A method and system for delineating a wildlife buffer zone of a habitat

Abstract

The application provides a habitat wild animal buffer zone delineation method and system, relates to the technical field of ecological environment division, and comprises the following steps: acquiring habitat information data; processing the habitat information data based on a quantum hybrid optimization algorithm and a preset multi-fractal coupling model to generate a habitat suitability distribution pattern; performing multi-scale landscape connectivity analysis according to the suitability distribution pattern to obtain a connectivity topology atlas of the habitat space; constructing and coupling a membrane calculation ecological network model for the connectivity topology atlas to obtain an initial wild animal buffer zone delineation scheme; generating an optimized buffer zone delineation scheme through a random programming and a preset robust optimization hybrid strategy; and performing collaborative evaluation by using a q-order fuzzy measure and a Choquet integral to obtain an intelligent wild animal buffer zone delineation scheme. The application objectively quantifies the complex trade-off relationship among multiple targets and ensures the optimal balance of the scheme in all aspects.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ecological environment division, and in particular, relates to a method and system for delineating a wild animal buffer zone of a habitat. BACKGROUND

[0002] In the prior art, scientifically delineating an ecological buffer zone is a key technical means for coordinating biodiversity protection and human activity interference. At present, the mainstream method in this field relies on static environmental data provided by geographic information systems (GIS) and remote sensing technology to evaluate habitat suitability. Although this traditional technology can achieve preliminary planning, it has significant limitations:

[0003] First, most models are based on linear assumptions and are difficult to depict the nonlinear, multiscale characteristics and complex interactions between ecological elements that exist universally in ecological environments;

[0004] Second, existing methods lack robustness in considering uncertain factors such as climate change and human activities, resulting in insufficient adaptability and easy failure of the planned buffer zone in a dynamic environment;

[0005] Third, the final decision of the scheme mostly relies on expert experience for weight allocation or simple weighted summation, making it difficult to objectively quantify the complex trade-offs and synergies between multiple benefit targets, thereby affecting the overall optimality and scientificity of the scheme.

[0006] Therefore, there is an urgent need for a method and system for delineating a wild animal buffer zone of a habitat to solve the above technical problems. SUMMARY

[0007] The present application aims to provide a method and system for delineating a wild animal buffer zone of a habitat to improve the above problems. To achieve the above purpose, the technical solution adopted by the present application is as follows:

[0008] In a first aspect, the present application provides a method for delineating a wild animal buffer zone of a habitat, comprising:

[0009] Obtaining habitat terrain, vegetation, climate, species distribution, human activity, and protection level data fused with multi-source remote sensing and dynamic monitoring of the Internet of Things, and constructing habitat information data;

[0010] Processing the habitat information data based on a quantum hybrid optimization algorithm and a preset multi-fractal coupling model to generate a habitat suitability distribution pattern;

[0011] According to the suitability distribution pattern, using topological data analysis and discrete exterior differential theory to analyze multiscale landscape connectivity, obtaining a connectivity topological atlas of the habitat space;

[0012] The connectivity topological atlas is subjected to a membrane computing ecological network model construction and coupling to obtain an initial wild animal buffer zone delineation scheme;

[0013] The initial wild animal buffer zone delineation scheme is processed by a hybrid strategy of random programming and preset robust optimization to generate an optimized buffer zone delineation scheme.

[0014] In the second aspect, the application further provides a wild animal buffer zone delineation system for a habitat, comprising:

[0015] An acquisition unit is configured to acquire habitat terrain, vegetation, climate, species distribution, human activity and protection level data fused with multi-source remote sensing and Internet of Things dynamic monitoring, and construct habitat information data;

[0016] A processing unit is configured to process the habitat information data based on a quantum hybrid optimization algorithm and a preset multi-fractal coupling model to generate a habitat suitability distribution pattern;

[0017] An analysis unit is configured to analyze a multi-scale landscape connectivity according to the suitability distribution pattern by using topological data analysis and discrete exterior differential theory to obtain a connectivity topological atlas of the habitat space;

[0018] A coupling unit is configured to subject the connectivity topological atlas to a membrane computing ecological network model construction and coupling to obtain an initial wild animal buffer zone delineation scheme;

[0019] An optimization unit is configured to subject the connectivity topological atlas to a membrane computing ecological network model construction and coupling to generate an optimized buffer zone delineation scheme.

[0020] The application has the following beneficial effects:

[0021] The application first realizes a cross-scale integration from a macro ecological pattern to a micro ecological process. Macro environmental data monitored by remote sensing are combined with ecological parameters such as species distribution, the spatiotemporal differentiation law of habitat suitability is revealed by quantum optimization and multi-fractal model, the multi-dimensional characteristics of landscape connectivity are analyzed by topological data analysis and discrete exterior differential theory, and finally the self-organizing process of ecological flow is simulated by membrane calculation to form a full-chain analysis capability of "pattern-process-function", which overcomes the limitations of single-scale analysis of traditional methods.

[0022] Secondly, the application establishes a dynamic optimization mechanism suitable for uncertain environment, converts uncertain factors such as climate change and human activity interference into quantitative constraint conditions through a mixed strategy of stochastic programming and robust optimization, so that the buffer zone scheme has the adaptability to respond to environmental changes. Meanwhile, the complex trade-off relationship between multiple objectives is objectively quantified by using q-order fuzzy measure and Choquet integral, so as to ensure the optimal balance between ecological benefit, economic benefit and social benefit.

[0023] Finally, in the application, a multi-dimensional, multi-scale, dynamic and future-oriented wildlife protection zone delineation method is reconstructed by habitat integrity, and by fusing quantum computing, topological data analysis, membrane computing, random optimization and other technologies, the method aims to intelligently and adaptively reconstruct a complete and vital habitat system that can long-term maintain biodiversity and promote species reproduction and exchange from three dimensions of spatial structure, ecological function and system robustness, and then generate an intelligent wildlife buffer zone delineation scheme with optimal comprehensive benefits.

