A method for ecological protection knowledge reasoning and generative design of artificial bird nests

Abstract

This invention discloses a method for ecological protection knowledge reasoning and generative design of artificial bird nests, belonging to the field of product design-assisted ecological protection. The method includes the following steps: collecting multi-source data on migratory bird protection, extracting key knowledge related to migratory bird behavior and habitat, and constructing a multidimensional knowledge dataset for migratory bird protection; locally deploying a large language model and performing customized optimization, performing knowledge reasoning using a knowledge base and inference engine, simulating the decision-making process of bird experts, and testing and applying the customized and optimized large language model; establishing a dynamic force model for the bird nest; calculating node weights, generating a point cloud model and a force surface model, and visualizing them; designing the bird nest base and creating a three-dimensional model of the nest base; performing generative design of the nest body; and manufacturing the nest body, deploying, and monitoring it. This invention can generate customized bird nests that adapt to the habitat needs and activity patterns of migratory birds, thereby improving the intelligence and scientific nature of ecological protection.

Description

Technical Field

[0001] The present invention belongs to the field of product design-assisted ecological protection, and specifically relates to an ecological protection knowledge reasoning and artificial bird nest generation design method. Background Art

[0002] Migratory birds, which migrate periodically with the seasons, are of great ecological value and conservation significance. They typically migrate to higher temperate latitudes to breed in the summer and return to lower tropical latitudes to overwinter. China's migratory bird populations include rare species such as the Oriental White Stork. Protecting their habitats and breeding grounds is a crucial measure for maintaining biodiversity and improving the ecological environment. With the deepening implementation of the green development strategy, the protection of migratory birds and their habitats has become a key issue in ecological development.

[0003] Currently, migratory bird observation and data processing primarily rely on the development of electronic bird databases and virtual bird libraries. By integrating observation records, photographs, and satellite banding data, migratory bird information is systematically stored and managed, providing fundamental support for related research. Regarding data analysis, early studies relied on biologists manually annotating track points in a GIS or manually counting distribution points to infer migration sites and routes. Traditional methods also involve monitoring by protected area observers using image recognition and sound source mapping, supplemented by empirical judgment. However, these methods suffer from inefficiencies and inaccuracies.

[0004] Constructing artificial nests is an effective means of protecting migratory bird nests. By providing foundations, iron nest baskets, and locally sourced nesting materials, artificial nests can effectively mitigate the risks of natural and human damage. However, the current site selection for artificial nests is largely subjective and ambiguous, with many locations concentrated in wetlands or near natural nests. Nest design often references natural nest morphology, and in recent years, numerous studies have examined the correlation between bird reproduction rates and artificial nest design parameters.

[0005] In recent years, machine learning-based intelligent question-answering systems (such as Chat-GPT), with their high knowledge integration and natural language interaction capabilities, have become key tools for information processing and decision support. Generative design (also known as generative design) uses algorithms to derive product forms based on set constraints. Currently, common generative design products are mostly connectors between two components, which can meet the movement requirements of the two components and are gradually being widely adopted across various industries. The cross-disciplinary integration of these technologies provides new ideas and solutions for migratory bird habitat protection and intelligent artificial nesting design.

[0006] Regarding migratory bird observation data processing, while electronic bird databases and virtual bird libraries can systematically organize migratory bird information, they lack specificity and are unable to provide specific conservation strategies and recommendations. Local migratory bird conservationists, lacking comprehensive knowledge, often struggle to formulate scientific conservation strategies based on specific environmental factors. Furthermore, while visual observations and image recognition or sound source mapping can generate real-world data, they lack further analysis and cannot form comprehensive conservation approaches tailored to specific regions. This information processing model primarily aggregates data and lacks the depth to support decision-making.

[0007] Constructing artificial nests is an important means of protecting migratory birds. However, inappropriate location, height, shape, material, and construction methods can lead to nest abandonment by migratory birds, potentially affecting their survival and reproductive efficiency. Research has shown that migratory bird nesting behavior is influenced by a variety of environmental factors, and these factors must be considered comprehensively when constructing artificial nests. Furthermore, while camera towers facilitate observation, they can also interfere with bird activity, further impacting the birds' breeding and habitat.

[0008] The generative design of the bird's nest faces significant limitations. Existing methods use generative design software to generate connectors between two components. However, compared to generating simple connectors, the design of independent components (such as the main body of the bird's nest) requires the precise location, shape, and mechanical parameters of complex load-bearing surfaces. Current algorithms struggle to accurately simulate these dynamic conditions, resulting in design deficiencies in structural stability and applicability, undermining the scientific and reliable nature of generative design for generating bird's nests. Summary of the Invention

[0009] To address the above-mentioned issues, the present invention proposes a method for reasoning about migratory bird protection knowledge and generative design of artificial bird nests. By constructing a professional data set in the field of migratory bird protection and introducing a data processing framework, the observation data, literature, and environmental factors are linked, and the domain adaptability and reasoning capabilities of the large language model are optimized, thereby generating precise migratory bird protection plans for specific scenarios, especially design and construction recommendations for artificial bird nests. In addition, the present invention establishes a dynamic force model for bird nests, combining bird activity behavior, habitat environment, and nest material properties to analyze the force distribution of bird nests under different conditions, and designs a visual calculation tool to simulate the force distribution of bird nests and determine the optimal solution for the shape, weight, and material of the nest. Ultimately, the optimization results are used as input conditions for generative design to generate customized bird nests that adapt to the habitat needs and activity patterns of migratory birds, thereby improving the intelligence and scientific nature of ecological protection.

[0010] The technical solutions of the present invention are as follows:

[0011] A method for ecological protection knowledge reasoning and artificial bird nest generative design includes the following steps:

[0012] Step 1: Collect multi-source data on migratory bird protection, extract key knowledge information related to migratory bird behavior and habitat, and construct a multidimensional knowledge dataset in the field of migratory bird protection;

[0013] Step 2: Deploy the large language model locally and perform customized optimization. Use the knowledge base and inference engine to perform knowledge reasoning, simulate the decision-making process of bird experts, and test and apply the customized and optimized large language model.

[0014] Step 3: Establish a dynamic force model of the bird's nest;

[0015] Step 4: Calculate the node weights, generate the point cloud model and force surface model, and visualize them;

[0016] Step 5: Design the bird's nest base and build a three-dimensional model of the bird's nest base;

[0017] Step 6: Conduct generative design of the main body of the bird's nest;

[0018] Step 7: Manufacture the main body of the bird's nest and carry out deployment and monitoring.

