Water conservancy facility layout and termite prevention and control method and system

Abstract

The invention discloses a water conservancy facility layout and termite prevention and control method and system, and the method comprises the steps: termite behavior data collection and feature integration, including termite path selection features, nesting and path construction, environmental adaptability, abnormal behavior and error correction, and construction of a multi-dimensional comprehensive data set; carrying out fine mapping processing and preprocessing on the behavior characteristic data so as to facilitate training and optimization of a subsequent algorithm model; constructing a multi-objective optimization and hierarchical genetic algorithm, and training the multi-objective optimization and hierarchical genetic algorithm by using the integrated feature data; a multi-level optimization and feedback mechanism is adopted, and optimization measures are continued to be deepened and perfected based on an optimization result. Compared with a traditional water conservancy facility layout method, by combining natural behaviors of termites and a multi-objective optimization and hierarchical genetic algorithm technology, more efficient and more flexible path optimization layout is realized, and an innovative solution is provided for long-term stable operation of water conservancy facilities and termite prevention and control.

Description

Technical Field

[0001] The present invention relates to the technical fields of water conservancy projects and biological prevention and control, and particularly to a method and system for the layout of water conservancy facilities and termite prevention and control. Background Art

[0002] In the fields of water conservancy projects and biological prevention and control, effective layout of water conservancy facilities and termite prevention and control are of great significance. Traditional water conservancy facility planning and prevention and control methods often rely on fixed design principles and empirical guidance. However, this method has many limitations and defects when facing complex terrains and dynamic environments. For example, it lacks the ability of dynamic adjustment and there are contradictions between global and local planning.

[0003] In recent years, optimization algorithms inspired by biological behaviors have gradually become a research hotspot. Termite behavior simulation, as a bio-inspired optimization method, provides inspiration for the layout of water conservancy facilities by simulating the nest-building and path selection behaviors of termites, making up for the deficiencies of traditional methods. However, there are still many problems in existing termite behavior simulation algorithms. For example, it is difficult for the algorithm to find the best balance between global planning and local adjustment, and its adaptability and stability in a dynamic environment are insufficient.

[0004] Traditional water conservancy facility layout methods have certain limitations. For example, it is difficult to take into account the requirements of global layout and local optimization, resulting in it being difficult to simultaneously achieve the optimal stability and efficiency of the path network. To this end, researchers have proposed multi-objective optimization and hierarchical biological algorithms, which improve the self-adaptability and overall performance of water conservancy facility layout by combining multi-dimensional feature data of termite behaviors with the optimization ability of genetic algorithms. However, in practical applications, how to effectively combine these improved algorithms to form a systematic method for the layout of water conservancy facilities and termite prevention and control still requires further research and exploration. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems in the related technologies to some extent.

[0006] The present invention proposes a method for the layout of water conservancy facilities and termite prevention and control, which realizes a more efficient and flexible path optimization layout by combining the natural behaviors of termites with multi-objective optimization and hierarchical genetic algorithm technologies, and provides an innovative solution for the long-term stable operation of water conservancy facilities and termite prevention and control.

[0007] Another object of the present invention is to propose a system for the layout of water conservancy facilities and termite prevention and control.

[0008] To achieve the above object, on the one hand, the present invention proposes a method for the layout of water conservancy facilities and termite prevention and control, including:

[0009] Construct a multi - dimensional comprehensive dataset based on termite path - selection characteristics, nesting and path - building data, environmental adaptability data, abnormal behavior and error - correction data;

[0010] Perform mapping processing and pre - processing on the multi - dimensional comprehensive dataset to obtain pre - processed data for model training;

[0011] Set multiple optimization objectives based on the multi - dimensional comprehensive dataset to obtain a multi - objective optimization formula, and calculate the comprehensive score of each path in the path network according to the multi - objective optimization formula for preliminary path optimization; Use the pre - processed data to train a hierarchical genetic algorithm model to decompose the path - planning task into multiple levels of optimization to obtain an optimization result;

[0012] Optimize the total length and environmental adaptability of the path network using a global optimization formula, and adjust the weights in the global optimization formula according to real - time environmental data to perform multiple levels of path correction based on the optimization result.

[0013] To achieve the above - mentioned object, on the other hand, the present invention proposes a water conservancy facility layout and termite prevention and control system, including:

[0014] A feature data acquisition module, which is used to construct a multi - dimensional comprehensive dataset based on termite path - selection characteristics, nesting and path - building data, environmental adaptability data, abnormal behavior and error - correction data;

[0015] A data pre - processing module, which is used to perform mapping processing and pre - processing on the multi - dimensional comprehensive dataset to obtain pre - processed data for model training;

[0016] An objective optimization and layering module, which is used to set multiple optimization objectives based on the multi - dimensional comprehensive dataset to obtain a multi - objective optimization formula, and calculate the comprehensive score of each path in the path network according to the multi - objective optimization formula for preliminary path optimization; Use the pre - processed data to train a hierarchical genetic algorithm model to decompose the path - planning task into multiple levels of optimization to obtain an optimization result;

[0017] A multi - layer optimization and feedback module, which is used to optimize the total length and environmental adaptability of the path network using a global optimization formula, and adjust the weights in the global optimization formula according to real - time environmental data to perform multiple levels of path correction based on the optimization result.

