Land planning auxiliary decision-making method and system based on multi-source data

Abstract

The application discloses a land planning auxiliary decision-making method and system based on multi-source data, and particularly relates to the field of land planning auxiliary decision-making, which comprises the following steps: acquiring multi-source land records of a target region and unifying a spatial reference, constructing a constraint triangulated network with source boundaries, mapping synonymous attribute values in a triangular face to land propositions, determining the quality of the propositions according to the proportion of covered area, determining the quality of unknown propositions according to the proportion of uncovered area, and obtaining a triangular face evidence set; performing knowledge fusion on the triangular face evidence set, substituting land propositions of different sources into a sixth proportional conflict redistribution algorithm according to the source order, and adding the quality product of intersecting land propositions to the intersection proposition; dividing the multi-source land records into constraint triangular faces and performing knowledge fusion, respectively performing land index allocation and project scheduling under the conditions of a reference interpretation and a replacement interpretation, generating a configuration conflict value and a segmentation line according to the difference in resource allocation, and merging configuration consistent regions into land units.

Description

Technical Field

[0001] This invention relates to the field of land planning auxiliary decision-making technology, and more specifically, to a land planning auxiliary decision-making method and system based on multi-source data. Background Technology

[0002] Existing land planning auxiliary decision-making methods mostly use parcels, map patches, or regular grids as the basic calculation objects. They unify coordinates, correspond fields, and spatially overlay data such as cadastral surveys, current land use status, planning control, remote sensing identification, infrastructure carrying capacity, and project applications to screen construction areas, calculate land use indicators, and determine project locations.

[0003] For the allocation of construction land near the urban development boundary, the same candidate area often crosses the boundaries of the land parcel, the current land category, the ecological control boundary, the road service area, and the proposed project land boundary. At the same time, it must also meet the conditions that the total amount of construction land quota cannot be exceeded, the ecological control area cannot be occupied, and the necessary functional areas of the project cannot be split.

[0004] Since the basic calculation objects are determined before the multi-source data fusion, local boundary differences will directly affect the entire parcel or raster, causing candidate areas with limited local scope to be excluded as a whole, or the limited scope to be not identified separately because it is calculated together with the surrounding areas; if all source boundaries are superimposed at once, a large number of fragmented map patches will be formed that do not affect the index allocation and project ranking, increasing the amount of subsequent calculations, while still not being able to determine which specific boundary difference has really changed the land allocation result;

[0005] The technical problem this application aims to solve is: how to determine the specific location where the land planning calculation object needs to be spatially segmented based on the differences in results formed by multi-source land data in indicator allocation, project access, and implementation ranking. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a land planning auxiliary decision-making method and system based on multi-source data. By dividing multi-source land records into constrained triangular surfaces and performing knowledge fusion, land index allocation and project scheduling are carried out under benchmark interpretation and alternative interpretation respectively. Configuration conflict values ​​and dividing lines are generated according to resource allocation differences, and then the areas with consistent configurations are merged into land units to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a land planning auxiliary decision-making method based on multi-source data, comprising:

[0008] S1. Obtain multi-source land records of the target area and unify the spatial benchmark. Construct a constrained triangular network based on the source boundary. Map synonymous attribute values ​​in the triangular facet to land propositions. Determine the quality of propositions according to the coverage area ratio and determine the quality of unknown propositions according to the uncovered area ratio to obtain the triangular facet evidence set.

[0009] S2. Perform knowledge fusion on the triangular evidence set, substitute land propositions from different sources into the sixth proportional conflict redistribution algorithm in the order of their sources, include the mass product of intersecting land propositions into the intersection proposition, and write back the mass product of non-intersecting land propositions according to the proportion of their respective mass to the sum of the masses of the two, and sum them up to obtain the fused proposition and fused mass.

[0010] S3. The first fusion proposition of each triangular face forms the benchmark interpretation, and the remaining fusion propositions are replaced one by one to form the alternative interpretation. Under each interpretation, the land use demand is read in the order of implementation. Triangular faces with matching uses and whose land indicators and infrastructure capacity after deduction are not negative are occupied in the order of overlapping area to obtain the resource allocation record.

[0011] S4. Compare the resource configuration records of the baseline interpretation and the alternative interpretation, accumulate the fusion quality corresponding to the alternative interpretation with different configurations as the configuration conflict value, generate vertex values ​​according to the average configuration conflict value of adjacent triangles, perform the discrete Morse-Small segmentation algorithm on the vertex values, and generate segmentation lines along the discrete gradient path passing through the triangles with non-zero configuration conflict values.

[0012] S5. Divide the land parcels by dividing lines, merge adjacent triangular faces that are not separated by dividing lines and have the same configuration into land units, repeat resource allocation and division until the new dividing line coincides with the existing boundary, use the land units with zero configuration conflict value for land index allocation and project scheduling, remove the remaining land units from the configuration and sort and verify them, and output the land planning auxiliary decision-making results.

[0013] In a preferred embodiment, S1 includes:

[0014] S11. Convert the multi-source land records to the unified spatial reference of the target area, disconnect the source boundary at the endpoints and intersections of the source boundary, and connect adjacent disconnections according to the order of their positions on the source boundary to obtain the constraint boundary map.

[0015] Multi-source land records refer to areal spatial records that originate from cadastral surveys, land use status surveys, planning control, remote sensing identification, infrastructure carrying capacity, and project application data. Each multi-source land record must include at least the source identifier, record number, formation time, code table version, spatial range, and attribute code.

[0016] The process of converting to a unified spatial reference for the target area involves reading the coordinate reference identifiers, coordinate axis order, length units, and boundary vertex coordinates of each multi-source land record; performing axis order adjustment, unit conversion, and coordinate transformation on the boundary vertices in sequence according to the coordinate transformation parameters registered between the source spatial reference and the target spatial reference; and then connecting the transformed boundary vertices in the original vertex order to obtain the multi-source land records under the unified spatial reference.

[0017] The source boundary refers to the planar geometric boundary that defines the spatial range in a multi-source land record, including the outer ring vertices and inner ring vertices connected in the order of the records; after the source boundary is converted to a unified spatial reference, it consists of line segments between adjacent boundary vertices.

[0018] S12. Read the vertices of each closed region in the constrained edge graph, select three adjacent vertices along the boundary of the closed region, divide the three vertices whose interior angle of the middle vertex faces the closed region and whose line connecting the two vertices is located in the closed region into triangular faces, delete the middle vertex and repeat the division to obtain the constrained triangular network, and write the triangular face number and triangular face area to each triangular face.

[0019] A closed region refers to a region in a constrained edge graph that is formed by connecting the beginning and end of constrained edges, whose boundaries do not intersect themselves, and whose interior does not contain constrained edges. The inner loop in the constrained edge graph is connected to the corresponding outer loop through a connecting edge that does not cross the constrained edge, and the vertices of the closed region are read according to a single closed boundary.

[0020] In a preferred embodiment, S1 further includes:

[0021] S13. Convert the attribute codes of each multi-source land record into a unified land attribute code, and calculate the overlapping area between the multi-source land record and each triangular face according to the source. When there are overlapping records in the same source, retain the record with the later formation time. When the formation time is the same, retain the record with the earlier record number. Then merge the overlapping areas with the same unified land attribute code into a land proposition area.

[0022] The attribute code refers to the discrete code recorded in the corresponding code table used by the source data. It is used to indicate the land use, land use status, planning control status or infrastructure carrying capacity within the spatial range of multi-source land records, and together with the source identifier and code table version, it determines the meaning of the corresponding attribute.

[0023] The process of converting to a unified land attribute code involves forming a mapping key using the source identifier, code table version, and attribute code, and reading the corresponding unified land attribute code from the unified attribute mapping table. When the same mapping key has multiple effective intervals, the effective interval containing the formation time of multiple land records is selected. If no corresponding item is found, an unmapped attribute code is written, and the unmapped attribute code is prohibited from participating in the merging of synonymous attribute values.

[0024] S14. Divide the area of ​​the land proposition area by the area of ​​the corresponding triangle to obtain the proposition quality. Divide the remaining area after deducting the area of ​​all land proposition areas from the triangle area by the triangle area to obtain the unknown proposition quality. Write the triangle number, land proposition, proposition quality, unknown proposition quality and triangle area into the triangle evidence set for use in land quota allocation and project scheduling.

[0025] In a preferred embodiment, S2 includes:

[0026] S21. Construct a source alignment graph using the land propositions and proposition quality from each source in the triangular face evidence set. Establish transmission edges only between semantically intersecting land propositions. Use the sum of semantic inclusion level difference and coverage area difference as the transmission cost. Iterate and allocate proposition quality through entropy regularization optimal transmission. In each round, normalize the row quality and column quality of each transmission edge until the row quality and column quality are equal to the proposition quality of the corresponding source and the non-zero transmission edges in two adjacent rounds are the same, thus obtaining the source alignment table.

