A method for constructing a refined ecological network by fusing functional traits and circuit theory

Abstract

The application discloses a fine ecological network construction method fusing functional traits and circuit theory, and relates to the technical field of intelligent ecological management. Ecological data input is acquired, an ecological network and a morphology-trait-function mapping are constructed; the construction of the ecological network comprises the following steps: acquiring ecological data to construct ecological nodes and ecological paths, and calculating ecological network connectivity; the construction of the morphology-trait-function mapping comprises the following steps: based on the ecological data input, patch shape-function traits are induced, and the corresponding relationship between functional traits and ecosystem functions is quantified; and ecological node priority is calculated through the ecological network connectivity and the morphology-trait-function mapping. The application can obtain better protection results at a lower cost.

Description

[0001] This application is the following application.

[0002] The application number is: 202511649321.9

[0003] Application date: November 12, 2025

[0004] The application is titled: "A Divisional Application of an Intelligent Decision-Making Method and System Based on Ecological-Social Dual-Network Coupling and Functional Trait Transmission". Technical Field

[0005] This invention relates to the field of intelligent ecological management technology, and more specifically to a refined method for constructing ecological networks that integrates functional traits and circuit theory. Background Technology

[0006] Currently, the current ecosystem management system has two main technical shortcomings: First, data synergy is poor. Ecological and social data are usually processed using separate modules, making it impossible to comprehensively consider ecological functions and social costs when making decisions.

[0007] Second, there is a disconnect in the transmission of functions. Existing protection and restoration plans only guide decision-making through landscape morphology indicators (such as the area of ​​ecological nodes) or vegetation types, without establishing a quantitative link with the actual functions of the ecosystem (such as carbon sinks and water conservation).

[0008] Therefore, how to propose an intelligent decision-making method and system based on ecological-social dual-network coupling and functional trait transmission, which decomposes ecological function assessment, social benefit accounting, and dynamic optimization decision-making into independent intelligent agent modules, and achieves cross-system collaboration to output the optimal solution that simultaneously meets ecological protection requirements and social feasibility constraints, is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0009] In view of this, the present invention provides a refined ecological network construction method that integrates functional traits and circuit theory.

[0010] To achieve the above objectives, the present invention adopts the following technical solution: A refined method for constructing ecological networks that integrates functional traits and circuit theory includes the following steps: Acquire ecological data inputs to construct ecological networks and morphology-trait-function mappings; The construction of the ecological network includes acquiring ecological data, constructing ecological nodes and ecological paths, and calculating the connectivity of the ecological network; The construction of the morphology-trait-function mapping includes summarizing patch shape-functional traits based on ecological data input, and quantifying the correspondence between functional traits and ecosystem functions; Ecological node priority is calculated through ecological network connectivity and morphology-characteristic-function mapping.

[0011] Optionally, building ecosystem nodes includes: Field quadrats were set up to obtain detailed data on vegetation communities; a stratified species distribution model was used to obtain hotspots of different vegetation distributions and the degree of interrelationships among plant species in the study area; by overlaying the distribution hotspots of multiple species, the results of plant habitat suitability in the study area were obtained. The probability maps of species distribution output from the stratified species distribution model were normalized, and the 0-1 standardization method was used to unify the dimensions. Spatial overlay analysis was used to weight and fuse the distribution hotspots of each species. The weighting coefficients were determined based on the species' conservation level and ecological function importance. The Habitat Suitability Index (HSI) was calculated. HSI = Σ(wi×Pi) / Σwi; Where wi is the weight of the i-th species and Pi is the distribution probability of the i-th species; finally, a continuous habitat suitability probability map is generated using spatial interpolation, and the spatial distribution results of plant habitat suitability in the study area are finally output; based on the habitat suitability results, the top 1% of the regions are extracted as ecological nodes and assigned a unique identifier.

[0012] Optionally, ecological pathway generation includes the following steps: Landscape resistance surface construction: The habitat suitability probability map before classification, i.e., the habitat suitability raster map, is transformed into a landscape resistance surface using a formula: Resistance value = 1 / (suitability value + ε); Where ε=0.01 is a correction coefficient to prevent division by zero, and the fitness value is derived from the output of the stratified species distribution model; Source point configuration: Select ecological nodes as current injection points; Criteria for determining node connectivity: Parallel computation is enabled when the distance between two nodes is ≤ the distance threshold D; The random walk model in circuit theory is used to generate the path using circuit theory analysis software. Connectivity edge extraction: Extract significant current paths and identify valid connections using scientific methods based on the relative current density distribution output by the model.

