Plateau afforestation habitat intelligent regulation method and system

Abstract

The application provides a highland afforestation habitat intelligent regulation method and system, and belongs to the technical field of ecological intelligent regulation. The method comprises the following steps: S1, obtaining multi-source data to construct a habitat database and quantifying habitat demand parameter of a target species; S2, calculating a comprehensive ecological function value and a species suitability index, and identifying a habitat demand conflict area; S3, designing a regulation scheme and pre-evaluating benefits; S4, taking the maximization of overall benefits of multiple species and the optimization of balance degree as targets, and outputting an optimal scheme through a multi-objective optimization algorithm; S5, obtaining feedback data through a monitoring network after the implementation of the scheme, comparing actual and expected benefits, and when the deviation exceeds a threshold, linking and correcting model parameters to form a closed-loop optimization. The highland afforestation habitat intelligent regulation method and system realize dynamic and accurate diagnosis and simulation optimization decision of habitat quality, build a real-time feedback adaptive management mechanism, and improve the diversity of highland mammals and birds.

Description

Technical Field

[0001] This invention relates to the field of ecological intelligent regulation technology, and in particular to an intelligent regulation method and system for plateau afforestation habitat. Background Technology

[0002] Plateau ecosystems are characterized by complex terrain and harsh climates, making biodiversity conservation and ecological restoration a challenging task. Traditional habitat restoration projects often focus on increasing vegetation cover, employing afforestation and vegetation management methods that emphasize plant species selection and survival rates, lacking a systematic consideration of the specific habitat needs of various wildlife species such as mammals and birds. Key habitat elements upon which animals depend for survival, such as shelter, food resources, reproductive conditions, and migration corridors, are often overlooked in existing planning and design. This results in ecological restoration projects failing to effectively promote animal population recovery and diversity maintenance while increasing vegetation cover, and may even hinder the development of specific species due to habitat homogenization.

[0003] Existing technologies have significant shortcomings in supporting refined habitat management. On the one hand, habitat monitoring largely relies on single data sources or periodic manual surveys, making it difficult to achieve integrated analysis and dynamic assessment of multi-dimensional environmental factors and spatiotemporal data of animal activity. This results in delayed and one-sided judgments on habitat quality. On the other hand, the lack of quantitative predictive models that link habitat structure parameters with the effects of control measures means that management decisions are mostly based on experience. It is impossible to simulate, compare, and optimize the habitat improvement benefits of different control schemes before implementation, let alone dynamically adjust and continuously improve them based on monitoring results after implementation. This results in insufficient intelligence and adaptability in the management process.

[0004] Therefore, existing methods generally face challenges in addressing the complex needs of multi-species coexistence and rapid habitat changes on the plateau, including inaccurate habitat assessment, lack of foresight in regulatory decisions, and difficulty in continuously optimizing management processes. A new method is urgently needed to solve these problems. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent method and system for regulating habitats in plateau afforestation projects, addressing the problems of extensive management, experience-based decision-making, and difficulty in synergistically improving the quality of habitats for multiple species in existing plateau ecological restoration projects. This method integrates multi-source monitoring data with species ecological needs to achieve dynamic diagnosis of habitat quality, simulation and pre-implementation of regulation schemes, and multi-objective collaborative optimization. Based on an implementation feedback adaptive correction model, it transforms habitat management from passive protection centered on trees to proactive habitat optimization aimed at improving the diversity of mammals and birds in the plateau.

[0006] To achieve the above objectives, this invention provides a method for intelligent regulation of plateau afforestation habitats, comprising the following steps: Step S1: Obtain multi-source data of the target area and construct a habitat database. Based on animal ecology knowledge, interpret and quantify the habitat requirement parameters of at least two target species. Step S2: Using the habitat database and habitat demand parameters as input, calculate the comprehensive ecological function value of the evaluation unit, calculate the potential suitability index of each species, and identify areas of conflict in habitat demand among multiple species. Step S3: Based on the multi-species habitat demand conflict zone and potential suitability index, design multiple control schemes, pre-evaluate the expected control benefits of each control scheme on each species, and calculate the overall synergistic benefit index. Step S4: Based on the expected comprehensive regulatory benefits and overall synergistic benefits of each regulation scheme, and with the goal of maximizing the overall expected benefits of multiple species and optimizing the balance of benefits among species, use a multi-objective optimization algorithm to find the optimal regulation scheme for each scheme and output the optimal habitat regulation optimization scheme. Step S5: Implement the optimal habitat regulation and optimization scheme. Obtain actual habitat status data through the monitoring network, calculate the actual comprehensive regulation benefits and compare them with the expected comprehensive regulation benefits obtained in step S3. When the deviation exceeds the threshold, feed back the actual habitat status data to steps S2 and S3, recalculate and update the model.

[0007] Preferably, step S1 specifically includes: Step S11: Collect remote sensing data, topographic data, meteorological data, soil data, and target species activity monitoring data of the target area; Step S12: Perform spatiotemporal matching and fusion of multi-source data to construct a habitat database; Step S13: For at least two target species, interpret and quantify the habitat requirements parameters of each species for concealment, food resources, water sources, reproductive conditions and disturbance distance.