[0024] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art upon examination of the following or can be learned by practice of the application. The objects and other advantages of the application can be realized and attained by means of the instrumentalities particularly pointed out in the written description and claims hereof as well as the appended drawings. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some of the embodiments of the present application, and therefore should not be regarded as a limitation on the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0026] Figure 1 For the wildlife buffer zone delineation method of the habitat described in the embodiments of the present application;

[0027] Figure 2 For the wildlife buffer zone delineation system structure diagram of the habitat described in the embodiments of the present application.

[0028] In the figure: 701, acquisition unit; 702, processing unit; 703, parsing unit; 704, coupling unit; 705, optimization unit. DETAILED DESCRIPTION

[0029] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0030] It should be noted that similar reference numerals and letters represent similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. Meanwhile, in the description of the present application, the terms "first", "second", and the like are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0031] Embodiment 1

[0032] The present embodiment provides a method for delineating a wildlife buffer zone of a habitat.

[0033] Referring to Figure 1 , the method includes steps S1, S2, S3, S4 and S5.

[0034] Step S1, acquiring habitat terrain, vegetation, climate, species distribution, human activity and protection level data fused with multi-source remote sensing and Internet of Things dynamic monitoring, and constructing habitat information data;

[0035] It can be understood that this step constructs a multi-dimensional, multi-scale dynamic habitat information database by integrating multi-source remote sensing technology (such as multispectral / hyperspectral remote sensing, laser radar, synthetic aperture radar) and Internet of Things ground sensing network (such as weather stations, camera traps, acoustic monitoring equipment). This step breaks through the limitations of traditional single data sources: remote sensing data provides large-scale, periodic land cover and terrain features, while Internet of Things devices capture fine-scale data such as microclimate conditions, species activity trajectories and human disturbance intensity in real time. Through data assimilation technology, heterogeneous data with different temporal and spatial resolutions are fused and standardized to form a multi-dimensional data cube with a unified geographic coordinate system and time stamp. This process not only includes static environmental factors, but more importantly, introduces "protection level dynamic monitoring data", such as periodically updated protection area management intensity index, human activity pressure change trend, etc., providing input basis reflecting the dynamic changes of the ecological system for subsequent models.

[0036] Step S2, based on the quantum hybrid optimization algorithm and the preset multifractal coupling model, processing the habitat information data to generate a habitat suitability distribution pattern;

[0037] It can be understood that this step realizes the accurate description of the habitat suitability distribution pattern through the synergistic effect of the quantum hybrid optimization algorithm and the multifractal coupling model. Firstly, the quantum tunneling effect and parallel computing advantage of the quantum annealing algorithm are used to perform global optimization search in a high-dimensional solution space composed of ecological parameters such as terrain undulation, vegetation cover heterogeneity, and climate factor volatility, effectively avoiding the problem that traditional optimization algorithms are prone to local optimization, and quickly determining the optimal parameter initial value of the multifractal model. Subsequently, based on the multifractal analysis of the box counting method and the matrix method, the singular spectrum function and the Holder index of the time series of ecological factors such as vegetation index sequence and terrain height sequence are calculated, quantifying the spatial variation characteristics of each ecological element. By embedding the multifractal spectrum characteristics (such as spectrum width, spectrum asymmetry) as nonlinear weight factors into the response function of the MaxEnt model, the suitability probability distribution function under the constraint of the fractal dimension is constructed. Finally, the quantum particle swarm optimization algorithm is used to iteratively fine-tune the model parameters, with the maximum spatial correlation between the suitability probability distribution and the known species distribution points as the objective function, to further improve the model accuracy. It can be understood that in this step, step S2 includes step S21, step S22, step S23 and step S24.

[0038] Step S21, according to the habitat information data, quantum encoding and initialization processing of model parameters are carried out, wherein the quantum annealing algorithm is used to search for the global optimal solution in the solution space composed of habitat terrain, vegetation, climate, species distribution, human activity and protection level, and the initial parameter set of the multifractal coupling model is obtained;

[0039] It can be understood that this step converts complex ecological parameters (terrain relief, vegetation cover heterogeneity, climate volatility, etc.) into quantum bit representation through quantum encoding, and uses the quantum tunneling effect of quantum annealing algorithm to search for the global optimal solution in the high-dimensional solution space composed of multiple ecological parameters. The process first establishes the mapping relationship between ecological parameters and quantum bits, converts continuous ecological variables into discrete quantum state superposition form through quantum encoding; then an energy function with parameter fitting degree as the target is constructed, and through the coherent tunneling effect of quantum bits in the quantum annealing process, the limitation of traditional optimization algorithm that is easy to fall into local optimum is broken through, and the characteristic parameter combination that makes the multi-fractal coupled model reach the best initial state is quickly found. This step provides a globally optimized high-precision initial parameter set for subsequent suitability modeling, significantly improves the model convergence speed and calculation efficiency, and lays a parameter foundation for accurately depicting the spatio-temporal heterogeneity of habitats. The energy function with parameter fitting degree as the target is as follows:

[0040] ; wherein, represents the core target to be minimized, q represents a sequence of quantum bits (0 or 1), n represents the number of observation samples, d represents the dth location, y d represents the true record of whether the species exists at the dth location (1 for existence and 0 for non-existence), represents the calculated species existence probability prediction value according to the environmental variables and model parameters e d and model parameters , λ represents a hyperparameter, m represents the number of model parameters, represents the zth model parameter, represents the initial estimate of the kth parameter; γ represents the fractal constraint strength coefficient. D r represents the predicted rth fractal dimension value, represents the rth fractal dimension target value calculated from the real landscape data (such as vegetation map, terrain map), p represents the total number of fractal dimensions.