[0013] Optionally, the threshold D is set based on ecological principles.

[0014] Optionally, the current density threshold can be dynamically determined based on the research object, and the following methods can be used as references: Sensitivity analysis: Testing the impact of different thresholds on corridor network structure; Cumulative distribution analysis: Select the inflection point of cumulative current density percentage; Ecological validation: Adjust thresholds by combining species migration data or habitat continuity; Record the propagation weights of significant paths.

[0015] Optionally, the ecological network connectivity index can be calculated, which includes the following steps: inputting the ecological nodes and ecological connection edges into the Conefor Sensinode software or an equivalent connectivity analysis tool to calculate the connectivity index of all nodes.

[0016] Optionally, quantifying the correspondence between functional traits and ecosystem functions specifically includes the following steps: By calling internationally recognized plant trait databases, we can directly obtain the quantitative correspondence between plant functional traits and ecosystem functions; or by conducting field surveys of selected ecological nodes, we can obtain the correspondence between vegetation cover and functional traits and establish localized prediction models.

[0017] Optionally, the priority level of ecological nodes is determined, specifically including the following steps: Based on the assessment results of the connectivity and function of all ecological nodes, the restoration priorities of different ecological nodes are determined, and the decision-making logic is as follows: Identify key nodes for ecological connectivity: Select the top 5% of nodes in the ecological connectivity index as key nodes for ecological connectivity; the percentage can be adjusted according to needs. High-risk areas for functional degradation are marked: For each ecological node, its functional value is calculated using a validated trait-function model based on vegetation cover and functional traits. Set a functional degradation threshold based on the research area background or literature standards; Based on the above, priority levels are assigned to all ecological nodes, with priority given to restoring ecological nodes with high ecological functions and low ecological connectivity.

[0018] As can be seen from the above technical solutions, compared with the prior art, the present invention provides a refined ecological network construction method that integrates functional traits and circuit theory, so as to obtain better protection results at a lower cost. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of an intelligent decision-making method based on ecological-social dual-network coupling and functional trait transmission provided by the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] This invention discloses an intelligent decision-making method based on ecological-social dual-network coupling and functional trait transmission, such as... Figure 1 As shown, it includes: An ecological network module and a social network module are constructed. The outputs of the ecological network module and the social network module are superimposed to form an ecological-social dual-network spatial coupling model. Based on the output of the ecological-social dual-network spatial coupling model, ecological intelligent agents and social intelligent agents are constructed sequentially. Based on the ecological intelligent agent and the social intelligent agent, a dual intelligent agent coordinator is constructed to optimize the ecological-social dual network spatial coupling model, and finally generate a decision scheme.

[0023] In a specific implementation, an intelligent decision-making method based on ecological-social dual-network coupling and functional trait transmission constructs an ecological-social collaborative management framework, specifically including: S1: Ecological Network Module: Achieving ecological function assessment through multi-stage processing. Step 1: Constructing Ecosystem Nodes Field quadrats were established to obtain detailed data on vegetation communities, such as species, canopy, and height. A hierarchical modeling of species communities (HMSC) was employed to identify hotspots of different vegetation distributions and the degree of inter-species relationships within the study area. By overlaying distribution hotspots of multiple species, the habitat suitability of the plant population within the study area was determined. HMSC is a joint modeling method based on a Bayesian framework that can simulate vegetation habitat suitability and biodiversity hotspots. It simultaneously considers the environmental response of species distribution and inter-species interactions, exhibiting stronger explanatory and predictive capabilities.

[0024] Specifically, the species distribution probability maps output by the HMSC model are normalized using a 0-1 standardization method to unify the dimensions; spatial overlay analysis is used to weight and fuse the distribution hotspots of each species. The weighting coefficients are determined based on the species' conservation level and ecological function importance; the Habitat Suitability Index (HSI) is calculated. HSI=Σ(wi ×P i ) / Σw i ; Among them, w i P represents the weight of the i-th species. i Let be the distribution probability of the i-th species; finally, a continuous habitat suitability probability map is generated using spatial interpolation, and the spatial distribution results of plant habitat suitability in the study area are output. Based on the habitat suitability results, the top 1% of the regions are extracted as ecological nodes and assigned unique identifiers.

[0025] Step 2: Ecological Pathway Generation

[0026] (1) Construction of resistance surface: The habitat suitability probability map before classification, i.e., the habitat suitability raster map (values ​​0-1), is transformed into a landscape resistance surface through a formula: Resistance value = 1 / (suitability value + ε); Where ε=0.01 is a correction coefficient to prevent division by zero, and the suitability value is derived from the output of the HMSC model.

