Resource capacity-oriented land space configuration optimization method and system

Abstract

The application discloses a land space configuration optimization method and system facing resource carrying capacity, and relates to the technical field of data processing.The method comprises the following steps: a continuous differentiable carrying field is constructed based on resource elements for a target land space; a space configuration module containing generation and optimization components is developed; an initial strategy is generated according to the dynamic carrying field and the carbon resistance surface, and is verified to determine the optimization instruction, and the optimization component is activated as needed to identify the intervention point and optimize; the scheme is displayed by hierarchical rendering and three-dimensional sectioning. The technical problem that the traditional configuration method relies on static threshold and discrete unit analysis, and it is difficult to accurately depict the spatial continuity and dynamic interaction of resource elements, resulting in insufficient adaptability of the configuration scheme to the resource carrying capacity is solved, the technical effect that the dynamic optimization is realized through the configuration generation and optimization component, the scheme intuitiveness and decision efficiency are improved by using hierarchical rendering and three-dimensional sectioning to generate a dynamic explosion diagram, and the fine and intelligent land space configuration is ensured.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and specifically to a method and system for optimizing land spatial allocation based on resource carrying capacity. Background Technology

[0002] In traditional land spatial planning and resource allocation, facing increasingly complex constraints of climate change and the need for multi-factor collaborative governance, existing methods often rely on static remote sensing image interpretation and discrete unit carrying capacity assessments, making it difficult to effectively characterize the dynamic coupling relationship between water resource pressure, carbon flow networks, and land suitability. Especially in complex scenarios involving inter-basin water transfer, urban agglomeration carbon sink interaction, and ecological red line delineation, the high spatial heterogeneity and nonlinearity of resource flows between land units lead to a lack of physical meaning in the continuous differentiable description of carrying capacity in existing planning models. The technical root cause lies in the failure to construct a continuous, differentiable resource field model, which cannot express the gradient changes and flow resistance of resource potential energy in space.

[0003] Furthermore, current planning often employs a one-way "generate-verify" process, with configuration optimization relying heavily on manual experience for localized fine-tuning. This fails to achieve dynamic game-theoretic optimization between the risk of carrying capacity overload and maximizing overall benefits. When addressing changes in carbon flow resistance or sudden tightening of resource constraints, the lack of sensitivity to key intervention points leads to low efficiency in optimization iterations and difficulty in converging to the global optimum. Simultaneously, existing results presentation methods typically rely on two-dimensional thematic maps, failing to intuitively represent the three-dimensional coupling relationship between underground resource carrying capacity layers, surface space utilization, and above-ground ecological corridors. Decision-makers, faced with multi-level spatial configuration schemes, cannot quickly backtrack, impacting the interpretability of multi-objective planning and the accuracy of interactive decision-making. Summary of the Invention

[0004] This application provides a method and system for optimizing land spatial allocation based on resource carrying capacity. The key is to address the technical obstacles that make it difficult to continuously quantify resource carrying capacity and dynamically optimize spatial allocation in land spatial planning and resource carrying capacity assessment scenarios, which are characterized by strong spatial heterogeneity, significant dynamic evolution, and complex multi-scale coupling of multi-source resource element data. By constructing a continuously differentiable resource carrying capacity field and introducing a spatial allocation optimization algorithm based on adversarial generation and negative entropy operators, and in conjunction with a spatial allocation processing flow oriented towards multi-source data fusion and dynamic updates, this application achieves refined characterization of resource carrying capacity, intelligent optimization of spatial allocation schemes, and efficient collaborative decision-making under multiple constraints.

[0005] The first aspect of this application provides a method for optimizing land spatial allocation based on resource carrying capacity, the method comprising: For the target land space, based on the defined resource elements of each land unit, resource field strength and resource potential energy are defined, and a continuously differentiable resource carrying capacity field is reconstructed. A spatial configuration module is developed on the land planning platform, which consists of a configuration generation component and a configuration optimization component. Based on the dynamically updated resource carrying capacity field, the configuration generation component generates an initial configuration strategy and performs lightweight verification using the carbon flow resistance surface as a condition, determines the configuration optimization instructions, and optionally activates the configuration optimization component. By identifying key intervention points and performing configuration optimization processing, a spatial configuration scheme is determined. The spatial configuration scheme is then rendered in layers and 3D sectioned to generate a dynamic exploded view, which is displayed on the terminal interface.

[0006] A second aspect of this application provides a land space allocation optimization system oriented towards resource carrying capacity, the system comprising: Resource Definition Module: For the target land space, based on the defined resource elements of each land unit, define resource field strength and resource potential energy, and reconstruct a continuously differentiable resource carrying capacity field; Module Development Module: Develop a spatial configuration module on the land planning platform, wherein the spatial configuration module consists of a configuration generation component and a configuration optimization component; Strategy Generation and Optimization Module: Based on the dynamically updated resource carrying capacity field, the configuration generation component generates an initial configuration strategy and performs lightweight verification with carbon flow resistance surface as a condition, determines the configuration optimization instruction, optionally activates the configuration optimization component, and determines the spatial configuration scheme by identifying key intervention points and performing configuration optimization processing; Scheme Rendering Module: Perform layered rendering and 3D sectioning of the spatial configuration scheme, generate a dynamic exploded view, and display it on the terminal interface.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: First, based on the resource elements of each unit in the target land space, resource field strength and resource potential energy are constructed, and a continuously differentiable resource carrying capacity field is further formed. Then, a spatial configuration module consisting of generation and optimization components is constructed in the land planning platform. Next, an initial configuration scheme is generated using dynamically updated carrying capacity fields and carbon flow resistance constraints. After verification, the optimization component is triggered as needed to intervene and adjust key areas, resulting in an optimized spatial configuration. Finally, the configuration results are visualized at multiple levels, outputting an intuitive and dynamic display effect through layered display and 3D sectioning, facilitating users' direct viewing of the optimization results. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 A schematic diagram of the land space allocation optimization method based on resource carrying capacity provided in the embodiments of this application.

[0010] Figure 2 A schematic diagram of the structure of a land space allocation optimization system based on resource carrying capacity provided in this application embodiment.

[0011] Explanation of reference numerals in the attached diagram: Resource definition module 11, Module development module 12, Strategy generation and optimization module 13, Scheme rendering module 14. Detailed Implementation

[0012] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0013] Example 1, as Figure 1 As shown, this application provides a method for optimizing land spatial allocation based on resource carrying capacity, wherein the method includes: For the target land space, based on the definition of resource elements of each land unit, the resource field strength and resource potential energy are defined, and a continuous and differentiable resource carrying capacity field is reconstructed.

[0014] In this embodiment, the target land space is first divided into gridded or plotted land units according to unified rules. Each of these units serves as the basic analysis object, and its corresponding resource element information, such as water resource pressure and carbon emission constraints, is extracted and integrated. Subsequently, various resource attributes are standardized and quantified. Then, by superimposing and weighting multiple resource attributes, a resource field strength reflecting the comprehensive resource status of each unit is constructed. Simultaneously, the spatial correlation between adjacent land units is combined to characterize the spatial transmission and influence of resources, and resource potential energy is defined to describe the interaction and constraint relationships between units. Finally, the resource field strength and resource potential energy are modeled in a unified manner to form a spatially continuously distributed resource carrying capacity field with differentiable characteristics, thus providing a mathematical basis for subsequent land space allocation based on gradient optimization, sensitivity analysis, and dynamic updates.

