Industrial land intelligent analysis system and method based on space coupling degree model

Abstract

The invention relates to the field of industrial land analysis, and particularly discloses an industrial land intelligent analysis system and method based on a spatial coupling degree model.The industrial land intelligent analysis method comprises the steps that firstly, by collecting basic geography, resource environment, social economy and other multi-source index data of an industrial land, the problem that high-dimensional heterogeneous data is difficult to directly and effectively utilize is solved; a low-dimensional embedded coding and aggregation analysis algorithm is innovatively introduced to carry out intelligent dimension reduction, feature extraction and aggregation on index data of each subsystem, and an efficacy value capable of accurately reflecting intrinsic characteristics of each subsystem is generated. On the basis, a coupling weight matrix (fusing geography and economic distance) is constructed, the interaction strength among different space units or elements is quantified, the overall efficacy and the overall coupling degree are calculated in combination with the efficacy value of each subsystem, and then the overall coordination degree of the industrial land is obtained. And finally, according to the comprehensive coordination degree, intelligent analysis of the industrial land is realized, and a proper industrial chain link is accurately matched.

Description

Technical Field

[0001] The present application relates to the field of industrial land analysis, and more specifically, to an industrial land intelligent analysis system and method based on a spatial coupling model. Background Art

[0002] The rational planning and efficient utilization of industrial land are the cornerstones of sustainable regional economic development and play a vital role in optimizing industrial structure and enhancing urban competitiveness. However, traditional industrial land planning and analysis often face numerous challenges, such as information asymmetry, a single evaluation dimension, and a lack of consideration of the complex interactions among various influencing factors. This leads to problems such as irrational industrial land allocation, resource waste, increased environmental pressure, and difficulty in coordinated industrial chain development. To achieve scientific, refined, and intelligent management of industrial land, it is urgent to develop an intelligent analysis solution that comprehensively considers multi-dimensional influencing factors and their spatial coupling relationships, thereby improving the matching accuracy between industrial land and introduced industries and the overall coordination of regional development.

[0003] While some research has begun to focus on evaluating industrial land using multi-source data, in-depth modeling and intelligent analysis of the complex spatial coupling effects between subsystems within the industrial land system and between industrial land and the external environment remain underdeveloped. For example, existing analytical methods often focus on the evaluation of single indicators or simple linear superposition, failing to reveal the nonlinear and dynamic interaction mechanisms between subsystems and failing to consider the interplay between geographic proximity and economic connections. These methods often struggle to cope with the high complexity of the industrial land system and the heterogeneous nature of indicator data. Consequently, their analytical results lack accuracy and guidance, making them ineffective in supporting precise industrial chain matching and coordinated regional development decisions.

[0004] Therefore, an optimized intelligent analysis solution for industrial land is desired. Summary of the Invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides an intelligent analysis system and method for industrial land based on a spatial coupling model. First, by collecting multi-source indicator data such as basic geography, resources, environment and social economy of industrial land, in order to solve the problem that high-dimensional heterogeneous data is difficult to directly and effectively use, low-dimensional embedded coding and aggregation analysis algorithms are innovatively introduced to perform intelligent dimensionality reduction, feature extraction and aggregation on the indicator data of each subsystem, and generate an efficacy value that can accurately reflect the intrinsic characteristics of each subsystem. This process not only overcomes data redundancy and noise interference, but also deeply explores the nonlinear correlation between indicators. On this basis, by constructing a coupling weight matrix (integrating geographical and economic distances), the interaction intensity between different spatial units or elements is quantified, and the global efficacy and global coupling degree are calculated in combination with the efficacy values ​​of each subsystem, thereby obtaining the overall coordination degree of the industrial land. Finally, based on this comprehensive coordination degree, intelligent analysis of industrial land is achieved, and appropriate industrial chain links are accurately matched. Through this complete chain from deep mining of data features to quantification of multi-system coupling effects and then to coordinated development assessment, this solution effectively solves the problems of information fragmentation, one-sided evaluation and insufficient consideration of spatial interaction mechanisms in traditional analysis, and realizes scientific and intelligent judgment of the potential and suitability of industrial land.

[0006] According to one aspect of the present application, a method for intelligent analysis of industrial land based on a spatial coupling model is provided, which includes:

[0007] Extract basic geographical indicator data, resource and environmental indicator data, and socioeconomic indicator data for primary industrial land;

[0008] Calculate the basic geographical subsystem efficacy value, resource and environmental subsystem efficacy value, and social and economic subsystem efficacy value of the primary industrial land based on the basic geographical indicator data, resource and environmental indicator data, and social and economic indicator data of the primary industrial land;

[0009] Calculate the global efficacy value and global coupling degree based on the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and socio-economic subsystem efficacy value of the first industrial land;

[0010] Calculating the coordination degree of the first industrial land based on the global efficacy value and the global coupling degree;

[0011] Based on the coordination degree of the first industrial land, the matching industrial chain link of the first industrial land is determined.

[0012] According to another aspect of the present application, an industrial land intelligent analysis system based on a spatial coupling model is provided, which includes:

[0013] The primary industry land index extraction module is used to extract the basic geographical index data, resource and environmental index data and socio-economic index data of the primary industry land;

[0014] a subsystem efficacy value calculation module, configured to calculate the basic geographic subsystem efficacy value, the resource and environmental subsystem efficacy value, and the socio-economic subsystem efficacy value of the primary industrial land based on the basic geographic indicator data, the resource and environmental indicator data, and the socio-economic indicator data of the first industrial land;

[0015] A global operator calculation module, configured to calculate a global efficacy value and a global coupling degree based on the basic geographic subsystem efficacy value, the resource and environmental subsystem efficacy value, and the socio-economic subsystem efficacy value of the first industrial land;

[0016] a primary industrial land coordination degree calculation module, configured to calculate the coordination degree of the first industrial land based on the global efficacy value and the global coupling degree;

[0017] The matching industrial chain determination module is used to determine the matching industrial chain link of the first industrial land based on the coordination degree of the first industrial land.

[0018] Compared with the prior art, the present application provides an intelligent analysis system and method for industrial land based on a spatial coupling model. First, by collecting multi-source indicator data such as basic geography, resources, environment, and socio-economics of industrial land, and addressing the problem that high-dimensional heterogeneous data is difficult to directly and effectively utilize, the application innovatively introduces low-dimensional embedding coding and aggregation analysis algorithms to perform intelligent dimensionality reduction, feature extraction, and aggregation on the indicator data of each subsystem, thereby generating an efficacy value that can accurately reflect the intrinsic characteristics of each subsystem. This process not only overcomes data redundancy and noise interference, but also deeply explores the nonlinear correlation between indicators. On this basis, by constructing a coupling weight matrix (integrating geographic and economic distances), the interaction intensity between different spatial units or elements is quantified, and the global efficacy and global coupling degree are calculated in combination with the efficacy values ​​of each subsystem, thereby obtaining the overall coordination degree of the industrial land. Finally, based on this comprehensive coordination degree, intelligent analysis of industrial land is achieved, and appropriate industrial chain links are accurately matched. Through this complete chain from deep mining of data features to quantification of multi-system coupling effects and then to coordinated development assessment, this solution effectively solves the problems of information fragmentation, one-sided evaluation and insufficient consideration of spatial interaction mechanisms in traditional analysis, and realizes scientific and intelligent judgment of the potential and suitability of industrial land. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0020] Figure 1 Flowchart of an intelligent analysis method for industrial land based on a spatial coupling degree model according to an embodiment of the present application;

[0021] Figure 2 Schematic diagram of data flow of an intelligent analysis method for industrial land based on a spatial coupling degree model according to an embodiment of the present application;

[0022] Figure 3 This is an example flow chart of calculating the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and social and economic subsystem efficacy value of the first industrial land based on the basic geographic indicator data, resource and environmental indicator data, and social and economic indicator data of the first industrial land according to the intelligent analysis method for industrial land based on the spatial coupling model in an embodiment of the present application;

[0023] Figure 4 Another exemplary flowchart of calculating the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and social and economic subsystem efficacy value of the primary industrial land based on the basic geographic indicator data, resource and environmental indicator data, and social and economic indicator data of the primary industrial land according to the intelligent analysis method for industrial land based on the spatial coupling model according to an embodiment of the present application;

[0024] Figure 5 This is a block diagram of an industrial land intelligent analysis system based on a spatial coupling model according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0026] As used in this application and the claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not intended to refer to the singular but may include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0027] Although the present application makes various references to certain modules in the system according to embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are illustrative only, and different aspects of the system and method can use different modules.