[0027] (2) Source point configuration: Select an ecological node as the current injection point (default current intensity 1A).

[0028] The criteria for determining node connectivity are: parallel computing is enabled when the distance between two nodes is ≤ the distance threshold (D).

[0029] Among them, the threshold D needs to be set based on ecological criteria (such as the maximum diffusion distance of the target species, the effective gene flow range of plants, and the fragmentation of landscape structure).

[0030] The random walk model in circuit theory is adopted, and the path generation is achieved through circuit theory analysis software, such as Circuiscape 4.0 or the gdistance package (function: shortestPath()) in R language.

[0031] (3) Connection edge extraction: extract significant current paths and identify effective connections based on the relative current density distribution output by the model using scientific methods.

[0032] The current density threshold needs to be determined dynamically based on the research object. Reference methods include: Sensitivity analysis: Testing the impact of different thresholds (e.g., 0.4–0.8) on the corridor network structure; Cumulative distribution analysis: Select the inflection point of the cumulative percentage of current density (e.g., the 85th percentile); Ecological validation: Adjust thresholds by combining species migration data or habitat continuity.

[0033] Record the conduction weights (current density values) of significant paths.

[0034] Step 3: Calculate the ecological network connectivity index

[0035] Input the ecological nodes and ecological connection edges into software Conefor Sensinode or an equivalent connectivity analysis tool (such as Graphab) to calculate the connectivity index of all nodes.

[0036] Specifically, the software includes various connectivity indices; for example, the following two connectivity indices can be used: ①Integral Index of Connectivity (IIC) The formula for integrating patch area (or other attributes) with topological location (such as shortest path steps) is as follows: ; Where n is the total number of ecological nodes; i and j are node indices, representing any two ecological nodes (i and j can be the same); a i a j These are the attribute values ​​of nodes i and j (such as patch area, habitat quality, etc.). nl ij Let $\frac{i}{j}$ be the shortest path distance (topological distance) from node $i$ to node $j$, which is the minimum number of links required to connect the two nodes. If the two nodes are not connected, then $\frac{i}{j}$ represents the shortest path distance from node $i$ to node $j$. nl ij =∞ (at this point, the fraction is 0); A L The maximum landscape attribute value (such as total landscape area) is used for standardization indices in the range of [0, 1].

[0037] ② Probability of Connectivity (PC)

[0038] Its key feature is that it considers the path with the highest probability, thus more accurately reflecting the species' dispersal capacity. The formula is as follows: ; Where n is the total number of ecological nodes; i and j are node indices, representing any two ecological nodes (i and j can be the same); a i a j These are the attribute values ​​of nodes i and j (such as patch area, habitat quality, etc.). p ij The path with the maximum product probability from node i to j, i.e., the maximum product of probabilities among all possible paths (path probability = direct diffusion probability p of each step in the path). ij The product of (i, j) and if i = j (the same node), then p ij =1 (self-reachability is 100%); if the two nodes are completely isolated, then p ij =0;A L The maximum landscape attribute value (such as total landscape area) is used for standardization indices in the range of [0, 1].

[0039] Step 4: Quantify the correspondence between functional traits and ecosystem functions

[0040] By accessing internationally recognized plant trait databases (such as the TRY database and the Chinese Plant Trait Database), the quantitative correspondence between plant functional traits and ecosystem functions can be directly obtained. Alternatively, field surveys can be conducted at selected ecological nodes to obtain the correspondence between vegetation cover and functional traits (such as specific leaf area and wood density), and localized prediction models can be established.

[0041] Specifically, the input data format includes: Ecological node ID, location information (geographic coordinates, altitude), included vegetation type, plant functional traits corresponding to each vegetation type (specific leaf area, wood density, etc. from the TRY database), coverage of each vegetation type, and functional value corresponding to each functional trait (carbon sequestration / water conservation, etc.). The conversion of the functional traits to ecological functions is achieved through any of the following methods: a. Based on trait parameters (specific leaf area, wood density, etc.) obtained from international plant trait databases (such as TRY), input them into a trait-function regression model that has been validated in the literature for calculation; b. A localized model calibrated using field measurement data; All models must meet the requirements of independent validation (e.g., R² ≥ 0.6) and pass sampling verification (e.g., sampling verification error ≤ 25%).

[0042] Step 5: Determine the priority level of ecological nodes

[0043] Based on the assessment results of the connectivity and function of all ecological nodes in the region, the restoration priorities for different ecological nodes are determined. The decision-making logic is as follows: (1) Identify key nodes of ecological connectivity Select the top 5% of nodes in the ecological connectivity index as key nodes for ecological connectivity. The percentage can be adjusted according to needs.

