Land space planning large-scale data storage method and system

Abstract

The invention relates to the technical field of data processing related to territorial space planning, in particular to a territorial space planning large-scale data storage method and system, and the method comprises the steps: obtaining territorial space planning data of a target region; converting the territorial space planning data into a space data distribution diagram; performing hierarchical clustering on each data point in the spatial data distribution diagram under a plurality of clustering scales to obtain a land use clustering result under each clustering scale; determining a selection index value corresponding to each clustering scale; selecting the clustering scale with the maximum index value as a target clustering scale; and based on the land use clustering result under the target clustering scale, carrying out classification marking and storage on the land space planning data of the target region. As the final target clustering scale is determined according to the selection index value, compared with a manual setting mode, the method is more accurate and effective, the clustering result has more reference value, and the reasonability and accuracy of data storage are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing related to territorial spatial planning, and particularly relates to a method and system for storing large-scale territorial spatial planning data. Background Art

[0002] Territorial spatial planning refers to making arrangements for the development and protection of territorial space within a certain area in terms of space and time. It is a guide for the country's spatial development and a spatial blueprint for sustainable development, and is the basic basis for various development, protection, and construction activities. The data involved in territorial spatial planning includes natural environment data, built environment data, and social environment data, representing natural environment data such as climate, soil, and landform, as well as economic and social data generated by human activities. Therefore, the data involved in territorial spatial planning is huge in quantity and diverse in types, and large-scale data storage is required.

[0003] In the related art, during the territorial spatial planning process, in order to minimize the amount of data as much as possible, hierarchical clustering is often performed on the territorial spatial planning data, and subsequent data in the same cluster are replaced with the same value, so as to reduce the storage amount while ensuring that the loss of territorial spatial planning data is minimized. However, in the traditional solution, the clustering scale needs to be set manually, usually directly determined according to previous division experience. The setting link of the clustering scale has a large human interference factor, resulting in a low reference value of the clustering result, and thus it is difficult to ensure the rationality and accuracy of data storage. Summary of the Invention

[0004] In order to solve the technical problems in the traditional solution that the clustering scale is set manually according to previous division experience, the reference value of the clustering result is low, and the rationality and accuracy of data storage are insufficient, the purpose of the present invention is to provide a method and system for storing large-scale territorial spatial planning data, and the specific technical solutions adopted are as follows: A method for storing large-scale territorial spatial planning data, the method comprising: Obtaining territorial spatial planning data of a target area; Converting the territorial spatial planning data into a spatial data distribution map; Performing hierarchical clustering on each data point in the spatial data distribution map at multiple clustering scales to obtain a land use clustering result at each clustering scale; Determining a selection index value corresponding to each clustering scale according to the land use clustering result at each clustering scale; Taking the clustering scale with the largest selection index value as the target clustering scale; Based on the land use clustering result at the target clustering scale, classifying and storing the territorial spatial planning data of the target area.

[0005] A method for storing large-scale data in territorial spatial planning according to the present invention determines a selection index value corresponding to each clustering scale based on the land use clustering result at each clustering scale, including: Calculate the clustering accuracy of each cluster at each clustering scale according to the land use clustering result; Take the average of the clustering accuracies of all clusters at each clustering scale to obtain the selection index value corresponding to each clustering scale.

[0006] A method for storing large-scale data in territorial spatial planning according to the present invention calculates the clustering accuracy of each cluster at each clustering scale according to the land use clustering result, including: Divide the data points in each cluster of the land use clustering result into regions to obtain multiple land blocks corresponding to each cluster; Determine the integrity of each land block in each cluster and determine the characteristic difference value between each land block and any adjacent block; Calculate the initial accuracy value of each cluster based on the integrity of each land block in each cluster; Determine the representative coefficient of each cluster based on the integrity of each land block, the integrity of adjacent blocks, and the characteristic difference value between each land block and any adjacent block; Based on the representative coefficient, correct the initial accuracy value of each cluster to obtain the clustering accuracy of each cluster at each clustering scale.

[0007] A method for storing large-scale data in territorial spatial planning according to the present invention determines the integrity of each land block in each cluster, including: In each cluster, determine multiple target blocks adjacent to each land block; Respectively determine the block sets for spatially dividing each land block and each target block corresponding to each land block; Obtain the first characteristic value of each land block, the second characteristic value of each target block, the third characteristic value of any element block in each block set, and the number of all element blocks in each block set; Based on the reciprocal of the number of all element blocks in each block set, determine the calculation coefficient value corresponding to each target block; Based on the first characteristic value of each land block, the second characteristic value of each target block, and the third characteristic value of any element block in each block set, calculate the characteristic difference summation value corresponding to each target block; Multiply the calculation coefficient value of each target block by the characteristic difference summation value to calculate the block integrity value of each land block relative to each target block; Calculate the mean of all the sub-block integrity values corresponding to each land block, and calculate the integrity of each land block in each cluster.

