Urban expansion simulation method based on multi-scale fusion of patch and grid rules

Abstract

The invention provides an urban expansion simulation method based on multi-scale fusion of patch and grid rules, which relates to the field of geographic information science and technology, and comprises: classifying and numbering original land use data, and obtaining an urban expansion patch distribution map after superposition; processing spatial factors with spatial analysis tools to obtain driving variable grid data, and then performing patch average processing and sampling on the driving variable grid data to obtain sample data and patch average driving variable grid data; training with the sample data and the patch average driving variable grid data to construct patch potential rules; in simulation iterations, matrixing the simulated land use grid data, and constructing grid interaction rules with a two-dimensional convolution algorithm; and calculating the probability of patch state transition and constructing patch area rules based on the patch potential rules, the grid interaction rules and the simulated space constraints, and then determining the urban expansion simulation results, thereby effectively improving the simulation accuracy of urban expansion.

Description

Technical Field

[0001] The present invention relates to the field of geographic information science and technology, and in particular to an urban expansion simulation method based on multi-scale fusion of patches and grid rules. Background Art

[0002] The cellular automaton (CA) model is the mainstream model for simulating urban land use changes in recent years. The cellular automaton is a grid dynamics model with discrete time, space, and state, and local spatial interactions and temporal causal relationships. It has the ability to simulate the spatiotemporal evolution of complex systems. Cellular automata can describe the evolution of complex systems through simple rules, so they have powerful spatiotemporal process modeling and simulation capabilities and are widely used in geographic simulation research. Since the acquisition, processing, and calculation of raster data are simpler than vector data, most of the current urban expansion CAs are raster data models, but they do not have the advantage of vector data models that can more realistically reflect the actual urban expansion process. Therefore, existing studies mostly adopt patch strategies in raster data models to describe the patch change process in vector data models. However, these studies have major limitations and it is difficult to achieve the fusion of patches and grid rules of different scales. Summary of the invention

[0003] The purpose of the present invention is to solve the problem that the current urban expansion CA is difficult to achieve the fusion of patches and grid rules of different scales. The present invention provides a method for simulating urban expansion based on multi-scale fusion of patches and grid rules.

[0004] The technical solution of the embodiment of the present application is implemented as follows:

[0005] The first aspect of the embodiment of the present application provides a method for simulating urban expansion based on multi-scale fusion of patches and grid rules, including:

[0006] Classifying the original land use raster data based on the land use type, identifying and numbering the patch structure of the original land use, and superimposing them to obtain an urban expansion patch distribution map;

[0007] Using spatial analysis tools to process spatial factors that affect urban expansion, obtaining driving variable raster data; the spatial factors include roads, stations, water areas, altitudes and slopes, and the grid size of the driving variable raster data is consistent with the original land use raster data;

[0008] The driving variable raster data is subjected to patch averaging and sampling to obtain sample data at the patch scale, and the random forest model is trained with the sample data to construct patch potential rules for urban expansion simulation; in each simulation iteration, the simulated land use raster data is binarized, and the neighborhood interaction effect is calculated using a two-dimensional convolution algorithm to construct grid interaction rules for urban expansion simulation; the simulated spatial constraints, patch potential and grid interaction are integrated to calculate the state transition probability of each patch, and the patch area rules for urban expansion simulation are constructed according to the state transition probability;

[0009] The expansion simulation result of the urban land is determined based on the patch potential rule, the grid interaction rule and the patch area rule.

[0010] Optionally, the original land use raster data is classified based on the land use type, the patch structure of the original land use is identified and numbered, and the patch distribution map of urban expansion is obtained by superimposing, including:

[0011] Reclassifying the original land use raster data to obtain classified data; the classified data includes urban land use data, water area land use data and other land use data;

[0012] Numbering the classified data according to the spatial position and category attribute differences of the patches in the original land use raster data to obtain the patch numbering layer;

[0013] The patch number layer and the classification data are overlaid and analyzed to obtain a superimposed distribution map of the urban expansion patches.

[0014] Optionally, the spatial analysis tool is used to process the spatial factors affecting urban expansion to obtain driving variable raster data, including:

[0015] Based on the difference in category attributes, the spatial factors affecting urban expansion are classified to obtain first-category factors and second-category factors; the first-category factors include roads, stations and water areas; the second-category factors include slope and altitude;

[0016] The first type of factors are calculated using Euclidean distance to obtain first type of driving variable grid data, and the second type of factors are processed by resampling to obtain second type of driving variable grid data.

[0017] Optionally, the step of performing patch averaging processing and sampling on the driving variable grid data to obtain patch-scale sample data, training a random forest model with the sample data, and constructing patch potential rules for urban expansion simulation includes:

[0018] Get the average value of each driving variable grid data in the patch to obtain the average driving variable grid data of the patch;

[0019] Combined with the urban expansion patch distribution map and the patch average driving variable grid data, sampling is performed in a manner that the urban expansion patch label is 1 and the other patch labels are 0, and the patch average driving variable grid data value at the sample patch is obtained to obtain the sample data at the patch scale;

[0020] The random forest model is trained with the sample data at the patch scale, and the patch-averaged driving variable grid data is input into the trained random forest model to obtain the development potential values ​​of all patches, and the development potential values ​​are used as the patch potential rules for constructing urban expansion simulation.