[0044] (2) High-risk areas for annotation function degradation

[0045] For each ecological node, based on vegetation cover and functional traits, its functional value (such as carbon storage and water conservation) is calculated using a validated trait-function model (such as a regression model). Based on the background of the study area or literature standards, set functional degradation thresholds (such as a decrease in carbon sequestration capacity of ≥30% or a decrease in water conservation capacity of 20% below the regional average).

[0046] Based on the above, priority levels are assigned to all ecological nodes, with priority given to restoring ecological nodes with high ecological functions and low ecological connectivity.

[0047] S2: Social Network Module Step Six: Constructing Social Nodes Based on the classification of management entities, the main stakeholder groups are government agencies (environmental protection / forestry and other functional departments), enterprises (developers / tourism companies and other entities), and community residents. Social nodes are constructed based on these protection and restoration entities.

[0048] Step 7: Construct an N×N dimensional relationship matrix (N is the number of participants), matrix element a _ij The assignment rules are as follows: +1 is assigned when both parties have engaged in joint protection and restoration activities within the past 5 years (e.g., government-enterprise joint construction of wetland parks); -1 is assigned when environmental litigation occurs (e.g., residents complaining about corporate pollution); and 0 is assigned when there is no direct interaction. For every 10,000 yuan of joint project investment, +0.1 is assigned; for every 10,000 yuan of litigation compensation, -0.1 is assigned.

[0049] The final output is a weighted social network topology graph, which is used for subsequent collaborative decision optimization.

[0050] Step 8: Determine the priority level of social nodes

[0051] Based on the topological relationships of the social network, the eigenvector centrality is calculated using Gephi software. A priority level is then assigned based on the centrality, with higher centrality levels generally having higher priority.

[0052] S3: Dual-Agent Collaboration Module

[0053] Step Nine: Construct an ecological-social dual-network spatial coupling model (action layer)

[0054] Data on conservation and restoration activities related to the target ecosystem within a specific period or in the planning stage are compiled, and ecological and social networks are constructed according to the aforementioned eight steps. The impact of conservation and restoration activities on ecological nodes is quantified, that is, the change in ecological function caused by changes in the area of ​​ecological nodes.

[0055] Specifically, it can be calculated from any of the following aspects: ① Specific indicators of the project, such as the planned area of ​​different vegetation to be restored, are compared with the area of ​​the original different vegetation in the ecological nodes in reality, in order to quantify the impact of protection and restoration activities on the ecological nodes.

[0056] ② Based on monitoring data before and after protection and restoration activities, quantify the changes in the area of ​​different vegetation in ecological nodes.

[0057] Furthermore, the ecological network and social network are overlaid based on geospatial location. The priority levels of ecological nodes and social nodes are added together to calculate the number of ecological nodes occupied by protection and restoration activities and the sum of their priority scores. First, the vector data of ecological nodes and social nodes are intersected using ArcGIS's Intersect tool. Then, the priority levels of the intersecting social nodes and ecological nodes are added together. Finally, the priority levels are assigned as weights to the nodes in the intersection area to guide the prioritization of protection and restoration activities. Nodes with higher priority levels indicate higher ecological connectivity and ecological functions, and will bring higher social benefits; therefore, they should be prioritized for protection and restoration.

[0058] Step 10: Construct an ecological intelligent entity

[0059] (1) Small sample modeling

[0060] ① Select 3%-5% of representative ecological nodes from the entire ecological network.

[0061] The selection criteria must simultaneously meet the following: ranking in the top 20% for landscape connectivity; and covering more than 80% of the target area's vegetation types.

[0062] ② Calculate the similarity between representative ecological nodes

[0063] Vegetation cover data within ecological nodes is obtained, and the data source can be field measurement, remote sensing inversion, or a third-party database. Using the coverage percentage of different vegetation types within an ecological node as weights, the similarity between different ecological nodes is calculated based on the similarity between different plants, and this similarity serves as the connection edge of the ecological network. The formula for measuring the similarity between ecological nodes utilizes the Bray-Curtis distance, and its formula is as follows: ; Among them, S A，i and S B，i w represents the absolute value of the cover of each type of vegetation in two ecological nodes A and B that have a connecting edge; i Indicates the importance weight of the plant.