[0008] Preferably, step S2 specifically includes: Step S21: Divide the study area into standard evaluation units; Step S22: Based on the constructed habitat database, calculate the comprehensive ecological function value of each evaluation unit using the ecological function quantification model. The calculation formula for each sub-function value is as follows: ; ; ; ; ; in, , , , These are the weighting coefficients for four functions: water conservation, food supply, shelter, and reproduction. This represents the water conservation function value. Provides functional value to food. For concealment and shelter function value, For reproductive and parenting function values, For the first Vegetation coverage of each evaluation unit For the first Leaf area index of each evaluation unit For the first Annual precipitation for each evaluation unit For the first Soil saturated hydraulic conductivity of each evaluation unit , , These are the weighting coefficients. For the first Normalized Difference Vegetation Index of each evaluation unit For the first Herbaceous biomass of each evaluation unit For the first Edible plant abundance in each evaluation unit , , These are the weighting coefficients. For the first Vegetation density of each evaluation unit, For the first Vegetation height of each evaluation unit, For the first Surface roughness of each evaluation unit, , , These are the weighting coefficients. For the first The terrain concealment level of each evaluation unit For the first The reciprocal of the distance from the interference source to each evaluation unit. For the first Microclimate suitability of each evaluation unit , , These are the weighting coefficients; Step S23: Based on the habitat requirement parameters interpreted in step S1, construct a species-specific habitat suitability model for each target species, and calculate the potential suitability index of each species in each evaluation unit. The specific process is as follows: Step S231, targeting the target species Select its key environmental variable set Each variable corresponds to a type of habitat requirement parameter in step S1. , , Representing the 1st, 2nd, and so on, respectively. The specific values ​​or types of key environmental variables; Step S232: Calculate the evaluation unit Potential suitability index : ; in, For the intercept term, For the first Environment variables The response coefficient, As an evaluation unit The Middle The values ​​of the environment variables, This represents the total number of environment variables. is the base of the natural logarithm; Step S233: Using the maximum likelihood estimation method, based on the target species activity monitoring data, the response coefficients are... Calibrate the model to maximize the match between the fitness index predicted by the model and the actual probability of species occurrence. Step S234: Calculate all evaluation units The values ​​were normalized to obtain the distribution of the potential fitness index of this species throughout the study area; Step S24: Identify areas of conflict in habitat requirements among multiple species by comparing and analyzing the suitability indices of each species.

[0009] Preferably, in step S24, the identification of habitat conflict zones among multiple species is achieved through the following method: Step S241: For each evaluation unit Calculate the potential fitness index of species A and species B. and ; Step S242: Define the conflict index : ; in, and These are the habitat requirement weights for species A and species B, respectively. Step S243: Set the conflict threshold ,when At that time, the evaluation unit was marked as a habitat demand conflict zone; Step S244: Perform spatial cluster analysis on all identified habitat demand conflict areas to form contiguous habitat demand conflict areas.

[0010] Preferably, step S3 specifically includes: Step S31: Using the identified habitat demand conflict areas as the target areas for regulation, design a variety of regulation schemes with different spatial configurations and vegetation structure parameters. Step S32: Input the parameters of each regulation scheme into the habitat regulation benefit evaluation model. Based on the potential suitability index and habitat demand parameters, simulate in advance the expected habitat quality value for each target species after the implementation of each regulation scheme. and expected comprehensive regulatory benefits Among them, the expected habitat quality value The calculation formula is: ; in, The first day after the implementation of the control plan The first evaluation unit Ecological function value, For the first The weighting coefficient of each ecological function for the target species. The total number of ecological functions; Expected comprehensive regulatory benefits The calculation formula is: ; in, The first day after the implementation of the control plan The expected habitat quality value for each evaluation unit. The total number of evaluation units, The average habitat quality value before regulation; Step S33: Based on the expected comprehensive regulation benefits, calculate the overall synergistic benefit index of the regulation scheme, including overall benefits and benefit balance.

[0011] Preferably, in step S33, the overall benefit The calculation formula is: ; in, For the first The expected comprehensive regulatory benefits for each species, Total number of species; The calculation process for the benefit balance degree is as follows: Step S331: Calculate the niche weight of each species. The calculation formula is: ; in, For species The endangered status coefficient, For species Ecological function coefficient, For species Protection priority coefficient, This is a temporary index variable used to iterate through all species; Step S332: Calculate the weighted average benefit. : ; Step S333: Calculate the weighted equilibrium degree : ; Step S334: Calculate the weighted balance degree. Used for multi-objective optimization decision-making in step S4.

[0012] Preferably, step S4 specifically includes: Step S41: Establish a multi-objective optimization model based on the expected comprehensive control benefits and overall synergistic benefits indicators of each control scheme; Step S42: Define the objective function and constraints of the multi-objective optimization model, where the objective function is: Optimization Objective 1: Maximize overall benefits ; Optimization Objective 2: Maximize the balance of benefits ; The constraints are: ,in To control the total cost of the plan, The upper limit of the budget; Step S43: Define the specific parameters of the multi-objective optimization algorithm, including: population size, maximum number of iterations, crossover probability, and mutation probability; Step S44: Use the second-generation non-dominated sorting genetic algorithm to optimize the solution; Step S45: Select the Pareto optimal solution set from the final population.