[0019] Furthermore, in step 1, multi-source data in the field of migratory bird protection are collected through observation records, literature, field interviews, professional data released by research institutions, and real-time monitoring results of protected areas; the constructed multidimensional knowledge dataset in the field of migratory bird protection includes the appearance characteristics of migratory birds, migratory bird behavior patterns, habitat characteristics, the time when migratory birds appear in a certain place and behavior records, the potential causes of migratory bird behavior and its ecological significance, bird nest parameters, and factors affecting migratory bird activities.

[0020] Furthermore, the specific process of step 2 is as follows:

[0021] Step 2.1: Locally deploy the open source large language model;

[0022] Step 2.2, convert the multidimensional knowledge dataset in the field of migratory bird protection into the input format of the large language model;

[0023] Step 2.3: Fine-tune the large language model through low-rank matrix adaptation to achieve customized optimization. The specific process is as follows:

[0024] Step 2.3.1. Represent the pre-trained weights of the large language model as the product of two low-rank matrices:

[0025] ΔW=A·B;

[0026] Among them, ΔW is the pre-training weight; represents the first low-rank matrix, d is the row dimension of the original weight matrix, and r is the rank of the low-rank matrix; represents the second low-rank matrix, k is the column dimension of the original weight matrix;

[0027] Step 2.3.2: During fine-tuning, optimize and update A and B. The formula is:

[0028]

[0029] Where A′ is the updated A; B′ is the updated B; η is the learning rate; L is the loss function;

[0030] Step 2.3.3, the pre-trained weight W after fine-tuning is expressed as:

[0031] W=W o +ΔW=W o +A′·B′;

[0032] Among them, W o is the initial pre-training weight;

[0033] Step 2.4: Test the customized and optimized large language model. Comprehensively test the performance of the customized and optimized large language model in combination with actual application scenarios. If the output results of the customized and optimized large language model fail to meet expectations, expand the multidimensional knowledge dataset in the field of migratory bird protection through an iterative optimization mechanism or adjust the training parameters. The training parameters include the rank of the low-rank matrix, the output scaling ratio, and the training batch size.

[0034] Step 2.5: Apply the customized and optimized large language model; input the current area's migratory bird habitat conditions into the customized and optimized large language model, and output customized migratory bird protection recommendations; input the current area's bird nest construction environmental conditions into the customized and optimized large language model, and output customized bird nest construction recommendations and specific parameters.

[0035] Furthermore, the specific process of step 3 is as follows:

[0036] Step 3.1: Establish the bird's nest dynamics model. The formula is as follows:

[0037] l(t)=f l (t);

[0038]

[0039] F gravity =m·g;

[0040] Where l is the bird's position; t is the time; f l (·) represents the trajectory function of the bird's position l; v is the velocity; a is the acceleration; F gravityis gravity; m is the mass of the bird; g is the acceleration due to gravity;

[0041] Step 3.2: Establish the force distribution equation of the bird's nest; the details are as follows:

[0042] The distribution of the force points is related to the material stiffness, and the distribution of the contact force is calculated using the node averaging method:

[0043]

[0044] Among them, F nor is the normal contact force; N is the number of nodes; F total (i) represents the total contact force value associated with the i-th node;

[0045] The distributed forces are then adjusted taking into account the material stiffness matrix:

[0046] F dis =K·F nitial ;

[0047] Among them, F dis is the adjusted distributed force; K is the stiffness matrix; F nirial is the basic force of the node;

[0048] Step 3.3, establish the friction equation:

[0049] F friction =μ·F nor ;

[0050] Among them, F friction is the friction force; μ is the friction coefficient;

[0051] Step 3.4: Establish the dynamic contact force equation, the formula is:

[0052]

[0053] Among them, F dynamic is the dynamic contact force; is the material stiffness; δ is the deformation; c is the damping coefficient; ∥v∥ is the velocity modulus;

[0054] Step 3.5, calculate the total contact force, the formula is:

[0055]

[0056] Among them, F total is the total contact force.

[0057] Furthermore, the specific process of step 4 is as follows:

[0058] Step 4.1, generate the bird's nest geometry model:

[0059]

[0060] Among them, x, y, and z represent the x-axis, y-axis, and z-axis coordinates of the node respectively; is the radius of the sphere; θ is the polar angle; φ is the azimuth angle;

[0061] Nodes with z>0 are selected to construct the point cloud model of the upper part of the bird's nest;

[0062] In order to consider the geometric height of the bird's nest Scale the z-axis coordinate:

[0063]

[0064] Where z′ is the z-axis coordinate after scaling adjustment;

[0065] Step 4.2: Calculate and screen node weights. The specific process is as follows:

[0066] Step 4.2.1. Calculate the force weight using the formula:

[0067]

[0068] Among them, ω force is the force weight; F nitial is the basic force of the node; σ(F nitial ) is F nitial The standard deviation of

[0069] Step 4.2.2, calculate the geometric weight, the formula is:

[0070]

[0071] Among them, ω geometry is the geometric weight; is the distance between the node and the center; α is the adjustment parameter;

[0072] Step 4.2.3. Calculate the material weight using the formula:

[0073]

[0074] Among them, ω material is the material weight;

[0075] Step 4.2.4, calculate the gradient weight, the formula is:

[0076]

[0077] in, is the gradient weight; is the gradient; is the vector from the node to the center;

[0078] Step 4.2.5: Calculate the comprehensive weight ω by multi-factor weighted product form. The formula is:

[0079]

[0080] Among them, a1, a2, a3, and a4 are different adjustment coefficients;

[0081] Step 4.2.6: Pre-set a threshold τ and delete nodes whose comprehensive weight is lower than the threshold τ;

[0082] Step 4.3: Assign a color to each node according to the value of the comprehensive weight to achieve visualization of the point cloud model:

[0083] c=ω·C max ;

[0084] Where c is the color value; C max is the maximum value of the color;

[0085] Display the spatial distribution and force intensity of nodes through 3D scatter plot;

[0086] Step 4.4: Normalize the force value of each node and use the triangulation algorithm to convert the point cloud model into a continuous force surface model:

[0087]

[0088] in, is the area of ​​the triangle; are different side vectors of the triangle;

[0089] Step 4.5: Highlight the stress concentration area to visualize the stress surface model; mark the high stress area, filter out the nodes with stress values ​​greater than the preset ratio, and highlight them in different colors in the 3D model; the specific judgment condition is: when F>γ·max(F), mark the node; where F is the stress value of the node; γ is the screening threshold.

[0090] Furthermore, the specific process of step 5 is: in a three-dimensional design platform, draw a basic platform for placing the bird's nest body and establish a three-dimensional model of the bird's nest base, and design different bird's nest body and bird's nest base connection structures according to different bird's nest construction environments.

