CN117131756A - Ground crack susceptibility evaluation method based on ground surface time sequence deformation and disaster-pregnancy background - Google Patents
Ground crack susceptibility evaluation method based on ground surface time sequence deformation and disaster-pregnancy background Download PDFInfo
- Publication number
- CN117131756A CN117131756A CN202310670192.6A CN202310670192A CN117131756A CN 117131756 A CN117131756 A CN 117131756A CN 202310670192 A CN202310670192 A CN 202310670192A CN 117131756 A CN117131756 A CN 117131756A
- Authority
- CN
- China
- Prior art keywords
- ground
- disaster
- evaluation
- crack
- factor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 65
- 238000000034 method Methods 0.000 claims abstract description 43
- 238000005516 engineering process Methods 0.000 claims abstract description 7
- 230000007246 mechanism Effects 0.000 claims abstract description 6
- 238000011161 development Methods 0.000 claims description 14
- 238000012549 training Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000002790 cross-validation Methods 0.000 claims description 6
- 238000011160 research Methods 0.000 claims description 6
- 230000035935 pregnancy Effects 0.000 claims description 5
- 230000001939 inductive effect Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000000611 regression analysis Methods 0.000 claims description 3
- 239000011435 rock Substances 0.000 claims description 3
- 230000035945 sensitivity Effects 0.000 claims description 3
- 238000010586 diagram Methods 0.000 claims description 2
- 238000011084 recovery Methods 0.000 claims 1
- 208000010392 Bone Fractures Diseases 0.000 description 27
- 206010017076 Fracture Diseases 0.000 description 27
- 238000010801 machine learning Methods 0.000 description 4
- 238000010276 construction Methods 0.000 description 3
- 238000013139 quantization Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 201000004569 Blindness Diseases 0.000 description 1
- 208000013201 Stress fracture Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000013210 evaluation model Methods 0.000 description 1
- 238000012854 evaluation process Methods 0.000 description 1
- 230000000763 evoking effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000005086 pumping Methods 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 238000011158 quantitative evaluation Methods 0.000 description 1
- 238000013077 scoring method Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
Abstract
The invention discloses a ground crack susceptibility evaluation method based on ground surface time sequence deformation and a disaster-prone background, which comprises the steps of determining the type of the ground crack to be evaluated and the disaster-prone background, and initially establishing a ground crack evaluation factor space database; grading and quantifying the ground fracture evaluation factors by using a deterministic factor method CF to obtain grading and quantifying values; acquiring an integral time sequence deformation result of an evaluation area through a TS-InSAR technology, establishing a coordinate system and resolution consistent with a ground crack evaluation factor in ArcGIS software, and acquiring an annual average deformation rate graph; determining a primary selected area of a ground crack non-disaster point by combining a ground crack occurrence mechanism, determining a selected area of the ground crack disaster point according to historical ground crack list data, and finally constructing a ground crack sample database; constructing an XGBoost model, taking a quantized value and a ground crack sample database as the input of the model, outputting a susceptibility evaluation result through the trained XGBoost model, and evaluating the model precision.
Description
Technical Field
The invention relates to the field of prevention and control of geological disasters of ground cracks, in particular to a ground crack susceptibility evaluation method based on ground surface time sequence deformation and a disaster-pregnant background.
Background
The evaluation of the susceptibility to the ground cracks is a method for obtaining the spatial distribution of the susceptibility to the ground cracks through quantitative evaluation and prediction of the factors affecting the ground cracks. The method has important guiding significance for identifying potential dangerous areas, disaster prevention and control and urban comprehensive management. At present, the common susceptibility evaluation methods of researchers at home and abroad are mainly divided into a knowledge-driven model and a data-driven model.