[0013] Preferably, in step S45, the entropy-weighted TOPSIS method is used to select the optimal solution from the Pareto front, specifically including: Step S451: Construct the decision matrix of the Pareto solution set, which includes two indicators: overall benefit and benefit balance. Step S452: Determine the objective weights of the two indicators using the entropy weight method; Step S453: Based on the species demand parameters interpreted in step S1, species demand satisfaction is introduced as the third decision indicator, and the calculation formula is as follows: ; in, As an evaluation unit For species Potential suitability index, As an evaluation unit The area; Step S454: Normalize the decision matrix; Step S455: Determine the positive ideal solution and the negative ideal solution; Step S456: Calculate the Euclidean distance between each solution and the positive and negative ideal solutions; Step S457: Calculate the relative closeness of each solution and select the solution with the highest closeness as the optimal control scheme.

[0014] Preferably, step S5 specifically includes: Step S51: Implement habitat regulation engineering based on the output optimal habitat regulation optimization scheme; Step S52: Obtain actual habitat status data after implementation through the integrated air-space-ground monitoring network; Step S53: Process the actual habitat status data and merge and update the processed data with the habitat database; Step S54: Based on the updated data, re-execute the calculation process of steps S2 and S3 to obtain the actual ecological function value. Actual species suitability index and actual comprehensive regulation benefits Among them, the actual comprehensive regulatory benefits The calculation formula is: ; in, The first one actually monitored Habitat quality values ​​for each evaluation unit; Step S55, Benefit Comparison and Deviation Analysis, specifically includes: Calculate the actual comprehensive regulation benefits Overall regulatory benefits relative error : ; when If the expected regulatory effect is achieved, the current model parameters are maintained. when When this happens, the parameter correction process begins; in For the threshold; Step S56: Model parameter linkage correction; Step S57: Apply the corrected model parameters to the next round of habitat assessment and regulation decisions.

[0015] This invention also provides an intelligent regulation system for plateau afforestation habitats, comprising: The data acquisition and database construction module is used to acquire multi-source data from the target area and build a habitat database. The habitat requirement parameter quantification module is used to interpret and quantify the habitat requirement parameters of at least two target species based on animal ecology knowledge. The habitat assessment and conflict identification module is used to calculate the comprehensive ecological function value of the evaluation unit and the potential suitability index of each species, and to identify areas of conflict in habitat demand among multiple species. The regulation scheme design and pre-evaluation module is used to design multiple regulation schemes based on multi-species habitat demand conflict areas and potential suitability indices, pre-evaluate the expected regulation benefits of each scheme, and calculate the overall synergistic benefit index. The multi-objective optimization decision module is used to optimize each regulation scheme by using a multi-objective optimization algorithm, with the goal of maximizing the overall expected benefits of multiple species and optimizing the balance of benefits among species, and outputting the optimal habitat regulation optimization scheme. The monitoring and closed-loop optimization module is used to implement the optimal plan. It obtains actual habitat status data through the monitoring network, calculates the actual comprehensive regulation benefits and compares them with the expected benefits. When the deviation exceeds the threshold, the actual data is fed back to the habitat assessment and conflict identification module and the regulation plan design and pre-evaluation module to recalculate and update the model parameters.

[0016] Therefore, the present invention employs the above-mentioned intelligent regulation method and system for plateau afforestation habitats, and the beneficial technical effects are as follows: (1) This invention realizes the transformation of habitat quality assessment from static evaluation to dynamic and precise diagnosis and simulation-based optimization decision-making, effectively overcoming the problems of lagging monitoring and experience-dependent decision-making in traditional methods. By integrating multi-source data and constructing ecological function quantification and species suitability models, this invention can perform real-time, quantitative, and dynamic diagnosis of high-altitude habitat structure and animal needs. Furthermore, by utilizing the regulation benefit evaluation model, the habitat improvement effects of different regulation schemes can be simulated and multi-objective collaboratively optimized before the implementation of the project, thereby transforming management decisions from experience-based judgments to scientific decisions based on quantitative simulation and benefit prediction, significantly improving the pertinence and effectiveness of ecological restoration measures.

[0017] (2) An adaptive management mechanism based on real-time feedback and model iteration was constructed, solving the problem that traditional linear management models in ecological engineering are difficult to dynamically adjust and continuously optimize. This invention continuously acquires feedback data on habitat and animal activities through a monitoring network after implementation, and compares and analyzes the actual regulation benefits with the model predictions. When the deviation exceeds the allowable range, the system can automatically trigger and link to correct the key parameters of the underlying evaluation and prediction models, enabling the models to continuously evolve and be used to guide the next round of management decisions. This mechanism ensures that habitat regulation strategies can be dynamically adjusted and iteratively optimized according to actual effects and ecosystem changes, thereby guaranteeing the long-term stability and sustainability of ecological restoration effects.

[0018] (3) This invention quantifies the habitat requirements of multiple species, identifies and optimizes conflict areas, and adopts optimization decision-making aimed at the overall benefits and balance of multiple species. This ensures that habitat regulation measures not only serve vegetation restoration but also directly address the key needs of animal communities. Compared with traditional afforestation methods centered on vegetation, after the implementation of this method, the species richness index, suitable habitat area for key species, and population activity frequency in the target area are significantly improved in a measurable and quantifiable manner. This achieves a fundamental shift from afforestation to habitat creation, providing an effective technical approach for the overall restoration of plateau ecosystems and the protection of biodiversity. Attached Figure Description

[0019] Figure 1 This is a flowchart of an intelligent regulation method for plateau afforestation habitat according to the present invention; Figure 2 Flowchart for habitat assessment and conflict identification; Figure 3 Flowchart for the design and pre-assessment of control measures; Figure 4 Flowchart for multi-objective optimization and solution selection; Figure 5 This is a flowchart for the implementation and closed-loop optimization of the solution. Detailed Implementation

[0020] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0021] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0022] Example 1 like Figures 1-5 As shown, a method for intelligent regulation of highland afforestation habitats includes the following steps: Step S1: Obtain multi-source data of the target area and construct a habitat database. Based on animal ecology knowledge, interpret and quantify the habitat requirement parameters of at least two target species.

