A county geological disaster risk analysis method and system

Abstract

The present application relates to the technical field of geological disaster analysis, and discloses a county geological disaster risk analysis method and system, which comprises the following steps: dividing a slope unit of a geological disaster area, screening out susceptibility analysis factors in the slope unit, selecting a training set and a test set from the geological disaster area and a non-geological disaster area; establishing an initial analysis model of the training set, optimizing the hyperparameters of the initial analysis model, outputting the susceptibility classification result of the training set by using the optimized analysis model, analyzing the classification accuracy of the susceptibility classification result, and determining the susceptibility analysis result of the slope unit; calculating the risk index of the slope unit, and dividing the risk subarea in the geological disaster area by using the risk index; screening out vulnerability analysis factors in the slope unit, performing hierarchical analysis on the vulnerability analysis factors by using a judgment matrix; and calculating the vulnerability index of the slope unit. The present application can improve the accuracy of the geological disaster analysis result to match the real geological disaster risk situation.

Description

TECHNICAL FIELD

[0001] The present application relates to a county geological disaster risk analysis method and system, belonging to the geological disaster analysis technical field. BACKGROUND

[0002] Geological disasters generally refer to events caused by a series of geological processes, which pose a serious threat to human life, property, nature and ecology. Geological disasters such as landslides, collapses and mudslides frequently occur in mountainous areas and have great destructive power. Based on the understanding of the occurrence regularity of previous geological disasters, carrying out geological disaster risk analysis can provide theoretical support and scientific basis for regional disaster prevention and mitigation. Geological disaster risk analysis is a quantitative analysis of the possibility of a risk area suffering from different intensity geological disasters and the possible disaster losses, including analysis of three levels of susceptibility, danger and vulnerability. Susceptibility refers to the possibility of geological disasters occurring in a certain region controlled by disaster-pregnant conditions. Danger refers to the possibility of a certain region occurring a certain scale and type of geological disaster in a certain time period under the action of a certain triggering factor. Vulnerability refers to the severity of the damage that the disaster-bearing body may suffer from geological disasters. Danger = susceptibility x inducing factor, risk = danger x vulnerability.

[0003] Geological disaster risk analysis is about the coupling results of the selection of analysis factors, the selection of analysis samples, the selection of analysis models, and the selection of analysis scales. A large number of geological disaster risk analysis methods have made important progress, but existing methods still have strong subjectivity and insufficient accuracy. How to select appropriate methods objectively in each link to improve the accuracy of the results and match the real geological disaster risk situation is a challenging technical problem. SUMMARY

[0004] The present application provides a county geological disaster risk analysis method and system, which mainly aims to match the real geological disaster risk situation by improving the accuracy of geological disaster analysis results.

[0005] To achieve the above-mentioned purpose, the present application provides a county geological disaster risk analysis method, which comprises:

[0006] Identify the geological disaster area and the non-geological disaster area of the county, divide the slope unit of the geological disaster area, screen the susceptibility analysis factors in the slope unit, and select the training set and the test set from the geological disaster area and the non-geological disaster area based on the susceptibility analysis factors;

[0007] An initial analysis model of the training set is established, hyperparameter optimization is performed on the initial analysis model to obtain optimized hyperparameters, an optimized analysis model corresponding to the optimized hyperparameters is obtained, the optimized analysis model is used to output the susceptibility classification results of the training set and the test set, the classification accuracy of the susceptibility classification results is analyzed, and the susceptibility analysis result of the slope unit is determined according to the classification accuracy;

[0008] A risk index of the slope unit is calculated, and the risk index is used to divide the risk subarea in the geological disaster area;

[0009] An easy-fragility analysis factor in the slope unit is screened, an index hierarchy of the easy-fragility analysis factor is generated, a judgment matrix of the index hierarchy is constructed, the easy-fragility analysis factor is analyzed hierarchically by using the judgment matrix, and a hierarchical analysis result is obtained;

[0010] An easy-fragility index of the slope unit is calculated by using the hierarchical analysis result, and an easy-fragility analysis result of the geological disaster area is determined according to the easy-fragility index;

[0011] Based on the susceptibility analysis result, the risk subarea, and the easy-fragility analysis result, a geological disaster risk analysis result of the slope unit is generated.

[0012] Optionally, the division of the slope unit in the geological disaster area comprises:

[0013] A topographic surface model of the geological disaster area is constructed;

[0014] The topographic surface model is subjected to depression filling to obtain a depression-free model;

[0015] A water flow direction is extracted from the depression-free model;

[0016] According to the water flow direction, a flow accumulation of the depression-free model is calculated;

[0017] Based on the flow accumulation, a river network map of the depression-free model is generated;

[0018] River connection information in the river network map is identified;

[0019] The river connection information is used to determine a catchment area and a reverse catchment area of the depression-free model;

[0020] The catchment area and the reverse catchment area are fused to obtain a slope unit model;

[0021] A slope unit corresponding to the slope unit model in the geological disaster area is queried.