The knowledge driving model mainly comprises an analytic hierarchy process AHP, an expert scoring method and the like, the method depends on expert experience, and the subjectivity of the evaluation process is high; the data driving method is mainly used for realizing prediction analysis of disasters by analyzing the relation between the obtained data and the ground crack disasters. Common methods include probabilistic methods, deep learning, machine learning, and the like. The existing ground crack susceptibility evaluation model is mainly a knowledge driving model, namely the models show larger subjectivity on the determination of the environmental factor weight and are abbreviated on the construction of the susceptibility model. Although these models are continuously perfected, some students introduce machine learning and deep learning models into the evaluation of the susceptibility of the ground fracture.
However, the model mainly adopts static evaluation, and ignores the dynamic evolution characteristics of the geological disaster of the ground fissure and the dynamic evolution characteristics of disaster points and non-disaster points. With the continuous development of research, a machine learning model is gradually developed into an ideal method for treating nonlinear geological disaster problems by the outstanding advantages of the machine learning model, such as higher model stability and prediction accuracy, and being capable of treating partial missing data sets. However, the problem of random non-disaster point selection exists in the susceptibility evaluation of the geological disaster of the ground cracks, so that the susceptibility evaluation precision of the ground cracks is not high.
Disclosure of Invention
Aiming at the problems in the field, the invention provides a method for evaluating the susceptibility of a ground fracture based on the ground surface time sequence deformation and the disaster-prone background, which can solve the technical problem that the susceptibility evaluation of the ground fracture geological disaster has non-disaster point random selection, so that the susceptibility evaluation precision of the ground fracture disaster is not high.
In order to solve the technical problems, the invention discloses a ground crack susceptibility evaluation method based on ground surface time sequence deformation and a disaster-prone background, which comprises the following steps:
determining the type of the ground cracks to be evaluated and the background of the pregnant disaster, and initially establishing a ground crack evaluation factor space database;
grading and quantifying the ground fracture evaluation factors by using a deterministic factor method CF to obtain grading and quantifying values of the ground fracture evaluation factors;
acquiring an integral time sequence deformation result of an evaluation area through a TS-InSAR technology, establishing a coordinate system and resolution consistent with a ground crack evaluation factor in ArcGIS software, and acquiring an annual average deformation rate graph; processing the annual average deformation rate graph by combining the ground fracture occurrence mechanism, and determining a primary selection area of a ground fracture non-disaster point by combining a disaster-tolerant background; determining a selected area of the ground crack disaster point according to the historical ground crack list data, and finally constructing a ground crack sample database;
constructing an XGBoost model, taking the quantized value and the ground crack sample database as the input of the XGBoost model, and outputting a susceptibility evaluation result through the trained XGBoost model.
Preferably, the determining the type of the ground fissure to be evaluated and the background of the pregnant disaster comprises the following steps:
extracting a main control factor for inducing the ground cracks according to the development characteristics, the cause types and different pregnancy factors of the ground cracks;
judging the collinearity and the relativity of each main control factor by using a Schmidt orthogonalization SO method and calculating a variance expansion factor VIF, and determining an evaluation factor;
the calculation formula of the schmitt orthogonalization SO method is as follows:
β 1 ＝α 1
wherein alpha is i Representing any linear independent vector group, beta, in Euclidean space j Representing the corresponding orthogonal vector group, j representing the number of factors;
the calculation formula of the variance expansion factor VIF is as follows:
wherein R is i Negative correlation coefficients for regression analysis of the remaining independent variables are performed for the independent variables.
Preferably, the step of establishing a space database of the ground fracture evaluation factors, wherein the influence of each ground fracture evaluation factor on the ground fracture disasters is evaluated by a single-factor ROC curve method, and the influence degree is ordered and falls into 1:1, removing factors in the line after analysis, and constructing a main control factor database;
and unifying the coordinate system and the resolution of all the main control factors through ArcGIS software, and finally unifying the main control factors into a grid unit.
