A Classification Method for Township Shrinkage Features Based on Multi-Scale Association

Abstract

This invention discloses a method for classifying township shrinkage characteristics based on multi-scale correlation: First, target townships are identified, and study townships are selected, with their spatial boundary data extracted. Second, the spatial distance between the target townships and the study townships is measured, constructing a spatial distance matrix. Third, the spatial distance matrix is ​​reclassified, generating first, second, and third-level spatial relationship matrices. Fourth, the population size change rate, socio-economic vitality change rate, and construction land size change rate for each target township at each spatial level during the study period are calculated. Fifth, a shrinkage characteristic vector set for the target townships is constructed, classifying the shrinkage characteristics. Finally, a shrinkage characteristic map of the target townships is drawn. This invention accurately identifies and classifies township shrinkage characteristics, assesses township development status, and effectively guides the allocation of construction land indicators, the configuration of public service facilities, and the formulation of infrastructure construction standards in territorial spatial planning.

Description

Technical Field

[0001] This invention relates to the field of geographic information technology, and more specifically, to a method for classifying township shrinkage features based on multi-scale correlation. Background Technology

[0002] Against the backdrop of urbanization entering a high-quality development stage, the development differentiation among townships is becoming increasingly prominent. Some townships are shrinking due to population loss, economic decline, and weakened functions, becoming a focal point that urgently needs attention in the coordinated development of urban and rural areas. Accurately identifying the characteristics of township shrinkage plays an important role in accurately assessing the development status of townships, scientifically allocating construction land quotas, rationally allocating public service facilities, and determining infrastructure construction standards.

[0003] Current classification of township shrinkage typically uses township administrative divisions as the unit of analysis. It utilizes data such as socio-economic statistics, census data, street view data, and remote sensing images, and employs methods such as interval numerical statistics and spatial morphology analysis to classify townships based on the magnitude of interval numerical changes or spatial morphological characteristics within a certain period.

[0004] However, existing methods for classifying township shrinkage ignore the shrinkage characteristics of townships and surrounding areas, resulting in inaccurate representation of shrinkage characteristics. This has limitations in assessing the development status of townships and in determining the allocation of construction land indicators, configuration of public service facilities, and standards for infrastructure construction in territorial spatial planning. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, this invention proposes a township shrinkage characteristic classification method based on multi-scale correlation. It focuses on the multi-dimensional shrinkage characteristics of the township's location, such as the population size change rate, construction land size change rate, and socio-economic vitality index change rate at multiple scales. Through cluster analysis, it accurately identifies the township shrinkage characteristic classification, thereby more accurately assessing the township's development status and more effectively guiding the allocation of construction land indicators, the configuration of public service facilities, and the formulation of infrastructure construction standards in territorial spatial planning.

[0006] The objective of this invention is achieved through the following technical solutions.

[0007] A township shrinkage feature classification method based on multi-scale correlation includes the following steps:

[0008] S1: Identify the target townships, select the townships to be studied, and extract their spatial boundary data respectively;

[0009] S2: Based on the above spatial boundary data, measure the spatial distance between the target township and the research township, and construct a spatial distance matrix between the target township and the research township.

[0010] S3: Reclassify the spatial distance matrix between the target townships and the research townships to generate first, second and third level spatial relationship matrices respectively;

[0011] S4: Determine the research period, obtain the population size data at the start and end times of the research object townships, and calculate the population size change rate of each target township at each level of space during the research period in combination with the spatial relationship matrix at each level mentioned above.

[0012] S5: Obtain the nighttime light intensity data at the start and end times of the study object townships respectively. Combine the above population size data and spatial relationship matrices at each level to calculate the rate of change of socio-economic vitality of each target township at each level of space during the study period.

[0013] S6: Obtain the construction land data at the start and end times of the research object townships respectively, and calculate the rate of change of construction land scale of each target township at each level of space during the research period by combining the spatial relationship matrices at each level mentioned above.

[0014] S7: Extract the population size change rate, socio-economic vitality change rate, and construction land size change rate of each level of space corresponding to the above target townships during the study period, construct the shrinkage feature vector set of the target townships, and use the K-means clustering analysis method to classify the shrinkage features of the target townships.

[0015] S8: Based on the classification results of the shrinkage characteristics of the target townships, draw a shrinkage characteristic map of the target townships.

[0016] Furthermore, in step S1, the target townships are clearly identified, the townships to be studied are selected, and their spatial boundary data are extracted. The specific process includes:

[0017] S11: Obtain spatial boundary data of townships nationwide. Using ArcGIS's "Projection" tool, transform the current coordinate system into a suitable projected coordinate system to obtain new spatial boundary data of townships nationwide. Load this data into an ArcMap document. Select the aforementioned projected coordinate system, choose the spatial boundary data of each target township, and export it as data for each target township. The target townships refer to the townships whose shrinkage characteristics need to be measured.

[0018] S12: Using ArcGIS's "Merge" tool, input the data of each target township and generate the range data of each target township respectively; using ArcGIS's "Buffer" tool, input the range data of each target township, set the distance value, and generate the buffer data of each target township respectively; using ArcGIS's "Select by Location" tool, select the new national township spatial boundary data as the target layer, select the target township buffer data as the source layer, and select "Intersect with Source Layer" as the spatial selection method for the target layer objects to determine the spatial boundary data of each research township and export it as the data of each research township; wherein, the research township refers to the townships covered by the range of the target township and the surrounding spatial range that will affect the shrinkage characteristics of the target township.

[0019] Furthermore, the process of measuring the spatial distance between the target township and the research township based on the aforementioned spatial boundary data, and constructing a spatial distance matrix between the target township and the research township in step S2, specifically includes:

[0020] S21: Using ArcGIS's "Feature to Point" tool, input the township data of each target object and the township data of each research object respectively to obtain the point data of each target object and the point data of each research object.