[0015] Furthermore, reconstructing a continuously differentiable resource carrying field includes: For each land unit in the target territory, resource attributes based on resource elements are defined, and resource field strength based on the superposition of resource attributes is determined. The resource elements include water resource pressure, carbon emission quota, and land suitability. The resource potential energy of neighboring land units relative to each land unit is defined. Based on the resource field strength and resource potential energy, the resource carrying capacity field of the target territory is defined.

[0016] Preferably, the target land area is first divided into basic spatial units. This division can be based on administrative parcels, regular grids, or planning zoning units, forming multiple independently analyzable land units. Each land unit corresponds to a clear spatial location, boundary range, and attribute record. Then, for each land unit, its corresponding resource element data is collected. These resource elements include at least water resource pressure, carbon emission quotas, and land suitability. After acquiring various resource elements, the collected raw data undergoes preprocessing, including data cleaning, missing value completion, and unit standardization. After preprocessing, the standard values ​​of resource elements for each land unit are defined as resource attributes. Next, to quantify the resource field strength of each land unit, the reserved weights for each resource attribute are read. These weights are set according to planning objectives; for example, water resources 0.4, carbon constraints 0.3, and land suitability 0.3. After reading the data, the water resource pressure value, carbon emission quota value, and land suitability value of each land unit are weighted and superimposed. For example, for a given unit, the standardized water resource pressure value is multiplied by the water resource weight, the standardized carbon emission quota value is multiplied by the carbon constraint weight, and the standardized land suitability value is multiplied by the land suitability weight. These products are then summed to obtain the resource field strength of that unit. The higher the resource field strength, the higher the carrying capacity. After defining the resource field strength, for each land unit, its neighborhood is determined using a fixed adjacency method (e.g., eight neighborhoods) or by setting an influence radius (e.g., all units within a 3km radius). For each land unit within the neighborhood, its influence value on the target land unit is calculated. This influence value is determined by a weighted average of two parts: the resource field strength of the neighboring units and the spatial distance between them and the target land unit. This spatial distance can be calculated using Euclidean distance or road distance, and a distance attenuation rule is set. Generally, the greater the distance, the smaller the influence. Then, the influence values ​​of all neighboring units are summed to obtain the resource potential energy of the target unit, which represents the intensity of the comprehensive effect of the surrounding environment. Finally, the resource field strength and resource potential energy are fused to construct a resource carrying capacity field. Specifically, the resource field strength of each unit is used as the base value, and its resource potential energy influence is linearly weighted and superimposed to form a new carrying capacity value. To avoid spatial abrupt changes, the results can be smoothed, for example, by using neighborhood averaging filtering or spatial interpolation methods to make the numerical changes between adjacent units more continuous. After the above processing, a continuously changing carrying capacity distribution surface will be formed throughout the entire national land space as a resource carrying capacity field. This resource carrying capacity field has no obvious spatial breaks and changes smoothly in value, thus possessing continuous differentiability, providing a direct data foundation and calculation basis for subsequent spatial configuration optimization.

[0017] A spatial configuration module is developed on the land planning platform, wherein the spatial configuration module consists of a configuration generation component and a configuration optimization component.

[0018] In one embodiment, a spatial configuration module for spatial configuration analysis and decision-making is developed and constructed within a land planning platform. This module, embedded as a core computing unit within the platform architecture, includes configuration generation and optimization components, and interacts with resource carrying capacity fields through a unified data interface. During development, the integration of spatial data and the model's operating environment is first completed on the platform side, including access to resource carrying capacity field data, basic geographic information data, and planning constraints, providing a unified data foundation for subsequent configuration calculations. Subsequently, based on the spatial distribution characteristics of resource carrying capacity fields and constraints such as carbon flow resistance surfaces, each land unit is initially classified into functional types. A configuration generation component that meets basic constraints is developed through training. This component is primarily used to quickly generate initial spatial configuration strategy schemes under given constraints. Furthermore, a spatial negative entropy operator is introduced, and a configuration optimization component is developed based on the critical transition between order and chaos. This component is used to further adjust and optimize the initial spatial configuration strategy schemes, thereby improving overall carrying capacity matching and spatial utilization efficiency. Finally, by combining and coordinating the configuration generation component and the configuration optimization component, the constructed spatial configuration module is equipped with the processing capability of first generating and then optimizing, realizing a complete process from initial configuration to optimized configuration. This solves the problems of low efficiency in generating initial configuration schemes, unreasonable configuration results, and difficulty in further fine-tuning caused by the complexity of multi-source data on national land space, diverse and coupled resource constraints, which make it difficult to systematically generate spatial configurations and lack a unified mechanism for the optimization process. This achieves the goal of automatically generating and iteratively optimizing spatial configuration schemes, improving the matching degree between configuration results and resource carrying capacity, and enhancing the scientific nature and decision-making efficiency of national land space planning.

[0019] Furthermore, the development space allocation module on the land planning platform includes: A carbon flow resistance surface is constructed, defined by the interaction characteristics of carbon sinks and carbon aggregates. These interaction characteristics include spatial connectivity, adjacency, and energy flow between carbon sink patches and carbon source land units. Using the resource carrying capacity as a baseline and the carbon flow resistance surface as a condition, a configuration generation component is constructed through adversarial supervised training, employing a combined architecture training method based on configuration generation and configuration judgment. A configuration optimization component is constructed by introducing a spatial negative entropy operator, aiming at identifying critical transition zones based on order and chaos and fine-tuning spatial configuration. This spatial negative entropy operator is a mathematical operator for the evolution of low-entropy states defined by land space boundaries, negative entropy sources, and flow channels. The configuration generation component and the configuration optimization component are combined to form a spatial configuration module.

[0020] Preferably, the land space is first classified based on land use type within the target land area. Forest land, grassland, wetlands, water bodies, and ecological corridor areas are marked as carbon sink patches, while construction land, industrial land, transportation land, and high-intensity development areas are marked as carbon source units. For carbon sink patches, the total carbon content can be calculated by multiplying the carbon sequestration coefficient per unit area by the patch area. The carbon sequestration coefficient per unit area can be obtained from a table or based on regional ecological assignment. The emission intensity of carbon source units can be spatially allocated through industry statistical data or carbon emission accounting results. Subsequently, carbon sink interaction characteristics are defined, including spatial connectivity, adjacency, and energy flow between carbon sink patches and carbon source land units. Specifically, connectivity analysis or minimum cost path analysis methods can be used to determine whether continuous paths exist between carbon sink patches and between carbon sink patches and carbon source units, and to count the number of connectable carbon sink patches and the length of connecting paths to determine spatial connectivity characteristics. Adjacency characteristics can be determined by constructing buffer zones (e.g., 500m or 1km) around carbon source units, counting the area proportion of carbon sink patches within the buffer zone, and combining this with topological contact thresholds to determine whether direct boundary contact exists. Finally, energy flow characteristics can be determined by combining topographic data (DEM), wind direction data, ecological corridor distribution, and water network data to determine the main direction and channel integrity of carbon exchange or ecological energy transfer. Subsequently, based on the interaction characteristics between total carbon sink and carbon sink, regions with strong carbon sink capacity and high connectivity are assigned low resistance values, while regions adjacent to carbon sources and with poor connectivity are assigned high resistance values. Furthermore, for units located on ecological corridors, water systems, or green transportation corridors, resistance is adjusted based on the integrity of the channel; generally, intact channels reduce resistance, while broken channels increase resistance. By spatially smoothing the resistance values ​​of each region, such as through a moving window averaging, a continuously distributed carbon flow resistance surface can be formed.