[0091] Furthermore, the specific process of step 6 is as follows:

[0092] Step 6.1: Based on the large language model recommendations and the force analysis results, determine the installation location, height, and overall dimensions of the artificial bird's nest;

[0093] Step 6.2: Use generative design tools to automatically design a nest that meets the habitat needs of migratory birds. The specific process is as follows:

[0094] Step 6.2.1. Based on the basic dimensions and shape range of the bird's nest, define the outer contour and the locations of each interface. In the generative design tool, set the core structural support of the bird's nest and generate the shape and dimensions of each structural component based on the force distribution data.

[0095] Step 6.2.2: Based on the load-bearing surface of the main body of the Bird's Nest, optimize the areas with high loads; adjust the thickness and stiffness of the structure based on the node weights and the mechanical properties of the materials;

[0096] Step 6.2.3. Identify the objects that the bird's nest needs to avoid and set them as obstacle avoidance objects. At the same time, determine the objects that the bird's nest must contact and define them as contact objects to ensure that the connections and forces between the various parts do not conflict during the design process.

[0097] Step 6.2.4: Set the optimization goal for the generative design and add the adaptability constraint for the installation environment.

[0098] Step 6.3: Review the bird's nest body generated by the generative design tool in detail to ensure that the design meets the preset strength, size, and material requirements; output detailed bird's nest body design files and 3D models to the manufacturing team for subsequent production and installation.

[0099] Furthermore, the specific process of step 7 is as follows:

[0100] Step 7.1: Based on the generative design results, a customized bird's nest body is produced using high-precision manufacturing processes. The production of the bird's nest body includes material processing, structural molding, and base assembly.

[0101] Step 7.2: After manufacturing is complete, the main body of the bird's nest is combined with the base and deployed on-site. During deployment, the height, tilt angle, and orientation of the base and main body of the bird's nest are adjusted based on the customized and optimized large language model recommendation strategy. The connection method between the main body of the bird's nest and the base is optimized based on the terrain and environmental characteristics of the site.

[0102] Step 7.3: Install sensors or cameras for follow-up monitoring to further verify the design effect of the Bird's Nest and provide data support for future large-scale language model optimization and design upgrades.

[0103] The beneficial technical effects brought about by the present invention are:

[0104] Based on intelligent knowledge reasoning and a large language model, this invention can combine the habitat requirements, activity patterns, and environmental factors of migratory birds to generate precise conservation plans for specific scenarios, particularly customized artificial bird nest designs. Through deep learning and a data processing framework, artificial intelligence knowledge reasoning can be used to provide customized bird nest designs based on the needs of different regions and species, improving the accuracy and efficiency of migratory bird conservation.

[0105] This invention integrates multidimensional data on migratory birds' appearance, behavior, habitat characteristics, roosting times, and behavioral patterns, along with material properties and dynamic stress analysis, providing comprehensive data support for the design of artificial bird nests. By establishing a dynamic stress model of the nest, it accurately simulates the stress distribution under different environments and bird activities, ensuring a scientifically sound design with strong practical application value.

[0106] Utilizing force models and visualization calculation tools, this invention optimizes the nest's shape, weight, material, and other design parameters, ensuring it can withstand the loads of bird activity while minimizing material waste and achieving a balance between lightweight and high strength. Furthermore, this optimized structural design effectively improves the nest's stability and durability, reducing the need for human intervention and subsequent maintenance.

[0107] This invention provides a scientific conservation strategy and solution for migratory bird habitats, offering strong support for ecological improvement, species conservation, and biodiversity conservation. By designing customized bird nests suitable for different migratory bird species, it is expected to further increase the reproduction rate of endangered migratory birds and promote the implementation and sustainable development of ecological protection measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0108] Figure 1 This is the overall flow chart of the ecological protection knowledge reasoning and artificial bird nest generative design method of the present invention.

[0109] Figure 2 This is a working process diagram of fine-tuning a large language model using a low-rank matrix adaptation method in the present invention.

[0110] Figure 3 This is a visualization diagram of the force point cloud of a bird's nest according to the solution of the present invention.

[0111] Figure 4 This is a high-stress point cloud visualization diagram of a bird's nest according to the solution of the present invention.

[0112] Figure 5 This is a visualization diagram of the stress-bearing surface of a bird's nest according to the present invention.

[0113] Figure 6 This is a schematic diagram of the bird's nest base structure of the present invention.

[0114] Figure 7 This is a schematic diagram of the force constraint setting of a generative design product according to the solution of the present invention.

[0115] Figure 8 An interface diagram is provided for generating goals and constraints for a generative design product of the invention solution.

[0116] Figure 9 An interface diagram of generating analysis materials for a generative design product of the solution of the present invention.

[0117] Figure 10 This is a schematic diagram of an artificial nest for migratory birds produced by the solution of the present invention.

[0118] Among them, 1-central axis connecting part; 2-bird's nest chassis; 3-fixing rope adjustment knob; 4-metal fixing rope; 5-bird droppings collection port; 6-fixing part; 7-bird's nest base fixing bracket; 8-fixing bracket; 9-angle steel; 10-bird's nest main body. DETAILED DESCRIPTION

[0119] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0120] This invention combines cutting-edge technologies from multiple fields, including bird ecology, artificial intelligence, big data analysis, mechanics, and generative design, and creatively proposes a fully digitalized method for migratory bird protection. This promotes the innovation and application of ecological protection technologies and provides useful reference and inspiration for research in related fields.

[0121] This paper proposes a dynamic force model for bird nests, combining bird behavior, habitat, and material properties to simulate the force distribution of a nest under different conditions. This model dynamically reflects the force changes during use, providing scientific theoretical support for the design of artificial nests.

[0122] This invention innovatively combines a large language model with multi-dimensional data in the field of migratory bird protection. Through a comprehensive analysis of the habitat needs and ecological environment of migratory birds, it realizes the personalized design of artificial bird nests under different habitat conditions.

[0123] like Figure 1As shown, the overall process of the present invention is divided into a large language model part, an algorithm and visualization part, and a generative design part. In the large language model part, a multidimensional knowledge data set in the field of migratory bird protection is constructed through data collection, literature research, etc., and the open source large language model is fine-tuned using LoRA technology to optimize the domain adaptability and reasoning ability of the large language model, thereby generating customized migratory bird protection recommendations. In the algorithm and visualization part, by constructing a dynamic force model of the bird's nest, the force conditions of the bird's nest structure during transportation, degradation, and rest are analyzed, and a point cloud model is generated in combination with the triangulation algorithm to visualize the force distribution, providing a scientific basis for design optimization. In the generative design part, based on the force analysis results and design constraints, a variety of bird's nest design schemes are generated using design software. The optimal form is determined through screening, and the production and on-site installation of the bird's nest are finally completed.

