0041] Fig. 9 This is a sample distribution diagram within a 15m×15m grid where monitoring point No. 1 is located in an embodiment of the present invention;

[0042] Fig.10 This is a graph showing the relationship between the consistency rate of point 1 and the size of the grid in an embodiment of the present invention;

[0043] Fig.11 This is a graph of the consistency rate of points 1-10 in the embodiment of the present invention;

[0044] Fig.12 This is a graph of the consistency rate of different grid sizes at points 1-10 in an embodiment of the present invention;

[0045] Fig.13 is a cumulative displacement diagram of monitoring points in an embodiment of the present invention;

[0046] Fig.14 This is a cluster sample distribution diagram in an embodiment of the present invention;

[0047] Fig.15 This is a sample distribution diagram of semi-supervised classification results after data expansion in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The following embodiments of the technical solution of the present invention are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore only used as examples, and cannot be used to limit the protection scope of the present invention.

[0049] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should have the common meanings understood by those skilled in the art to which the present invention belongs.

[0050] Example

[0051] This paper takes a certain landslide as the research area and proposes a targeted analysis method for landslide stability considering displacement traction based on GIS platform and data analysis. Figure 1 As shown;

[0052] Firstly, the regular grid unit is used as the evaluation unit, and the displacement direction data of the GNSS monitoring points are extracted and analyzed in combination with the rainfall data. That is, the influence of displacement traction on the scale division of the grid unit is considered, and the optimal evaluation unit is determined; secondly, based on the availability of data and the development characteristics of landslides in the study area, 11 evaluation factors suitable for subsequent model training are prepared (surface roughness, surface undulation, terrain moisture index, elevation, slope, slope direction, distance to water system, distance to road, distance to building, distance to crack, and stratum lithology); then, a rich training labeled sample set is constructed through cluster analysis of the characteristic data of the GNSS monitoring point displacement; finally, the accuracy of each model and the influence of different grid unit scale division on the model accuracy are compared and analyzed, and a suitable grid unit scale division method and landslide stability evaluation model are obtained, which provide basic data and theoretical basis for landslide risk assessment and prevention work in the study area.

[0053] The working principle of the embodiment is described in detail below:

[0054] 1. Determination of evaluation unit

[0055] The present invention performs a rough division of the grid size according to a preliminary grid size delineation formula to determine the initial value range of the best evaluation unit, then conducts a targeted analysis of the traction correlation of the landslide monomer displacement, and finally refines the grid unit size to determine the best evaluation unit.

[0056] 1.1 Determine the scale range of regular grid cells:

[0057] The present invention preliminarily divides the size of the regular grid unit according to the preliminary grid size demarcation formula to find out the preliminary value range of the grid size. That is, for the terrain scale of the experimental area of ​​1:2000, the result value calculated according to the formula is about 9m. According to the error range of the calculation formula, the optimal grid unit value range is 5-15m. Then, the grid size is determined according to the landslide unit displacement traction association. The formula is expressed as follows:

[0058] G = 7.49 + 6 × 10 -4 ×M-2×10 -9 ×M 2 +2.9×10 -15 M 3

[0059] Where G is the appropriate grid unit size, and M is the denominator of the topographic map scale. The above grid size preliminary delineation formula is obtained by the inventor based on logistic regression of big data.

[0060] 1.2 Consider displacement traction and obtain sample data with displacement direction labels

[0061] The present invention proposes a method for determining the size of a landslide evaluation unit taking displacement traction into consideration, that is, according to the displacement direction obtained from the landslide displacement monitoring point, the evaluation unit size is determined from the perspective of the association between the displacement monitoring points, that is, the displacement direction of the landslide monitoring point is used as the attribute data to represent the geographical attribute data within each evaluation unit, and by statistically analyzing the consistency of the landslide displacement direction in each evaluation unit, a grid evaluation unit that satisfies the high consistency of the displacement direction of each landslide point in the grid size is a more ideal grid evaluation unit.

[0062] By extracting the displacement size and displacement direction of the GNSS monitoring point, the displacement vector of each GNSS monitoring point along the slope is plotted according to the time series. A rectangular coordinate system targeted analysis is established with the monitoring point as the origin. The displacement trajectory of the monitoring point can be divided into four categories according to the four quadrants of the coordinate axis, Ⅰ, Ⅱ, Ⅲ, and Ⅳ. That is, if the displacement trajectory is in quadrant Ⅰ, the displacement trajectory of this point can be classified into category Ⅰ. Repeat the above classification operation to obtain a set of GNSS monitoring point sample data with displacement direction labels. The displacement trajectories of monitoring points 1-6 are as follows: Figure 2 The distribution of all GNSS point displacement trajectories is shown in Figure 3 shown.