[0022] Optionally, the screening of the susceptibility analysis factor in the slope unit comprises:

[0023] dividing the grid unit in the slope unit;

[0024] selecting a topographic factor, a geological condition factor, a hydrological condition factor, an engineering activity factor and a vegetation normalized index in the grid unit;

[0025] quantitatively grading the topographic factor, the geological condition factor, the hydrological condition factor, the engineering activity factor and the vegetation normalized index to obtain a quantitatively graded factor;

[0026] calculating the Pearson coefficient of the quantitatively graded factor by using the following formula:

[0027]

[0028] wherein, ρ x,y represents the Pearson coefficient, cov(x,y) represents the covariance between the x quantitatively graded factor and the y quantitatively graded factor, σ x represents the standard deviation of the x quantitatively graded factor, σ y represents the standard deviation of the y quantitatively graded factor, μ x represents the mean of the x quantitatively graded factor, μ y represents the mean of the y quantitatively graded factor, E[(x-μ x )(y-μ y )] represents the mathematical expectation of (x-μ x )(y-μ y ), x represents the quantitatively graded factor, and y represents the quantitatively graded factor other than x;

[0029] eliminating the quantitatively graded factor with the absolute value of the Pearson coefficient greater than a preset factor threshold to obtain a remaining graded factor;

[0030] calculating the factor median of the remaining graded factor;

[0031] generating a factor zoning map of the slope unit by using the factor median to obtain the susceptibility analysis factor in the slope unit;

[0032] wherein, the topographic factor comprises an elevation, a slope, a slope direction, a curvature, a surface roughness and a terrain relief index, the geological condition factor comprises a stratum lithology, an engineering geological rock group, a slope structure and a fault distance, the hydrological condition factor comprises a terrain humidity index, a water flow intensity index and a water system distance, and the engineering activity factor comprises a medium-high risk cutting slope house building density and a highway distance.

[0033] Optionally, the establishing of the initial analysis model of the training set comprises:

[0034] randomly splitting the training set to obtain a split training set;

[0035] calculating training set similarity between each split training set;

[0036] training set screening is performed on the split training set through the training set similarity to obtain a screened training set;

[0037] randomly extracting a training sample from the screened training set;

[0038] After sample splitting is performed on the training sample by using sample features of the training sample, a decision tree of the training set is determined;

[0039] An initial analysis model of the training set is established by using the decision tree.

[0040] Optionally, the initial analysis model is subjected to hyperparameter optimization to obtain optimized hyperparameters, including:

[0041] partitioning a to-be-optimized subspace of the initial analysis model;

[0042] uniformly sampling to-be-optimized hyperparameters in the to-be-optimized subspace;

[0043] analyzing an objective function of the to-be-optimized hyperparameters;

[0044] calculating a posterior distribution of the objective function;

[0045] According to the posterior distribution, an acquisition function value of the to-be-optimized hyperparameters is calculated by using the following formula:

[0046] AC(z) = μ(z) + kσ(z)

[0047] wherein, AC(z) represents the acquisition function value, μ(z) represents a mean value of the posterior distribution z, σ(z) represents a variance of the posterior distribution, and k represents a tuning parameter;

[0048] Based on the acquisition function value, a target subspace in the to-be-optimized subspace is identified;

[0049] The hyperparameters optimization is performed on the to-be-optimized hyperparameters in the target subspace through the objective function corresponding to the target subspace to obtain the optimized hyperparameters.

[0050] Optionally, the classification accuracy of the susceptibility classification result is analyzed, including:

[0051] An confusion matrix of the susceptibility classification result is constructed;

[0052] The true positive rate and the false positive rate of the susceptibility classification result are queried from the confusion matrix.

[0053] constructing a ROC curve of the true positive rate and the false positive rate;

[0054] calculating a curve area of the ROC curve;

[0055] analyzing classification accuracy of the susceptibility classification result by using the curve area.

[0056] Optionally, the calculating the risk index of the slope unit comprises:

[0057] calculating the instability probability of the slope unit by using the following formula:

[0058]

[0059] wherein, P i represents the instability probability of the i th slope unit in a preset time period under rainfall conditions, L max / day represents the maximum daily rainfall since the slope unit has monitoring records, L i represents the annual average maximum daily rainfall under different rainfall conditions;

[0060] according to the instability probability, calculating the risk index of the slope unit by using the following formula:

[0061]

[0062] wherein, H i represents the risk index of the i th slope unit under rainfall conditions, P i represents the instability probability of the i th slope unit in a preset time period under rainfall conditions, Y i represents the susceptibility index of the i th slope unit, Y max represents the maximum susceptibility index of the slope unit.

[0063] Optionally, the screening the vulnerability analysis factor in the slope unit comprises:

[0064] selecting the population number, the building number, the traffic road number, the life facility number and the important facility number in the slope unit;

[0065] quantitatively grading the population number, the building number, the traffic road number, the life facility number and the important facility number to obtain the vulnerability analysis factor;

[0066] wherein, the life facility number comprises the oil and gas pipeline number, the water pipeline number, the power line number and the communication line number, and the important facility number comprises the school number, the hospital number, the nursing home number and the power station number.

[0067] Optionally, the calculating the vulnerability index of the slope unit by using the hierarchical analysis result comprises:

[0068] obtaining an index weight value in the hierarchical analysis result;

[0069] calculating the vulnerability index of the slope unit by using the following formula according to the index weight value:

[0070]

[0071] wherein V i represents the vulnerability index, i represents the serial number of the slope unit, ω j represents the index weight value of the jth vulnerability analysis factor of the ith slope unit, y j represents the value of the jth vulnerability analysis factor of the ith slope unit after standardization, S i represents the slope damage coefficient of the ith slope unit.

[0072] In order to solve the above problems, the application further provides a county geological disaster risk analysis system, which comprises:

[0073] a training set selection module, which is used for identifying geological disaster areas and non-geological disaster areas in a county, dividing slope units in the geological disaster areas, screening vulnerability analysis factors in the slope units, selecting a training set and a test set from the geological disaster areas and the non-geological disaster areas based on the vulnerability analysis factors;

[0074] a vulnerability analysis module, which is used for establishing an initial analysis model of the training set, optimizing hyperparameters of the initial analysis model to obtain optimized hyperparameters, obtaining an optimized analysis model corresponding to the optimized hyperparameters, outputting vulnerability classification results of the training set and the test set by using the optimized analysis model, analyzing classification accuracy of the vulnerability classification results, and determining a vulnerability analysis result of the slope units according to the classification accuracy;

[0075] a partition division module, which is used for calculating a danger index of the slope units and dividing danger partitions in the geological disaster areas by using the danger index;

[0076] a hierarchical analysis module, which is used for screening vulnerability analysis factors in the slope units, generating an index hierarchy of the vulnerability analysis factors, constructing a judgment matrix of the index hierarchy, performing hierarchical analysis on the vulnerability analysis factors by using the judgment matrix, and obtaining a hierarchical analysis result;

[0077] a vulnerability analysis module configured to calculate a vulnerability index of the slope unit by using the hierarchical analysis result, and determine a vulnerability analysis result of the geological disaster area according to the vulnerability index;

[0078] a risk analysis module configured to generate a geological disaster risk analysis result of the slope unit based on the proneness analysis result, the danger zoning result and the vulnerability analysis result.

[0079] Compared with the problems in the background art, the method takes the proneness-danger-vulnerability evaluation as the idea, constructs a random forest-ARE-AHP coupling model (RF-ARE-AHP) based on the random forest model, the Bayesian optimization method, the extreme rainfall hazard zoning, the chromatography analysis method and the slope gradient discount analysis, and is used for the geological disaster risk evaluation of the slope unit at the county scale. BRIEF DESCRIPTION OF DRAWINGS

[0080] Figure 1 a flowchart of a county geological disaster risk analysis method provided by an embodiment of the present application;

[0081] Figure 2 a confusion matrix diagram for implementing the county geological disaster risk analysis method provided by an embodiment of the present application;

[0082] Figure 3 a module diagram for implementing the county geological disaster risk analysis method provided by an embodiment of the present application.

[0083] The object implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0084] It should be understood that the specific embodiments described herein are merely intended to explain the present application, and are not intended to limit the present application.

[0085] The embodiments of the present application provide a county geological disaster risk analysis method. An execution subject of the county geological disaster risk analysis method includes but is not limited to at least one of electronic devices such as a server and a terminal which can be configured to execute the method provided by the embodiments of the present application. In other words, the county geological disaster risk analysis method can be executed by software or hardware installed in a terminal device or a server device. The server includes but is not limited to a single server, a server cluster, a cloud server or a cloud server cluster.

[0086] Embodiment 1

[0087] Referring to Figure 1 Fig. 1 is a flowchart of a county geological disaster risk analysis method according to an embodiment of the present application. In this embodiment, the county geological disaster risk analysis method comprises the following steps.

[0088] S1, identifying a geological disaster area and a non-geological disaster area in a county, dividing slope units in the geological disaster area, screening susceptibility analysis factors in the slope units, and selecting a training set and a test set from the geological disaster area and the non-geological disaster area based on the susceptibility analysis factors.

[0089] In this embodiment, the geological disaster area refers to an area where geological disasters have occurred, and the non-geological disaster area refers to an area where no geological disasters have occurred.

[0090] Further, in this embodiment, the slope units in the geological disaster area are divided to analyze the risk of geological sources causing geological disasters. The slope unit refers to a landform unit defined by a ridge line and a valley line, which is related to hydrological conditions and processes that shape the topography and geomorphology, and is a basic unit of geological disaster development.

[0091] In an embodiment of the present application, the division of the slope units in the geological disaster area comprises: constructing a topographic surface model of the geological disaster area; performing depression filling on the topographic surface model to obtain a depression-free model; extracting a water flow direction from the depression-free model; calculating a flow accumulation of the depression-free model according to the water flow direction; generating a river network map of the depression-free model based on the flow accumulation; identifying river connection information in the river network map; determining a catchment area and a reverse catchment area of the depression-free model using the river connection information; fusing the catchment area and the reverse catchment area to obtain a slope unit model; and querying a slope unit corresponding to the slope unit model in the geological disaster area.

[0092] The topographic surface model refers to a DEM model.

[0093] Optionally, the process of filling the low-lying area of the terrain surface model is realized by the "filling low-lying area" operation of the ArcGIS software, further, the process of extracting the water flow direction from the low-lying area model is realized by the "flow direction" operation of the ArcGIS software, further, the process of calculating the flow accumulation of the low-lying area model according to the water flow direction refers to assigning different water levels to each grid node of the DEM, according to the natural law that water flows from high to low, calculating the water flow at each node according to the regional topography and water flow direction, thereby obtaining the flow accumulation of the region, further, the process of generating the river network graph of the low-lying area model based on the flow accumulation refers to setting the threshold value of the flow accumulation according to the test results and topographic data and the like for auxiliary inspection, wherein different levels of valleys correspond to different threshold values, and the threshold values of the valleys of the same level but in different regions are also different, the grid attribute greater than the threshold value is assigned a value of 1, and the grid attribute less than or equal to the threshold value is set as no data (null), thereby obtaining the grid file of the river network, and then the river network graph is generated through the raster-to-vector conversion, the process of identifying the river connection information in the river network graph refers to identifying the connection information between the nodes in the river network graph, for example, each arc segment connects two nodes as water outlet points or confluence points, or connects a node as a water outlet point and a starting point of the river network, thereby obtaining the starting point and the ending point of each river network arc segment, the process of determining the catchment area and the reverse catchment area of the low-lying area model by using the river connection information refers to extracting the ridge line by hydrological analysis of the positive DEM and extracting the valley line by hydrological analysis of the inverted DEM, and the ridge line and the valley line represent the catchment area and the reverse catchment area, and the process of fusing the catchment area and the reverse catchment area refers to taking the positions of the ridge line and the valley line in the DEM as the slope unit model.

[0094] Further, the embodiment of the present application screens the factors that can be used for the susceptibility risk analysis by screening the susceptibility analysis factors in the slope unit.

[0095] In an embodiment of the present application, the screening of the susceptibility analysis factors in the slope unit comprises: dividing the grid unit in the slope unit; selecting the topographic factor, the geological condition factor, the hydrological condition factor, the engineering activity factor and the normalized difference vegetation index in the grid unit; quantitatively grading the topographic factor, the geological condition factor, the hydrological condition factor, the engineering activity factor and the normalized difference vegetation index to obtain the quantitatively graded factors; and calculating the Pearson coefficient of the quantitatively graded factors by using the following formula:

[0096]

[0097] wherein, ρ x,ydenotes the Pearson coefficient, cov(x, y) denotes the covariance between the x quantile factor and the y quantile factor, σ x denotes the standard deviation of the x quantile factor, σ y denotes the standard deviation of the y quantile factor, μ x denotes the mean of the x quantile factor, μ y denotes the mean of the y quantile factor, E[(x-μ x )(y-μ y )] denotes the mathematical expectation of (x-μ x )(y-μ y ), x denotes a quantile factor, and y denotes a quantile factor other than x.

[0098] quantile factors whose absolute values of the Pearson coefficient are greater than a preset factor threshold value, to obtain residual quantile factors; a factor median of the residual quantile factors is calculated; a factor zoning map of the ramp unit is generated using the factor median, to obtain a susceptibility analysis factor in the ramp unit; wherein the topographic and geomorphic factors include elevation, slope, slope direction, curvature, surface roughness, and topographic relief index, the geological condition factors include stratum lithology, engineering geology rock group, slope structure, and fault distance, the hydrological condition factors include topographic humidity index, water flow intensity index, and water system distance, and the engineering activity factors include medium-high risk cut slope building density and highway distance.

[0099] The grid unit refers to a regular grid into which space is divided, each grid is referred to as a unit, and the preset factor threshold value refers to a threshold value of the absolute value of the Pearson coefficient, for example, 0.9, which aims to reduce data redundancy and improve the relative independence between the susceptibility evaluation factors. The correlation between the evaluation factors is evaluated by calculating the Pearson correlation coefficient, and factors whose absolute values of the correlation coefficient are all greater than 0.5 are removed, thereby avoiding the problem of multicollinearity.

[0100] Optionally, the process of dividing the grid units in the ramp unit, for example, converts all susceptibility evaluation factors into factors within a 10m*10m grid unit, the process of quantitatively grading the topographic and geomorphic factors, the geological condition factors, the hydrological condition factors, the engineering activity factors, and the vegetation normalization index to obtain quantile factors refers to quantitatively grading multiple factors in the grid unit by using the natural breakpoint method, and the process of generating a factor zoning map of the ramp unit using the factor median refers to calculating the median of the susceptibility evaluation factors in each ramp unit by using the "show zoning statistics in table" tool in ArcGIS 10.3 software, finally generating a susceptibility evaluation factor zoning map based on the ramp unit, and taking the susceptibility evaluation factor zoning map as the susceptibility analysis factor.

[0101] Optionally, the process of selecting the training set and the test set from the geological disaster area and the non-geological disaster area based on the susceptibility analysis factor, for example: in order to construct a binary classification model for machine learning, it is necessary to select geological disaster positive samples and non-geological disaster negative samples to learn and train the random forest model, take the slope unit where the geological disaster is located as the geological disaster positive sample, randomly select the same number of non-geological disaster slope units as the negative sample, divide the sample set composed of the positive sample and the negative sample into a training set and a test set, 70% of the samples are used for training the model, and 30% of the samples are used for testing the model. It should be noted that the training set and the test set are data of the slope unit in the historical period, and in the subsequent process of determining the susceptibility analysis result of the slope unit, the susceptibility analysis result is determined based on the data of the slope unit in the current period.

[0102] S2, an initial analysis model of the training set is established, the initial analysis model is subjected to hyperparameter optimization to obtain optimized hyperparameters, an optimized analysis model corresponding to the optimized hyperparameters is acquired, the optimized analysis model is used to output the susceptibility classification results of the training set and the test set, the classification accuracy of the susceptibility classification results is analyzed, and the susceptibility analysis result of the slope unit is determined according to the classification accuracy.

[0103] In the embodiment of the present application, the initial analysis model of the training set is established to establish a random forest model by using the training set.

[0104] In an embodiment of the present application, the initial analysis model of the training set is established, including: randomly dividing the training set to obtain a divided training set; calculating the training set similarity between each divided training set; training set screening is performed on the divided training set by using the training set similarity to obtain a screened training set; training samples are randomly extracted from the screened training set; after sample splitting is performed on the training samples by using the sample features of the training samples, a decision tree of the training set is determined; and the initial analysis model of the training set is established by using the decision tree.

[0105] Optionally, the process of randomly dividing the training set to obtain a divided training set refers to that the training set of each decision tree is composed of n training samples randomly and with replacement from the original training set by the bootstrap method, and the training set of each decision tree is different, and the training set may also contain repeatedly selected training samples, therefore, it is necessary to randomly divide the training set to obtain a divided training set, and it is also necessary to calculate the training set similarity to ensure that each divided training set is different, so as to establish a random forest model for each divided training set, further, the process of randomly extracting training samples from the screening training set refers to that if the screening training set contains M features, m features (m < M) are randomly and non-repeatedly selected from the M features when each node is generated, the samples are divided by using the m features, and the best feature is found to split the generated node, further, the process of determining the decision tree of the training set after sample splitting of the training sample by using the sample features of the training sample refers to that the foregoing step K is repeated K times, that is, K decision trees are obtained, and finally the classification results of the K decision trees (that is, the susceptibility classification results corresponding to each training set) are integrated to obtain an initial analysis model.

[0106] Further, the embodiment of the present application optimizes the hyperparameters of the initial analysis model to optimize the parameters of the established random forest model and improve the output accuracy of the random forest model.

[0107] In an embodiment of the present application, the hyperparameter optimization of the initial analysis model to obtain the optimized hyperparameters comprises: dividing a to-be-optimized subspace of the initial analysis model; uniformly sampling to-be-optimized hyperparameters in the to-be-optimized subspace; analyzing an objective function of the to-be-optimized hyperparameters; calculating a posterior distribution of the objective function; and calculating a collection function value of the to-be-optimized hyperparameters according to the posterior distribution by using the following formula:

[0108] AC(z) = μ(z) + kσ(z)

[0109] wherein, AC(z) represents the collection function value, μ(z) represents the mean of the posterior distribution z, σ(z) represents the variance of the posterior distribution, and k represents a tuning parameter.

[0110] Based on the collection function value, a target subspace in the to-be-optimized subspace is identified, and the hyperparameters in the target subspace are optimized by using the objective function corresponding to the target subspace to obtain the optimized hyperparameters.

[0111] wherein, the to-be-optimized subspace refers to a continuous forest region containing N hyperparameters randomly selected in the initial analysis model, and the objective function refers to a loss value corresponding to the to-be-optimized hyperparameters.

[0112] Optionally, the process of calculating the posterior distribution of the objective function refers to using a Gaussian process to estimate the posterior distribution of the objective function of each sampled hyperparameter combination, further, the process of identifying the target subspace in the to-be-optimized subspace based on the collected function values refers to using the collected function values to identify the hyperparameter forest area that can optimize the objective function, and the process of performing hyperparameter optimization on the to-be-optimized hyperparameters in the target subspace by using the objective function corresponding to the target subspace refers to using a Bayesian optimization algorithm to optimize the objective function corresponding to the target subspace.

[0113] Further, in the embodiment of the present application, the susceptibility classification result refers to the susceptibility evaluation result of the optimization analysis model on the susceptibility analysis factors in the training set and the test set.

[0114] Further, the embodiment of the present application analyzes the classification accuracy of the susceptibility classification result to evaluate the classification accuracy of the optimization analysis model, and displays the classification ability of the model by plotting the true positive rate (TPR) and the false positive rate (FPR) under different thresholds. The area under the curve (AUC) of the ROC curve is a quantitative index for measuring the performance of the random forest model. The closer the AUC value is to 1, the better the classification performance of the random forest model. Since the ROC curve is not sensitive to the distribution of positive and negative samples, it can provide a relatively comprehensive evaluation of the model performance.

[0115] In an embodiment of the present application, the analysis of the classification accuracy of the susceptibility classification result comprises: constructing a confusion matrix of the susceptibility classification result; querying the true positive rate and the false positive rate of the susceptibility classification result from the confusion matrix; constructing an ROC curve of the true positive rate and the false positive rate; calculating the curve area of the ROC curve; and analyzing the classification accuracy of the susceptibility classification result by using the curve area.

[0116] Referring to Figure 2 Fig. 1 shows a confusion matrix for implementing the county geological disaster risk analysis method according to an embodiment of the present application. Figure 2 In the figure, the true positive rate is the positive sample sensitivity TPR, the false positive rate is the positive sample specificity FPR, the true negative rate is the negative sample sensitivity TNR, the false negative rate is the negative sample specificity FNR, P is the number of medium-high susceptibility slope units obtained by field investigation, N is the number of low susceptibility slope units obtained by field investigation, the true positive (TP) and the true negative (TN) are the numbers of correctly classified positive and negative samples respectively, the false positive (FP) and the false negative (FN) are the numbers of incorrectly classified positive and negative samples respectively, and the true positive rate TPR represents the proportion of correctly classified positive samples.

[0117] Optionally, the process of determining the susceptibility analysis result of the slope unit according to the classification accuracy refers to continuing to optimize the random forest model when the classification accuracy is insufficient, and using the optimized random forest model to evaluate the susceptibility of the slope unit in the current period when the classification accuracy is sufficient.

[0118] S3, calculating a risk index of the slope unit, and dividing the risk zoning in the geological disaster area by using the risk index.

[0119] The embodiment of the present application calculates the risk index of the slope unit to evaluate the risk in the geological disaster area by using the risk index, adopts a geological disaster risk evaluation method based on the extreme rainfall assumption method, fully considers the historical facts of extreme rainfall inducing geological disasters in the area, combines the extreme rainfall records and dynamic rainfall intensity, highlights the relationship between the geological disaster risk evaluation and the inducing factor probability, and embodies the dynamic evaluation process of the geological disaster risk.

[0120] In an embodiment of the present application, the process of calculating the risk index of the slope unit comprises: calculating the instability probability of the slope unit by using the following formula:

[0121]

[0122] Wherein, P i represents the instability probability of the i th slope unit in the preset period under the rainfall working condition, L max / day represents the daily maximum rainfall since the slope unit has monitoring records, L i represents the annual average maximum daily rainfall under different rainfall working conditions.

[0123] According to the instability probability, the risk index of the slope unit is calculated by using the following formula:

[0124]

[0125] Wherein, H i represents the risk index of the i th slope unit under the rainfall working condition, P i represents the instability probability of the i th slope unit in the preset period under the rainfall working condition, Y i represents the susceptibility index of the i th slope unit, Y max represents the maximum susceptibility index of the slope unit.

[0126] Optionally, the process of dividing the risk zoning in the geological disaster area by using the risk index refers to obtaining the risk zoning under different rainfall working conditions by calculating the risk index of each slope unit.

[0127] S4, screen the vulnerability analysis factors in the slope unit, generate an index hierarchy of the vulnerability analysis factors, construct a judgment matrix of the index hierarchy, perform hierarchical analysis on the vulnerability analysis factors by using the judgment matrix, and obtain a hierarchical analysis result.

[0128] The embodiment of the present application selects the factors capable of evaluating the vulnerability of the slope unit by screening the vulnerability analysis factors in the slope unit.

[0129] In an embodiment of the present application, the screening of the vulnerability analysis factors in the slope unit comprises: selecting the population number, the building number, the traffic road number, the life facility number and the important facility number in the slope unit; and quantitatively grading the population number, the building number, the traffic road number, the life facility number and the important facility number to obtain the vulnerability analysis factors; wherein the life facility number comprises the oil and gas pipeline number, the water pipeline number, the power transmission line number and the communication line number, and the important facility number comprises the school number, the hospital number, the nursing home number and the power station number.

[0130] Optionally, the process of quantitatively grading the population number, the building number, the traffic road number, the life facility number and the important facility number to obtain the vulnerability analysis factors is performed by using the natural break method to quantitatively grade the vulnerability evaluation factors based on the grid unit.

[0131] Optionally, the process of generating the index hierarchy of the vulnerability analysis factors comprises: decomposing the elements related to decision-making into multiple levels such as target, criterion, scheme and the like, taking the vulnerability evaluation result as the target layer, taking the population vulnerability, the material vulnerability and the important facility vulnerability as the criterion layer, taking the population, the building, the traffic road, the life facility and the important facility as the scheme layer, and further, the process of constructing the judgment matrix of the index hierarchy comprises: determining the weight of each criterion layer to the target layer by subjectively comparing each vulnerability analysis factor with each other according to the analysis and research, and further, the process of performing hierarchical analysis on the vulnerability analysis factors by using the judgment matrix to obtain the hierarchical analysis result comprises, for example: using the analytic hierarchy process to perform the evaluation method, combining the qualitative method with the quantitative method to calculate the relative importance between each evaluation factor, and thereby calculating the weight value of the vulnerability analysis factor.

[0132] S5, calculating the vulnerability index of the slope unit by using the hierarchical analysis result, and determining the vulnerability analysis result of the geological disaster area according to the vulnerability index.

[0133] The embodiment of the present application calculates the vulnerability index of the slope unit by using the hierarchical analysis result, and calculates the vulnerability index of the slope unit by weighted summation, so that when large-scale regional vulnerability evaluation is performed, the overall terrain fluctuation is large, a large number of population buildings are gathered in the county and town settlement areas with flat terrain, and these areas are often classified as high vulnerability or extremely high vulnerability in vulnerability evaluation. However, the risk evaluation method based on the traditional formula "risk (R) = hazard (H) x vulnerability (V)" will inevitably lead to some flat areas which are not actually threatened by geological disasters being classified into medium-high risk areas, and therefore the vulnerability evaluation is performed according to the loss coefficient.

[0134] In an embodiment of the present application, the calculation of the vulnerability index of the slope unit by using the hierarchical analysis result comprises: obtaining an index weight value in the hierarchical analysis result; and calculating the vulnerability index of the slope unit by using the following formula according to the index weight value:

[0135]

[0136] wherein V i represents the vulnerability index, i represents the serial number of the slope unit, ω j represents the index weight value of the jth vulnerability analysis factor of the ith slope unit, y j represents the value of the jth vulnerability analysis factor of the ith slope unit after standardization, S i represents the slope loss coefficient of the ith slope unit.

[0137] The vulnerability analysis result is the vulnerability index.

[0138] It should be noted that when the slope of the slope unit is less than 5°, the slope unit has no instability risk, and the slope loss coefficient is 0; when the slope is between 5° and 15°, only the buildings near the slope toe are threatened by geological disasters, and the slope loss coefficient is 0.1; and when the slope is greater than 35°, all threatened objects in the entire slope unit are threatened by geological disasters, and the slope loss coefficient is 1. The slope loss coefficient is determined by expert experience method, so that the object of vulnerability evaluation is focused on the area actually threatened by geological disasters.

[0139] S6, generating a geological disaster risk analysis result of the slope unit based on the susceptibility analysis result, the hazard zoning and the vulnerability analysis result.

[0140] Optionally, the process of generating the geological disaster risk analysis result of the slope unit based on the susceptibility analysis result, the risk zoning and the vulnerability analysis result is that: through the evaluation of the geological disaster susceptibility, risk and vulnerability of the evaluation region, the evaluation results are standardized, and based on the risk evaluation model "risk (R) = risk (H) x vulnerability (V)" and the susceptibility analysis result, the geological disaster risk zoning of different rainfall conditions is divided by the natural breakpoint method, that is, into a low risk area, a medium risk area, a high risk area and an extremely high risk area.

[0141] Compared with the problems in the background art, the method takes the susceptibility-risk-vulnerability evaluation as the idea, based on the random forest model, the Bayesian optimization method, the extreme rainfall condition risk zoning, the chromatography analysis method and the slope discount analysis method, a random forest-extreme rainfall assumption-coupling model (RF-ARE-AHP) is constructed for the county scale slope unit geological disaster risk evaluation, and the area (AUC) value under the ROC curve, the confusion matrix parameters and the field survey results are used for model verification, a high-precision risk zoning map based on the slope unit is made, and the verification results show that the problems of strong subjectivity and insufficient accuracy of the traditional method are better solved, and reference is provided for county scale geological disaster risk evaluation and disaster prevention and reduction.

[0142] Embodiment 2:

[0143] As shown in Figure 3 It is a county geological disaster risk analysis system function module diagram of the present application.

[0144] The county geological disaster risk analysis system 300 can be installed in an electronic device. According to the functions realized, the county geological disaster risk analysis system can include a training set selection module 301, a susceptibility analysis module 302, a zoning division module 303, a hierarchical analysis module 304, a vulnerability analysis module 305 and a risk analysis module 306. The modules of the present application can also be referred to as units, which refer to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0145] In the embodiments of the present application, the functions of each module / unit are as follows:

[0146] The training set selection module 301 is used to identify the geological disaster area and the non-geological disaster area of the county, divide the slope unit of the geological disaster area, screen the susceptibility analysis factors in the slope unit, select the training set and the test set from the geological disaster area and the non-geological disaster area based on the susceptibility analysis factors;

[0147] The proneness analysis module 302 is configured to establish an initial analysis model of the training set, perform hyperparameter optimization on the initial analysis model to obtain optimized hyperparameters, acquire an optimized analysis model corresponding to the optimized hyperparameters, output proneness classification results of the training set and the test set by using the optimized analysis model, analyze classification accuracy of the proneness classification results, and determine a proneness analysis result of the slope unit according to the classification accuracy.

[0148] The partition division module 303 is configured to calculate a risk index of the slope unit, and divide a risk partition in the geological disaster area by using the risk index.

[0149] The hierarchical analysis module 304 is configured to screen an easy-fragility analysis factor in the slope unit, generate an index hierarchical structure of the easy-fragility analysis factor, construct a judgment matrix of the index hierarchical structure, perform hierarchical analysis on the easy-fragility analysis factor by using the judgment matrix, and obtain a hierarchical analysis result.

[0150] The easy-fragility analysis module 305 is configured to calculate an easy-fragility index of the slope unit by using the hierarchical analysis result, and determine an easy-fragility analysis result of the geological disaster area according to the easy-fragility index.

[0151] The risk analysis module 306 is configured to generate a geological disaster risk analysis result of the slope unit based on the proneness analysis result, the risk partition, and the easy-fragility analysis result.

[0152] In detail, the modules in the county geological disaster risk analysis system 300 in the embodiments of the present application use the same technical means as the county geological disaster risk analysis method in the above Figure 1 , and can produce the same technical effects, which will not be described here.

[0153] It is obvious for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0154] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application but not limit the present application, and although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application.

Claims

1. A county geological disaster risk analysis method, characterized in that, The method comprises: Identify the geological disaster area and the non-geological disaster area of the county area, divide the slope unit of the geological disaster area, screen the susceptibility analysis factors in the slope unit, select the training set and the test set from the geological disaster area and the non-geological disaster area based on the susceptibility analysis factors; Establish the initial analysis model of the training set, specifically including: randomly dividing the training set to obtain a divided training set; calculating the training set similarity between each divided training set; training set screening is performed on the divided training set through the training set similarity to obtain a screened training set; training samples are randomly extracted from the screened training set; after sample splitting of the training samples using sample characteristics of the training samples, a decision tree of the training set is determined; and an initial analysis model of the training set is established using the decision tree; Optimize the hyperparameters of the initial analysis model to obtain optimized hyperparameters, specifically including: dividing the to-be-optimized subspace of the initial analysis model; uniformly sampling the to-be-optimized hyperparameters in the to-be-optimized subspace; analyzing the objective function of the to-be-optimized hyperparameters; calculating the posterior distribution of the objective function; and calculating the acquisition function value of the to-be-optimized hyperparameters according to the posterior distribution using the following formula: AC(z) = μ(z) + kσ(z) Wherein, AC(z) represents the acquisition function value, μ(z) represents the mean of the posterior distribution z, σ(z) represents the variance of the posterior distribution, and k represents a tuning parameter; Based on the acquisition function value, identify the target subspace in the to-be-optimized subspace; and through the objective function corresponding to the target subspace, optimize the hyperparameters of the to-be-optimized hyperparameters in the target subspace to obtain optimized hyperparameters; Obtain the optimized analysis model corresponding to the optimized hyperparameters, output the susceptibility classification results of the training set and the test set using the optimized analysis model, analyze the classification accuracy of the susceptibility classification results, determine the susceptibility analysis result of the slope unit according to the classification accuracy; Calculate the risk index of the slope unit, and divide the risk zoning in the geological disaster area using the risk index; Screen the vulnerability analysis factors in the slope unit, generate the index hierarchy of the vulnerability analysis factors, construct the judgment matrix of the index hierarchy, perform hierarchical analysis on the vulnerability analysis factors using the judgment matrix, and obtain a hierarchical analysis result; Calculate the vulnerability index of the slope unit using the hierarchical analysis result, and determine the vulnerability analysis result of the geological disaster area according to the vulnerability index; Based on the susceptibility analysis result, the risk zoning and the vulnerability analysis result, generate the geological disaster risk analysis result of the slope unit.

2. The county geologic hazard risk analysis method of claim 1, wherein, The division of the slope unit of the geological disaster area comprises: Construct a topographic surface model of the geological disaster area; Perform depression filling on the topographic surface model to obtain a depression-free model; Extract the water flow direction from the depression-free model; According to the water flow direction, calculate the flow accumulation of the depression-free model; generating a river network graph of the depression-free model based on the confluence accumulation; identifying river connection information in the river network graph; determining a catchment basin and a reverse catchment basin of the depression-free model by using the river connection information; fusing the catchment basin and the reverse catchment basin to obtain a slope unit model; querying a slope unit corresponding to the slope unit model in the geological disaster area.

3. The county-based geological disaster risk analysis method of claim 1, wherein, The screening of the susceptibility analysis factor in the slope unit comprises: dividing a grid unit in the slope unit; selecting a topographic factor, a geological condition factor, a hydrological condition factor, an engineering activity factor and a vegetation normalized index in the grid unit; quantitatively grading the topographic factor, the geological condition factor, the hydrological condition factor, the engineering activity factor and the vegetation normalized index to obtain a quantitatively graded factor; calculating a Pearson coefficient of the quantitatively graded factor by using the following formula: wherein p x,y represents the Pearson coefficient, cov(x, y) represents the covariance between the quantization scale factor x and the quantization scale factor y, σ x represents the standard deviation of the quantization scale factor x, σ y represents the standard deviation of the quantization scale factor y, μ x represents the mean of the quantization scale factor x, μ y represents the mean of the quantization scale factor y, E[(x-μ x )(y-μ y )] represents the mathematical expectation of (x-μ x )(y-μ y ), x represents the quantization scale factor, and y represents the quantization scale factor other than x. eliminating a quantitatively graded factor with an absolute value of the Pearson coefficient greater than a preset factor threshold to obtain a remaining graded factor; calculating a factor median of the remaining graded factor; generating a factor zoning map of the slope unit by using the factor median to obtain a susceptibility analysis factor in the slope unit; The topographic factor comprises an elevation, a slope, a slope direction, a curvature, a surface roughness and a terrain relief index, the geological condition factor comprises a stratum lithology, an engineering geology rock group, a slope structure and a fault distance, the hydrological condition factor comprises a terrain humidity index, a water flow intensity index and a water system distance, and the engineering activity factor comprises a medium-high risk cutting slope house building density and a highway distance.