Preferably, the step of quantifying the ground fault assessment factor by using a deterministic factor method CF includes the following steps:
considering the linear development characteristics of ground crack disasters, taking the ratio of ground crack length to each grading area under different grades in different ground crack evaluation factors as a main index, evaluating the relation between the conditional probability p (H|S) under different grading conditions and the prior probability p (F) of ground crack development density in an integral research area, and assigning a value to each independent factor according to the size of a CF value:
wherein p (F|S) represents the linear density of occurrence of the ground cracks in the specific ground crack evaluation factor value interval, and p (F) represents the linear density of the ground cracks in the whole research area.
Preferably, the whole time sequence deformation result of the evaluation area is obtained through a TS-InSAR technology, a coordinate system and resolution which are consistent with the ground crack evaluation factors are established in ArcGIS software, and an annual average deformation rate graph is obtained; processing the annual average deformation rate graph by combining the ground fracture occurrence mechanism, and determining a primary selection area of non-disaster points of the ground fracture by combining the pregnant disaster background, wherein the selection of the non-disaster points must meet the following four conditions at the same time:
a. the non-disaster point should be selected at a position at least 1km away from the existing ground crack data;
b. the non-disaster point should select the non-base rock area at the middle and low altitudes;
c. the non-disaster point should be selected to be at a location outside the range of at least 1km from the boundary of the relief;
d. and processing the acquired annual average deformation rate map by using an ArcGIS platform, removing the area with the annual average deformation rate of more than 10mm/a, and only reserving the part with the annual average deformation rate of less than 10 mm/a.
Preferably, the XGBoost model is constructed, that is, the XGBoost library in python is utilized, grid search GridSearch and CV cross validation are utilized to perform optimization on the hyper parameters of the model, the selected cross validation CV value is 5, and the optimized parameters through grid search are n_optimators=70, max_depth=10, subsamples=1, colsample_byte=0.79, gamma=0 and min_child_weight=1.
Preferably, the method further comprises the step of evaluating model precision according to the evaluation results corresponding to the training set and the vulnerability evaluation results output through the trained XGBoost model.
Preferably, the outputting of the susceptibility evaluation result by the trained XGBoost model, the model precision evaluation, includes the following calculation steps:
the ROC curve takes specificity as false positive rate and sensitivity as true positive rate, and respectively takes the specificity as the abscissa and the ordinate of the ROC curve, and takes values from a confusion matrix formed by the ground fracture evaluation factor spatial database distribution and the XGBoost model prediction result distribution to form curves under different threshold conditions;
area under curve AUC, the fitting degree and prediction precision of the training sample by the test model are improved;
the Kappa coefficient was calculated as:
wherein P is 0 Is the sum of the number of correctly classified samples of each class divided by the total number of samples, i.e., the overall classification accuracy; p (P) c Product sum of predicted number and actual number for each category divided by n 2 ：
Where TP is the number of positive samples successfully predicted, TN is the number of negative samples successfully predicted, FP is the number of negative samples incorrectly predicted as positive, and FN is the number of positive samples incorrectly predicted as negative.
Compared with the prior art, the invention has the following beneficial effects:
1. according to the method, the whole time sequence deformation result of the evaluation area is obtained through a TS-InSAR technology, and the area with the annual average deformation less than 10mm/a outside a 1km buffer area of a medium-low latitude base rock area, outside a 1km buffer area of an existing ground fracture, outside a 1km buffer area of a ground appearance boundary is taken as a primary selection area of a non-disaster point of the ground fracture, so that blindness of non-disaster point selection is reduced, and inaccurate model prediction caused by the fact that the non-disaster point falls in a high-probability area of the ground fracture is avoided.
2. The method is characterized in that the deterministic factor method CF is used for carrying out grading quantification on the ground fracture evaluation factors, the characteristics of the ground fracture linear geological disasters are considered, when the ground fracture evaluation factors are graded and assigned, the conditional probability of the occurrence of the ground fracture in different grades of the ground fracture evaluation factors is compared with the priori probability of the total fracture length in each factor, and the scientificity and rationality of the factor assignment are realized.