[0041] Step S22, according to the initial parameter set, the multi-fractal singularity spectrum calculation processing of the habitat ecological elements is carried out, wherein based on the multi-fractal analysis of box counting method and matrix method, the singularity spectrum function and Holder index of vegetation index sequence and terrain height sequence are calculated respectively, and the multi-fractal spectrum characteristics are obtained;

[0042] It can be understood that this step is based on the initial parameter set obtained by quantum optimization, and the multi-fractal analysis technology combining box counting method and moment method is used to analyze the multi-scale characteristics of ecological elements such as vegetation index sequence and terrain height sequence. This step first divides the ecological element data into different scale grid units through box counting method, calculates the probability distribution function at each scale, and reveals the spatial heterogeneity characteristics of ecological parameters; then the moment method is used to analyze the scale behavior under different order moments, and the singular spectrum function and Holder index representing the multi-fractal characteristics of the system are obtained through Legendre transformation. Among them, the width of the singular spectrum function reflects the amplitude range of the fluctuation of the ecological elements, and the Holder index quantifies the local singularity strength of each point. This multi-fractal analysis method can effectively capture the self-similar structure and scale-dependent characteristics widely existing in the ecological system, and through calculation, the multi-fractal spectrum characteristics describing the spatial variation pattern and complexity of the habitat ecological elements are obtained, which provides a key mathematical representation basis for constructing the suitability probability distribution model under the constraint of fractal dimension.

[0043] Step S23, performing suitability probability calculation processing based on fractal dimension constraint according to the obtained multi-fractal spectrum characteristics, wherein the multi-fractal spectrum characteristics are used as nonlinear weight factors to construct a suitability probability distribution function under the constraint of fractal dimension, and the suitability probability distribution of the habitat is calculated;

[0044] It can be understood that this step is based on the obtained multi-fractal spectrum characteristics, and the width of the singular spectrum function and the spectrum asymmetry are converted into nonlinear weight factors and embedded into the response function of the traditional maximum entropy model to construct a suitability probability distribution function with fractal dimension constraint. This processing process first quantifies the spatial variation intensity of the ecological factor through the singular spectrum width, which is used as an adaptive parameter to adjust the curvature of the response function; then the spectrum asymmetry is used to represent the bias of the ecological factor fluctuation, which is used as a weight factor to adjust the contribution of different ecological variables; finally, the distribution characteristics of the Holder index are used as a spatial constraint condition to ensure that the generated suitability probability distribution can maintain the multi-scale structure characteristics consistent with the original ecological data. This probability modeling method under the constraint of fractal dimension can effectively capture the nonlinear response relationship and scale-dependent characteristics in the ecological system, and produces a habitat suitability probability distribution that is more consistent with the ecological mechanism, providing a spatial probability surface with clear ecological significance for subsequent landscape connectivity analysis.

[0045] Among them, the suitability probability distribution function with fractal dimension constraint is as follows: ; wherein, is the suitability probability with fractal dimension constraint, y=1 is the dependent variable, c is the independent variable, β a is the basic suitability level, β kthe response coefficient of the kth environmental variable, the environmental characteristic function: the transformed form of vegetation index, terrain index, etc. the spectral width of the kth environmental variable, B k the spectral asymmetry of the kth environmental variable, K represents the total number of environmental variables.

[0046] Step S24, constructing a fitness function aiming to maximize the spatial correlation between the suitability probability distribution and the species distribution based on the habitat suitability probability, and iteratively adjusting the parameters of the fractal coupling model using a quantum particle swarm optimization algorithm to obtain the habitat suitability distribution pattern with spatiotemporal heterogeneity characteristics.

[0047] It can be understood that this step constructs a fitness function aiming to maximize the spatial correlation, and uses a particle swarm optimization algorithm to finely adjust the parameters of the fractal coupling model. This process first constructs a target function considering spatial autocorrelation based on the spatial distribution characteristics of the species distribution point data and the suitability probability surface. The target function is to quantify the spatial matching degree of the two by using Moran's index statistical quantity. Then, the parallel search capability of the particle swarm optimization algorithm is used to efficiently optimize in the parameter solution space. The particle position update mechanism of quantum coding is used to synchronously adjust multiple fractal parameters and response function coefficients. The quantum rotation gate operation is introduced in the iteration process to update the particle state. This optimization process can effectively handle nonlinear optimization problems in high-dimensional parameter space, generating habitat suitability distribution patterns with significant spatiotemporal heterogeneity characteristics. This pattern not only maintains the multiple fractal characteristics of the ecosystem, but also presents high spatial consistency with the actual species distribution data, providing reliable spatial input data for subsequent landscape connectivity analysis.

[0048] Step S3, based on the suitability distribution pattern, using topological data analysis and discrete exterior differential theory to analyze the multi-scale landscape connectivity, obtaining the connectivity topological atlas of the habitat space;

[0049] It can be understood that this step uses a method combining topological data analysis and discrete exterior differential theory to realize multi-scale analysis of landscape connectivity, breaking through the limitation of traditional landscape connectivity indices that can only provide single-scale information, and realizing full-scale connectivity representation from micro-patch connection to macro-landscape pattern, providing a spatial framework with mathematical rigor and ecological rationality for subsequent ecological corridor design. In this step, step S3 includes step S31, step S32, and step S33.

[0050] Step S31, according to the habitat suitability distribution pattern, the topological feature extraction processing based on persistent homology theory is carried out, and a topological invariant feature set describing the multi-scale connectivity of habitat patches and the structure of ecological barriers is obtained;

[0051] It can be understood that this step is based on the habitat suitability distribution pattern generated in the previous step, and multi-scale topological feature extraction is carried out by using persistent homology theory. This process first converts the continuous probability distribution surface into a series of binary habitat patch layers (the threshold range is usually set to [0.3, 0.7], and the step is 0.05) through dynamic threshold method, forming a topological space filter with parameterized characteristics. By calculating the simple complex homology group of each topological space, the Betti number (β0 represents the number of connected components, and β1 represents the number of ring structures) at different scales is obtained, and a persistent barcode graph is generated.

[0052] Among them, by analyzing the death time distribution of the barcode, the spatial position and intensity characteristics of the ecological barrier are identified - long persistence interval corresponds to stable core habitat block (such as the area with a connected component number persistence interval exceeding 0.2 threshold span), and short interval reflects transitional characteristics or noise (such as the hollow structure with a ring structure number persistence interval less than 0.05 threshold span). This step breaks through the scale limitation of traditional landscape index method, and for the first time realizes the quantitative representation of multi-scale connectivity from micro patch (hundred meter scale) to macro landscape (kilometer scale). The topological invariant feature set generated by it provides a spatial topological constraint condition with clear mathematical meaning for subsequent ecological flow field modeling.