[0023] Step S11: Collect remote sensing data, topographic data, meteorological data, soil data, and target species activity monitoring data of the target area.

[0024] Remote sensing data includes at least one of Landsat 8 OLI imagery, Sentinel-2 MSI imagery, and GF-2 PMS imagery, used to acquire vegetation index and land cover information; The terrain data is an ASTER GDEM V3 digital elevation model, used to calculate slope, aspect, topographic relief, and surface roughness. Meteorological data, including annual precipitation, annual average temperature, annual accumulated temperature, and annual evaporation, are sourced from site observation data of the China Meteorological Data Network (http: / / data.cma.cn) and spatial distribution maps are generated using Kriging interpolation. Soil data, including soil type, soil depth, and soil organic matter content, are sourced from the World Soil Database (HWSD) and the China Soil Database. The target species activity monitoring data consists of animal activity images automatically captured by infrared cameras, including species name, activity time, and activity location information.

[0025] Step S12: Use data assimilation technology to perform spatiotemporal matching and fusion of the above multi-source data to construct a habitat database.

[0026] Step S13: Based on animal ecology knowledge, for at least two target species, interpret and quantify the core habitat requirements parameters of each species for concealment, food resources, water sources, reproductive conditions and disturbance distance.

[0027] Step S2: Using the habitat database and habitat demand parameters constructed in Step S1 as input, calculate the comprehensive ecological function value of the evaluation unit, calculate the potential suitability index of each species, and identify the habitat demand conflict areas of multiple species.

[0028] Step S21: Divide the study area into The standard evaluation unit.

[0029] Step S22: Based on the habitat database constructed in step S1, calculate the comprehensive ecological function value of each evaluation unit using the ecological function quantification model. The calculation formula for each sub-function value is as follows: ; in, , , , These are the weighting coefficients for four functions: water conservation, food supply, shelter, and reproduction, to meet the requirements. ; This represents the water conservation function value. Provides functional value to food. For concealment and shelter function value, Values ​​related to reproductive and child-rearing functions; Water conservation function value The calculation formula is: ; in, For the first Vegetation coverage of each evaluation unit For the first Leaf area index of each evaluation unit For the first Annual precipitation for each evaluation unit For the first Soil saturated hydraulic conductivity of each evaluation unit; , , Let be the weighting coefficient, satisfying ; Food supply function value The calculation formula is: ; in, For the first Normalized Difference Vegetation Index of each evaluation unit For the first Herbaceous biomass of each evaluation unit For the first Edible plant abundance in each evaluation unit; , , Let be the weighting coefficient, satisfying ; Concealment and Shelter Function Value The calculation formula is: ; in, For the first Vegetation density of each evaluation unit, For the first Vegetation height of each evaluation unit, For the first Surface roughness of each evaluation unit; , , Let be the weighting coefficient, satisfying ; Reproduction and Child Rearing Function Value The calculation formula is: ; in, For the first The terrain concealment level of each evaluation unit For the first The reciprocal of the distance from the interference source to each evaluation unit. For the first Microclimate suitability of each evaluation unit; , , Let be the weighting coefficient, satisfying .

[0030] Step S23: Based on the habitat requirement parameters interpreted in step S1, construct a species-specific habitat suitability model for each target species, and calculate the potential suitability index of each species in each evaluation unit. The specific process is as follows: Step S231, targeting the target species Select its key environmental variable set Each variable corresponds to a type of habitat requirement parameter in step S1. , , Representing the 1st, 2nd, and so on, respectively. The specific values ​​or types of key environmental variables; Step S232: Calculate the evaluation unit Potential suitability index : ; in, For the intercept term, For the first Environment variables The response coefficient, As an evaluation unit The Middle The values ​​of the environment variables, This represents the total number of environment variables. Step S233: Using the maximum likelihood estimation method, based on the target species activity monitoring data, the response coefficients are... Calibrate the model to maximize the match between the fitness index predicted by the model and the actual probability of species occurrence. Step S234: Calculate all evaluation units The values ​​are normalized and mapped to the [0, 1] interval to obtain the potential fitness index distribution of the species throughout the study area.

[0031] Step S24: Identify areas of conflict in habitat requirements among multiple species by comparing and analyzing the suitability indices of each species. Step S241: For each evaluation unit Calculate the potential fitness index of species A and species B. and ; Step S242: Define the conflict index : ; in, and These are the habitat requirement weights for species A and species B, respectively. Step S243: Set the conflict threshold ,when At that time, the evaluation unit was marked as a habitat demand conflict zone; Step S244: Perform spatial cluster analysis on all conflict areas identified in step S243 to form contiguous habitat demand conflict areas.