Drawings
FIG. 1 is a flow chart of the overall method of the present invention;
FIG. 2 is a flow chart of a sample database construction strategy of the present invention;
FIG. 3 is a graph of the results of a single-factor ROC analysis of the present invention;
FIG. 4 is a graph showing the result of evaluating the vulnerability to a severe-south-earth-crack disaster based on the XGBoost method of the present invention;
FIG. 5 (a) is a graph of the ROC model of the training effect of the CF-XGboost model of the present invention;
FIG. 5 (b) is a graph of a training effect model test confusion matrix for the CF-XGboost model of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to fig. 1 to 5 in the embodiments of the present invention. It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
Examples
As shown in fig. 1, an embodiment of the present invention provides a method for evaluating susceptibility of a ground fracture based on surface time-series deformation and a disaster-prone background, the method comprising:
a Wei river basin is selected as a typical crack development area, namely, a Wei south city of Shaanxi, and ground crack susceptibility evaluation of fusion ground surface time sequence deformation and a pregnant disaster background is carried out. The main flow is as follows:
1. according to the type of the ground fissure and the disaster-tolerant background, initially establishing a ground fissure environment factor space database:
(1) Extracting main control factors for inducing ground cracks according to development characteristics, cause types and different pregnancy factors of ground cracks
According to development characteristics of Weinan ground cracks, spatial relations of different pregnancy factors and typical crack causes, main crack types such as active fracture creeping ground cracks, earthquake induced ground cracks, regional micro-fracture open ground cracks, pumping subsidence ground cracks and the like are comprehensively considered, and the pregnancy factors are primarily selected: landform type, formation lithology, structural fracture, seismic intensity, fourth degree of lineage, evoked factors: the maximum annual average rainfall and the development degree of ground subsidence are seven factors to analyze.
(2) Judging the collinearity and the relativity of each main control factor by using a Schmidt orthogonalization SO method and a variance expansion factor VIF:
specifically, the schmitt orthogonalization SO method is implemented by the following formula:
β 1 ＝α 1
wherein alpha is i Representing any linear independent vector group, beta, in Euclidean space j The corresponding orthogonal vector group is represented, and j represents the number of factors.
The variance expansion factor VIF is realized by the following formula:
wherein R is i Negative correlation coefficients for regression analysis of the remaining independent variables are performed for the independent variables.
The larger the coefficient of variance expansion VIF, the greater the likelihood that co-linearity will exist between the independent variables. Generally, if the variance expansion factor exceeds 10, then the regression model has severe multiple collinearity.
And verifying the collinearity and availability of each master control factor. The VIF calculation results are shown in table 1:
TABLE 1 results of factor co-linearity analysis
The collinearity among the factors passes the inspection and the next step of hierarchical quantization is carried out.
(3) Grading and quantifying the evaluation factors according to a deterministic factor method CF
Considering the linear development characteristics of ground crack disasters, taking the ratio of the lengths of the ground cracks at different levels in different factors to the classification areas as a main index, evaluating the relation between the conditional probability p (H|S) under different classification conditions and the prior probability p (F) of the ground crack development density in the whole research area, and finally assigning a value to each independent factor according to the size of the CF value.
P (f|s) represents the linear density of occurrence of ground cracks in a specific factor value interval, and P (F) represents the linear density of ground cracks in the entire investigation region.
The hierarchical quantization results are shown in table 2:
table 2 factor hierarchical quantization
(4) And evaluating the influence of each factor on the ground crack disaster by a single-factor ROC curve method, removing the influence factors with smaller influence degree, and constructing a final master control factor database.
As shown in fig. 3, the maximum annual average rainfall, the thickness of the fourth line, and the seismic intensity in the left portion (i.e., when the confusion matrix judgment threshold is high) are lower than the diagonal reference line, and especially the maximum annual average rainfall is more obvious. The distribution of the maximum annual average rainfall does not have obvious influence on the development of ground cracks, so that the maximum annual average rainfall is removed, and finally six indexes including landform type, stratum lithology, structural fracture, earthquake intensity, fourth-line thickness and ground subsidence are selected to establish a factor space database.