[0053] Step S32, according to the topological invariant feature set and habitat information data, the ecological flow field modeling processing based on discrete exterior differential theory is carried out, and a discrete exterior differential model describing the ecological flow movement characteristics of the landscape surface is obtained;

[0054] It can be understood that this step is based on the topological invariant feature set, combined with high-precision terrain data, vegetation resistance and hydrological characteristics and other multi-source habitat information, and uses discrete exterior differential theory to construct an ecological flow field model. This process first discretizes the landscape surface into triangular mesh elements, where each mesh vertex is assigned a niche potential value derived from the suitability distribution, and the mesh edge is defined as an ecological resistance coefficient according to parameters such as terrain slope and vegetation density. By defining the discrete first form to represent the niche potential gradient field, the first form is differentiated using the exterior derivative operator to obtain the second form, which describes the ecological flow vorticity and flux characteristics. In particular, the persistent homology barcode information in the topological invariant feature set is used as a constraint condition: the core habitat blocks corresponding to the persistent interval of the number of connected components are used as the source points of the ecological flow, and the ecological barrier areas identified by the persistent interval of the ring structure are assigned a high resistance coefficient. Finally, by solving the discrete Poisson equation, a complete discrete exterior differential model containing the potential field, flow field and vorticity field is generated. This step breaks through the limitations of traditional minimum cost path analysis and for the first time realizes the three-dimensional dynamic characterization of ecological flow movement, which can simulate the path selection, flow intensity and direction change of species migration, and provides a physically meaningful flow field dynamics basis for subsequent ecological corridor design.

[0055] Step S33, according to the discrete exterior differential model and the topological invariant feature set, a multi-scale topological network coupling and graph generation process is performed to obtain a connectivity topological graph of the habitat space that characterizes the structural connectivity and functional connectivity of the habitat.

[0056] It can be understood that this step constructs a connectivity topological graph with ecological mechanical significance by coupling the discrete exterior differential model with the topological invariant feature set at multiple scales. This process first determines the key nodes of the network using the persistent homology barcode information in the topological invariant: core habitat blocks with a persistent interval of the number of connected components exceeding 0.3 are defined as source-sink nodes, and ecological barrier areas identified by the persistent interval of the ring structure are used as network block points. Subsequently, the calculated ecological flow field data is used to construct a weighted adjacency matrix using spectral graph theory methods, where the connection weight between nodes is determined by the flow field flux value, vorticity intensity and topological distance, specifically by performing a tensor product operation on the flow field second form integral result and the persistent homology generation distance. Finally, by calculating the eigenvalues and eigenvectors of the graph Laplacian matrix, a spectral clustering graph reflecting the multi-scale connectivity characteristics is generated: the eigenvalue size represents the connectivity strength, and the eigenvector distribution indicates the dominant direction of ecological flow. This step realizes the mathematical unification of structural connectivity (represented by topological invariants) and functional connectivity (represented by ecological flow field), and the generated topological graph not only identifies key corridors and barriers, but also quantifies the connectivity intensity gradient at different spatial scales, providing a spatial optimization framework for ecological corridor design that combines theoretical rigor and practical guidance.

[0057] Step S4, constructing and coupling the membrane computing ecological network model of the connectivity topology map to obtain an initial wild animal buffer zone delineation scheme;

[0058] It can be understood that this step first maps the nodes (core habitat blocks) and edges (potential connection paths) in the topology map to membrane structures in membrane computing: each core habitat is defined as a basic membrane object, and the species diversity, habitat quality and other attributes contained therein are converted into multiple membranes, and the connection relationship between the membranes is set according to the transmission rules of the weighted adjacency matrix in the topology map. In particular, the reaction-diffusion-convection mechanism is used to construct the rule system in the membrane: the species diffusion ability is used as the diffusion coefficient, the habitat suitability is used as the reaction term generation rate, and the terrain and hydrological oriented ecological flow is used as the convection term, to form a system of partial differential equations to describe the migration process of species in the membrane system. Through parallel computing of the rules of each membrane evolution, when the species probability flow intensity exceeds the ecological threshold (such as the minimum viable population size required migration flow), the membrane dissolution or membrane generation rule is triggered, and the membrane structure is dynamically reconstructed - for example, when the ecological flow intensity between two membrane objects continuously exceeds the threshold, the intermediate barrier membrane is dissolved and a new connection membrane is generated. Through the self-organization evolution of the membrane system, an initial buffer zone scheme composed of a series of ecological corridors is finally formed, and the corridor width is quantitatively determined by the ecological flow intensity between the membranes. In this step, step S4 includes step S41, step S42 and step S43.

[0059] Step S41, performing environment initialization processing based on the membrane system structure according to the connectivity topology map, defining the habitats in the topology map as a plurality of membrane objects according to the suitability distribution pattern, and constructing the membrane structure and the initial rule set based on the connectivity information in the map to obtain a membrane computing ecological network model simulating the transmission and conversion of ecological flow in the landscape;

[0060] It can be understood that this step is based on the connectivity topological map generated in the previous step, and an initialization framework of the ecological network model is constructed by the membrane computing theory. First, the node elements in the topological map are analyzed for ecological function: the core habitat blocks with a β0 persistence interval exceeding 0.3 are defined as main membrane objects, and the ecological properties such as species diversity index and habitat suitability index contained therein are converted into initial elements of the membrane; the ecological barrier areas identified in the topological map are defined as inhibitory membranes, and the resistance coefficient thereof is converted into an inhibitory factor in the membrane. Subsequently, the inter-membrane connection relationship is constructed according to the weighted adjacency matrix of the topological map: for strong connection edges with a flux value greater than 0.7, an active transmission channel is established and a high-rate transmission rule is set; for medium connection edges with a flux value between 0.3 and 0.7, a condition-triggered transmission rule is established; for weak connection edges with a flux value less than 0.3, a to-be-activated channel is set. In particular, the death time data of the persistence barcode are used to set the priority of the membrane rule - the rule corresponding to the long persistence interval has a higher execution priority. The final membrane computing ecological network model contains three layers of structure: the outer membrane corresponds to the landscape boundary condition, the middle layer membrane represents the core habitat block, and the inner layer membrane represents the ecological flow exchange node, and all membrane rules are formalized by using a random Petri net with a time delay parameter. This step converts the abstract topological connection relationship into a computable biological-inspired ecological network model, and realizes the discrete event simulation of the ecological flow transmission process through the parallel computing architecture of the membrane system.