[0124] The present invention specifically comprises the following steps:

[0125] Step 1: Collect multi-source data on migratory bird conservation through observation records, literature, field interviews, professional data released by research institutions, and real-time monitoring results from protected areas. Extract key knowledge related to migratory bird behavior and habitats, and construct a multidimensional knowledge dataset for migratory bird conservation. This dataset is a structured knowledge question-answering dataset that can be used for large language model learning. The specific content of the multidimensional knowledge dataset for migratory bird conservation includes the following aspects:

[0126] (1) Appearance of migratory birds: record the high-precision description of the body shape, color and typical characteristics of migratory birds; for example: "What are the characteristics of the appearance of the Oriental White Stork?", "The Oriental White Stork is large in size, about 1200mm in length, and its body feathers are white except for the black primary flight feathers. The skin around the eyes and the bare skin before the eyes is red, and the eyes are also red. The beak is black, very thick, and the lower beak is tilted downward. When the beak is closed, the middle part cannot be completely closed, leaving a gap. When flying, the head, neck and legs are all straight, and the legs are obviously extended behind the tail feathers. The black primary and secondary flight feathers are in sharp contrast with the pure white body feathers."

[0127] (2) Migratory bird behavior patterns: migratory routes, habitat behaviors, and dynamic characteristics of migratory birds at different stages of life, especially the typical behaviors in different regions and at different life stages. For example, "What are the behavioral characteristics of the Oriental White Stork when flying?" "The Oriental White Stork flies steadily and slowly, and its seven primary flight feathers can be clearly seen separated from each other. Looking down at its back, the black edges of its wings are perpendicular to its body. The outer edges of the secondary flight feathers are grayish white, and the black parts appear black and white. Looking up at its belly, a black band can be seen separated by its white body."

[0128] (3) Habitat characteristics: covering geographical location, climate conditions, topography and biological resources. For example, "What are the natural geographical characteristics of Heilongjiang Honghe Nature Reserve as one of the important habitats of Oriental White Stork?" "The geographical coordinates of Heilongjiang Honghe Nature Reserve are 133°34′38″~133°46′29″ east longitude, 47°42′18″~47°52′00″ north latitude, with an area of ​​21,836 hm2, an altitude of 55-65 m, an average annual temperature of 1.9℃, and an average annual precipitation of about 500-600 mm. The landform vegetation is mainly herbaceous swamps and aquatic vegetation, with island forests distributed in between. Honghe Nature Reserve retains the original swamp landscape of Sanjiang Plain before it was cultivated. The reserve is relatively closed, with little human activity and low interference intensity. The sparse reed swamps, winding river shallows and large lakes provide a comfortable habitat for Oriental White Stork."

[0129] (4) Recording the time and behavior of migratory birds in a certain place: Based on monitoring data, record the time and behavior characteristics of migratory birds in their habitats. For example, "What are the activity behaviors and time periods of Oriental White Stork chicks observed in Xingkai Lake National Nature Reserve, Heilongjiang Province in 2015?" "Oriental White Stork chicks mainly roost from 22:00 to 4:00 the next day; after 4:00, their activity behaviors, such as standing, preening, and walking, increase significantly. 6:00-8:00 and 16:00-18:00 are the main periods for chicks to stand; 6:00-8:00 and 18:00-20:00 are the main periods for chicks to preen; and 14:00-18:00 is the main period for chicks to walk."

[0130] (5) Analysis of migratory bird behavior: Integrate literature data to summarize the potential causes of migratory bird behavior and its ecological significance. For example, "What is the possible reason for the Oriental White Stork to eat between 4:00 and 18:00 in the Xingkai Lake National Nature Reserve in Heilongjiang Province in 2015?", "14:00 to 18:00 every day is one of the peak times for adult Oriental White Storks to leave the nest and forage, and chicks generally eat their last meal at 20:00. Studies have shown that as the temperature drops at night, the resting metabolic rate of birds increases, and they maintain a constant body temperature by increasing their metabolic level. During the observation period of this study, the night temperature in the Xingkai Lake National Nature Reserve was about 20℃, so the Oriental White Stork needs to eat before dark to meet its energy consumption."

[0131] (6) Bird nest parameters: Sorting out the structural, material and functional parameters of existing natural and artificial bird nests. For example, “What are the parameters of Oriental White Stork nests in the Yellow River Delta National Nature Reserve in 2009?”, “The average nest height in the Dawenliu nesting area is 13.25m±2.07m, and the nest spacing is 647.22m±1086.49m; the average nest height in the Yellow River Estuary nesting area is 25.50m±7.97m, and the nest spacing is 42640.00m±62838.80m.

[0132] (7) Factors affecting migratory bird activities: sort out the interference and impact of various environmental factors on the behavior and habitat of migratory birds. For example, "What are the main factors affecting the site selection and distribution of Oriental White Storks observed in the Yellow River Delta from 2016 to 2022?", "In the process of habitat selection, Oriental White Storks in the Yellow River Delta are restricted by the local altitude range on the one hand, and are also affected by the local isothermality, seasonal changes in precipitation, distance from rivers and lakes, daily range of average temperature, and land use type on the other hand."

[0133] Step 2: Locally deploy a large language model and perform customized optimization. Use the knowledge base and inference engine to perform knowledge reasoning, simulate the decision-making process of bird experts, and test and apply the customized and optimized large language model. The specific process is as follows:

[0134] Step 2.1: Locally deploy a large language model. Locally deploy a publicly available open-source large language model, such as the llama3 language model. Enter simple questions and verify the accuracy of the answers.

[0135] Step 2.2: Dataset format conversion: Convert the multidimensional knowledge dataset in the field of migratory bird protection into the input format of the large language model.

[0136] Step 2.3: Custom optimization of large language models.

[0137] The large language model is efficiently fine-tuned through the Low-Rank Adaptation (LoRA) method to enhance its professionalism and accuracy in identifying migratory birds and generating protection recommendations. The multidimensional knowledge dataset in the field of migratory bird protection is converted into JSON format and loaded into the large language model to ensure that the generalization ability of the original model is retained during training and to achieve the integration of targeted optimization and knowledge supplementation. Ultimately, the large language model can not only expand the migratory bird protection knowledge base through knowledge reasoning, but also adjust decisions based on new knowledge in different ecological environments. Generate an optimized language model that adapts to the needs of the migratory bird ecological field and provides intelligent support for subsequent protection plans.

[0138] The following is an example of the JSON format of a large language model training dataset:

[0139]

[0140] like Figure 2 As shown in Figure 2, the low-rank matrix adaptation method works as follows to fine-tune a large language model:

[0141] Step 2.3.1, weight matrix decomposition; low-rank matrix adaptation method represents the pre-trained weights of the large language model as the product of two low-rank matrices:

[0142] ΔW=A·B;

[0143] Among them, ΔW is the pre-training weight; Denotes the first low-rank matrix, d is the row dimension of the original weight matrix, and r is the rank of the low-rank matrix (r<<d). represents the second low-rank matrix, and k is the column dimension of the original weight matrix.