(5) Unifying the coordinate system and resolution of all factors by ArcGIS software, and finally unifying to form grid unit
2. Acquiring an integral time sequence deformation result of an evaluation area through a TS-InSAR technology, and establishing a coordinate system and resolution consistent with an evaluation factor in an ArcGIS:
the example selects a time sequence InSAR method-SBAS-InSAR suitable for urban area monitoring, uses Sentinel-1A as a data source, selects SRTM DEM and POD as external data, and obtains a model of the average deformation rate of the Weinan city 2015-2022 year and a time sequence deformation evolution diagram based on an SBAS-InSAR module provided in ENVI-SARscape, as shown in figure 4.
And setting the same grid size and space coordinates of the InSAR and each evaluation factor in the ArcGIS. And only the area with the annual average deformation rate within the range of 10mm/a is reserved by utilizing the function of the grid calculator and is used as the selection basis of the non-ground crack disaster point.
3. Construction of a sample database
(1) Constructing a ground crack sample from the historical ground crack list. The ground cracks not only cover the traditional ground surface cracks, but also comprise holes and the like which are found through investigation, and the ground cracks are shared by 492 disaster sample points.
(2) As shown in fig. 2, a selection library area of non-ground fissure disaster points is constructed in combination with the ground fissure disaster background.
The selection of non-disaster points must satisfy the following four conditions simultaneously:
a. the non-disaster point should be selected at a position at least 1km away from the existing ground crack data
b. The non-disaster point should be selected from the non-bedrock area at medium and low altitudes
c. The non-disaster point should be selected to be a location outside the range of at least 1km from the boundary of the relief
d. And processing the acquired annual average deformation rate map by using an ArcGIS platform, removing the area with the annual average rate of more than 10mm/a, and only reserving the part with the annual average deformation rate of less than 10 mm/a.
Non-disaster points must be selected at the above locations. Considering that the possibility of expanding the ground cracks in the range of the existing ground crack 1km buffer area is higher, firstly, establishing a 1km buffer area around the existing ground cracks in ArcGIS software, and marking as an area A; in addition, from the viewpoint of mechanism analysis, the ground cracks are easy to occur on the demarcation line of the change of the landform, so that the landform boundary is extracted in the ArcGIS, and a buffer zone of 1km is established and is marked as a zone B; classifying the acquired DEM data in a grading manner, and marking a medium-low altitude bedrock area as a C area; the stable deformation region identified by the InSAR result is marked as a D region. Thereby establishingAnd pre-selecting areas, wherein non-crack disaster points are randomly selected from the pre-selected areas, and the number of the selected points is consistent with that of the disaster points and is 492.
4. Constructing a training set and a testing set of the model, and evaluating the susceptibility of the ground crack disaster by using the XGBoost model
The hyper parameters of the model were optimized using grid search (GridSearch) and CV cross validation using XGBoost library in python, with selected cross validation CV values of 5, and grid search optimization parameters of n_detectors=70, max_depth=10, subsamples=1, column_byte=0.79, gamma=0, min_child_weight=1.
5. Model accuracy assessment
The prediction accuracy of the model was evaluated by calculating the following index.
(1) ROC curve and AUC
As shown in fig. 5 (a), ROC curves are constructed under different threshold conditions with the false positive rate (FPR, 1-specificity) and true positive rate (TPR, sensitivity) plotted on the abscissa. The values are derived from a confusion matrix consisting of a true sample distribution and a model prediction result distribution. Area Under Curve (AUC) is an important evaluation index of the fitting degree and prediction accuracy of the test model to the training sample.
