Ecological protection knowledge reasoning and artificial nest derivative design method

Abstract

The invention discloses an ecological protection knowledge reasoning and artificial nest derivative design method, which belongs to the field of product design assisted ecological protection, and comprises the following steps: collecting multi-source data in the field of migrant bird protection, extracting key knowledge information related to migrant bird behaviors and habitats, and constructing a multi-dimensional knowledge data set in the field of migrant bird protection; the method comprises the following steps: locally deploying a large language model, performing customization optimization, performing knowledge reasoning through a knowledge base and a reasoning engine, simulating a bird expert decision-making process, and testing and applying the customized and optimized large language model; establishing a dynamic stress model of the bird nest; node weights are calculated, and a point cloud model and a stress surface model are generated and visualized; designing a bird nest base, and establishing a three-dimensional model of the bird nest base; carrying out derivative design on the bird nest main body; and manufacturing a bird nest main body, and performing deployment and monitoring. According to the method, the customized bird nest adaptive to the habitat requirements and activity rules of the migrant birds can be generated, so that the intelligent level and scientificity of ecological protection are improved.

Description

Technical Field

[0001] The present invention belongs to the field of product design assistance for ecological protection, and specifically relates to a method for ecological protection knowledge reasoning and generative design of artificial bird nests. Background Art

[0002] Migratory birds are birds that migrate periodically with the change of seasons, and their ecological value and protection significance have attracted much attention. Generally, they migrate to temperate regions with higher latitudes for breeding in summer and return to tropical regions with lower latitudes for wintering in winter. Among the migratory bird populations in China, there are rare species such as the Oriental Stork, and the protection of their habitats and breeding grounds is an important measure to maintain biodiversity and improve the ecological environment. With the in-depth promotion of the green development strategy, the protection of migratory birds and their habitats has become a key issue in ecological construction.

[0003] Currently, the observation and data processing of migratory birds mainly rely on the construction of electronic bird databases and virtual bird libraries. By integrating observation records, photographed images and satellite tracking data, migratory bird migration information is systematically stored and managed to provide basic support for relevant research. In terms of data analysis, early studies mostly relied on biologists to manually mark trajectory points in GIS or count the number of distribution points manually to infer migration areas and migration routes. Traditional methods also include reserve observers using image recognition and sound source atlas analysis, supplemented by empirical judgment for monitoring, but such methods have problems of low efficiency and accuracy.

[0004] In terms of the protection of migratory bird nests, the construction of artificial bird nests is an effective means. By providing basic brackets, iron nest baskets and local nest materials, artificial bird nests can effectively reduce the risks of natural and human damage. However, at present, the site selection of artificial bird nests mostly relies on subjective experience, which has a certain degree of ambiguity and mostly focuses on wetland environments or near natural nests. In terms of nest design, the morphology of natural nests is usually referred to, and in recent years, there have been many literatures studying the correlation between the breeding rate of birds and the design parameters of artificial bird nests.

[0005] In recent years, intelligent question-and-answer systems (such as Chat-GPT) based on machine learning, with high knowledge integration and natural language interaction capabilities, have become key tools for information processing and decision-making support. Generative design (also known as generative design) is a product form derived by algorithms on the basis of setting condition constraints. Currently, common generative design products are mostly connecting parts between two components, which can meet the movement requirements of the two components and are gradually widely used in various industries. The cross-field integration of these technologies provides new ideas and solutions for the intelligent protection of migratory bird habitats and the design of artificial nests.

[0006] In the aspect of processing migratory bird observation data, although electronic bird databases and virtual bird libraries can systematically organize migratory bird information, there are problems of insufficient pertinence and the inability to provide specific protection strategies and suggestions. In the absence of comprehensive knowledge reserves, local migratory bird protection workers often have difficulty formulating scientific protection strategies in combination with specific environmental factors. In addition, although real data can be obtained through visual observation and by means of image recognition or sound source map recognition, there is a lack of further analysis, and a comprehensive protection method adapted to a specific area cannot be formed. This information processing mode is more about data aggregation and lacks the depth of decision-making support.

[0007] In terms of migratory bird protection, the construction of artificial bird nests is an important means of protecting migratory birds. However, if the location, height, shape, material and construction method are not suitable, the bird nests may be abandoned by migratory birds, and even affect their survival and reproduction efficiency. Research shows that the nesting behavior of migratory birds is affected by various environmental factors, and various influencing factors also need to be considered comprehensively when building artificial nests. In addition, although camera towers are convenient for observation, they may also interfere with bird activities, further affecting the reproduction and habitat of migratory birds.