[0021] Next, using the resource carrying capacity as a baseline and the carbon flow resistance surface as a constraint, a configuration generation component is constructed. Specifically, using land units as basic sample units, the resource carrying capacity value, carbon flow resistance value, current land use type (using one-hot encoding), planning control boundary (e.g., constraint mask form), and neighborhood features are concatenated to form the feature vector of each sample. Simultaneously, a supervised training dataset is constructed using historically preferred planning schemes as labels. In terms of model structure, the configuration generation component is built upon a configuration generator and a configuration judge. The configuration generator employs a multi-layer feedforward neural network or lightweight convolutional network structure. Its input is the aforementioned feature vector, and its output is the configuration strategy for the corresponding space. Typically, the network has 3-5 layers, with hidden layers of 64 or 128 dimensions, and the activation function is ReLU. The configuration judge is a scoring model used to evaluate the rationality of the generated results. Its input is a combination of the generated results and the original features, and its output is a rationality score. During training, the generator is first pre-trained based on supervised learning, using a weighted cross-entropy loss function, and optimized through backpropagation of the error between the generated results and the true labels. After the configuration generator outputs its configuration results, the configuration judge is input to obtain scoring feedback, and a joint loss function is constructed, for example, a weighted result of the generation error loss and the judge's scoring constraint loss. During training and optimization, the Adam optimization algorithm is used, with the learning rate (e.g., 0.001) and batch size (e.g., 32 or 64) set for iterative updates. Through the above alternating optimization process, the generator not only fits historical schemes but also gradually learns the implicit spatial rationality rules. When training converges, the generator parameters are fixed, and the configuration judge is removed, retaining only the trained configuration generator as the configuration generation component.

[0022] Then, a spatial negative entropy operator is introduced. This operator is built upon three elements: national spatial boundaries, negative entropy sources, and circulation channels. National spatial boundaries can be selected from regional boundaries with organizational constraints, such as urban agglomeration boundaries, river basin boundaries, metropolitan area boundaries, ecological function zone boundaries, and administrative division boundaries. Negative entropy sources characterize spatial objects that can enhance regional order, strengthen spatial coordination, and stabilize resource flow, including ecological corridors, water systems, transportation networks, important green space systems, and regional public service frameworks. Circulation channels characterize the paths through which matter, energy, and information are exchanged, transmitted, and linked within the national spatial area, including rivers, water networks, road networks, wind corridors, and logistics channels. Based on these three elements, the spatial negative entropy operator is a set of rules for negative entropy injection, whose core function is to guide the formation and expansion of ordered structures within the spatial network. Specifically, the entire national territory is first abstracted into a spatial network structure composed of national land units, where each unit is a node and adjacent relationships or functional connections are edges. Then, based on the relationships between each node and the aforementioned boundaries, negative entropy sources, and flow channels, the current degree of order and chaos is calculated, thereby identifying the critical transition zone between order and chaos. These areas typically exhibit significant functional conflicts, fragmented spatial structures, discontinuous channels, or obstructed resource flow. Next, negative entropy injection is performed on these critical transition zones. That is, based on the distance relationship between each unit and the negative entropy source, the degree of coupling with flow channels, and its position within the boundary structure, the spatial location and intensity of negative entropy injection are determined. For example, for areas near ecological corridors but blocked by construction land, the intensity of negative entropy injection can be increased to guide them towards ecological or low-intensity utilization; for areas located at transportation backbone nodes but with chaotic spatial organization, negative entropy injection can strengthen their functional agglomeration and structural order; for areas at the intersection of urban clusters or watershed boundaries, negative entropy injection can optimize the boundary transition morphology and reduce abrupt changes and conflicts. In the specific adjustment process, negative entropy injection will be applied to spatial configuration parameters, including but not limited to the functional type allocation of land units, the shape and location of spatial boundaries, and the connectivity and accessibility of circulation channels. By adjusting these parameters, the local spatial structure can be reconstructed, for example, by repairing broken ecological corridors, opening up obstructed water systems or transportation channels, optimizing the transition relationship between construction land and ecological land, and reducing the development intensity of high-conflict areas.

[0023] The aforementioned negative entropy injection process is progressively advanced through multi-level iterations. In each iteration, the degree of orderliness of each unit and the overall negative entropy level of the system are reassessed based on the current spatial configuration results. The location and intensity of the negative entropy injection are dynamically adjusted, allowing the spatial structure to gradually evolve from a disordered or inefficient state to a highly efficient and ordered state. During the iteration process, maximizing the overall negative entropy is taken as the optimization objective, that is, guided by improving the overall connectivity, coordination, and stability of the spatial system. When the increase in negative entropy tends to stabilize or reaches a preset threshold after several consecutive iterations, the negative entropy injection distribution corresponding to the current spatial configuration state is considered to be in a relatively optimal state. Finally, through the aforementioned negative entropy injection and iterative optimization process based on the spatial negative entropy operator, a configuration optimization component with the spatial negative entropy operator as its core is constructed, which can automatically identify key regions and make fine-tuning adjustments, thereby achieving an overall spatial structure order improvement driven by local fine-tuning. Finally, the configuration generation component and the configuration optimization component are integrated to form a spatial configuration module, which enables the initial configuration results to be generated quickly under the premise of meeting resource carrying capacity and carbon flow constraints. Through refined optimization of key areas, the overall spatial structure is coordinated and orderly, thereby effectively improving the rationality, stability and comprehensive benefits of land spatial configuration.

[0024] Based on the dynamically updated resource carrying capacity, the configuration generation component generates an initial configuration strategy and performs lightweight verification based on the carbon flow resistance surface, determines the configuration optimization instructions, and optionally activates the configuration optimization component. By identifying key intervention points and performing configuration optimization processing, the spatial configuration scheme is determined.

[0025] In one embodiment, based on the continuous dynamic updating of the resource carrying capacity field, a configuration generation component first participates in spatial configuration calculation. Specifically, the configuration generation component uses the current resource carrying capacity field as the basic input, and introduces a carbon flow resistance surface as an important constraint to characterize the ease or difficulty of carbon flow between different spatial units and their spatial connectivity characteristics. Based on this, the component performs functional allocation and layout deduction for each land unit using the adversarial responses of multiple stakeholders in a land planning scenario, thereby forming a set of initial configuration strategies. After the initial configuration strategies are generated, they undergo lightweight verification processing. This verification does not involve complex optimization calculations, but rather uses rapid rule verification to preliminarily screen the schemes, for example, determining whether they meet the requirements for carrying capacity entropy change and coupled oscillation. Based on the verification results, if any preset condition is found not to be met, a corresponding configuration optimization instruction is generated to indicate the area or direction requiring further optimization. Once the configuration optimization command is generated, it activates the configuration optimization component. Based on the initial configuration strategy, this component analyzes the impact of each land unit on the overall objective. By assessing the changing trends of carrying capacity, it identifies key intervention points that significantly impact the overall spatial pattern—those areas where adjustments can substantially improve overall carrying capacity matching or resource utilization efficiency. Subsequently, using these key points as the core, the functional types or spatial layouts of relevant land units are adjusted, and the configuration results are gradually optimized through iteration. Ultimately, after completing the adjustments to key areas, a spatial configuration scheme that is overall coordinated, meets resource carrying capacity constraints, and possesses high utilization efficiency is formed. This effectively reduces the risk of resource overload, improves the continuity and orderliness of spatial layout, and enhances the scientific rigor and feasibility of the land spatial configuration results.