(2) Kappa coefficient
Wherein P is 0 Is the sum of the number of correctly classified samples of each class divided by the total number of samples, i.e., the overall classification accuracy; p (P) c Product sum of predicted number and actual number for each category divided by n 2 ：
Where TP is the number of positive samples successfully predicted, TN is the number of negative samples successfully predicted, FP is the number of negative samples incorrectly predicted as positive, and FN is the number of positive samples incorrectly predicted as negative.
As shown in fig. 5 (b), the AUC value is calculated to be 0.96, the confusion matrix is calculated by using a judgment threshold value of 0.5 as shown in table 3, the test accuracy calculated according to the confusion matrix result is calculated to be 0.90, and the kappa coefficient is calculated to be 0.80, which indicates that the fitting degree and the consistency degree of the model to the sample data set are extremely high.
TABLE 3 Multi-index based model test results
The foregoing is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art, who is within the scope of the present invention, should make equivalent substitutions or modifications according to the technical scheme of the present invention and the inventive concept thereof, and should be covered by the scope of the present invention.
In addition, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. All documents mentioned in this specification are incorporated by reference for the purpose of disclosing and describing the methodologies associated with the documents. In case of conflict with any incorporated document, the present specification will control.
Claims (8)
1. The ground crack susceptibility evaluation method based on the ground surface time sequence deformation and the disaster-prone background is characterized by comprising the following steps:
determining the type of the ground cracks to be evaluated and the disaster-tolerant background, and establishing a ground crack evaluation factor space database;
grading and quantifying the ground fracture evaluation factors by using a deterministic factor method CF to obtain grading and quantifying values of the ground fracture evaluation factors;
acquiring an integral time sequence deformation result of an evaluation area through a TS-InSAR technology, establishing a coordinate system and resolution consistent with a ground crack evaluation factor in ArcGIS software, and acquiring an annual average deformation rate graph; processing the annual average deformation rate graph by combining the ground fracture occurrence mechanism, and determining a primary selection area of a ground fracture non-disaster point by combining a disaster-tolerant background; determining a selected area of the ground crack disaster point according to the historical ground crack list data, and finally constructing a ground crack sample database;
constructing an XGBoost model, taking the quantized value and the ground crack sample database as the input of the XGBoost model, and outputting a susceptibility evaluation result through the trained XGBoost model.
2. The method for evaluating the susceptibility of a ground crack based on the surface time sequence deformation and the disaster-pregnancy background according to claim 1, wherein the determining the type of the ground crack to be evaluated and the disaster-pregnancy background comprises the following steps:
extracting a main control factor for inducing the ground cracks according to the development characteristics, the cause types and different pregnancy factors of the ground cracks;
judging the collinearity and the relativity of each main control factor by using a Schmidt orthogonalization SO method and calculating a variance expansion factor VIF, and determining an evaluation factor;
the calculation formula of the schmitt orthogonalization SO method is as follows:
β 1 ＝α 1
wherein alpha is i Representing any linear independent vector group, beta, in Euclidean space j Representing the corresponding orthogonal vector group, j representing the number of factors;
the calculation formula of the variance expansion factor VIF is as follows:
wherein R is i Negative correlation coefficients for regression analysis of the remaining independent variables are performed for the independent variables.
3. The method for evaluating the susceptibility of the ground cracks based on the surface time sequence deformation and the pregnant disaster background according to claim 2, wherein the step of establishing a space database of the ground crack evaluation factors is characterized in that the influence of each ground crack evaluation factor on the ground crack disasters is evaluated by a single-factor ROC curve method, the influence degree is ordered, and the influence degree is 1:1, removing factors in the line after analysis, and constructing a main control factor database;
and unifying the coordinate system and the resolution of all the main control factors through ArcGIS software, and finally unifying the main control factors into a grid unit.