[0064] Specifically, appropriate weights can be added in the following three scenarios: a. Key Species Protection: If the vegetation cover on which endangered species depend is below a certain percentage, such as 5% (the specific value can be adjusted according to the region and ecosystem), an additional weighting coefficient (e.g., w) needs to be added. i =2.0), with weights based on IUCN Red List tiers; b. Invasive species control: For invasive species such as Spartina alterniflora, negative weights (e.g., w) can be set. i =-1.0) to reduce the resulting "false similarity", and the weight of invasive species refers to the hazard level in Appendix C of the "Regulations on the Management of Invasive Alien Species"; c. Functional group differences: When modeling herbaceous / shrub / tree groups, group weights can be assigned according to hierarchy.

[0065] (2) Global functional extrapolation based on transfer learning

[0066] First, the computer learns the patterns of the surveyed ecological nodes (e.g., larger leaf area corresponds to stronger carbon sequestration). Then, based on the similarity between two ecological nodes (close location, similar vegetation type, etc.), it generalizes these patterns to unsurveyed ecological nodes. The prediction results are labeled with their reliability (e.g., ecological nodes with low similarity will be marked "need verification").

[0067] Specifically, 1) Input data: (1) Small sample modeling data of the investigated ecological nodes.

[0068] (2) Basic information of ecological nodes not investigated: ecological node number, location information (geographic coordinates, altitude), vegetation type, plant functional traits (specific leaf area, wood density, etc. from TRY database) for each vegetation type, and coverage of each vegetation type.

[0069] (3) Ecological node relationship data: Which ecological nodes are adjacent to each other; ecological similarity between ecological nodes (Bray-Curtis index).

[0070] 2) Input data structure: Line: Ecological nodes (n surveyed + m unsurveyed); Column: Feature variables (standardized): [longitude, latitude, altitude, coverage, trait 1, trait 2, ..., Bray-Curtis similarity].

[0071] 3) Algorithm architecture: Domain-adaptive neural networks employing dual-channel feature extraction: Channel 1: Processing spatial features (coordinates + altitude); Channel 2: Treat trait characteristics (such as specific leaf area + wood density + ...).

[0072] 4) Domain discriminator: Calculates the distribution difference between surveyed and unsurveyed nodes.

[0073] 5) Loss function L: L=α·L_pred+β·L_domain (α=0.7, β=0.3); Where L_pred is the prediction loss (represented by mean squared error) and L_domain is the domain difference loss, represented by MMD distance.

[0074] 6) Output layer: a. Functional value prediction; b. The reliability assessment module outputs three levels of reliability rating: high, medium, and low. c. Credibility judgment rule: When Bray-Curtis similarity > 0.5, perform transfer learning prediction; when Bray-Curtis similarity < 0.5, skip the node and continue to search other nodes.

[0075] 7) Output of overall ecological functions: The functional values ​​of all ecological nodes (including migration prediction nodes) obtained in step a of step 6 are assigned weights w based on their network topology importance. k (Weight sources can be selected: current density, eigenvector centrality, etc.), calculate the total global performance value using the following formula: ; Where n is the number of surveyed ecological nodes (including field measurement data), m is the number of unsurveyed ecological nodes (predicted through transfer learning), and f k w represents the ecological function value (measured or predicted) of node k. k The topology integration weight for node k (generated by any method: Circuitscape propagation weight; network centrality (eigenvector centrality); node area ratio, etc.).

[0076] Step 11: Constructing a Social Intelligent Agent

[0077] Establish a historical protection and restoration case database for each group in the social network, and quantify the benefit value of their past ecological intervention behaviors. For example, the carbon sink gain coefficient of government projects is +0.82.

[0078] This paper proposes a method for calculating social benefits in social intelligent agents, which combines "ecosystem service value + social adjustment coefficient". While mature assessment standards exist for ecosystem service value, such as water conservation, carbon sequestration, and biodiversity maintenance, simply calculating ecological value fails to reflect differences in the social dimension, such as the degree of benefit to different groups, policy preferences, and time costs. Therefore, a social adjustment coefficient is needed to make social benefits more scientific and objective.

[0079] 1. Quantifying the value of ecosystem services (basic calculations)

[0080] (1) Methods for quantifying the value of ecosystem services: The various functions of the ecosystem (such as forest carbon sequestration and wetland water purification) are quantified using a multi-standard integrated system, specifically including: Domestic standards: Basic calculations are based on relevant standards and literature, such as the "Technical Specification for Gross Ecosystem Product (GEP) Accounting in Ecosystem Assessment", the "Technical Specification for National Ecological Status Survey and Assessment - Ecosystem Service Function Assessment", and the equivalent factor method (Journal of Natural Resources, 2015). The unit value is based on yuan / hectare / year.