[0027] The process of constructing the source alignment graph is as follows: taking any two sources within the same triangular face as a group, setting the land propositions with non-zero proposition quality in the first source as the left node, and setting the land propositions with non-zero proposition quality in the second source as the right node, and writing the source number, land proposition, proposition quality, and coverage area to each node; establishing a transmission edge between semantically intersecting left and right nodes; for land propositions without semantically intersecting nodes, connecting them to the unknown proposition node of another source, and writing the transmission cost and transmission quality to the transmission edge, thus obtaining a bipartite source alignment graph;

[0028] Semantic intersection refers to the fact that two land propositions correspond to the same unified land attribute code, or that one unified land attribute code belongs to the lower code of the other unified land attribute code in the unified attribute hierarchy tree, so that the land attribute range represented by the two land propositions has a common part; unknown propositions do not participate in semantic intersection determination, but are only used to accept the quality of propositions that cannot be corresponding to land propositions from another source.

[0029] The difference between the level and the difference in the coverage area refer to the absolute value of the difference in the number of levels of two semantically intersecting unified land attribute codes in the unified attribute hierarchy tree, divided by the total number of levels in the hierarchy tree. When the unified land attribute codes are the same, it is recorded as zero. The difference in the coverage area refers to the absolute value of the difference in the ratio obtained by dividing the coverage area of ​​two land propositions in the same triangle by the area of ​​the triangle.

[0030] The process of allocating proposition quality through entropy regularized optimal transmission iteratively involves generating transmission weights based on the transmission cost of each transmission edge, with lower costs resulting in higher values. The product of the transmission weight and the proposition quality of the corresponding left and right nodes is written as the initial value of the transmission quality. Subsequently, the transmission quality of associated transmission edges is scaled row by row according to the left node, so that the sum of the transmission quality of each row equals the proposition quality of the corresponding left node. Then, the transmission quality of associated transmission edges is scaled column by column according to the right node, so that the sum of the transmission quality of each column equals the proposition quality of the corresponding right node. This row scaling and column scaling are repeated. The proposition quality is converted into an integer quality share according to a unified area unit of measurement. The iteration stops when the transmission quality share of each row and each column is equal to the proposition quality share of the corresponding node, and the sets of transmission edge numbers with non-zero transmission quality in two adjacent rounds are consistent. The transmission quality of each transmission edge is then written into the source alignment table.

[0031] S22. Read the sources in ascending order of transmission cost according to the source alignment table. Execute the sixth proportional conflict redistribution algorithm on the aligned land propositions. Include the quality product of semantically intersecting land propositions in the intersection proposition. Write back the quality product of semantically non-intersecting land propositions according to the proportion of the quality of each land proposition allocated through the source alignment table to the sum of the quality of the two. Add up the written-back quality of the same land proposition to obtain the initial fusion proposition and the initial fusion quality.

[0032] The process of executing the sixth proportional conflict redistribution algorithm is as follows: According to the source order determined by the source alignment table, the current fusion result is combined with the next source item by item; when the semantics of two land propositions intersect, the product of the quality obtained by the two is accumulated and added to the corresponding intersection proposition; when the semantics of two land propositions do not intersect, the product of the quality obtained by the two is recorded as the conflict quality, and then the conflict quality is multiplied by the proportion of the quality obtained by each to the sum of the quality obtained by the two, and written back to the corresponding land proposition; the intersection quality and the written back quality obtained by the same land proposition are accumulated respectively; when an unknown proposition is combined with any land proposition, the quality product is included in the land proposition, and the source number and written value of each quality are recorded until the combination of all sources is completed, and the initial fusion proposition and the initial fusion quality are obtained.

[0033] In a preferred embodiment, S2 further includes:

[0034] S23. Map each initial fusion proposition to the acceptance status of whether or not land occupation indicators are allowed. After removing one source in sequence, re-execute S21 and S22. When the acceptance status of the first initial fusion proposition before and after removing the source is different, transfer the quality written to the corresponding first initial fusion proposition by the removed source to the unknown proposition. Repeat the removal of sources and quality transfer until removing any source does not change the acceptance status of the first initial fusion proposition or all unstable qualities have been transferred to the unknown proposition, and obtain the corrected fusion proposition and corrected fusion quality.

[0035] The process of mapping the acceptance status of whether or not land quotas are allowed to be occupied is as follows: the first initial fusion proposition is determined according to the descending order of the initial fusion quality and the ascending order of the land proposition code. The unified land attribute code corresponding to the first initial fusion proposition is read, and the land quota acceptance table is queried with the unified land attribute code and the current planning period number. When the land quota acceptance table records that the unified land attribute code can accept the land quotas of the current planning period, it is written as allowed to accept. When it records that acceptance is prohibited or no corresponding record is found, it is written as prohibited to accept. Unknown propositions are written as pending verification and do not occupy land quotas.

[0036] S24. Verify that the sum of all corrected fusion quality and unknown proposition quality within the same triangular face is one. Write the difference generated when the verification is not equal to one into the unknown proposition, and output the fusion proposition and fusion quality in descending order of corrected fusion quality and ascending order of land proposition code.

[0037] In a preferred embodiment, S3 includes:

[0038] S31. Read the first fusion proposition of each triangle face in descending order of fusion quality and ascending order of fusion proposition coding to form the baseline interpretation. Replace the first fusion proposition of the other fusion propositions of the same triangle face in turn to form the replacement interpretation. Record the replaced triangle face corresponding to each replacement interpretation to obtain the interpretation table.

[0039] S32. Under the benchmark interpretation, read the land use requirements according to the implementation order, connect the projects with matching uses to the triangular facets, and use the overlapping area, unoccupied area of ​​the triangular facets, land quota acceptance area and infrastructure capacity acceptance area within the proposed land use area of ​​the project as the connection capacity after deducting them respectively, and each positive increment will make at least one of the deductions negative. Read the connection relationship in descending order of overlapping area and ascending order of triangular facet number, and use the connection capacity and the remaining requirements of the project as the allocation area after deducting them respectively, and simultaneously deduct the remaining requirements of the project, land quota and infrastructure capacity to obtain the benchmark resource allocation record.

[0040] Positive increment refers to an area share corresponding to a unified area measurement unit; when the connection capacity or allocation area increases by an area share, and any of the following deduction results becomes negative: overlapping area, unoccupied area of ​​triangular face, land quota acceptance area, infrastructure capacity acceptance area, or remaining project demand, the increase will stop and the value before the increase will be determined as the connection capacity or allocation area.

[0041] S33. For each replacement interpretation, find the first project that the replaced triangle can participate in, retain the baseline resource allocation record before the corresponding project, withdraw the baseline resource allocation record of the corresponding project and all subsequent projects and return the corresponding land indicators and infrastructure capacity, update the connection relationship of the replaced triangle according to the replacement fusion proposition, and then repeatedly allocate land use requirements from the corresponding project to obtain the resource allocation record corresponding to each replacement interpretation.

[0042] In a preferred embodiment, S4 includes:

[0043] S41. According to the triangular face number, compare the resource allocation records of the benchmark interpretation and each alternative interpretation. Record the difference of the undertaken projects as the project difference, divide the absolute value of the allocated area difference by the corresponding project land use requirement as the area difference, divide the absolute value of the implementation order difference by the number of projects as the order difference, multiply the sum of the project difference, area difference and order difference by the fusion quality of the corresponding alternative interpretation, and accumulate them as the configuration conflict value of the corresponding triangular face to obtain the configuration conflict table.

[0044] The configuration conflict value refers to the cumulative amount of resource allocation changes that occur on the same triangular face under the benchmark interpretation and various alternative interpretations. When one of the benchmark interpretation and the alternative interpretation does not record the undertaking project, the allocated area of ​​the unrecorded party is recorded as zero and the project difference is recorded as one. When both parties undertake the same project, the project difference is recorded as zero. Then, the sum of the project difference, area difference and order difference is multiplied by the fusion quality of the corresponding alternative interpretation and accumulated according to the alternative interpretation. The resulting value is used to represent the degree of impact of the change in the fusion proposition on the land index allocation and project implementation scheduling of the triangular face.

[0045] S42. Read the configuration conflict value and triangle area of ​​each vertex's adjacent triangle face. Multiply each configuration conflict value by the corresponding triangle face area and sum them up. Then divide by the sum of the adjacent triangle face areas to obtain the vertex value. Establish the lower star sequence according to the vertex value in ascending order, vertex number in ascending order, and simplex dimension in ascending order to obtain the discrete scalar field.

[0046] The simplex dimension is the integer obtained by subtracting one from the number of independent vertices in the simplex of a constrained triangular network. The simplex dimension of a vertex is zero, the simplex dimension of an edge connecting two vertices is one, and the simplex dimension of a triangle formed by three edges is two.