[0021] Optionally, in each simulation iteration, the simulated land use grid data is binarized, a two-dimensional convolution algorithm is used to calculate the neighborhood interaction effect, and a grid interaction rule for urban expansion simulation is constructed, including:

[0022] The simulated land use raster data obtained in each iteration is matrixed and binarized according to the raster attribute values, with the city grid attribute value set to 1 and other grid attribute values ​​set to 0 to obtain a binary land use matrix;

[0023] The convolution kernel matrix required for the two-dimensional convolution algorithm operation is set according to the adopted extended Moore-type neighborhood size, and the convolution kernel matrix is ​​binarized according to the neighborhood cell category, the central matrix element value is set to 0, and the other matrix elements are set to 1, so as to obtain the binarized target convolution kernel matrix;

[0024] A two-dimensional convolution operation is performed on the binarized land use matrix to determine the neighborhood effect value of each grid, thereby obtaining a grid neighborhood effect matrix, and a grid interaction rule for urban expansion simulation is constructed based on the grid neighborhood effect matrix.

[0025] Optionally, the fusion simulation of spatial constraints, patch potential and grid interaction, calculating the state transition probability of each patch, and constructing a patch area rule for urban expansion simulation according to the state transition probability, includes:

[0026] Taking the existing urban patches and water patches as the spatial constraints, the spatial constraints of urban expansion simulation are set and realized through the conditional judgment function, that is, Z i = con ( C i = City OR waters?), where C i Characterizing plaques i Category, con(·) is a conditional judgment function. When the judgment content is true, the return value is 0, that is, Z i = 0, when the judgment content is false, the return value is 1, that is Z i = 1;

[0027] In each iteration, the grid neighborhood effect matrix is ​​averaged by patch, and the scale is increased from grid to patch to obtain the average grid neighborhood effect value of the patch;

[0028] According to the formula PatchP i = PatchS i × PM_GridN i × Z i Calculate the state transition probability of each patch, and set the patch area rule based on the state transition probability as follows: select PatchP For the largest patch, update its attribute from non-urban patch to urban patch, record the patch area, and stop the iteration when the area of ​​the updated patch in all iterations reaches the set target.

[0029] Optionally, the determining of the expansion simulation result of urban land based on the patch potential rule, the grid interaction rule and the patch area rule includes:

[0030] According to the urban land area in the original land use grid data of the target year or the predicted urban land area, the termination condition of the urban expansion simulation is set; the termination condition includes that the cumulative area of ​​the newly added urban patches reaches a preset threshold;

[0031] The area of ​​the updated patch is recorded and accumulated in each iteration, and the iteration is stopped when the termination condition is met, and the simulation result is output.

[0032] The second aspect of the embodiment of the present application provides a device for simulating urban expansion based on multi-scale fusion of patch and grid rules, including: a classification module, a processing module, a training module and a determination module, wherein:

[0033] The classification module is configured to classify the original land use raster data based on the land use type, identify the patch structure of the original land use and number it, and superimpose it to obtain the urban expansion patch distribution map;

[0034] The processing module is configured to process the spatial factors affecting urban expansion using spatial analysis tools to obtain driving variable raster data; the spatial factors include roads, stations, water areas, altitudes and slopes, and the grid size of the driving variable raster data is consistent with the original land use raster data;

[0035] The training module is configured to perform patch averaging processing and sampling on the driving variable grid data to obtain sample data at the patch scale, train the random forest model with the sample data, and construct the patch potential rules for urban expansion simulation; in each simulation iteration, binarize the simulated land use grid data, use a two-dimensional convolution algorithm to calculate the neighborhood interaction effect, and construct the grid interaction rules for urban expansion simulation; integrate the simulated spatial constraints, patch potential and grid interaction, calculate the state transition probability of each patch, and construct the patch area rules for urban expansion simulation according to the state transition probability;

[0036] The determination module is configured to determine the expansion simulation result of urban land based on the patch potential rule, the grid interaction rule and the patch area rule.

[0037] A third aspect of an embodiment of the present application provides an electronic device, including a processor and a memory; the memory stores a computer program, wherein the computer program, when executed by the processor, implements the method for simulating urban expansion based on multi-scale fusion of patch and grid rules as described in the first aspect.

[0038] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in the first aspect.