[0061] Step S42, according to the membrane computing ecological network model, an ecological flow simulation process of the fusion reaction-diffusion-convection mechanism is carried out to obtain a dynamic probability density field of species migration and habitat;

[0062] It can be understood that this step is based on the membrane computing ecological network model, and realizes the dynamic simulation of ecological flow by fusing the reaction-diffusion-convection mechanism. Firstly, the various elements (species population quantity, genetic diversity, etc.) in the membrane are defined as reaction terms, wherein the reaction rate constant is dynamically adjusted by the habitat suitability index. The diffusion process adopts an anisotropic diffusion model, and the diffusion coefficient matrix is determined by the species movement ability (such as the maximum daily migration distance) and the landscape permeability, wherein the diagonal elements represent the diffusion strength of the main flow direction, and the non-diagonal elements represent the diffusion deflection effect caused by the terrain. The convection term is driven by the ecological flow field data calculated by the discrete external differential model, and the terrain gradient, hydrological flow direction and other factors are converted into a convection vector field. In particular, the parallel computing characteristics of the membrane system are used to solve the reaction-diffusion-convection partial differential equation synchronously at each time step. Finally, the Monte Carlo method is used to simulate the individual migration behavior, and a spatial and temporal distribution field of the species existence probability is generated, which contains the double characteristics of the migration path probability (convection dominated) and the habitat suitability probability (reaction-diffusion dominated). This step realizes the dynamic coupling of ecological processes and the simulation of species distribution, and provides an ecological mechanical basis with spatio-temporal evolution characteristics for the design of buffer zones.

[0063] wherein the reaction-diffusion-convection partial differential equation is as follows: ; wherein, denotes the rate of change of the probability density u with respect to time t, denotes the behavior of random walk or exploratory migration of the species. It will cause the population to diffuse from high-density areas to low-density areas, and finally make the distribution tend to be smooth. denotes the directional migration of the species under the action of the velocity field v, denotes the growth or decay of the population at the local site.

[0064] Step S43, according to the dynamic probability density field of species migration and habitat and the membrane computing ecological network model, an ecological corridor identification process is performed, and an initial wild animal buffer zone delineation scheme is constructed through all the ecological corridors identified.

[0065] It can be understood that this step first analyzes the spatio-temporal characteristics of the probability density field: the probability contour surface is extracted by using the regularization level set method, and the geometric morphology of the migration path is identified by calculating the curvature characteristics of the contour surface; the streamline characteristics of the probability field are analyzed by using the particle image velocimetry technology, and the flow intensity and direction stability of the migration path are quantified. At the same time, according to the trigger record of the membrane calculation model, the hot path and bottleneck area of ecological flow transmission are extracted. The above characteristics are fused with the connectivity constraint in the topological graph, and the adaptive width algorithm is introduced to determine the range of the corridor space: taking the path centerline as the reference, the corridor width is dynamically adjusted according to the probability field gradient descent rate and terrain complexity, wherein the core area width is determined by the area with probability density greater than 0.7, and the transition area width is determined according to the decay characteristics of the probability density between 0.3 and 0.7. The corridor network is converted into a polygon layer with hierarchical protection strength, and an initial buffer zone scheme is formed.

[0066] Step S5, processing the initial wild animal buffer zone delineation scheme by a hybrid strategy of random programming and preset robust optimization to generate an optimized buffer zone delineation scheme;

[0067] It can be understood that in the case of coexistence of unknown distribution and extreme disturbance, the buffer zone boundary generated in this step has quantifiable reliability and toughness, can maintain multi-scale functional connectivity and ecological flow robust through with small area cost, significantly reduces the risk of topological disconnection and corridor failure, and avoids the conservative expansion of traditional robust scheme, realizes the verifiable protection of real uncertainty. In this step, step S5 includes step S51, step S52 and step S53.

[0068] Step S51, performing uncertainty scenario generation processing based on Monte Carlo simulation according to the initial wild animal buffer zone delineation scheme, generating a set of possible scenarios in a preset time period by sampling the preset climate fluctuation and human activity disturbance random data;

[0069] It can be understood that this step first abstracts the climate fluctuation factors (such as precipitation extreme events, temperature anomalies, and drought frequency) and human activity factors (such as land use expansion, road construction pulse, and tourism activity intensity) into random variables with long-tailed distribution or mixed distribution characteristics; in order to ensure the rationality of sampling, these variables are captured through empirical distribution fitting to capture the correlation structure between variables, thereby avoiding the underestimation of synergistic extreme events under the independent assumption. Subsequently, in the Monte Carlo process, a large number of iterations are performed for a predetermined time period, a set of climate and human disturbance parameters are extracted each time and mapped to the buffer zone space to generate a possible ecological pressure distribution map of the buffer zone under different scenarios. By introducing a hierarchical importance sampling mechanism, the frequency of low-probability but high-risk scenarios is increased, so that the scenario set can not only reflect the range of normal conditions, but also cover potential impacts under extreme conditions. The final scenario set is a time-space coupled uncertainty set, providing a rich simulation basis for subsequent robust optimization. This step effectively captures the dynamic risk boundary of the buffer zone under future multi-dimensional disturbance, enabling the model to face multiple futures rather than a single prediction, significantly improving the adaptability and robustness of buffer zone delineation in complex environments.

[0070] Step S52, based on the initial wild animal buffer zone delineation scheme and the possible scenario set of the predetermined time period, an optimization model is constructed with the objective of minimizing cost and the constraint of the probability of ecological connectivity being lower than the predetermined threshold.