[0008] For the generative design of bird nests, there are significant limitations in the generative design when generating bird nests. The existing method is to use generative design software to generate connectors between two components. Compared with generating such a simple structure as a connector, the design of independent components (such as the main body of the bird nest) requires clarifying the position, shape and mechanical parameters of complex stress surfaces. However, the current algorithms are difficult to accurately simulate these dynamic conditions, resulting in deficiencies in the structural stability and applicability of the design results, and weakening the scientific nature and reliability of the generative design of bird nests. Summary of the Invention

[0009] To solve the above problems, the present invention proposes a method for reasoning 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 materials and environmental factors are associated to optimize the domain adaptability and reasoning ability of the large language model, so as to generate accurate migratory bird protection plans in specific scenarios, especially the design and construction suggestions of artificial bird nests. In addition, the present invention establishes a dynamic stress model of the bird nest, combines the bird activity behavior, habitat environment and material characteristics of the bird nest, analyzes the stress distribution of the bird nest under different conditions, and designs a visual calculation tool to simulate the stress distribution of the bird nest and determine the optimal solutions for the shape, weight and material of the bird nest. Finally, the optimized results are used as the input conditions for the generative design to generate customized bird nests that adapt to the habitat needs and activity rules of migratory birds, thereby improving the intelligent level and scientific nature of ecological protection.

[0010] The technical solution of the present invention is as follows:

[0011] An ecological protection knowledge reasoning and generative design method for artificial bird nests, comprising the following steps:

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

[0013] Step 2: Locally deploy a large language model, customize and optimize it, perform knowledge reasoning through a knowledge base and an inference engine, 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 nest;

[0015] Step 4: Calculate the node weights, generate a point cloud model and a force-bearing surface model, and perform visualization;

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

[0017] Step 6: Perform generative design of the main body of the bird nest;

[0018] Step 7: Manufacture the main body of the bird nest and perform deployment and monitoring.

[0019] Furthermore, in the above Step 1, multi-source data in the field of migratory bird protection are collected by means of observation records, literature, on-site interviews, professional data released by research institutions, and real-time monitoring results of protected areas; the multi-dimensional knowledge dataset in the field of migratory bird protection includes the appearance characteristics of migratory birds, migratory bird behavior patterns, habitat characteristics, the time and behavior records of migratory birds appearing in a certain place, the potential causes and ecological significance of migratory bird behavior, bird nest parameters, and migratory bird activity influencing factors.

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

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

[0022] Step 2.2: Convert the multi-dimensional 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 the low-rank matrix adaptation method to achieve customization and 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] where ΔW is the pre-trained 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, and k is the column dimension of the original weight matrix;

[0027] Step 2.3.2: During the fine-tuning process, optimize and update A and B, and 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] where W o is the initial pre-trained 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 the actual application scenario; if the output result of the customized and optimized large language model fails to meet the expectation, expand the multi-dimensional knowledge dataset in the field of migratory bird protection through the 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; after inputting the current situation of migratory bird habitats in the region into the customized and optimized large language model, output customized migratory bird protection suggestions; after inputting the environmental conditions for nest building in the current region into the customized and optimized large language model, output customized nest building suggestions and specific parameters.

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

[0036] Step 3.1: Establish a nest dynamics model, and the formula is as follows:

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

[0038]

[0039] F gravity = m·g;

[0040] where l is the position of the bird; t is the time; f l (·) represents the trajectory function of the bird position l; v is the velocity; a is the acceleration; F gravityF is 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; specifically as follows:

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

[0043]

[0044] where, F nor is the normal contact force; N is the number of nodes; F total (i) represents the total contact force value related to the i-th node;

[0045] Subsequently, consider the material stiffness matrix to adjust the distributed force:

[0046] F dis = K·F nitial ;

[0047] where, F dis is the adjusted distributed force; K is the stiffness matrix; F nirial is the basic force on the node;

[0048] Step 3.3: Establish the friction force equation:

[0049] F friction = μ·F nor ;

[0050] where, 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] where, F dynamic is the dynamic contact force; is the material stiffness; δ is the deformation; c is the damping coefficient; ∥v∥ is the magnitude of the velocity;

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

[0055]

[0056] where, F total is the total contact force.

[0057] Furthermore, the specific process of the said Step 4 is:

[0058] Step 4.1: Generate the geometric model of the bird's nest:

[0059]

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

[0061] Select nodes with z > 0 to construct the point cloud model of the upper half of the bird's nest;

[0062] To consider the geometric height of the bird's nest Adjust the z-axis coordinate by scaling:

[0063]

[0064] Among them, z′ is the z-axis coordinate after scaling adjustment;

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

[0066] Step 4.2.1: Calculate the force weight, and the formula is:

[0067]

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

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

[0070]

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

[0072] Step 4.2.3: Calculate the material weight, and the formula is:

[0073]

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

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

[0076]

[0077] Among them, is the gradient weight; is the gradient; The vector from the node to the center;