[0032] Step S3: Based on the habitat demand conflict areas and potential suitability indices identified in Step S2, design multiple control schemes, pre-evaluate the expected control benefits of each scheme on each species, and calculate the overall synergistic benefit index.

[0033] Step S31: Using the habitat demand conflict area identified in step S2 as the control target area, design a variety of control schemes with different spatial configurations and vegetation structure parameters.

[0034] Step S32: Input the parameters of each regulation scheme into the habitat regulation benefit evaluation model. Based on the suitability index and habitat demand parameters calculated in step S2, simulate in advance the expected habitat quality value for each target species after the implementation of each regulation scheme. and expected comprehensive regulatory benefits Among them, the expected habitat quality value The calculation formula is: ; in, The first day after the implementation of the control plan The first evaluation unit Ecological function value, For the first The weighting coefficient of each ecological function for the target species. The total number of ecological functions; Expected comprehensive regulatory benefits The calculation formula is: ; in, The first day after the implementation of the control plan The expected habitat quality value for each evaluation unit. The total number of evaluation units, This represents the average habitat quality value before regulation.

[0035] Step S33: Based on the expected comprehensive regulation benefits of step S32, calculate the overall synergistic benefit index of the regulation scheme, including overall benefits and benefit balance.

[0036] Overall benefits The calculation formula is: ; in, For the first The expected comprehensive regulatory benefits for each species, Total number of species; The calculation of benefit equilibrium adopts an improved equilibrium model based on species niche weights, which specifically includes: Step S331: Calculate the niche weight of each species. The calculation formula is: ; in, For species The endangered status coefficient, For species Ecological function coefficient, For species Protection priority coefficient, This is a temporary index variable used to iterate through all species; Step S332: Calculate the weighted average benefit. : ; Step S333: Calculate the weighted equilibrium degree : ; Step S334: Calculate the weighted balance degree. Used for multi-objective optimization decision-making in step S4.

[0037] Step S4: Based on the expected comprehensive regulatory benefits and overall synergistic benefits of each regulation scheme obtained in Step S3, and with the goal of maximizing the overall expected benefits of multiple species and optimizing the balance of benefits among species, a multi-objective optimization algorithm is used to find the optimal regulation scheme for each scheme and output the optimal habitat regulation optimization scheme.

[0038] Step S41: Establish a multi-objective optimization model based on the expected comprehensive control benefits and overall synergistic benefit indicators of each control scheme obtained in step S3.

[0039] Step S42: Define the objective function and constraints of the multi-objective optimization model, where the objective function is: Optimization Objective 1: Maximize overall benefits ; Optimization Objective 2: Maximize the balance of benefits ; The constraints are: ,in To control the total cost of the plan, This is the budget ceiling.

[0040] Step S43: Define the specific parameters of the multi-objective optimization algorithm, including: The population size is set to 100; The maximum number of iterations is set to 200; The crossover probability is set to 0.8; The mutation probability is set to 0.1.

[0041] Step S44: Optimize the solution using the second-generation non-dominated sorting genetic algorithm. The specific process includes: a) Initialize the population: Randomly generate 100 combinations of control schemes as the initial population; b) Non-dominated sorting: Based on the optimization objective function in step S42, the individuals in the population are sorted in a non-dominated manner; c) Crowding Calculation: Calculate the crowding distance for each individual; d) Selection, crossover, and mutation: Generate offspring populations through tournament selection, simulated binary crossover, and polynomial mutation; e) Elite preservation strategy: merge the parent and offspring populations and select the best 100 individuals to enter the next generation; f) Iterative optimization: Repeat steps b-e until the maximum number of iterations is reached.

[0042] Step S45: Select the Pareto optimal solution set from the final population; Step S451: Construct the decision matrix of the Pareto solution set, which includes two indicators: overall benefit and benefit balance. Step S452: Determine the objective weights of the two indicators using the entropy weight method; Step S453: Based on the species demand parameters interpreted in step S1, species demand satisfaction is introduced as the third decision indicator, and the calculation formula is as follows: ; in, As an evaluation unit For species Potential suitability index, As an evaluation unit The area; Step S454: Normalize the decision matrix; Step S455: Determine the positive ideal solution and the negative ideal solution; Step S456: Calculate the Euclidean distance between each solution and the positive and negative ideal solutions; Step S457: Calculate the relative closeness of each solution and select the solution with the highest closeness as the optimal control scheme.

[0043] Step S5: Implement the optimal habitat regulation optimization scheme output in step S4, obtain actual habitat status data through the monitoring network, calculate the actual comprehensive regulation benefit and compare it with the expected comprehensive regulation benefit obtained in step S3. When the deviation exceeds the threshold, feed back the actual habitat status data to steps S2 and S3, recalculate and update the model.

[0044] Step S51: Implement habitat regulation engineering based on the optimal habitat regulation optimization scheme output in step S4, specifically including: Determine the vegetation configuration type and density of each region based on the spatial configuration map; Replanting or thinning of trees, shrubs and herbaceous plants according to design parameters; Construct artificial nests, watering points, and concealed structures.

[0045] Step S52: Obtain actual habitat status data after implementation through an integrated air-space-ground monitoring network, specifically including: Aerial surveys were conducted weekly using a DJI Matrice M300 RTK drone to acquire multispectral imagery and lidar point cloud data at a resolution of 0.05 meters. Animal activity images are automatically captured and uploaded every 24 hours via the Ltl-Acorn infrared camera network; Environmental sound data was recorded every 30 minutes using the Wildlife Acoustics SM4 acoustic monitoring station; Soil temperature and humidity data are collected every 6 hours using the Decagon 5™ soil moisture sensor.