[0071] It can be understood that this step first converts the ecological connectivity constraint into an opportunity constraint programming problem, setting the probability of the ecological connectivity index (the number of connected components based on the topological graph) being lower than the threshold of 0.7 not exceeding 10%. In the objective function design, a weighted method of sub-costs is adopted: the buffer zone construction cost is refined into land acquisition cost (calculated according to the unit cost of different land types), ecological restoration cost (classified according to the difficulty of vegetation restoration), and long-term maintenance cost (discounted at a 20-year cycle), with weights determined by the entropy weight method as 0.4:0.3:0.3. In particular, scenario tree modeling technology is introduced to handle the time-space correlation: 500 random scenarios are organized into a scenario tree with time branching structure, each node containing the joint probability distribution of climate fluctuation and human disturbance. 3000 possible paths are generated through Monte Carlo forward simulation, and the conditional value at risk is used to measure the risk cost under extreme scenarios. The optimization model finally established contains 158 decision variables (including 45 spatial layout variables and 113 management strategy variables) and 279 constraint conditions (including 89 opportunity constraints), and a sample average approximation-based solving strategy is used to convert the original problem into a deterministic equivalent form. This step realizes the balance between economy and ecological reliability through a risk quantification mechanism, providing a strict mathematical programming basis for subsequent robust optimization.

[0072] Step S53, according to the optimization model, a Benders pair decomposition-based solving process is performed to generate an optimized buffer zone demarcation scheme.

[0073] It can be understood that this step of the process first decomposes the original problem into a main problem (integer programming) containing spatial layout decisions and a sub-problem (linear programming) for each random scenario. In the main problem solving stage, the branch and bound method is used to determine the spatial layout scheme of the buffer zone, while receiving Benders cut constraints from all sub-problems; in the sub-problem solving stage, 500 random scenarios are calculated in parallel, and the optimal cut and feasible cut are generated by solving the dual problem, where the feasible cut is specifically used to handle the probability constraint violation when the ecological connectivity is below the 0.7 threshold. In particular, the Pareto optimal cut generation technique is introduced to accelerate convergence: 500 scenarios are reduced to 12 typical scenario patterns using scenario clustering analysis, and reinforced Benders cuts are generated based on the central scenario of each pattern. After 23 iterations, the algorithm converges to the optimal solution, and finally outputs an optimal layout scheme containing 45 spatial units, with the core protection area ratio, adaptive management area ratio, and dynamic adjustment area ratio. This step solves the computational complexity problem of large-scale random programming, shortens the solving time of the original problem from the estimated 48 hours to 3.5 hours through the decomposition strategy, while ensuring the global optimality of the solution, providing an optimized scheme for buffer zone construction that is strictly mathematically verified.

[0074] In this step, step S5 is followed by step S6.

[0075] Step S6, using q-order fuzzy measure and Choquet integral to evaluate the multi-objective synergy of the optimized buffer zone demarcation scheme, and obtaining the intelligent buffer zone demarcation scheme with optimal comprehensive benefits.

[0076] It can be understood that, by introducing the Choquet measure and integral mechanism, this step converts the multi-dimensional benefits of buffer zone design from traditional linear superposition to nonlinear evaluation that can reflect interactions, not only improving the ecological rationality of evaluation, but also achieving adaptive balance of the scheme under multiple objectives, thereby obtaining an intelligent buffer zone demarcation result with optimal comprehensive benefits. In this step, step S6 includes step S61, step S62, and step S63.

[0077] Step S61, according to the optimized buffer zone demarcation scheme and the preset multi-dimensional benefit target, performing Shapley value interaction index and q-order fuzzy measure calculation based on the fuzzy measure calculation model to generate a fuzzy measure calculation model;

[0078] It can be understood that this step extracts 6 core benefit indicators from the scheme: ecological benefit (species protection value improvement degree, corridor connectivity gain), economic benefit (land use cost, long-term maintenance cost) and social benefit (community acceptance, management feasibility), and calculates the marginal contribution of each indicator under all possible combinations. The interaction index is calculated by the Shapley value formula: ;

[0079] Wherein, represents the Shapley value of indicator a, the weight factor, represents the probability of the appearance of the alliance S , which ensures that all possible rankings are equally probable, wherein!represents factorial. N represents the set of all elements, S is a subset of N, which represents the indicator combination without considering indicator a, and a is the benefit indicator to be calculated, represents the benefit value added when indicator a is added to the indicator combination S , represents the scheme benefit evaluation value when the indicator combination is S .

[0080] Then, the Shapley value is converted into a q-order fuzzy measure by using the Hammersly-Clifford theorem, and the optimal q value is determined by solving the following optimization problem: ; wherein, μ is the optimal q value, μ( A ) represents the comprehensive contribution of the total target when the indicator combination A performs well, represents the independent contribution of each indicator in the indicator combination A, A represents the indicator combination, and N represents the set of all elements, represents the value of the parameter q when the error reaches the minimum.

[0081] This step breaks the linear assumption of the traditional weighted summation method, accurately quantifies the complex interaction between benefit indicators (such as the synergistic effect between ecological benefit and social benefit) through fuzzy measure, and provides a mathematical basis for subsequent nonlinear integral evaluation.

[0082] Step S62, each optimized buffer zone is regarded as a "scheme-indicator" matrix, and the Choquet integral is used to integrate the scores of each scheme under the indicator set path defined by the fuzzy measure, to obtain the comprehensive benefit evaluation value of each candidate scheme;

[0083] It can be understood that this step first normalizes the values of the 6 benefit indicators of each scheme to form a decision matrix. Then, according to the q-order fuzzy measure μ obtained in S61, all indicators of each scheme are rearranged in descending order according to their numerical values. Then, the Choquet integral value is calculated: ,in, Scheme X i The comprehensive benefit evaluation value, where μ is the fuzzy measure, X i(k) Scheme X i The standardized value on the Kth indicator, The standardized value at the (k+1)th index, Indicates the fuzzy measure μ in the subset The value that can be taken on.

[0084] The Choquet integral calculation in this step is essentially a weighted summation along the path of the index set defined by the fuzzy measure. When μ is an additivity measure, it degenerates into a linear weighted sum, while when μ is a Choquet measure, it can capture the interaction effects between indicators. For example, when there is a synergistic effect between ecological and social benefits (i.e., μ(ecological, social) > μ(ecological) + μ(social)), the scheme that achieves high values ​​for both indicators simultaneously will gain excess gains. Finally, each scheme obtains a comprehensive evaluation value that takes into account the complex relationships between indicators, providing a quantitative basis for subsequent scheme ranking. This step overcomes the limitations of the traditional linear weighting method, accurately reflecting the nonlinear interaction between ecological, economic, and social benefits, and providing a more scientific decision-making basis for scheme selection.

[0085] Step S63: Calculate the Fisher information distance of the comprehensive benefit evaluation value of all candidate schemes relative to the ideal point, and verify the ranking results by Monte Carlo simulation of perturbation fuzzy measure to obtain the intelligent delineation scheme of wildlife buffer zone with the best comprehensive benefit.

[0086] Understandably, this step employs a combination of information geometry theory and Monte Carlo simulation to perform final selection and robustness verification of candidate solutions. This process first constructs a statistical manifold space of evaluation indicators, mapping the comprehensive benefit evaluation value of each solution to a point on the manifold, and selecting an ideal point (a vector composed of the theoretical maximum values ​​of each indicator) as a reference benchmark. The Fisher information distance between each solution point and the ideal point is then calculated. ;in, This represents the Fisher information distance between each alternative point and the ideal point. For Fisher information matrix elements, Let p and q be the difference in the i-th dimension parameter. Let θ be the difference between schemes p and q in the j-th dimension parameter, and let θ represent the coordinate parameter on the statistical manifold.

[0087] Then the fuzzy measure μ is disturbed by Monte Carlo simulation: 2000 groups of random measure samples conforming to Dirichlet distribution are generated, and the Kendall coordination coefficient of the scheme ranking under each group of measure is calculated. When the coordination coefficient is greater than 0.85, the ranking is considered robust, otherwise return to step S61 to recalibrate the fuzzy measure. Finally, the scheme with the minimum Fisher information distance and passing the robustness test is selected as the optimal delineation scheme. This step combines geometric distance measurement with statistical simulation to ensure that the preferred scheme is not only optimal in a theoretical sense, but also has strong robustness to parameter fluctuations, providing a strictly verified scientific basis for ecological protection decision-making.

[0088] Embodiment 2:

[0089] As shown in Figure 2 The embodiment provides a wild animal buffer zone delineation system for a habitat, which is shown in Figure 2 The system shown in the figure includes an acquisition unit 701, a processing unit 702, an analysis unit 703, a coupling unit 704, and an optimization unit 705.

[0090] The acquisition unit 701 is configured to acquire habitat terrain, vegetation, climate, species distribution, human activity, and protection level data fused with multi-source remote sensing and Internet of Things dynamic monitoring, and construct habitat information data.

[0091] The processing unit 702 is configured to process the habitat information data based on a quantum hybrid optimization algorithm and a preset multi-fractal coupling model, and generate a suitability distribution pattern of the habitat.

[0092] The analysis unit 703 is configured to analyze the suitability distribution pattern by using topological data analysis and discrete exterior differential theory to obtain a connectivity topological atlas of the habitat space.

[0093] The coupling unit 704 is configured to construct and couple a membrane computing ecological network model based on the connectivity topological atlas to obtain an initial wild animal buffer zone delineation scheme.

[0094] The optimization unit 705 is configured to construct and couple a membrane computing ecological network model based on the connectivity topological atlas to generate an optimized buffer zone delineation scheme.

[0095] It should be noted that, as for the system in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment related to the method, and will not be described in detail here.

[0096] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0097] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for delineating wildlife buffer zones in habitats, characterized in that, include: Acquire habitat topography, vegetation, climate, species distribution, human activities, and protection level data by integrating multi-source remote sensing and IoT dynamic monitoring, and construct habitat information data; The habitat information data is processed based on a quantum hybrid optimization algorithm and a preset multifractal coupling model to generate a suitable distribution pattern of habitats. Based on the suitability distribution pattern, topological data analysis and discrete exterior differential theory are used to perform multi-scale landscape connectivity analysis to obtain a connectivity topology map of habitat space. The connectivity topology map is used to construct and couple a membrane computing ecological network model to obtain an initial wildlife buffer zone delineation scheme. An optimized buffer zone delineation scheme is generated by processing the initial wildlife buffer zone delineation scheme through a hybrid strategy of stochastic programming and pre-defined robust optimization. The habitat information data is processed based on a quantum hybrid optimization algorithm and a pre-defined multifractal coupling model, including: The model parameters are quantized and initialized based on habitat information data. Specifically, the quantum annealing algorithm is used to search for the global optimal solution in the solution space consisting of habitat topography, vegetation, climate, species distribution, human activities and protection level to obtain the initial parameter set of the multifractal coupling model. Based on the initial parameter set, multifractal singular spectrum calculations are performed on habitat ecological elements. Among them, multifractal analysis based on box counting and moment method is used to calculate the singular spectral functions and Holder index of vegetation index sequence and topographic elevation sequence, respectively, to obtain multifractal spectrum features. Based on the obtained multifractal spectrum features, the suitability probability calculation is performed based on the fractal dimension constraint. Specifically, by using the multifractal spectrum features as nonlinear weighting factors, a suitability probability distribution function under the fractal dimension constraint is constructed, and the suitability probability distribution of the habitat is calculated. A fitness function is constructed based on the suitability probability of habitats, with the goal of maximizing the spatial correlation between the suitability probability distribution and species distribution. The parameters of the fractal coupling model are iteratively adjusted using the quantum particle swarm optimization algorithm to obtain the suitability distribution pattern of habitats with spatiotemporal heterogeneity. Based on the aforementioned suitability distribution pattern, multi-scale landscape connectivity analysis is performed using topological data analysis and discrete exterior differential theory, including: Based on the suitability distribution pattern of the habitat, topological feature extraction processing based on the continuous homology theory is performed to obtain a set of topological invariant features describing the multi-scale connectivity and ecological barrier structure of habitat patches. Based on the topological invariant feature set and habitat information data, ecological flow field modeling based on discrete exterior differential theory is performed to obtain a discrete exterior differential model describing the ecological flow movement characteristics of the landscape surface. Based on the discrete external differential model and the topological invariant feature set, multi-scale topological network coupling and graph generation processing are performed to obtain a connectivity topological graph of habitat space that characterizes the structural connectivity and functional connectivity of habitat. Specifically, the connectivity topology map is used to construct and couple a membrane computational ecological network model to obtain an initial wildlife buffer zone delineation scheme, including: Based on the connectivity topology map, environmental initialization processing based on membrane system structure is performed. By defining the habitat as multiple membrane objects according to the suitability distribution pattern in the topology map, and constructing the membrane structure and initial rule set based on the connectivity information in the map, a membrane computational ecological network model simulating the transmission and transformation of ecological flow in the landscape is obtained. Based on the membrane computational ecological network model, ecological flow simulation processing of the fusion reaction-diffusion-convection mechanism is performed to obtain the dynamic probability density field of species migration and habitat. Ecological corridors are identified based on the dynamic probability density field of species migration and habitat and the membrane computation ecological network model. An initial wildlife buffer zone delineation scheme is then constructed using all the identified ecological corridors.