[0144] Step 2.3.2, optimization process; during fine-tuning, optimize and update A and B without changing the initial pre-training weight W o The specific update formula is:

[0145]

[0146] Among them, A′ is the updated A; B′ is the updated B; η is the learning rate; L is the loss function, and common loss functions such as the cross entropy loss function can be used.

[0147] Step 2.3.3, weight merging;

[0148] The pre-trained weight W after fine-tuning can be expressed as:

[0149] W=W o +ΔW=W o +A′·B′;

[0150] Among them, W o are the initial pre-training weights, which remain fixed.

[0151] Figure 2 In , x represents the input feature of the data in the large language model, and h represents the transformation result of the input feature x after passing the LoRA method. Then:

[0152] h=W0x+ΔWx=W0x+(A′·B′)x;

[0153] LoRA uses the initial pre-trained weights W of the original model o Provides initial generalization capability, achieves customized optimization by fine-tuning the newly added weight part (ΔW = A·B), and optimizes the low-rank matrices A and B while maintaining the pre-trained weight W. o Under the premise of unchanged, rapid adaptation and efficient training of domain knowledge can be achieved.

[0154] Step 2.4: Test the customized and optimized large language model.

[0155] The performance of the customized and optimized large language model is fully tested in combination with actual application scenarios. For example, when asked about suggestions for building artificial bird nests, the model is verified to be able to provide comprehensive and scientific suggestions, including nest location, height, diameter, material properties, structural parameters, and environmental adaptability. If the output results of the customized and optimized large language model fail to meet expectations, the multidimensional knowledge dataset in the field of migratory bird protection is expanded through an iterative optimization mechanism, or training parameters such as the rank of the LoRA low-rank matrix, output scaling ratio, and training batch size are adjusted to further improve the model's understanding and reasoning capabilities for complex problems. The customized and optimized large language model has been quantized and should support multi-terminal deployment and rapid response.

[0156] Step 2.5: Apply the customized and optimized large language model; input the current area's migratory bird habitat conditions into the customized and optimized large language model, and output customized migratory bird protection recommendations; input the current area's bird nest construction environmental conditions into the customized and optimized large language model, and output customized bird nest construction recommendations and specific parameters.

[0157] Step 3: Build a dynamic force model for the bird's nest. Input the specific nest construction parameters obtained in Step 2.5 into the model to calculate the force applied to a specific point in the nest during takeoff, landing, and a few seconds of nesting. Specific nest construction parameters include the bird's weight, the bird's velocity and acceleration in the x, y, and z directions over a 2- to 5-second period, the nest's material stiffness, contact damping, friction coefficient, and the nest's test nodes.

[0158] The construction of the dynamic force model of the bird's nest needs to be based on the physical interaction between the bird and the nest. In combination with observation data and material parameters, the dynamic force model of the bird's nest is used to analyze the force and distribution applied by the bird on the nest under different dynamic states. The specific construction process of the dynamic force model of the bird's nest is as follows:

[0159] Step 3.1. Build a bird's nest dynamics model. Use the trajectory function to describe the bird's position, velocity, and acceleration at time t, and then calculate the bird's motion state. The formula is as follows:

[0160] l(t)=f l (t);

[0161]

[0162] F gravity =m·g;

[0163] Where l is the bird's position; f l (·) represents the trajectory function of the bird's position l; v is the velocity; a is the acceleration; Fgravity is gravity; m is the mass of the bird; g is the acceleration due to gravity, g = 9.81 m / s 2 .

[0164] Step 3.2: Establish the force distribution equation of the bird's nest; the details are as follows:

[0165] The distribution of the force points is related to the material stiffness, and the distribution of the contact force is calculated using the node averaging method:

[0166]

[0167] Among them, F nor is the normal contact force; N is the number of nodes; F total (i) represents the total contact force value associated with the i-th node.

[0168] The distributed forces are then adjusted taking into account the material stiffness matrix:

[0169] F dis =K·F nitial ;

[0170] Among them, F dis is the adjusted distributed force; K is the stiffness matrix, which describes the elastic interaction between different nodes. nitial is the basic force of the node;

[0171] Step 3.3, establish the friction equation:

[0172] F friction =μ·F nor ;

[0173] Among them, F friction is the friction force; μ is the friction coefficient; the friction force is proportional to the normal contact force and the friction coefficient:

[0174] Step 3.4: Establish the dynamic contact force equation. The dynamic contact force is the sum of the normal contact force and the damping force, and is used to represent the reaction force between the object and the contact surface. Its calculation formula is:

[0175]

[0176] Among them, F dynamic is the dynamic contact force; is the material stiffness; δ is the deformation, which indicates the degree to which the object is compressed or stretched. c is the damping coefficient, which indicates the material's damping capacity. v is the velocity, and ∥v∥ is the velocity modulus (i.e., the magnitude of the velocity).

[0177] Step 3.5. Calculate the total contact force. The total contact force is the sum of the dynamic contact force and the friction force, representing the combined force between the object and the contact surface. The calculation formula is:

[0178]

[0179] Among them, F total is the total contact force. At this point, the time variation of the force at each point can be output based on the initially set bird's nest force points.

[0180] Step 4: Calculate the node weights, generate the point cloud model and force surface model, and visualize them.

[0181] Based on the dynamic stress model of the Bird's Nest, a stress visualization algorithm and program were developed. By calculating the node weights, a point cloud model and a stress surface model were generated to provide scientific support for the design of the Bird's Nest. The specific steps are as follows:

[0182] Step 4.1. Generate the bird's nest geometric model; Generate the geometric shape of the bird's nest point cloud based on the hemisphere:

[0183]

[0184] Among them, x, y, and z represent the x-axis, y-axis, and z-axis coordinates of the node respectively; is the radius of the sphere; θ is the polar angle, which represents the angle between the point and the z-axis, and the range is 0≤θ≤π; φ is the azimuth angle, which represents the angle between the projection of the point on the plane formed by the x-axis and the y-axis and the x-axis, and the range is 0≤φ<2π;

[0185] Nodes with z>0 are filtered to construct the point cloud model of the upper part of the bird's nest.