[0046] Step S53, data processing and analysis, specifically includes: Orthorectification and classification of UAV imagery are performed to extract vegetation coverage, height, and density. Species identification and population statistics are performed on infrared camera images; Soundscape index calculation and species call identification are performed on acoustic monitoring data; The processed data is then merged and updated with the habitat database constructed in step S1.

[0047] Step S54: Based on the updated data, re-execute the calculation process of steps S2 and S3 to obtain the actual ecological function value. Actual species suitability index and actual comprehensive regulation benefits Among them, the actual comprehensive regulatory benefits The calculation formula is: ; in, The first one actually monitored The habitat quality value of each evaluation unit.

[0048] Step S55, Benefit Comparison and Deviation Analysis, specifically includes: Calculate the actual comprehensive regulation benefits Combined with the expected comprehensive regulatory benefits obtained in step S3 relative error : ; Set deviation threshold ; when If the expected regulatory effect is achieved, the current model parameters are maintained. when When this happens, the parameter correction process begins; in The threshold value is used.

[0049] Step S56, Model parameter linkage correction, specifically includes: The model parameter linkage correction employs a reinforcement learning algorithm based on multi-model collaboration, specifically including: Step S561: Define the collaborative state space, which includes the parameter states of three models: the parameter state of the ecological function quantification model. Species suitability model parameter status Parameter status of regulation benefit evaluation model ; Step S562: Define the cooperative action space, which includes combinations of three types of actions: (Adjusting parameters of the ecological function quantification model) (Adjusting the parameters of the species suitability model) (Adjusting the parameters of the regulation and control benefit evaluation model); Step S563: Design the collaborative reward function : ; in, and These are the actual and predicted ecological function values, respectively. and These are the actual and predicted species fitness indices, respectively. and For collaborative weighting coefficients; Step S564: Collaborative training is performed using a multi-agent reinforcement learning framework. The specific process is as follows: Step S5641: Model the ecological function quantification model, species suitability model, and regulation benefit evaluation model as intelligent agents respectively. , , ; Step S5642: Each agent, based on the current state... Select Action ,in , ; Step S5643: Perform the combined action Afterwards, the system receives an immediate reward. and transition to the new state. ; Step S5644: Each agent updates its Q-value function using a distributed Q-learning algorithm; Step S5645: Agents learn collaboratively through a parameter sharing mechanism, with each agent considering the strategies of other agents when updating its Q-value; Step S5646: Repeat steps S5642 to S5645 for multiple rounds of iterative training until the Q-value function converges or the maximum number of training rounds is reached. Step S565: Determine the optimal parameter combination based on the collaborative training results. The specific process is as follows: Step S5651: After training is complete, each agent obtains its optimal policy. , indicating the state The action to be chosen below; Step S5652: For the given initial model parameter state By iteratively executing the optimal joint action sequence, the final parameter adjustment scheme is obtained; Step S5653: Calculate the collaborative prediction bias after parameter adjustment. : ; Step S5654, Select to use The minimum parameter combination is taken as the optimal parameter and applied to the next round of habitat regulation decisions.

[0050] Step S57, closed-loop iterative optimization, specifically includes: The revised model parameters will be applied to the next round of habitat assessment and regulation decisions; Continue to monitor, evaluate, and correct to form a continuously optimized adaptive closed-loop management system.

[0051] Example 2 A smart habitat regulation system for plateau afforestation includes: The data acquisition and database construction module is used to acquire multi-source data from the target area and build a habitat database. The habitat requirement parameter quantification module is used to interpret and quantify the habitat requirement parameters of at least two target species based on animal ecology knowledge. The habitat assessment and conflict identification module is used to calculate the comprehensive ecological function value of the evaluation unit and the potential suitability index of each species, and to identify areas of conflict in habitat demand among multiple species. The regulation scheme design and pre-evaluation module is used to design multiple regulation schemes based on multi-species habitat demand conflict areas and potential suitability indices, pre-evaluate the expected regulation benefits of each scheme, and calculate the overall synergistic benefit index. The multi-objective optimization decision module is used to optimize each regulation scheme by using a multi-objective optimization algorithm, with the goal of maximizing the overall expected benefits of multiple species and optimizing the balance of benefits among species, and outputting the optimal habitat regulation optimization scheme. The monitoring and closed-loop optimization module is used to implement the optimal plan. It obtains actual habitat status data through the monitoring network, calculates the actual comprehensive regulation benefits and compares them with the expected benefits. When the deviation exceeds the threshold, the actual data is fed back to the habitat assessment and conflict identification module and the regulation plan design and pre-evaluation module to recalculate and update the model parameters.

[0052] It is worth noting that all contents not described in detail in this invention are existing technologies and are well known to those skilled in the art.

[0053] Therefore, the present invention adopts the above-mentioned intelligent regulation method and system for plateau afforestation habitat, which monitors key habitat parameters in real time through a deployed sensor network, and dynamically drives the regulation system based on an intelligent decision model, thereby realizing refined and dynamic regulation and continuous optimization of habitat quality in afforestation areas and improving the diversity of mammals and birds in the plateau.