[0028] Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the various steps may be processed in reverse order or simultaneously, as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0029] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0030] Facing the shortcomings of traditional methods in multidimensional information fusion, complex relationship modeling, and high-dimensional data processing, this approach first lays the foundation for a comprehensive assessment by collecting multi-source indicator data on industrial land, including basic geography, resources, environment, and socioeconomic indicators. Next, addressing the difficulty of directly and effectively utilizing high-dimensional, heterogeneous data, this approach innovatively introduces low-dimensional embedding coding and aggregation analysis algorithms to intelligently reduce the dimensionality, extract features, and aggregate the indicator data of each subsystem (such as the basic geography subsystem), generating efficacy values ​​that accurately reflect the inherent characteristics of each subsystem. This process not only overcomes data redundancy and noise interference but also deeply explores nonlinear correlations between indicators. Furthermore, a coupling weight matrix (integrating geographic and economic distances) is constructed to quantify the interaction strength between different spatial units or elements. The global efficacy and global coupling degree are calculated by combining the efficacy values ​​of each subsystem, thereby deriving the overall coordination degree of industrial land. Finally, based on this comprehensive coordination degree, intelligent analysis of industrial land is achieved and precise matching of appropriate industrial chain links is achieved. Through this complete chain from deep mining of data features to quantification of multi-system coupling effects and then to coordinated development assessment, this solution effectively solves the problems of information fragmentation, one-sided evaluation and insufficient consideration of spatial interaction mechanisms in traditional analysis, and realizes scientific and intelligent judgment of the potential and suitability of industrial land.

[0031] In the technical solution of this application, an intelligent analysis method for industrial land based on a spatial coupling model is proposed. Figure 1 This is a flowchart of an intelligent analysis method for industrial land based on a spatial coupling model according to an embodiment of the present application. Figure 2 Schematic diagram of data flow of the industrial land intelligent analysis method based on the spatial coupling model according to the embodiment of the present application. Figure 1and Figure 2 As shown, the intelligent analysis method for industrial land based on the spatial coupling model according to the embodiment of the present application includes the following steps: S100, extracting the basic geographical indicator data, resource and environmental indicator data, and socio-economic indicator data of the first industrial land; S200, calculating the basic geographical subsystem efficacy value, resource and environmental subsystem efficacy value, and socio-economic subsystem efficacy value of the first industrial land based on the basic geographical indicator data, resource and environmental indicator data, and socio-economic indicator data of the first industrial land; S300, calculating the global efficacy value and the global coupling degree based on the basic geographical subsystem efficacy value, resource and environmental subsystem efficacy value, and socio-economic subsystem efficacy value of the first industrial land; S400, calculating the coordination degree of the first industrial land based on the global efficacy value and the global coupling degree; S500, determining the matching industrial chain link of the first industrial land based on the coordination degree of the first industrial land.

[0032] Specifically, in step S100, basic geographic indicator data, resource and environmental indicator data, and socioeconomic indicator data for primary industrial land are extracted. It should be understood that intelligent analysis of industrial land does not rely solely on single-dimensional information; rather, it requires a comprehensive consideration of multiple factors influencing its suitability, potential, and sustainability. Therefore, extracting basic geographic indicator data, resource and environmental indicator data, and socioeconomic indicator data for primary industrial land is the cornerstone of building a comprehensive assessment system. These data represent the objective physical conditions of industrial land, its interaction with the ecological environment, and the human activities and economic development potential it supports. Together, they form a complete framework for understanding and evaluating the value of industrial land. Specifically, basic geographic indicator data provides information on the spatial location, topography, hydrology, and transportation accessibility of industrial land. For example, terrain slope determines construction costs and suitable industry types; distance from transportation hubs such as roads, railways, and ports directly affects logistics costs and transportation efficiency; and water system distribution affects water security and sewage disposal conditions. These data are essential for understanding the inherent properties and spatial layout constraints of industrial land. Resource and environmental indicators reflect the carrying capacity and sensitivity of the natural ecosystems in which industrial land is located. For example, climate conditions (precipitation, temperature), soil type, and vegetation cover directly impact environmental capacity and resource consumption by different industries. The availability of energy resources (electricity, natural gas), water availability, air quality, and potential environmental risks (such as pollution sources) are key factors in assessing the environmental suitability and sustainable development potential of industrial land. These data help identify environmentally sensitive areas and prevent industrial layouts from causing ecological damage or resource depletion. Socioeconomic indicators provide insights from the perspectives of human and economic development. These include population density, labor supply structure, industrial base, GDP, tax contributions, output per unit area, distribution of scientific research institutions, and policy support. For example, population density and labor structure determine the availability of human resources; regional economic development level and industrial agglomeration effects influence industrial chain synergy and market hinterland; and government policy orientation and fiscal support directly influence the industrial investment environment. These data help assess the potential contribution of industrial land to the regional economy and its synergy with upstream and downstream industrial chains.

[0033] More specifically, in a specific example of this application, extracting such data generally relies on a comprehensive data collection and preprocessing process. First, geospatial data (such as topographic maps, image maps, transportation network maps, water system maps, etc.) can be integrated through remote sensing technology, geographic information system (GIS) data acquisition and related databases. For example, slope and elevation information can be extracted using a public digital elevation model (DEM); and buffer zones from transportation hubs can be generated by vectorizing the transportation network. Secondly, resource and environmental data come from various environmental monitoring station networks, meteorological departments, water conservancy departments, land use status surveys, etc., or are obtained through remote sensing image interpretation (such as land use type and vegetation cover index). For example, air quality indicators are obtained through environmental monitoring data, and groundwater levels are obtained through water resources census data. Finally, socioeconomic data mainly comes from statistical yearbooks, economic census data, industry reports, government public information, and big data platforms (such as business registration information, employment data, etc.). For example, GDP and population density data are obtained through statistical departments, and per-mu tax information is obtained through tax departments. After being aggregated, these heterogeneous data usually undergo a series of preprocessing steps, including data cleaning, missing value processing, standardization, unified spatial reference (projection transformation), etc., to ensure data quality and comparability, and provide high-quality input for subsequent intelligent analysis models.

[0034] Specifically, in step S200, the primary industrial land's basic geographic indicator data, resource and environmental indicator data, and socioeconomic indicator data are used to calculate the primary industrial land's basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and socioeconomic subsystem efficacy value. It should be understood that simply superimposing or comparing raw, heterogeneous, and high-dimensional basic geographic indicator data, resource and environmental indicator data, and socioeconomic indicator data is difficult to accurately reflect the true effectiveness of each subsystem, and cannot effectively capture the complex nonlinear relationships between indicators. These raw data often have varying dimensions, redundant information, and even contain noise. Without refinement and integration, the accuracy and efficiency of subsequent coupling analysis will be seriously affected. Therefore, it is necessary to transform these multidimensional, heterogeneous, and potentially correlated raw data into a unified quantitative indicator, namely, an "efficacy value," that can comprehensively represent the overall status of each subsystem. In other words, first, the complex high-dimensional indicator information must be compressed into a low-dimensional representation that is easy to understand and manipulate, providing standardized and information-rich input for the subsequent calculation of global coupling and coordination. Secondly, by quantifying the efficacy values ​​of each subsystem, we can clearly and independently reveal the performance of specific industrial land in terms of geographical conditions, environmental endowments and socio-economic development potential, thus laying the foundation for in-depth analysis of the interactions and coordination among the subsystems.