[0021] S22: Using ArcGIS's "Point Distance" tool, input the target township point data as features and the study township point data as neighboring features to obtain the distances between the centers of all target townships and the centers of all study townships. Then, construct a spatial distance matrix between the target townships and the study townships, as shown in the following formula:

[0022] (1),

[0023] In the formula, D is the spatial distance matrix between the target township and the research township. Target townships Townships and villages as research subjects Spatial distance, , The number of target townships , The number of townships included in the study.

[0024] Furthermore, in step S3, the spatial distance matrix between the target townships and the research townships is reclassified to generate first, second, and third level spatial relationship matrices, respectively. The specific process includes:

[0025] S31: Set the first-level spatial threshold to 0, reclassify the spatial distance matrix between the target township and the research township, reassign the elements equal to 0 in the matrix to 1, and reassign the elements greater than 0 to 0, to obtain the first-level spatial relationship matrix.

[0026] S32: Set the second-level spatial threshold to A2, reclassify the spatial distance matrix between the target township and the research township, assign 1 to elements in the matrix that are less than or equal to A2, and assign 0 to elements that are greater than A2, to obtain the second-level spatial relationship matrix.

[0027] S33: Set the third-level spatial threshold to A3, reclassify the spatial distance matrix between the target township and the research township, assign a value of 1 to elements in the matrix that are less than or equal to A3, and assign a value of 0 to elements that are greater than A3, to obtain the third-level spatial relationship matrix.

[0028] Further, in step S4, the research period is determined, and population size data at the start and end times of the research object townships are obtained. Combining this with the aforementioned spatial relationship matrices at each level, the population size change rate of each target township at each spatial level during the research period is calculated. The specific process includes:

[0029] S41: Determine the research period as needed, obtain raster-format population size data for the starting and ending time nodes of the townships within the research object, and load them into an ArcGIS document;

[0030] S42: Using ArcGIS's "Show Zonal Statistics in Tables" tool, input raster or feature area data, selecting the spatial boundary data of the towns and villages under study. Select FID for the segment name field. Input raster data, selecting population size data at the start and end time points respectively. Obtain the population size of each town and village at the start and end time points. Construct population size matrices for the towns and villages at the start and end time points respectively. The matrix format is as follows:

[0031] (2),

[0032] (3),

[0033] In the formula, and These are the population size matrices of the townships at the start and end of the research period, respectively. and Townships that are the subjects of the study Population size at the start and end time points, , The number of townships included in the study;

[0034] S43: Using the spatial relationship matrices of the first, second, and third levels mentioned above, and the population size matrices of the target townships at the starting and ending times, calculate the population size matrices of the target townships at each spatial level, respectively. The calculation formulas are as follows:

[0035] (4),

[0036] (5),

[0037] In the formula, For the first Hierarchical spatial relationship matrix ; The target townships correspond to the first The hierarchical spatial starting time node population size matrix and the target township corresponding to the first Population size matrix at the end of the hierarchical space and These are the population size matrices of the townships under study at the start and end of their respective study times. and Target townships respectively Corresponding to the The population size of the hierarchical space at the start and end time points. , The number of target townships;

[0038] S44: Calculate the population size change rate of each target township at each spatial level during the study period. The calculation formula is as follows:

[0039] (6),

[0040] In the formula, Target townships Corresponding to the The rate of change in population size in the hierarchical space during the study period. and Target townships respectively Corresponding to the The population size of the hierarchical space at the start and end time points.

[0041] Further, in step S5, nighttime light intensity data at the start and end times of the study target townships are obtained respectively. Combined with the aforementioned population size data and spatial relationship matrices at each level, the rate of change in socioeconomic vitality at each spatial level corresponding to each target township during the study period is calculated. The specific process includes:

[0042] S51: Obtain the nighttime light intensity data at the start and end times of the research subjects' townships, and load them into an ArcMap document;

[0043] S52: Using ArcGIS's "Resampling" tool, input the nighttime light intensity data at the start and end times of the study object's township area, set the cell size, select the average value as the sampling technique, and generate new nighttime light intensity data at the start and end times of the study object's township area, respectively.

[0044] S53: Using ArcGIS's "Resampling" tool, input the population size data at the start time node and the population size data at the end time node of the research object's township area, respectively. Output the same pixel size selection as the nighttime light intensity data at the start time node of the new research object's township area, and generate the population size data at the start time node and the population size data at the end time node of the new research object's township area, respectively.

[0045] S54: Using ArcMap's "Raster Calculator" tool, input and select "Nighttime light intensity data for the starting time point of the new research subject's township area". The starting time point population size data of the new research subject township area is used to generate the socio-economic vitality index data of the starting time point of the research subject township area; using ArcMap's "Raster Calculator" tool, input and select "the ending time point nighttime light intensity data of the new research subject township area * the ending time point population size data of the new research subject township area" to generate the socio-economic vitality index data of the ending time point of the research subject township area.

[0046] S55: Using ArcGIS's "Display Zonal Statistics in Tables" tool, input raster or feature area data, selecting the spatial boundary data of the townships under study. Select FID for the segment name field. Input raster data for the socio-economic vitality index at the start and end times of the township's geographical area. Generate the socio-economic vitality of each township at both the start and end times, and construct socio-economic vitality matrices for each township at both times. The matrix format is as follows:

[0047] (7),

[0048] (8),

[0049] In the formula, and These are the socio-economic vitality matrices of the townships under study at the start and end of their respective study times; and Townships that are the subjects of the study Socioeconomic vitality at the start and end time points , The number of townships included in the study;

[0050] S56: Using the above-mentioned first, second, and third-level spatial relationship matrices, and the socio-economic vitality matrices of the target townships at the start and end times, calculate the socio-economic vitality matrices of each spatial level corresponding to the target townships at the start and end times, respectively. The calculation formulas are as follows:

[0051] (9),

[0052] (10)

[0053] In the formula, For the first Hierarchical spatial relationship matrix ; The target townships correspond to the first The hierarchical space's socio-economic vitality matrix at the starting time point and the corresponding target townships are as follows: The socio-economic vitality matrix of the hierarchical space at the end of the time node, and These are the socioeconomic vitality matrices of the townships under study at the beginning and end of their respective study periods. and Target townships respectively Corresponding to the The socioeconomic vitality of hierarchical space at the start and end time points. , The number of target townships;

[0054] S57: Calculate the rate of change in socioeconomic vitality at each spatial level corresponding to each target township during the study period. The calculation formula is as follows:

[0055] (11),

[0056] In the formula, Target townships Corresponding to the The rate of change of socioeconomic vitality in the hierarchical space during the research period. and Target townships respectively Corresponding to the The socio-economic vitality of hierarchical space at the start and end time points.