[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for intelligent regulation of highland afforestation habitats, characterized in that, Includes the following steps: Step S1: Obtain multi-source data of the target area and construct a habitat database. Based on animal ecology knowledge, interpret and quantify the habitat requirement parameters of at least two target species. Step S2: Divide the study area into standard evaluation units. Based on the constructed habitat database, use the ecological function quantification model to calculate the function values ​​and comprehensive ecological function values ​​of the four ecological functions of water conservation, food supply, shelter and protection, and reproduction and rearing in each evaluation unit. Based on the habitat demand parameters interpreted in Step S1, construct species-specific habitat suitability models for each target species, calculate the potential suitability index of each species in each evaluation unit, and identify multi-species habitat demand conflict areas by comparing and analyzing the suitability indices of each species. The comprehensive ecological function value is a weighted sum of the water conservation function value, food supply function value, shelter and protection function value, and reproduction and rearing function value. Step S3: Based on the multi-species habitat demand conflict zone and potential suitability index, design a variety of afforestation habitat regulation schemes with different spatial configurations and vegetation structure parameters. Input the parameters of each regulation scheme into the habitat regulation benefit evaluation model, pre-evaluate the expected comprehensive regulation benefits of each regulation scheme on each species, and calculate the overall synergistic benefit index. Expected comprehensive regulatory benefits The calculation formula is: ； in, The first day after the implementation of the control plan The expected habitat quality value for each evaluation unit. The total number of evaluation units, The average habitat quality value before regulation; Expected habitat quality value The calculation formula is: ； in, The first day after the implementation of the control plan The first evaluation unit Ecological function value, For the first The weighting coefficient of each ecological function for the target species. The total number of ecological functions; The overall synergistic benefit index includes overall benefit and benefit balance. The overall benefit is the sum of the expected comprehensive regulatory benefits of each species, and the benefit balance is a weighted balance calculated based on the niche weight of each species. The calculation process for the benefit balance degree is as follows: Step S331: Calculate the niche weight of each species. The calculation formula is: ； in, For species The endangered status coefficient, For species Ecological function coefficient, For species Protection priority coefficient, This is a temporary index variable used to iterate through all species. Total number of species; Step S332: Calculate the weighted average benefit. : ； in, For the first The expected comprehensive regulatory benefits for each species; Step S333: Calculate the weighted equilibrium degree : ； Step S4: Establish a multi-objective optimization model based on the expected comprehensive regulatory benefits and overall synergistic benefits of each regulation scheme. With the goal of maximizing the overall expected benefits of multiple species and optimizing the balance of benefits among species, use a multi-objective optimization algorithm to find the best regulation scheme and output the optimal habitat regulation optimization scheme. Step S5: Implement the optimal habitat regulation and optimization scheme. Obtain actual habitat status data through the monitoring network, calculate the actual comprehensive regulation benefits and compare them with the expected comprehensive regulation benefits obtained in step S3. When the deviation exceeds the threshold, feed back the actual habitat status data to steps S2 and S3, recalculate and update the model.

2. The intelligent regulation method for plateau afforestation habitat according to claim 1, characterized in that, Step S1 specifically includes: Step S11: Collect remote sensing data, topographic data, meteorological data, soil data, and target species activity monitoring data of the target area; Step S12: Perform spatiotemporal matching and fusion of multi-source data to construct a habitat database; Step S13: For at least two target species, interpret and quantify the habitat requirements parameters of each species for concealment, food resources, water sources, reproductive conditions and disturbance distance.

3. The intelligent regulation method for plateau afforestation habitat according to claim 1, characterized in that, Step S2 specifically includes: Step S21: Divide the study area into standard evaluation units; Step S22: Based on the constructed habitat database, calculate the comprehensive ecological function value of each evaluation unit using the ecological function quantification model. The calculation formula for each sub-function value is as follows: ； ； ； ； ； in, , , , These are the weighting coefficients for four functions: water conservation, food supply, shelter, and reproduction. This represents the water conservation function value. Provides functional value to food. For concealment and shelter function value, For reproductive and parenting function values, For the first Vegetation coverage of each evaluation unit For the first Leaf area index of each evaluation unit For the first Annual precipitation for each evaluation unit For the first Soil saturated hydraulic conductivity of each evaluation unit , , These are the weighting coefficients. For the first Normalized Difference Vegetation Index of each evaluation unit For the first Herbaceous biomass of each evaluation unit For the first Edible plant abundance in each evaluation unit , , These are the weighting coefficients. For the first Vegetation density of each evaluation unit, For the first Vegetation height of each evaluation unit, For the first Surface roughness of each evaluation unit, , , These are the weighting coefficients. For the first The terrain concealment level of each evaluation unit For the first The reciprocal of the distance from the interference source to each evaluation unit. For the first Microclimate suitability of each evaluation unit , , These are the weighting coefficients; Step S23: Based on the habitat requirement parameters interpreted in step S1, construct a species-specific habitat suitability model for each target species, and calculate the potential suitability index of each species in each evaluation unit. The specific process is as follows: Step S231, targeting the species Select its key environmental variable set Each variable corresponds to a type of habitat requirement parameter in step S1. , , Representing the 1st, 2nd, and so on, respectively. The specific values ​​or types of key environmental variables; Step S232: Calculate the evaluation unit Potential suitability index : ； in, For the intercept term, For the first Environment variables The response coefficient, As an evaluation unit Middle The values ​​of the environment variables, This represents the total number of environment variables. is the base of the natural logarithm; Step S233: Using the maximum likelihood estimation method, based on the target species activity monitoring data, the response coefficients are... Perform calibration; Step S234: Calculate all evaluation units The values ​​were normalized to obtain the distribution of the potential fitness index of this species throughout the study area; Step S24: Identify areas of conflict in habitat requirements among multiple species by comparing and analyzing the suitability indices of each species.