[0035] Figure 3This is an example flow chart of calculating the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and social and economic subsystem efficacy value of the first industrial land based on the basic geographic indicator data, resource and environmental indicator data, and social and economic indicator data of the first industrial land according to the industrial land intelligent analysis method based on the spatial coupling model of the embodiment of the present application. Figure 3 As shown, according to the intelligent analysis method for industrial land based on the spatial coupling model of the embodiment of the present application, step S200 includes: S201, calculating the weighted sum of each basic geographical indicator in the basic geographical indicator data of the first industrial land to obtain the basic geographical subsystem efficacy value; S202, calculating the weighted sum of each resource and environmental indicator in the resource and environmental indicator data of the first industrial land to obtain the resource and environmental subsystem efficacy value; S203, calculating the weighted sum of each socio-economic indicator in the socio-economic indicator data of the first industrial land to obtain the socio-economic subsystem efficacy value.

[0036] Figure 4 This is another example flow chart of calculating the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and social and economic subsystem efficacy value of the first industrial land based on the basic geographic indicator data, resource and environmental indicator data, and social and economic indicator data of the first industrial land according to the industrial land intelligent analysis method based on the spatial coupling model according to the embodiment of the present application. Figure 4 As shown, according to the intelligent analysis method of industrial land based on the spatial coupling degree model of the embodiment of the present application, step S200 includes: S210, low-dimensional embedding coding of the basic geographical indicators of each first industrial land in the basic geographical indicator data of the first industrial land to obtain a set of basic geographical indicator low-dimensional embedding coding vectors; S220, constructing a coupling weight matrix; S230, coupling optimization of each basic geographical indicator low-dimensional embedding coding vector in the set of basic geographical indicator low-dimensional embedding coding vectors based on the coupling weight matrix to obtain a set of optimized basic geographical indicator low-dimensional embedding coding vectors; S240, aggregation analysis of the set of basic geographical indicator low-dimensional embedding coding vectors to obtain a basic geographical indicator low-dimensional embedding aggregation coding vector; S250, feature decoding of the basic geographical indicator low-dimensional embedding aggregation coding vector to obtain the basic geographical subsystem efficacy value.

[0037] Specifically, in step S210, low-dimensional embedding coding is performed on the basic geographical indicators of each primary industrial land in the basic geographical indicator data of the first industrial land to obtain a set of low-dimensional embedded coding vectors of basic geographical indicators. It should be understood that since the original basic geographical indicators (such as slope, elevation, distance to water system, distance to transportation artery, land use type, etc.) are usually numerous and high-dimensional, and there may be collinearity or redundant information between the indicators, directly analyzing in such a high-dimensional space is not only computationally inefficient and susceptible to the "curse of dimensionality", leading to data sparsity problems, but may also introduce noise and blur key features, thereby reducing the accuracy and robustness of subsequent model construction. In addition, using simple methods to perform a simple weighted summation of these indicators makes it difficult to effectively capture the nonlinear relationship between the indicators and their complex joint impact on the suitability of industrial land. For example, a moderate slope combined with good drainage conditions may constitute the optimal geographical form, rather than a simple superposition of a single indicator. Based on this, in the technical solution of the present application, the basic geographical indicators of each primary industrial land in the basic geographical indicator data of the first industrial land are further low-dimensionally embedded and coded to map the high-dimensional, possibly sparse and noisy original basic geographical indicator data to a low-dimensional, dense and more information-rich vector space. In this new space, similar geographical feature data points will be close to each other in the vector space, making the data structure and internal pattern clearer. In this way, the most representative and discriminative potential features in the data can be effectively extracted, information redundancy can be compressed, and a simpler, more abstract but more insightful input can be provided for subsequent complex models (such as coupling optimization and aggregation analysis). This not only improves computational efficiency, but also enhances the model's ability to understand the inherent correlation of geographical features. More specifically, in a specific example of the present application, the basic geographical indicators of each primary industrial land in the basic geographical indicator data of the first industrial land (such as slope, elevation, distance from water system, distance from transportation artery, land use type, etc.) are passed through an embedding coding model based on Word2Vec to obtain a set of low-dimensional embedded coding vectors of basic geographical indicators.

[0038] Specifically, in step S220, a coupling weight matrix is ​​constructed. It should be understood that industrial land and the economic activities it carries do not exist in isolation, and there are complex spatial interactions and economic connections between them. Traditional industrial land analysis methods often ignore this interactivity and only evaluate based on a single attribute or independent indicator of the land, resulting in the evaluation results being out of touch with the actual situation and unable to effectively guide the optimization of industrial layout. For example, even if two industrial lands are superior in their own conditions, if they are too far apart geographically or lack economic connection, it is difficult to form an industrial agglomeration effect or coordinated development. In order to truly reflect this inherent connection and incorporate it into the intelligent analysis system, a mechanism must be designed to quantify the intensity of this interaction, that is, to construct a weight matrix that can reflect "who is closer to whom".

[0039] More specifically, in an embodiment of the present application, constructing a coupling weight matrix includes: constructing a geographic distance matrix; constructing an economic distance matrix; and generating the coupling weight matrix based on the geographic distance matrix and the economic distance matrix. By constructing a geographic distance matrix, the purpose is to capture spatial proximity, because geographical proximity usually means lower transportation costs, more convenient communication and a higher possibility of agglomeration. At the same time, by constructing an economic distance matrix, the purpose is to measure the similarity or complementarity between economic activities or industrial entities in terms of output value, industrial chain association, resource sharing, etc., because close economic ties can promote the optimal allocation of resources, form collaborative networks and enhance overall competitiveness. Ultimately, generating a coupling weight matrix based on these two distance matrices is to organically integrate the "proximity" in physical space with the "connection" in economic functions, thereby forming a comprehensive measure that can fully reflect the coupling intensity between industrial lands. This matrix will serve as the key basis for the subsequent coupling optimization process, guiding the model to consider interactions rather than independent attributes when exploring land potential.

[0040] Specifically, the geographic distance matrix (W d ) is calculated as:

[0041]

[0042] where d i.j represents the geographical distance between the i-th enterprise and the j-th enterprise.

[0043] Economic distance matrix (W e ) is calculated as:

[0044]

[0045] where e i.j Represents the difference in output value between the i-th enterprise and the j-th enterprise.

[0046] The coupling weight matrix (W) is the fusion of geographical and economic distances, and its formula is:

[0047] W=W e ×W d

[0048] Specifically, in step S230, the coupling optimization is performed on each of the basic geographic indicator low-dimensional embedding coding vectors in the set of basic geographic indicator low-dimensional embedding coding vectors based on the coupling weight matrix to obtain a set of optimized basic geographic indicator low-dimensional embedding coding vectors. It should be understood that the coupling weight matrix optimizes the expression of the embedded features of each basic geographic indicator by integrating spatial and economic interactions to improve the globality and robustness of the basic geographic indicator feature expression. Although the original basic geographic indicator low-dimensional embedding coding vectors have captured the intrinsic characteristics of each indicator (such as slope, elevation, distance from water system, distance from transportation artery, land use type, etc.), these vectors only reflect independent attributes and ignore the synergistic or restrictive effects derived from the geographical proximity and economic correlation between enterprises in the industrial land system. The coupling weight matrix (generated by the fusion of the geographical distance matrix and the economic distance matrix) quantifies the interaction intensity between different enterprises. The larger the element value, the closer the coupling relationship between enterprises (small geographical distance or small output value difference). Through coupling optimization, this spatial-economic dependency relationship can be injected into the feature vector, so that the optimized features not only retain the characteristics of the indicator itself, but also encode the association pattern with related enterprises, thereby more accurately representing the overall spatial coupling state of industrial land and laying the foundation for subsequent aggregation analysis and coordination calculation.

[0049] Specifically, in a specific example of the present application, first, each basic geographic indicator low-dimensional embedding coding vector is matrix-multiplied by the coupling weight matrix to obtain a set of basic geographic indicator low-dimensional embedding coding vectors after information mapping. Then, the set of basic geographic indicator low-dimensional embedding coding vectors after information mapping is subjected to a multi-layer perceptron (MLP) including a nonlinear activation function (such as ReLU or tanh) to perform a nonlinear transformation on the set of basic geographic indicator low-dimensional embedding coding vectors after information mapping to obtain the set of optimized basic geographic indicator low-dimensional embedding coding vectors.