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

[0079]

[0080] where a 1 , a 2 , a 3 , a 4 are different adjustment coefficients;

[0081] Step 4.2.6: Preset the threshold τ and delete the nodes with comprehensive weight lower than the threshold τ;

[0082] Step 4.3: Assign colors to each node according to the value of the comprehensive weight to realize the 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 the nodes through a 3D scatter plot;

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

[0087]

[0088] where is the triangle area; are the different side vectors of the triangle;

[0089] Step 4.5: Highlight the stress concentration area to realize the visualization of the force surface model; Mark the high stress area, screen 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 the 3D design platform, draw the basic platform for placing the main body of the Bird's Nest and establish a 3D model of the Bird's Nest base. For different construction environments of the Bird's Nest, design different connection structures between the main body of the Bird's Nest and the Bird's Nest base.

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

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

[0093] Step 6.2: Conduct automated design based on the generative design tool to generate the main body of the bird nest that meets the habitat needs of migratory birds; the specific process is as follows:

[0094] Step 6.2.1: Combine the basic dimensions and shape range of the bird nest to define the outer contour and the positions of each interface; in the generative design tool, set the core structural support part of the bird nest, and generate the shapes and dimensions of each structural component according to the force distribution data;

[0095] Step 6.2.2: Optimize the areas with large forces according to the force-bearing surface of the main body of the bird nest; combine the node weights and the mechanical properties of the material to adjust the thickness and stiffness of the structure;

[0096] Step 6.2.3: Identify the objects that the bird nest needs to avoid and set them as obstacles; at the same time, determine the objects that the bird nest must contact and define them as touch objects to ensure that there are no conflicts in the connection and force of each part during the design process;

[0097] Step 6.2.4: Set the optimization objectives of the generative design, and at the same time, add the adaptability constraints of the installation environment;

[0098] Step 6.3: Conduct a detailed review of the main body of the bird nest generated by the generative design tool and ensure that the design meets the preset strength, dimension, and material requirements; output the detailed design document and 3D model of the main body of the bird nest, and provide them to the manufacturing team for subsequent production and installation.

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

[0100] Step 7.1: According to the generative design results, manufacture the customized main body of the bird nest through high-precision manufacturing processes; the manufacture of the main body of the bird nest includes material processing, structure forming, and base assembly;

[0101] Step 7.2: After manufacturing, combine the main body of the bird nest with the bird nest base and conduct on-site deployment; during the deployment process, based on the customized and optimized large language model suggestion strategy, adjust the height, tilt angle, and direction of the bird nest base and the main body of the bird nest; combine the topographic and environmental characteristics of the site to optimize the connection method between the main body of the bird nest and the base;

[0102] Step 7.3: Install sensors or camera devices for subsequent monitoring, further verify the design effect of the bird nest, and provide data support for future optimization and design upgrade of the large language model.

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

[0104] Based on intelligent knowledge reasoning and large language models, the present invention can combine the habitat requirements, activity patterns of migratory birds, and environmental factors of the habitat to generate precise protection plans for specific scenarios, especially personalized artificial bird nest designs. Through deep learning and data processing frameworks, the artificial intelligence knowledge reasoning process can provide customized bird nest design plans according to the needs of different regions and species, improving the accuracy and efficiency of migratory bird protection.

[0105] The present invention integrates multi-dimensional data such as the appearance, behavior, habitat characteristics, roosting time, and behavior patterns of migratory birds, as well as 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 bird nest, the stress distribution of the bird nest under different environments and bird activities is accurately simulated, ensuring that the design results are scientific and reasonable and have strong practical application value.

[0106] Using the stress model and visualization calculation tools, the present invention can optimize design parameters such as the shape, weight, and material of the bird nest, ensuring that while the bird nest bears the load of bird activities, material waste is minimized, achieving a balance between lightweight and high strength. In addition, the optimized design of the structure can effectively improve the stability and durability of the bird nest, reducing the need for human intervention and post-maintenance.

[0107] The present invention provides scientific protection strategies and solutions for migratory bird habitats, and can provide strong support for ecological environment improvement, species protection, and biodiversity protection. By designing customized bird nests suitable for different migratory bird species, it is expected to further increase the breeding rate of endangered migratory birds and promote the implementation and sustainable development of ecological protection measures. BRIEF DESCRIPTION OF THE DRAWINGS

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

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

[0110] Figure 3 It is a visualization diagram of the stress point cloud of a bird nest in the solution of the present invention.

[0111] Figure 4 It is a visualization diagram of the high-stress point cloud of a bird nest in the solution of the present invention.

[0112] Figure 5 It is a visualization diagram of the stress surface of a bird nest in the solution of the present invention.