[0026] Furthermore, the resource carrying capacity is dynamically updated based on changes in the target land space, re-division of land units, and changes in resource attributes and resource potential energy; based on the configuration generation component, an initial configuration strategy is generated with the adversarial response of multiple subjects in the land planning scenario; the initial configuration strategy is lightly verified, and the configuration optimization component is activated and the strategy is optimized based on the verification results.

[0027] Preferably, during the operation of the land planning platform, it continuously receives updated information from multiple data sources, including but not limited to adjustments in the scope of land space, changes in administrative or planning boundaries, re-division of land units, updates to resource element data, and changes in neighboring relationships. When any type of data changes, the resource carrying capacity field update mechanism is automatically triggered. This recalculates the resource attributes, resource field strength, and neighboring resource potential energy of the affected land units, and partially or entirely reconstructs the carrying capacity field of the entire region. Simultaneously, smoothing processing ensures the spatial continuity of the carrying capacity field before and after the update, ensuring that each land unit always corresponds to its current carrying capacity value, providing a real-time basis for subsequent configuration generation. After the resource carrying capacity field is updated, the configuration generation component is invoked to generate the initial configuration strategy. Unlike traditional single-rule-driven systems, the system introduces a multi-stakeholder adversarial response mechanism based on the land planning scenario, incorporating the behavioral preferences and decision-making tendencies of different stakeholders into the configuration generation process. Specifically, different types of space use needs are first abstracted into several behavioral subjects, such as public management orientation, industrial development orientation, and residential and ecological orientation. Among them, public management orientation mainly reflects preferences for regional development goals and balanced public services; industrial development orientation mainly reflects preferences for land development intensity, location benefits, and transportation accessibility; and residential and ecological orientation mainly reflects preferences for residential stability, ecological environment quality, land rights protection, and living convenience. Based on this, corresponding preference rules are defined for each type of subject and mapped to the national land unit level.

[0028] Subsequently, during the configuration generation process, for each land unit, a multi-agent adversarial decision-making process is constructed by comprehensively considering resource carrying capacity constraints, carbon flow resistance surface constraints, and the preferences of multiple stakeholders. Specifically, this can be achieved by iteratively simulating the decision-making competition among stakeholders on that unit within the configuration generation component. For example, an industrial development orientation tends to allocate highly accessible areas as construction land, a public management orientation tends to optimize industrial and spatial structures while meeting carrying capacity constraints, and a residential and ecological orientation tends to maintain the status quo or improve the ecological environment. The system determines the optimal or suboptimal functional type allocation result for that unit under current conditions by weighted coordination of the stakeholders' preferences, conflict resolution, or adversarial game solving. By executing the above process on all land units, an initial configuration strategy incorporating the multi-agent adversarial results is generated, ensuring that the strategy not only meets resource constraints but also reflects the balance between different stakeholders.

[0029] After obtaining the initial configuration strategy, a lightweight verification is performed. This lightweight verification uses the first and second verification conditions as the basis to evaluate the carrying capacity and multi-factor oscillations of the initial configuration strategy. If the verification results show that all verification conditions are met, there is no need to enter the optimization stage, and it can be directly output as a spatial configuration scheme. If one or more verification conditions are not met, a configuration optimization instruction is generated, and the configuration optimization component is activated according to the instruction. The configuration optimization component then focuses on the problem area, further optimizing the initial strategy by identifying key intervention points, adjusting functional allocation, optimizing boundary morphology, and improving channel connectivity, thereby obtaining the final spatial configuration scheme. Through the above process, dynamic response of the resource carrying capacity field, initial configuration generation driven by multi-stakeholder confrontation, and adaptive optimization decision-making based on verification results are realized. This enables the spatial configuration result to achieve a dynamic balance between resource constraints and the interests of multiple parties, improving the rationality, stability, and feasibility of the planning scheme.

[0030] Furthermore, the first verification condition is the entropy change of carrying capacity in the target land space, and the second verification condition is the coupled oscillation of multiple spatial elements. If at least one of the first and second verification conditions is not met, a configuration optimization instruction is generated.

[0031] Optionally, after generating the initial configuration strategy, an entropy change analysis is first performed on the resource carrying capacity of the target land space. The carrying capacity distribution corresponding to the current spatial configuration result is compared with the baseline state, such as the updated resource carrying capacity field, to characterize the degree of change of the system from an ordered state to a disordered state. Specifically, the carrying capacity utilization level can be extracted for each land unit. This carrying capacity utilization level is the ratio of resource occupancy intensity to carrying capacity. Statistical distribution analysis is then performed across the entire region to determine whether there is excessive dispersion in carrying capacity distribution, localized concentrated overload, or exacerbation of overall imbalance. If the analysis results indicate that the overall trend is towards disorder, the carrying capacity pressure distribution is deteriorating, or the entropy value in a local area is abnormally high, then the first verification condition is deemed not met. Subsequently, an oscillation analysis was conducted on the spatial multi-element coupling state. This multi-element includes key spatial elements such as ecological land, agricultural land, construction land, and carbon transport pathways. For each land unit and its neighboring relationships, the degree of spatial coupling and the trend of change among different elements were analyzed. Specifically, by comparing the frequency of functional type changes, the degree of boundary morphology changes, and the continuity of channels between adjacent units, it was possible to identify whether there were frequent switching, repeated occurrence of local conflicts, or structural instability. For example, if ecological and construction land are alternately distributed and constantly adjusted in a certain area, channels are repeatedly blocked and restored, and functional boundaries continue to oscillate, it can be determined that the spatial coupling oscillation is obvious. If such oscillation exceeds a preset threshold, the second verification condition is determined to be unsatisfactory. After completing the above two types of verification, if both the first and second verification conditions are met, it indicates that the current initial configuration strategy is in a relatively stable and reasonable state in terms of load distribution and spatial structure, and further optimization is not required. If either verification condition is not met, a configuration optimization instruction is generated. This configuration optimization instruction includes at least the problem type identifier, the corresponding spatial location range, and the degree of impact. The instruction is then passed to the configuration optimization component to trigger subsequent optimization processing for key areas.

[0032] Furthermore, by identifying key intervention points and performing configuration optimization processes, including: According to the configuration optimization instruction, the configuration optimization component is activated; for each national land unit, the sensitivity gradient of the bearing state to the global bearing target is measured based on the bearing state and partial derivative; according to the direction of the sensitivity gradient, key intervention points are identified for the initial configuration strategy, wherein key intervention points are determined based on the national land unit with the contribution to the improvement of global bearing capacity; and configuration optimization processing is performed with the key intervention points as the intervention subjects.