[0063] Due to the limited number of GNSS landslide displacement monitoring points deployed on a single landslide, the requirements of directional consistency statistical data under different grid sizes cannot be met. Therefore, semi-supervised learning is selected to expand the labeled samples. The present invention generates a large number of random points in the study area as unlabeled sample data, uses the detection point data with displacement direction labels as labeled data, performs semi-supervised learning on the two samples at the same time, transfers the displacement direction labels to the random point data, and finally obtains a sample data with a large number of displacement direction labels for subsequent experimental analysis. Finally, 6000 unlabeled sample point data are randomly generated, and the semi-supervised learning algorithm is used to spread the labels of 39 sample data with displacement direction labels to the selected 6000 unlabeled samples, and finally 6039 sample data with displacement direction labels are obtained.

[0064] 1.3 Within the scale of regular grid cells, targeted analysis is performed on sample data with displacement direction labels to determine the best evaluation unit

[0065] When determining the optimal scale of the grid unit, the two constraints of data homogeneity and data volume should be considered comprehensively. For example: the optimal grid unit size has been determined to have a range of 5-15m. Next, the grids are divided into 5m, 6m, 7m, 8m, 9m, 11m, 13m, and 15m. The semi-supervised expanded sample point data with displacement direction labels are superimposed and analyzed with grid data of different sizes. The consistency of displacement direction in evaluation units of different sizes where the 39 GNSS monitoring points are located is counted. The consistency of displacement direction is determined by the consistency rate formula, which is expressed as:

[0066]

[0067] Take monitoring point 1 as an example: the samples in different grid sizes at monitoring point 1 are as follows: Figures 4 to 9 As shown, the functional relationship is as follows Fig.10 As shown. From the functional relationship diagram of data consistency rate and grid size, it can be seen that the grid corresponding to the inflection point is the best evaluation unit. For monitoring point 1, this grid size value is selected as the best evaluation unit, which can not only meet the homogeneity of the data in the evaluation unit as much as possible, but also avoid the problem of too much data caused by simply taking the minimum value. As above, the relationship between different grid sizes and data consistency rates of all monitoring points is obtained, among which the relationship diagram of monitoring points 1-10 is shown as follows Fig.11 and 12 shown.

[0068] Since the displacement traction is different in each regular grid unit, the optimal evaluation unit of each monitoring point obtained through this step is not completely the same in size and shape.

[0069] 2. Extract evaluation factors

[0070] The evaluation factor data in each evaluation unit were extracted from the aspects of topography, geological conditions and hydrology, and the evaluation factor system of the subsequent model was constructed. That is, based on the availability of data and the development characteristics of landslides in the study area, 11 evaluation factors suitable for subsequent model training were prepared (surface roughness, surface undulation, terrain moisture index, elevation, slope, slope aspect, distance to water system, distance to road, distance to building, distance to crack, and stratum lithology).

[0071] 3. Labeled sample set construction

[0072] The present invention takes landslide monomer as the research object, pre-processes the data of GNSS monitoring points arranged in the research area, and extracts displacement feature data. Based on the ISSA-K-means algorithm, cluster analysis is performed on the displacement feature data, a labeled sample set based on GNSS monitoring points is constructed, and the labeled sample set is expanded using a semi-supervised learning method.

[0073] 3.1 Displacement characteristic data based on rainfall

[0074] In order to fully utilize the data of all detection points, the present invention selects the intersection of the monitoring time periods of all monitoring points and extracts the characteristic data such as daily displacement and cumulative displacement of each GNSS monitoring point within this period. Fig.13 As shown, the study found that greater rainfall is highly correlated with the occurrence of landslides. Therefore, in this study, the displacement size feature extraction was carried out during the period when more rainfall occurred.

[0075] The feature extraction of landslide displacement data can obtain different stability data of each landslide point location. The present invention calculates the number of times the displacement of 39 monitoring points is greater than the average value, the monthly cumulative displacement, the daily average displacement, the maximum daily displacement, and the total displacement during the period of heavy rainfall (February-August). The feature data is subjected to correlation analysis, and features with correlation greater than 0.85 are eliminated.

[0076] 3.2 Cluster analysis of evaluation sample sets

[0077] Based on the displacement feature data of each GNSS monitoring point extracted by the present invention, cluster analysis is performed using the ISSA-K-means algorithm to aggregate the feature data and attach different labels to four different categories, and mark the evaluation units where the monitoring points are located as corresponding categories to complete the preliminary construction of the evaluation sample set. Fig.14 Based on the semi-supervised learning algorithm, the label information of 39 labeled sample data is diffused to the selected 1000 unlabeled samples, and finally 1039 labeled sample data are obtained. The distribution is shown in Fig.15 As shown. Fig.15 It can be seen that the number of samples in the four categories varies greatly. Therefore, the sample size of all categories is selected according to the minimum number of samples, that is, the samples of all categories are selected according to the number of high-risk samples. That is, 125 sample data with high risk, medium risk, low risk and very low risk labels are selected from the samples expanded by semi-supervised classification, for a total of 500 labeled sample data to meet the requirements of the model training data volume.