[0113] Figure 6 It is a schematic diagram of the base structure of a bird nest in the solution of the present invention.

[0114] Figure 7 Schematic diagram of force constraint setting for a derivative design product of the solution of the present invention.

[0115] Figure 8 Interface diagram for generating goals and constraints of a derivative design product of the solution of the present invention.

[0116] Figure 9 Interface diagram for generating analysis materials of a derivative design product of the solution of the present invention.

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

[0118] Among them, 1 - central axis connecting piece; 2 - bird's nest chassis; 3 - fixing cable adjustment knob; 4 - metal fixing cable; 5 - bird droppings collection port; 6 - fixing piece; 7 - fixing bayonet for bird's nest base; 8 - fixing bayonet; 9 - angle steel; 10 - bird's nest main body. Specific implementation manners

[0119] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners:

[0120] The present invention combines cutting-edge technologies in multiple fields such as avian ecology, artificial intelligence, big data analysis, mechanics, and generative design, and creatively proposes a full-process digital method for migratory bird protection, promoting the innovation and application of ecological protection technologies, and providing useful references and inspirations for research in related fields.

[0121] The present invention proposes a dynamic force model for bird's nests, which combines the activity behaviors, habitats, and material properties of birds to simulate the force distribution of bird's nests under different conditions. This model can dynamically reflect the force changes of bird's nests during use, providing scientific theoretical support for the design of artificial nests.

[0122] The present invention innovatively combines large language models with multi-dimensional data in the field of migratory bird protection, and through a comprehensive analysis of the habitat requirements and ecological environment of migratory birds, realizes personalized artificial nest design under different habitat conditions.

[0123] Such as 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 derivative design part. In the large language model part, a multidimensional knowledge data set in the field of migratory bird protection is constructed by means of 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 derivative 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 best form is determined through screening, and finally the production and on-site installation of the bird's nest are completed.

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

[0125] Step 1: Collect multi-source data in the field of migratory bird protection through observation records, literature, field interviews, professional data released by research institutions, and real-time monitoring results of protected areas, extract key knowledge information related to migratory bird behavior and habitat, and construct a multidimensional knowledge dataset in the field of migratory bird protection. 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 in the field of migratory bird protection includes the following aspects:

[0126] (1) Appearance of migratory birds: High-precision descriptions 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 1200 mm in length. Its body feathers are white except for the black primary flight feathers. The skin around and before the eyes is red, and the eyes are also red. The beak is black and 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: the dynamic characteristics of migratory bird migration routes, habitat behaviors, and different life stages, especially the typical behaviors in different regions and 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 the 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. Because the outer edges of the secondary flight feathers are grayish white, the black part is black and white. Looking up at the belly, a black band can be seen separated by the white body."

[0128] (3) Habitat characteristics: covering geographical location, climatic conditions, topography and biological resources, etc. For example, "What are the natural geographical characteristics of Honghe National Nature Reserve in Heilongjiang as one of the important habitats of the Oriental Stork?", "The geographical coordinates of Honghe National Nature Reserve in Heilongjiang are 133°34′38″ - 133°46′29″ east longitude and 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 °C, and an average annual precipitation of about 500 - 600 mm. The landform plants are mainly herbaceous swamps and aquatic vegetation, with island forests distributed intermittently. Honghe Reserve retains the original swamp landscape before the reclamation of the Sanjiang Plain. The reserve is relatively closed, with less human activity and low interference intensity. The sparse reed swamps, winding river shallows and large areas of lakes provide a comfortable habitat for the Oriental Stork."

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

[0130] (5) Analysis of migratory bird behavior: Integrate literature data to summarize the potential causes and ecological significance of migratory bird behavior. For example, "What might be the reason for the feeding time of the Oriental Stork from 4:00 to 18:00 observed in Xingkai Lake National Nature Reserve in Heilongjiang in 2015?", "Every day from 14:00 to 18:00 is one of the peak periods for adult Oriental Storks to leave the nest to forage, and the chicks generally have their last meal at 20:00. Research shows that due to the drop in night temperature, 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 Xingkai Lake National Nature Reserve was about 20 °C. Therefore, the Oriental Stork needs to feed before dark to meet its energy consumption."

[0131] (6) Nest parameters: Sort out the structure, materials, and functional parameters of existing natural and artificial bird nests. For example, "What are the parameters of the Oriental Stork nests in the Yellow River Delta National Nature Reserve in 2009?", "In the Dawenliu nesting area, the average nest height is 13.25 m ± 2.07 m, and the nest spacing is 647.22 m ± 1086.49 m; in the Yellow River Estuary nesting area, the average nest height is 25.50 m ± 7.97 m, and the nest spacing is 42640.00 m ± 62838.80 m.