[0033] Optionally, after generating the configuration optimization instruction through lightweight verification, the information in the configuration optimization instruction is first read, including the problem type identifier, corresponding spatial location range, and impact level, and the configuration optimization component is activated accordingly. Then, for each land unit within the target land space, its sensitivity gradient to the global carrying capacity objective is calculated one by one. Specifically, the current carrying capacity utilization level of the land unit is extracted first, and its functional type, the functional distribution of neighboring units, and their position in the resource carrying capacity field are recorded. Then, a perturbation simulation is performed on the unit; that is, without changing the overall structure, a small adjustment is made to the unit, for example, temporarily changing its functional type to one of the alternative types, or reducing / increasing its development intensity by a preset level, and the adjusted global carrying capacity is recalculated. Next, the ratio of the change in global carrying capacity before and after the adjustment to the change in unit configuration is calculated to obtain an approximate value of the partial derivative of the global carrying capacity objective function with respect to the configuration variable of the land unit, thereby characterizing the degree of impact of the unit change on the overall system. By performing the above-mentioned perturbation simulation on multiple candidate adjustment directions, such as from construction to ecology, from construction to agriculture, and from agriculture to ecology, and calculating the corresponding partial derivative approximations, the impact of the unit under different adjustment directions can be determined. Thus, the optimal adjustment direction and the corresponding impact magnitude can be extracted, and the optimal adjustment direction can be used as the direction of the sensitivity gradient, and the corresponding partial derivative approximation can be used as the magnitude of the sensitivity gradient.

[0034] After calculating the sensitivity gradients of all land units, units with positive sensitivity gradients are selected, meaning their adjustments can improve the overall carrying capacity. These units are then ranked from largest to smallest impact. This ranking is then overlaid with the problem areas in the configuration optimization instructions, prioritizing units located in carrying capacity over-limit zones, structural conflict zones, or channel blockage zones, and with larger sensitivity gradients, as key intervention points. Furthermore, to avoid overly dispersed optimization, spatial clustering rules can be set to merge spatially adjacent or functionally related groups of highly sensitive units into one intervention area, forming a set of key intervention points. These points are then categorized and labeled according to their contribution to improving the overall carrying capacity. After identifying the key intervention points, configuration optimization is performed using these points as the intervention subjects, through steps such as low-entropy flow evolution, spatial configuration fine-tuning, and optimization result fitting. The final spatial configuration scheme is output, which, compared to the initial configuration strategy, achieves targeted optimization in key areas, significantly reduces the risk of carrying capacity over-limit, and improves the continuity and coordination of the spatial structure, thus realizing refined spatial configuration optimization oriented towards resource carrying capacity.

[0035] Furthermore, taking the aforementioned key intervention points as the intervention subjects, configuration optimization processing is performed, including: For the intervention subject, a first configuration optimization result is determined by low-entropy flow evolution based on the negative entropy flow operator, wherein the optimization iteration is guided by the spatial configuration mode based on maximizing negative entropy; for the intervention subject, a second configuration optimization result is determined by locating the critical transition zone and fine-tuning the spatial configuration; the spatial configuration scheme is determined by fitting the first configuration optimization result and the second configuration optimization result.

[0036] Preferably, after identifying key intervention points, each intervention subject and its neighboring related units are extracted to form a local optimization sub-region. The current spatial configuration status, resource carrying capacity, boundary relationships, circulation channel relationships, and negative entropy source distribution of this sub-region are used as inputs to obtain a first configuration optimization result and a second configuration optimization result. The first configuration optimization result focuses on improving overall orderliness by using the low-entropy flow evolution of the negative entropy flow operator to perform directional optimization of the region where the intervention subject is located. The second configuration optimization result focuses on correcting local boundary conflicts and structural abrupt changes by identifying critical transition zones and fine-tuning spatial configuration to perform refined correction of the region where the intervention subject is located. Specifically, for the intervention subject, the information on the land space boundary, negative entropy source, and circulation channel within the local optimization sub-region where the intervention subject is located is first read. The land space boundary can include urban agglomeration boundaries, watershed boundaries, functional zoning boundaries, and rigid control boundaries. Negative entropy sources can include ecological corridors, water system networks, transportation frameworks, and other structural sources that can improve spatial organization and orderliness. Circulation channels can include material exchange paths, energy transmission paths, information connection paths, and composite corridors. Subsequently, based on the spatial relationships between the intervention subject and the aforementioned elements, negative entropy injection rules are constructed to determine where to inject negative entropy from, in which direction it is transmitted, the injection intensity, and which units it preferentially affects within a local sub-region. In this process, the strength of the connection between the intervention subject and the nearest negative entropy source, the degree of coupling with the main flow channels, and the degree of conflict with boundary constraints can be assessed first. If the intervention subject is close to an ecological corridor, main waterway, or main transportation network, and the current spatial configuration hinders the continuation of such structures, the negative entropy injection intensity in the corresponding direction is increased, guiding the region towards a more continuous and orderly configuration pattern. If several pre-subjects are located in boundary fragmentation zones, functional conflict zones, or high-resistance blocking locations, negative entropy allocation is preferentially carried out along directions where connectivity can be restored, and the tendency to allocate in directions that would cause further fragmentation is reduced. Next, multiple rounds of low-entropy flow evolution iterations are conducted on the intervention subject and its neighboring units. In each iteration, the amount of negative entropy injected and the transmission path are reallocated based on the current configuration, and the functional types, boundary morphologies, channel orientations, and local structural combinations of units within the region are adjusted simultaneously. For example, construction units that block ecological corridors are adjusted to ecological buffer units, land boundaries that disrupt water system continuity are set back, and construction units that spread disorderly along the transportation framework are reorganized into a more compact layout. After each round of adjustment, the improvement in the degree of orderliness of the local sub-region, the improvement in the continuity of the circulation channels, and the degree of coordination with the global carrying capacity are recalculated. If the adjustment in this round increases the level of negative entropy, the result of this round is retained and iteration continues; if it does not improve or new conflicts arise, the adjustment in this round is reverted and a suboptimal injection direction is used or the injection intensity is reduced.Through multiple iterations, the spatial configuration pattern that maximizes negative entropy is gradually approximated, and the local configuration result corresponding to this pattern is recorded as the first configuration optimization result.

[0037] On the other hand, for the intervention subject, within the local optimization sub-region where the intervention subject is located, the critical transition zone between the ordered state and the chaotic state is identified. That is, by analyzing the degree of functional differences, boundary mutation degree, resource carrying capacity gradient changes, spatial connectivity changes and conflict frequency of each land unit in the region, the transition zones that are located at the strong intersection of two or more types of functions, have rapid structural changes, unstable configuration states and are prone to conflict spread are identified. For example, the sharp contact boundary between construction land and ecological land, the fragmented zone formed by agricultural land being cut by construction patches, the conflict zone formed by the superposition of the urban agglomeration expansion front and the watershed ecologically sensitive area, and the fluctuation zone formed by functional incoordination around the water system corridor can all be identified as critical transition zones. Subsequently, spatial configuration fine-tuning is performed on the identified critical transition zones. This fine-tuning refers to making small, continuous, and directional corrections to unit categories, boundary positions, patch combinations, and channel connectivity at a local scale without significantly altering the overall layout framework. Specifically, this may include converging overly prominent construction boundaries, filling in fragmented ecological buffer zones, merging broken agricultural units, replacing individual high-conflict units that obstruct circulation channels, and inserting buffer layers or setting gradient transitions at functional interfaces. During the fine-tuning process, each adjustment only affects the main intervention entity and one or more adjacent layers of units with limited amplitude. After each adjustment, the boundary smoothness, conflict reduction, coupling coordination, and improvement in the carrying capacity of the transition zone are immediately recalculated. If the oscillation of the critical transition zone decreases, the boundary becomes smoother, and the functional connection becomes more natural after the fine-tuning, the result of that fine-tuning is retained; otherwise, it is revoked, and other fine-tuning methods are tried. Through several rounds of continuous local fine-tuning, a configuration result that is superior in terms of boundary connectivity, structural continuity, and local stability is obtained and recorded as the second configuration optimization result.

