Wild animal migration path prediction method based on environmental factor data and monitoring site data

Abstract

The invention relates to the technical field of wildlife monitoring data prediction and simulation, in particular to a wildlife migration path prediction method based on environmental factor data and monitoring site data, which comprises the following steps: S1, carrying out systematic digital extraction work of site and path information; s2, obtaining a digital orthograph (DOM) and preprocessing the DOM; s3, environment factor data are sorted out; s4, integrating the environment factor data and the wild animal group appearing and disappearance point and migration path data to form a wild animal group migration database; s5, establishing a migration path simulation model by adopting a mode of constructing a resource selection probability-resource steering path algorithm (RSF-RGPA); s6, analyzing point locations on the basis of the model obtained in the S5, and continuously iterating until an end point to form a path; according to the method, environmental factor data and wild animal appearing and diving point data are systematically integrated, and the accuracy of migration path prediction is greatly improved by utilizing a constructed resource selection probability-resource guide path algorithm (RSF-RGPA).

Description

Technical Field

[0001] The present invention relates to the technical field of wildlife monitoring data prediction and simulation, and in particular to a method for predicting wildlife migration paths based on environmental factor data and monitoring point data. Background Art

[0002] Wildlife migration route prediction has important significance and application value in ecological protection, species protection, ecological research, agriculture and resource management, and climate change research. It plays an important role in promoting sustainable development, protecting natural ecosystems and human well-being.

[0003] Existing animal migration route predictions are often made by researchers such as scientists based on experience combined with some monitoring tools, such as:

[0004] 1. GPS tracking: By giving animals GPS devices, the location of animals can be tracked and recorded in real time. These devices can transmit data via satellite to obtain detailed animal migration routes and behavior data.

[0005] 2. Satellite remote sensing: Use satellite remote sensing technology to obtain surface information, such as vegetation type, ocean temperature, etc. By analyzing satellite images and remote sensing data, the environmental preferences and route choices of animal migration can be revealed.

[0006] 3. Biomarkers: By implanting or binding various biomarker devices such as rings, wireless tags, and chips on animals, the location information of animals can be tracked and recorded in real time. These marking devices can provide valuable migration routes and behavior data. This method is very dependent on the experience level of researchers, which will lead to large differences in the accuracy of the paths predicted by different personnel and cannot be unified. Summary of the invention

[0007] The purpose of the present invention is to provide a method for predicting the migration path of wild animals based on environmental factor data and monitoring point data.

[0008] In order to achieve the above technical objectives and the above technical effects, the present invention is implemented through the following technical solutions:

[0009] A method for predicting wildlife migration paths based on environmental factor data and monitoring point data comprises the following steps:

[0010] S1: Systematically digitize and extract the location and path information of wild animals in the area;

[0011] S2: Obtain the digital orthophoto map DOM with an initial resolution of 2m obtained by the China Resources Satellite Application Center and perform remote sensing image preprocessing.

[0012] S3: Organize environmental factor data from remote sensing images through visual interpretation, including but not limited to topography, population, NDVI, road network, and food source distribution; establish a high-precision basic data set;

[0013] S4: Integrate the environmental factor data obtained in S3 with the data on the locations of wild animals and their migration paths to form a wild animal migration database;

[0014] S5: Based on the database obtained in S4, a migration path simulation model is established by using the resource selection probability-resource-oriented path algorithm;

[0015] S6: Based on the model obtained in S5, the probability of the existence of the animal at each location under the influence of various environmental factors is analyzed according to the actual migration points of wild animals, and the probability of the six points around each point is determined. The score value is obtained by combining the probability and the migration cost, and the point with the high score value is determined as the next point. Iterate continuously until the end point to form a path;

[0016] Furthermore, the step S1 specifically includes the following steps:

[0017] S1.1: Determine the migration behavior characteristics of species by clarifying the target animal species for research and identifying environmental factors;

[0018] S1.2: Select reliable data sources, including but not limited to national monitoring databases, scientific literature, and satellite tracking data;

[0019] S1.3: Convert data to a unified format, including GeoJSON or Shapefile, to ensure compatibility across GIS platforms;

[0020] S1.4: Extract the point information collected and recorded, and remove abnormal points through analysis, strict screening and verification of the points; perform position accuracy correction on the collected data according to known geographical landmarks to eliminate measurement errors;

[0021] S1.5: Based on the confirmed point information, integrate data from different time periods and sources to form a coherent historical data set, and record the data source and timestamp.

[0022] Furthermore, the step S2 specifically includes the following steps:

[0023] S2.1: Obtain the digital orthophoto map DOM with an initial resolution of 2m obtained by the China Resources Satellite Application Center, and preprocess the data from the remote sensing image of Ziyuan-3. First, prepare the data and download the original image data package containing panchromatic (PAN) images and multispectral (MS) images from the Ziyuan-3 satellite. The data of Ziyuan-3 satellite usually includes stereo imaging images, which are suitable for DEM extraction. Then perform radiometric correction on the image to reduce the atmospheric impact and adjust the brightness and contrast of the image to ensure the accuracy of the spectral information of the image. And perform geometric correction on the image, using the known ground control points (GCP) and satellite orbit data to correct the spatial offset of the image so that it is aligned with the geographic coordinate system. In order to improve the spatial resolution of the image, PAN (panchromatic image, high resolution) is fused with MS (multispectral image, low resolution). Use IHS (hue-saturation-brightness) fusion or other methods to generate high-resolution color images. Use DEM to orthorectify the image to eliminate image distortion caused by terrain undulations. After the orthorectification is completed, save the result as a digital orthophoto (DOM), usually in GeoTIFF format. The Ziyuan-3 satellite has stereo imaging capabilities, and the forward and backward images obtained can be used for stereo image matching. Use photogrammetry software (such as ERDASIMAGINE or PCI Geomatica) to perform stereo matching. Image stereo matching generates a disparity map, which is a displacement difference map between two images. Use the disparity map to calculate elevation data and generate a preliminary DEM. The software will calculate the elevation value of each pixel based on the disparity between the stereo images using the principle of photogrammetry. And obtain the digital orthophoto map DOM of the area with a final accuracy of 2m and a 5-meter digital elevation model (DEM), which shows the terrain distribution in the study area. Through further calculations on the DEM data, the DEM is smoothed to remove noise and outliers to improve accuracy. Use known ground elevation data (such as GPS measurement data) to verify and correct the DEM for accuracy. Manually edit the DEM to repair areas with obvious errors to ensure the accuracy of the DEM data. Output the DEM. Save the final corrected DEM data in a commonly used format (such as GeoTIFF) for use in GIS software to obtain terrain relief data;

[0024] S2.2: Verify the image acquisition time, sensor type, and resolution information to ensure the freshness and applicability of the data.