[0018] The water conservancy facility layout and termite prevention and control method and system of the embodiments of the present invention introduce multi - objective optimization and hierarchical genetic algorithm technologies, aiming to improve the overall efficiency and stability of water conservancy facility layout by dynamically adapting to complex environmental conditions, and solve the problems that traditional methods are difficult to balance between global layout and local optimization, and lack the ability of real - time adjustment in a dynamic environment.

[0019] Additional aspects and advantages of the present invention will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and / or additional aspects and advantages of the present invention will become apparent and be readily understood from the following description of embodiments in conjunction with the accompanying drawings, where:

[0021] Figure 1 It is a flowchart of a water conservancy facility layout and termite control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0022] Figure 2 It is a path selection diagram of a water conservancy facility layout and termite control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0023] Figure 3 It is a pheromone concentration diagram of a water conservancy facility layout and termite control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0024] Figure 4 It is a path smoothness diagram of a water conservancy facility layout and termite control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0025] Figure 5 It is a connectivity diagram between path points of a water conservancy facility layout and termite control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0026] Figure 6 It is a multiple simulation score trend diagram of a water conservancy facility layout and termite control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0027] Figure 7 It is a different-level optimization score diagram of a water conservancy facility layout and termite control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0028] Figure 8 It is a change trend diagram of the total length of the path and environmental adaptability of a water conservancy facility layout and termite control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0029] Figure 9 It is a score comparison diagram between path A and path B of a water conservancy facility layout and termite control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0030] Figure 10A trend chart of path deviation and adjustment time for a water conservancy facility layout and termite prevention and control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0031] Figure 11 An adaptive closed-loop optimization system diagram for a water conservancy facility layout and termite prevention and control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0032] Figure 12 A heat map of on-site testing for a water conservancy facility layout and termite prevention and control method based on multi-objective optimization and hierarchical biological algorithms provided by the present invention;

[0033] Figure 13 A structural diagram of a water conservancy facility layout and termite prevention and control system based on multi-objective optimization and hierarchical biological algorithms provided by the present invention. Detailed implementation manners

[0034] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments may be combined with each other. The present invention will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0035] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0036] The water conservancy facility layout and termite prevention and control method and system proposed according to the embodiments of the present invention will be described below with reference to the drawings.

[0037] Figure 1 is a flowchart of a water conservancy facility layout and termite prevention and control method according to an embodiment of the present invention. As Figure 1 shown, the method includes:

[0038] S1, constructing a multi-dimensional comprehensive data set based on termite path selection characteristics, nesting and path construction data, environmental adaptability data, and abnormal behavior and error correction data;

[0039] S2, performing mapping processing and preprocessing on the multi-dimensional comprehensive data set to obtain preprocessed data for model training;

[0040] S3. Set multiple optimization objectives based on the multi-dimensional comprehensive data set to obtain a multi-objective optimization formula, and calculate the comprehensive score of each path in the path network according to the multi-objective optimization formula for preliminary path optimization; train a hierarchical genetic algorithm model using the preprocessing data to decompose the path planning task into multiple levels of optimization to obtain an optimization result;

[0041] S4. Optimize the total length and environmental adaptability of the path network using the global optimization formula, and adjust the weights in the global optimization formula according to real-time environmental data to perform multiple levels of path correction based on the optimization result.

[0042] In this embodiment, S1 includes the following sub-steps:

[0043] S11. The foraging behavior and path efficiency characteristics of termites reflect their natural mechanism for finding the optimal path in complex terrains. By collecting the decision-making data of termites' path selection in complex terrains, including path length, pheromone concentration and distribution, path smoothness, etc., deeply analyze the path selection patterns of termites under different environmental conditions, so as to provide accurate path optimization references for the model and ensure that the algorithm can effectively simulate the natural foraging behavior of termites and achieve the optimization of global and local path planning.

[0044] First, a complex terrain was set up in the experimental environment to simulate the foraging behavior of termites in the natural environment. In order to accurately capture the movement trajectory of termites, a high-resolution camera was used to record the path data of termites from the nest to the food source. At this stage, special attention should be paid to the calculation of path length, because the path length can directly reflect the efficiency of termites' path selection under different terrains. For this reason, the present invention proposes a calculation formula for path length:

[0045]

[0046] where L represents the total length of the path selected by the termite, and d(x i , x i+1 ) represents the Euclidean distance between adjacent points x i and x i+1 on the path.

[0047] As Figure 2 shown, it shows the path selection process of termites. Each blue dot represents a node on the path, and the distances between these nodes are calculated to obtain the total length of the path. This illustration helps to better understand how termites find a suitable path in complex terrains.