[0186] Furthermore, in order to consider the geometric height of the bird's nest The z-axis coordinate can be scaled:

[0187]

[0188] Where z′ is the z-axis coordinate after scaling adjustment;

[0189] Step 4.2, node weight calculation and screening: In order to optimize the point cloud data and delete unnecessary stress points, a multi-factor node weight calculation method was established. The steps are as follows:

[0190] Step 4.2.1. Calculate the force weight of the node. By normalizing and combining the fluctuation characteristics of the force distribution, the formula is:

[0191]

[0192] Among them, ω force is the force weight; F nitial is the basic force of the node; σ(F nitial ) is F nitial The standard deviation of .

[0193] Step 4.2.2: Calculate geometric weights. Considering the distance from the node to the center of the bird's nest, an exponentially decreasing weight distribution strategy is used. The formula is:

[0194]

[0195] Among them, ω geometry is the geometric weight; is the distance between the node and the center, and the calculation formula is α is an adjustment parameter used to control the influence of geometric distance on weight.

[0196] Step 4.2.3. Calculate the material weights based on the influence of the stiffness matrix:

[0197]

[0198] Among them, ω material is the material weight;

[0199] Step 4.2.4: Gradient weight calculation. The gradient weight of the node introduces directional information and describes the geometric distribution of the gradient by the dot product of the force gradient and the position vector. The formula is:

[0200]

[0201] in, is the gradient weight; is the gradient; is the vector from the node to the center;

[0202] Step 4.2.5: Calculate the comprehensive weight ω by multiplying the weighted products of multiple factors to reflect the nonlinear interaction of the weight factors. The formula is:

[0203]

[0204] Among them, a1, a2, a3, and a4 are different adjustment coefficients used to adjust the influence weight of each factor.

[0205] Step 4.2.6: Pre-set the threshold τ, perform weight selection, and delete nodes whose comprehensive weight is lower than the threshold τ.

[0206] Step 4.3: Assign a color to each node according to the value of the comprehensive weight to achieve visualization of the point cloud model:

[0207] c=ω·C max ;

[0208] Where c is the color value; C max is the maximum value of the color;

[0209] The spatial distribution and force intensity of nodes are displayed through 3D scatter plots.

[0210] Step 4.4: Normalize the force value of each node and use the triangulation algorithm to convert the point cloud model into a continuous force surface model:

[0211]

[0212] in, is the area of ​​the triangle; are different side vectors of the triangle.

[0213] Step 4.5: Highlight stress concentration areas to visualize the stress surface model; mark high stress areas, select nodes with stress values ​​greater than a pre-set ratio, and highlight them in different colors in the 3D model. The specific judgment condition is: when F>γ·max(F), mark the node. Where F is the stress value of the node; γ is the screening threshold;

[0214] Figure 3 This is a visualization of a point cloud model of the Bird's Nest, a possible solution of the present invention. Each point in the image represents a stress point, and the color of the point indicates the strength of the stress. Areas with dense point clouds represent areas of concentrated stress, while areas with sparse point clouds represent areas with less stress.

[0215] Figure 4 This is a possible visualization of a high-stress point cloud of a bird's nest. Each point in the image represents a stress point. The stress values ​​of the bird's nest points in the high-stress area exceed the set threshold τ, while the remaining bird's nest points represent stress points below the threshold τ.

[0216] Figure 5 This is a possible visualization image of the load-bearing surface of the Bird's Nest. According to the distribution of the point cloud highlights, the point cloud data is processed using a triangulation algorithm and converted into a triangular mesh, thereby drawing a continuous load-bearing surface model.

[0217] Step 5: Design and model the bird's nest base. On a 3D design platform, draw the basic platform for the nest and create a 3D model of the nest base. Design different connection structures between the nest and the base for different nest construction environments. For example, if the construction environment involves angle steel frames or columns, use metal plates and plastic parts, along with metal cables and slots, to ensure the nest can be tightly mounted on the base.

[0218] Step 6: Generative design of the main body of the bird's nest.

[0219] Step 6.1: Based on the large language model recommendations and force analysis results, determine the installation location, height, and overall dimensions of the artificial bird's nest.

[0220] Step 6.2: Generative Design Process. Generative design tools are used to automatically design a nest that meets the habitat needs of migratory birds based on the mechanical parameters, dimensional and geometric parameters, design goals, and constraints. This process proceeds through the following steps:

[0221] Step 6.2.1, structure outline definition:

[0222] Based on the basic size and shape range of the bird's nest, define the outer contour of the bird's nest and the positions of each interface.

[0223] In the generative design tool, the main structural support parts of the bird's nest are set, and the shapes and sizes of each structural component are generated based on the force distribution data.

[0224] Step 6.2.2: Delineation and optimization of stress areas:

[0225] Based on the load-bearing surface of the Bird's Nest, the areas with greater loads were optimized. By calculating the areas with concentrated loads, the strength of these areas was strengthened during the design process.

[0226] Combining the node weights and the mechanical properties of the material, the thickness and stiffness of the structure are adjusted to ensure that the bird's nest can stably withstand bird activities under dynamic loads.

[0227] Step 6.2.3, obstacle avoidance and touch object settings:

[0228] Identify the objects that the Bird's Nest needs to avoid (such as base support structures, external equipment, etc.) and set them as obstacle avoidances. At the same time, determine the objects that the Bird's Nest must contact (such as support poles or cameras) and define them as touch objects to ensure that the connections and forces between the various parts do not conflict during the design process.

[0229] Step 6.2.4: Generate objectives and constraints:

[0230] Generative design optimization goals were set to ensure that the main body of the bird's nest was designed while minimizing material consumption while meeting strength, safety, and adaptability requirements.

[0231] At the same time, adaptability constraints of the installation environment are added to consider possible interference factors during the actual installation process, such as installation angle and wind impact.

[0232] Step 6.3, Final Design Output: The Bird's Nest structure generated by the generative design tool is carefully reviewed to ensure that the design meets the preset strength, size, and material requirements. Detailed design files and a 3D model of the Bird's Nest structure are output and provided to the manufacturing team for subsequent production and installation.

[0233] Step 7: Manufacture the main body of the bird's nest and carry out deployment and monitoring.

[0234] Step 7.1: Based on the generative design results, a customized Bird's Nest body is produced using high-precision manufacturing processes. Fabrication of the Bird's Nest body includes material processing, structural forming, and base assembly to ensure it meets the design requirements for strength, force distribution, and environmental compatibility. After fabrication, the Bird's Nest body is combined with the base for on-site deployment.

[0235] Step 7.2: During deployment, based on the customized and optimized large language model recommendation strategy, the height, tilt angle, and orientation of the nest base and main body are precisely adjusted to ensure adaptability and stability within the migratory bird range. Based on the site's topography and environmental characteristics, the connection between the main body and the nest base is optimized to minimize disruption to the surrounding ecosystem and extend the nest's lifespan.

[0236] Step 7.3: Install sensors or cameras for follow-up monitoring to further verify the design of the Bird's Nest and provide data support for future large-scale language model optimization and design upgrades.

[0237] Figure 6 This is a schematic diagram of a possible structure of an Oriental White Stork's nest base according to the present invention. The nest base includes a central axis connector 1, a nest chassis 2, a fixing rope adjustment knob 3, a metal fixing rope 4, a bird droppings collection port 5, a fixing member 6, a base cover plate 7, a fixing bayonet 8 and an angle steel 9; the central axis connector 1 is fixedly connected to the center position of the front of the nest chassis 2, and the center position of the back of the nest chassis 2 is connected to the base cover plate 7 through a fixing member 6. Two fixing rope adjustment knobs 3 are symmetrically arranged on the front of the base cover plate 7, and a fixing bayonet 8 is provided in the middle of the two fixing rope adjustment knobs 3. Two metal fixing ropes 4 are symmetrically provided at the bottom of the angle steel diagram 9, and each metal fixing rope 4 contains three parallel and pullable metal ropes. The connection between the fixing rope adjustment knob 3 and the metal fixing rope 4 is used to realize the connection between the base cover plate 7 and the angle steel 9.

[0238] The central axis connector 1 is used to connect the bird's nest base to the generatively designed bird's nest body. The bird's nest body is fixed to the bird's nest base via the connector 1. The central axis connector 1 serves as the main generated touch object during generative design.

[0239] The bird's nest chassis 2 is disc-shaped, smaller than the maximum diameter of the generatively designed bird's nest. The disc converges toward the center to collect bird droppings and debris. During the generative design process, the bird's nest chassis 2 serves as an avoidance object.

[0240] The base of the bird's nest base is fixed to the bird's nest placement location with a metal rope. Twisting the fixing rope adjustment knob 3 clockwise can tighten the metal fixing rope 4 to tighten the bird's nest base.

[0241] After the bird's nest base is installed, pull out the metal rope ring, pass it through the fixing rope adjusting knob 3, and wrap it around the structure at the fixed position. Then, tighten the metal fixing rope 4 through the fixing rope adjusting knob 3.

[0242] The bird droppings collection port 5 is located on the side of the protrusion of each fan-shaped chassis. If bird protection requires the collection and analysis of bird droppings, the collection port can achieve this purpose;

[0243] If the bird's nest chassis 2 does not need to be designed, the connection between the bird's nest base and the generatively designed bird's nest body can be directly connected to the fixing member 6;

[0244] Considering that the installation environment of the bird's nest may be a single right-angle steel frame or a cross right-angle steel frame, the bird's nest base fixing bayonet 8 is designed to facilitate the fixing of the bird's nest in various scenarios. The bird's nest base fixing bayonet 8 is a groove that facilitates the product to be fixed to the steel frame.

[0245] The angle steel 9 can be designed as an L-shaped structure.

[0246] Figure 7 A schematic diagram of force constraint settings for a generatively designed product using the present invention demonstrates the force constraint setting method used in the generative design process. By connecting triangulated force surfaces with point clouds, key parameters such as the force distribution, force magnitude, force direction, geometric obstacles, gravity, and environmental factors can be determined. This allows designers to ensure that the Bird's Nest can withstand various dynamic loads and external influences in actual use.

[0247] Figure 8 and Figure 9 Generate objectives, constraints, and material analysis for a possible generative design product. Based on customer and production requirements, set minimum mass, maximum stiffness, and safety factors, and select materials and processing methods.

[0248] Figure 10 This is a possible artificial nest for migratory birds, including Figure 6 The bird's nest base and main body 10 are constructed in step 7.2. The main body 10 is provided with a central axis that mates with the central axis connector 1. The central axis and central axis connector 1 are connected by a latch. The bird's nest base 2 is threadedly connected to the fixings 6 and the base cover 7. The bird's nest base and the entire nest are secured to the tower, pole angle 9, or pole end by tightening the metal cable base 4. Screws are used for reinforcement if necessary. The nest is then placed in a nature reserve.

[0249] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the essential scope of the present invention should also fall within the scope of protection of the present invention.

Claims

1. A method for ecological protection knowledge reasoning and artificial bird nest generative design, characterized by: The steps include: Step 1: Collect multi-source data on migratory bird protection, extract key knowledge information related to migratory bird behavior and habitat, and construct a multidimensional knowledge dataset in the field of migratory bird protection; Step 2: Deploy the large language model locally and perform customized optimization. Use the knowledge base and inference engine to perform knowledge reasoning, simulate the decision-making process of bird experts, and test and apply the customized and optimized large language model. Step 3: Establish a dynamic force model of the bird's nest; Step 4: Calculate the node weights, generate the point cloud model and force surface model, and visualize them; Step 5: Design the bird's nest base and build a three-dimensional model of the bird's nest base; Step 6: Conduct generative design of the main body of the Bird's Nest. The specific process is as follows: Step 6.1: Based on the large language model recommendations and the force analysis results, determine the installation location, height, and overall dimensions of the artificial bird's nest; Step 6.2: Automated design using generative design tools generates a nest body that meets the habitat needs of migratory birds. Step 6.3: Review the Bird's Nest structure generated by the generative design tool in detail to ensure that the design meets the preset strength, size, and material requirements. Output detailed Bird's Nest structure design files and a 3D model to provide to the manufacturing team for subsequent production and installation. Step 7: Manufacture the main body of the bird's nest and carry out deployment and monitoring.

2. The ecological protection knowledge reasoning and artificial bird nest generative design method according to claim 1 is characterized by: In step 1, multi-source data in the field of migratory bird protection are collected through observation records, literature, field interviews, professional data released by research institutions, and real-time monitoring results of protected areas; the constructed multidimensional knowledge dataset in the field of migratory bird protection includes the appearance characteristics of migratory birds, migratory bird behavior patterns, habitat characteristics, the time when migratory birds appear in a certain place and behavior records, potential causes of migratory bird behavior and its ecological significance, bird nest parameters, and factors affecting migratory bird activities.

3. The ecological protection knowledge reasoning and artificial bird nest generative design method according to claim 1 is characterized by: The specific process of step 2 is: Step 2.1: Locally deploy the open source large language model; Step 2.2, convert the multidimensional knowledge dataset in the field of migratory bird protection into the input format of the large language model; Step 2.3: Fine-tune the large language model through low-rank matrix adaptation to achieve customized optimization. The specific process is as follows: Step 2.3.