[0008] According to a large-scale data storage method for territorial space planning provided by the present invention, based on the integrity of each land block in each cluster, calculate the initial accuracy value of each cluster, including: Calculate the mean of the integrity of each land block in each cluster, and calculate the mean integrity of each cluster; Calculate the standard deviation of the integrity of each land block in each cluster, and calculate the standard deviation of the integrity of each cluster; Divide the mean integrity of each cluster by the standard deviation of the integrity of each cluster to calculate the initial accuracy value of each cluster.

[0009] According to a large-scale data storage method for territorial space planning provided by the present invention, based on the integrity of each land block, the integrity of adjacent blocks, and the characteristic difference value between each land block and any adjacent block, determine the representative coefficient of each cluster, including: Based on the integrity of each land block, the integrity of adjacent blocks, and the characteristic difference value between each land block and any adjacent block, calculate the block correlation degree between each land block and any adjacent block; Based on the block correlation degree between each land block and any adjacent block, calculate the spatial weight between each land block and any adjacent block; Based on the spatial weight, calculate the association index between each land block and any adjacent block; Calculate the mean of all the association indexes under each cluster to calculate the representative coefficient of each cluster.

[0010] According to a large-scale data storage method for territorial space planning provided by the present invention, based on the block correlation degree between each land block and any adjacent block, calculate the spatial weight between each land block and any adjacent block, including: For each land block and any adjacent block, subtract the block correlation degree from 1 to obtain a first intermediate quantity; Respectively determine the first integrity of the cluster where the land block is located and the second integrity of the cluster where the adjacent block is located; Sum the first integrity and the second integrity to obtain a second intermediate quantity; Divide the first intermediate quantity by the second intermediate quantity to calculate the spatial weight between each land block and any adjacent block.

[0011] A method for storing large-scale data of territorial space planning provided by the present invention, based on the representative coefficient, corrects the initial accuracy value of each type of cluster to obtain the clustering accuracy of each cluster at each clustering scale, including: Add the representative coefficient to 1 to calculate a correction coefficient; Divide the initial accuracy value of each type of cluster by the correction coefficient and perform normalization processing to obtain the clustering accuracy of each cluster at each clustering scale.

[0012] A method for storing large-scale data of territorial space planning provided by the present invention, based on the land use clustering result at the target clustering scale, classifies and stores the territorial space planning data of the target area, including: According to the land use clustering result at the target clustering scale, classify and label the territorial space planning data belonging to the same cluster with the same cluster label; Store the cluster label and spatial location information of the territorial space planning data after classification labeling.

[0013] On the other hand, the present invention also provides a large-scale data storage system for territorial space planning, and the system includes: An acquisition module for acquiring territorial space planning data of a target area; A conversion module for converting the territorial space planning data into a spatial data distribution map; A clustering module for performing hierarchical clustering on each data point in the spatial data distribution map at multiple clustering scales to obtain the land use clustering result at each clustering scale; A calculation module for determining the selection index value corresponding to each clustering scale according to the land use clustering result at each clustering scale; A processing module for taking the clustering scale with the largest selection index value as the target clustering scale; A storage module for classifying and storing the territorial space planning data of the target area based on the land use clustering result at the target clustering scale.

[0014] The present invention has the following beneficial effects: By obtaining the land space planning data of the target area and converting it into a spatial data distribution map, each data point in the spatial data distribution map is hierarchically clustered at multiple clustering scales to obtain the land use clustering results at each clustering scale. Then, according to the land use clustering results at each clustering scale, the selection index value corresponding to each clustering scale is determined. Then, the clustering scale with the largest selection index value is used as the target clustering scale. Finally, based on the land use clustering results at the target clustering scale, the land space planning data of the target area is classified, marked and stored. Since the final target clustering scale is determined based on the selection index value, it is more accurate and effective than the artificial setting method, and the clustering result has more reference value, thereby improving the rationality and accuracy of data storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 A method flow chart of a method for storing large-scale data of land space planning provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of the division principle of the block set; Figure 3 A system structure diagram of a large-scale data storage system for national land space planning provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation method, structure, features and effects of a large-scale data storage method and system for land space planning proposed by the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0018] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0019] Combine the following Figures 1 to 3 The specific scheme of a large-scale data storage method and system for national land space planning provided by the present invention is specifically described.