[0039] Compared with the prior art, the technical solution provided by this application has the following beneficial effects:

[0040] The present invention provides a method for simulating urban expansion based on multi-scale fusion of patch and grid rules. The method comprises the following steps: classifying and numbering original land use data, and obtaining a patch distribution map of urban expansion after superposition; processing spatial factors by using spatial analysis tools to obtain driving variable grid data, and then performing patch average processing and sampling on the driving variable grid data to obtain sample data and driving variable grid data of patch average; training is performed by using the sample data and driving variable grid data of patch average to construct patch potential rules; in simulation iterations, the simulated land use grid data is matrixed, and a two-dimensional convolution algorithm is used to construct a grid interaction rule; based on the patch potential rule, the grid interaction rule and the simulated space constraint, the probability of patch state transition is calculated and a patch area rule is constructed, and then the simulation result of urban expansion is determined, thereby effectively taking into account the patch and grid characteristics of urban expansion and improving the simulation accuracy of urban expansion. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A schematic flow chart of a method for simulating urban expansion based on multi-scale fusion of patches and grid rules provided in an embodiment of the present application;

[0042] Figure 2 A schematic diagram of land use reclassification results provided in an embodiment of the present application;

[0043] Figure 3 A schematic diagram of driving variable grid data provided by an embodiment of the present application;

[0044] Figure 4 A schematic diagram of obtaining a plaque driving variable value provided in an embodiment of the present application;

[0045] Figure 5 A schematic diagram of the plaque potential rule provided in the embodiments of the present application;

[0046] Figure 6 A schematic diagram of the calculation principle of the matrix two-dimensional convolution operation provided in the embodiment of the present application;

[0047] Figure 7 A schematic diagram showing the comparison between the simulation results of applying different rules provided in the embodiment of the present application and the actual observed urban expansion;

[0048] Figure 8 A schematic diagram of the structure of a device for simulating urban expansion based on multi-scale fusion of patches and grid rules provided in an embodiment of the present application;

[0049] Fig. 9 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0050] Below, embodiments of the present application will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present application. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present application. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present application.

[0051] The terms used herein are only for describing specific embodiments and are not intended to limit the present application. The terms "include", "comprising", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0052] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification, and should not be interpreted in an idealized or overly rigid manner.

[0053] Some block diagrams and / or flow charts are shown in the accompanying drawings. It should be understood that some blocks or combinations thereof in the block diagrams and / or flow charts may be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that these instructions, when executed by the processor, may create a device for implementing the functions / operations described in these block diagrams and / or flow charts.

[0054] In some embodiments, see Figure 1 , Figure 1 A flow chart of a method for simulating urban expansion based on multi-scale fusion of plaques and grid rules provided in an embodiment of the present application; The method for simulating urban expansion based on multi-scale fusion of plaques and grid rules provided in an embodiment of the present application includes:

[0055] S110, classifying the original land use raster data based on the land use type, identifying and numbering the patch structure of the original land use, and overlaying to obtain an urban expansion patch distribution map.

[0056] For example, see Figure 2 , Figure 2 Schematic diagram of land use reclassification results provided in the embodiment of the present application; here, the original land use raster data is first classified according to the land use type, and each type of land is represented by different colors, patterns or symbols. During the drawing process, the legend, scale, annotation and other elements can be adjusted to make the map clearer and easier to read. In this process, it is necessary to maintain consistency in the representation of various types of land, such as the selection and use of colors, patterns or symbols.

[0057] In some embodiments, S110, classifying the original land use grid data based on the land use type, identifying the patch structure of the original land use and numbering it, and superimposing it to obtain an urban expansion patch distribution map, including:

[0058] Reclassify the original land use raster data to obtain classified data; the classified data includes urban land use data, water area land use data and other land use data;

[0059] The classified data are numbered according to the spatial location and category attribute differences of the patches in the original land use raster data, and the patch number layer is obtained;

[0060] The patch number layer and classification data are overlaid and analyzed to obtain the overlaid urban expansion patch distribution map.

[0061] In this embodiment, the classified land use raster data is patched, and the patches are numbered according to their spatial positions and category attribute differences to obtain a patch number layer, which is overlaid with the classified land use raster and the results of the overlay analysis are saved as a new patch layer, i.e., an urban expansion patch distribution map.

[0062] S120, using spatial analysis tools to process spatial factors that affect urban expansion to obtain driving variable raster data; the spatial factors include roads, stations, water areas, altitudes and slopes, and the grid size of the driving variable raster data is consistent with the original land use raster data.

[0063] In some embodiments, see Figure 3 , Figure 3 Schematic diagram of driving variable grid data provided in an embodiment of the present application; S120, using spatial analysis tools to process spatial factors affecting urban expansion to obtain driving variable grid data, including:

[0064] Based on the difference in category attributes, the spatial factors that affect urban expansion are classified into the first and second category factors; the first category factors include roads, stations and water areas; the second category factors include slope and altitude;

[0065] The first type of factor is calculated using Euclidean distance to obtain the first type of driving variable raster data, and the second type of factor is processed by resampling to obtain the second type of driving variable raster data.

[0066] Here, for vector spatial factor data such as roads, stations and waters, the Euclidean distance tool is used to calculate the driving variables, keeping the grid size of the driving variable raster data consistent with the land use raster data; for other raster data such as slope and altitude, the driving variable data is obtained by resampling, keeping the grid size of the driving variable raster data consistent with the land use raster data.

[0067] S130, perform patch averaging processing and sampling on the driving variable raster data to obtain sample data at the patch scale, use the sample data to train the random forest model, and construct patch potential rules for urban expansion simulation; in each simulation iteration, binarize the simulated land use raster data, use a two-dimensional convolution algorithm to calculate the neighborhood interaction effect, and construct grid interaction rules for urban expansion simulation; integrate the simulated spatial constraints, patch potential, and grid interaction, calculate the state transition probability of each patch, and construct the patch area rules for urban expansion simulation based on the state transition probability.

[0068] In some embodiments, S130, performing patch averaging processing and sampling on the driving variable grid data to obtain patch-scale sample data, training a random forest model with the sample data, and constructing patch potential rules for urban expansion simulation, including:

[0069] Get the average value of each driving variable grid data in the patch to obtain the average driving variable grid data of the patch;

[0070] Combined with the urban expansion patch distribution map and the average driving variable grid data of the patches, sampling is performed in the way that the urban expansion patch label is 1 and the other patch labels are 0, and the average driving variable grid data value of the sample patch is obtained to obtain the sample data of the patch scale;

[0071] The random forest model is trained with patch-scale sample data, and the patch-average driving variable raster data is input into the trained random forest model to obtain the development potential values ​​of all patches. The development potential values ​​are used as the patch potential rules for constructing urban expansion simulation.

[0072] In this example, see Figure 4 and Figure 5 , Figure 4 A schematic diagram of obtaining a plaque driving variable value provided in an embodiment of the present application; Figure 5 A schematic diagram of the patch potential rule provided in the embodiment of the present application; according to the patch number layer, the average value of each driving variable grid data in the patch is counted in turn to obtain the average driving variable grid data of the patch, 2000 samples are randomly selected from the expansion patch and the label is set to 1, 2000 samples are selected from other land patches and water patches and the label is set to 0, the average driving variable grid data value of the patch at these 4000 sample patches is extracted as a feature, and the sample data is constructed; the random forest model is trained using the above sample data, and then the average driving variable grid data set of all patches is input into the trained model to obtain the development potential value of all patches PatchS , as a patch potential rule for expansion simulation.

[0073] In some embodiments, see Figure 6 , Figure 6 Schematic diagram of the calculation principle of the matrix two-dimensional convolution operation provided in the embodiment of the present application; S130, in each simulation iteration, the simulated land use raster data is binarized, and the neighborhood interaction effect is calculated using a two-dimensional convolution algorithm to construct a grid interaction rule for urban expansion simulation, including:

[0074] The simulated land use raster data obtained in each iteration is matrixed and binarized according to the raster attribute values, with the city grid attribute value set to 1 and other grid attribute values ​​set to 0 to obtain a binary land use matrix;

[0075] The convolution kernel matrix required for the two-dimensional convolution algorithm operation is set according to the adopted extended Moore-type neighborhood size, and the convolution kernel matrix is ​​binarized according to the neighborhood cell category, the central matrix element value is set to 0, and the other matrix elements are set to 1, so as to obtain the binarized target convolution kernel matrix;

[0076] A two-dimensional convolution operation is performed on the binary land use matrix to determine the neighborhood effect value of each grid and obtain the grid neighborhood effect matrix. Based on the grid neighborhood effect matrix, the grid interaction rules for urban expansion simulation are constructed.

[0077] In some embodiments, S130, integrating the simulated spatial constraints, patch potential, and grid interaction, calculating the state transition probability of each patch, and constructing a patch area rule for urban expansion simulation according to the state transition probability, includes:

[0078] Taking the existing urban patches and water patches as the spatial constraints, the spatial constraints of urban expansion simulation are set and realized through the conditional judgment function, that is, Z i = con ( C i =City OR waters?), where C i Characterizing plaques i Category, con (·) is a conditional judgment function. When the judgment content is true, the return value is 0, that is, Z i =0, when the judgment content is false, the return value is 1, that is Z i =1;

[0079] In each iteration, the grid neighborhood effect matrix is ​​averaged by patch, and the scale is raised from grid to patch to obtain the average grid neighborhood effect value of the patch;

[0080] According to the formula PatchP i = PatchS i × PM_GridN i × Z i Calculate the state transition probability of each patch, and set the patch area rule based on the state transition probability: select PatchP For the largest patch, update its attribute from non-urban patch to urban patch, record the patch area, and stop the iteration when the area of ​​the updated patch in all iterations reaches the set target.

[0081] S140, determining the expansion simulation result of the urban land based on the patch potential rule, the grid interaction rule and the patch area rule.

[0082] Here, the expansion simulation results are determined by iterative calculation. In each iteration, the patch and grid rules of different scales are integrated according to the method in the previous step to construct the urban expansion CA simulation rules. After the set termination conditions are met through iterative simulation, the final urban expansion simulation results are obtained.

[0083] In some embodiments, S140, determining the expansion simulation result of urban land based on the patch potential rule, the grid interaction rule and the patch area rule includes:

[0084] According to the urban land area in the original land use grid data of the target year or the predicted urban land area, the termination conditions of the urban expansion simulation are set; the termination conditions include that the cumulative area of ​​the newly added urban patches reaches the preset threshold;

[0085] In each iteration, the area of ​​the updated patch is recorded and accumulated. When the termination condition is met, the iteration is stopped and the simulation results are output.

[0086] In some optional embodiments, the simulation performance of the commonly used grid-based urban expansion CA (Grid-CA), the patch-based urban expansion CA (Patch-CA) and the urban expansion simulation CA (Patch-Grid Multi-scale Fusion CA, PGMF-CA) based on multi-scale fusion of patch and grid rules constructed in this application is compared to demonstrate the advantages of the proposed method. The above CA was used in the embodiment for simulation experiments, and the experiments were all carried out on a computer. The operating system was Windows 11 64-bit Professional Edition, the CPU processor was ten cores and twenty threads, the model was Intel i9-10900T CPU 1.90GHz, and the memory was 64GB. The required software environment is ArcGIS 10.8 and Matlab R2020. Data processing, analysis and mapping expression are performed in ArcGIS 10.8, and program calculation and iterative simulation are performed in Matlab R2020. Under the support of the same embodiment and the same equipment, the urban expansion simulation accuracy of the above CA can be compared to evaluate its simulation performance. The currently more commonly used urban expansion simulation accuracy test method is the FoM (Figure of Merit) indicator, and the calculation formula is:

[0087] FoM = Hits / ( Hits + Misses + FalseAlarms )

[0088] in, Hitsis the actual and simulated area of ​​urban expansion, Misses is the area that is actually urban expansion but simulated as non-urban expansion, FalseAlarms It is the area that is actually not urban expansion but is simulated as urban expansion. The FoM indicator focuses on the part where the land use type has changed, which can better evaluate the simulation accuracy and is widely used in geographical research. The experimental data of the embodiment mainly include three-phase land use raster data in 2000, 2010 and 2020, traffic network vector data that characterize the driving force of urban expansion, point of interest data and topographic data, etc. In this embodiment, the CA model is calibrated with the urban expansion process from 2000 to 2010 to evaluate its calibration accuracy; the calibrated CA model is applied to simulate the urban expansion process from 2010 to 2020 with the land use data in 2010 as the initial stage, and its verification accuracy is evaluated. Figure 7 The schematic diagram of the comparison between the simulation results of the above CA in the calibration and verification phase and the actual observed urban expansion is shown. Table 1 shows the calibration and verification simulation accuracy of Grid-CA, Patch-CA and PGMF-CA constructed by the method of the present invention.

[0089] Table 1 Calibration and verification simulation accuracy of Grid-CA, Patch-CA and PGMF-CA

[0090]

[0091] Compared with the commonly used Grid-CA and Patch-CA, the simulation results of PGMF-CA constructed by the method of the present invention are more similar to actual observations, and the simulation accuracy is significantly improved, and this improvement is very significant in both the calibration and verification simulation stages. This comparison result fully demonstrates that the simulation performance of PGMF-CA constructed by the method of the present invention is better and the simulation results are more accurate.

[0092] The embodiment of the present invention classifies and numbers the original land use data, and obtains the urban expansion patch distribution map after superposition; uses the spatial analysis tool to process the spatial factors to obtain the driving variable grid data, and then performs patch average processing and sampling on the driving variable grid data to obtain the sample data and the driving variable grid data of the patch average; uses the sample data and the driving variable grid data of the patch average to train and construct the patch potential rule; in the simulation iteration, the simulated land use grid data is matrixed, and the grid interaction rule is constructed by using the two-dimensional convolution algorithm; based on the patch potential rule, the grid interaction rule and the simulation space constraint, the patch state transition probability is calculated and the patch area rule is constructed, and then the urban expansion simulation result is determined, which effectively takes into account the patch and grid characteristics of the urban expansion and improves the simulation accuracy of the urban expansion.

[0093] In some embodiments, see Figure 8 , Figure 8 A schematic diagram of the structure of a device for simulating urban expansion based on multi-scale fusion of plaques and grid rules provided in an embodiment of the present application; The present application embodiment provides a device 800 for simulating urban expansion based on multi-scale fusion of plaques and grid rules, including: a classification module 810, a processing module 820, a training module 830 and a determination module 840, wherein:

[0094] The classification module 810 is configured to classify the original land use raster data based on the land use type, identify the original land use patch structure and number it, and overlay it to obtain the urban expansion patch distribution map;

[0095] The processing module 820 is configured to process the spatial factors affecting urban expansion using spatial analysis tools to obtain driving variable raster data; the spatial factors include roads, stations, water areas, altitudes and slopes, and the grid size of the driving variable raster data is consistent with the original land use raster data;

[0096] The training module 830 is configured to perform patch averaging processing and sampling on the driving variable raster data to obtain sample data at the patch scale, train the random forest model with the sample data, and construct the patch potential rules for urban expansion simulation; in each simulation iteration, binarize the simulated land use raster data, use a two-dimensional convolution algorithm to calculate the neighborhood interaction effect, and construct the grid interaction rules for urban expansion simulation; integrate the simulated spatial constraints, patch potential and grid interaction, calculate the state transition probability of each patch, and construct the patch area rules for urban expansion simulation based on the state transition probability;

[0097] The determination module 840 is configured to determine the expansion simulation result of the urban land based on the patch potential rule, the grid interaction rule and the patch area rule.

[0098] In some embodiments, the classification module 810 is specifically configured to:

[0099] Reclassify the original land use raster data to obtain classified data; the classified data includes urban land use data, water area land use data and other land use data;

[0100] The classified data are numbered according to the spatial location and category attribute differences of the patches in the original land use raster data, and the patch number layer is obtained;

[0101] The patch number layer and classification data are overlaid and analyzed to obtain the overlaid urban expansion patch distribution map.

[0102] In some embodiments, the processing module 820 is specifically configured as follows:

[0103] Based on the difference in category attributes, the spatial factors affecting urban expansion are classified into the first and second category factors; the first category factors include roads, stations and water areas; the second category factors include slope and altitude;

[0104] The first type of factor is calculated using Euclidean distance to obtain the first type of driving variable raster data, and the second type of factor is processed by resampling to obtain the second type of driving variable raster data.

[0105] In some embodiments, the training module 830 is specifically configured as follows:

[0106] Get the average value of each driving variable grid data in the patch to obtain the average driving variable grid data of the patch;

[0107] Combined with the urban expansion patch distribution map and the average driving variable grid data of the patches, sampling is performed in the way that the urban expansion patch label is 1 and the other patch labels are 0, and the average driving variable grid data value of the sample patch is obtained to obtain the sample data at the patch scale;

[0108] The random forest model is trained with patch-scale sample data, and the patch-averaged driving variable raster data is input into the trained random forest model to obtain the development potential values ​​of all patches. The development potential values ​​are used as the patch potential rules for constructing urban expansion simulation.

[0109] In some embodiments, the training module 830 is specifically configured as follows:

[0110] The simulated land use raster data obtained in each iteration is matrixed and binarized according to the raster attribute values, with the city grid attribute value set to 1 and other grid attribute values ​​set to 0 to obtain a binary land use matrix;

[0111] The convolution kernel matrix required for the two-dimensional convolution algorithm operation is set according to the adopted extended Moore-type neighborhood size, and the convolution kernel matrix is ​​binarized according to the neighborhood cell category, the central matrix element value is set to 0, and the other matrix elements are set to 1, so as to obtain the binarized target convolution kernel matrix;

[0112] A two-dimensional convolution operation is performed on the binary land use matrix to determine the neighborhood effect value of each grid and obtain the grid neighborhood effect matrix. Based on the grid neighborhood effect matrix, the grid interaction rules for urban expansion simulation are constructed.

[0113] In some embodiments, the training module 830 is specifically configured as follows:

[0114] Taking the existing urban patches and water patches as the spatial constraints, the spatial constraints of urban expansion simulation are set and realized through the conditional judgment function, that is, Zi = con ( C i =City OR waters?), where C i Characterizing plaques i Category, con (·) is a conditional judgment function. When the judgment content is true, the return value is 0, that is, Z i = 0, when the judgment content is false, the return value is 1, that is Z i =1;

[0115] In each iteration, the grid neighborhood effect matrix is ​​averaged by patch, and the scale is raised from grid to patch to obtain the average grid neighborhood effect value of the patch;

[0116] According to the formula PatchP i = PatchS i × PM_GridN i × Z i Calculate the state transition probability of each patch, and set the patch area rule based on the state transition probability: select PatchP For the largest patch, update its attribute from non-urban patch to urban patch, record the patch area, and stop the iteration when the area of ​​the updated patch in all iterations reaches the set target.

[0117] In some embodiments, the determination module 840 is specifically configured to:

[0118] According to the urban land area in the original land use grid data of the target year or the predicted urban land area, the termination conditions of the urban expansion simulation are set; the termination conditions include that the cumulative area of ​​the newly added urban patches reaches the preset threshold;

[0119] In each iteration, the area of ​​the updated patch is recorded and accumulated. When the termination condition is met, the iteration is stopped and the simulation results are output.

[0120] The device for simulating urban expansion based on multi-scale fusion of patches and grid rules provided in the embodiment of the present application can realize the various processes in the embodiments corresponding to the method for simulating urban expansion based on multi-scale fusion of patches and grid rules mentioned above, and will not be described again here to avoid repetition.

[0121] It should be noted that the device for simulating urban expansion based on multi-scale fusion of patches and grid rules provided in the embodiment of the present application and the method for simulating urban expansion based on multi-scale fusion of patches and grid rules provided in the embodiment of the present application are based on the same application concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned method for simulating urban expansion based on multi-scale fusion of patches and grid rules, and the repeated parts will not be repeated.

[0122] In some embodiments, see Fig. 9 , Fig. 9 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. An electronic device 900 provided in an embodiment of the present application includes a processor 910 and a memory 920; the memory 920 stores a computer program, wherein the computer program implements the above-mentioned method of urban expansion simulation based on multi-scale fusion of patch and grid rules when executed by the processor.

[0123] Specifically, the processor 910 may include, for example, a general-purpose microprocessor, an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 910 may also include an onboard memory for cache purposes. The processor 910 may be a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present application.

[0124] The memory 920 may be any medium capable of containing, storing, conveying, propagating or transmitting instructions. For example, the memory 920 may include, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared or semiconductor system, device, component or propagation medium. Specific examples of the memory 920 include: a magnetic storage device, such as a magnetic tape or a hard disk (HDD); an optical storage device, such as a compact disk (CD-ROM); a random access memory (RAM) or flash memory; and / or a wired / wireless communication link.

[0125] The present application also provides a computer-readable medium on which a computer program is stored. When the program is executed by a processor, the method for simulating urban expansion based on multi-scale fusion of patch and grid rules is implemented. The computer-readable medium may be included in the device / apparatus / system described in the above embodiment; or it may exist independently without being assembled into the device / apparatus / system. The above computer-readable medium carries one or more programs. When the above one or more programs are executed, the method according to the embodiment of the present application is implemented.

[0126] According to an embodiment of the present application, a computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a computer-readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, optical cable, radio frequency signal, etc., or any suitable combination of the above.

[0127] Those skilled in the art will appreciate that the features described in the various embodiments and / or claims of the present application may be combined and / or combined in a variety of ways, even if such combinations or combinations are not explicitly described in the present application. In particular, without departing from the spirit and teachings of the present application, the features described in the various embodiments and / or claims of the present application may be combined and / or combined in a variety of ways. All of these combinations and / or combinations fall within the scope of the present application. Therefore, the scope of the present application should not be limited to the above-described embodiments, but should be determined not only by the appended claims, but also by the equivalents of the appended claims.

Claims

1. A method for simulating urban expansion based on multi-scale fusion of patch and grid rules, characterized in that: include: Classifying the original land use raster data based on the land use type, identifying and numbering the patch structure of the original land use, and superimposing them to obtain an urban expansion patch distribution map; Using spatial analysis tools to process spatial factors that affect urban expansion, obtaining driving variable raster data; the spatial factors include roads, stations, water areas, altitudes and slopes, and the grid size of the driving variable raster data is consistent with the original land use raster data; The driving variable raster data is subjected to patch averaging and sampling to obtain sample data at the patch scale, and the random forest model is trained with the sample data to construct patch potential rules for urban expansion simulation; in each simulation iteration, the simulated land use raster data is binarized, and the neighborhood interaction effect is calculated using a two-dimensional convolution algorithm to construct grid interaction rules for urban expansion simulation; the simulated spatial constraints, patch potential and grid interaction are integrated to calculate the state transition probability of each patch, and the patch area rules for urban expansion simulation are constructed according to the state transition probability; Obtain patch potential rules for constructing urban expansion simulations, including: Get the average value of each driving variable grid data in the patch to obtain the average driving variable grid data of the patch; Combined with the urban expansion patch distribution map and the patch average driving variable grid data, sampling is performed in a manner that the urban expansion patch label is 1 and the other patch labels are 0, and the patch average driving variable grid data value at the sample patch is obtained to obtain the sample data at the patch scale; Training a random forest model with the sample data at the patch scale, inputting the patch-averaged driving variable grid data into the trained random forest model to obtain development potential values ​​of all patches, and using the development potential values ​​as patch potential rules for constructing urban expansion simulation; The expansion simulation result of the urban land is determined based on the patch potential rule, the grid interaction rule and the patch area rule.

2. The method for simulating urban expansion based on multi-scale fusion of patch and grid rules according to claim 1, characterized in that: The original land use raster data is classified based on the land use type, the patch structure of the original land use is identified and numbered, and the urban expansion patch distribution map is obtained by superimposing, including: Reclassifying the original land use raster data to obtain classified data; the classified data includes urban land use data, water area land use data and other land use data; Numbering the classified data according to the spatial position and category attribute differences of the patches in the original land use raster data to obtain the patch numbering layer; The patch number layer and the classification data are overlaid and analyzed to obtain a superimposed distribution map of the urban expansion patches.

3. The method for simulating urban expansion based on multi-scale fusion of patch and grid rules according to claim 1, characterized in that: The spatial analysis tool is used to process the spatial factors that affect urban expansion to obtain driving variable raster data, including: Based on the difference in category attributes, the spatial factors affecting urban expansion are classified to obtain first-category factors and second-category factors; the first-category factors include roads, stations and water areas; the second-category factors include slope and altitude; The first type of factors are calculated using Euclidean distance to obtain first type of driving variable grid data, and the second type of factors are processed by resampling to obtain second type of driving variable grid data.

4. The method for simulating urban expansion based on multi-scale fusion of patch and grid rules according to claim 1, characterized in that: In each simulation iteration, the simulated land use grid data is binarized, and the neighborhood interaction effect is calculated using a two-dimensional convolution algorithm to construct grid interaction rules for urban expansion simulation, including: The simulated land use raster data obtained in each iteration is matrixed and binarized according to the raster attribute values, with the city grid attribute value set to 1 and other grid attribute values ​​set to 0 to obtain a binary land use matrix; The convolution kernel matrix required for the two-dimensional convolution algorithm operation is set according to the adopted extended Moore-type neighborhood size, and the convolution kernel matrix is ​​binarized according to the neighborhood cell category, the central matrix element value is set to 0, and the other matrix elements are set to 1, so as to obtain the binarized target convolution kernel matrix; A two-dimensional convolution operation is performed on the binarized land use matrix to determine the neighborhood effect value of each grid, thereby obtaining a grid neighborhood effect matrix, and a grid interaction rule for urban expansion simulation is constructed based on the grid neighborhood effect matrix.

5. The method for simulating urban expansion based on multi-scale fusion of patch and grid rules according to claim 4, characterized in that: The fusion simulation of spatial constraints, patch potential and grid interaction, calculating the state transition probability of each patch, and constructing the patch area rule for urban expansion simulation based on the state transition probability, include: Taking the existing urban patches and water patches as the spatial constraints, the spatial constraints of urban expansion simulation are set and realized through the conditional judgment function, that is, Z i = con ( C i = City OR waters?), where C i Characterizing plaques i Category, con (·) is a conditional judgment function. When the judgment content is true, the return value is 0, that is, Z i = 0, when the judgment content is false, the return value is 1, that is Z i = 1; In each iteration, the grid neighborhood effect matrix is ​​averaged by patch, and the scale is increased from grid to patch to obtain the average grid neighborhood effect value of the patch; According to the formula PatchP i = PatchS i × PM_GridN i × Z i Calculate the state transition probability of each patch, and set the patch area rule based on the state transition probability as follows: select PatchP For the largest patch, update its attribute from non-urban patch to urban patch, record the patch area, and stop the iteration when the area of ​​the updated patch in all iterations reaches the set target.

6. The method for simulating urban expansion based on multi-scale fusion of patch and grid rules according to claim 1, characterized in that: The determining of the expansion simulation result of urban land based on the patch potential rule, the grid interaction rule and the patch area rule includes: According to the urban land area in the original land use grid data of the target year or the predicted urban land area, the termination condition of the urban expansion simulation is set; the termination condition includes that the cumulative area of ​​the newly added urban patches reaches a preset threshold; The area of ​​the updated patch is recorded and accumulated in each iteration, and the iteration is stopped when the termination condition is met, and the simulation result is output.

7. A device for simulating urban expansion based on multi-scale fusion of patch and grid rules, characterized in that: include: A classification module, a processing module, a training module and a determination module, wherein: The classification module is configured to classify the original land use raster data based on the land use type, identify the patch structure of the original land use and number it, and superimpose it to obtain the urban expansion patch distribution map; The processing module is configured to process the spatial factors affecting urban expansion using spatial analysis tools to obtain driving variable raster data; the spatial factors include roads, stations, water areas, altitudes and slopes, and the grid size of the driving variable raster data is consistent with the original land use raster data; The training module is configured to perform patch averaging processing and sampling on the driving variable grid data to obtain patch-scale training sample data, input the training sample data into the random forest model, and obtain patch potential rules for constructing urban expansion simulation; in each simulation iteration, the simulated land use grid data is binarized, and the neighborhood interaction effect is calculated using a two-dimensional convolution algorithm to construct grid interaction rules for urban expansion simulation; the simulated spatial constraints, patch potential and grid interaction are integrated to calculate the state transition probability of each patch, and the patch area rules for urban expansion simulation are constructed according to the state transition probability; Obtain patch potential rules for constructing urban expansion simulations, including: Get the average value of each driving variable grid data in the patch to obtain the average driving variable grid data of the patch; Combined with the urban expansion patch distribution map and the patch average driving variable grid data, sampling is performed in a manner that the urban expansion patch label is 1 and the other patch labels are 0, and the patch average driving variable grid data value at the sample patch is obtained to obtain the sample data at the patch scale; Training a random forest model with the sample data at the patch scale, inputting the patch-averaged driving variable grid data into the trained random forest model to obtain development potential values ​​of all patches, and using the development potential values ​​as patch potential rules for constructing urban expansion simulation; The determination module is configured to determine the expansion simulation result of urban land based on the patch potential rule, the grid interaction rule and the patch area rule.

8. An electronic device comprising a processor and a memory; the memory stores a computer program, wherein: When the computer program is executed by the processor, the computer program implements the method for simulating urban expansion based on multi-scale fusion of patch and grid rules according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images