2. The method for delineating wildlife buffer zones in habitats according to claim 1, characterized in that... The initial wildlife buffer zone delineation scheme is processed through a hybrid strategy of stochastic programming and pre-defined robust optimization, including: Based on the initial wildlife buffer zone delineation scheme, uncertainty scenario generation processing based on Monte Carlo simulation is carried out. By sampling the preset random data of climate fluctuations and human activity interference, a set of possible scenarios for a preset time period is generated. Based on the initial wildlife buffer zone delineation scheme and the set of possible scenarios for the preset time period, an optimization model is constructed with the goal of minimizing costs and the constraint that the probability of ecological connectivity being lower than a preset threshold is used. The optimized model is solved using Benders dual decomposition to generate an optimized buffer zone delineation scheme.

3. A wildlife buffer zone delineation system for habitats, characterized in that, include: The acquisition unit is used to acquire habitat topography, vegetation, climate, species distribution, human activities and protection level data by integrating multi-source remote sensing and IoT dynamic monitoring, and to construct habitat information data; The processing unit is used to process the habitat information data based on the quantum hybrid optimization algorithm and the preset multifractal coupling model to generate a suitable distribution pattern of the habitat. The analysis unit is used to perform multi-scale landscape connectivity analysis based on the suitability distribution pattern using topological data analysis and discrete exterior differential theory to obtain a connectivity topology map of the habitat space. The coupling unit is used to construct and couple the connectivity topology map into a membrane computing ecological network model to obtain an initial wildlife buffer zone delineation scheme. The optimization unit is used to construct and couple the connectivity topology map into a membrane computing ecological network model, and generate an optimized buffer zone delineation scheme. The processing unit includes: The first processing subunit is used to perform quantum encoding and initialization of model parameters based on habitat information data. Specifically, the quantum annealing algorithm is used to search for the global optimal solution in the solution space consisting of habitat topography, vegetation, climate, species distribution, human activities and protection level to obtain the initial parameter set of the multifractal coupling model. The second processing subunit is used to perform multifractal singular spectrum calculation processing of habitat ecological elements according to the initial parameter set. Among them, based on the box counting method and the moment method, the singular spectrum function and Holder index of the vegetation index sequence and the topographic elevation sequence are calculated respectively to obtain the multifractal spectrum features. The third processing subunit is used to perform suitability probability calculation based on fractal dimension constraints according to the obtained multifractal spectrum features. Specifically, by using the multifractal spectrum features as nonlinear weighting factors, a suitability probability distribution function under fractal dimension constraints is constructed, and the suitability probability distribution of the habitat is calculated. The fourth processing subunit is used to construct a fitness function based on the suitability probability of the habitat, with the goal of maximizing the spatial correlation between the suitability probability distribution and the species distribution. The parameters of the fractal coupling model are iteratively adjusted using the quantum particle swarm optimization algorithm to obtain the suitability distribution pattern of the habitat with spatiotemporal heterogeneity. The parsing unit includes: The first analytical subunit is used to perform topological feature extraction based on the continuous homology theory according to the suitability distribution pattern of the habitat, so as to obtain a set of topological invariant features describing the multi-scale connectivity and ecological barrier structure of the habitat patches. The second analytical subunit is used to perform ecological flow field modeling based on the discrete exterior differential theory according to the topological invariant feature set and habitat information data, so as to obtain a discrete exterior differential model describing the ecological flow movement characteristics of the landscape surface. The third analytical subunit is used to perform multi-scale topological network coupling and graph generation processing based on the discrete external differential model and the topological invariant feature set to obtain a connectivity topological graph of habitat space that characterizes the structural connectivity and functional connectivity of habitat. The coupling unit includes: The first coupling subunit is used to perform environmental initialization processing based on the membrane system structure according to the connectivity topology map. By defining the habitat as multiple membrane objects according to the suitability distribution pattern in the topology map, and constructing the membrane structure and initial rule set based on the connectivity information in the map, a membrane computational ecological network model simulating the transmission and transformation of ecological flow in the landscape is obtained. The second coupling subunit is used to perform ecological flow simulation processing of the fusion reaction-diffusion-convection mechanism based on the membrane computational ecological network model to obtain the dynamic probability density field of species migration and habitat. The third coupling subunit is used to identify ecological corridors based on the dynamic probability density field of species migration and habitat and the membrane calculation ecological network model, and to construct an initial wildlife buffer zone delineation scheme through all the identified ecological corridors.

4. The wildlife buffer zone delineation system for habitats according to claim 3, characterized in that, The optimization unit includes: The first optimization subunit is used to perform uncertainty scenario generation processing based on Monte Carlo simulation according to the initial wildlife buffer zone delineation scheme. By sampling the preset random data of climate fluctuations and human activity interference, a set of possible scenarios for a preset time period is generated. The second optimization subunit is used to construct an optimization model based on the initial wildlife buffer zone delineation scheme and the set of possible scenarios for the preset time period, with the goal of minimizing costs and the probability of ecological connectivity being lower than a preset threshold as a constraint. The third optimization subunit is used to perform Benders dual decomposition-based solution processing on the optimization model to generate an optimized buffer zone delineation scheme.

Images