[0020] Please refer to Figure 1 , which shows the method flow chart of a large-scale data storage method for territorial space planning provided by an embodiment of the present invention. As Figure 1 shown, the above-mentioned large-scale data storage method for territorial space planning specifically includes: Step 110: Obtain the territorial space planning data of the target area.

[0021] In this embodiment, the territorial space planning data usually can represent the characteristic parameters of each position point in the spatial position, such as satellite remote sensing data, precipitation data, and temperature data, etc.

[0022] Step 120: Convert the territorial space planning data into a spatial data distribution map.

[0023] It can be understood that since the territorial space planning data has a strong association with the spatial position, therefore, the territorial space planning data can be converted into a form of image expression in combination with the spatial position conditions of different regions, that is, converted into a spatial data distribution map. For example, satellite remote sensing data can be converted into a remote sensing image, precipitation data can be converted into a precipitation distribution map, and temperature data can be converted into a temperature distribution map.

[0024] Step 130: Perform hierarchical clustering on each data point in the spatial data distribution map at multiple clustering scales to obtain the land use clustering results at each clustering scale.

[0025] It can be understood that the hierarchical clustering algorithm is an unsupervised learning algorithm for constructing a data hierarchical structure, mainly by calculating the similarity or distance between data points, gradually merging or splitting clusters to form a clustering tree with a hierarchical structure.

[0026] Taking the spatial data distribution map as a satellite remote sensing image as an example, since the territorial space planning data of different uses appears as different colors in the satellite remote sensing image, such as rivers, cultivated land, forests, etc. having different colors. Specifically, rivers appear as blue in the satellite remote sensing image, cultivated land as yellow, and forests as green. Therefore, the territorial space planning data of different uses in the target area can be represented by color information. Using the color information as the clustering index, the individual data points are thus hierarchically clustered. Similarly, if it is a precipitation distribution map, hierarchical clustering can be performed according to the precipitation information.

[0027] In this embodiment, the data points in the spatial data distribution map can be hierarchically clustered according to different uses at different clustering scales, so as to obtain the land use clustering results at different clustering scales.

[0028] Step 140: Determine the selection index value corresponding to each clustering scale according to the land use clustering results at each clustering scale.

[0029] It should be noted that the selection index value can represent the clustering accuracy at the corresponding clustering scale. The higher the selection index value, the more accurate the clustering result obtained at that clustering scale.

[0030] Step 150: Take the clustering scale with the largest selection index value as the target clustering scale.

[0031] In this embodiment, the clustering scale with the largest selection index value, that is, the optimal clustering scale, is taken as the target clustering scale, so that the optimal clustering scale can be reasonably selected according to the clustering accuracy situation.

[0032] Step 160: Based on the land use clustering results at the target clustering scale, classify and mark the national territorial space planning data of the target area and store it.

[0033] It can be understood that after obtaining the final land use clustering results, the distribution areas of the national territorial space planning data of different uses can be determined according to the land use clustering results, and then the national territorial space planning data of different uses can be classified and marked, realizing the reasonable storage of the national territorial space planning data.

[0034] In one embodiment, determining the selection index value corresponding to each clustering scale according to the land use clustering results at each clustering scale specifically includes: First, calculate the clustering accuracy of each cluster at each clustering scale according to the land use clustering results.

[0035] It can be understood that the clustering accuracy can represent the similarity of the uses of each data point under the corresponding cluster. The higher the clustering accuracy, the higher the similarity of the uses of the data points under that cluster, and the more accurate the division of the cluster.

[0036] Then, calculate the average value of the clustering accuracies of all clusters at each clustering scale to obtain the selection index value corresponding to each clustering scale.

[0037] In this embodiment, by calculating the average value of the clustering accuracies of all clusters at each clustering scale, the selection index value corresponding to each clustering scale is calculated, so that the selection index value can effectively represent the accuracy of the division of each cluster at that clustering scale.

[0038] In one embodiment, calculating the clustering accuracy of each cluster at each clustering scale according to the land use clustering results specifically includes: The first step is to divide the data points in each cluster in the land use clustering results to obtain multiple land blocks corresponding to each cluster.

[0039] In practical applications, since it is not clear how many categories the land in the target area should be divided into, different final cluster numbers will lead to different division results of the national territorial space planning data, and ultimately result in differences in the judgment of the true use of the land. Although there are some methods for adaptively determining the number of clustering clusters, such as the silhouette method, the elbow method, etc. However, since the land of the same use is not continuous in space, such as the division of land by rivers, the division of cultivated land and residential areas, there are large errors in the above methods for judging the number of clusters, which in turn affects the accuracy of the selected clustering scale.

[0040] Ideally, under a certain clustering scale, the national territorial space planning data divided into the same cluster belongs to the same use. Taking satellite remote sensing images as an example, different uses can be distinguished by colors. Therefore, the colors of the data points under the same cluster should be similar in order to be divided into the same cluster. Although the national territorial space planning data of different uses is not continuous in spatial position, this can only be reflected after reaching a certain scale. If strong discontinuity in spatial position is allowed, then not clustering in the hierarchical clustering tree is more accurate for the judgment and division of land uses. However, this obviously violates the purpose of reducing the storage scale of national territorial space planning data through hierarchical clustering. Therefore, this embodiment needs to ensure the spatial continuity of the land use clustering results while preventing the problem of excessive continuity. For this, this embodiment provides a description scheme for the spatial continuity characteristics of different clusters.

[0041] In this embodiment, a certain clustering scale in hierarchical clustering can be selected to obtain all the clusters divided under this clustering scale. Select one of the clusters, which represents the data set of the same use under this clustering scale. Mark these clusters on the spatial data distribution map to obtain the corresponding spatial positions of each cluster. Select a data point in this cluster as the analysis object and perform region growing in an 8-connected manner, requiring that the growing data must be data points of the same cluster as this data point to obtain the growing region.

[0042] After that, remove the data points that have been grown from this cluster, and use the same method as above for the remaining data points to obtain the corresponding growing regions. Repeat the above process to complete the regional division of each data point under this cluster. It can be understood that each growing region can represent the national territorial space planning data of the same use adjacent in space, representing the block division of the national territorial space planning data of this use in space, that is, the land block.

[0043] In the second step, determine the integrity of each land block in each cluster, and determine the characteristic difference value between each land block and any adjacent block.

[0044] It can be understood that the feature difference value can characterize the difference between this land block and adjacent blocks. For example, for satellite remote sensing images, the feature difference value can be characterized by the difference in color information between this land block and adjacent blocks.

[0045] In the third step, based on the integrity of each land block in each cluster, the initial accuracy value of each cluster is calculated.

[0046] In this embodiment, the initial accuracy value can characterize the possibility that all data points in this cluster belong to the same land use. The higher the initial accuracy value, the greater the possibility that all data in this cluster belong to the same land use.

[0047] In the fourth step, based on the integrity of each land block, the integrity of adjacent blocks, and the feature difference value between each land block and any adjacent block, the representative coefficient of each cluster is determined.

[0048] It can be understood that the representative coefficient can characterize the association of all land sub-blocks in this cluster. The larger the representative coefficient, the stronger the association of all land sub-blocks in this cluster.

[0049] In the fifth step, based on the representative coefficient, the initial accuracy value of each cluster is corrected to obtain the clustering accuracy of each cluster at each clustering scale.

[0050] In this embodiment, by correcting the initial accuracy value through the representative coefficient, a clustering accuracy with higher reliability can be obtained, so as to more accurately characterize the accuracy of cluster division at different clustering scales, and can provide an effective data basis for the selection of subsequent target clustering scales.

[0051] In an embodiment, determining the integrity of each land block in each cluster specifically includes: In the first step, in each cluster, multiple target blocks adjacent to each land block are determined.

[0052] In the second step, the block sets for spatially dividing each land block and each corresponding target block of each land block are respectively determined.

[0053] As Figure 2 shown, a land block A can be selected to construct the neighborhood of land block A in the same cluster. Specifically, the distances between land block A and the remaining land blocks in the same cluster are calculated, and the k remaining land blocks with the smallest distances are selected as the neighborhood of the current land block. By selecting a target block B in the neighborhood, the block set M passed by the minimum distance connection line L between the two land blocks can be obtained. Figure 2In it, A and B are two land blocks of the same cluster, and the straight line L is the minimum distance connection line; the land blocks 1 and 2 under other clusters passed by the minimum distance connection line L can form a block set M. The block set M represents the set of land blocks that spatially divide the land block A and the target block B.

[0054] If the land blocks under this cluster belong to the same national land use, the characteristic difference between the land block A and the target block B should be small, and different clusters represent lands of different uses, then the difference between the land block A and the block set M should be large, indicating that the land that spatially divides the land block A and the target block B belongs to the normal division of land of other uses.

[0055] Step 3: Obtain the first eigenvalue of each land block, the second eigenvalue of each target block, the third eigenvalue of any element block in each block set, and the number of all element blocks in each block set.

[0056] Step 4: Based on the reciprocal of the number of all element blocks in each block set, determine the calculation coefficient value corresponding to each target block.

[0057] Step 5: Based on the first eigenvalue of each land block, the second eigenvalue of each target block, and the third eigenvalue of any element block in each block set, calculate the sum value of the characteristic differences corresponding to each target block.

[0058] Step 6: Multiply the calculation coefficient value of each target block by the sum value of the characteristic differences to calculate the block integrity value of each land block relative to each target block.

[0059] In this embodiment, taking the land block A, the target block B, and the block set M as examples, the block integrity value of the land block A relative to the target block B can be calculated as follows: (1) Among them, represents the block integrity value of the land block A relative to the target block B; represents the first eigenvalue of the land block A, represents the second eigenvalue of the target block B, represents the third eigenvalue of an element block in the block set M. It should be noted that the first eigenvalue, the second eigenvalue, and the third eigenvalue can all be represented by the mean value of the eigenvalues of all data points in the corresponding block; represents the number of all element blocks in the block set M; c represents a non-zero fixed value, which is used to prevent the formula from being meaningless due to a denominator of 0.

[0060] In the seventh step, calculate the mean of all the block integrity values corresponding to each land block, and obtain the integrity degree of each land block in each cluster.

[0061] It can be understood that the smaller the difference between land block A and target block B, and the larger the difference between land block A and block set M, it indicates that under the current clustering scale, the spatial segmentation between land block A and target block B belongs to normal segmentation, and the higher the block integrity value.

[0062] If the spatial segmentation between land block A and the remaining land blocks in the same cluster all belongs to normal segmentation, the integrity of land block A is stronger. Based on this, in this embodiment, calculate the mean of all the block integrity values of land block A relative to the remaining target blocks in the same cluster, and use it as the integrity degree of land block A, denoted as .

[0063] In one embodiment, based on the integrity degree of each land block in each cluster, calculate the initial accuracy value of each cluster, specifically including: First, calculate the mean of the integrity degrees of each land block in each cluster to obtain the mean integrity degree of each cluster.

[0064] Then, calculate the standard deviation of the integrity degrees of each land block in each cluster to obtain the standard deviation of the integrity degree of each cluster.

[0065] Finally, divide the mean integrity degree of each cluster by the standard deviation of the integrity degree to calculate the initial accuracy value of each cluster.

[0066] In this embodiment, the initial accuracy value of the i-th cluster can be expressed as follows: (2) Among them, represents the initial accuracy value of the i-th cluster; respectively represent the mean integrity degree and the standard deviation of the integrity degree of all land blocks under the i-th cluster; d represents a non-zero fixed value, which is used to prevent the formula from being meaningless due to a denominator of 0.

[0067] It can be understood that the higher the mean integrity degree, the more it indicates that the spatial positions of all land blocks under this cluster are normally segmented; the smaller the standard deviation of the integrity degree, the more it indicates that these land blocks have higher segmentation accuracy, and thus the greater the possibility that this cluster represents the same national land use.

[0068] In one embodiment, based on the integrity degree of each land block, the integrity degree of adjacent blocks, and the characteristic difference value between each land block and any adjacent block, determine the representative coefficient of each cluster, specifically including: In the first step, based on the integrity of each land block, the integrity of adjacent blocks, and the characteristic difference value between each land block and any adjacent block, the block correlation degree between each land block and any adjacent block is calculated.

[0069] In practical applications, the initial accuracy value can preliminarily characterize the possibility that all data points under the same cluster have the same land use. However, the scale of hierarchical clustering may be too small, resulting in multiple clusters representing the same land use, thus affecting the effectiveness of national territorial space planning. Although the continuous feature analysis can, to a certain extent, consider the discontinuous problems caused by the small scale, the similarity comparison of features between each land block depends on the selection of the clustering center. In the actual clustering process, it is very easy to have a situation where the characteristic difference between the divided land block A and the element blocks in the block set M is small, while the difference between the clustering centers of the two clusters is large; or the difference between the clustering centers is small, but the characteristic difference between the land blocks in the two clusters is large. And because it is compared with all the element blocks in the block set M, it will cause some land blocks that should belong to the same use to be interfered by the remaining land blocks at the minimum distance and be regarded as a reasonable spatial segmentation.

[0070] In some applications, when the clustering scale is too small, there will be a problem of overly fine land division, such that the adjacent blocks of the land blocks under a certain cluster actually belong to the same use in national territorial space planning but are divided into different clusters. Therefore, when the clustering scale is too small, there will be a problem that adjacent land blocks with the same actual use are divided into different clusters.

[0071] Therefore, if there is a high correlation between a certain land block and the land blocks adjacent in spatial position, it indicates that although the adjacent positions are divided into different clusters, these land blocks actually represent the land of the same use, thus indicating that the current selection of the clustering scale is inaccurate and the accuracy of the actual cluster division result is low; the worse the correlation, the more it indicates that the adjacent land blocks in the spatial position are divided into different clusters to represent land of different uses. Therefore, a small correlation belongs to a normal phenomenon, and at this time, it indicates that the accuracy of the cluster division is higher.

[0072] Based on the above situation, this embodiment introduces a calculation link for the block correlation degree to characterize the correlation between each land block and any adjacent block through the block correlation degree.

[0073] In this embodiment, the land block and the adjacent block The characteristic difference value can be calculated as follows: (3) Among them, represents the land block and the adjacent block The characteristic difference value, for example, can be the color difference value; Are respectively represented as land blocks And adjacent blocks The characteristic values, for example, can be the color values of the blocks; Are respectively represented as land blocks The characteristic mean value of the cluster to which the land block belongs and the adjacent block The characteristic mean value of the said cluster.

[0074] Furthermore, the relevance of the land block And the adjacent block Can be expressed as follows: (4) Wherein, Represents the relevance of the land block And the adjacent block The relevance of the block, Represents the correlation coefficient between two objects, Represents the land block And the adjacent block The characteristic difference value of, Represents the integrity of the land block The integrity of, Represents the adjacent block The integrity of.

[0075] In the second step, based on the relevance of each land block to any adjacent block, the spatial weight of each land block to any adjacent block is calculated.

[0076] In a specific implementation, based on the relevance of each land block to any adjacent block, the spatial weight of each land block to any adjacent block is calculated, specifically including: First, for each land block and any adjacent block, subtract the relevance from 1 to obtain a first intermediate quantity.

[0077] Then, respectively determine the first integrity of the cluster where the land block is located and the second integrity of the cluster where the adjacent block is located.

[0078] Next, sum the first integrity and the second integrity to obtain a second intermediate quantity.

[0079] Finally, divide the first intermediate quantity by the second intermediate quantity to calculate the spatial weight of each land block to any adjacent block.

[0080] In this embodiment, the spatial weight of the land block And the adjacent block Can be specifically expressed as follows: (5) Among them, represents the land parcel and the adjacent parcels of the spatial weight, represents the land parcel where the first integrity of the cluster is located, represents the adjacent parcels where the second integrity of the cluster is located, represents the land parcel and the adjacent parcels of the block correlation degree.

[0081] Step 3: Based on the spatial weight, calculate the correlation index between each land parcel and any adjacent parcel.

[0082] In this embodiment, after obtaining the spatial weight, the calculation process of Moran's index can be used to calculate the correlation index between each land parcel and any adjacent parcel. This correlation index is between -1 and 1. It can be understood that Moran's index is a commonly used spatial statistical index for measuring the spatial autocorrelation between elements, that is, the degree of aggregation or dispersion of elements in space. In this embodiment, two land parcels can be used as two elements to be analyzed, and the Moran's index between the two land parcels can be further obtained by using the calculated spatial weight, and then the obtained Moran's index is used as the correlation index between the two land parcels.

[0083] Accordingly, there is a correlation index between each land parcel and each adjacent parcel. The larger the correlation index, the stronger the spatial correlation between the two land parcels, and the lower the accuracy of cluster division.

[0084] Step 4: Calculate the average value of all the correlation indexes under each cluster to obtain the representative coefficient of each cluster.

[0085] It can be understood that the representative coefficient can characterize the overall correlation situation between all the land parcels under this cluster.

[0086] In one embodiment, based on the representative coefficient, the initial accuracy value of each cluster is corrected to obtain the clustering accuracy of each cluster at each clustering scale, which specifically includes: First, add the representative coefficient to 1 to calculate the correction coefficient; Then, divide the initial accuracy value of each cluster by the correction coefficient and perform normalization processing to obtain the clustering accuracy of each cluster at each clustering scale.

[0087] In this embodiment, the clustering accuracy of the i-th cluster at the current clustering scale can be calculated as follows: (6) Among them, represents the clustering accuracy of the i-th cluster, represents the initial value of the accuracy of the i-th cluster, represents the representative coefficient of the i-th cluster, ( ) represents the normalization function.

[0088] After obtaining the clustering accuracy, the average value of the clustering accuracies of all clusters at each clustering scale is calculated to obtain the selection index value corresponding to each clustering scale. The target clustering scale corresponding to the maximum selection index value is selected, and the land use clustering result obtained by clustering at the target clustering scale is used as the final hierarchical clustering result, so as to provide an accurate classification basis for subsequent effective data storage.

[0089] In one embodiment, based on the land use clustering result at the target clustering scale, the national territorial space planning data of the target area is classified and marked and stored, specifically including: First, according to the land use clustering result at the target clustering scale, the national territorial space planning data belonging to the same cluster is classified and marked with the same cluster label.

[0090] Then, the cluster label and spatial location information of the national territorial space planning data after classification and marking are stored.

[0091] In practical applications, the national territorial space planning data belonging to the same cluster can be represented by the same cluster label, and different clusters use different cluster labels. During the process of storing the national territorial space planning data, only the location information needs to be recorded, and the specific data value is represented by the cluster label, so that it is not necessary to store the specific data value for each national territorial space planning data, thereby realizing the large-scale storage of national territorial space planning data.

[0092] Based on the same general inventive concept, the present invention also protects a large-scale data storage system for national territorial space planning. The large-scale data storage system for national territorial space planning provided by the present invention is described below, and the large-scale data storage system for national territorial space planning described below can be mutually referred to with the large-scale data storage method for national territorial space planning described above.

[0093] Please refer to Figure 3 , which shows the system structure diagram of a large-scale data storage system for national territorial space planning provided by an embodiment of the present invention. As Figure 3 shown, the above-mentioned large-scale data storage system for national territorial space planning specifically includes: An acquisition module 210, configured to acquire the national territorial space planning data of the target area.

[0094] A conversion module 220, configured to convert the national territorial space planning data into a spatial data distribution map.

[0095] The clustering module 230 is configured to perform hierarchical clustering on each data point in the spatial data distribution map at multiple clustering scales to obtain the land use clustering results at each clustering scale.

[0096] The calculation module 240 is configured to determine the selection index value corresponding to each clustering scale according to the land use clustering results at each clustering scale.

[0097] The processing module 250 is configured to use the clustering scale with the largest selection index value as the target clustering scale.

[0098] The storage module 260 is configured to classify and store the territorial spatial planning data of the target area based on the land use clustering results at the target clustering scale.

[0099] In the territorial spatial planning large-scale data storage system provided by the embodiments of the present invention, since the final target clustering scale is determined according to the selection index value, compared with the artificially set method, it is more accurate and effective, and the clustering results are more valuable for reference, thereby improving the rationality and accuracy of data storage.

[0100] Regarding the system in the above embodiments, the specific manners in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated herein.

[0101] It should be noted that: the above sequence of the embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0102] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A method for storing large-scale data of national land space planning, characterized in that: The method comprises: Obtain the land and space planning data of the target area; Converting the national land space planning data into a spatial data distribution map; Performing hierarchical clustering on each data point in the spatial data distribution map at multiple clustering scales to obtain a land use clustering result at each clustering scale; Performing regional division on the data points in each cluster in the land use clustering result to obtain a plurality of land blocks corresponding to each cluster; determining the integrity of each land block in each cluster, and determining the characteristic difference value between each land block and any adjacent block; calculating the initial accuracy value of each cluster based on the integrity of each land block in each cluster; determining the representative coefficient of each cluster based on the integrity of each land block, the integrity of adjacent blocks, and the characteristic difference value between each land block and any adjacent block; correcting the initial accuracy value of each cluster based on the representative coefficient to obtain the clustering accuracy of each cluster at each clustering scale; averaging the clustering accuracy of all clusters at each clustering scale to obtain the selection index value corresponding to each clustering scale; The clustering scale with the largest selection index value is used as the target clustering scale; Based on the land use clustering results at the target clustering scale, the national land space planning data of the target area is classified, marked and stored.

2. According to claim 1, a method for storing large-scale data of national land space planning is characterized in that: Determine the completeness of each land parcel within each cluster, including: In each cluster, a plurality of target blocks within a preset neighborhood of each land block are determined; Determine a set of blocks for spatially segmenting each land block and each target block corresponding to each land block; Obtaining a first characteristic value of each land block, a second characteristic value of each target block, a third characteristic value of any element block in each block set, and the number of all element blocks in each block set; Determine a calculation coefficient value corresponding to each target block based on the inverse of the number of all element blocks in each block set; Based on the first eigenvalue of each land block, the second eigenvalue of each target block, and the third eigenvalue of any element block in each block set, a sum of characteristic differences corresponding to each target block is calculated; Multiplying the calculation coefficient value of each target block by the sum of the characteristic differences to calculate the block integrity value of each land block relative to each target block; The average of all the block integrity values ​​corresponding to each land block is calculated to obtain the integrity of each land block in each cluster.

3. According to claim 1, a large-scale data storage method for national land space planning is characterized in that: Based on the integrity of each land block in each cluster, the initial accuracy value of each cluster is calculated, including: averaging the integrity of each land block in each cluster, and calculating the integrity mean of each cluster; Calculating the standard deviation of the integrity of each land block in each cluster, and obtaining the standard deviation of the integrity of each cluster; The initial accuracy value of each cluster is calculated by taking the quotient of the integrity mean value and the integrity standard deviation of each cluster.

4. According to claim 1, a large-scale data storage method for national land space planning is characterized in that: Based on the integrity of each land block, the integrity of the adjacent blocks, and the characteristic difference value between each land block and any adjacent block, the representative coefficient of each cluster is determined, including: Based on the integrity of each land block, the integrity of the adjacent blocks, and the characteristic difference value between each land block and any adjacent block, the block correlation between each land block and any adjacent block is calculated; Based on the block correlation between each land block and any adjacent block, the spatial weight of each land block and any adjacent block is calculated; Based on the spatial weight, a correlation index between each land block and any adjacent block is calculated; The average of all the association indexes under each type of cluster is calculated to obtain the representative coefficient of each type of cluster.

5. A method for storing large-scale data of national land space planning according to claim 4, characterized in that: Based on the block correlation between each land block and any adjacent block, the spatial weight of each land block and any adjacent block is calculated, including: For each land block and any adjacent block, subtract 1 from the block correlation to obtain a first intermediate value; respectively determining a first completeness of the cluster where the land block is located and a second completeness of the cluster where the adjacent block is located; Summing the first integrity and the second integrity to obtain a second intermediate quantity; The first intermediate quantity is divided by the second intermediate quantity to calculate the spatial weight of each land block and any adjacent block.

6. The method for storing large-scale data of national land space planning according to claim 1, characterized in that: Based on the representative coefficient, the initial accuracy value of each cluster is corrected to obtain the clustering accuracy of each cluster at each clustering scale, including: Adding the representative coefficient to 1, and calculating the correction coefficient; The initial accuracy value of each cluster is divided by the correction coefficient and normalized to obtain the clustering accuracy of each cluster at each clustering scale.

7. The method for storing large-scale data of national land space planning according to claim 1, characterized in that: Based on the land use clustering results at the target clustering scale, the land space planning data of the target area is classified, marked and stored, including: According to the land use clustering results at the target clustering scale, the land space planning data belonging to the same cluster are classified and marked using the same cluster label; The cluster number and spatial location information of the classified and marked national land space planning data are stored.

8. A large-scale data storage system for national land space planning, characterized in that: The system comprises: The acquisition module is used to obtain the land space planning data of the target area; A conversion module, used to convert the national land space planning data into a spatial data distribution map; A clustering module, used to perform hierarchical clustering on each data point in the spatial data distribution map at multiple clustering scales to obtain a land use clustering result at each clustering scale; A calculation module is used to divide the data points in each cluster in the land use clustering result into regions to obtain multiple land blocks corresponding to each cluster; determine the integrity of each land block in each cluster, and determine the characteristic difference value between each land block and any adjacent block; based on the integrity of each land block in each cluster, calculate the initial accuracy value of each cluster; based on the integrity of each land block, the integrity of adjacent blocks and the characteristic difference value between each land block and any adjacent block, determine the representative coefficient of each cluster; based on the representative coefficient, correct the initial accuracy value of each cluster to obtain the clustering accuracy of each cluster at each clustering scale; average the clustering accuracy of all clusters at each clustering scale to obtain the selection index value corresponding to each clustering scale; A processing module, used for taking the clustering scale with the largest selection index value as the target clustering scale; The storage module is used to classify, mark and store the land space planning data of the target area based on the land use clustering results at the target clustering scale.

Images