[0050] Specifically, in step S240, the set of low-dimensional embedded coding vectors of basic geographical indicators is subjected to aggregation analysis to obtain low-dimensional embedded aggregate coding vectors of basic geographical indicators. It should be understood that although the original low-dimensional embedded coding vectors of basic geographical indicators have captured the semantic characteristics of individual indicators (such as slope, elevation, water system distance, etc.) through low-dimensional coding and coupling optimization, these scattered vectors cannot directly represent the overall geographical attributes of industrial land units. The synergistic effects and spatial coupling relationships between indicators are separated in independent vectors, which makes it difficult to support the subsequent accurate calculation of the global coupling degree. Therefore, in the technical solution of the present application, the set of low-dimensional embedded coding vectors of basic geographical indicators is further subjected to aggregation analysis to obtain low-dimensional embedded aggregate coding vectors of basic geographical indicators. Through aggregation analysis, the commonalities and specific deviations of a set of basic geographic indicators can be organically integrated. K-Means clustering is first used to extract initial cluster centers representing spatial distribution density, forming the basic pattern primitives of geographic indicators. A self-attention mechanism is then used to explore the global correlations between cluster centers, generating a core integrated representation vector that reflects the intrinsic structure of the geographic attributes of industrial land. The excited-state components of each original indicator vector relative to the ground-state characteristics are then calculated, and the individual deviation expressions under boundary geometry constraints are reconstructed using induced metric theory to isolate key specific information. Finally, the ground-state commonalities and weighted excited-state deviations are integrated to generate a compact aggregate representation that combines global statistical properties with local significant differences. The resulting low-dimensional embedded aggregate encoding vector of basic geographic indicators breaks through the limitations of traditional weighted averaging and preserves nonlinear interactions in spatial coupling through quantized feature decomposition. For example, the unique topographic patterns of industrial clusters can be identified through the synergistic relationship between slope and transportation distance.

[0051] More specifically, in an embodiment of the present application, an aggregation analysis is performed on the set of the basic geographic indicator low-dimensional embedded coding vectors to obtain the basic geographic indicator low-dimensional embedded aggregate coding vectors, including: performing core feature integration processing based on cluster analysis on the set of the basic geographic indicator low-dimensional embedded coding vectors to obtain the basic geographic indicator set core integrated representation vector; based on the basic geographic indicator set core integrated representation vector, calculating the dynamic compensation representation vector of each basic geographic indicator low-dimensional embedded coding vector in the set of the basic geographic indicator low-dimensional embedded coding vector relative to the basic geographic indicator set core integrated representation vector to obtain a set of basic geographic indicator feature compensation dynamic representation vectors; calculating Calculate the dynamic significance adjustment weight factor of each basic geographical indicator feature compensation dynamic representation vector in the set of basic geographical indicator feature compensation dynamic representation vectors to obtain a set of basic geographical indicator dynamic significance adjustment weight factors; based on the set of basic geographical indicator dynamic significance adjustment weight factors, perform weighted modulation on the set of basic geographical indicator feature compensation dynamic representation vectors to obtain a set of basic geographical indicator feature compensation dynamic representation significant coding vectors; input the set of basic geographical indicator feature compensation dynamic representation significant coding vectors and the core integrated representation vector of the basic geographical indicator set into a feature enhancement fusion network to obtain the basic geographical indicator low-dimensional embedding aggregation coding vector.

[0052] In particular, in a specific example of the present application, the set of low-dimensional embedded coding vectors of the basic geographic indicators is subjected to core feature integration processing based on clustering analysis to obtain a core integrated representation vector of the basic geographic indicator set, including: inputting the set of low-dimensional embedded coding vectors of the basic geographic indicators into a K-Means clustering network to obtain K basic geographic indicator initial cluster center coding vectors; performing self-attention coding on each of the K basic geographic indicator initial cluster center coding vectors to obtain K basic geographic indicator self-attention coding vectors; and performing positional aggregation on the K basic geographic indicator self-attention coding vectors and the K basic geographic indicator initial cluster center coding vectors to obtain the core integrated representation vector of the basic geographic indicator set.

[0053] Specifically, the set of low-dimensional embedded coding vectors of the basic geographical indicators is input into the K-Means clustering network to obtain K basic geographical indicator initial cluster center coding vectors, which can be expressed as follows:

[0054] X={x1,x2,...,x i ,...,x n}

[0055] C=K_Means(X)={c1,c2,...,c k}

[0056] Q=CW Q

[0057] K=CW k

[0058] V=CW v

[0059] Where X is the set of low-dimensional embedding coding vectors of the basic geographic indicators, x1, x2, x i ,x n are the first, second, i-th and n-th basic geographical indicator low-dimensional embedding coding vectors in the set of basic geographical indicator low-dimensional embedding coding vectors, K_Means is K-Means clustering network processing, C is the initial clustering center coding vector of K basic geographical indicators, c1, c2, c k They are respectively the first, second and kth basic geographic indicator initial cluster center coding vectors among the K basic geographic indicator initial cluster center coding vectors.

[0060] It's understandable that although basic geographic indicator data has been low-dimensionally embedded, forming a set of multiple vectors (each representing a feature representation of a basic geographic indicator such as slope, elevation, and distance to a water system), this set itself can still be large and complex. Directly applying global feature aggregation analysis (such as self-attention mechanisms) to this set can be computationally expensive, difficult to capture key common patterns, and susceptible to noise or redundant information. For example, a large number of geographic indicator vectors may imply typical spatial distribution patterns or common geographic feature combinations (such as the "steep slope-short travel distance" pattern or the "gentle slope-near water system" pattern). These patterns are key to understanding regional geographic characteristics, but extracting these common structures directly from the original set of embedding vectors is challenging. Therefore, we utilize the classic unsupervised learning algorithm, K-Means, to perform a preliminary and efficient structured induction and information compression on this set of low-dimensional embedding encoding vectors of basic geographic indicators. Through iterative optimization, K-Means aims to find areas with dense data points in a high-dimensional feature space (here, the space formed by low-dimensional embedding vectors) and identify the centers (centroids) of these areas to form K representative initial cluster center encoding vectors for basic geographic indicators. These cluster centers can be regarded as "pattern primitives" or "prototypes" in the feature space of basic geographic indicators, representing the K most important combinations or trends of geographic features in the dataset. This can compress and refine the complex and original set of low-dimensional embedding encoding vectors of basic geographic indicators, providing a simpler and more structured input for subsequent more complex feature aggregation analysis (such as collective ground-state feature aggregation based on the self-attention mechanism).

[0061] Specifically, self-attention encoding is performed on each of the K basic geographic indicator initial cluster center encoding vectors to obtain K basic geographic indicator self-attention encoding vectors, which can be expressed as follows:

[0062] Q=CW Q

[0063] K=CW k

[0064] V=CW v

[0065]

[0066] Among them, W Q 、W k and W v are the basic geographic indicator query weight matrix, basic geographic indicator key weight matrix and basic geographic indicator value weight matrix respectively. Q, K and V are the sets of basic geographic indicator query vectors, the sets of basic geographic indicator key vectors and the sets of basic geographic indicator value vectors respectively. d is the scale of K, Softmax is the Softmax function, Attention is the self-attention mechanism, and Attention(Q,K,V) is the set of basic geographic indicator self-attention encoding vectors.