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

[0133] Step 2: Locally deploy a large language model and perform customization and optimization. Conduct knowledge reasoning through a knowledge base and an inference engine to 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: Local deployment of the large language model. Locally deploy an open-source large language model that has been launched on the market, such as the llama3 language model. Input simple questions to verify the accuracy of its answers.

[0135] Step 2.2: Conversion of the dataset format. Convert the multi-dimensional knowledge dataset in the field of migratory bird protection into the input format of the large language model.

[0136] Step 2.3: Customization and optimization of the large language model.

[0137] Efficiently fine-tune the large language model through the Low-Rank Adaptation (LoRA) method to enhance its professionalism and accuracy in migratory bird identification and protection advice generation. Convert the multi-dimensional knowledge dataset in the field of migratory bird protection into JSON format and load it into the large language model to ensure the generalization ability of the original model is retained during the training process, and achieve the integration of directional optimization and knowledge supplementation. Finally, the large language model can not only expand the migratory bird protection knowledge base through knowledge reasoning but also adjust its decisions according to new knowledge in different ecological environments. Generate an optimized language model that meets the needs of the migratory bird ecological field to provide intelligent support for subsequent protection plans.

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

[0139]

[0140] As Figure 2 shown, the working process of fine-tuning the large language model by the low-rank matrix adaptation method is as follows:

[0141] Step 2.3.1, weight matrix decomposition; the 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] where ΔW is the pre-trained weight; represents the first low-rank matrix, d is the row dimension of the original weight matrix, 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 the fine-tuning process, A and B are optimized and updated without changing the initial pre-trained weight W o . The specific update formula is:

[0145]

[0146] where 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] where W o is the initial pre-trained weight and remains 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 through the LoRA method. Then:

[0152] h = W 0 x+ΔWx = W 0 x+(A′·B′)x;

[0153] LoRA uses the initial pre-trained weight W of the original model o to provide initial generalization ability, realizes customized optimization by fine-tuning the newly added weight part (ΔW = A·B), and by optimizing the low-rank matrices A and B, while keeping the pre-trained weight W oOn the premise of remaining unchanged, achieve the rapid adaptation and efficient training of domain knowledge.

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

[0155] Comprehensively test the performance of the customized and optimized large language model in combination with the actual application scenario. For example, when asking for suggestions on building an artificial bird's nest, verify whether the model can provide comprehensive and scientific suggestions, including the location, height, diameter, material properties, structural parameters, and environmental adaptability of the bird's nest. If the output result of the customized and optimized large language model fails to meet the expectations, expand the multi-dimensional knowledge dataset in the field of migratory bird protection through the iterative optimization mechanism, or adjust the training parameters, such as the rank of the LoRA low-rank matrix, the output scaling ratio, the training batch size, etc., to further improve the model's understanding and reasoning ability for complex problems. After quantization processing, the customized and optimized large language model should support multi-terminal deployment and fast response.

[0156] Step 2.5: Apply the customized and optimized large language model; after inputting the current situation of migratory bird habitats in the area into the customized and optimized large language model, output customized migratory bird protection suggestions; after inputting the environmental conditions for building a bird's nest in the current area into the customized and optimized large language model, output customized bird's nest building suggestions and specific parameters.

[0157] Step 3: Establish a dynamic force model of the bird's nest. Input the specific parameters for building the bird's nest obtained in Step 2.5 into the dynamic force model of the bird's nest to calculate the force situation at a certain point of the bird's nest within a few seconds when the bird takes off, lands, and stays in the nest. The specific parameters for building the bird's nest include the weight of the bird, the velocity and acceleration of the bird in the x, y, and z directions within the interval of 2 to 5 seconds, the stiffness of the bird's nest material, the contact damping, the friction coefficient, the test nodes of the bird's nest, etc.

[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 bird's nest, combined with the observed data and material parameters. The dynamic force model of the bird's nest is used to analyze the force situation and distribution exerted by the bird on the bird's 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: Establish a dynamic model of the bird's nest. Describe the position, velocity, and acceleration of the bird at time t through a trajectory function, and then calculate the motion state of the bird. The formula is as follows:

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

[0161]

[0162] F gravity = m·g;

[0163] where l is the position of the bird; fl (·) represents the trajectory function of the position l of the bird; v is the velocity; a is the acceleration; F gravity is the 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; specifically as follows:

[0165] The distribution of the force application points is related to the material stiffness, and the distribution of the contact force is calculated by 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 related to the i-th node.