1. Represent the pre-trained weights of the large language model as the product of two low-rank matrices: ΔW=A·B; Among them, ΔW is the pre-training weight; represents the first low-rank matrix, d is the row dimension of the original weight matrix, and r is the rank of the low-rank matrix; represents the second low-rank matrix, k is the column dimension of the original weight matrix; Step 2.3.2: During fine-tuning, optimize and update A and B. The formula is: Where A′ is the updated A; B′ is the updated B; η is the learning rate; L is the loss function; Step 2.3.3, the pre-trained weight W after fine-tuning is expressed as: W=W o +ΔW=W o +A′ B′; Among them, W o is the initial pre-training weight; Step 2.4: Test the customized and optimized large language model. Comprehensively test the performance of the customized and optimized large language model in combination with actual application scenarios. If the output results of the customized and optimized large language model fail to meet expectations, expand the multidimensional knowledge dataset in the field of migratory bird protection through an iterative optimization mechanism or adjust the training parameters. The training parameters include the rank of the low-rank matrix, the output scaling ratio, and the training batch size. Step 2.5: Apply the customized and optimized large language model; input the current area's migratory bird habitat conditions into the customized and optimized large language model, and output customized migratory bird protection recommendations; input the current area's bird nest construction environmental conditions into the customized and optimized large language model, and output customized bird nest construction recommendations and specific parameters.

4. The ecological protection knowledge reasoning and artificial bird nest generative design method according to claim 1 is characterized by: The specific process of step 3 is as follows: Step 3.1: Establish the bird's nest dynamics model. The formula is as follows: l(t)=f l (t); F gravity =m·g; Where l is the bird's position; t is the time; f l (·) represents the trajectory function of the bird's position l; v is the velocity; a is the acceleration; F gravity is gravity; m is the mass of the bird; g is the acceleration due to gravity; Step 3.2: Establish the force distribution equation of the bird's nest; the details are as follows: The distribution of the force points is related to the material stiffness, and the distribution of the contact force is calculated using the node averaging method: Among them, F nor is the normal contact force; N is the number of nodes; F total (i) represents the total contact force value associated with the i-th node; The distributed forces are then adjusted taking into account the material stiffness matrix: F dis =K·F nitial ; Among them, F dis is the adjusted distributed force; K is the stiffness matrix; F nitial is the basic force of the node; Step 3.3, establish the friction equation: F friction =μ·F nor ; Among them, F friction is the friction force; μ is the friction coefficient; Step 3.4: Establish the dynamic contact force equation, the formula is: Among them, F dynamic is the dynamic contact force; is the material stiffness; δ is the deformation; c is the damping coefficient; ||v|| is the velocity modulus; Step 3.5, calculate the total contact force, the formula is: Among them, F total is the total contact force.

5. The ecological protection knowledge reasoning and artificial bird nest generative design method according to claim 4 is characterized in that: The specific process of step 4 is as follows: Step 4.1, generate the bird's nest geometry model: Among them, x, y, and z represent the x-axis, y-axis, and z-axis coordinates of the node respectively; is the radius of the sphere; θ is the polar angle; φ is the azimuth angle; Nodes with z>0 are selected to construct the point cloud model of the upper part of the bird's nest; In order to consider the geometric height of the bird's nest Scale the z-axis coordinate: Where z′ is the z-axis coordinate after scaling adjustment; Step 4.2: Calculate and screen node weights. The specific process is as follows: Step 4.2.

1. Calculate the force weight using the formula: Among them, ω force is the force weight; F nitial is the basic force of the node; σ(F nitial ) is F nitial The standard deviation of Step 4.2.2, calculate the geometric weight, the formula is: Among them, ω geometry is the geometric weight; is the distance between the node and the center; α is the adjustment parameter; Step 4.2.

3. Calculate the material weight using the formula: Among them, ω material is the material weight; Step 4.2.4, calculate the gradient weight, the formula is: in, is the gradient weight; is the gradient; is the vector from the node to the center; Step 4.2.5: Calculate the comprehensive weight ω by multi-factor weighted product form. The formula is: Among them, a1, a2, a3, and a4 are different adjustment coefficients; Step 4.2.6: Pre-set a threshold τ and delete nodes whose comprehensive weight is lower than the threshold τ; Step 4.3: Assign a color to each node according to the value of the comprehensive weight to achieve visualization of the point cloud model: c=ω·C max ; Where c is the color value; C max is the maximum value of the color; Display the spatial distribution and force intensity of nodes through 3D scatter plot; Step 4.4: Normalize the force value of each node and use the triangulation algorithm to convert the point cloud model into a continuous force surface model: in, is the area of ​​the triangle; are different side vectors of the triangle; Step 4.5: Highlight the stress concentration area to visualize the stress surface model; mark the high stress area, filter out the nodes with stress values ​​greater than the preset ratio, and highlight them in different colors in the 3D model; the specific judgment condition is: when F>γ·max(F), mark the node; where F is the stress value of the node; γ is the screening threshold.

6. The ecological protection knowledge reasoning and artificial bird nest generative design method according to claim 1 is characterized by: The specific process of step 5 is: in a three-dimensional design platform, draw a basic platform for placing the bird's nest body and establish a three-dimensional model of the bird's nest base, and design different connection structures of the bird's nest body and the bird's nest base according to different construction environments of the bird's nest.

7. The ecological protection knowledge reasoning and artificial bird nest generative design method according to claim 1 is characterized by: The specific process of step 6.2 is as follows: Step 6.2.

1. Based on the basic dimensions and shape range of the bird's nest, define the outer contour and the locations of each interface. In the generative design tool, set the core structural support of the bird's nest and generate the shape and dimensions of each structural component based on the force distribution data. Step 6.2.2: Based on the load-bearing surface of the main body of the Bird's Nest, optimize the areas with high loads; adjust the thickness and stiffness of the structure based on the node weights and the mechanical properties of the materials; Step 6.2.

3. Identify the objects that the bird's nest needs to avoid and set them as obstacle avoidance objects. At the same time, determine the objects that the bird's nest must contact and define them as contact objects to ensure that the connections and forces between the various parts do not conflict during the design process. Step 6.2.4: Set the optimization goal for generative design and add the adaptability constraints of the installation environment.

8. The ecological protection knowledge reasoning and artificial bird nest generative design method according to claim 1 is characterized by: The specific process of step 7 is as follows: Step 7.1: Based on the generative design results, a customized bird's nest body is produced using high-precision manufacturing processes. The production of the bird's nest body includes material processing, structural molding, and base assembly. Step 7.2: After manufacturing is complete, the main body of the bird's nest is combined with the base and deployed on-site. During deployment, the height, tilt angle, and orientation of the base and main body of the bird's nest are adjusted based on the customized and optimized large language model recommendation strategy. The connection method between the main body of the bird's nest and the base is optimized based on the terrain and environmental characteristics of the site. Step 7.3: Install sensors or cameras for follow-up monitoring to further verify the design effect of the Bird's Nest and provide data support for future large-scale language model optimization and design upgrades.

Images