[0038] After obtaining the first and second configuration optimization results, the two types of results are aligned, that is, mapped to the same set of land units, the same spatial boundary benchmark, and the same functional classification system. Then, the configuration differences between the two types of results are compared unit by unit. For units with consistent results, the consistent configuration is directly retained. For units with inconsistent results, they are fused according to a preset fitting rule. This fitting rule may include: prioritizing the retention of configuration choices that contribute more to the improvement of global negative entropy, while also taking into account the local stability benefits brought by the fine-tuning of the critical transition zone; for units located in the main channel, main boundary, key ecological node, or highly sensitive carrying capacity area, priority is given to the global orderliness improvement reflected in the first configuration optimization result; for units located at the boundary interface, local conflict frequent point, or structural abrupt point, priority is given to the local smoothing and buffering effect reflected in the second configuration optimization result; for units with similar contributions, a weighted approach is used to compromise and obtain a final configuration that takes into account both overall orderliness and local coordinability. After fitting is completed, the fitting results are comprehensively checked to see if they exceed rigid boundaries, cause new resource overload, disrupt the continuity of the main channel, or reduce the stability of critical regions. If the check passes, the fitting result is determined as the spatial configuration scheme. If the check fails, the fitting unit with conflict is backtracked, and the fitting values ​​of that part are readjusted until the constraints are met. Through the above process, the first configuration optimization result, which focuses on global ordered evolution, and the second configuration optimization result, which focuses on local fine-tuning, can be effectively integrated. This ensures that the spatial configuration responds to the goal of maximizing negative entropy while also taking into account the stable transition and boundary coordination of the critical transition zone, thus obtaining a final spatial configuration scheme that combines overall integrity and fine-tuning.

[0039] The spatial configuration scheme is rendered in layers and sectioned in three dimensions to generate a dynamic exploded view, which is then displayed on the terminal interface.

[0040] In one embodiment, after obtaining the spatial configuration scheme, the spatial information of above-ground, surface, underground, and different scale ranges is first classified and organized according to the functional type, resource carrying capacity, optimization and adjustment results, and mutual coupling relationships of each land unit in the spatial configuration scheme. Then, it is drawn layer by layer according to preset display rules to form a visual layer that can reflect the spatial structural characteristics of different levels. On this basis, for key intervention areas, important passage nodes, functional boundary areas, or areas with significant changes in resource carrying capacity, three-dimensional sections are performed along selected directions or selected locations to deconstruct and display the multi-layered structure that was originally superimposed in the same space, so that the underground resource support relationship, surface land use organization relationship, and above-ground ecological or corridor connection relationship can be intuitively distinguished. Afterwards, based on the sectioning results, each layer is stretched, separated, staggered, and dynamically associated to generate a dynamic exploded map that can reflect the inter-layer correspondence, adjustment path, and spatial coupling state. Finally, the dynamic exploded view is displayed on the terminal interface, allowing users to intuitively view the structural features and optimization results of the spatial configuration scheme at different levels, locations, and time sequences through playback, zooming, rotation, layer switching, or selection viewing, thereby improving the readability, display effect, and decision support capabilities of the spatial configuration scheme.

[0041] Furthermore, the spatial configuration scheme is rendered in layers and 3D sectioned to generate a dynamic exploded view, which is then displayed on the terminal interface, including: For the aforementioned spatial configuration scheme, resource configuration states at different spatial scales are rendered in layers to determine a layered view. The different spatial scales include macro-level watersheds, meso-level urban clusters, and micro-level plots. For the layered view, a three-dimensional section is performed along key intervention points to generate a dynamic exploded map. The sectioning is based on the spatial coupling relationship and negative entropy flow path of the underground resource carrying layer, the surface space utilization layer, and the above-ground ecological corridor layer. The dynamic exploded map is then displayed on a terminal interface.

[0042] Preferably, after obtaining the spatial configuration plan, the plan is first divided by scale and organized by structure. The functional types, resource carrying capacity, optimization and adjustment information, and spatial relationships of each land unit are uniformly coded and organized hierarchically according to different spatial scales. At the macro scale, water resource allocation patterns, ecological security patterns, and overall carrying capacity distribution are summarized, forming a macro-basin level resource allocation view, using a watershed or region as the unit. At the meso scale, construction land layout, industrial spatial structure, transportation backbone, and ecological network structure are extracted, forming a meso-city cluster level spatial organization view, using urban agglomerations or functional zones as units. At the micro scale, the functional types, carrying capacity changes, and optimization and adjustment results of each unit are displayed in detail, forming a micro-plot level refined configuration view, using specific plots or subdivided land units as units. Subsequently, based on data at different scales, each level is graphically rendered, including color mapping, symbol labeling, boundary depiction, and transparency settings, so that spatial structural features at different scales can be clearly distinguished and superimposed to form a layered visual view. After constructing the layered visualization, the identified key intervention points are selected as the cutting center or cutting path for 3D cutting processing. Specifically, the cutting direction and range are first determined in the layered visualization. Cutting surfaces can be established along the main channel direction, boundary transition direction, or direction of significant change in bearing gradient where the key intervention points are located. Then, the underground resource carrying capacity layer, surface space utilization layer, and above-ground ecological corridor layer are used as the basic layered units of the 3D structure, spatially separating the multi-layered information that was originally superimposed on the same plane. During the cutting process, the connection relationship between different layers is synchronously marked according to the spatial coupling relationship and negative entropy flow path between each layer. For example, it shows how underground resource support affects the distribution of surface land use, how surface land use connects with above-ground ecological corridors, and the transmission path of negative entropy flow between different layers. Afterward, the cut layers are stretched, separated, and staggered to create gaps between each layer in the vertical direction. The interlayer association is maintained by connecting lines, paths, or flow direction markers, thereby generating a dynamic exploded diagram with interlayer correspondence and structural evolution characteristics. Finally, the dynamic exploded diagram is loaded onto the terminal interface for display, and interactive control functions are provided in the visualization terminal, including layer switching, scale switching, section position adjustment, view rotation, zoom browsing, and time series playback. This allows users to view the spatial configuration status at different levels (macro, meso, and micro) as needed, and intuitively observe the coupling relationships and optimization paths between multi-layer structures at key intervention points. Through the above process, the spatial configuration scheme is transformed from a two-dimensional static expression into a multi-scale, multi-level, interactive three-dimensional dynamic display, thereby improving the interpretability of the results and their decision support capabilities.

[0043] Furthermore, the method also includes: In response to the layer selection command from the user, the load-bearing gradient change curve and negative entropy contribution ranking of key intervention points within the target layer are dynamically associated; based on the load-bearing gradient change curve and negative entropy contribution ranking, the spatial configuration parameters and optimization adjustment paths are traced back and highlighted for display on the user interface.