[0168] Subsequently, consider adjusting the distributed force by 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. F nitial is the basic force on the node;

[0171] Step 3.3: Establish the friction force 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, which is used to represent the reaction force between the object and the contact surface, and its calculation formula is:

[0175]

[0176] Among them, F dynamic is the dynamic contact force; is the material stiffness; δ is the deformation amount, which represents the degree to which the object is compressed or stretched. c is the damping coefficient, which represents the damping ability of the material. v is the velocity, and ∥v∥ is the modulus of the velocity (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 frictional force, representing the comprehensive force between the object and the contact surface. Its calculation formula is:

[0178]

[0179] where F total is the total contact force. At this time, the time variation of the force at each point can be output according to the initially set force application points of the bird's nest.

[0180] Step 4: Calculate the node weights, generate the point cloud model and the force-bearing surface model, and perform visualization.

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

[0182] Step 4.1: Generation of the geometric model of the bird's nest; generate the geometric shape of the point cloud of the bird's nest based on a hemispherical shape:

[0183]

[0184] where 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, representing the angle between the point and the z-axis, with a range of 0 ≤ θ ≤ π; φ is the azimuth angle, representing the angle between the projection of the point on the plane formed by the x-axis and the y-axis and the x-axis, with a range of 0 ≤ φ < 2π;

[0185] Select the nodes with z > 0 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 and adjusted:

[0187]

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

[0189] Step 4.2: Calculation and screening of node weights: To optimize the point cloud data and delete unnecessary force application points, a multi-factor node weight calculation method is established. The steps are as follows:

[0190] Step 4.2.1: Calculation of the force weight of the node. It is calculated through normalization processing and combined with the fluctuation characteristics of the force distribution. The formula is:

[0191]

[0192] where ω force is the force weight; Fnitial is the basic force on the node; σ(F nitial ) is the standard deviation of F nitial .

[0193] Step 4.2.2, Geometric weight calculation. Combining the distance from the node to the center of the bird's nest, an exponentially decreasing weight assignment strategy is adopted, and the formula is:

[0194]

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

[0196] Step 4.2.3, Calculate the material weight in combination with the influence of the stiffness matrix:

[0197]

[0198] where 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 through the dot product of the force gradient and the position vector. The formula is:

[0200]

[0201] where 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 in the form of a multi-factor weighted product to reflect the non-linear interaction of the weight factors. The formula is:

[0203]

[0204] where a 1 , a 2 , a 3 , a 4 are different adjustment coefficients used to adjust the influence weights of each factor.

[0205] Step 4.2.6, Preset the threshold τ, perform weight point screening, and delete the nodes with a comprehensive weight lower than the threshold τ.

[0206] Step 4.3, Assign colors to each node according to the value of the comprehensive weight to realize the 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 stress intensity of the nodes are displayed through a 3D scatter plot.

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

[0211]

[0212] where is the triangle area; are the different side vectors of the triangle.

[0213] Step 4.5: Highlight the stress concentration area to realize the visualization of the stress surface model; Mark the high-stress area, screen the nodes with stress values greater than a preset ratio, and highlight them in different colors in the 3D model. The specific judgment condition is: when F > γ·max(F), the nodes are marked. Where F is the stress value of the node; γ is the screening threshold;

[0214] Figure 3 This is a possible visualization image of the stress point cloud model of a bird's nest in the solution of the present invention. Each point in the figure represents a stress point, and the color of the point represents the stress intensity. The area with dense point cloud represents the stress concentration area, while the area with sparse point cloud represents the part with less stress.

[0215] Figure 4 This is a possible visualization image of the high-stress point cloud of a bird's nest in the solution of the present invention. Each point in the figure represents a stress point, and the stress value of the points in the high-stress area of the bird's nest exceeds the set threshold τ, and the remaining bird's nest points represent the stress points below the threshold τ.

[0216] Figure 5 This is a possible visualization image of the stress surface of a bird's nest in the solution of the present invention. According to the distribution of the high-brightness points of the point cloud, the triangulation algorithm is used to process the point cloud data, convert it into a triangular mesh, and thus draw a continuous stress surface model.

[0217] Step 5: Design and model the bird's nest base. In the 3D design platform, draw the basic platform for placing the bird's nest main body and establish the 3D model of the bird's nest base. For different construction environments of the bird's nest, design different connection structures between the bird's nest main body and the bird's nest base. For example, when the construction environment is an angle steel frame, a column, etc., metal plates and plastic parts can be used, and metal cables and card slots can be designed so that the bird's nest main body can be tightly installed on the bird's nest base subsequently.

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

[0219] Step 6.1, Determine the installation location, height, and overall dimensions of the artificial bird's nest based on the suggestions of the large language model and the results of the force analysis.

[0220] Step 6.2, Generative design process. Using the generative design tool, based on various mechanical parameters, dimension and geometric shape parameters, design goals, and constraints, input them into the generative design tool for automated design to generate the main body of the bird's nest that meets the habitat needs of migratory birds. This process is carried out through the following steps:

[0221] Step 6.2.1, Definition of the structural contour:

[0222] Combined with the basic dimensions 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, set the main structural support parts of the bird's nest, and generate the shapes and dimensions of each structural component according to the force distribution data.