[0025] S2.3: Use radiation correction models to reduce atmospheric effects, taking into account the effects of aerosols, clouds, and water vapor during the correction process.

[0026] S2.4: Stitch and crop multiple adjacent images to ensure that the resulting image covers the entire study area.

[0027] Furthermore, the step S3 specifically includes the following steps:

[0028] S3.1: Collect environmental factor data related to wildlife migration path analysis through various channels, including but not limited to terrain, vegetation cover, human activities, water sources, and road network factors. Identify important environmental factors, including vegetation cover type, human activities, water sources, and road networks. In order to ensure high accuracy and spatial consistency of data, use remote sensing image processing, digital elevation model (DEM) calculation, and other spatial analysis methods to extract and process multiple key environmental factors, and use expert knowledge and references to determine the impact weight of each factor.

[0029] S3.2: Environmental factor data reflects the comprehensive ecological environment that affects the migration of wild animals. Based on the public data of the Open StreetMap platform, combined with visual interpretation, reading and field observation, the distance to the water source is obtained, and further correction is performed to ensure the accuracy and spatial consistency of the water source data;

[0030] S3.3: Obtain wildlife food source data from the Plant Science Data Center of the Chinese Academy of Sciences, including but not limited to corn, bamboo forests, orchards, and sugarcane. Through field exploration and DOM reading and identification, and through visual interpretation and correction, obtain the food source data with a final resolution of 2m;

[0031] S3.4: Obtain the 1km plantation data of the area from the China 1km grid land use degree dataset, including the land type data of tea plantations and coffee plantations. Through field exploration and DOM reading and identification, and through visual interpretation and correction, obtain the plantation data with a final resolution of 2m.

[0032] S3.5: Obtain grassland and shrub data from the national 1m land use data, and obtain grassland and shrub data with a final resolution of 2m;

[0033] S3.6: Obtain natural mixed forest data from the national 1m land use data, obtain the natural mixed forest data with a final resolution of 2m through field exploration and DOM reading and identification, and correct it through visual interpretation;

[0034] S3.7: Obtain rubber forest data from the China 1km grid land use degree dataset, obtain the rubber forest data with a final resolution of 2m through field exploration and DOM reading and identification, and correct it through visual interpretation;

[0035] S3.8: Use classification algorithms to classify images, verify the accuracy of classification results, and perform manual corrections;

[0036] S3.9: Convert the interpretation results into vector format to facilitate integration with other geographic data, and check the resolution and accuracy of the data to ensure that it meets the research needs;

[0037] S3.10: Use field survey data to verify the interpretation results and make necessary corrections, establish metadata descriptions of factor data, and record the data acquisition, processing, and verification process.

[0038] Furthermore, the step S4 specifically includes the following steps:

[0039] S4.1: Design the database structure, including tables, fields, and indexes, determine the spatial data format, and the storage method of non-spatial data.

[0040] S4.2: Import environmental factor data and animal movement data into the database and verify the integrity and consistency of the data through SQL query and spatial analysis tools;

[0041] S4.3: Associate different data sets through spatial joins, determine the topological relationship of spatial data, and ensure the spatial consistency of data.

[0042] Furthermore, the step S5 specifically includes the following steps:

[0043] S5.1: Load the data of the points where wild animals have passed and have not passed;

[0044] S5.2: Merge two point data;

[0045] S5.3: Extract environmental factor data from environmental raster files;

[0046] S5.4: Standardize the environmental factor data to obtain standardized continuous variables;

[0047] S5.5: Define the independent environmental variables x and select the dependent variables y;

[0048] S5.6: Fit a logistic regression model to simulate wildlife migration paths, using known migration points as "used" points and rarely existing points as "unused" points, and determine the impact of various environmental factors through logistic regression based on standardized environmental factors;

[0049] A log-linear model was used to quantify the attractiveness and hindrance of each factor:

[0050] w(x)=pred_prob=exp(β 1 x 1i +β 2 x 2i +...+β n x ni)(1)

[0051] Among them, represents the resource selection weight, x ni is the value of the i-th environmental factor at the n-th point, β n is the corresponding regression coefficient.

[0052] w(x): represents the resource selection function; pred_prob: represents the resource selection probability;

[0053] exp: represents the natural exponential function, that is, e x , where e is the base of natural logarithms, approximately equal to 2.71828;

[0054] β 1 , β 2 , ... β n : are model parameters, which are coefficients to be estimated and represent the degree of influence of different environmental variables or covariates on the probability of resource selection;

[0055] x 1i 、x 2i ,……x ni : are covariates, that is, environmental characteristics that influence animals' selection of specific resource units, including vegetation type, terrain slope, and distance to water sources.

[0056] The form of this function is an exponential model that maps the linear combination of covariates to a probability. In resource selection research, it is usually assumed that the selection probability of a resource unit is exponentially related to the value of the covariate. If the coefficient of a covariate is positive, then an increase in the covariate will increase the probability of the resource unit being selected: if the coefficient is negative, the opposite is true;

[0057] S5.7: Output the resource selection probability formula and the model test results. During the fitting process, it is necessary to ensure that the model can fully explain the variation in the data and is not overfitted. Overfitting refers to the situation where the model performs well on the training data but performs poorly on new data. To avoid this situation, techniques such as cross-validation can be used to evaluate the generalization ability of the model.

[0058] After fitting is completed, the goodness of fit of the model needs to be evaluated. This can be achieved by calculating statistical indicators such as Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). These indicators take into account the complexity and goodness of fit of the model and can help select the best model;

[0059] S5.8: Construct a resource-guided path algorithm (RGPA). The RGPA algorithm is a new algorithm built for the simulation of wildlife migration paths. It is formed on the basis of the resource selection function algorithm. The logical idea of ​​the algorithm is: based on the probability of the resource selection function and the current location of the wildlife, analyze the surroundings and gradually move toward the migration destination; calculate the path cost of wildlife migration at each step, and consider the influence of positive attraction factors and negative obstacles, so as to simulate the actual migration path of wildlife;

[0060] S5.9: When constructing RGPA, the influence of spatial scale also needs to be considered. Different spatial scales (such as 1m*1m and 30m*30m, etc.) will have a significant impact on the simulation effect. In order to determine the optimal spatial scale, try to use grid points of different sizes to divide the study area, and use RGPA to simulate the route respectively, and then compare the performance of these models in terms of prediction accuracy and explanatory power. Select the best performing grid size for calculation. The selected grid size is not restricted. Select a grid size suitable for the applied wild animal (if the selected wild animal is small in size or has a small range of action, select a smaller grid size). Here we choose 30m*30m as the optimal scale of the final pathfinding grid.

[0061] S5.10: Use the RSF model to calculate the probability value of each grid point being selected on a 30m*30m grid. These probability values ​​will serve as the basis for subsequent path generation.

[0062] S5.11: Determine the starting point (P start ) and end point (P end ) in the grid system, identified by ID, and path planning is performed for these two points;

[0063] S5.12: Construct a hexagonal honeycomb structure of neighboring points, that is, search and obtain the six neighboring points around the current point. Determine whether the neighboring points are empty, if yes, end the algorithm, if no, continue the algorithm; take the starting point (P start ) is the current point (P current ), consider its six adjacent grid points as neighboring points (P neighbors )

[0064]

[0065]

[0066] P current : Current point; P neighbors : Neighboring points

[0067] d 1: The distance between the current point and the adjacent point;

[0068] S5.13: Score the 6 neighboring points: resource selection probability × migration cost. Score the points based on the utilization probability pred_prob of each neighboring point calculated by the RSF model and the direction information of these points relative to the endpoint (i.e., which direction is closer to the endpoint).

[0069]

[0070] score neighbor =pred_prob neighbor ×d(6)

[0071] Prioritize the adjacent points with high scores (higher probability and closer to the end point) as the next candidate points.

[0072] Where: d 2 : The distance from the end point, i.e. the migration cost; d: Migration cost coefficient

[0073] pred_prob neighbor : The probability of resource selection for each neighboring point

[0074] score neighbor : The score value of each neighboring point;

[0075] S5.14: Iteration and termination conditions. Select the neighboring point P with high score neighbor As the next current point P current , repeat the above search process

[0076] P next =ChooseNext(P current ,P neighbor1 ,P neighbor2 ,...)(7)

[0077] Until the current point is the end point P end Or it is impossible to find the next point P that meets the conditions next The search is terminated when ; if the end point is successfully reached or is close enough to the end point, the algorithm ends and the sequence of path points stored in the path list is output.

[0078] L = LineString(P start ,P next1 ,P next2 ,...P end )(8)

[0079] S5.15: The generated path may contain some unnecessary turns or redundant points. Smoothing processing (such as linear interpolation, Bezier curves, etc.) is used to optimize the path to make it more natural and efficient. At the same time, the path is further analyzed and evaluated according to actual needs, such as calculating the total length of the path, the proportion of different habitat types, etc.

[0080] Furthermore, the step S6 specifically includes the following steps:

[0081] S6.1: Set the starting point and potential end point, initialize the current point as the starting point, set the initial path, and define the starting conditions;

[0082] S6.2: Get the six neighboring points around the current point of the wild animal group and determine whether the six neighboring points are empty. If the neighboring points are empty, the algorithm ends; if the neighboring points are not empty, analyze the surroundings and gradually move toward the migration destination. Use Monte Carlo simulation to analyze the environmental adaptability probability of each point, implement neighborhood analysis on the GIS platform, and calculate the migration probability of the six surrounding points.

[0083] S6.3: In the early stage, algorithms such as A* or Dijkstra or circuit theory were used to find the best path, but the path-finding effect was not good. For example, when the circuit theory algorithm was applied to long-distance biological migration, its path-finding accuracy could not be achieved. The A* algorithm is mainly suitable for path planning in static road networks. It needs to be replanned after environmental changes, and because it cannot effectively use the information calculated last time, the calculation efficiency is low. In addition, it has low applicability in high-dimensional space and is not suitable for high-dimensional space because of the large amount of calculation. In some complex maps, the path searched by the A* algorithm may produce some redundant turning points, causing the mobile robot to turn frequently and increase the loss. In addition, the global optimality of the A* algorithm is limited. Due to the constraints of the 8-neighborhood search node strategy, the planned path may not meet the global optimality, which reduces the path-finding efficiency. Therefore, the RSF algorithm was selected, and the RGPA algorithm innovation was made based on the algorithm. The structure of 8 neighboring search nodes was abandoned, and the honeycomb structure of 6 neighboring search nodes was innovated to ensure that the distance from the current point to the surrounding 6 neighboring points is equal. Consider the probability of resource selection and the comprehensive migration cost to score the neighboring points and select the optimal neighboring point. The most important thing is to ensure that the probability of resource selection is large and close to the end point; calculate the migration path cost at each step, and consider the impact of positive attraction factors and negative obstacles;

[0084] S6.4: Add the optimal point to the path, update the current point to the optimal point, determine whether the current point is the end point, if the current point is not the end point, obtain the 6 neighboring points around the current point, and repeat the operations from S6.2 to S6.4; if the current point is the end point, use the path point set to create a path. During the iteration process, record the conditions and results of each path selection, and save the path.

[0085] S6.4: The actual migration paths of wildlife populations are simulated, the generated paths are tested for ecological plausibility, compared with actual observation data, and the path prediction results are revised through expert review and further field investigations.

[0086] Beneficial effects of the present invention:

[0087] By systematically integrating environmental factor data with wildlife occurrence point data, the present invention utilizes the constructed resource selection probability-resource-guided path algorithm (RSF-RGPA) to greatly improve the accuracy of migration path prediction. The RSF model quantifies the effects of environmental factors through logistic regression, and can identify and evaluate the impact of various environmental variables (such as vegetation type, terrain, roads, human activities, etc.) on animal migration behavior. This scientific quantitative method avoids the limitations of traditional reliance on human experience and judgment. Through the maximum likelihood estimation method, the model can be adaptively adjusted and is applicable to different species and environmental conditions, ensuring the scientificity and consistency of the prediction results.

[0088] Using digital orthophotos (DOM) with an initial resolution of 2 meters, and through a series of pre-processing techniques (including geometric correction, radiation correction, image stitching, etc.), we ensure the high accuracy of image data, which is the basis for path prediction. Through spatial database technology (such as PostGIS or SpatiaLite), we can achieve efficient integration and management of environmental factors and animal movement data, ensuring the integrity and consistency of the data. This high-precision data not only improves the reliability of model input, but also provides detailed spatial background support for analyzing the behavior of wild animals in different environments.

[0089] The RSF-RGPA model effectively enhances the robustness of the model by flexibly adjusting to different environments and species characteristics. Through the logistic regression model, the behavioral probability of animals under given environmental conditions can be quantified and analyzed, and key influencing factors and their relative importance can be identified. This adaptive ability makes the model less susceptible to changes in a single factor, so that it can maintain stable performance in different geographical and ecological situations and provide reliable migration route predictions.

[0090] In the process of path generation, combining resource selection probability and neighborhood analysis, comprehensive migration cost, A* or Dijkstra or circuit theory algorithms were used in the early stage to find the best path, but due to poor path finding effect, the above algorithms were abandoned and the RSF algorithm was selected. On the basis of this algorithm, the RGPA algorithm was innovated, the structure of 8 neighboring search nodes was abandoned, and the honeycomb structure of 6 neighboring search nodes was innovated to ensure that the distance from the current point to the surrounding 6 neighboring points is equal. This method analyzes the probability and cost of multiple potential path points and intelligently selects the optimal path point, which not only ensures the shortest or optimal distance of the path, but also considers ecological rationality. At the same time, through comparison and verification with actual observation data and expert review, the generated path conforms to biological logic and can effectively guide ecological research and species protection. In this way, path prediction is not only reasonable in mathematical model, but also has practical application value in ecology and conservation practice, providing a scientific basis for conservation decision-making.

[0091] Xishuangbanna and Pu'er are the main activity areas for wildlife migration. The existing ecological environment research and wildlife monitoring system are relatively mature, and the data sources of wildlife migration paths and related environmental factors are rich and reliable. The present invention will use the existing ecological monitoring data and completed field surveys, including food sources, water sources and other data and wildlife trace surveys. In addition, by processing the remote sensing images of the Resources-3 satellite, high-precision DOM data with a resolution of 2 meters is obtained, which can provide accurate terrain information and environmental information, and improve the road network survey in Yunnan Province. These provide a solid foundation for building a high-precision migration database.

[0092] By building a high-precision simulation model of wildlife migration paths and a three-dimensional model visualization, the research results will provide a scientific basis for reducing conflicts between humans and wildlife and optimizing ecological protection strategies. The dynamic prediction function and visualization display of the model can not only help managers better understand the migration behavior of wildlife groups, but also provide strong support for the formulation of regional ecological protection policies. In addition, the introduction of digital twins will promote technological innovation in the field of ecological protection and has broad application prospects and academic value.

[0093] This technology still has many shortcomings. The current accuracy of path simulation can be guaranteed within the range of 100 meters, but further improvement to the 10-meter level still faces challenges. This difficulty may be due to the small scope of the designated experimental area, resulting in strong spatial correlation. In addition, the environment in which biological pathways are simulated is usually complex and changeable, which also increases the difficulty of simulation.

[0094] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0096] Figure 1 It is a schematic diagram of the overall flow chart of the algorithm of the present invention;

[0097] Figure 2 This is a schematic diagram of the environmental factor data (part) of the study area of ​​the present invention;

[0098] Figure 3 This is a schematic diagram of the simulation path diagram of the experimental area of ​​the present invention. DETAILED DESCRIPTION

[0099] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0100] Example 1

[0101] The method for predicting the migration path of wild animals based on environmental factor data and monitoring point data described in this embodiment, in which the wild animals in this embodiment are Asian elephants, includes the following steps:

[0102] S1: In order to further explore the activity patterns of Asian elephants, the collected information on the locations and migration routes of Asian elephants in the area was used to systematically digitize and extract the location and route information;

[0103] S2: Obtain the digital orthophoto map DOM with an initial resolution of 2m obtained by the China Resources Satellite Application Center and perform remote sensing image preprocessing.

[0104] S3: Through visual interpretation of remote sensing images, environmental factor data including topography, population, NDVI, road network, food source distribution, etc. are sorted out to establish a high-precision basic data set;

[0105] S4: Integrate the environmental factor data obtained in S3 with the Asian elephant's location and migration path data to form an Asian elephant migration database to ensure the spatial accuracy and integrity of the data;

[0106] S5: Based on data collection and organization, a migration path simulation model is established by using the Resource Selection Function-Resource-Guided Path Algorithm (RSF-RGPA) method;

[0107] S6: Based on the model obtained in S5, the probability of the existence of the animal at each location under the influence of various environmental factors is analyzed according to the actual migration points of Asian elephants, and the probability of the six points around each point is determined. The score value is obtained by combining the probability and the migration cost, and the point with the high score value is determined as the next point. Iterate continuously until the end point to form a path;

[0108] In this embodiment, step S1 specifically includes the following steps:

[0109] S1.1: By clarifying the target animal species of the study, the Asian elephant, the junction of Pu'er and Xishuangbanna is identified as the study area, specifically the Simao-Jinghong-Jiangcheng junction area, and deeply exploring the activity patterns of the Asian elephant population in this area. Through literature retrieval and expert consultation, it is shown that the migration and habitat selection of Asian elephants at the junction of Pu'er and Xishuangbanna are affected by a variety of environmental factors, including the distribution of food resources, terrain characteristics, and the demand for minerals, which constitute their unique ecological behavior and migration pattern, and determine the migration behavior characteristics of this species;

[0110] S1.2: Select reliable data sources, such as national monitoring databases, scientific literature, satellite tracking data, etc.

[0111] S1.3: Convert data to a unified format, such as GeoJSON or Shapefile, to ensure compatibility across GIS platforms.

[0112] S1.4: Extract the recorded point information from the collection, and extract a total of 245 Asian elephant sites. Then, through analysis, strict screening and verification of the sites, finally retain 233 Asian elephant sites. The location accuracy of the collected data is corrected with known geographical landmarks to eliminate measurement errors.

[0113] S1.5: Based on the confirmed point information, data from different time periods and sources are integrated to form a coherent historical data set, and the data source and timestamp are recorded. This embodiment identifies and maps 26 main migration paths of Asian elephants, which clearly show the activity tracks of Asian elephants in the study area.

[0114] In this embodiment, step S2 specifically includes the following steps:

[0115] S2.1: The acquisition mainly includes the digital orthophoto DOM with an initial resolution of 2m obtained by the China Resources Satellite Application Center, and the data preprocessing of the Resource-3 remote sensing image, and the final digital orthophoto DOM of the area with a precision of 2m and a 5-meter digital elevation model (DEM), which shows the terrain distribution in the study area. Through further calculation of the DEM data, the terrain relief data was obtained, which reflects the terrain change range at different locations in the study area, which helps to understand the impact of terrain changes on the migration path of Asian elephants.

[0116] S2.2: Verify image acquisition time, sensor type, resolution, and other information to ensure data freshness and applicability.

[0117] S2.3: Use radiation correction models (such as FLAASH) to reduce atmospheric effects and consider the effects of aerosols, clouds, and water vapor during the correction process.

[0118] S2.4: Stitch and crop multiple adjacent images to ensure that the resulting image covers the entire study area.

[0119] In this embodiment, step S3 specifically includes the following steps:

[0120] S3.1: Environmental factor data related to the analysis of Asian elephant migration paths are collected through various channels, covering terrain, vegetation cover, human activities, water sources, road networks and other factors. Important environmental factors such as vegetation cover type, human activities, water sources, road networks, etc. are determined. In order to ensure high accuracy and spatial consistency of data, this embodiment uses remote sensing image processing, digital elevation model (DEM) calculation and other spatial analysis methods to extract and process multiple key environmental factors, and uses expert knowledge and references to determine the impact weight of each factor.

[0121] S3.2: Environmental factor data mainly include terrain undulation, slope, slope aspect, vegetation cover (NDVI), distance to water source, distribution of food sources such as orchards and bamboo forests, population density, road network distribution, etc., aiming to reflect the comprehensive ecological environment that affects elephant migration. Water sources provide key living conditions for Asian elephants, especially in the dry season or during migration, they will give priority to areas close to water sources. The existence of road networks may hinder or guide the migration path of the herd, especially the density of roads and traffic flow will affect the range of elephant activities. Based on the public data of the Open Street Map platform (2016), combined with visual interpretation, reading and identification and field observation, the distance to the water source was obtained, and further correction was performed to ensure the accuracy and spatial consistency of the water source data.

[0122] S3.3: Obtain the wildlife food source data published in the Vegetation of China by the Plant Science Data Center of the Chinese Academy of Sciences in 2022, including corn, bamboo forests, orchards, sugarcane, etc. Through field investigation and DOM reading and identification, and through visual interpretation and correction, obtain the food source data with a final resolution of 2m.

[0123] S3.4: Obtain 1km plantation data for the area from the 2020 China 1km grid land use degree dataset, including tea plantations, coffee plantations and other land type data. Through field exploration and DOM reading and identification, and through visual interpretation and correction, the final plantation data with a resolution of 2m is obtained.

[0124] S3.5: Obtain grassland and shrub data from the national 1m land use data (SinoLC-1) to obtain grassland and shrub data with a final resolution of 2m.

[0125] S3.6: Obtain natural mixed forest data from the national 1m land use data (SinoLC-1), obtain the natural mixed forest data with a final resolution of 2m through field exploration and DOM reading and identification, and correct it through visual interpretation.

[0126] S3.7: Obtain rubber forest data from the 2020 China 1km grid land use degree dataset. Through field exploration and DOM reading and identification, and through visual interpretation and correction, the rubber forest data with a final resolution of 2m is obtained.

[0127] S3.8: Use classification algorithms (such as maximum likelihood method and support vector machine) to classify images, verify the accuracy of the classification results, and perform manual corrections.

[0128] S3.9: Convert the interpreted data into vector format for integration with other geographic data. Check the resolution and accuracy of the data to ensure that it meets the research needs.

[0129] S3.10: Use field survey data to verify the interpretation results and make necessary corrections, establish metadata descriptions of factor data, and record the data acquisition, processing, and verification process.

[0130] In this embodiment, step S4 specifically includes the following steps:

[0131] S4.1: Design the database structure, including tables, fields, and indexes, and determine the spatial data format (such as PostGIS or SpatiaLite) and the storage method for non-spatial data.

[0132] S4.2: Import environmental factor data and animal movement data into the database, and verify the integrity and consistency of the data through SQL queries and spatial analysis tools.

[0133] S4.3: Associate different data sets through spatial join, determine the topological relationship of spatial data, and ensure the spatial consistency of data.

[0134] In this embodiment, step S5 specifically includes the following steps:

[0135] S5.1: Load the data of the points that Asian elephants have passed and have not passed;

[0136] S5.2: Merge two point data;

[0137] S5.3: Extract environmental factor data from environmental raster files;

[0138] S5.4: Standardize the environmental factor data to obtain standardized continuous variables;

[0139] S5.5: Define the independent environmental variables x and select the dependent variables y;

[0140] S5.6: Fit the logistic regression model. For the simulation of Asian elephant migration paths, since we are concerned with the probability of Asian elephants choosing habitats under different environmental conditions, GLM is a common choice. In particular, the logistic regression model is suitable for processing binary response variables (i.e., "used" and "unused" resource units), and can estimate model parameters through maximum likelihood estimation, taking known migration points as "used" points and rarely existing points as "unused" points, and judging the influence of various environmental factors through logistic regression based on standardized environmental factors (such as water sources, food sources, roads, human activity distribution, etc.);

[0141] The log-linear model (as shown below) is used to quantify the attractiveness and hindering effects of each factor:

[0142] w(x)=pred_prob=exp(β 1 x 1i +β 2 x 2i +...+β n x ni )(1)

[0143] Among them, represents the resource selection weight, x ni is the value of the i-th environmental factor at the n-th point, β n is the corresponding regression coefficient.

[0144] w(x): represents the resource selection function; pred_prob: represents the resource selection probability;

[0145] exp: represents the natural exponential function, that is, e x , where e is the base of natural logarithms, approximately equal to 2.71828;

[0146] β 1 , β 2 , ... β n : are model parameters, which are coefficients to be estimated and represent the degree of influence of different environmental variables (or covariates) on the probability of resource selection;

[0147] x 1i 、x 2i ,……x ni : It is a covariate, that is, an environmental characteristic that affects animals' selection of specific resource units, such as vegetation type, terrain slope, distance from water source, etc.

[0148] The form of this function is an exponential model that maps the linear combination of covariates to a probability. In resource selection research, it is usually assumed that the selection probability of a resource unit is exponentially related to the value of the covariate. If the coefficient of a covariate is positive, then an increase in the covariate will increase the probability of the resource unit being selected: if the coefficient is negative, the opposite is true;

[0149] S5.7: Output the resource selection probability formula and the model test results. During the fitting process, it is necessary to ensure that the model can fully explain the variation in the data and is not overfitted. Overfitting refers to the situation where the model performs well on the training data but performs poorly on new data. To avoid this situation, techniques such as cross-validation can be used to evaluate the generalization ability of the model.

[0150] After fitting is completed, the goodness of fit of the model needs to be evaluated. This can be achieved by calculating statistical indicators such as Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). These indicators take into account the complexity and goodness of fit of the model and can help select the best model;

[0151] S5.8: Construct a Resource-Guided Path Algorithm (RGPA). The RGPA algorithm is a new algorithm built for simulating the migration path of Asian elephants. It is formed on the basis of the resource selection function algorithm. The logical idea of ​​the algorithm is: based on the probability of the resource selection function and the current location of the Asian elephant, analyze the surroundings and gradually move toward the migration destination; calculate the path cost of the Asian elephant migration at each step, and consider the influence of positive attraction factors and negative obstacles, so as to simulate the actual migration path of wild animals;

[0152] S5.9: When constructing RGPA, the influence of spatial scale also needs to be considered. Different spatial scales (such as 1m*1m and 30m*30m, etc.) will have a significant impact on the simulation effect. In order to determine the optimal spatial scale, try to use grid points of different sizes to divide the study area, and use RGPA to simulate the route respectively, and then compare the performance of these models in terms of prediction accuracy and explanatory power, and select the best performing grid size for calculation. The selected grid size is not restricted, and a grid size suitable for the applied wild animal is selected (if the selected wild animal is small in size or has a small range of movement, a smaller grid size is selected). Here we choose 30m*30m as the optimal scale of the final pathfinding grid;

[0153] S5.10: Use the RSF model to calculate the probability value of each grid point being selected on a 30m*30m grid. These probability values ​​will serve as the basis for subsequent path generation.

[0154] S5.11: Determine the starting point (P start ) and end point (P end ) in the grid system, identified by ID, and path planning is performed for these two points;

[0155] S5.12: Search and obtain the 6 neighboring points around the current point. Determine whether the neighboring points are empty. If yes, the algorithm ends. If no, the algorithm continues. start ) is the current point (P current ), consider its six adjacent grid points as neighboring points (P neighbors )

[0156]

[0157] P current : Current point; P neighbors : Neighboring points

[0158] d 1 : The distance between the current point and the adjacent point;

[0159] S5.13: Score the 6 neighboring points: resource selection probability × migration cost. Score the points based on the utilization probability pred_prob of each neighboring point calculated by the RSF model and the direction information of these points relative to the endpoint (i.e., which direction is closer to the endpoint).

[0160]

[0161] score neighbor =pred_prob neighbor ×d(6)

[0162] Prioritize the adjacent points with high scores (higher probability and closer to the end point) as the next candidate points.

[0163] Where: d 2 : The distance from the end point, i.e. the migration cost; d: Migration cost coefficient

[0164] pred_prob neighbor : The probability of resource selection for each neighboring point

[0165] score neighbor : The score value of each neighboring point;

[0166] S5.14: Iteration and termination conditions. Select the neighboring point P with high score neighbor As the next current point P current , repeat the above search process

[0167] P next =ChooseNext(P current ,P neighbor1 ,P neighbor2 ,…)(7)

[0168] Until the current point is the end point P end Or it is impossible to find the next point P that meets the conditions next The search is terminated when ; if the end point is successfully reached or is close enough to the end point, the algorithm ends and the sequence of path points stored in the path list is output.

[0169] L = LineString(P start ,P next1 ,P next2 ,…P end )(8)

[0170] S5.15: The generated path may contain some unnecessary turns or redundant points. Smoothing processing (such as linear interpolation, Bezier curves, etc.) is used to optimize the path to make it more natural and efficient. At the same time, the path is further analyzed and evaluated according to actual needs, such as calculating the total length of the path, the proportion of different habitat types, etc.

[0171] In this embodiment, step S6 specifically includes the following steps:

[0172] S6.1: Set the starting point and potential end point, initialize the current point as the starting point, set the initial path, and define the starting conditions;

[0173] S6.2: Get the six neighboring points around the current location of the elephant herd, and determine whether the six neighboring points are empty. If the neighboring points are empty, the algorithm ends; if the neighboring points are not empty, analyze the surroundings and gradually move toward the migration destination, use Monte Carlo simulation to analyze the environmental adaptability probability of each point, implement neighborhood analysis on the GIS platform, and calculate the migration probability of the six surrounding points.

[0174] S6.3: The RSF algorithm was selected, and the RGPA algorithm was innovated based on the algorithm. The structure of 8 neighboring search nodes was abandoned, and the honeycomb structure of 6 neighboring search nodes was innovated to ensure that the distance from the current point to the surrounding 6 neighboring points is equal. Consider the probability of resource selection and the comprehensive migration cost to score the neighboring points and select the optimal neighboring point. The most important thing is to ensure that the probability of resource selection is large and close to the end point. The path cost of the elephant migration is calculated at each step, and the influence of positive attraction factors and negative obstacles is considered.

[0175] S6.4: Add the optimal point to the path, update the current position of the elephant group to the optimal point, determine whether the current point is the end point, if the current point is not the end point, obtain the 6 neighboring points around the current point, and repeat the operations from S6.2 to S6.4; if the current point is the end point, use the path point set to create a path. During the iteration process, record the conditions and results of each path selection, and save the path.

[0176] S6.4: The actual migration path of elephants can be simulated, the ecological plausibility of the generated path can be tested, compared with the actual observation data, and the path prediction results can be revised through expert review and further field investigation.

[0177] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for predicting wildlife migration paths based on environmental factor data and monitoring point data, characterized in that: The following steps are involved: S1: Systematically digitize and extract the location and path information of wild animals in the area; S2: Obtain the digital orthophoto map DOM with an initial resolution of 2m obtained by the China Resources Satellite Application Center and perform remote sensing image preprocessing; S3: Organize environmental factor data from remote sensing images through visual interpretation, including but not limited to topography, population, NDVI, road network, and food source distribution; establish a high-precision basic data set; S4: Integrate the environmental factor data obtained in S3 with the data on the locations of wild animals and their migration paths to form a wild animal migration database; S5: Based on the database obtained in S4, a migration path simulation model is established by using the resource selection probability-resource-oriented path algorithm; S6: Based on the model obtained in S5, the probability of the existence of the animal at each location under the influence of various environmental factors is analyzed according to the actual migration points of wild animals, and the probability of the six points around each point is determined. The score value is obtained by combining the probability and the migration cost, and the point with the high score value is determined as the next point. Iterate continuously until the end point to form a path.

2. The method for predicting wild animal migration paths based on environmental factor data and monitoring point data according to claim 1, characterized in that: The step S2 specifically includes the following steps: S2.1: Obtain the digital orthophoto map DOM with an initial resolution of 2m obtained by the China Resources Satellite Application Center, and preprocess the data from the remote sensing image of the Resource-3 satellite. First, prepare the data and download the original image data package containing the panchromatic PAN image and the multispectral MS image from the Resource-3 satellite; the data of the Resource-3 satellite includes stereo imaging images, which are suitable for DEM extraction; then perform radiation correction on the image to reduce the atmospheric impact, adjust the brightness and contrast of the image to ensure the accuracy of the spectral information of the image; and perform geometric correction on the image, using the known ground control point GCP and satellite orbit data to correct the spatial offset of the image so that it is aligned with the geographic coordinate system; Fuse PAN with MS, use IHS fusion or other methods to generate high-resolution color images; use DEM to orthorectify the images to eliminate image distortion caused by terrain undulations; after orthorectification, save the results as digital orthoimages DOM in GeoTIFF format; Ziyuan-3 satellite has stereo imaging capability, and the acquired forward and backward images are matched for stereo images; use photogrammetry software to perform stereo matching of images and generate disparity maps, which are displacement difference maps between two images; use disparity maps to calculate elevation data and generate preliminary DEM; the software calculates the elevation value of each pixel based on the disparity between stereo images using the principle of photogrammetry; The final digital orthophoto map DOM of the area with a final accuracy of 2m and the digital elevation model DEM of 5 meters were obtained, showing the terrain distribution in the study area; through further calculation of DEM data, the DEM was smoothed to remove noise and outliers to improve accuracy; the DEM accuracy was verified and corrected using known ground elevation data; the DEM was manually edited to repair areas with obvious errors to ensure the accuracy of the DEM data; the DEM was output and the final corrected DEM data was saved in GeoTIFF format for use in GIS software to obtain terrain relief data; S2.2: Verify the acquisition time, sensor type, and resolution information of the image to ensure the freshness and applicability of the data; S2.3: Use radiation correction models to reduce atmospheric influences, taking into account the influence of aerosols, clouds and water vapor during the correction process; S2.4: Stitch and crop multiple adjacent images to ensure that the resulting image covers the entire study area.

3. The method for predicting wildlife migration paths based on environmental factor data and monitoring point data according to claim 1, characterized in that: The step S3 specifically comprises the following steps: S3.1: Collect environmental factor data related to wildlife migration path analysis through various channels, including but not limited to terrain, vegetation cover, human activities, water sources, and road network factors; determine important environmental factors, including vegetation cover type, human activities, water sources, and road networks. In order to ensure high accuracy and spatial consistency of data, multiple key environmental factors are extracted and processed using remote sensing image processing, digital elevation model calculation, and other spatial analysis methods. Expert knowledge and references are used to determine the impact weight of each factor; S3.2: Environmental factor data reflects the comprehensive ecological environment that affects wildlife migration. Based on the public data of the Open Street Map platform, combined with visual interpretation, reading and field observation, the distance to the water source is obtained, and further correction is performed to ensure the accuracy and spatial consistency of the water source data; S3.3: Obtain wildlife food source data from the Plant Science Data Center of the Chinese Academy of Sciences, including but not limited to corn, bamboo forests, orchards, and sugarcane. Through field exploration and DOM reading and identification, and through visual interpretation and correction, obtain the food source data with a final resolution of 2m; S3.4: Obtain the 1km plantation data of the area from the China 1km grid land use degree dataset, including the land type data of tea plantations and coffee plantations. Through field exploration and DOM reading and identification, and through visual interpretation and correction, obtain the plantation data with a final resolution of 2m. S3.5: Obtain grassland and shrub data from the national 1m land use data, and obtain grassland and shrub data with a final resolution of 2m; S3.6: Obtain natural mixed forest data from the national 1m land use data, obtain the natural mixed forest data with a final resolution of 2m through field exploration and DOM reading and identification, and correct it through visual interpretation; S3.7: Obtain rubber forest data from the China 1km grid land use degree dataset, obtain the rubber forest data with a final resolution of 2m through field exploration and DOM reading and identification, and correct it through visual interpretation; S3.8: Use classification algorithms to classify images, verify the accuracy of the classification results, and perform manual corrections; S3.9: Convert the interpretation results into vector format to facilitate integration with other geographic data, and check the resolution and accuracy of the data to ensure that it meets the research needs; S3.10: Use field survey data to verify the interpretation results and make necessary corrections, establish metadata descriptions of factor data, and record the data acquisition, processing, and verification process.

4. The method for predicting wild animal migration paths based on environmental factor data and monitoring point data according to claim 1, characterized in that: The step S4 specifically comprises the following steps: S4.1: Design the database structure, including tables, fields, and indexes, determine the spatial data format, and the storage method of non-spatial data; S4.2: Import environmental factor data and animal movement data into the database and verify the integrity and consistency of the data through SQL query and spatial analysis tools; S4.3: Associate different data sets through spatial joins, determine the topological relationship of spatial data, and ensure the spatial consistency of data.

5. The method for predicting wildlife migration paths based on environmental factor data and monitoring point data as claimed in claim 1, characterized in that: The step S5 specifically comprises the following steps: S5.1: Load the data of the points where wild animals have passed and have not passed; S5.2: Merge two point data; S5.3: Extract environmental factor data from environmental raster files; S5.4: Standardize the environmental factor data to obtain standardized continuous variables; S5.5: Define the independent environmental variables x and select the dependent variables y; S5.6: Fit a logistic regression model to simulate wildlife migration paths, taking known migration points as "used" points and rarely existing points as "unused" points, and using logistic regression to determine the impact of various environmental factors based on standardized environmental factors; A log-linear model is used to quantify the attractiveness and hindrance of each factor: w(x)=pred_prob=exp(β1x 1i +β2x 2i +...+β n x ni )(1) Among them, represents the resource selection weight, x ni is the value of the i-th environmental factor at the n-th point, β n is the corresponding regression coefficient; w(x): represents the resource selection function; pred_prob: represents the resource selection probability; exp: represents the natural exponential function, that is, e x , where e is the base of natural logarithms, approximately equal to 2.71828; β1, β2, ... β n : are model parameters, which are coefficients to be estimated and represent the degree of influence of different environmental variables or covariates on the probability of resource selection; x 1i 、x 2i ,……x ni : are covariates, i.e., environmental characteristics that influence animals’ selection of specific resource units, including vegetation type, terrain slope, and distance to water sources; Map the linear combination of covariates to a probability; set the selection probability of a resource unit to be exponentially related to the value of the covariate; if the coefficient of a covariate is positive, then the increase of the covariate will increase the probability of the resource unit being selected; if the coefficient is negative, the opposite is true; S5.7: Output the resource selection probability formula and model test results. During the fitting process, it is necessary to ensure that the model can fully explain the variation in the data and is not overfitted; use cross-validation to evaluate the generalization ability of the model; After the fitting is completed, the goodness of fit of the model needs to be evaluated; this can be achieved by calculating the Akaike information criterion and the Bayesian information criterion statistical indicators; S5.8: Construct a resource-oriented path algorithm. The RGPA algorithm is a new algorithm built for simulating wildlife migration paths. It is formed on the basis of the resource selection function algorithm. Based on the probability of the resource selection function and the current location of the wildlife, the surrounding area is analyzed and gradually moved toward the migration destination. The path cost of wildlife migration is calculated at each step, and the influence of positive attraction factors and negative obstacles is considered, so as to simulate the actual migration path of wildlife. S5.9: Try to use grid points of different sizes to divide the study area and simulate routes with RGPA respectively, then compare the performance of these models in terms of prediction accuracy and explanatory power, and choose the grid size with the best performance. The selected grid size is not restricted. Choose a grid size that is suitable for the wildlife to be applied. Here we choose 30m*30m as the optimal scale of the final pathfinding grid; S5.10: Use the RSF model to calculate the probability value of each grid point being selected on a 30m*30m grid. These probability values ​​will serve as the basis for subsequent path generation. S5.11: Determine the starting point P start and the end point P end The specific location in the grid system is identified by ID, and path planning is performed for these two points; S5.12: Construct a honeycomb structure of neighboring points, that is, search and obtain the six neighboring points around the current point; determine whether the neighboring points are empty, if so, end the algorithm, otherwise, continue the algorithm; take the starting point P start is the current point P current , consider its six adjacent grid points as neighboring points P neighbors ; P current : Current point; P neighbors : Neighboring points d1: the distance between the current point and the adjacent point; S5.13: Score the 6 neighboring points: resource selection probability × migration cost; score the use probability pred_prob of each neighboring point calculated by the RSF model, combined with the direction information of these points relative to the end point; score neighbor =pred_prob neighbor ×d(6) Prioritize the adjacent points with high scores as the next candidate points; Where: d2: distance from the end point, i.e. migration cost; d: migration cost coefficient pred_prob neighbor : The probability of resource selection for each neighboring point score neighbor : The score value of each neighboring point; S5.14: Iteration and termination conditions; select neighboring points P with high scores neighbor As the next current point P current , repeat the above search process P next =ChooseNext(P current ,P neighbor1 ,P neighbor2 ,...)(7) Until the current point is the end point P end Or it is impossible to find the next point P that meets the conditions next The search is terminated when ; if the end point is successfully reached or is close enough to the end point, the algorithm ends and the path point sequence stored in the path list is output; L=LineString(P start ,P next1 ,P next2 ,...P end )(8) S5.15: The generated path may contain some unnecessary turns or redundant points. Smoothing is used to optimize the path to make it more natural and efficient. At the same time, the path is further analyzed and evaluated according to actual needs, including calculating the total length of the path and the proportion of the path passing through different habitat types.

6. The method for predicting wildlife migration paths based on environmental factor data and monitoring point data according to claim 1, characterized in that: The step S6 specifically comprises the following steps: S6.1: Set the starting point and potential end point, initialize the current point as the starting point, set the initial path, and define the starting conditions; S6.2: Get the six neighboring points around the current point and determine whether the six neighboring points are empty; if the neighboring points are empty, the algorithm ends; if the neighboring points are not empty, analyze the surroundings and gradually move toward the migration end point, use Monte Carlo simulation to analyze the environmental adaptability probability of each point, implement neighborhood analysis on the GIS platform, and calculate the migration probability of the six surrounding points; S6.3: In the early stage, the RSF algorithm is used to find the best path, including the honeycomb structure of 6 neighboring search nodes, which ensures that the distances from the current point to the surrounding 6 neighboring points are equal; the neighboring points are scored considering the resource selection probability and the comprehensive migration cost, and the optimal neighboring point is selected to ensure that the resource selection probability is large and close to the end point; the migration path cost is calculated at each step, and the influence of positive attraction factors and negative obstacles is considered; S6.4: Add the optimal point to the path, update the current point to the optimal point, determine whether the current point is the end point, if the current point is not the end point, obtain the 6 neighboring points around the current point, and repeat the operations from S6.2 to S6.4; if the current point is the end point, create a path using the path point set. During the iteration process, record the conditions and results of each path selection, and save the path; S6.4: Actual migration paths are simulated, the generated paths are tested for ecological plausibility, compared with actual observation data, and the path prediction results are revised through expert review and further field investigations.

Images