[0067] It should be understood that although these vectors effectively summarize the data density peaks of local areas, they are essentially isolated individual representations and fail to capture the potential global, high-order correlations and synergistic effects between these core geographical patterns. For example, a cluster center representing a "port flat area" and another cluster center representing an "inland transportation hub" often have topographic complementarity and economic functional interdependence in the actual industrial land system. This cross-modal correlation cannot be revealed by K-Means itself. Therefore, in the technical solution of this application, each of the K basic geographical indicator initial cluster center encoding vectors is further self-attention encoded to obtain K basic geographical indicator self-attention encoding vectors. This makes up for the limitations brought about by the independence of K-Means cluster centers, that is, as representatives of local patterns, they cannot reflect the complex spatial coupling mechanism inherent in the industrial land system as a whole. Simply put, K-Means gives "what typical geographical features there are", while the self-attention mechanism needs to understand "what kind of relationship exists between these typical geographical features". By applying the self-attention mechanism, we can more comprehensively reflect the status of a specific combination of geographical features in the entire industrial land system and the intensity of its interaction with other features. For example, a self-attention encoding vector of "moderate slope-proximity to transportation hubs" may be strengthened due to its strong correlation with the cluster center of "urban core area-complete infrastructure", thereby better representing the potential industrial carrying capacity of the region and making the final aggregate representation more insightful and accurate.

[0068] Specifically, the K basic geographic indicator self-attention encoding vectors and the K basic geographic indicator initial cluster center encoding vectors are aggregated by position to obtain the core integrated representation vector of the basic geographic indicator set, which is expressed as follows:

[0069] g=layerNorm(C+Attention(Q,K,V))

[0070] Among them, (C+Attention(Q,K,V)) is the positional addition of C and Attention(Q,K,V), layerNorm is the layer normalization processing, and g is the core integrated representation vector of the basic geographic indicator set.

[0071] It's understandable that the self-attention encoding vectors of basic geographic indicators capture the interactions and contextual relationships between different geographic patterns. However, while the self-attention encoding vectors can capture global correlations, such as the topographic synergy between the port area and the inland industrial belt, over-reliance on context can dilute the local density characteristics of the original cluster centers, for example, overlooking the independent geological features of steep-slope industrial zones. Furthermore, while the original initial cluster center encoding vectors faithfully preserve these local characteristics, they lack a macroscopic perspective on the networked coupling of the entire industrial land system. Therefore, the network requires a composite representation that balances both local essence and global connections, ensuring that key individual geographic features are not lost in subsequent analysis while fully leveraging the synergies between patterns. Simply put, it's crucial to see both the "trees" (the local characteristics of the initial cluster center encoding vectors of basic geographic indicators) and the "forest" (the global connections of the self-attention encoding vectors of basic geographic indicators), and to understand each tree in the forest within its relationship to other trees. Therefore, this solution aggregates the K basic geographic indicator self-attention encoding vectors and the K basic geographic indicator initial cluster center encoding vectors by position, fusing the global context perception capability imparted by self-attention with the original, precise local feature density information retained by the initial cluster centers. This enables the final integrated vector to simultaneously encode the independent characteristics of each typical geographic pattern as well as their interactions and synergistic relationships within the entire geographic system, thereby forming a more comprehensive and robust set-level core representation. This representation is no longer a single-dimensional information summary, but rather has the ability to deeply characterize the complex coupling patterns of geographic features.

[0072] Specifically, based on the core integrated representation vector of the basic geographic indicator set, the dynamic compensation representation vector of each basic geographic indicator low-dimensional embedded coding vector in the set of the basic geographic indicator low-dimensional embedded coding vector is calculated relative to the core integrated representation vector of the basic geographic indicator set to obtain a set of basic geographic indicator feature compensation dynamic representation vectors, which is expressed as follows:

[0073] e i =Sigmoid(W e [x i ;g]+b e )⊙(x i -g)

[0074] ξ={e1,e2,...,e i ,...,e n}

[0075] Among them, W e and b eare the modulated trainable weight matrix and the modulated trainable bias vector, respectively, ⊙ is the position point multiplication, Sigmoid is the Sigmoid function, ξ is the set of basic geographic indicator feature compensation dynamic representation vectors, e1, e2, e i ,e n They are respectively the 1st, 2nd, i-th and n-th basic geographical indicator feature compensation dynamic representation vectors in the set of basic geographical indicator feature compensation dynamic representation vectors.

[0076] It should be understood that the core integrated representation vector of the basic geographic indicator set represents the "base state" or "collective statistical characteristics" of the entire basic geographic indicator set. It well summarizes the global correlation between common geographic feature combinations and patterns. However, in actual industrial land scenarios, each specific low-dimensional embedding encoding vector of the original geographic indicator (for example, the slope, elevation, distance to the water system, and other characteristics of a particular plot of land) may have its own unique characteristics or deviations that are not fully explained by this "base state". For example, a plot of land may have unusual underground caves, extreme geological structures, or small hills that appear abruptly in a generally flat area. These are all "individual-specific information" that deviates from the average behavior. If we rely solely on the base state vector, we will not be able to capture those key individual uniqueness or potential anomalies, and these "deviations" often carry information that is critical to downstream tasks, such as potential risks, special utilization potential, etc. Therefore, in the technical solution of this application, a dynamic compensation representation vector is further calculated for each low-dimensional embedded coding vector of the basic geographic indicator in the set relative to the core integrated representation vector of the basic geographic indicator set. This aims to conceptually decompose each original low-dimensional embedded coding vector of the basic geographic indicator into its implicit projection on the "ground state" (i.e., the portion explained by the ground state) and the dynamic compensation representation that is not explained by the ground state. Its essence lies in separating the portion of each original feature vector that is not explained by the ground state, namely, the deviation or perturbation of the original low-dimensional embedded coding vector of the basic geographic indicator relative to the collective average behavior. Through this process, the network can explicitly process and quantify the uniqueness or anomaly of each data point, allowing its individual-specific information to be clearly expressed rather than being obscured by the average. In this way, whether in assessing the uniqueness of a specific plot of land, identifying potential geographic risks, or in the subsequent fusion stage, this composite information that takes into account both the global and individual, commonalities and differences, can be more fully utilized to generate a high-information-density aggregate representation, providing more comprehensive and detailed insights for intelligent analysis of industrial land.

[0077] Specifically, the dynamic significance adjustment weight factor of each basic geographical indicator feature compensation dynamic representation vector in the set of basic geographical indicator feature compensation dynamic representation vectors is calculated to obtain a set of basic geographical indicator dynamic significance adjustment weight factors, which is expressed as follows:

[0078]

[0079] A={α1,α2,...,α i ,...,α n}

[0080] Among them, W s is the excitation modulation trainable weight matrix, ReLU is the ReLU activation function, w is the excitation modulation trainable vector, exp is the exponential function value with the natural constant e as the base, A is the set of dynamic significance adjustment weight factors of basic geographical indicators, α1, α2, α i ,α n They are respectively the 1st, 2nd, i-th and n-th basic geographical indicator dynamic significance adjustment weight factors in the set of basic geographical indicator dynamic significance adjustment weight factors.

[0081] It should be understood that these basic geographic indicator feature compensation dynamic representation vectors quantify the difference between each geographic indicator and the ensemble average behavior. However, not all individual deviations have equal importance or decision-making value. For example, the slope of a certain plot may fluctuate slightly relative to the regional average, but this may be just insignificant noise; while another plot, although its slope does not change much, has special geological structures (such as caves or faults) underground, which may be a key, highly significant deviation that has a decisive impact on the suitability of industrial land. Simply quantifying the magnitude of the deviation cannot distinguish between these "useful signals" and "useless noise." If not screened and weighted, meaningless random noise or unimportant details may be amplified, resulting in inaccurate and inefficient subsequent aggregate representation, or even over-expansion of representation, and failure to refine truly valuable information. Therefore, in the technical solution of the present application, the dynamic significance adjustment weight factor of each basic geographical indicator feature compensation dynamic representation vector in the set of the basic geographical indicator feature compensation dynamic representation vector is further calculated. Through this process, the network ensures that the subsequent weighted modulation (applying these weight factors to the corresponding excited state vectors) will only highlight those truly meaningful individual differences, thereby effectively removing noise and improving the purity and refinement of information. This enables the final aggregate representation to focus more accurately on the core characteristics and key anomalies of industrial land, avoids unnecessary information expansion, and achieves a compact, high-information-density overall representation, providing high-quality, intelligently screened individual difference information for the final feature gain superposition and efficacy value calculation.