[0224] Step 6.2.2, Demarcation and optimization of the stress area:

[0225] According to the stress surface of the main body of the bird's nest, optimize the areas with greater stress. By calculating the stress concentration areas, ensure that the strength of these areas is focused on strengthening during the design process.

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

[0227] Step 6.2.3, Obstacle avoidance and contact object setting:

[0228] Identify the objects that the bird's nest needs to avoid (such as the base support structure, external equipment, etc.) and set them as obstacles. At the same time, determine the objects that the bird's nest must contact (such as support rods or cameras) and define them as contact objects to ensure that there are no conflicts in the connection and force of each part during the design process.

[0229] Step 6.2.4, Generation of goals and constraints:

[0230] Set the optimization goals of the generative design to ensure that the designed main body of the bird's nest minimizes material consumption while meeting the requirements of strength, safety, and adaptability.

[0231] At the same time, add the adaptability constraints of the installation environment, considering possible interference factors during the actual installation process, such as installation angle, wind influence, etc.

[0232] Step 6.3, Final Design Output: Conduct a detailed review of the main body of the bird's nest generated by the generative design tool and ensure that the design meets the preset strength, size, and material requirements. Output a detailed design document and 3D model of the main body of the bird's nest and provide them to the manufacturing team for subsequent production and installation.

[0233] Step 7, Manufacture the main body of the bird's nest and conduct deployment and monitoring.

[0234] Step 7.1, Based on the generative design results, fabricate a customized main body of the bird's nest through high-precision manufacturing processes. The fabrication of the main body of the bird's nest includes material processing, structure forming, and base assembly to ensure that it meets the requirements of strength, force distribution, and environmental adaptability in the design. After manufacturing, combine the main body of the bird's nest with the bird's nest base and conduct on-site deployment.

[0235] Step 7.2, During the deployment process, based on the customized and optimized large language model recommendation strategy, precisely adjust the height, tilt angle, and direction of the bird's nest base and the main body of the bird's nest to ensure its adaptability and stability within the migratory bird activity area. Combine the on-site terrain and environmental characteristics to optimize the connection method between the main body of the bird's nest and the bird's nest base, reduce interference with the surrounding ecology, and extend the service life of the bird's nest.

[0236] Step 7.3, Conduct subsequent monitoring by installing sensors or camera devices to further verify the design effect of the bird's nest and provide data support for future optimization and design upgrade of the large language model.

[0237] Figure 6 This is a schematic diagram of a possible structure of the Oriental Stork bird's nest base for the solution of the present invention. The bird's nest base includes a central axis connector 1, a bird's nest chassis 2, a fixing cable adjustment knob 3, a metal fixing cable 4, a bird feces collection port 5, a fixing member 6, a base upper 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 surface of the bird's nest chassis 2. The center position of the back surface of the bird's nest chassis 2 is connected to the base upper cover plate 7 through the fixing member 6. Two fixing cable adjustment knobs 3 are symmetrically arranged on the front surface of the base upper cover plate 7. A fixing bayonet 8 is arranged between the two fixing cable adjustment knobs 3. Two metal fixing cables 4 are symmetrically arranged at the bottom of the angle steel 9. Each metal fixing cable 4 contains 3 parallel and retractable metal ropes inside. The connection between the base upper cover plate 7 and the angle steel 9 is realized through the connection between the fixing cable adjustment knob 3 and the metal fixing cable 4.

[0238] Among them, the central axis connector 1 is used to connect the bird's nest base with the main body of the bird's nest designed by generative design. The main body of the bird's nest is inserted and fixed on the bird's nest base through the connector 1. During generative design, the central axis connector 1 serves as the main object for generation and contact.

[0239] The bird's nest chassis 2 is in the shape of a disk, and its size is smaller than the maximum diameter of the generative design bird's nest. The disk gathers toward the center and is used to receive bird droppings and debris. The bird's nest chassis 2 is used as an object to avoid in the generative design.

[0240] The base of the bird's nest base is fixed to the bird's nest placement location with a metal rope, and the metal fixing rope 4 can be tightened by twisting the fixing rope adjustment knob 3 clockwise to tighten the bird's nest base;

[0241] After the bird's nest base is installed, the metal rope ring is pulled out and passed through the fixing rope adjusting knob 3 to be wound around the structure at the fixed position, and then the metal fixing rope 4 is tightened 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 needs to collect and analyze 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 part 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, etc., the bird's nest base fixing bayonet 8 is designed to facilitate fixing 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 a possible force constraint setting for a generative design product in the solution of the present invention is shown. The force constraint setting method in the generative design process is demonstrated. By connecting the triangular force surface and the point cloud, the key parameters such as the point distribution of the force on the bird's nest, the magnitude of the force, the direction of the force, geometric obstacles, gravity, and environmental factors are clarified. Designers can ensure that the bird's nest can withstand various dynamic loads and external influences in actual applications.