[0044] Preferably, after the spatial configuration scheme is displayed hierarchically on the terminal interface, a layer selection instruction is received from the user. This instruction allows the user to click or check a specific spatial scale layer or a specific functional thematic layer on the interface. Upon receiving the instruction, the system parses the selected target layer, extracts the corresponding set of land units and the set of marked key intervention points within that layer, and retrieves the historical calculation data recorded during the optimization process for these key intervention points. Subsequently, for each key intervention point, its carrying capacity gradient change curve is dynamically generated. Specifically, the carrying capacity state value, sensitivity gradient value, and impact value on the global carrying capacity of the point before and after each round of optimization are extracted from the stored optimization iteration process records and arranged in chronological order to form a continuously changing data sequence. Based on this data sequence, a corresponding carrying capacity gradient change curve is generated to reflect the change trend of the point from the initial state to the final state during the optimization process, such as a gradual decrease in carrying capacity pressure and a gradual increase in carrying capacity equilibrium. Simultaneously, the role of each key intervention point in the negative entropy optimization process is quantitatively evaluated. Specifically, by comparing the change in the overall spatial negative entropy state value before and after the adjustment of each point, its contribution to the overall negative entropy improvement in each round of optimization is calculated. All points are then ranked to obtain a negative entropy contribution ranking, reflecting the importance of each point in improving spatial orderliness. After obtaining the bearing gradient change curve and the negative entropy contribution ranking, a backtracking analysis is performed based on this information. Starting from the key intervention points in the selected layer, the adjustment process in each round of optimization is gradually traced back along their corresponding optimization record links. This includes changes in the point's functional type, adjustments in development intensity, corrections to boundary morphology, improvements in channel connectivity, and adjustments in linkage with neighboring units. Simultaneously, the parameter changes corresponding to each adjustment are extracted and organized. After the backtracking is completed, the aforementioned spatial configuration parameters and optimization adjustment paths are highlighted on the terminal interface. For example, color gradients are used to represent changes in load-bearing status, arrows or lines are used to represent adjustment directions and paths, and annotation information is used to display the changes in key parameters for each round of optimization. The corresponding load-bearing gradient change curves and negative entropy contribution rankings are then simultaneously displayed on the side of the interface or in a pop-up window, allowing users to intuitively view the optimization effects of different points and their impact on the overall spatial configuration. Through this process, users can achieve traceable analysis of the spatial configuration optimization process and visualize key information, enabling them to clearly understand the mechanisms of action and optimization paths of key intervention points, thereby improving the system's interpretability and interactive analysis capabilities.

[0045] In summary, the embodiments of this application have at least the following technical effects: First, for the target land space, based on the defined resource elements of each land unit, resource field strength and resource potential energy are defined to reconstruct a continuously differentiable resource carrying capacity field. Then, a spatial configuration module is developed on the land planning platform, consisting of a configuration generation component and a configuration optimization component. Next, based on the dynamically updated resource carrying capacity field, the configuration generation component generates an initial configuration strategy and performs lightweight verification, using the carbon flow resistance surface as a condition, to determine configuration optimization instructions. Optionally, the configuration optimization component is activated, and by identifying key intervention points and performing configuration optimization processing, a spatial configuration scheme is determined. Finally, the spatial configuration scheme is rendered in layers and 3D sectioned to generate a dynamic exploded view, which is then displayed on the terminal interface. This invention addresses the technical problem that traditional land spatial allocation methods rely on static thresholds and discrete unit analysis, making it difficult to accurately depict the spatial continuity and dynamic interaction of resource elements, resulting in insufficient adaptability between allocation schemes and resource carrying capacity. It achieves the technical effect of constructing a continuous and differentiable resource carrying field to accurately depict resource distribution, realizing dynamic optimization through configuration generation and optimization components, and using layered rendering and 3D sectioning to generate dynamic exploded maps to improve the intuitiveness of the scheme and decision-making efficiency, thus ensuring the refinement and intelligence of land spatial allocation.

[0046] Example 2, based on the same inventive concept as the resource-carrying capacity-oriented land space allocation optimization method in the foregoing examples, such as... Figure 2 As shown, this application provides a land spatial allocation optimization system based on resource carrying capacity, wherein the system includes: Resource Definition Module 11: For the target land space, based on the defined resource elements of each land unit, define resource field strength and resource potential energy, and reconstruct a continuously differentiable resource carrying capacity field; Module Development Module 12: Develop a spatial configuration module on the land planning platform, wherein the spatial configuration module consists of a configuration generation component and a configuration optimization component; Strategy Generation and Optimization Module 13: Based on the dynamically updated resource carrying capacity field, the configuration generation component generates an initial configuration strategy and performs lightweight verification with carbon flow resistance surface as a condition, determines the configuration optimization instruction, optionally activates the configuration optimization component, and determines the spatial configuration scheme by identifying key intervention points and performing configuration optimization processing; Scheme Rendering Module 14: Perform layered rendering and three-dimensional sectioning on the spatial configuration scheme, generate a dynamic exploded view and display it on the terminal interface.

[0047] Furthermore, the resource definition module 11 is used to perform the following methods: For each land unit in the target territory, resource attributes based on resource elements are defined, and resource field strength based on the superposition of resource attributes is determined. The resource elements include water resource pressure, carbon emission quota, and land suitability. The resource potential energy of neighboring land units relative to each land unit is defined. Based on the resource field strength and resource potential energy, the resource carrying capacity field of the target territory is defined.

[0048] Furthermore, the module development module 12 is used to perform the following methods: A carbon flow resistance surface is constructed, defined by the interaction characteristics of carbon sinks and carbon aggregates. These interaction characteristics include spatial connectivity, adjacency, and energy flow between carbon sink patches and carbon source land units. Using the resource carrying capacity as a baseline and the carbon flow resistance surface as a condition, a configuration generation component is constructed through adversarial supervised training, employing a combined architecture training method based on configuration generation and configuration judgment. A configuration optimization component is constructed by introducing a spatial negative entropy operator, aiming at identifying critical transition zones based on order and chaos and fine-tuning spatial configuration. This spatial negative entropy operator is a mathematical operator for the evolution of low-entropy states defined by land space boundaries, negative entropy sources, and flow channels. The configuration generation component and the configuration optimization component are combined to form a spatial configuration module.

[0049] Furthermore, the strategy generation and optimization module 13 is used to perform the following method: The resource carrying capacity is dynamically updated based on changes in the target land space, re-division of land units, and changes in resource attributes and resource potential energy. Based on the configuration generation component, an initial configuration strategy is generated with the adversarial response of multiple subjects in the land planning scenario. The initial configuration strategy is lightly verified, and the configuration optimization component is activated and the strategy is optimized based on the verification results.

[0050] Furthermore, the strategy generation and optimization module 13 is used to perform the following method: The first verification condition is the entropy change of carrying capacity in the target land space, and the second verification condition is the coupled oscillation of multiple spatial elements. If at least one of the first and second verification conditions is not met, a configuration optimization instruction is generated.

[0051] Furthermore, the strategy generation and optimization module 13 is used to perform the following method: According to the configuration optimization instruction, the configuration optimization component is activated; for each national land unit, the sensitivity gradient of the bearing state to the global bearing target is measured based on the bearing state and partial derivative; according to the direction of the sensitivity gradient, key intervention points are identified for the initial configuration strategy, wherein key intervention points are determined based on the national land unit with the contribution to the improvement of global bearing capacity; and configuration optimization processing is performed with the key intervention points as the intervention subjects.

[0052] Furthermore, the strategy generation and optimization module 13 is used to perform the following method: For the intervention subject, a first configuration optimization result is determined by low-entropy flow evolution based on the negative entropy flow operator, wherein the optimization iteration is guided by the spatial configuration mode based on maximizing negative entropy; for the intervention subject, a second configuration optimization result is determined by locating the critical transition zone and fine-tuning the spatial configuration; the spatial configuration scheme is determined by fitting the first configuration optimization result and the second configuration optimization result.