4. The method for evaluating the susceptibility of the ground cracks based on the surface time sequence deformation and the pregnant disaster background according to claim 3, wherein the step of quantifying the ground crack evaluation factors in a grading manner by using a deterministic factor method CF comprises the following steps:
considering the linear development characteristics of ground crack disasters, taking the ratio of ground crack length to each grading area under different grades in different ground crack evaluation factors as a main index, evaluating the relation between the conditional probability p (H|S) under different grading conditions and the prior probability p (F) of ground crack development density in an integral research area, and assigning a value to each independent factor according to the size of a CF value:
wherein p (F|S) represents the linear density of occurrence of the ground cracks in the specific ground crack evaluation factor value interval, and p (F) represents the linear density of the ground cracks in the whole research area.
5. The method for evaluating the susceptibility of the ground cracks based on the ground surface time sequence deformation and the disaster-pregnant background according to claim 4, wherein the method is characterized in that the whole time sequence deformation result of an evaluation area is obtained through a TS-InSAR technology, a coordinate system and a resolution which are consistent with ground crack evaluation factors are established in ArcGIS software, and an annual average deformation rate diagram is obtained; processing the annual average deformation rate graph by combining the ground fracture occurrence mechanism, and determining a primary selection area of non-disaster points of the ground fracture by combining the pregnant disaster background, wherein the selection of the non-disaster points must meet the following four conditions at the same time:
a. the non-disaster point should be selected at a position at least 1km away from the existing ground crack data;
b. the non-disaster point should select the non-base rock area at the middle and low altitudes;
c. the non-disaster point should be selected to be at a location outside the range of at least 1km from the boundary of the relief;
d. and processing the acquired annual average deformation rate map by using an ArcGIS platform, removing the area with the annual average deformation rate of more than 10mm/a, and only reserving the part with the annual average deformation rate of less than 10 mm/a.
6. The method for evaluating the susceptibility to ground cracks based on surface time sequence deformation and a pregnant disaster background according to claim 5, wherein an XGBoost model is constructed, namely, a grid search GridSearch and CV cross validation method are utilized to optimize the superparameter of the model, the selected cross validation CV value is 5, and the optimization parameters of grid search are n_evastizer=70, max_depth=10, subsample=1, sample_byte=0.79, gamma=0 and min_child_weight=1.
7. The method for evaluating the susceptibility to ground cracks based on the surface time sequence deformation and the disaster-pregnant background according to claim 6, further comprising the step of evaluating model precision according to the evaluation results corresponding to the training set and the output of the susceptibility evaluation results through the trained XGBoost model.
8. The method for evaluating the susceptibility to ground cracks based on the surface time sequence deformation and the disaster recovery background according to claim 7, wherein the method for evaluating the susceptibility to ground cracks by outputting the susceptibility evaluation result through the trained XGBoost model is characterized by comprising the following calculation steps:
the ROC curve takes specificity as false positive rate and sensitivity as true positive rate, and respectively takes the specificity as the abscissa and the ordinate of the ROC curve, and takes values from a confusion matrix formed by the ground fracture evaluation factor spatial database distribution and the XGBoost model prediction result distribution to form curves under different threshold conditions;
area under curve AUC, the fitting degree and prediction precision of the training sample by the test model are improved;
the Kappa coefficient was calculated as:
wherein P is 0 Is the sum of the number of correctly classified samples of each class divided by the total number of samples, i.e., the overall classification accuracy; p (P) c Product sum of predicted number and actual number for each category divided by n 2 ：