[0082] Specifically, based on the set of basic geographical indicator dynamic significance adjustment weight factors, the set of basic geographical indicator feature compensation dynamic representation vectors is weighted modulated to obtain a set of basic geographical indicator feature compensation dynamic representation significant coding vectors, which can be expressed as follows:

[0083]

[0084] in, A set of dynamic representation significant coding vectors for basic geographic indicator feature compensation, They are respectively the 1st, 2nd, i-th and n-th basic geographical indicator feature compensation dynamic representation significant coding vectors in the set of basic geographical indicator feature compensation dynamic representation significant coding vectors.

[0085] It should be understood that by multiplying the dynamic representation vector of each basic geographic indicator feature compensation with its corresponding basic geographic indicator dynamic significance adjustment weight factor, individual deviations judged to be "significant" are selectively amplified, while deviations judged to be "insignificant" or "noise" are suppressed or weakened. The goal of this is to generate a filtered and focused excitation state representation, ensuring that only individual-specific information that is of practical significance to industrial land assessment and can truly reflect the region's unique potential risks or opportunities is effectively transmitted to subsequent stages. This allows the network to more accurately capture "individual highlights" rather than being distracted by irrelevant details.

[0086] Specifically, the set of the basic geographic indicator feature compensation dynamic representation significant coding vector and the core integrated representation vector of the basic geographic indicator set are input into the feature enhancement fusion network to obtain the basic geographic indicator low-dimensional embedding aggregation coding vector, which is expressed as follows:

[0087]

[0088] Where n is the number of vectors in the set of dynamic representation significant coding vectors of the basic geographical indicator feature compensation, MLP is a multi-layer perceptron, and v f A low-dimensional embedding aggregation encoding vector is generated for the basic geographic indicators.

[0089] It should be understood that simply splicing or averaging the set of significant encoding vectors of the basic geographic indicator feature compensation dynamic representation and the core integrated representation vector of the basic geographic indicator set cannot effectively mine the complex nonlinear relationship between them, nor can it ensure the information integrity and compactness of the final basic geographic indicator representation. Therefore, an advanced fusion mechanism is needed that can understand the deep connection between the ground state and the excited state and intelligently fuse them into a single, more expressive vector. In the technical solution of this application, the task of the feature enhancement fusion network is to learn how to optimally combine this information to generate a final representation that includes both the overall average portrait (provided by the ground state) and the most noteworthy individual highlights (provided by the dynamic representation). This fusion process goes beyond simple linear combination and can capture the complex interaction between the ground state and the dynamics, for example, how a unique geographic feature (dynamic) appears more prominent or has special significance in the overall geographic background (ground state). Its core function is to construct a final, information-complete aggregate representation, aiming to encapsulate the multi-level information of the entire input set in a compact manner. The resulting low-dimensional embedded aggregate coding vectors of basic geographic indicators not only reflect the universal patterns and macroscopic characteristics of the entire geographic region or type collection, but also highlight and quantify the key deviations exhibited by each unique geographic entity or feature within the region. This compact and comprehensive representation not only overcomes the redundancy and noise of the original high-dimensional data but also provides high-quality and efficient input for subsequent more advanced analytical tasks (such as industrial land matching and coordination calculation), greatly enhancing the analytical capabilities and decision-making accuracy of the entire system.

[0090] Preferably, when calculating the dynamic compensation representation of the low-dimensional embedded coding vector of each basic geographical indicator relative to the core integrated representation vector of the basic geographical indicator set, (x i -g) essentially defines the ground-state space-induced metric of the low-dimensional embedding coding vector of each basic geographic indicator relative to the core integrated representation vector of the basic geographic indicator set, that is, the relatively high-dimensional original feature vectors x i The attribute space of the basic geographic indicator set is used as the measurement benchmark to define the dimensionality reduction space measurement criteria of the core integrated representation vector in the basic geographic indicator set through guided mapping.

[0091] In this way, it is necessary to consider the boundary measure problem under the guided measure criterion, that is, to make the metric of dynamic compensation limited by the geometric restrictions of spatial transformation under the guided mapping, rather than a simple high-low dimensional space transformation.

[0092] Based on this, first calculate x i The low-order activation coefficient α relative to g = ||x i ||1-||g||1 and β=||x i||2-||g||2, and introduce the low-order initial space form x 1i =x i -α and x 2i =x i -β, thereby solving the guided map in the form of the initial space so that it is projected onto an orthogonal basis in the ground state space.

[0093] Then, the low-order initial space form is used as the measure representation under the local submanifold construction to reconstruct the boundary-space geometry as follows:

[0094]

[0095] The adaptive correction complexity is converted from the high-dimensional set expression determined by the low-dimensional embedding coding vector of the basic geographic indicators to the edge geometric characteristics of the initial spatial measurement criterion. Where ω is the adjustment coefficient used to compensate for the excessive spatial morphological parameter n n-1 The impact brought about.

[0096] In this way, x' i Correction i for e' i =x' i ⊙e i , under the guidance measure criterion from the submanifold measure criterion to the ground state, by making the adaptive correction measure depend on the boundary rather than the spatial transformation, the normalization of the fixed points in the tangential measure flow can be achieved to achieve effective measurement of the boundary and avoid excessive expansion of the representation of the dynamic representation vector of each basic geographic indicator feature compensation.

[0097] Specifically, in step S250, the basic geographic indicator low-dimensional embedded aggregate coding vector is feature decoded to obtain the basic geographic subsystem efficacy value. It should be understood that the basic geographic indicator low-dimensional embedded aggregate coding vector, as a composite embedding of the set, not only reflects the global statistical characteristics, but also retains the key individual difference information, but it is essentially an abstract, high-dimensional feature representation, and is not a "efficacy value" that can be directly used for decision-making or intuitive understanding. In other words, although the basic geographic indicator low-dimensional embedded aggregate coding vector integrates rich information, its multidimensional characteristics make it difficult to serve as a single, standardized basis for judging the quality of the geographic subsystem. It needs to be "translated" into a specific, quantitative indicator to support subsequent coordination calculations and industrial chain matching. Therefore, in an embodiment of the present application, the basic geographic indicator low-dimensional embedded aggregate coding vector can be feature decoded by a basic geographic subsystem efficacy value calculator based on a decoder to obtain the basic geographic subsystem efficacy value. This calculator is essentially a feedforward neural network (or multi-layer perceptron, MLP). Its operational process can be summarized as follows: The entire feature decoding process involves embedding highly abstract yet information-rich basic geographic indicators into a low-dimensional aggregate encoding vector. Through a carefully trained neural network, this dimensionality is reduced and mapped into a quantifiable utility value with clear business implications. This enables the complex feature representation within the machine learning model to ultimately serve the needs of practical intelligent analysis and decision-making for industrial land use, completing a closed loop from deep data understanding to ultimate benefit evaluation.

[0098] Specifically, in step S300, a global efficacy value and global coupling degree are calculated based on the first industrial land's basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and socioeconomic subsystem efficacy value. It should be understood that the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and socioeconomic subsystem efficacy value each independently quantify the suitability of industrial land in specific dimensions. However, an industrial land is not simply the sum of these independent attributes. The rise and fall of an industry, as well as its sustainable development, are often the complex result of the interactions and synergistic coupling of multiple subsystems. For example, a piece of land with favorable geographical conditions but scarce resources may have its true value significantly reduced due to resource bottlenecks. Similarly, a region with a good socioeconomic foundation but poor environmental carrying capacity will find it difficult to achieve long-term sustainable development. Therefore, to achieve a highly abstract and refined representation of the complex industrial land system and enable cross-regional comparison and optimal matching, the system requires a set of macro-indicators that reflect the overall comprehensive benefits and the interaction between subsystems. Based on this, the global efficacy value and global coupling degree are further calculated.