[0047] In a preferred embodiment, S4 further includes:

[0048] S43. Perform the Discrete Morse-Small Segmentation Algorithm on the discrete scalar field. Pair the unpaired simplex in the lower star sequence with the unpaired upper simplex that is in the order of the first simplex and contains the corresponding simplex. The simplex that cannot be paired is recorded as the critical simplex. Connect the critical simplex along the pairing relationship to form a discrete gradient path. If there is only one discrete gradient path between a pair of critical simplexes and the triangular faces on both sides of the corresponding path have the same project, area allocation and implementation order in all resource allocation records, cancel this pair of critical simplexes and reconnect the corresponding pairing relationship. Repeat the cancellation until there are no critical simplexes that meet the cancellation conditions to obtain the preserved gradient path.

[0049] The process of executing the Discrete Morse-SmithKline (DMS) algorithm is as follows: A vertex total order is formed by ascending vertex values ​​and vertex numbers. For each vertex, all simplexes with that vertex as the last vertex in the total order are read and formed into a corresponding lower star. The simplexes within the lower star are arranged in ascending order of simplex dimension and simplex number. Each unpaired simplex is read sequentially. Among the unpaired simplexes containing the current simplex and whose simplex dimension increases by one, the one with the higher order is paired. If no corresponding simplex exists, the current simplex is recorded as a critical simplex. Starting from the critical simplex, the paired simplexes are alternately connected to their common faces until another critical simplex or the boundary of the cadastral parcel is reached, forming a discrete gradient path. When there is only one discrete gradient path between a pair of critical simplexes and all resource configuration records of the triangular faces on both sides of the path are consistent, the pair of critical simplexes and their path pairing are cancelled, and adjacent simplexes are reconnected according to the original pairing directions at both ends of the cancellation position until no more critical simplexes can be cancelled.

[0050] A simplex is a basic geometric object in a constrained triangulation network that participates in the discrete Morse-Smith subdivision operation. It includes vertices, edges formed by connecting two vertices, and triangular faces enclosed by three vertices and three edges. Each simplex has a simplex number and maintains an inclusion relationship with the constrained triangulation network.

[0051] S44. Read the paths that pass through the triangular facets with non-zero configuration conflict values ​​from the retained gradient paths, delete the path segments that enter the same side and then return to the corresponding side, and verify that there is at least one different resource configuration record on both sides of the path. Connect the paths that pass the verification to the critical simplex or cadastral land boundary along the boundary of the constrained triangular network to obtain the dividing line used to distinguish the land index allocation and project implementation scheduling results.

[0052] The boundary of a cadastral parcel refers to the legal boundary line in the cadastral survey data that is identified by the parcel number and encloses the spatial range of the parcel. It consists of an outer ring boundary connected in sequence according to the boundary points and an inner ring boundary corresponding to the deduction area inside the parcel. It is used to limit the extension range of the dividing line within the corresponding parcel.

[0053] In a preferred embodiment, S5 includes:

[0054] S51. Divide the corresponding cadastral parcels by dividing lines and cadastral parcel boundaries. Establish adjacency relationships for triangular faces that share complete edges and are not separated by dividing lines. Group the adjacent triangular faces that have the same project number, project implementation order, land quota account, and infrastructure service scope in all resource allocation records into the same connected group. Accumulate the allocated area, land quota deduction, and infrastructure capacity deduction within the connected group to obtain the land unit and unit allocation record.

[0055] A shared complete edge refers to two triangular faces that have the same two endpoint numbers, and the edge of the triangular mesh formed by connecting these two endpoints serves as a component edge of both triangular faces; triangular faces that only have their vertices connected, their edges partially overlap, or are truncated by dividing lines are not considered shared complete edges.

[0056] S52. Replace the corresponding cadastral parcel with land units, retain resource allocation records that are not affected by changes in land units, reallocate land use demand and regenerate dividing lines starting from the first affected project, and write the newly added dividing lines into the boundary set according to the triangular mesh edge numbers and the start and end positions within the edges. If there are dividing lines in the boundary set that have not been recorded in the existing land unit boundaries, execute S51 and this step until all dividing lines in the boundary set have been recorded in the existing land unit boundaries, and obtain a stable land unit set.

[0057] S53. Centralize the stable land units and deduct land indicators and infrastructure capacity according to the unit configuration record for land units with zero conflict value, and write them into the project implementation order. Cancel the corresponding resource occupation for land units with non-zero conflict value, generate the verification order according to the descending order of the configuration conflict value and the ascending order of the land unit number, and output the land planning auxiliary decision-making results.

[0058] A land planning auxiliary decision-making system based on multi-source data, comprising an evidence construction module, a knowledge fusion module, a resource allocation module, a conflict segmentation module, and a decision output module.

[0059] The evidence construction module is used to acquire multi-source land records of the target area and unify the spatial benchmark. It constructs a constrained triangular network based on the source boundary, maps synonymous attribute values ​​within the triangular facets to land propositions, determines the quality of propositions based on the coverage area ratio, and determines the quality of unknown propositions based on the uncovered area ratio, thus obtaining the triangular facet evidence set.

[0060] The knowledge fusion module is used to perform knowledge fusion on the triangular evidence set. It substitutes land propositions from different sources into the sixth proportional conflict redistribution algorithm in the order of their sources, adds the mass product of intersecting land propositions to the intersection proposition, and writes back the mass product of non-intersecting land propositions according to the proportion of their respective mass to the sum of the masses of the two, and accumulates them to obtain the fused proposition and fused mass.

[0061] The resource allocation module is used to form a baseline interpretation by merging the first and second fusion propositions of each triangular face, and to form a replacement interpretation by replacing the remaining fusion propositions one by one. Under each interpretation, the land use demand is read in the order of implementation. Triangular faces with matching uses and whose land indicators and infrastructure capacity after deduction are not negative are occupied in the order of overlapping area to obtain the resource allocation record.

[0062] The conflict segmentation module is used to compare the resource configuration records of the baseline interpretation and the alternative interpretation, accumulate the fusion quality corresponding to the alternative interpretation with different configurations as the configuration conflict value, generate vertex values ​​according to the average configuration conflict value of adjacent triangles, perform the discrete Morse-Small segmentation algorithm on the vertex values, and generate segmentation lines along the discrete gradient path passing through the triangles with non-zero configuration conflict values.

[0063] The decision output module is used to divide the cadastral land parcels by dividing lines, merge adjacent triangular faces that are not separated by dividing lines and have the same configuration into land units, repeat resource allocation and division until the newly added dividing line coincides with the existing boundary, use the land units with zero configuration conflict value for land index allocation and project scheduling, remove the remaining land units from the configuration and sort and verify them, and output the land planning auxiliary decision results.

[0064] The technical effects and advantages of this invention are as follows:

[0065] 1. This solution compares the resource allocation differences between the baseline interpretation and the alternative interpretation and forms configuration conflict values, which can locate the spatial location that truly changes the allocation of indicators and the order of projects, and reduce the segmentation caused by irrelevant boundaries;

[0066] 2. Constructing the multi-source boundary into a constrained triangular network and generating land propositions and proposition quality according to the triangular faces can limit the scope of the effect of local attribute differences and alleviate the problem of exclusion of the entire parcel due to local constraints;

[0067] 3. By performing source alignment and conflict quality redistribution on land propositions from different sources, the differences and quality contributions of each source can be preserved, reducing the direct impact of single-source bias on project access results;

[0068] 4. Extracting discrete gradient paths and generating dividing lines within regions with non-zero conflict values ​​can reduce fragmented patches unrelated to resource configuration, thereby reducing the processing load for subsequent indicator calculations and project scheduling.

[0069] 5. Land units are formed by adjacent triangular faces with consistent resource allocation and not separated by dividing lines. Only the affected projects are recalculated, which can preserve the original allocation results of stable areas and reduce duplicate allocation and full-area recalculation. Attached Figure Description

[0070] Figure 1This is a flowchart of the land planning auxiliary decision-making method of the present invention;

[0071] Figure 2 This is a diagram of the land planning auxiliary decision-making system of the present invention. Detailed Implementation

[0072] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0073] Refer to the instruction manual appendix Figure 1-2 The land planning auxiliary decision-making method based on multi-source data of the present invention includes:

[0074] S1. Obtain multi-source land records of the target area and unify the spatial benchmark. Construct a constrained triangular network based on the source boundary. Map synonymous attribute values ​​in the triangular facet to land propositions. Determine the quality of propositions according to the coverage area ratio and determine the quality of unknown propositions according to the uncovered area ratio to obtain the triangular facet evidence set.

[0075] To ensure that data from land ownership surveys, land use status surveys, planning control, remote sensing identification, infrastructure carrying capacity, and project application are incorporated into a unified spatial unit, this implementation method sequentially completes spatial normalization, boundary network construction, attribute merging, and quality writing, thereby establishing a definite correspondence between land-related issues and their sources, attribute dimensions, and triangular facets. This implementation process includes the following steps:

[0076] Before each multi-source land record enters geometric processing, the source identifier, record number, formation time, code table version, spatial range, attribute code, coordinate reference identifier, coordinate axis order, length unit, and boundary vertex coordinates are read.

[0077] Linear and point-based infrastructure records are generated into areal spatial records according to the registered service width, service zone, service radius, or service boundary, respectively. Records without registered service ranges are only associated with the infrastructure capacity account.

[0078] Spatial normalization is performed according to S11. Based on the transformation parameters registered between the source spatial reference and the target spatial reference, the coordinate axis order is adjusted, the length unit is converted, and the boundary vertex coordinates are transformed. Then, the vertices are organized according to the unified coordinate precision.

[0079] Records with missing coordinate reference identifiers are written into the verification table, duplicate vertices are merged, unclosed boundaries are patched with start and end edges, and self-intersecting boundaries are extracted after being broken at the intersection point. Records with zero area and zero length edges are no longer processed.

[0080] Then, break the edges at the boundary endpoints, intersections, endpoint locations, and endpoints of collinear overlapping edges. Vertices in the same position share the same number, overlapping edges are divided into segments and the source identifier is retained. Connect adjacent break points according to the original boundary order to obtain the constrained edge graph.

[0081] After the constrained edge graph is formed, the closed region with no self-intersections on the boundary and no constrained edges inside is extracted. The outer ring is read in a counterclockwise direction, and the inner ring is read in a clockwise direction.

[0082] When an inner loop exists, select the vertex with the first vertex number in the inner loop and connect it to the vertex with the first vertex number in the outer loop whose line does not cross the constraint edge.

[0083] Triangulation is performed according to S12. The three adjacent vertices are read sequentially along the closed region. A triangular face is formed when the interior angle of the middle vertex faces the inside of the region and the line connecting the two vertices is within the region and does not cross the constraint edge. The middle vertex is deleted and the triangulation continues.

[0084] Multiple candidate vertices are read according to vertex number. Collinear three vertices are deleted, and the middle vertex is not formed into a triangle. Triangles with zero area are written into the degeneracy record.

[0085] The partitioning is completed when only three vertices remain in the closed region, and the triangle identifier and triangle area are written; regions that cannot be partitioned are written to the geometry anomaly table.

[0086] Attribute processing uses source identifier, code table version and attribute code as mapping key, reads attribute dimension code and unified land attribute code from unified attribute mapping table, and uses independent attribute dimension code for land use, land use status, planning control status and infrastructure carrying capacity status respectively.

[0087] S13 first determines the effective mapping based on the record formation time. If the mapping intervals overlap or no mapping item is found, it is written to the unmapped record table. Then, the overlapping area between multi-source land records and triangular faces is obtained according to the source and attribute dimensions.

[0088] When records from the same source overlap, only the record with the later creation time is used within the overlapping area; if the creation times are the same, the record with the earlier record number is used; the original records are retained in the remaining areas.

[0089] Overlapping areas with identical source identifiers, attribute dimension codes, and unified land attribute codes are merged into land proposition areas, and areas corresponding to unmapped attribute codes are transferred to unknown proposition areas.

[0090] After completing the land proposition area sorting, the single source, single triangle facet, and single attribute dimension are used as the quality calculation unit; S14 calculates the proposition quality according to the ratio of the land proposition area area to the triangle facet area, and calculates the unknown proposition quality by subtracting the union area of ​​the land proposition area from the triangle facet area; areas without mapped attribute codes, uncovered areas, and areas with indeterminate attributes are all included in the unknown proposition area.

[0091] Areas are uniformly converted to integer area shares, and the conversion remainder is written into the unknown proposition; when the area of ​​the union of land proposition regions exceeds the area of ​​the triangular face, the corresponding source, triangular face, and attribute dimension are written into the area anomaly table.

[0092] After verifying that the sum of all proposition quality and unknown proposition quality is one, the source identifier, triangle identifier, attribute dimension code, land proposition, proposition quality, unknown proposition quality, and triangle area are written into the triangle evidence set;

[0093] Taking the area near the urban development boundary as an example, the land parcels, current land categories, planning control, remote sensing identification, water supply service area and the proposed land use area of ​​the project are first converted to the same spatial benchmark, and the water supply pipeline generates a planar spatial record according to the registered service zone; after the boundary intersections, endpoints and some overlapping places are uniformly cut off, a constraint edge map and a constraint triangulation are formed.

[0094] Current status survey codes and remote sensing identification codes are written into the land use status dimension, and planning control codes are written into the planning control status dimension.

[0095] When a construction land proposition from the same source forms 70% of the area within a certain triangle, and the remaining area does not form a valid mapping, the quality of the construction land proposition is recorded as 70%, the quality of the unknown proposition is recorded as 30%, and other attribute dimensions are calculated separately, thus forming a triangular evidence set for subsequent knowledge integration, land quota allocation, and project scheduling.

[0096] S2. Perform knowledge fusion on the triangular evidence set, substitute land propositions from different sources into the sixth proportional conflict redistribution algorithm in the order of their sources, include the mass product of intersecting land propositions into the intersection proposition, and write back the mass product of non-intersecting land propositions according to the proportion of their respective mass to the sum of the masses of the two, and sum them up to obtain the fused proposition and fused mass.

[0097] Regarding the semantic differences and quality allocation relationships of land propositions from different sources within the same triangular facet, this implementation method first completes source pairing within the same attribute dimension, then determines the fusion order according to the pairing results and performs conflict quality write-back, and then checks the land index acceptance status of the first fused proposition by eliminating sources one by one, finally forming a fused proposition with quality conservation and traceable source effects; the specific processing is as follows;

[0098] When entering source pairing, any two sources within the same triangle face and the same attribute dimension form a group. S21 sets the non-zero land proposition of the previous source as the left node and the non-zero land proposition of the subsequent source as the right node, and sets unknown proposition nodes respectively.

[0099] Nodes with the same unified land attribute code or with hierarchical relationships establish transmission edges; when ordinary land propositions do not have semantically intersecting nodes, they only connect to unknown proposition nodes on the other side; unknown proposition nodes on both sides are connected to each other and respectively connected to nodes on the opposite side that have not obtained semantic correspondence;

[0100] The semantic content difference is obtained by dividing the difference in the number of levels of the two unified land attribute codes in the unified attribute hierarchy tree by the total number of levels, and the coverage area difference is obtained by the absolute value of the difference in the coverage ratio of the two propositions. The two are added together to form the transmission cost.

[0101] The negative value of the transmission cost is divided by the registered entropy regularization coefficient, and the resulting value is taken as the transmission weight. The entropy regularization coefficient is determined by the total number of attribute hierarchy tree levels and the total number of uniform quality shares in the planning period configuration table.

[0102] The transmission weight is multiplied by the proposition quality of the left and right nodes to form the initial value of the transmission quality. Then, each row and column is alternately scaled to make the row quality and column quality match the proposition quality of the corresponding node.

[0103] After converting real-valued transmission quality into integer quality shares, missing shares in rows and columns are added according to the transmission edge number, and excess shares are transferred out. Each correction eliminates one row or column surplus until all row and column shares match and the sets of non-zero transmission edge numbers in two adjacent rounds are consistent. Then, the transmission edge, transmission cost, and transmission quality are written into the source alignment table.

[0104] After source pairing is completed, the transmission cost of each pair of sources is multiplied by the corresponding transmission quality and summed to obtain the source alignment cost. The source with the earlier source number is taken as the starting source, and the remaining sources are added in ascending order of source alignment cost with the already read source set. If the costs are the same, they are added in ascending order of source number.

[0105] S22 performs the recursive sixth proportional conflict redistribution in this order: the current fusion result is combined with the non-zero land propositions from the next source item by item. When the semantics intersect, the mass product is included in the intersection proposition. When the unified land attribute code has a hierarchical relationship, the lower code is used as the intersection proposition. When the semantics do not intersect, the mass product is recorded as the conflict mass, and then written back according to the proportion of the mass obtained by the two participating propositions in the sum of their masses.

[0106] The quality obtained by combining an unknown proposition with a land proposition is included in the land proposition; the quality obtained by combining two unknown propositions is included in the unknown proposition.

[0107] Each intersection quality and write-back quality records the source set, participating proposition coding, and written value. The quality of the same land proposition is accumulated and used as the current fusion result when the next source is added, until all sources are processed, and the initial fusion proposition and initial fusion quality are obtained.

[0108] To verify whether a single source changes the land quota acceptance result, S23 first determines the first initial fusion proposition according to the initial fusion quality in descending order and the land proposition code in ascending order, and then queries the land quota acceptance table with its unified land attribute code and the current planning period number, and writes the status as allowed acceptance, prohibited acceptance, or pending verification.

[0109] Then, the evidence set is copied and recalculated from the complete triangular evidence set. Only one source is deleted at a time, and source pairing and conflict reallocation are re-executed. After the source verification is completed, the recalculated evidence set is discarded, and the next source is verified from the complete triangular evidence set.

[0110] When the initial fusion proposition's state differs before and after source removal, the source set is read from the quality source record, containing only the write quality of the removed source. This quality is then registered as unstable and transferred to an unknown proposition. If the source set contains the quality of multiple sources, the quality difference of the same proposition before and after removal is written to the joint change record.

[0111] After all source verifications are completed, the next round begins with the transferred quality. The round stops when no new unstable quality is generated in a single round of verification, or when the writing quality of each single source has been transferred to an unknown proposition, and the corrected fusion proposition and corrected fusion quality are obtained.

[0112] After completing the source removal correction, S24 summarizes the quality of all corrections and fusion and the quality of unknown propositions by source, triangular facet, and attribute dimension respectively;

[0113] When the total mass is missing an integer mass share, the missing share will be added to the unknown proposition; when the total mass exceeds an integer mass share, the share added to the unknown proposition in this round will be deducted first. If there is still an excess share, the merged mass will be deducted and corrected in reverse order of the land proposition code until the total mass is restored to one.

[0114] After quality conservation verification, the fusion proposition and fusion quality are written in descending order of corrected fusion quality and ascending order of land proposition coding, and the source set, joint change record and unknown proposition quality are retained for reference interpretation, substitution interpretation and land index acceptance calculation.

[0115] Taking the arrangement of urban construction land as an example, when the same triangular facet obtains the propositions of cadastral survey, current status survey and remote sensing identification in the land use dimension at the same time, three sets of source alignment relationships are first formed based on the unified land attribute code level and coverage area, and then the recursive fusion order is determined according to the source alignment cost.

[0116] When a conflict arises between a remote sensing identification proposition and a land ownership survey proposition, the conflict quality is written back to the two land propositions involved in the conflict. After removing the sources one by one, if removing the remote sensing identification source causes the first fusion proposition to change from an allowed acceptance state to a prohibited acceptance state, then only the quality written to the first fusion proposition by that source is transferred to the unknown proposition.

[0117] The integrated propositions formed after multiple rounds of testing can maintain a fixed total quality and clearly distinguish between the effects of single and combined sources, providing verifiable inputs for subsequent land quota allocation and project implementation scheduling.

[0118] S3. The first fusion proposition of each triangular face forms the benchmark interpretation, and the remaining fusion propositions are replaced one by one to form the alternative interpretation. Under each interpretation, the land use demand is read in the order of implementation. Triangular faces with matching uses and whose land indicators and infrastructure capacity after deduction are not negative are occupied in the order of overlapping area to obtain the resource allocation record.

[0119] To transform the uncertainty of the fusion proposition into comparable land quota allocation and project scheduling results, this implementation method first establishes an interpretation table containing only single proposition substitutions, then forms a baseline resource allocation record in an independent resource ledger, and subsequently recalculates each substitution interpretation starting from the first affected project, ensuring that resource changes under different interpretations have the same calculation caliber; the specific processing is as follows:

[0120] The explanation table is used to fix the spatial location and quality source of each proposition change, and to avoid configuration differences that cannot be attributed due to the simultaneous changes of multiple triangular faces;

[0121] The fusion propositions within each triangle are read in descending order of fusion quality. If the fusion quality is the same, they are read in ascending order of fusion proposition codes. S31 writes the first fusion proposition into the baseline interpretation. The remaining fusion propositions of the same triangle replace the first fusion proposition. Each replacement interpretation contains only one replaced triangle and one replacement fusion proposition. The remaining triangles follow the baseline interpretation.

[0122] When replacing a fusion proposition with an unknown proposition, the replaced triangle will be written to a pending verification state and no project connection will be established.

[0123] The interpretation table stores the interpretation identifier, the replaced triangle, the first and second fusion propositions, the replacement fusion propositions, and the replacement fusion quality, which are read for resource allocation and configuration conflict calculation.

[0124] The benchmark resource allocation, through the standardization of usage, area, and resource account definition, assigns project requirements to triangular faces with the necessary conditions for undertaking them.

[0125] Before implementing S32, the initial balances of the formal land quota account and infrastructure capacity account are copied to form the benchmark interpretation resource account; land use demand is read in ascending order according to the implementation order, and when the implementation order is the same, it is read in ascending order according to the project number, and the project use acceptance table is queried by the unified land attribute code of the project use code and the first and second digits of the triangular facet.

[0126] When a record of permitted acceptance is found and the proposed land area of ​​the project overlaps with the triangular face, a connection is established. The infrastructure acceptance area is obtained by dividing the remaining capacity of each required facility by the capacity consumption per unit land area of ​​the project. Facilities not required by the project are not included in the calculation.

[0127] For each connection, the area share is increased one by one from zero until adding another area share will cause the overlapping area, the unoccupied area of ​​the triangular face, the corresponding land quota balance, or the balance of any required infrastructure capacity to become negative. The value before the increase is written as the connection capacity. When the connection capacity is zero, no allocable connection is written.

[0128] Each connection is read in descending order of overlapping area and ascending order of triangular face number, and then the allocated area is increased one by one until the next area share will cause the connection capacity margin or the remaining demand of the project to become negative.

[0129] When the allocated area is non-zero, the baseline resource allocation record is written with the interpretation identifier, triangle number and project number as keys, and the unoccupied area of ​​the triangle, the remaining demand of the project, the corresponding land quota account and the capacity accounts of each facility are deducted simultaneously.

[0130] Multiple project configuration records can be generated on the same triangular face; if there are still remaining needs after all connections of a project have been read, projects with phased identifiers retain the allocated area and record the remaining needs, while projects without phased identifiers withdraw all configuration records of the project and return the corresponding resources.

[0131] The replacement interpretation recalculation is used to limit the scope of resource adjustments, so that previous irrelevant configurations can be preserved, while covering two types of changes in the triangle face: from allowed to prohibited and from prohibited to allowed.

[0132] For each replacement interpretation, S33 compares the usage matching results, connection capacity, and allocated area of ​​the replaced triangular face under the baseline interpretation and the replacement interpretation sequentially from the starting point of the project implementation sequence. The project that first shows a difference is identified as the affected project; if no difference is found, the baseline resource configuration record is directly copied.

[0133] When there are affected projects, retain the baseline resource allocation records before their implementation sequence, withdraw the allocation records of the project and subsequent projects one by one, and return the land quotas, infrastructure capacity and unoccupied triangular area according to the land quota account number, facility number and allocated area in the original records.

[0134] The replacement fusion proposition is then written into the replaced triangle, its acceptance status and connection relationship are redefined, and then the allocation continues from the affected project according to the same project order, connection order, capacity calculation and insufficient acceptance rules as the baseline interpretation, to obtain the resource allocation record of the replacement interpretation.

[0135] For example, if the first fusion proposition of a certain triangle allows industrial projects to be undertaken, and the second fusion proposition belongs to the prohibited construction attribute, the benchmark interpretation first allocates the industrial project area to the triangle in the resource account copy; after the alternative interpretation is formed, the subsequent allocation is withdrawn from the first project that has a difference in use matching or spatial overlap with the triangle, and the land quota and facility capacity are returned to the original account, and then redistributed according to the alternative undertaking status.

[0136] Therefore, only the resource configuration differences caused by the replaced fusion proposition are retained among the interpretations, providing a traceable baseline resource configuration record and replacement resource configuration record for subsequent configuration conflict value calculation.

[0137] S4. Compare the resource configuration records of the baseline interpretation and the alternative interpretation, accumulate the fusion quality corresponding to the alternative interpretation with different configurations as the configuration conflict value, generate vertex values ​​according to the average configuration conflict value of adjacent triangles, perform the discrete Morse-Small segmentation algorithm on the vertex values, and generate segmentation lines along the discrete gradient path passing through the triangles with non-zero configuration conflict values.

[0138] The differences in land allocation results caused by the changes in the fusion proposition need to be first converted into triangular face allocation conflict values, and then the dividing lines are generated by using the topological relationship of the constrained triangular network.

[0139] This implementation method sequentially completes resource allocation record comparison, vertex value propagation, simplex pairing, critical path screening and boundary processing, so that the dividing line can locate the spatial location where land index allocation and project implementation scheduling have changed.

[0140] First, the resource configuration records in the baseline interpretation and each alternative interpretation are grouped by triangular face number. When the same triangular face corresponds to multiple project records, the union of the project numbers under the two interpretations is read, and each project number is compared one by one.

[0141] Configuration difference quantification is performed in S41: when a project exists in both interpretations, the project difference is recorded as zero; when a project exists only on one side, the project difference is recorded as one, the area allocated to the missing side is recorded as zero, and the implementation order is recorded as the total number of projects plus one.

[0142] The absolute value of the allocated area difference is divided by the original land requirement of the project to obtain the area difference position, and the absolute value of the implementation order difference is divided by the total number of projects to obtain the order difference position; projects with a required area of ​​zero are written into the requirement exception table and do not participate in the configuration conflict calculation.

[0143] After the three types of differences of each project are accumulated, the result is multiplied by the fusion quality of the replacement fusion proposition in the replaced triangle, and the resulting value is accumulated to the triangle to form the configuration conflict value. The replacement interpretation identifier, project number set, various types of differences and configuration conflict value are written into the configuration conflict table.

[0144] When the configuration conflict value of the triangle face is propagated to the vertices of the constrained triangular mesh, the configuration conflict value and the area of ​​the triangle face adjacent to each vertex are read, the two are multiplied and summed and then divided by the sum of the areas of the adjacent triangle faces to obtain the vertex value of that vertex.

[0145] Vertices not adjacent to valid triangles are written into the isolated vertex table and do not enter subsequent segmentation; S42 further establishes a vertex total order according to the vertex value in ascending order and the vertex number in ascending order. The edge number is composed of the two endpoint numbers in ascending order, and the triangle number is composed of the three vertex numbers in ascending order.

[0146] Each simplex is assigned to the lower star of the last vertex in the total order of its contained vertices. Within the same lower star, they are further arranged according to the simplex dimension and simplex number, where the simplex dimensions of vertices, edges and triangular faces are zero, one and two respectively, thus obtaining a discrete scalar field with a unique reading order.

[0147] After entering the topological decomposition, the unpaired simplexes in each lower star are read in a predetermined order, and the unpaired upper-dimensional simplexes with the first order position, one increased dimension and containing the current simplex are selected for pairing within the same lower star.

[0148] S43 stipulates that each simplex can only participate in pairing once; when a candidate pairing would cause the discrete gradient path to repeatedly pass through an existing pairing, the candidate is discarded and the next candidate is read; when all candidates are unusable, the current simplex is recorded as the critical simplex.

[0149] Then, starting from the critical simplex, the path alternates between low-dimensional simplexes, paired high-dimensional simplexes, and common surfaces until it reaches another critical simplex or the boundary of the cadastral parcel, forming a discrete gradient path.

[0150] The cancellation test is performed only on critical simplexes with a dimension difference of one. The path traversal confirms that there is only one discrete gradient path between the two, and the project number sequence, allocated area and implementation order of the triangular faces on the left and right sides of the path are compared side by side.

[0151] When all records are consistent, cancel the critical simplex pair, reverse the original pairing direction along the unique path, so that each simplex still maintains a unique pairing;

[0152] Each undo operation reduces two critical simplexes, and this process is repeated until no critical simplexes satisfy the conditions are found, thus obtaining a gradient-preserving path.

[0153] After the gradient path is formed, S44 reads the path that passes through the triangles with non-zero conflict values ​​and determines the triangles on the left and right sides of each directed edge according to the path travel direction; when the same edge appears adjacent in opposite directions, the pair of folded edge segments are deleted and the paths before and after are reconnected.

[0154] When the path passes through the interior of the triangle, splitting vertices are generated at the entry and exit edges. Then, two candidate edge chains are read along the boundary of the triangle. The edge chain with the earlier edge number sequence is selected and written into the splitting path. When the end of the path has not yet reached the critical simplex or the boundary of the cadastral parcel, the adjacent edge that does not return to the previous edge and has the earlier edge number sequence is read.

[0155] The resource configuration records on both sides of the path are verified segment by segment. If there are differences in project number, allocated area or implementation order, they are retained. The path generates an intersection point and disconnects at the intersection. Line segments with overlapping positions share the same dividing edge number. Suspended line segments with consistent configuration on both sides are deleted. Finally, a dividing line without repeated edge segments is formed and written into the constrained triangulation network.

[0156] Taking a land parcel that simultaneously undertakes industrial projects and public facility projects as an example, a certain triangular facet is allocated to an industrial project under the baseline interpretation and converted to a public facility project under the alternative interpretation. The project difference, area difference, and order difference of this triangular facet are weighted by the alternative fusion quality to form a non-zero configuration conflict value. After this value is propagated to the adjacent vertices and participates in simplex pairing, the preserved gradient path passing through this triangular facet is converted into a dividing line on the boundary of the constrained triangular network.

[0157] Through the above processing, the differences in resource allocation can be transformed into spatial boundaries with clear endpoints, segments, and lateral configuration relationships, providing definite input for subsequent cadastral parcel division and land unit generation.

[0158] S5. Divide the cadastral land parcels with dividing lines, merge adjacent triangular faces that are not separated by dividing lines and have the same configuration into land units, repeat resource allocation and division until the new dividing line coincides with the existing boundary, use the land units with zero configuration conflict value for land index allocation and project scheduling, remove the remaining land units from the configuration and sort and verify them, and output the land planning auxiliary decision-making results.

[0159] To transform the configuration conflict boundary into a spatial object that can directly participate in land quota allocation and project implementation scheduling, this implementation method first writes the dividing line into a constrained triangular network and generates land units, then recalculates the affected projects based on the change range of the land units, and finally writes the stable configuration into the formal resource account to form a land planning auxiliary decision-making result.

[0160] When the segmentation results are used to generate land units, the intersection points of the segmentation lines and the edges of the triangular mesh are first calculated, and the segmentation vertices are generated at the intersection points.

[0161] When the dividing line passes through the interior of the triangle, the original triangle is redivided into sub-triangles along the dividing line. The sub-triangles inherit the land proposition, integration quality and resource allocation records of the original triangle. The original allocated area is split according to the ratio of the area of ​​the sub-triangle to the area of ​​the original triangle. The remaining area is written to the sub-triangle with the highest area.

[0162] After the dividing line is entered into the network, S51 recognizes the shared complete edge as the edge of the triangular network that has the same two endpoint numbers and belongs to two triangular faces at the same time. Triangular faces that are cut off by the dividing line, only have vertices connected, or only have some edge segments overlapping are not established as adjacency relationships.

[0163] For the remaining adjacent triangles, the resource allocation records are arranged according to the project implementation order and project number, and the project number, allocated area, land quota account number and facility service scope number are compared in turn. When all are consistent, they are grouped into the same connectivity group.

[0164] Subsequently, the land quota deduction amount is accumulated according to the land quota account number, and the infrastructure capacity deduction amount is accumulated according to the facility number. The project allocation area is then summarized to form a land unit, a set of bottom triangular face numbers, and a unit configuration record.

[0165] Before a land unit participates in reallocation, a resource dependency table is established between the land unit and the project, land quota account, and facility capacity account.

[0166] S52 replaces the original cadastral parcel with the newly formed land unit as the allocation object, and retains the resource allocation records that do not spatially overlap with the changed land unit, do not call the same land indicator account, and do not call the same facility account;

[0167] For the remaining projects, starting from the earlier ones in the implementation sequence, the configuration records in the explanation resource account will be withdrawn one by one, and the land quotas, infrastructure capacity and unoccupied area of ​​land units will be returned according to the original records, and then reconfigured according to the S3 rules of use matching, connection capacity and allocation area.

[0168] The regenerated dividing lines are recorded according to the triangulation edge numbers. The positions within the edges are uniformly adopted using the integer length share calculated from the earlier endpoint number. The start and end positions are written into the boundary set in ascending numerical order.

[0169] When there is a boundary code that is not covered by the existing land unit boundary, land unit generation and resource recalculation are performed again;

[0170] At least one new dividing edge is added during each round of processing. The number of available edge segments in the constrained triangulation is fixed. The process stops after all newly added dividing lines are covered by the boundaries of existing land units, resulting in a stable set of land units.

[0171] After a stable land unit is formed, the configuration conflict value of each triangular face within the unit is multiplied by the area of ​​the corresponding triangular face and accumulated. Then, it is divided by the area of ​​the land unit to obtain the configuration conflict value of the land unit.

[0172] When the configuration conflict values ​​of all triangular faces within a unit are zero, the configuration conflict value of that land unit is recorded as zero.

[0173] The formal accounting is executed by S53, which reads the initial balance from the land quota account and infrastructure capacity account, reads the unit configuration records with zero configuration conflict value according to the project implementation order and land unit number, verifies that all resource deductions are not negative, and then writes the allocated area, land quota deduction amount, facility capacity deduction amount and project implementation order into the formal resource account.

[0174] Land units with non-zero configuration conflict values ​​are not written into the formal resource account, and the verification order is generated according to the descending order of configuration conflict values ​​and the ascending order of land unit numbers;

[0175] When a project is jointly undertaken by multiple land units, the land units with a total configuration conflict value of zero are allocated area; when a project is not marked as allowing phased implementation and the total area does not meet the land use requirements, all formal resource occupation of the project is withdrawn; when marking as allowing phased implementation, the corresponding implementation period is written according to the registered phased area, and the remaining area is retained as the remaining requirements of the project.

[0176] Taking a certain land parcel as an example, which is divided into three land units by two dividing lines, two of the land units undertake the same industrial project under various alternative interpretations, call the same land indicator account and water supply facility account, and have the same allocated area record. The two are merged into a connected group; the project undertaken by the third land unit changes with the fusion proposition, and its land unit configuration conflict value is non-zero. Therefore, it is removed from the formal configuration and enters the verification sequence.

[0177] After the resource recalculation is completed, the project area undertaken by the first two land units meets the land use requirements, and the relevant land indicators and facility capacity are written into the formal resource account. This results in a land planning auxiliary decision-making result with clear land unit boundaries, project implementation order, resource deduction records, and verification order.

[0178] exist Figure 1 It should be noted that: the multi-source land records sequentially go through spatial benchmark unification and triangular facet evidence construction, land proposition knowledge integration, resource allocation under benchmark interpretation and alternative interpretation, allocation conflict calculation and dividing line generation, and then the cadastral parcels are divided to form land units; the solid arrows indicate the data transfer and execution order between each processing stage, and the negative arrow of the judgment node returns to the resource allocation stage, indicating that if the newly added dividing line does not coincide with the existing boundary, the resource allocation and segmentation will continue to be executed, and the land planning auxiliary decision-making results will be output after the coincidence is confirmed.

[0179] exist Figure 2 It should be noted that the system is mainly composed of an evidence construction module, a knowledge fusion module, a resource allocation module, a conflict segmentation module, and a decision output module. Multi-source land data enters from the input end and sequentially forms a triangular evidence set, fusion proposition, resource allocation record, configuration conflict value, dividing line, and land unit. Solid lines with arrows indicate the main data flow and calling order between modules, while dashed lines indicate that the unified attribute mapping table, land index account, and infrastructure capacity account provide rules or resource data to relevant modules. Dashed lines without arrows between the center and each module indicate that the modules together form a land planning auxiliary decision-making system.

[0180] Furthermore, the present invention also includes a land planning auxiliary decision-making system based on multi-source data, the system comprising an evidence construction module, a knowledge fusion module, a resource allocation module, a conflict segmentation module, and a decision output module:

[0181] The evidence construction module is used to acquire multi-source land records of the target area and unify the spatial benchmark. It constructs a constrained triangular network based on the source boundary, maps synonymous attribute values ​​within the triangular facets to land propositions, determines the quality of propositions based on the coverage area ratio, and determines the quality of unknown propositions based on the uncovered area ratio, thus obtaining the triangular facet evidence set.

[0182] The knowledge fusion module is used to perform knowledge fusion on the triangular evidence set. It substitutes land propositions from different sources into the sixth proportional conflict redistribution algorithm in the order of their sources, adds the mass product of intersecting land propositions to the intersection proposition, and writes back the mass product of non-intersecting land propositions according to the proportion of their respective mass to the sum of the masses of the two, and accumulates them to obtain the fused proposition and fused mass.

[0183] The resource allocation module is used to form a baseline interpretation by merging the first and second fusion propositions of each triangular face, and to form a replacement interpretation by replacing the remaining fusion propositions one by one. Under each interpretation, the land use demand is read in the order of implementation. Triangular faces with matching uses and whose land indicators and infrastructure capacity after deduction are not negative are occupied in the order of overlapping area to obtain the resource allocation record.

[0184] The conflict segmentation module is used to compare the resource configuration records of the baseline interpretation and the alternative interpretation, accumulate the fusion quality corresponding to the alternative interpretation with different configurations as the configuration conflict value, generate vertex values ​​according to the average configuration conflict value of adjacent triangles, perform the discrete Morse-Small segmentation algorithm on the vertex values, and generate segmentation lines along the discrete gradient path passing through the triangles with non-zero configuration conflict values.

[0185] The decision output module is used to divide the cadastral land parcels by dividing lines, merge adjacent triangular faces that are not separated by dividing lines and have the same configuration into land units, repeat resource allocation and division until the newly added dividing line coincides with the existing boundary, use the land units with zero configuration conflict value for land index allocation and project scheduling, remove the remaining land units from the configuration and sort and verify them, and output the land planning auxiliary decision results.

[0186] Overall Implementation and Explanation: This plan first unifies the data on land ownership survey, current land use, planning control, remote sensing identification, infrastructure carrying capacity, and project application to the same spatial benchmark, and divides the data into triangular faces according to the source boundaries, converting the records from each source into land propositions and proposition quality;

[0187] Subsequently, land information from different sources is integrated through source alignment and the sixth proportional conflict redistribution algorithm to form a baseline interpretation and an alternative interpretation. Land indicators and infrastructure capacity are then allocated sequentially under each interpretation. The system then compares the differences in project acceptance, allocated area and implementation order generated by different interpretations, calculates the configuration conflict value, uses the discrete Morse-Small segmentation algorithm to determine the configuration change boundary, and then segments the cadastral parcels and merges adjacent triangular faces with consistent configurations.

[0188] After resource recalculation and boundary iteration, land units with stable configurations are entered into project scheduling, while land units with conflicts are entered into the verification sequence, thus forming the land planning auxiliary decision-making results.

[0189] For example, a certain planning area may simultaneously have land parcels with registered ownership, current land type, planned use, remote sensing identification results, water and power supply capacity, and multiple construction project applications;

[0190] This solution first overlays these data into a unified triangular network to determine the land use and credibility of each triangular facet. Then, it calculates the projects and resource usage that the triangular facet can undertake when using the primary land proposition and other land propositions.

[0191] If certain triangular faces consistently undertake the same project under different interpretations and the resource deduction results are consistent, they are merged to form stable land units, and land indicators and project implementation order are arranged. If certain locations may experience project changes, area adjustments, or changes in scheduling order due to differences in data sources, a dividing line is formed along the configuration differences, and the relevant land units are temporarily removed from the configuration and verified first to avoid disputed land information directly affecting project arrangements.

[0192] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A land planning auxiliary decision-making method based on multi-source data, characterized in that, include: S1. Obtain multi-source land records of the target area and unify the spatial benchmark. Construct a constrained triangular network based on the source boundary. Map synonymous attribute values ​​in the triangular facet to land propositions. Determine the quality of propositions according to the coverage area ratio and determine the quality of unknown propositions according to the uncovered area ratio to obtain the triangular facet evidence set. S2. Perform knowledge fusion on the triangular evidence set, substitute land propositions from different sources into the sixth proportional conflict redistribution algorithm in the order of their sources, include the mass product of intersecting land propositions into the intersection proposition, and write back the mass product of non-intersecting land propositions according to the proportion of their respective mass to the sum of the masses of the two, and sum them up to obtain the fused proposition and fused mass. S3. The first fusion proposition of each triangular face forms the benchmark interpretation, and the remaining fusion propositions are replaced one by one to form the alternative interpretation. Under each interpretation, the land use demand is read in the order of implementation. Triangular faces with matching uses and whose land indicators and infrastructure capacity after deduction are not negative are occupied in the order of overlapping area to obtain the resource allocation record. S4. Compare the resource configuration records of the baseline interpretation and the alternative interpretation, accumulate the fusion quality corresponding to the alternative interpretation with different configurations as the configuration conflict value, generate vertex values ​​according to the average configuration conflict value of adjacent triangles, perform the discrete Morse-Small segmentation algorithm on the vertex values, and generate segmentation lines along the discrete gradient path passing through the triangles with non-zero configuration conflict values. S5. Divide the land parcels by dividing lines, merge adjacent triangular faces that are not separated by dividing lines and have the same configuration into land units, repeat resource allocation and division until the new dividing line coincides with the existing boundary, use the land units with zero configuration conflict value for land index allocation and project scheduling, remove the remaining land units from the configuration and sort and verify them, and output the land planning auxiliary decision-making results.

2. The land planning auxiliary decision-making method based on multi-source data according to claim 1, characterized in that: S1 includes: S11. Convert the multi-source land records to the unified spatial reference of the target area, disconnect the source boundary at the endpoints and intersections of the source boundary, and connect adjacent disconnections according to the order of their positions on the source boundary to obtain the constraint boundary map. S12. Read the vertices of each closed region in the constrained edge graph, select three adjacent vertices along the boundary of the closed region in sequence, divide the three vertices with the interior angle of the middle vertex facing the closed region and the line connecting the two vertices in the closed region into triangular faces, delete the middle vertex and repeat the division to obtain the constrained triangular network, and write the triangular face number and triangular face area to each triangular face.

3. The land planning auxiliary decision-making method based on multi-source data according to claim 2, characterized in that: S1 also includes: S13. Convert the attribute codes of each multi-source land record into a unified land attribute code, and calculate the overlapping area between the multi-source land record and each triangular face according to the source. When there are overlapping records in the same source, retain the record with the later formation time. When the formation time is the same, retain the record with the earlier record number. Then merge the overlapping areas with the same unified land attribute code into a land proposition area. S14. Divide the area of ​​the land proposition region by the area of ​​the corresponding triangle to obtain the proposition quality. Divide the remaining area after deducting the area of ​​all land proposition regions from the triangle area by the triangle area to obtain the unknown proposition quality. Write the triangle number, land proposition, proposition quality, unknown proposition quality and triangle area into the triangle evidence set.

4. The land planning auxiliary decision-making method based on multi-source data according to claim 3, characterized in that: S2 includes: S21. Construct a source alignment graph using the land propositions and proposition quality from each source in the triangular face evidence set. Establish transmission edges only between semantically intersecting land propositions. Use the sum of semantic inclusion level difference and coverage area difference as the transmission cost. Iterate and allocate proposition quality through entropy regularization optimal transmission. In each round, normalize the row quality and column quality of each transmission edge until the row quality and column quality are equal to the proposition quality of the corresponding source and the non-zero transmission edges in two adjacent rounds are the same, thus obtaining the source alignment table. S22. Read the sources in ascending order of transmission cost according to the source alignment table. Execute the sixth proportional conflict redistribution algorithm on the aligned land propositions. Include the quality product of semantically intersecting land propositions in the intersection proposition. Write back the quality product of semantically non-intersecting land propositions according to the proportion of the quality of each land proposition allocated through the source alignment table to the sum of the quality of the two. Add up the written-back quality of the same land proposition to obtain the initial fusion proposition and the initial fusion quality.

5. The land planning auxiliary decision-making method based on multi-source data according to claim 4, characterized in that: S2 also includes: S23. Map each initial fusion proposition to the acceptance status of whether or not land occupation indicators are allowed. After removing one source in sequence, re-execute S21 and S22. When the acceptance status of the first initial fusion proposition before and after removing the source is different, transfer the quality written to the corresponding first initial fusion proposition by the removed source to the unknown proposition. Repeat the removal of sources and quality transfer until removing any source does not change the acceptance status of the first initial fusion proposition or all unstable qualities have been transferred to the unknown proposition, and obtain the corrected fusion proposition and corrected fusion quality. S24. Verify that the sum of all corrected fusion quality and unknown proposition quality within the same triangular face is one. Write the difference generated when the verification is not equal to one into the unknown proposition, and output the fusion proposition and fusion quality in descending order of corrected fusion quality and ascending order of land proposition code.

6. The land planning auxiliary decision-making method based on multi-source data according to claim 5, characterized in that: S3 includes: S31. Read the first fusion proposition of each triangle face in descending order of fusion quality and ascending order of fusion proposition coding to form the baseline interpretation. Replace the first fusion proposition of the other fusion propositions of the same triangle face in turn to form the replacement interpretation. Record the replaced triangle face corresponding to each replacement interpretation to obtain the interpretation table. S32. Under the benchmark interpretation, read the land use requirements according to the implementation order, connect the projects with matching uses to the triangular facets, and use the overlapping area, unoccupied area of ​​the triangular facets, land quota acceptance area and infrastructure capacity acceptance area within the proposed land use area of ​​the project as the connection capacity after deducting them respectively, and each positive increment will make at least one of the deductions negative. Read the connection relationship in descending order of overlapping area and ascending order of triangular facet number, and use the connection capacity and the remaining requirements of the project as the allocation area after deducting them respectively, and simultaneously deduct the remaining requirements of the project, land quota and infrastructure capacity to obtain the benchmark resource allocation record. S33. For each replacement interpretation, find the first project that the replaced triangle can participate in, retain the baseline resource allocation record before the corresponding project, withdraw the baseline resource allocation record of the corresponding project and all subsequent projects and return the corresponding land indicators and infrastructure capacity, update the connection relationship of the replaced triangle according to the replacement fusion proposition, and then repeatedly allocate land use requirements from the corresponding project to obtain the resource allocation record corresponding to each replacement interpretation.

7. The land planning auxiliary decision-making method based on multi-source data according to claim 6, characterized in that: S4 includes: S41. According to the triangular face number, compare the resource allocation records of the benchmark interpretation and each alternative interpretation. Record the difference of the undertaken projects as the project difference, divide the absolute value of the allocated area difference by the corresponding project land use requirement as the area difference, divide the absolute value of the implementation order difference by the number of projects as the order difference, multiply the sum of the project difference, area difference and order difference by the fusion quality of the corresponding alternative interpretation, and accumulate them as the configuration conflict value of the corresponding triangular face to obtain the configuration conflict table. S42. Read the configuration conflict value and triangle area of ​​each vertex's adjacent triangle face. Multiply each configuration conflict value by the corresponding triangle face area and sum them up. Then divide by the sum of the areas of the adjacent triangle faces to obtain the vertex value. Establish the lower star sequence according to the ascending order of vertex value, vertex number, and simplex dimension to obtain the discrete scalar field.

8. The land planning auxiliary decision-making method based on multi-source data according to claim 7, characterized in that: S4 also includes: S43. Perform the Discrete Morse-Small Segmentation Algorithm on the discrete scalar field. Pair the unpaired simplex in the lower star sequence with the unpaired upper simplex that is in the order of the first simplex and contains the corresponding simplex. The simplex that cannot be paired is recorded as the critical simplex. Connect the critical simplex along the pairing relationship to form a discrete gradient path. If there is only one discrete gradient path between a pair of critical simplexes and the triangular faces on both sides of the corresponding path have the same project, area allocation and implementation order in all resource allocation records, cancel this pair of critical simplexes and reconnect the corresponding pairing relationship. Repeat the cancellation until there are no critical simplexes that meet the cancellation conditions to obtain the preserved gradient path. S44. Read the paths that pass through the triangular facets with non-zero configuration conflict values ​​from the preserved gradient paths, delete the path segments that enter the same side and then return to the corresponding side, and verify that there is at least one different resource configuration record on both sides of the path. Connect the paths that pass the verification to the critical simplex or cadastral parcel boundary along the boundary of the constrained triangular network to obtain the dividing line.

9. The land planning auxiliary decision-making method based on multi-source data according to claim 8, characterized in that: S5 includes: S51. Divide the corresponding cadastral parcels by dividing lines and cadastral parcel boundaries. Establish adjacency relationships for triangular faces that share complete edges and are not separated by dividing lines. Group the adjacent triangular faces that have the same project number, project implementation order, land quota account, and infrastructure service scope in all resource allocation records into the same connected group. Accumulate the allocated area, land quota deduction, and infrastructure capacity deduction within the connected group to obtain the land unit and unit allocation record. S52. Replace the corresponding cadastral parcel with land units, retain resource allocation records that are not affected by changes in land units, reallocate land use demand and regenerate dividing lines starting from the first affected project, and write the newly added dividing lines into the boundary set according to the triangular mesh edge numbers and the start and end positions within the edges. If there are dividing lines in the boundary set that have not been recorded in the existing land unit boundaries, execute S51 and this step until all dividing lines in the boundary set have been recorded in the existing land unit boundaries, and obtain a stable land unit set. S53. Centralize the stable land units and deduct land indicators and infrastructure capacity according to the unit configuration record for land units with zero conflict value, and write them into the project implementation order. Cancel the corresponding resource occupation for land units with non-zero conflict value, generate the verification order according to the descending order of the configuration conflict value and the ascending order of the land unit number, and output the land planning auxiliary decision-making results.

10. A land planning auxiliary decision-making system based on multi-source data, used to implement the land planning auxiliary decision-making method based on multi-source data as described in any one of claims 1-9, the system comprising an evidence construction module, a knowledge fusion module, a resource allocation module, a conflict segmentation module, and a decision output module, characterized in that: The evidence construction module is used to acquire multi-source land records of the target area and unify the spatial benchmark. It constructs a constrained triangular network based on the source boundary, maps synonymous attribute values ​​within the triangular facets to land propositions, determines the quality of propositions based on the coverage area ratio, and determines the quality of unknown propositions based on the uncovered area ratio, thus obtaining the triangular facet evidence set. The knowledge fusion module is used to perform knowledge fusion on the triangular evidence set. It substitutes land propositions from different sources into the sixth proportional conflict redistribution algorithm in the order of their sources, adds the mass product of intersecting land propositions to the intersection proposition, and writes back the mass product of non-intersecting land propositions according to the proportion of their respective mass to the sum of the masses of the two, and accumulates them to obtain the fused proposition and fused mass. The resource allocation module is used to form a baseline interpretation by merging the first and second fusion propositions of each triangular face, and to form a replacement interpretation by replacing the remaining fusion propositions one by one. Under each interpretation, the land use demand is read in the order of implementation. Triangular faces with matching uses and whose land indicators and infrastructure capacity after deduction are not negative are occupied in the order of overlapping area to obtain the resource allocation record. The conflict segmentation module is used to compare the resource configuration records of the baseline interpretation and the alternative interpretation, accumulate the fusion quality corresponding to the alternative interpretation with different configurations as the configuration conflict value, generate vertex values ​​according to the average configuration conflict value of adjacent triangles, perform the discrete Morse-Small segmentation algorithm on the vertex values, and generate segmentation lines along the discrete gradient path passing through the triangles with non-zero configuration conflict values. The decision output module is used to divide the cadastral land parcels by dividing lines, merge adjacent triangular faces that are not separated by dividing lines and have the same configuration into land units, repeat resource allocation and division until the newly added dividing line coincides with the existing boundary, use the land units with zero configuration conflict value for land index allocation and project scheduling, remove the remaining land units from the configuration and sort and verify them, and output the land planning auxiliary decision results.

Images