Where TP is the number of positive samples successfully predicted, TN is the number of negative samples successfully predicted, FP is the number of negative samples incorrectly predicted as positive, and FN is the number of positive samples incorrectly predicted as negative.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310670192.6A CN117131756A (en) | 2023-06-07 | 2023-06-07 | Ground crack susceptibility evaluation method based on ground surface time sequence deformation and disaster-pregnancy background |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310670192.6A CN117131756A (en) | 2023-06-07 | 2023-06-07 | Ground crack susceptibility evaluation method based on ground surface time sequence deformation and disaster-pregnancy background |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117131756A true CN117131756A (en) | 2023-11-28 |
Family
ID=88855416
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310670192.6A Pending CN117131756A (en) | 2023-06-07 | 2023-06-07 | Ground crack susceptibility evaluation method based on ground surface time sequence deformation and disaster-pregnancy background |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117131756A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114186413A (en) * | 2021-12-10 | 2022-03-15 | 中国地质科学院探矿工艺研究所 | Landslide susceptibility evaluation method based on surface deformation and pregnant disaster environment conditions |
| US20230141886A1 (en) * | 2021-03-02 | 2023-05-11 | Hohai University | Method for assessing hazard on flood sensitivity based on ensemble learning |
-
2023
- 2023-06-07 CN CN202310670192.6A patent/CN117131756A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230141886A1 (en) * | 2021-03-02 | 2023-05-11 | Hohai University | Method for assessing hazard on flood sensitivity based on ensemble learning |
| CN114186413A (en) * | 2021-12-10 | 2022-03-15 | 中国地质科学院探矿工艺研究所 | Landslide susceptibility evaluation method based on surface deformation and pregnant disaster environment conditions |
Non-Patent Citations (2)
| Title |
|---|
| 刘强: "时序InSAR 技术在中型城市地表形变时空特征应用及预测分析", 中国优秀硕士学位论文全文数据库基础科学辑, no. 2022 * |
| 郭山川: "时序InSAR在城市地铁工程区形变监测中的应用", 测绘通报, no. 2017 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109034582B (en) | Tunnel crossing fault water inrush and mud inrush risk evaluation method based on cloud model and combined weighting | |
| CN108595414B (en) | Soil heavy metal enterprise pollution source identification method based on source-sink space variable reasoning | |
| CN111898315B (en) | Landslide susceptibility assessment method based on fractal-machine learning hybrid model | |
| CN115688404B (en) | Rainfall landslide early warning method based on SVM-RF model | |
| WO2023168781A1 (en) | Soil cadmium risk prediction method based on spatial-temporal interaction relationship | |
| CN111080356A (en) | Method for calculating residence price influence factors by using machine learning regression model | |
| CN111914943B (en) | Information vector machine method and device for comprehensively judging stability of dumping type karst dangerous rock | |
| CN114330812A (en) | Landslide disaster risk assessment method based on machine learning | |
| CN110889440A (en) | Rockburst grade prediction method and system based on principal component analysis and BP neural network | |
| CN114926067A (en) | Information quantity model-based geological disaster risk evaluation method by hierarchical analysis | |
| CN111507972A (en) | Tunnel surface defect detection method combining convolutional neural network and support vector machine | |
| US20220049606A1 (en) | Automated Updating of Geological Model Boundaries for Improved Ore Extraction | |
| CN111127465A (en) | Automatic generation method and system for bridge detection report | |
| CN115879654A (en) | Method for predicting stability and evaluating reliability of underground roadway | |
| CN113191605A (en) | House risk assessment method and device | |
| CN113836999A (en) | Tunnel construction risk intelligent identification method and system based on ground penetrating radar | |
| CN105184050A (en) | Landslide hazard degree evaluation method under support of GIS (Geographic Information System) and artificial intelligence technology | |
| CN111797188B (en) | Urban functional area quantitative identification method based on open source geospatial vector data | |
| CN112508336B (en) | Space and environmental efficiency correlation measurement method based on structural equation model | |
| CN114398951A (en) | Land use change driving factor mining method based on random forest and crowd-sourced geographic information | |
| CN116611592B (en) | Prediction method for geothermal abnormal region along railway corridor based on deep learning | |
| JP2000002769A (en) | Method and device for predicting the spatial distribution of geological structure to prepare geological map | |
| CN116992522A (en) | Deep foundation pit support structure deformation prediction method, device, equipment and storage medium | |
| CN109636194B (en) | Multi-source cooperative detection method and system for major change of power transmission and transformation project | |
| CN111984701A (en) | Method, device, equipment and storage medium for predicting village settlement evolution |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |