Method for constructing multi-scale dynamic grid model for disaster assessment

Abstract

The present invention discloses a method for constructing a multi-scale dynamic grid model for disaster assessment, which reads geographical (or geological) information data of the sites in the research area for disaster assessment from a memory; automatically divides the research area into several research scenarios; performs a first grid division on the research area by a processor based on the distribution of the geographical (or geological) information data of the research area; on the basis of the first grid, performs a multi-scale second grid division on the research area according to the importance degree of the research scenarios to obtain a preprocessing grid model; dynamically adjusts the preprocessing grid model according to the distance between the research area and the disaster occurrence center point to obtain a disaster research grid; reads multi-source geographical (or geological) information data from the memory; performs gridification on the multi-source geographical (or geological) information data by using the disaster research grid to obtain a gridified data model; and outputs the gridified data model in a visualization manner on a display interface.

Description

Technical Field

[0001] The present invention relates to the field of natural disaster and man-made disaster loss assessment, and particularly relates to a method for constructing a multi-scale dynamic grid model for disaster assessment. Background Art

[0002] Disaster loss assessment is the basis for disaster relief and post-disaster restoration and reconstruction. In the establishment of various other disaster assessment models, the most commonly used loss calculation method is the grid method. The regional grid is the basic calculation unit for risk assessment and an important basis for visualizing the affected area and degree. The grid method based on macroeconomic indicators in the current earthquake disaster assessment system has proven that the grid-based data can better reflect the disaster characteristics of the region.

[0003] However, the research on multi-scale grids and dynamic grid division is not comprehensive and effective enough. In the calculation process, due to the problem that the specific values are relatively vague and the adaptability to the scenario is not comprehensively considered, the calculation efficiency of disaster situation assessment is low, and a large amount of calculation is required to obtain clear data, resulting in waste of computing resources. Therefore, there is still great room for development and research value. Summary of the Invention

[0004] Based on this, the present invention proposes a method for constructing a multi-scale dynamic grid model for disaster assessment. The established multi-scale dynamic grid model is an important basis for realizing functions such as determining the degree of disaster damage, outputting specific loss values, and visualizing results.

[0005] According to one aspect of the present invention, a method for constructing a multi-scale dynamic grid model for disaster assessment executed by an electronic device includes:

[0006] Reading geographical (or geological) information data of the site of the research area from the memory;

[0007] Automatically dividing the above-mentioned research area into a number of research scenarios;

[0008] Performing a first grid division on the above-mentioned research area by a processor based on the distribution of the geographical (or geological) information data of the above-mentioned research area;

[0009] On the basis of the first grid, performing a multi-scale second grid division on the above-mentioned research area according to the importance of the above-mentioned research scenarios to obtain a preprocessing grid model;

[0010] Dynamically adjusting the above-mentioned preprocessing grid model according to the distance between the above-mentioned research area and the disaster occurrence center point to obtain a disaster research grid;

[0011] Reading multi-source geographical (or geological) information data from the above-mentioned memory;

[0012] Gridify the above-mentioned multi-source geographic (or geological) information data using the above-mentioned disaster research grid to obtain a gridified data model;

[0013] Output the above-mentioned gridified data model in a display interface through a visualization method and connect to an external interface for various disaster calculation methods or software calls;

[0014] Among them, the above-mentioned output gridified data model is converted from a two-dimensional image file into a multi-dimensional function file containing grid information and attribute values.

[0015] According to an embodiment of the present invention, wherein the dynamically adjusting the above-mentioned preprocessing grid model according to the distance between the above-mentioned research area and the center point of the disaster occurrence includes:

[0016] Determine the above-mentioned center point of the disaster occurrence;

[0017] Establish a bivariate normal distribution model based on the distances between each point in the above-mentioned research area and the above-mentioned center point;

[0018] Adjust the above-mentioned second grid using the weights of each point in the above-mentioned research area in the bivariate normal distribution model.

[0019] According to an embodiment of the present invention, wherein the gridifying the above-mentioned multi-source geographic (or geological) information data using the above-mentioned disaster research grid includes:

[0020] For point-type vector data, calculate the amount of point data in each cell of the above-mentioned disaster research grid, and use this value or a multiple or logarithm of this value as the point attribute value corresponding to each above-mentioned cell in the above-mentioned disaster research grid;

[0021] For line-type vector data, calculate the length of the line segment in each cell of the above-mentioned disaster research grid, and use this value or a multiple or logarithm of this value as the line attribute value corresponding to each above-mentioned cell in the above-mentioned disaster research grid;

[0022] For surface-type vector data, calculate the overlapping area between each cell of the above-mentioned disaster research grid and the surface data, and use this value or a multiple or logarithm of this value as the surface attribute value corresponding to each above-mentioned cell in the above-mentioned disaster research grid;

[0023] For raster data, sum the values corresponding to all rasters in the above-mentioned disaster research grid, and use this value or a multiple or logarithm of this value as the raster attribute value corresponding to the above-mentioned disaster research grid.

[0024] According to an embodiment of the present invention, wherein the above-mentioned gridified data model includes:

[0025] All the geometric information of all the above units within the second grid and the attribute values corresponding to each of the above units, where the attribute values include the above point attribute values, the above line attribute values, the above surface attribute values, and the above grid attribute values.

[0026] According to an embodiment of the present invention, wherein the automatically dividing the research area into a number of research scenarios includes:

[0027] Invoking the population distribution data, administrative division data, river basin data, economic belt data, building distribution data, important infrastructure data, vegetation data, geological structure distribution data, and geological attribute distribution data stored in the database;

[0028] Based on the above population distribution data, the above administrative division data, the above river basin data, the above economic belt data, the above building distribution data, the above important infrastructure data, the above vegetation data, the above geological structure distribution data, and the above geological attribute distribution data;

[0029] Taking each partition as an evaluation unit for disaster assessment to obtain the above research scenarios.

[0030] According to an embodiment of the present invention, wherein the first grid division of the research area by the processor based on the distribution of the geographical (or geological) information data of the research area is a grid size that can roughly reflect the losses within the research area.

[0031] According to an embodiment of the present invention, wherein the multi-scale second grid division of the research area based on the importance level of the above research scenarios includes:

[0032] Dividing the importance levels according to the importance level of the above research scenarios;

[0033] Selecting grids of multiple different scales to perform the above second grid division on the above research scenarios of different above importance levels.

[0034] It can be seen from the above technical solutions that the fault uncertainty modeling method and training method provided by the present invention have the following beneficial effects:

[0035] The present invention discloses a method for constructing a multi-scale dynamic grid model for disaster assessment. By hierarchically dividing the affected area multiple times, the visualization fineness within the research area is reasonably allocated, and fine analysis is performed on key areas, improving the efficiency of disaster assessment and overcoming problems such as low computational efficiency, low visualization fineness, and low scene adaptability in existing assessment models.

[0036] The present invention discloses a method for constructing a multi-scale dynamic grid model for disaster assessment. The established multi-scale dynamic grid model is an important basis for realizing functions such as determining the degree of disaster damage, outputting specific loss values, and visualizing results. Due to fully considering the importance of the scene and the invisible disaster-affected range, the loss calculation is made more efficient and accurate. At the same time, it provides an important basis for effectively reflecting the loss distribution, thereby facilitating decision-makers to quickly judge the damage range and post-disaster losses, and formulate targeted response strategies, such as the demand for shelters, the distribution of relief resources, and post-disaster reconstruction. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is the construction process of the grid data model according to an embodiment of the present invention Figure Ⅰ ;

[0038] Figure 2 is the construction process of the grid data model according to an embodiment of the present invention Figure Ⅱ ;

[0039] Figure 3 is the construction process of the grid data model according to an embodiment of the present invention Figure Ⅲ ;

[0040] Figure 4 is the process diagram of the first grid division within the research area according to an embodiment of the present invention;

[0041] Figure 5 is the process diagram of the second grid division within the research area according to an embodiment of the present invention;

[0042] Figure 6 is the process diagram of the dynamic adjustment of the second grid within the research area according to an embodiment of the present invention;

[0043] Figure 7 is the schematic diagram of the bivariate normal distribution model within the research area according to an embodiment of the present invention;

[0044] Figure 8 is the schematic diagram of the gridification of multi-source geographic (or geological) information data within the research area according to an embodiment of the present invention;

[0045] Figure 9 is the structural composition diagram of the multi-dimensional function file output by the grid data model of the research area according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] In order to make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the following further describes the present invention in detail with reference to specific embodiments and the accompanying drawings.

[0047] In view of the problem of relatively single scale in mesh generation technology, the current method of using variable meshes can evaluate the conceptual framework by combining deterministic mesh modeling and sub-grid variability models, which can solve the problem of a high degree of sub-grid variability in the mesh model.

[0048] Or a method using the probability density function of sub-grid resolution observation data to represent the non-uniform development density of the grid scale of the urban climate model. This method provides multi-scale and spatial guidance, as well as a method for influencing simulation predictions at various aggregated grid scales, which provides a certain basis for the calculation of multi-scale grid models.

[0049] And a stacked bidirectional gated recurrent unit neural network model to effectively predict the traffic speed of highways at different estimated time intervals. By establishing a multi-scale grid model, it is possible to derive a set of key traffic parameters at different scales in less time to predict the traffic speed of roads at different scales.

[0050] In the process of implementing the present invention, it is found that the research on multi-scale grids and dynamic mesh generation is not comprehensive and effective enough. In the calculation process, due to the problem that the specific numerical values specified are relatively vague and the adaptability to the scenario is not comprehensively considered, the calculation efficiency of disaster situation assessment is low, and a large amount of calculation is required to obtain clear data, resulting in waste of computing resources, which is also the main reason for the difficulty in quickly judging the damage range and post-disaster losses.

[0051] Figure 1 The construction process of the grid data model for the embodiments of the present invention Figure Ⅰ 。

[0052] As Figure 1 shown, according to the general inventive concept of one aspect of the present invention, there is provided a method for constructing a multi-scale dynamic grid model for disaster assessment executed by a computer, including:

[0053] S100: Read geographical (or geological) information data of the research area from the memory;

[0054] S200: Automatically divide the research area into a number of research scenarios;

[0055] S300: Perform a first grid division on the research area by the processor based on the distribution of geographical (or geological) information data of the research area;

[0056] S400: On the basis of the first grid, perform a multi-scale second grid division on the research area according to the importance of the research scenarios to obtain a preprocessing grid model;

[0057] S500: Dynamically adjust the pre - processed grid model according to the distance between the research area and the center point of the disaster occurrence to obtain the disaster research grid;

[0058] S600: Read multi - source geographic (or geological) information data from the memory;

[0059] S700: Grid the multi - source geographic (or geological) information data using the disaster research grid to obtain a gridded data model;

[0060] S800: Output the gridded data model in a visual way on the display interface and connect to the external interface for various disaster calculation methods or software calls;

[0061] Among them, the above - mentioned output gridded data model is obtained by converting a two - dimensional image file into a multi - dimensional function file containing grid information and attribute values.

[0062] Through the embodiments of the present invention, the data on the disaster situation of the site in the research area can be obtained by means such as satellite maps or drones.

[0063] Through the embodiments of the present invention, a method for constructing a multi - scale dynamic gridding model for disaster assessment is provided. The established multi - scale dynamic gridding model is an important basis for realizing functions such as disaster damage degree determination, specific loss value output, and result visualization. Since the importance of the scene and the invisible disaster - affected range are fully considered, the scale of grid division is more in line with reality, thus greatly improving the calculation efficiency.

[0064] By uniformly outputting the clear data calculated by the model as a file of a multi - dimensional function, the loss calculation is made more efficient and accurate, reducing the consumption of computing resources to a certain extent. At the same time, it provides an important basis for effectively reflecting the loss distribution, and further facilitates decision - makers to quickly judge the damage range and post - disaster losses, and formulate targeted response strategies, such as the need for shelters, the distribution of relief resources, and post - disaster reconstruction.

[0065] Figure 2 This is the construction process of the gridded data model of the embodiments of the present invention Figure Ⅱ 。

[0066] Figure 3 This is the construction process of the gridded data model of the embodiments of the present invention Figure Ⅲ 。

[0067] As Figure 2 and Figure 3 shown, according to the embodiments of the present invention, the construction of the gridded data model is specifically as follows:

[0068] Obtain the site disaster situation data of the affected research area. The processor sets the size and parameters of the initial grid according to the basic situation of the research area, and demarcates the first grid as the initial grid;

[0069] The processor demarcates the second grid according to the importance of the research scenarios at each geopolitical location; for important scenarios, high-magnitude encryption is performed, for general scenarios, low-magnitude encryption is performed, and for secondary scenarios, the size remains the same as that of the first grid;

[0070] The processor encrypts and adjusts the initial grid according to the importance of the research scenarios at each geopolitical location and the distance from the center point of the disaster situation;

[0071] For scenarios closer to the center point, high-magnitude encryption is performed, for scenarios at a general distance, low-magnitude encryption is performed, and for scenarios farther from the center point, the size remains the same as that of the first grid, obtaining a "grid map" with multiple scales;

[0072] Subsequently, directly call the multi-source geographical (or geological) information data of the research area in the memory, integrate the multi-source geographical (or geological) information data into the "grid map", and the disaster situation data that fits the natural environment and social environment can be obtained, making the loss calculation more efficient and accurate, and directly improving the calculation efficiency.

[0073] According to the embodiment of the present invention, in S200, automatically dividing the research area into several research scenarios includes:

[0074] Call the population distribution data, administrative division data, river basin data, economic belt data, building distribution data, important infrastructure data, vegetation data, geological structure distribution data, and geological attribute distribution data stored in the database;

[0075] Divide the research area according to the population distribution data, administrative division data, river basin data, economic belt data, building distribution data, important infrastructure data, vegetation data, geological structure distribution data, and geological attribute distribution data;

[0076] Take each division as an evaluation unit to obtain the research scenario.

[0077] According to the embodiment of the present invention, in S200, the population distribution data includes the average population distribution data in each research scenario in the research area;

[0078] The administrative division data includes data such as administrative divisions (provincial, municipal, district-level, etc.) in each research scenario in the research area;

[0079] The river basin data includes information data such as the river basins and the areas where the water systems are distributed in each research scenario in the research area;

[0080] The economic belt data includes the national economic belt data (such as the eastern coastal belt, the central belt, the western belt, etc.) to which each research scenario in the research area belongs.

[0081] The building distribution data includes the distribution data of steel-concrete structures and wood-concrete structures in each research scenario in the research area.

[0082] The important infrastructure data includes the data of reservoirs, nuclear power plants, etc. in each research scenario in the research area.

[0083] The vegetation data includes the vegetation distribution data mainly dominated by forests in each research scenario in the research area.

[0084] The geological structure distribution data includes the distribution data of faults, folds, and strata in each research scenario in the research area.

[0085] The geological attribute distribution data includes the mechanical index data of bedrock and overburden in each research scenario in the research area.

[0086] Figure 4 This is the process diagram of the first grid division within the research area of the embodiment of the present invention.

[0087] As Figure 4 shown, according to the embodiment of the present invention, in S300, the processor performs the first grid division on the research area based on the distribution of geographical (or geological) information data of the research area, and its specific size is determined according to the data refinement degree within the research area, and is set to be able to roughly reflect the grid size of the losses within the research area;

[0088] Figure 5 This is the process diagram of the second grid division within the research area of the embodiment of the present invention.

[0089] According to the embodiment of the present invention, in S400, the multi-scale second grid division of the research area according to the importance degree of the research scenario includes:

[0090] Dividing the importance level according to the importance degree of the research scenario;

[0091] Selecting grids of multiple different scales to perform the second grid division on research scenarios of different importance levels.

[0092] According to the embodiment of the present invention, in S400, the higher the importance level of the research scenario, the higher the encryption multiple of the selected grid.

[0093] As Figure 5 shown, according to the embodiment of the present invention, S400 is specifically:

[0094] Divide the importance levels according to the importance of the research scenarios. In the figure, I-IV indicate the classification of the importance levels of the research scenarios, and I represents the most important scenario.

[0095] Divide the levels according to the importance of each scenario within the disaster-affected area. Select multiple grids of different scales. Based on the initial maximum grid size, the higher the importance of the research scenario area, the higher the encryption multiple of the grid. The specific numerical values can be adjusted at any time according to the situation.

[0096] According to the embodiment of the present invention, in S400, if two types of research scenarios appear in the same maximum-size grid, the grid encryption multiple of the more important scenario will be selected. For example, Figure 3 for scenarios I and IV in , if they appear in the same maximum-size grid at the same time, then the grid encryption method of scenario I will be used for scenario IV.

[0097] Figure 6 It is a process diagram of the dynamic adjustment of the second grid within the research area of the embodiment of the present invention.

[0098] Figure 7 It is a schematic diagram of the bivariate normal distribution model within the research area of the embodiment of the present invention.

[0099] For example, Figure 6 - 7 As shown in , according to the embodiment of the present invention, in S500, the dynamic adjustment of the preprocessed grid model according to the distance between the research area and the center point of the disaster occurrence includes:

[0100] Determine the center point of the disaster occurrence;

[0101] Establish a bivariate normal distribution model based on the distances between each point in the research area and the center point;

[0102] Adjust the second grid using the weights of each point in the research area in the bivariate normal distribution model.

[0103] The closer to the center point of the disaster occurrence, the higher the degree of grid encryption required. For example, Figure 5 for the part closer to the center of the circle within the circle in , the higher the encryption degree. As Figure 6 shown, the specific encryption method is determined with the help of the bivariate normal distribution. The specific numerical values can be adjusted at any time according to the situation. The function of the bivariate normal distribution is as shown in Equation (1):

[0104]

[0105] where x represents the degree of disaster impact in the east-west direction, y represents the degree of disaster impact in the north-south direction, p X,Y (x,y) is the weight, σ X and σ YThe standard deviations are x and y respectively, and ρ is the correlation coefficient. Grid division is performed according to this weight, and the calculation formula of the weight is shown in Equation (2):

[0106] N = κ·int(p X,Y (x,y) α ) (2)

[0107] Among them, N represents the number of encrypted grids, κ is a coefficient customized by the user according to the disaster situation, α is a weight deformation coefficient, which depends on the situation and can usually be set to 1.

[0108] Figure 8 is a schematic diagram of the gridification of multi-source geographic (or geological) information data within the research area of the embodiment of the present invention.

[0109] As Figure 8 shown, according to the embodiment of the present invention, in S700, the gridification of multi-source geographic (or geological) information data using disaster research grids includes:

[0110] For point-type vector data, such as building distribution data, calculate the amount of point data in each unit of the disaster research grid, and use this value or a multiple or logarithm of this value as the point attribute value corresponding to each unit in the disaster research grid. The calculation process is shown in Equation (3):

[0111]

[0112] Among them, V pts is the value corresponding to the grid, k is a coefficient, pts i represents the i-th point data, G is the area corresponding to the grid, and I(i) is an indicator function, that is, it takes 1 when the point is within the grid and 0 otherwise.

[0113] For line-type vector data, such as road distribution data, calculate the length of the line segment in each unit of the disaster research grid, and use this value or a multiple or logarithm of this value as the line attribute value corresponding to each unit in the disaster research grid. The calculation process is shown in Equation (4):

[0114]

[0115] Among them, V line is the value corresponding to the grid, k is a coefficient, L i represents the i-th line data, G is the area corresponding to the grid, and ∩ is the intersection operation.

[0116] For surface-type vector data, such as administrative regions or geological regions, calculate the area of overlap between each cell in the disaster research grid and the surface data, and use this value, or a multiple or logarithm of this value, as the surface attribute value corresponding to each cell in the disaster research grid. The calculation process is shown in Equation (5):

[0117]

[0118] Where V area is the value corresponding to the grid, k is a coefficient, A i represents the area of the i-th surface data, G is the area corresponding to the grid, and ∩ is the intersection operation.

[0119] For raster data, such as forest vegetation, sum the values corresponding to all rasters in the disaster research grid, and use this value, or a multiple or logarithm of this value, as the raster attribute value corresponding to the disaster research grid. The calculation process is shown in Equation (6):

[0120]

[0121] Where V grid is the value corresponding to the grid, k is a coefficient, v x,y represents the raster value corresponding to the (x, y) position, G is the area corresponding to the grid, and I(x, y) is an indicator function, which takes 1 when the point (x, y) is within the grid and 0 otherwise.

[0122] According to an embodiment of the present invention, in S600, the multi-source geographic (or geological) information data includes population data and economic data.

[0123] According to an embodiment of the present invention, in S600, the population data is mainly population density, and the economic data can be further divided into situations such as GDP per unit area, building distribution, and the importance degree of the regional ecological environment, to determine the population quantity and economic total value per unit area.

[0124] Figure 9 This is the structural composition diagram of the multi-dimensional function file output by the grid data model of the research area in the embodiment of the present invention.

[0125] According to an embodiment of the present invention, in S800, the grid data model includes:

[0126] The geometric information of all units within the second grid and the attribute values corresponding to each unit. The attribute values include point attribute values, line attribute values, surface attribute values, and raster attribute values.

[0127] Such as Figure 9As shown, in order to perform subsequent loss calculations, the gridded point, line, and surface vector data and raster data are read and output in a unified file format for easy invocation of subsequent loss calculations. A multi-dimensional function file with an extended word of.mgird can be directly generated. This file contains multi-scale gridded data converted from the input data, that is, the geometric information of all grids and the attribute values corresponding to each grid. Specifically, it includes: the first grid identification code GUID, the set of identification codes of the second grid Children, the grid serial number Index (used to indicate the number of the parent grid of the second grid in the first grid), the division scales X and Y of the second grid, and the attribute numerical values Value corresponding to all grids (including population People, economic GDP, building distribution Building, and forest distribution Forest in this example). All this information is organized according to the.mgrid format.

[0128] In the specific embodiments described above, the purpose, technical solutions, and beneficial effects of the present invention have been further described in detail. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for constructing a multi-scale dynamic grid model for disaster assessment executed by a computer, comprising: Reading geographical or geological information data of the sites in the study area from a memory; Automatically dividing the study area into a number of study scenarios; Based on the distribution of the geographical or geological information data of the study area, performing a first grid division on the study area by a processor to obtain a plurality of first grids; On the basis of the first grids, performing a multi-scale second grid division on the plurality of first grids in the study area according to the importance degree of the study scenarios to obtain a preprocessing grid model, where the preprocessing grid model includes a plurality of second grids; Determining the center point where a disaster occurs in the preprocessing grid model, and establishing a bivariate normal distribution model based on the distances between each point in the preprocessing grid model and the center point; adjusting the second grids with the weights of each point in the preprocessing grid model in the bivariate normal distribution model to obtain a disaster research grid; Among them, x represents the degree of disaster impact in the east-west direction, and y represents the degree of disaster impact in the north-south direction. is the weight, and are the standard deviations of x and y respectively, and ρ is the correlation coefficient. Among them, the weight is determined based on the following formula: Wherein, N represents the number of encrypted grids, κ is a coefficient customized by the user according to the disaster situation, and α is a weight deformation coefficient; Reading multi-source geographical or geological information data from the memory; Using the disaster research grid to grid the multi-source geographical or geological information data to obtain a gridded data model; Outputting the gridded data model in a visual manner on a display interface and connecting to an external interface for calling various disaster calculation methods or software; Wherein, the above output of the gridded data model is to convert a two-dimensional image file into a multi-dimensional function file containing grid information and attribute values.

2. The model construction method according to claim 1, wherein, The using the disaster research grid to grid the multi-source geographical or geological information data includes: For point-type vector data, calculating the amount of point data in each unit in the disaster research grid, and using this value or a multiple or logarithm of this value as the point attribute value corresponding to each unit in the disaster research grid; For line-type vector data, calculating the length of the line segment in each unit in the disaster research grid, and using this value or a multiple or logarithm of this value as the line attribute value corresponding to each unit in the disaster research grid; For surface-type vector data, calculating the area of overlap between each unit in the disaster research grid and the surface data, and using this value or a multiple or logarithm of this value as the surface attribute value corresponding to each unit in the disaster research grid; For raster data, summing up the values corresponding to all rasters in the disaster research grid, and using this value or a multiple or logarithm of this value as the raster attribute value corresponding to the disaster research grid.

3. The model construction method according to claim 1, wherein, The gridded data model includes: Geometric information of all units within the second grid and the corresponding attribute values of each unit, where the attribute values include point attribute values, line attribute values, surface attribute values, and raster attribute values.

4. The model construction method according to claim 1, wherein, Automatically dividing the study area into a number of study scenarios includes: Call the population distribution data, administrative division data, river basin data, economic belt data, building distribution data, important infrastructure data, vegetation data, geological structure distribution data, and geological property distribution data stored in the database; Partition the research area according to the population distribution data, the administrative division data, the river basin data, the economic belt data, the building distribution data, the important infrastructure data, the vegetation data, the geological structure distribution data, and the geological property distribution data; Use each partition as an evaluation unit to obtain the research scenario.

5. The model construction method according to claim 1, wherein The first grid has a grid size that can roughly reflect the losses in the research area.

6. The model construction method according to claim 1, wherein, The multi-scale second grid division of the multiple first grids in the research area according to the importance level of the research scenario includes: Divide the importance levels according to the importance level of the research scenario; Select grids of multiple different scales to perform the second grid division on the multiple first grids in the research scenarios of different importance levels.

Images