4. The intelligent regulation method for plateau afforestation habitat according to claim 3, characterized in that, In step S24, the identification of habitat demand conflict areas among multiple species is achieved through the following methods: Step S241: For each evaluation unit Calculate the potential fitness index of species A and species B. and ; Step S242: Define the conflict index : ； in, and These are the habitat requirement weights for species A and species B, respectively. Step S243: Set the conflict threshold ,when At that time, the evaluation unit was marked as a habitat demand conflict zone; Step S244: Perform spatial cluster analysis on all identified habitat demand conflict areas to form contiguous habitat demand conflict areas.

5. The intelligent regulation method for plateau afforestation habitat according to claim 4, characterized in that, Step S4 specifically includes: Step S41: Establish a multi-objective optimization model based on the expected comprehensive control benefits and overall synergistic benefits indicators of each control scheme; Step S42: Define the objective function and constraints of the multi-objective optimization model, where the objective function is: Optimization Objective 1: Maximize overall benefits ; Optimization Objective 2: Maximize the balance of benefits ; The constraints are: ,in To control the total cost of the plan, The upper limit of the budget; Step S43: Define the specific parameters of the multi-objective optimization algorithm, including: population size, maximum number of iterations, crossover probability, and mutation probability; Step S44: Use the second-generation non-dominated sorting genetic algorithm to optimize the solution; Step S45: Select the Pareto optimal solution set from the final population.

6. The intelligent regulation method for plateau afforestation habitat according to claim 5, characterized in that, In step S45, the entropy-weighted TOPSIS method is used to select the optimal solution from the Pareto front, specifically including: Step S451: Construct the decision matrix of the Pareto solution set, which includes two indicators: overall benefit and benefit balance. Step S452: Determine the objective weights of the two indicators using the entropy weight method; Step S453: Based on the species demand parameters interpreted in step S1, species demand satisfaction is introduced as the third decision indicator, and the calculation formula is as follows: ； in, As an evaluation unit For species Potential suitability index, As an evaluation unit The area; Step S454: Normalize the decision matrix; Step S455: Determine the positive ideal solution and the negative ideal solution; Step S456: Calculate the Euclidean distance between each solution and the positive and negative ideal solutions; Step S457: Calculate the relative closeness of each solution and select the solution with the highest closeness as the optimal control scheme.

7. The intelligent regulation method for plateau afforestation habitat according to claim 6, characterized in that, Step S5 specifically includes: Step S51: Implement habitat regulation engineering based on the output optimal habitat regulation optimization scheme; Step S52: Obtain actual habitat status data after implementation through the integrated air-space-ground monitoring network; Step S53: Process the actual habitat status data and merge and update the processed data with the habitat database; Step S54: Based on the updated data, re-execute the calculation process of steps S2 and S3 to obtain the actual ecological function value. Actual species suitability index and actual comprehensive regulation benefits Among them, the actual comprehensive regulatory benefits The calculation formula is: ； in, The first one actually monitored Habitat quality values ​​for each evaluation unit; Step S55, Benefit Comparison and Deviation Analysis, specifically includes: Calculate the actual comprehensive regulation benefits Overall regulatory benefits relative error : ； when If the adjustment effect is considered to have met expectations, the current model parameters will be maintained. when When this happens, the parameter correction process begins; in For the threshold; Step S56: Model parameter linkage correction; Step S57: Apply the corrected model parameters to the next round of habitat assessment and regulation decisions.

8. A smart regulation system for plateau afforestation habitats, characterized in that, The method for intelligent regulation of plateau afforestation habitat as described in any one of claims 1-7 includes: The data acquisition and database construction module is used to acquire multi-source data from the target area and build a habitat database. The habitat requirement parameter quantification module is used to interpret and quantify the habitat requirement parameters of at least two target species based on animal ecology knowledge. The habitat assessment and conflict identification module is used to calculate the comprehensive ecological function value of the evaluation unit and the potential suitability index of each species, and to identify areas of conflict in habitat demand among multiple species. The regulation scheme design and pre-evaluation module is used to design multiple regulation schemes based on multi-species habitat demand conflict areas and potential suitability indices, pre-evaluate the expected regulation benefits of each scheme, and calculate the overall synergistic benefit index. The multi-objective optimization decision module is used to optimize each regulation scheme by using a multi-objective optimization algorithm, with the goal of maximizing the overall expected benefits of multiple species and optimizing the balance of benefits among species, and outputting the optimal habitat regulation optimization scheme. The monitoring and closed-loop optimization module is used to implement the optimal plan. It obtains actual habitat status data through the monitoring network, calculates the actual comprehensive regulation benefits and compares them with the expected benefits. When the deviation exceeds the threshold, the actual data is fed back to the habitat assessment and conflict identification module and the regulation plan design and pre-evaluation module to recalculate and update the model parameters.

Images