[0099] More specifically, in an embodiment of the present application, based on the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and socio-economic subsystem efficacy value of the first industrial land, a global efficacy value and a global coupling degree are calculated, including: calculating the global efficacy value based on the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and socio-economic subsystem efficacy value of the first industrial land using the following global efficacy calculation formula, the global efficacy calculation formula being:

[0100]

[0101] Among them, U1 is the effectiveness value of the basic geographical subsystem, U2 is the effectiveness value of the resource and environment subsystem, U3 is the effectiveness value of the social and economic subsystem, α, β and γ are weighted weights, is the global effectiveness value. Here, the global effectiveness value aims to provide a weighted, comprehensive, aggregated indicator representing the overall advantages or potential value of industrial land. Its goal is to aggregate the independent contributions of each subsystem into a quantitative representation of overall effectiveness. By introducing adjustable weight parameters, the model can flexibly adjust the contribution of each subsystem to the overall evaluation based on different industry needs, regional development priorities, or policy orientations, thereby adapting to diverse assessment objectives.

[0102] More specifically, in an embodiment of the present application, based on the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and socio-economic subsystem efficacy value of the first industrial land, a global efficacy value and a global coupling degree are calculated, including: calculating the global coupling degree using the following global coupling degree calculation formula based on the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and socio-economic subsystem efficacy value of the first industrial land, the global coupling degree calculation formula is:

[0103]

[0104] Among them, U1 is the effectiveness value of the basic geographical subsystem, U2 is the effectiveness value of the resource and environment subsystem, and U3 is the effectiveness value of the social and economic subsystem. Global coupling is the degree of global coupling. It aims to quantify the degree of synergy and balance in the mutual influence and constraints among subsystems. Its goal is to reveal the harmony and inherent resilience of the different dimensions of industrial land. A high degree of coupling indicates that the subsystems function well together, contributing to overall efficiency, rather than a single subsystem being unable to support itself or becoming a bottleneck. This is crucial for the long-term healthy development of industrial land and to avoid the lame duck effect.

[0105] Specifically, in step S400, the coordination degree of the first industrial land is calculated based on the global efficacy value and global coupling degree. It should be understood that the global efficacy value represents the comprehensive performance or potential of the land across the three major systems of basic geography, resources and environment, and socio-economics, while the global coupling degree measures the degree of interaction and coordination among these subsystems. While both indicators individually provide important insights, individually, they are insufficient to provide a comprehensive and accurate assessment of the overall optimal matching or long-term sustainable development potential of industrial land. A piece of land may have an extremely high global efficacy value, but if there are serious imbalances or bottlenecks between its subsystems (i.e., low coupling), its true carrying capacity and development value will be greatly reduced. Conversely, even if the subsystems are highly coordinated and unified (high coupling), if the overall potential is insufficient (low global efficacy value), it will not be able to support efficient industrial development. The coordination degree of the first industrial land is further calculated. More specifically, in an embodiment of the present application, the coordination degree of the first industrial land is calculated based on the global efficacy value and the global coupling degree, including: calculating the coordination degree of the first industrial land based on the global efficacy value and the global coupling degree using the following coordination degree formula, wherein the coordination degree formula is:

[0106]

[0107] in, is the global coupling degree, is the global efficacy value, The coordination degree of land used for primary industry.

[0108] Specifically, in step S500, the matching industrial chain link for the first industrial land is determined based on the coordination degree of the first industrial land. It should be understood that the coordination degree is the highest-level, most comprehensive quantitative expression of the value and health of a piece of industrial land. However, its numerical value itself remains an abstract metric. To truly play its guiding role, it must be clearly associated with specific industrial formats (i.e., industrial chain links). Therefore, the matching industrial chain link for the first industrial land is determined based on the coordination degree of the first industrial land. Different industrial chain links have very different requirements for industrial land. For example, a research and development center may have extremely high requirements for the socioeconomic environment and talent density, but a higher tolerance for certain aspects of the geographical location or resource environment. Heavy industry or raw material processing may place greater emphasis on geographic carrying capacity, resource availability, and environmental capacity, while having relatively lower requirements for certain sophisticated socioeconomic supporting facilities. Simply having a high coordination degree does not directly indicate which industry a piece of land is most suitable for. The system needs to translate this high-dimensional, comprehensive land adaptability into specific guidance for specific industry needs, to achieve accurate docking of industrial resources and land resources, and to avoid blind investment or inefficient use. Specifically, in a specific example of this application, the coordination level is divided into: low coordination There is a serious disconnect between land use and industrial demand; moderate coordination Partial matching, needs optimization; benign coordination Dynamic balance between land use and industry; high degree of coordination Optimal resource allocation. Furthermore, industrial adaptation uses a decision tree model to match industry chain links according to coordination levels (e.g., high-coordination areas prioritize vehicle manufacturing).

[0109] In summary, the intelligent analysis method of industrial land based on the spatial coupling degree model according to the embodiment of the present application is explained. It first collects multi-source indicator data such as basic geography, resource environment and social economy of industrial land, and innovatively introduces low-dimensional embedded coding and aggregation analysis algorithms to perform intelligent dimensionality reduction, feature extraction and aggregation on the indicator data of each subsystem to generate an efficacy value that can accurately reflect the intrinsic characteristics of each subsystem. This process not only overcomes data redundancy and noise interference, but also deeply explores the nonlinear correlation between indicators. On this basis, by constructing a coupling weight matrix (integrating geographical and economic distances), quantifying the interaction intensity between different spatial units or elements, and combining the efficacy values ​​of each subsystem to calculate the global efficacy and global coupling degree, the overall coordination degree of the industrial land is obtained. Finally, based on this comprehensive coordination degree, the intelligent analysis of industrial land is realized, and the appropriate industrial chain links are accurately matched. Through this complete chain from deep mining of data features to quantification of multi-system coupling effects and then to coordinated development assessment, this solution effectively solves the problems of information fragmentation, one-sided evaluation and insufficient consideration of spatial interaction mechanisms in traditional analysis, and realizes scientific and intelligent judgment of the potential and suitability of industrial land.

[0110] Furthermore, an intelligent analysis system for industrial land based on a spatial coupling model is also provided.

[0111] Figure 5 FIG is a block diagram of an industrial land intelligent analysis system based on a spatial coupling degree model according to an embodiment of the present application. Figure 5 As shown, the industrial land intelligent analysis system 500 based on the spatial coupling model according to the embodiment of the present application includes: a primary industrial land index extraction module 510, which is used to extract the basic geographical index data, resource and environmental index data and socio-economic index data of the primary industrial land; a subsystem efficacy value calculation module 520, which is used to calculate the basic geographical subsystem efficacy value, resource and environmental subsystem efficacy value and socio-economic subsystem efficacy value of the primary industrial land based on the basic geographical index data, resource and environmental index data and socio-economic index data of the first industrial land; a global operator calculation module 530, which is used to calculate the global efficacy value and the global coupling degree based on the basic geographical subsystem efficacy value, resource and environmental subsystem efficacy value and socio-economic subsystem efficacy value of the first industrial land; a primary industrial land coordination degree calculation module 540, which is used to calculate the coordination degree of the first industrial land based on the global efficacy value and the global coupling degree; a matching industrial chain determination module 550, which is used to determine the matching industrial chain link of the first industrial land based on the coordination degree of the first industrial land.

[0112] As described above, the industrial land intelligent analysis system 500 based on the spatial coupling model according to the embodiment of the present application can be implemented in various wireless terminals, such as a server having an industrial land intelligent analysis algorithm based on the spatial coupling model. In one possible implementation, the industrial land intelligent analysis system 500 based on the spatial coupling model according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or hardware module. For example, the industrial land intelligent analysis system 500 based on the spatial coupling model can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the industrial land intelligent analysis system 500 based on the spatial coupling model can also be one of the many hardware modules of the wireless terminal.