[0081] International standards: Cross-border ecological functions can refer to the equivalent conversion methods of TEEB or IPCC, and be converted into RMB through exchange rates and purchasing power parity.

[0082] Specifically, the integration standard is as follows: ① Priority: Domestic standards should be the primary focus, with international data supplemented only for two scenarios: a. Cross-border services (such as carbon sinks); b. Services for which there is no domestic data (such as marine genetic resources).

[0083] ② Difference Correction: If the difference between domestic and international values ​​for the same service exceeds 30%, a weighted average is used. V final =w dom V dom +w int V int When domestic data quality is high dom =0.7.

[0084] 2. Social adjustment coefficient (dynamically adjusted)

[0085] Because the distribution of ecological benefits is not equal, the aforementioned ecosystem service value serves as a basic input, and the following technological linkages are used to dynamically adjust social benefits. A social adjustment coefficient is introduced to reflect the impact of different groups, spaces, and times: (1) Spatial allocation coefficient (S) The benefits of ecosystem services are not evenly distributed. For example, the water conservation function of upstream forests is more important to downstream cities; the air purification function of urban green spaces has a greater impact on surrounding residents. The beneficiary distance decay model recommended by the TEEB framework (an international initiative launched in 2007 by the United Nations Environment Programme, titled "The Economics of Ecosystems and Biodiversity") (revised according to Chapter 3.2 of the TEEB-2010 report) is double-calibrated with the spatial allocation coefficients of the GEP specification. The distance decay model formula is as follows: S = 1 / (1+α·d); Where d is the distance from the ecological node to the benefit area, and α is the attenuation parameter (different values ​​for different ecological functions).

[0086] (2) Group preference coefficient (P)

[0087] Different interest groups (government, businesses, and residents) place varying degrees of importance on ecological value. For example, the government may be more concerned with carbon sequestration (policy assessment), businesses may be more concerned with landscape value (commercial development), and residents may be more concerned with air quality (health impact). Willingness to Pay (WTP) surveys or policy document analysis can quantify the preference weights of different groups.

[0088] (3) Time discount factor (T)

[0089] Ecological benefits often take many years to manifest (e.g., significant carbon sequestration only occurs 10 years after afforestation), while social decision-making focuses more on short-term gains. A dynamic discounted cash flow model can be used, such as: T = 1 / (1 + r) t ; Where r is the discount rate (referencing green finance standards), and t is the number of years required to realize the benefits.

[0090] 3. Calculation of final social benefits

[0091] Taking all the above factors into account, the formula for calculating social benefits is: ; in, Indicates the first Spatial units (such as ecological nodes and restoration plots). SB represents the total number of spatial units. g For the social benefits of a specific group g, ESV i For unit The value of ecosystem services, S i For unit Spatial allocation coefficient, P g T represents the group preference coefficient for a specific group g. i For unit The time discount factor, g, represents groups such as government, businesses, and residents. The final social benefits are weighted and aggregated: SB = ∑ (w_g × SB_g); Where SB represents the overall social benefits, and w_g is the group weight, such as government 0.4 / enterprise 0.3 / resident 0.3.

[0092] Step 12: Data Processing

[0093] Input the protection and restoration activities to be implemented for the target ecosystem within a certain timeframe or as planned, including the types, areas, and locations of the plants to be planted.

[0094] Step Thirteen: Setting up the Dual-Agent Coordinator

[0095] (1) Input data structure specifications

[0096] The coordinator receives two types of standardized input data: Ecological priority data: includes unique identifiers for ecological nodes and their corresponding priority scores. The score is a real number between 0 and 100, with higher values ​​indicating greater urgency for restoration. Social constraint data includes three types of information: Budget constraint: The maximum amount of funds available during the planning period, in real form (unit: yuan), used to limit the total cost of the two-agent coordinator.

[0097] Social priority data: includes a unique identifier for each social node and its corresponding priority score. The score is a real number between 0 and 100, with a higher score indicating a greater feasibility of repair. Spatial constraints: A set of geographic boundary coordinates for undevelopable areas, a set of geographic polygons (format: GeoJSON or Shapefile), used to automatically exclude legally undevelopable areas.

[0098] (2) Construction of multi-objective optimization model

[0099] Establish an optimization problem with dual objectives: Objective 1: Maximize the ecological function gain (achieved by accumulating the priority scores of selected ecological nodes); Objective 2: Maximize overall social acceptance (calculated by weighted average support rates from all stakeholders) and social benefits.

[0100] Constraints: Condition 1: The total implementation cost shall not exceed 90% of the budget (10% shall be reserved as contingency funds). Condition 2: No development activities are permitted within legally protected areas.