[0078] 4. Model Training

[0079] The Stacking model and other single models were trained using the labeled sample set, and then the accuracy of each model was compared and analyzed to obtain the best accuracy of the Stacking model; on the basis of selecting the Stacking model, the classification accuracy of the model under the conventional and displacement traction grid unit scale division methods was compared horizontally. According to the final accuracy evaluation results in Table 1, Table 1 shows the comparison data of the top six model accuracy rates:

[0080] Table 1:

[0081] Model Accuracy (%) F1-score(%) Kappa coefficient (%) Stacking Model (7m) 89.32 88.95 82.78 SVC Model 87.31 87.97 79.24 Stacking Model (9m) 87.17 87.65 79.03 Gaussian Process Model 83.89 83.11 72.89 Random Forest Model 82.17 81.97 69.23 XGB Model 80.54 80.16 67.36

[0082] According to the final accuracy evaluation results in Table 1, the accuracy of the Stacking model under the displacement traction grid unit scale division method proposed in the present invention is higher than that of other learner models and the Stacking model under the conventional grid unit scale division method. It has good classification performance and is suitable for the landslide stability evaluation in the study area.

[0083] In one embodiment, an electronic device is provided, which, in a second possible implementation, includes:

[0084] one or more processors;

[0085] A storage device for storing one or more programs;

[0086] When one or more programs are executed by one or more processors, the one or more processors implement the targeted landslide stability analysis method considering displacement traction described in the first aspect.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A targeted landslide stability analysis method considering displacement traction, characterized in that: The method comprises the following steps: Considering the influence of displacement traction on the scale of grid units, targeted analysis is performed on displacement data to determine the best evaluation unit; including: obtaining the displacement size and displacement direction of GNSS monitoring points, plotting the displacement vectors of each GNSS monitoring point along the slope according to the time series, establishing a rectangular coordinate system with the monitoring point as the origin for targeted analysis, dividing the displacement trajectory of the monitoring point into four categories according to the four quadrants Ⅰ, Ⅱ, Ⅲ, and Ⅳ of the coordinate axis, repeating the classification operation, and obtaining GNSS monitoring point sample data with displacement direction labels; generating a number of random points as unlabeled sample data, taking the detection point data with displacement direction labels as labeled data, performing semi-supervised learning on the two samples at the same time, transferring the displacement direction labels to the random point data, and obtaining a number of GNSS monitoring point sample data with displacement direction labels, wherein the GNSS monitoring point sample data with displacement direction labels are landslide displacement monitoring point data with displacement direction labels; based on the landslide displacement monitoring point data with displacement direction labels, analyzing the consistency of the displacement direction of each landslide displacement monitoring point data under evaluation units of different sizes, determining the evaluation unit size, and determining the best evaluation unit; the size and shape of the best evaluation unit are not completely the same; Extracting evaluation factors under the condition of considering the best evaluation unit; Based on the global navigation satellite system GNSS, the displacement feature data of the GNSS monitoring points in the landslide area under the correlation analysis of rainfall data are extracted, and the clustering algorithm is used to perform clustering analysis on the displacement feature data to construct an initial evaluation sample set; the initial evaluation sample set is expanded to obtain training marked sample set data; Inputting the evaluation factors and training sample set data into a machine learning model for training to obtain a landslide stability evaluation model; The landslide stability evaluation model is used to evaluate the landslide stability.

2. A targeted landslide stability analysis method considering displacement traction according to claim 1, characterized in that: The extraction of evaluation factors includes: extracting data in each evaluation unit from the aspects of topography, geological conditions and hydrology.

3. A targeted landslide stability analysis method considering displacement traction according to claim 2, characterized in that: The extracted evaluation factors include: surface roughness, surface undulation, terrain moisture index, elevation, slope, slope direction, distance to water system, distance to road, distance to building, distance to crack and stratum lithology.

4. The method for targeted analysis of landslide stability considering displacement traction according to claim 1 is characterized in that: The displacement characteristic data include: cumulative displacement during the rainfall period, average daily displacement, maximum daily displacement, and total displacement.

5. The method for targeted analysis of landslide stability considering displacement traction according to claim 1 is characterized in that: The ISSA-K-means K-means clustering algorithm based on the improved sparrow search algorithm is used to perform cluster analysis on the displacement feature data. The displacement feature data are aggregated into four different risk categories, and the evaluation units where the monitoring points are located are marked as four different categories: extremely low risk, low risk, medium risk, and high risk, completing the construction of the initial evaluation sample set.

6. The method for targeted analysis of landslide stability considering displacement traction according to claim 1, characterized in that: The initial evaluation sample set is expanded by a semi-supervised learning method to obtain training labeled sample set data.

7. The method for targeted analysis of landslide stability considering displacement traction according to claim 1, characterized in that: The machine learning models include: Stacking model, SVC model, Gaussian process model, random forest model, XGB model and logistic regression model; during training, the accuracy of each model is compared and analyzed to determine the optimal accuracy model, and the classification accuracy of the optimal accuracy model under conventional and displacement traction grid unit scale division methods is horizontally compared. According to the comparison results, the optimal machine learning model is determined as the landslide stability evaluation model.

8. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the landslide stability targeted analysis method considering displacement traction as described in any one of claims 1 to 7.

Images