[0057] Further, in step S6, the construction land data at the start and end times of the study object townships are obtained respectively. Combined with the spatial relationship matrices at each level mentioned above, the rate of change of construction land scale at each spatial level corresponding to each target township during the study period is calculated. The specific process includes:

[0058] S61: Obtain the raster-format construction land data for the starting and ending time nodes within the township scope of the research object, and load them into an ArcGIS document;

[0059] S62: Using ArcGIS's "Display Zonal Statistics in Tables" tool, input raster or feature area data, selecting the construction land data of the townships under study. Select FID for the segment name field. Input raster data, selecting the construction land data of the starting and ending time nodes for the townships under study in raster form. Obtain the construction land scale of each township under study at the starting and ending time nodes. Construct construction land scale matrices for the townships under study at the starting and ending time nodes, respectively. The matrix format is as follows:

[0060] (12)

[0061] (13)

[0062] In the formula, and These are the construction land scale matrices for the starting time nodes and the construction land scale matrices for the ending time nodes of the towns and villages under study, respectively. and Townships that are the subjects of the study The scale of construction land at the start and end time points, , The number of townships included in the study;

[0063] S63: Using the spatial relationship matrices of the first, second, and third levels mentioned above, and the construction land scale matrix of the target township at the start and end times, calculate the construction land scale matrix of each spatial level corresponding to the target township at the start and end times, respectively. The calculation formula is as follows:

[0064] (14)

[0065] (15)

[0066] In the formula, For the first Hierarchical spatial relationship matrix ; The target townships correspond to the first The hierarchical spatial construction land scale matrix at the starting time node and the corresponding target township The matrix of construction land scale in the hierarchical space at the end of the time node. and These are the construction land scale matrices for the townships under study at the start and end of their respective study times. and Target townships respectively Corresponding to the The scale of construction land at the start and end time points of the hierarchical space. , The number of target townships;

[0067] S64: Calculate the rate of change of construction land scale at each spatial level for each target township during the study period. The calculation formula is as follows:

[0068] (16)

[0069] In the formula, Target townships Corresponding to the The rate of change in the scale of construction land in the hierarchical space during the study period. and Target townships respectively Corresponding to the The scale of construction land at the start and end time points of the hierarchical space.

[0070] Further, in step S7, the population size change rate, socio-economic vitality change rate, and construction land size change rate of each level of space corresponding to the above-mentioned target townships are extracted during the study period to construct a shrinkage feature vector set of the target townships. K-means clustering analysis is then used to classify the shrinkage features of the target townships. The specific process includes:

[0071] S71: Extract the population size change rate, socio-economic activity change rate, and construction land scale change rate of each target township in the first, second, and third spatial levels during the study period, and construct a shrinkage feature vector set for the target township. ,

[0072] (17)

[0073] (18)

[0074] In the formula, Target townships The shrinking eigenvector, , The number of target townships; , , Target townships respectively Corresponding to the population size change rates of the first, second, and third spatial levels during the study period, , , Target townships Corresponding to the rates of change in socioeconomic vitality in the first, second, and third spatial levels during the research period, , , Target townships The rate of change in the scale of construction land corresponding to the first, second, and third spatial levels during the study period;

[0075] S72: Perform statistical analysis on the shrinkage feature vector set of the above target townships, and use the K-means clustering analysis method to calculate the shrinkage feature classification results of the target townships.

[0076] Furthermore, in step S8, based on the classification results of the shrinkage characteristics of the target townships, a shrinkage characteristic map of the target townships is drawn and applied to the scientific decision-making of township resource allocation. The specific process includes:

[0077] S81: Import the target township shrinkage characteristic classification results calculated above into Excel software and save it;

[0078] S82: Using the "Join Fields" tool in ArcMap, select the target township data as the input table, select FID as the input join field, select the shrinkage feature classification result of the target township as the join table, and select FID as the output join field to obtain an attribute table containing the shrinkage feature classification result of the target township after joining.

[0079] S83: In the ArcMap document, open the properties of the target township data, select the unique value of the category in the symbol, select the target township shrinkage feature classification result for the field value, and select Add All Values ​​to obtain the target township shrinkage feature map and export the map; use this map as the basis for judging the development status of townships and determining the allocation of construction land indicators, configuration of public service facilities, and infrastructure construction standards in territorial spatial planning.

[0080] Compared with the prior art, the beneficial effects of the technical solution of the present invention are:

[0081] (1) The present invention extends the identification of township shrinkage features from the identification of individual township features to the identification of multi-scale spatial features, which can more comprehensively and accurately reflect the actual types of township shrinkage.

[0082] (2) The township shrinkage characteristic classification results formed by the present invention can more effectively reflect the multi-scale and multi-dimensional characteristics of each classification, and thus more effectively judge the development status of townships and accurately guide the allocation of construction land indicators, configuration of public service facilities and infrastructure construction standards in territorial spatial planning.

[0083] Instruction manual illustrations

[0084] Figure 1 This is a flowchart of the township shrinkage feature classification method based on multi-scale correlation of the present invention. Detailed Implementation

[0085] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and specific examples. It should be understood that the specific embodiments described herein are merely intended to explain the present invention and not to limit it. For those skilled in the art, the present invention can be implemented without some of these specific details. The technical solutions of the present invention will now be described in detail with reference to the accompanying drawings.

[0086] To address the technical problems existing in the background art, this invention provides a township shrinkage feature classification method based on multi-scale correlation, such as... Figure 1 As shown, the specific steps include S1 to S8:

[0087] S1: Identify the target townships, select the townships to be studied, and extract their spatial boundary data respectively;

[0088] S2: Based on the above spatial boundary data, measure the spatial distance between the target township and the research township, and construct a spatial distance matrix between the target township and the research township.

[0089] Among them, the spatial distance between towns refers to the distance between different towns in geographical space. This invention measures it by the Euclidean length between the geometric centers of towns.

[0090] S3: Reclassify the spatial distance matrix between the target townships and the research townships to generate first, second and third level spatial relationship matrices respectively;

[0091] S4: Determine the research period, obtain the population size data at the start and end times of the research object townships, and calculate the population size change rate of each target township at each level of space during the research period in combination with the spatial relationship matrix at each level mentioned above.

[0092] S5: Obtain the nighttime light intensity data at the start and end times of the study object townships respectively. Combine the above population size data and spatial relationship matrices at each level to calculate the rate of change of socio-economic vitality of each target township at each level of space during the study period.

[0093] Among them, nighttime light intensity data refers to the information on the intensity of nighttime light emission from the Earth's surface obtained through satellite remote sensing. The data can be obtained in raster form, and the data information includes parameters such as coordinate system, pixel size, and pixel value.

[0094] Among them, socio-economic vitality refers to the ability to represent the activity of socio-economic activities. This invention uses indicators such as nighttime light intensity data and population density to calculate it.

[0095] S6: Obtain the construction land data at the start and end times of the research object townships respectively, and calculate the rate of change of construction land scale of each target township at each level of space during the research period by combining the spatial relationship matrices at each level mentioned above.

[0096] S7: Extract the population size change rate, socio-economic vitality change rate, and construction land size change rate of each level of space corresponding to the above target townships during the study period, construct the shrinkage feature vector set of the target townships, and use the K-means clustering analysis method to classify the shrinkage features of the target townships.

[0097] K-means clustering analysis is a classic unsupervised machine learning algorithm that classifies elements by analyzing the similarity of different vector data. This invention uses this method to classify the multidimensional shrinkage characteristics of townships.

[0098] Among them, the characteristics of township shrinkage are the overall characteristics of population, social, economic and spatial changes during the township shrinkage process. This invention uses multi-scale and multi-dimensional indicators for characterization.

[0099] S8: Based on the classification results of the shrinkage characteristics of the target townships, draw a shrinkage characteristic map of the target townships and apply it to the scientific decision-making of township resource allocation.

[0100] In one possible implementation, preferably, the specific process of step S1 above includes:

[0101] S11: Obtain national township spatial boundary data (.shp format). Using ArcGIS's "Project" tool, transform the current coordinate system into a suitable projected coordinate system (unit: meters) to obtain new national township spatial boundary data. Load this data into an ArcMap document, select the aforementioned projected coordinate system, choose the spatial boundary data of each target township, and export it as data for each target township. The target townships refer to those whose shrinkage characteristics need to be measured.

[0102] S12: Using ArcGIS's "Merge" tool, input the data of each target township to generate the range data of each target township; using ArcGIS's "Buffer" tool, input the range data of each target township, set the distance value (e.g., 50km, a threshold set according to current scientific consensus to reasonably select research townships, which can be adjusted according to actual needs), and generate the buffer data of each target township; using ArcGIS's "Select by Location" tool, select the new national township spatial boundary data as the target layer, select the target township buffer data as the source layer, and select "Intersect with Source Layer" as the spatial selection method for the target layer objects to determine the spatial boundary data of each research township, and export it as the data of each research township; wherein, the research township refers to the townships covered by the range of the target township and the surrounding spatial range that will affect the shrinkage characteristics of the target township.

[0103] In one possible implementation, preferably, the specific process of step S2 above includes:

[0104] S21: Using ArcGIS's "Feature to Point" tool, input the township data of each target object and the township data of each research object respectively to obtain the point data of each target object and the point data of each research object.

[0105] S22: Using ArcGIS's "Point Distance" tool, input the target township point data as features and the study township point data as neighboring features to obtain the distances between the centers of all target townships and the centers of all study townships. Then, construct a spatial distance matrix between the target townships and the study townships, as shown in the following formula:

[0106] (1),

[0107] In the formula, D is the spatial distance matrix between the target township and the research township. Target townships Townships and villages as research subjects Spatial distance, , The number of target townships , The number of townships included in the study.

[0108] In one possible implementation, preferably, step S3 above can reclassify the spatial distance matrix between the target township and the research township based on three levels of spatial thresholds, generating first, second, and third level spatial relationship matrices respectively. The specific process includes:

[0109] S31: Set the first-level spatial threshold to 0, reclassify the spatial distance matrix between the target township and the research township, reassign the elements equal to 0 in the matrix to 1, and reassign the elements greater than 0 to 0, to obtain the first-level spatial relationship matrix.

[0110] S32: Set the second-level spatial threshold to A2, reclassify the spatial distance matrix between the target township and the research township, assign 1 to elements in the matrix that are less than or equal to A2, and assign 0 to elements that are greater than A2, to obtain the second-level spatial relationship matrix.

[0111] S33: Set the third-level spatial threshold to A3, reclassify the spatial distance matrix between the target township and the research township, assign a value of 1 to elements in the matrix that are less than or equal to A3, and assign a value of 0 to elements that are greater than A3, to obtain the third-level spatial relationship matrix.

[0112] For example, A2 takes the value 200,000, and A3 takes the value 500,000. 0 meters, 200,000 meters, and 500,000 meters are three levels of spatial thresholds set according to scientific consensus, which can be adjusted according to actual needs.

[0113] In one possible implementation, preferably, the specific process of step S4 above includes:

[0114] S41: Determine the research period as needed, obtain raster-format population size data for the starting and ending time nodes of the townships within the research object, and load them into an ArcGIS document;

[0115] S42: Using ArcGIS's "Zonal Statistics as Table" tool, input raster or feature area data, selecting the spatial boundary data of the towns and villages under study (obtained in step S12). Select FID for the segment name field. Input raster data, selecting population size data at the start and end time points respectively. Obtain the population size of each town and village at the start and end time points. Construct the population size matrix of each town and village at the start and end time points respectively. The matrix format is as follows:

[0116] (2),

[0117] (3),

[0118] In the formula, and These are the population size matrices of the townships at the start and end of the research period, respectively. and Townships that are the subjects of the study Population size at the start and end time points, , The number of townships included in the study;

[0119] S43: Using the spatial relationship matrices of the first, second, and third levels mentioned above, and the population size matrices of the target townships at the starting and ending times, calculate the population size matrices of the target townships at each spatial level, respectively. The calculation formulas are as follows:

[0120] (4),

[0121] (5),

[0122] In the formula, For the first Hierarchical spatial relationship matrix ; The target townships correspond to the first The hierarchical spatial starting time node population size matrix and the target township corresponding to the first Population size matrix at the end of the hierarchical space and These are the population size matrices of the townships under study at the start and end of their respective study times. and Target townships respectively Corresponding to the The population size of the hierarchical space at the start and end time points. , The number of target townships;

[0123] S44: Based on the population size matrices of the target townships at the beginning and end of each spatial level, the population size change rate of each target township at each spatial level during the study period is calculated using the following formulas:

[0124] (6),

[0125] In the formula, Target townships Corresponding to the The rate of change in population size in the hierarchical space during the study period. and Target townships respectively Corresponding to the The population size of the hierarchical space at the start and end time points.

[0126] In one possible implementation, preferably, the specific process of step S5 above includes:

[0127] S51: Obtain the nighttime light intensity data at the start and end times of the research subjects' townships, and load them into an ArcMap document;

[0128] S52: Using ArcGIS's "Resample" tool, input the nighttime light intensity data for the starting and ending times within the study area of ​​the township, and set the cell size (e.g., cell size X value 1000, Y value 1000, 1000). 1000 is one option, but other values ​​are also possible (this invention does not limit this). The sampling technique is to select the average value, and to generate nighttime light intensity data at the start time node and the end time node of the new research object township area respectively.

[0129] S53: Using ArcGIS's "Resample" tool, input the population size data at the start and end times of the study area's township (obtained in step S41). The output cell size should be the same as the nighttime light intensity data at the new start time of the study area's township, for example, using 1000 pixels. With a pixel size of 1000, population size data at the start and end times of the new research subject township area are generated respectively.

[0130] S54: Using ArcMap's "Raster Calculator" tool, input the nighttime light intensity data for the starting time point of the new research subject's township area. The starting point population size data of the new research subject township area is used to generate the socioeconomic vitality index data of the starting point township area. Using ArcMap's "Raster Calculator" tool, input and select "nighttime light intensity data of the new research subject township area at the end time point * population size data of the new research subject township area at the end time point" to generate the socioeconomic vitality index data of the research subject township area at the end time point.

[0131] S55: Using ArcGIS's "Zonal Statistics as Table" tool, input raster or feature area data to select the spatial boundary data of the townships under study (obtained in step S12). Select FID for the segment name field. Input the socio-economic vitality index data of the townships under study at the start and end times respectively. Generate the socio-economic vitality of each township under study at the start and end times respectively, and construct the socio-economic vitality matrices of the townships under study at the start and end times respectively. The matrix form is as follows:

[0132] (7),

[0133] (8),

[0134] In the formula, and These are the socio-economic vitality matrices of the townships under study at the start and end of their respective study times; and Townships that are the subjects of the study Socioeconomic vitality at the start and end time points , The number of townships included in the study;

[0135] S56: Using the above-mentioned first, second, and third-level spatial relationship matrices, and the socio-economic vitality matrices of the target townships at the start and end times, calculate the socio-economic vitality matrices of each spatial level corresponding to the target townships at the start and end times, respectively. The calculation formulas are as follows:

[0136] (9),

[0137] (10)

[0138] In the formula, For the first Hierarchical spatial relationship matrix ; The target townships correspond to the first The hierarchical space's socio-economic vitality matrix at the starting time point and the corresponding target townships are as follows: The socio-economic vitality matrix of the hierarchical space at the end of the time node, and These are the socioeconomic vitality matrices of the townships under study at the beginning and end of their respective study periods. and Target townships respectively Corresponding to the The socioeconomic vitality of hierarchical space at the start and end time points. , The number of target townships;

[0139] S57: Based on the socio-economic vitality matrices of the target townships at each spatial level at the start and end time points, calculate the rate of change of socio-economic vitality of each target township at each spatial level during the study period. The calculation formula is as follows:

[0140] (11),

[0141] In the formula, Target townships Corresponding to the The rate of change of socioeconomic vitality in the hierarchical space during the research period. and Target townships respectively Corresponding to the The socio-economic vitality of hierarchical space at the start and end time points.

[0142] In one possible implementation, preferably, the specific process of step S6 above includes:

[0143] S61: Determine the research period as needed (consistent with the research period determined in step S41), obtain the raster data of the start time node and the end time node of the construction land within the township of the research object, and load them into the ArcGIS document;

[0144] S62: Using ArcGIS's "Zonal Statistic as Table" tool, input raster or feature area data, selecting the construction land data of the townships under study. Select FID for the segment name field. Input raster data, selecting the construction land data at the start and end times for the townships under study in raster format. Obtain the construction land scale of each township under study at the start and end times. Construct construction land scale matrices for the townships under study at the start and end times, respectively. The matrix format is as follows:

[0145] (12)

[0146] (13)

[0147] In the formula, and These are the construction land scale matrices for the starting time nodes and the construction land scale matrices for the ending time nodes of the towns and villages under study, respectively. and Townships that are the subjects of the study The scale of construction land at the start and end time points, , The number of townships included in the study;

[0148] S63: Using the spatial relationship matrices of the first, second, and third levels mentioned above, and the construction land scale matrix of the target township at the start and end times, calculate the construction land scale matrix of each spatial level corresponding to the target township at the start and end times, respectively. The calculation formula is as follows:

[0149] (14)

[0150] (15)

[0151] In the formula, For the first Hierarchical spatial relationship matrix ; The target townships correspond to the first The hierarchical spatial construction land scale matrix at the starting time node and the corresponding target township The matrix of construction land scale in the hierarchical space at the end of the time node. and These are the construction land scale matrices for the townships under study at the start and end of their respective study times. and Target townships respectively Corresponding to the The scale of construction land at the start and end time points of the hierarchical space. , The number of target townships;

[0152] S64: Based on the construction land scale matrix of each spatial level corresponding to the target townships at the start and end time nodes, calculate the rate of change of construction land scale of each spatial level corresponding to each target township during the study period. The calculation formula is as follows:

[0153] (16)

[0154] In the formula, Target townships Corresponding to the The rate of change in the scale of construction land in the hierarchical space during the study period. and Target townships respectively Corresponding to the The scale of construction land at the start and end time points of the hierarchical space.

[0155] In one possible implementation, preferably, the specific process of step S7 above includes:

[0156] S71: Extract the population size change rate of each target township in the first, second, and third spatial levels during the study period (obtained in step S44), the socio-economic vitality change rate of each target township in the first, second, and third spatial levels during the study period (obtained in step S57), and the construction land scale change rate of each target township in the first, second, and third spatial levels during the study period (obtained in step S64). Construct a shrinkage feature vector set for the target townships. ,

[0157] (17)

[0158] (18)

[0159] In the formula, Target townships The shrinking eigenvector, , The number of target townships; , , Target townships respectively Corresponding to the population size change rates of the first, second, and third spatial levels during the study period, , , Target townships Corresponding to the rates of change in socioeconomic vitality in the first, second, and third spatial levels during the research period, , , Target townships The rate of change in the scale of construction land corresponding to the first, second, and third spatial levels during the study period;

[0160] S72: Perform statistical analysis on the shrinkage feature vector set of the above target townships, and use the K-means clustering analysis method to calculate the shrinkage feature classification results of the target townships.

[0161] For example, the shrinkage feature vector set of the target townships is loaded into statistical analysis software. Using the K-means clustering analysis module, the number of random initial points is set to 10, and the number of clusters k is set to 6. The classification results of the shrinkage features of the target townships are then calculated. The number of clusters is determined to be 6, which is a threshold set according to scientific consensus. Statistical analysis software includes SPSS, R, Python, etc., and this invention does not limit the scope of the software.

[0162] In one possible implementation, preferably, the specific process of step S8 above includes:

[0163] S81: Import the target township shrinkage characteristic classification results calculated above into Excel software and save it (.xls format).

[0164] S82: Using the "Join-Field" tool in ArcMap, select the target township data as the input table (obtained in step S11), select FID as the input join field (this data is obtained from step S11), select the target township shrinkage feature classification result as the join table (obtained from step S81), and select FID as the output join field to obtain an attribute table containing the target township shrinkage feature classification result after joining. The selection of the join field can be set according to the ability to match the target township data and the target shrinkage feature classification result, and this invention does not impose any limitations.

[0165] S83: In the ArcMap document, open the Properties of the target township data, select Unique Values ​​for Categories in the Symbols section, select the target township shrinkage feature classification result for the field value (added to the attribute table in step S82), and select Add All Values ​​to obtain the target township shrinkage feature map and export the map; use this map as the basis for judging the development status of townships and determining the allocation of construction land indicators, configuration of public service facilities, and infrastructure construction standards in territorial spatial planning. Although the functions and working processes of the present invention have been described above in conjunction with the accompanying drawings, the present invention is not limited to the specific functions and working processes described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art, under the guidance of the present invention, can make many other forms without departing from the spirit and scope of the claims of the present invention, and these all fall within the protection scope of the present invention.

Claims

1. A township shrinkage feature classification method based on multi-scale association, characterized in that, Includes the following steps: S1: Identify the target townships, select the townships to be studied, and extract their spatial boundary data respectively; S2: Based on the above spatial boundary data, measure the spatial distance between the target township and the research township, and construct a spatial distance matrix between the target township and the research township. S3: Reclassify the spatial distance matrix between the target townships and the research townships to generate first, second and third level spatial relationship matrices respectively; The specific process includes: S31: Set the first-level spatial threshold to 0, reclassify the spatial distance matrix between the target township and the research township, reassign the elements equal to 0 in the matrix to 1, and reassign the elements greater than 0 to 0, to obtain the first-level spatial relationship matrix. S32: Set the second-level spatial threshold to A2, reclassify the spatial distance matrix between the target township and the research township, assign 1 to elements in the matrix that are less than or equal to A2, and assign 0 to elements that are greater than A2, to obtain the second-level spatial relationship matrix. S33: Set the third-level spatial threshold to A3, reclassify the spatial distance matrix between the target township and the research township, assign 1 to elements in the matrix that are less than or equal to A3, and assign 0 to elements that are greater than A3, to obtain the third-level spatial relationship matrix. S4: Determine the research period, obtain the population size data at the start and end times of the research object townships, and calculate the population size change rate of each target township at each level of space during the research period in combination with the spatial relationship matrix at each level mentioned above. S5: Obtain the nighttime light intensity data at the start and end times of the study object townships respectively. Combine the above population size data and spatial relationship matrices at each level to calculate the rate of change of socio-economic vitality of each target township at each level of space during the study period. S6: Obtain the construction land data at the start and end times of the research object townships respectively, and calculate the rate of change of construction land scale of each target township at each level of space during the research period by combining the spatial relationship matrices at each level mentioned above. S7: Extract the population size change rate, socio-economic vitality change rate, and construction land size change rate of each level of space corresponding to the above target townships during the study period, construct the shrinkage feature vector set of the target townships, and use the K-means clustering analysis method to classify the shrinkage features of the target townships. S8: Based on the classification results of the shrinkage characteristics of the target townships, draw a shrinkage characteristic map of the target townships.

2. The township shrinkage feature classification method based on multi-scale correlation according to claim 1, characterized in that, Step S1 involves identifying the target townships, selecting the townships for study, and extracting their spatial boundary data. The specific process includes: S11: Obtain spatial boundary data of townships nationwide. Using ArcGIS's "Projection" tool, transform the current coordinate system into a suitable projected coordinate system to obtain new spatial boundary data of townships nationwide. Load this data into an ArcMap document. Select the aforementioned projected coordinate system, choose the spatial boundary data of each target township, and export it as data for each target township. The target townships refer to the townships whose shrinkage characteristics need to be measured. S12: Using ArcGIS's "Merge" tool, input the data of each target township and generate the range data of each target township respectively; using ArcGIS's "Buffer" tool, input the range data of each target township, set the distance value, and generate the buffer data of each target township respectively; using ArcGIS's "Select by Location" tool, select the new national township spatial boundary data as the target layer, select the target township buffer data as the source layer, and select "Intersect with Source Layer" as the spatial selection method for the target layer objects to determine the spatial boundary data of each research township and export it as the data of each research township; wherein, the research township refers to the townships covered by the range of the target township and the surrounding spatial range that will affect the shrinkage characteristics of the target township.

3. The township shrinkage feature classification method based on multi-scale correlation according to claim 1, characterized in that, Step S2, which involves measuring the spatial distance between the target township and the research township based on the aforementioned spatial boundary data and constructing a spatial distance matrix between them, specifically includes the following steps: S21: Using ArcGIS's "Feature to Point" tool, input the township data of each target object and the township data of each research object respectively to obtain the point data of each target object and the point data of each research object. S22: Using ArcGIS's "Point Distance" tool, input the target township point data as features and the study township point data as neighboring features to obtain the distances between the centers of all target townships and the centers of all study townships. Then, construct a spatial distance matrix between the target townships and the study townships, as shown in the following formula: （1）， In the formula, D is the spatial distance matrix between the target township and the research township. Target townships Townships and villages as research subjects Spatial distance, , The number of target townships , The number of townships included in the study.

4. The township shrinkage feature classification method based on multi-scale correlation according to claim 1, characterized in that, In step S4, the research period is determined, and population size data at the start and end times of the research target townships are obtained. Combining this with the aforementioned spatial relationship matrices at each level, the population size change rate for each target township at each spatial level during the research period is calculated. The specific process includes: S41: Determine the research period as needed, obtain raster-format population size data for the starting and ending time nodes of the townships within the research object, and load them into an ArcGIS document; S42: Using ArcGIS's "Show Zonal Statistics in Tables" tool, input raster or feature area data, selecting the spatial boundary data of the towns and villages under study. Select FID for the segment name field. Input raster data, selecting population size data at the start and end time points respectively. Obtain the population size of each town and village at the start and end time points. Construct population size matrices for the towns and villages at the start and end time points respectively. The matrix format is as follows: （2）， （3）， In the formula, and These are the population size matrices of the townships at the start and end of the research period, respectively. and Townships that are the subjects of the study Population size at the start and end time points, , The number of townships included in the study; S43: Using the spatial relationship matrices of the first, second, and third levels mentioned above, and the population size matrices of the target townships at the starting and ending times, calculate the population size matrices of the target townships at each spatial level, respectively. The calculation formulas are as follows: （4）， （5）， In the formula, For the first Hierarchical spatial relationship matrix ; The target townships correspond to the first The hierarchical spatial starting time node population size matrix and the target township corresponding to the first Population size matrix at the end of the hierarchical space and These are the population size matrices of the townships under study at the start and end of their respective study times. and Target townships respectively Corresponding to the The population size of the hierarchical space at the start and end time points. , The number of target townships; S44: Calculate the population size change rate of each target township at each spatial level during the study period. The calculation formula is as follows: （6）， In the formula, Target townships Corresponding to the The rate of change in population size in the hierarchical space during the study period. and Target townships respectively Corresponding to the The population size of the hierarchical space at the start and end time points.

5. The township shrinkage feature classification method based on multi-scale correlation according to claim 1, characterized in that, In step S5, nighttime light intensity data at the start and end times of the study target townships are obtained. Combined with the aforementioned population size data and spatial relationship matrices at various levels, the rate of change in socioeconomic vitality at each spatial level corresponding to each target township during the study period is calculated. The specific process includes: S51: Obtain the nighttime light intensity data at the start and end times of the research subjects' townships, and load them into an ArcMap document; S52: Using ArcGIS's "Resampling" tool, input the nighttime light intensity data at the start and end times of the study object's township area, set the cell size, select the average value as the sampling technique, and generate new nighttime light intensity data at the start and end times of the study object's township area, respectively. S53: Using ArcGIS's "Resampling" tool, input the population size data at the start time node and the population size data at the end time node of the research object's township area respectively, and output the same pixel size selection as the nighttime light intensity data at the start time node of the new research object's township area, and generate new population size data at the start time node and the new population size data at the end time node of the new research object's township area respectively. S54: Using ArcMap's "Raster Calculator" tool, input and select "Nighttime light intensity data for the starting time point of the new research subject's township area". The data includes population size data at the start time of the new research township area, generating socioeconomic vitality index data at the start time of the research township area; and nighttime light intensity data at the end time of the new research township area using ArcMap's "Raster Calculator" tool. "Population size data at the end of the new research subject township scope", generating socioeconomic vitality index data at the end of the research subject township scope; S55: Using ArcGIS's "Display Zonal Statistics in Tables" tool, input raster or feature area data, selecting the spatial boundary data of the townships under study. Select FID for the segment name field. Input raster data for the socio-economic vitality index at the start and end times of the township's geographical area. Generate the socio-economic vitality of each township at both the start and end times, and construct socio-economic vitality matrices for each township at both times. The matrix format is as follows: （7）， （8）， In the formula, and These are the socio-economic vitality matrices of the townships under study at the start and end of their respective study times; and Townships that are the subjects of the study Socioeconomic vitality at the start and end time points , The number of townships included in the study; S56: Using the above-mentioned first, second, and third-level spatial relationship matrices, and the socio-economic vitality matrices of the target townships at the start and end times, calculate the socio-economic vitality matrices of each spatial level corresponding to the target townships at the start and end times, respectively. The calculation formulas are as follows: （9）， （10）， In the formula, For the first Hierarchical spatial relationship matrix ; The target townships correspond to the first The hierarchical space's socio-economic vitality matrix at the starting time point and the corresponding target townships are as follows: The socio-economic vitality matrix of the hierarchical space at the end of the time node, and These are the socioeconomic vitality matrices of the townships under study at the beginning and end of their respective study periods. and Target townships respectively Corresponding to the The socioeconomic vitality of hierarchical space at the start and end time points. , The number of target townships; S57: Calculate the rate of change in socioeconomic vitality at each spatial level corresponding to each target township during the study period. The calculation formula is as follows: （11）， In the formula, Target townships Corresponding to the The rate of change of socioeconomic vitality in the hierarchical space during the research period. and Target townships respectively Corresponding to the The socio-economic vitality of hierarchical space at the start and end time points.

6. The township shrinkage feature classification method based on multi-scale correlation according to claim 1, characterized in that, In step S6, the construction land data at the start and end times of the study object townships are obtained respectively. Combined with the spatial relationship matrices at each level mentioned above, the rate of change of construction land scale at each spatial level corresponding to each target township during the study period is calculated. The specific process includes: S61: Obtain the raster-format construction land data for the starting and ending time nodes within the township scope of the research object, and load them into an ArcGIS document; S62: Using ArcGIS's "Display Zonal Statistics in Tables" tool, input raster or feature area data, selecting the construction land data of the townships under study. Select FID for the segment name field. Input raster data, selecting the construction land data of the starting and ending time nodes for the townships under study in raster form. Obtain the construction land scale of each township under study at the starting and ending time nodes. Construct construction land scale matrices for the townships under study at the starting and ending time nodes, respectively. The matrix format is as follows: （12）， （13）， In the formula, and These are the construction land scale matrices for the starting time nodes and the construction land scale matrices for the ending time nodes of the towns and villages under study, respectively. and Townships that are the subjects of the study The scale of construction land at the start and end time points, , The number of townships included in the study; S63: Using the spatial relationship matrices of the first, second, and third levels mentioned above, and the construction land scale matrix of the target township at the start and end times, calculate the construction land scale matrix of each spatial level corresponding to the target township at the start and end times, respectively. The calculation formula is as follows: （14）， （15）， In the formula, For the first Hierarchical spatial relationship matrix ; The target townships correspond to the first The hierarchical spatial construction land scale matrix at the starting time node and the corresponding target township The matrix of construction land scale in the hierarchical space at the end of the time node. and These are the construction land scale matrices for the townships under study at the start and end of their respective study times. and Target townships respectively Corresponding to the The scale of construction land at the start and end time points of the hierarchical space. , The number of target townships; S64: Calculate the rate of change of construction land scale at each spatial level for each target township during the study period. The calculation formula is as follows: （16）， In the formula, Target townships Corresponding to the The rate of change in the scale of construction land in the hierarchical space during the study period. and Target townships respectively Corresponding to the The scale of construction land at the start and end time points of the hierarchical space.

7. The township shrinkage feature classification method based on multi-scale correlation according to claim 1, characterized in that, Step S7 extracts the population size change rate, socio-economic vitality change rate, and construction land size change rate for each target township at each spatial level during the study period, constructs a shrinkage feature vector set for the target townships, and uses K-means clustering analysis to classify the shrinkage features of the target townships. The specific process includes: S71: Extract the population size change rate, socio-economic activity change rate, and construction land scale change rate of each target township in the first, second, and third spatial levels during the study period, and construct a shrinkage feature vector set for the target township. , （17） （18） In the formula, Target townships The shrinking eigenvector, , The number of target townships; , , Target townships respectively Corresponding to the population size change rates of the first, second, and third spatial levels during the study period, , , Target townships Corresponding to the rates of change in socioeconomic vitality in the first, second, and third spatial levels during the research period, , , Target townships The rate of change in the scale of construction land corresponding to the first, second, and third spatial levels during the study period; S72: Perform statistical analysis on the shrinkage feature vector set of the above target townships, and use the K-means clustering analysis method to calculate the shrinkage feature classification results of the target townships.

8. The township shrinkage feature classification method based on multi-scale correlation according to claim 1, characterized in that, In step S8, based on the classification results of the shrinkage characteristics of the target townships, a shrinkage characteristic map of the target townships is drawn and applied to the scientific decision-making of township resource allocation. The specific process includes: S81: Import the target township shrinkage characteristic classification results calculated above into Excel software and save it; S82: Using the "Join Fields" tool in ArcMap, select the target township data as the input table, select FID as the input join field, select the shrinkage feature classification result of the target township as the join table, and select FID as the output join field to obtain an attribute table containing the shrinkage feature classification result of the target township after joining. S83: In the ArcMap document, open the properties of the target township data, select the unique value of the category in the symbol, select the target township shrinkage feature classification result for the field value, and select Add All Values ​​to obtain the target township shrinkage feature map and export the map; use this map as the basis for judging the development status of townships and determining the allocation of construction land indicators, configuration of public service facilities, and infrastructure construction standards in territorial spatial planning.

Images