[0113] While various embodiments of the present disclosure have been described above, the above descriptions are illustrative, non-exhaustive, and not intended to be limiting of the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is selected to best explain the principles of the embodiments, their practical applications, or improvements to existing technologies, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. An intelligent analysis method for industrial land based on a spatial coupling model, characterized in that: include: Extract basic geographical indicator data, resource and environmental indicator data, and socioeconomic indicator data for primary industrial land; Calculate the basic geographical subsystem efficacy value, resource and environmental subsystem efficacy value, and social and economic subsystem efficacy value of the primary industrial land based on the basic geographical indicator data, resource and environmental indicator data, and social and economic indicator data of the primary industrial land; Calculate the global efficacy value and global coupling degree based on the basic geographic subsystem efficacy value, resource and environmental subsystem efficacy value, and socio-economic subsystem efficacy value of the first industrial land; Calculating the coordination degree of the first industrial land based on the global efficacy value and the global coupling degree; Based on the coordination degree of the first industrial land, the matching industrial chain link of the first industrial land is determined.

2. The industrial land intelligent analysis method based on the spatial coupling model according to claim 1 is characterized in that: Based on the basic geographical indicator data, resource and environmental indicator data, and socio-economic indicator data of the first industrial land, the basic geographical subsystem efficacy value, resource and environmental subsystem efficacy value, and socio-economic subsystem efficacy value of the first industrial land are calculated, including: Calculating the weighted sum of each basic geographical indicator in the basic geographical indicator data of the first industrial land to obtain the basic geographical subsystem efficacy value; Calculating the weighted sum of each resource and environmental indicator in the resource and environmental indicator data of the first industrial land to obtain the resource and environmental subsystem efficacy value; The weighted sum of each socioeconomic indicator in the socioeconomic indicator data of the first industrial land is calculated to obtain the socioeconomic subsystem efficacy value.

3. The industrial land intelligent analysis method based on the spatial coupling model according to claim 2 is characterized in that: Calculating a global efficacy value and a global coupling degree based on the basic geographic subsystem efficacy value, the resource and environmental subsystem efficacy value, and the socio-economic subsystem efficacy value of the first industrial land includes: calculating the global efficacy value based on the basic geographic subsystem efficacy value, the resource and environmental subsystem efficacy value, and the socio-economic subsystem efficacy value of the first industrial land using the following global efficacy calculation formula, wherein the global efficacy calculation formula is: Among them, U1 is the effectiveness value of the basic geographical subsystem, U2 is the effectiveness value of the resource and environment subsystem, U3 is the effectiveness value of the social and economic subsystem, α, β and γ are weighted weights, is the global efficacy value.

4. The industrial land intelligent analysis method based on the spatial coupling model according to claim 2 is characterized in that: Calculating a global efficacy value and a global coupling degree based on the basic geographic subsystem efficacy value, the resource and environmental subsystem efficacy value, and the socio-economic subsystem efficacy value of the first industrial land, including: calculating the global coupling degree using the following global coupling degree calculation formula based on the basic geographic subsystem efficacy value, the resource and environmental subsystem efficacy value, and the socio-economic subsystem efficacy value of the first industrial land, wherein the global coupling degree calculation formula is: Among them, U1 is the effectiveness value of the basic geographical subsystem, U2 is the effectiveness value of the resource and environment subsystem, and U3 is the effectiveness value of the social and economic subsystem. is the global coupling degree.

5. The industrial land intelligent analysis method based on the spatial coupling model according to claim 1 is characterized in that: Calculating the coordination degree of the first industrial land based on the global efficacy value and the global coupling degree includes: calculating the coordination degree of the first industrial land based on the global efficacy value and the global coupling degree using the following coordination degree formula, wherein the coordination degree formula is: in, is the global coupling degree, is the global efficacy value, The coordination degree of land used for primary industry.

6. The industrial land intelligent analysis method based on the spatial coupling model according to claim 1 is characterized in that: Based on the basic geographical indicator data, resource and environmental indicator data, and socio-economic indicator data of the first industrial land, the basic geographical subsystem efficacy value, resource and environmental subsystem efficacy value, and socio-economic subsystem efficacy value of the first industrial land are calculated, including: Performing low-dimensional embedding coding on each basic geographical indicator of the first industrial land in the basic geographical indicator data to obtain a set of low-dimensional embedding coding vectors of the basic geographical indicators; Construct coupling weight matrix; Performing coupling optimization on each basic geographic indicator low-dimensional embedding coding vector in the set of basic geographic indicator low-dimensional embedding coding vectors based on the coupling weight matrix to obtain a set of optimized basic geographic indicator low-dimensional embedding coding vectors; Performing aggregation analysis on the set of basic geographic indicator low-dimensional embedded coding vectors to obtain a basic geographic indicator low-dimensional embedded aggregate coding vector; Feature decoding is performed on the low-dimensional embedded aggregate coding vector of the basic geographic indicator to obtain the basic geographic subsystem efficacy value.

7. The industrial land intelligent analysis method based on the spatial coupling model according to claim 6 is characterized in that: Construct the coupling weight matrix, including: Construct geographic distance matrix; Construct economic distance matrix; The coupling weight matrix is ​​generated based on the geographic distance matrix and the economic distance matrix.

8. The method for intelligent industrial land analysis based on a spatial coupling model according to claim 7, wherein the aggregation analysis is performed on the set of low-dimensional embedded coding vectors of basic geographic indicators to obtain a low-dimensional embedded aggregate coding vector of basic geographic indicators, comprising: Performing a core feature integration process based on cluster analysis on the set of low-dimensional embedded coding vectors of the basic geographic indicators to obtain a core integrated representation vector of the basic geographic indicator set; Based on the core integrated representation vector of the basic geographic indicator set, calculating the dynamic compensation representation vector of each basic geographic indicator low-dimensional embedded coding vector in the set of basic geographic indicator low-dimensional embedded coding vectors relative to the core integrated representation vector of the basic geographic indicator set to obtain a set of basic geographic indicator feature compensation dynamic representation vectors; Calculating the dynamic significance adjustment weight factor of each basic geographical indicator feature compensation dynamic representation vector in the set of basic geographical indicator feature compensation dynamic representation vectors to obtain a set of basic geographical indicator dynamic significance adjustment weight factors; Based on the set of basic geographical indicator dynamic significance adjustment weight factors, weighted modulation is performed on the set of basic geographical indicator feature compensation dynamic representation vectors to obtain a set of basic geographical indicator feature compensation dynamic representation significance coding vectors; The set of the basic geographic indicator feature compensation dynamic representation significant coding vectors and the basic geographic indicator set core integrated representation vector are input into the feature enhancement fusion network to obtain the basic geographic indicator low-dimensional embedding aggregation coding vector.

9. The method for intelligent industrial land analysis based on a spatial coupling model according to claim 8, wherein the core feature integration processing based on cluster analysis is performed on the set of low-dimensional embedded coding vectors of basic geographic indicators to obtain a core integrated representation vector of the basic geographic indicator set, including: Inputting the set of low-dimensional embedding coding vectors of the basic geographic indicators into a K-Means clustering network to obtain K basic geographic indicator initial cluster center coding vectors; Performing self-attention encoding on each of the K basic geographic indicator initial cluster center encoding vectors to obtain K basic geographic indicator self-attention encoding vectors; The K basic geographic indicator self-attention encoding vectors and the K basic geographic indicator initial cluster center encoding vectors are aggregated by position to obtain the core integrated representation vector of the basic geographic indicator set.

10. An industrial land intelligent analysis system based on a spatial coupling model, characterized in that: include: The primary industry land index extraction module is used to extract the basic geographical index data, resource and environmental index data and socio-economic index data of the primary industry land; a subsystem efficacy value calculation module, configured to calculate the basic geographic subsystem efficacy value, the resource and environmental subsystem efficacy value, and the socio-economic subsystem efficacy value of the primary industrial land based on the basic geographic indicator data, the resource and environmental indicator data, and the socio-economic indicator data of the first industrial land; A global operator calculation module, configured to calculate a global efficacy value and a global coupling degree based on the basic geographic subsystem efficacy value, the resource and environmental subsystem efficacy value, and the socio-economic subsystem efficacy value of the first industrial land; a primary industrial land coordination degree calculation module, configured to calculate the coordination degree of the first industrial land based on the global efficacy value and the global coupling degree; The matching industrial chain determination module is used to determine the matching industrial chain link of the first industrial land based on the coordination degree of the first industrial land.

Images