[0247] Figure 8 and Figure 9 Generate targets and constraints for a possible generative design product of the invention, and analyze the interface diagram of the material. According to the needs of customers and production, set the generated minimum mass, maximum stiffness, safety factor, and select the materials and processing methods to be used.

[0248] Figure 10 This is a possible artificial nest for migratory birds, including Figure 6The bird's nest base and the bird's nest main body 10. The bird's nest main body 10 is generated through step 7.2. A central axis adapted to the central axis connecting member 1 is provided below it. The central axis and the central axis connecting member 1 are connected and fixed by a pin. The bird's nest bottom plate 2 is threadedly connected to both the fixing member 6 and the base upper cover plate 7. The bird's nest base and the whole bird's nest are fixed to the angle steel 9 of the electric tower or the vertical pole or the pole end through the tensioning metal cable base 4. When necessary, screws are used for reinforcement and it is placed in the 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 those skilled in the art within the essence of the present invention should also fall within the protection scope of the present invention.

Claims

1. An ecological protection knowledge reasoning and artificial bird nest generative design method, characterized in that: The steps include: Step 1: Collect multi-source data in the field of 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. Perform knowledge reasoning through the knowledge base and reasoning engine, simulate the decision-making process of bird experts, and test and apply the customized and optimized large language model. Step 3, establishing a dynamic force model of the bird's nest; Step 4: Calculate the node weights, generate the point cloud model and the 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; 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 in that: In the 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 data set 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.

3. The ecological protection knowledge reasoning and artificial bird nest generative design method according to claim 1 is characterized in that: The specific process of step 2 is: Step 2.1: Locally deploy the open source large language model; Step 2.2, converting the multidimensional knowledge dataset in the field of migratory bird protection into a large language model input format; Step 2.3: Fine-tune the large language model through the low-rank matrix adaptation method to achieve customized optimization; the specific process is: 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, A and B are optimized and updated. The formula is: Among them, 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; conduct a comprehensive test on the performance of the customized and optimized large language model in combination with actual application scenarios; if the output result of the customized and optimized large language model fails to meet expectations, expand the multidimensional knowledge data set 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 habitat conditions of migratory birds in the current area into the customized and optimized large language model, and output customized migratory bird protection suggestions; input the environmental conditions for building bird nests in the current area into the customized and optimized large language model, and output customized bird nest building suggestions and specific parameters.

4. The ecological protection knowledge reasoning and artificial bird nest generative design method according to claim 1 is characterized in that: 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 by the node average 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: Among them, 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, the formula is: 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, the formula is: 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: Preset 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 transform 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 realize the visualization of the load-bearing 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 three-dimensional 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 in that: 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.

7. The ecological protection knowledge reasoning and artificial bird nest generative design method according to claim 1 is characterized in that: The specific process of step 6 is as follows: Step 6.1, based on the large language model suggestion and the force analysis results, determine the installation location, height and overall size of the artificial bird's nest; Step 6.2: Perform automated design based on generative design tools to generate a bird's nest that meets the habitat needs of migratory birds. The specific process is as follows: Step 6.2.

1. Based on the basic size and shape range of the bird's nest, define the outer contour and the positions of each interface of the bird's nest; in the generative design tool, set the core structural support part of the bird's nest, and generate the shape and size of each structural component according to the force distribution data; Step 6.2.2, optimize the area with high stress according to the stress-bearing surface of the main body of the bird's nest; adjust the thickness and stiffness of the structure in combination with the node weight and the mechanical properties of the material; Step 6.2.3, identify the objects that the bird's nest needs to avoid and set them as obstacles to avoid; at the same time, determine the objects that the bird's nest must contact with and define them as touching objects to ensure that the connection and force of each part do not conflict during the design process; Step 6.2.4, set the optimization goal of the generative design and add the adaptability constraint of the installation environment; Step 6.3: Review the main body of the bird's nest generated by the generative design tool in detail and ensure that the design meets the preset strength, size and material requirements; Output detailed main design documents and 3D models of the Bird's Nest and provide them to the manufacturing team for subsequent production and installation.

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

1. According to the generative design results, a customized bird's nest body is manufactured through high-precision manufacturing technology. The manufacturing of the bird's nest body includes material processing, structural molding and base assembly. Step 7.2: After manufacturing is completed, the main body of the bird's nest is combined with the base of the bird's nest for on-site deployment. During the deployment process, the height, tilt angle and direction of the base and the 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 in combination with 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 language model optimization and design upgrades.

Images