[0053] Furthermore, the scheme rendering module 14 is used to perform the following method: For the aforementioned spatial configuration scheme, resource configuration states at different spatial scales are rendered in layers to determine a layered view. The different spatial scales include macro-level watersheds, meso-level urban clusters, and micro-level plots. For the layered view, a three-dimensional section is performed along key intervention points to generate a dynamic exploded map. The sectioning is based on the spatial coupling relationship and negative entropy flow path of the underground resource carrying layer, the surface space utilization layer, and the above-ground ecological corridor layer. The dynamic exploded map is then displayed on a terminal interface.

[0054] Furthermore, the scheme rendering module 14 is used to perform the following method: In response to the layer selection command from the user, the load-bearing gradient change curve and negative entropy contribution ranking of key intervention points within the target layer are dynamically associated; based on the load-bearing gradient change curve and negative entropy contribution ranking, the spatial configuration parameters and optimization adjustment paths are traced back and highlighted for display on the user interface.

[0055] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for optimizing land spatial allocation based on resource carrying capacity, characterized in that, The method includes: For the target land space, based on the definition of resource elements of each land unit, define resource field strength and resource potential energy, and reconstruct a continuous and differentiable resource carrying capacity field; A spatial configuration module is developed on the land planning platform, wherein the spatial configuration module consists of a configuration generation component and a configuration optimization component; Based on the dynamically updated resource carrying capacity, the configuration generation component generates an initial configuration strategy and performs lightweight verification based on the carbon flow resistance surface, determines the configuration optimization instructions, and optionally activates the configuration optimization component. By identifying key intervention points and performing configuration optimization processing, the spatial configuration scheme is determined. The spatial configuration scheme is rendered in layers and sectioned in three dimensions to generate a dynamic exploded view, which is then displayed on the terminal interface.

2. The land spatial allocation optimization method based on resource carrying capacity as described in claim 1, characterized in that, Reconstructing a continuously differentiable resource carrying field includes: For each land unit of the target territory, resource attributes based on resource elements are defined, and resource field strength based on the superposition state of resource attributes is determined. The resource elements include water resource pressure, carbon emission quota and land suitability. Define the resource potential of neighboring territorial units relative to each territorial unit; Based on the resource field strength and resource potential energy, the resource carrying capacity of the target land space is defined.

3. The land spatial allocation optimization method based on resource carrying capacity as described in claim 1, characterized in that, Develop spatial configuration modules on the land planning platform, including: A carbon flow resistance surface is constructed, wherein the carbon flow resistance surface is defined by the total amount of carbon sinks and the interaction characteristics of carbon sinks. The interaction characteristics of carbon sinks are the spatial connectivity, adjacency relationship and energy flow between carbon sink patches and carbon source land units. Using the resource carrying field as a baseline and the carbon flow resistance surface as a condition, a configuration generation component is constructed through adversarial supervised training, wherein the training method is a combined architecture training based on configuration generation and configuration judgment. By introducing a spatial negative entropy operator, a configuration optimization component is constructed with the goal of identifying critical transition zones based on order and chaos and fine-tuning spatial configuration. The spatial negative entropy operator is a mathematical operator for the evolution of low-entropy states based on the definition of national spatial boundaries, negative entropy sources, and circulation channels. The configuration generation component and the configuration optimization component are combined to form a spatial configuration module.

4. The land spatial allocation optimization method based on resource carrying capacity as described in claim 1, characterized in that, The resource carrying capacity is dynamically updated based on changes in the target land space, the re-division of land units, and changes in resource attributes and resource potential energy. Based on the configuration generation component, an initial configuration strategy is generated to reflect the adversarial responses of multiple stakeholders in a land planning scenario. The initial configuration strategy is lightly validated, and the configuration optimization components are activated and the strategy is optimized based on the validation results.

5. The land spatial allocation optimization method based on resource carrying capacity as described in claim 4, characterized in that, The first verification condition is the entropy change of carrying capacity in the target land space, and the second verification condition is the coupled oscillation of multiple spatial elements. If at least one of the first and second verification conditions is not met, a configuration optimization instruction is generated.

6. The land spatial allocation optimization method based on resource carrying capacity as described in claim 5, characterized in that, By identifying key intervention points and performing configuration optimization processes, including: Activate the configuration optimization component according to the configuration optimization instruction; For each land unit, the sensitivity gradient of the bearing state to the global bearing target is measured based on the bearing state and partial derivatives. Based on the direction of the sensitivity gradient, key intervention points are identified for the initial configuration strategy, wherein key intervention points are determined using land units based on the contribution of global carrying capacity enhancement. The key intervention points are used as the intervention subjects, and configuration optimization processing is performed.

7. The land spatial allocation optimization method based on resource carrying capacity as described in claim 6, characterized in that, Using the aforementioned key intervention points as the intervention subjects, configuration optimization processing is performed, including: For the intervention subject, the first configuration optimization result is determined by low-entropy flow evolution based on negative entropy flow operator, wherein the optimization iteration is guided by the spatial configuration mode based on maximizing negative entropy. For the intervention subject, the second configuration optimization result is determined by locating the critical transition zone and making spatial configuration fine-tuning. The space configuration scheme is determined by fitting the first configuration optimization result with the second configuration optimization result.

8. The land spatial allocation optimization method based on resource carrying capacity as described in claim 1, characterized in that, The spatial configuration scheme is rendered in layers and sectioned in three dimensions to generate a dynamic exploded view, which is then displayed on the terminal interface, including: For the aforementioned spatial configuration scheme, resource configuration states at different spatial scales are rendered in layers to determine the layered view. The different spatial scales include macro-level watersheds, meso-level urban clusters, and micro-level plots. For the layered view, a three-dimensional section is performed along the key intervention points to generate a dynamic exploded view. The spatial coupling relationship and negative entropy flow path of the underground resource carrying layer, the surface space utilization layer and the above-ground ecological corridor layer are used as the sectioning basis. The dynamic exploded view is displayed on a terminal interface.

9. The land spatial allocation optimization method based on resource carrying capacity as described in claim 1, characterized in that, The method further includes: In response to the layer selection command on the user's end, the load gradient change curve and negative entropy contribution ranking of key intervention points within the target layer are dynamically correlated. Based on the bearing gradient change curve and the ranking of negative entropy contribution, the spatial configuration parameters and optimization adjustment paths are traced back and highlighted for display on the user interface.

10. A land spatial allocation optimization system based on resource carrying capacity, characterized in that, The system is used to implement the land spatial allocation optimization method based on resource carrying capacity as described in any one of claims 1-9, the system comprising: Resource definition module: For the target land space, based on the defined resource elements of each land unit, define resource field strength and resource potential energy, and reconstruct a continuous and differentiable resource carrying capacity field; Module Development Module: Develop a spatial configuration module on the land planning platform, wherein the spatial configuration module consists of a configuration generation component and a configuration optimization component; Strategy generation and optimization module: Based on the dynamically updated resource carrying capacity, the configuration generation component generates an initial configuration strategy and performs lightweight verification with carbon flow resistance surface as a condition, determines the configuration optimization instructions, and optionally activates the configuration optimization component. By identifying key intervention points and performing configuration optimization processing, the spatial configuration scheme is determined. Scheme rendering module: Performs layered rendering and 3D sectioning of the spatial configuration scheme, generates a dynamic exploded view, and displays it on the terminal interface.

Images