[0101] (3) Optimize the solution process: use Monte Carlo simulation iterative solution: ① Initial solution generation: The following two cases can constitute an initial solution: a. Based on the spatial location and area of ​​the protection and restoration activities in the plans or programs of interest in the study, determine the amount of change these activities bring to the ecological network (identify the affected ecological nodes and their changed areas). b. Randomly select n ecological nodes from those with high social and ecological priority levels to form an initial scheme, ensuring that the average topological distance between nodes does not exceed 3 hops (based on ecological network connectivity).

[0102] ② Neighborhood search strategy. Each iteration performs one of the following three types of operations: Type A operation: Replace one ecological node in the current scheme (replacement probability 30%); Type B operation: Adjust the implementation area by 10%-15% (adjustment probability 50%). Type C operation: reallocate the weight coefficients of ecological and social goals (adjustment not exceeding ±10% of the initial value).

[0103] ③ The decision rule is to retain non-dominated solutions that simultaneously satisfy the following conditions: a. There is no other solution in the existing solution set that is superior to the current solution in both ecological function and social benefits; b. The functional complementarity index between ecological nodes is greater than 0.7 (calculated by Bray-Curtis similarity).

[0104] c. Satisfies: Implementation standard = Social benefits (SB) / Total cost >= θ; Wherein, total cost is the budget constraint value, and θ is the cost-benefit ratio threshold (it is recommended that θ=1.2, that is, every 1 yuan invested should generate at least 1.2 yuan of benefit). If the above rules are not met, return to the action layer for adjustment.

[0105] (4) Output results

[0106] The optimization scheme for the coordinator output includes the following elements: a. Implementation Recommendation List: Ecological node number and recommended implementation area (accurate to 0.1 hectares); expected ecological function improvement (standardized as a percentage); corresponding cost estimate.

[0107] b. Benefit evaluation report: Eco-Social Balance Index (ESI): ESI = ω1 × Ecological Gain + ω2 × Social Benefit; Among them, ω1 and ω2 are weighting factors greater than 0 and less than 1, and can be assigned appropriate weights to ecology and society according to the actual situation. For example, in regions that prioritize ecological development, a relatively high value can be assigned to ecological gains.

[0108] (5) Parameter sensitivity analysis

[0109] Perform ±10% perturbation analysis on the following core parameters, including but not limited to: a) Eco-social balance weight ω1, see step thirteen; b) Group preference coefficient P_g, see step eleven; c) Budget occupancy rate cap, see constraints in step thirteen; d) Functional complementarity index threshold, see Bray-Curtis standard.

[0110] In a specific implementation, an intelligent decision-making system based on ecological-social dual-network coupling and functional trait transmission includes: Basic model building module: used to build ecological network modules and social network modules; Coupling module: used to overlay the outputs of the ecological network module and the social network module to construct an ecological-social dual-network spatial coupling model; Agent-Gain Prediction Module: Used to sequentially construct ecological agents and social agents based on the output of the ecological-social dual-network spatial coupling model; Decision module: Used to optimize the ecological-social dual-network spatial coupling model by constructing a dual-agent coordinator based on ecological and social agents, and finally generate a decision scheme.

[0111] This invention addresses the problems of data silos and inefficient decision-making in the field of ecological management through three groundbreaking design features: Decoupled modular architecture: Ecological function assessment, social benefit accounting, and dynamic optimization decision-making are separated into independent intelligent agent modules, and cross-system collaboration is achieved.

[0112] Self-feedback decision-making closed loop: By balancing the state of ecological and social indicators through a coordinator, the scheme is automatically triggered for re-optimization, forming a continuous improvement mechanism of "evaluation-decision-calibration".

[0113] Dual-drive collaborative optimization engine: The ecological driving module (based on functional characteristics) and the social driving module (based on cost-effectiveness) output the optimal solution that simultaneously meets the requirements of ecological protection and social feasibility constraints through an intelligent negotiation mechanism.

[0114] This invention is applicable to carbon sink-tourism revenue balance decisions in national park ecological restoration projects, or to the synergistic optimization of biodiversity and resident satisfaction in urban green space renewal.

[0115] This invention focuses on how optimizing a single subsystem can have unintended negative impacts on the entire system. For example, increasing resilience to external pressures in an ecological network can reduce the overall stability of the coupled system, leading to a phenomenon known as the optimizer's tragedy. This highlights how measures to stabilize one part of a system can inadvertently cause global collapse.