4. The county geologic hazard risk analysis method of claim 1, wherein, The analysis of the classification accuracy of the susceptibility classification result comprises: constructing a confusion matrix of the susceptibility classification result; querying a true positive rate and a false positive rate of the susceptibility classification result from the confusion matrix; constructing a ROC curve of the true positive rate and the false positive rate; calculating an area of the ROC curve; analyzing the classification accuracy of the susceptibility classification result by using the area.

5. The county-based geological disaster risk analysis method of claim 1, wherein, The calculation of the risk index of the slope unit comprises: calculating an instability probability of the slope unit by using the following formula: where P i represents the instability probability of the i-th slope unit in a preset time period under rainfall conditions, L max / day represents the daily maximum rainfall since the slope unit has been monitored, L i represents the annual average maximum daily rainfall under different rainfall conditions; calculating a risk index of the slope unit by using the following formula according to the instability probability: wherein H i represents the hazard index of the i-th slope unit under rainfall conditions, P i represents the instability probability of the i-th slope unit in a preset time period under rainfall conditions, Y i represents the susceptibility index of the i-th slope unit, Y max represents the maximum susceptibility index of the slope unit.

6. The county geologic hazard risk analysis method of claim 1, wherein, The screening of the vulnerability analysis factor in the slope unit comprises: selecting a population number, a building number, a traffic road number, a life facility number and an important facility number in the slope unit; quantitatively grading the population number, the building number, the traffic road number, the life facility number and the important facility number to obtain a vulnerability analysis factor; The life facility number comprises an oil and gas pipeline number, a water pipeline number, a power line number and a communication line number, and the important facility number comprises a school number, a hospital number, a nursing home number and a power station number.

7. The county geologic hazard risk analysis method of claim 1, wherein, The calculation of the vulnerability index of the slope unit by using the hierarchical analysis result comprises: obtaining an index weight value in the hierarchical analysis result; calculating a vulnerability index of the slope unit by using the following formula according to the index weight value: wherein V i represents the vulnerability index, i represents the serial number of the slope unit, ω j represents the index weight value of the jth vulnerability analysis factor of the ith slope unit, y j represents the value after standardization of the jth vulnerability analysis factor of the ith slope unit, S i represents the slope discount coefficient of the ith slope unit.

8. A county area geological disaster risk analysis system, characterized in that, The system comprises: A training set selection module is used to identify geological disaster areas and non-geological disaster areas in a county, divide the slope units in the geological disaster areas, screen the susceptibility analysis factors in the slope units, and select training sets and test sets from the geological disaster areas and the non-geological disaster areas based on the susceptibility analysis factors; The susceptibility analysis module is used to establish an initial analysis model for the training set, specifically comprising: randomly splitting the training set to obtain split training sets; calculating the training set similarity between each split training set; screening the split training sets based on the training set similarity to obtain a screened training set; randomly extracting training samples from the screened training set; performing sample splitting on the training samples using the sample characteristics of the training samples, and then determining a decision tree for the training set; and establishing an initial analysis model for the training set using the decision tree. The initial analysis model is subjected to hyperparameter optimization to obtain optimized hyperparameters, specifically comprising: dividing the subspace to be optimized of the initial analysis model; uniformly sampling the hyperparameters to be optimized in the subspace to be optimized; analyzing the objective function of the hyperparameters to be optimized; calculating the posterior distribution of the objective function; and calculating the acquisition function value of the hyperparameters to be optimized using the following formula based on the posterior distribution: AC(z)=μ(z)+kσ(z) Where AC(z) represents the acquisition function value, μ(z) represents the mean of the posterior distribution z, σ(z) represents the variance of the posterior distribution, and k represents the tuning parameter; Based on the acquisition function value, identifying a target subspace in the subspace to be optimized; performing hyperparameter optimization on the hyperparameter to be optimized in the target subspace using an objective function corresponding to the target subspace to obtain optimized hyperparameters; Obtaining an optimization analysis model corresponding to the optimization hyperparameter, outputting susceptibility classification results of the training set and the test set using the optimization analysis model, analyzing the classification accuracy of the susceptibility classification results, and determining the susceptibility analysis result of the slope unit based on the classification accuracy; A zoning module is used to calculate the hazard index of the slope unit and divide the hazard zones in the geological disaster area using the hazard index; a hierarchical analysis module, configured to screen vulnerability analysis factors in the slope unit, generate an indicator hierarchy of the vulnerability analysis factors, construct a judgment matrix of the indicator hierarchy, perform a hierarchical analysis on the vulnerability analysis factors using the judgment matrix, and obtain a hierarchical analysis result; a vulnerability analysis module, configured to calculate a vulnerability index of the slope unit using the hierarchical analysis result, and determine a vulnerability analysis result of the geological hazard area based on the vulnerability index; A risk analysis module is used to generate a geological hazard risk analysis result of the slope unit based on the susceptibility analysis result, the hazard zoning and the vulnerability analysis result.

Images