[0116] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0117] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A refined method for constructing ecological networks that integrates functional traits and circuit theory, characterized in that, Includes the following steps: Acquire ecological data inputs to construct ecological networks and morphology-trait-function mappings; The construction of the ecological network includes acquiring ecological data, constructing ecological nodes and ecological paths, and calculating the connectivity of the ecological network; The construction of the morphology-trait-function mapping includes summarizing patch shape-functional traits based on ecological data input, and quantifying the correspondence between functional traits and ecosystem functions; Ecological node priority is calculated through ecological network connectivity and morphology-characteristic-function mapping.

2. The refined ecological network construction method integrating functional traits and circuit theory according to claim 1, characterized in that, Building ecosystem nodes includes: Field quadrats were set up to obtain detailed data on vegetation communities; a stratified species distribution model was used to obtain hotspots of different vegetation distributions and the degree of interrelationships among plant species in the study area; by overlaying the distribution hotspots of multiple species, the results of plant habitat suitability in the study area were obtained. The probability maps of species distribution output from the stratified species distribution model were normalized, and the 0-1 standardization method was used to unify the dimensions. Spatial overlay analysis was used to weight and fuse the distribution hotspots of each species. The weighting coefficients were determined based on the species' conservation level and ecological function importance. The Habitat Suitability Index (HSI) was calculated. HSI=Σ(w i ×P i ) / Σw i ; Among them, w i P represents the weight of the i-th species. i Let be the distribution probability of the i-th species; finally, use spatial interpolation to generate a continuous habitat suitability probability map, and finally output the spatial distribution results of plant habitat suitability in the study area; based on the habitat suitability results, extract the top 1% of the areas as ecological nodes and assign them unique identifiers.

3. The refined ecological network construction method integrating functional traits and circuit theory according to claim 1, characterized in that, Ecosystem pathway generation includes the following steps: Landscape resistance surface construction: The habitat suitability probability map before classification, i.e., the habitat suitability raster map, is transformed into a landscape resistance surface using a formula: Resistance value = 1 / (suitability value + ε); Where ε=0.01 is a correction coefficient to prevent division by zero, and the fitness value is derived from the output of the stratified species distribution model; Source point configuration: Select ecological nodes as current injection points; Criteria for determining node connectivity: Parallel computation is enabled when the distance between two nodes is ≤ the distance threshold D; The random walk model in circuit theory is used to generate the path using circuit theory analysis software. Connectivity edge extraction: Extract significant current paths and identify valid connections using scientific methods based on the relative current density distribution output by the model.

4. The refined ecological network construction method integrating functional traits and circuit theory according to claim 3, characterized in that, in, The threshold D is set based on ecological principles.

5. The refined ecological network construction method integrating functional traits and circuit theory according to claim 3, characterized in that, The current density threshold is dynamically determined based on the research object. Reference methods include: Sensitivity analysis: Testing the impact of different thresholds on corridor network structure; Cumulative distribution analysis: Select the inflection point of cumulative current density percentage; Ecological validation: Adjust thresholds by combining species migration data or habitat continuity; Record the propagation weights of significant paths.

6. The refined ecological network construction method integrating functional traits and circuit theory according to claim 1, characterized in that, The calculation of the ecological network connectivity index includes the following steps: inputting the ecological nodes and ecological connection edges into the software Conefor Sensinode or an equivalent connectivity analysis tool to calculate the connectivity index of all nodes.

7. The refined ecological network construction method integrating functional traits and circuit theory according to claim 1, characterized in that, Quantifying the correspondence between functional traits and ecosystem functions includes the following steps: By calling internationally recognized plant trait databases, we can directly obtain the quantitative correspondence between plant functional traits and ecosystem functions; or by conducting field surveys of selected ecological nodes, we can obtain the correspondence between vegetation cover and functional traits and establish localized prediction models.

8. The refined ecological network construction method integrating functional traits and circuit theory according to claim 1, characterized in that, Determining the priority level of ecological nodes includes the following steps: Based on the assessment results of the connectivity and function of all ecological nodes, the restoration priorities of different ecological nodes are determined, and the decision-making logic is as follows: Identify key nodes for ecological connectivity: Select the top 5% of nodes in the ecological connectivity index as key nodes for ecological connectivity; the percentage can be adjusted according to needs. High-risk areas for functional degradation are marked: For each ecological node, its functional value is calculated using a validated trait-function model based on vegetation cover and functional traits. Set a functional degradation threshold based on the research area background or literature standards; Based on the above, priority levels are assigned to all ecological nodes, with priority given to restoring ecological nodes with high ecological functions and low ecological connectivity.

Images