[0048] Then, the concentration of pheromones released by termites on the path was measured using highly sensitive chemical sensors. These pheromones guide other termites to move along a specific path. The calculation formula for the pheromone concentration C proposed in the present invention is as follows:

[0049]

[0050] where C represents the total sum of pheromone concentrations at all points on the path, and c(x i ) represents the local pheromone concentration at point x i on the path. These concentration values show how termites select and reinforce certain paths through chemical signals.

[0051] As Figure 3 shown, it shows the average pheromone concentration for each segment on the path. These data reveal the preferences of termites in path selection and the importance of the path, which helps to further understand the selection mechanism of termites.

[0052] To understand more deeply the path selection strategy of termites in complex terrains, the smoothness of the path was then evaluated. Smoothness refers to the continuity and curvature change of the path, which reflects the feasibility and safety of the path during the movement of termites. The calculation formula for the path smoothness S proposed in the present invention is as follows:

[0053]

[0054] where S represents the smoothness of the path, θ i-1,i,i+1 represents the angle between the midpoints x i-1 , x i+1 , x i on the path, and dist(x i-1 , x i+1 ) represents the distance between x i-1 and x i+1 .

[0055] As Figure 4 shown, it shows the angles between each segment on the path, and these angles reveal the smoothness of the path. Larger angles may mean that the path is not smooth, while smaller angles indicate that the path is more straight and smooth. This part of the analysis helps to understand how termites ensure the feasibility and safety of the path in path selection.

[0056] Finally, all the collected characteristic data (path length L, pheromone concentration C, path smoothness S) were integrated into the evaluation function proposed in the present invention to comprehensively evaluate the optimization effect of the path and ensure that the selected path can maximize the benefits in practical applications. The evaluation function is expressed as:

[0057]

[0058] Among them, P is the optimized score of the path. C is the pheromone concentration. L is the path length. S is the path smoothness. w 1 and w 2 are the importance weight coefficients that measure the balance between the pheromone concentration and the path length, and the path smoothness respectively.

[0059] By adjusting the relationship between the pheromone concentration C and the path length L and the path smoothness S, as well as different weights w 1 and w 2 , the optimization effect of the path in termite behavior can be more precisely reflected. The higher the score P, the more in line with the natural selection preference of termites the path is after comprehensively considering the pheromone concentration, path length, and smoothness.

[0060] Through the calculation of the evaluation function, the path selection preference of termites in complex terrain can be specifically obtained. As shown in the table:

[0061] Table 1

[0062]

[0063] The above table calculates the following for each path: The highest for Path 3 is 0.3824, indicating that it is the best path choice. Although its smoothness is slightly poor, its shorter path length and higher pheromone concentration make it superior in the overall evaluation. The second highest is for Path 1, which is 0.3573, and it is still a viable path choice. The lowest is for Path2, which is 0.2833, so this path should be avoided.

[0064] To sum up, termites tend to choose paths with shorter lengths and higher pheromone concentrations in complex terrain, although the smoothness is lower. This also suggests that when optimizing the layout of water conservancy facilities, the path length and the effectiveness of guiding information should be given priority to achieve optimal design.

[0065] The characteristics of the nest-building and path construction processes of termites directly simulate the gradual optimization process of path planning. By obtaining the behavioral data of the gradual construction of paths by termites during nest-building, recording the selection basis, construction order, and connectivity with other paths of each path point, the balance between local optimization and global planning in the path selection and nest-building processes of termites can be analyzed in detail;

[0066] First, set multiple possible nest locations and connecting paths in the experimental environment. To accurately record the nest-building behavior of termites, track the movement trajectories of termites through sensors and record how they select path points and gradually build the nest network. During this process, use the path selection function to calculate the selection probability of each path point:

[0067]

[0068] Among them, P 选择 represents the selection probability of path points, d represents the distance between path points and the current nest or other paths, C represents the pheromone concentration of path points, E represents environmental characteristics (such as temperature, humidity, etc.), and α, β, γ are weight coefficients that balance the influence of these factors. The key to this step lies in analyzing how termites select path points in a complex environment through a quantitative formula.

[0069] Then, based on the record of path point selection, the path construction order of termites is further tracked. By observing the behavior of termites, the construction time and order of each path point are recorded in sequence, and these data are summarized in a table, as shown in Table 2:

[0070] Table 2

[0071]

[0072] Observing this table, in the global planning stage, the model not only values the single selection probability but also ensures the overall efficiency and adaptability of the path network. In this case, the construction of Path 1 can provide a better foundation for the entire network in subsequent local and micro-optimizations, so it is preferentially selected.

[0073] Next, focus on the connectivity between path points of termites. The purpose of this step is to analyze how termites establish connections between different path points, especially the strategies when expanding new paths on the basis of existing paths. In this process, the path connection formula is used:

[0074]

[0075] Among them, F 连接 represents the optimization score of path connection, P 选择 represents the selection probability of path points, and l 连接强度 is an index that measures the connection tightness between path points. Through this formula, the behavior of termites in the path connection process can be quantitatively evaluated and provide a reference for the optimization of the path network.

[0076] Finally, all the collected data are integrated to construct a comprehensive prediction model. This model includes the following key parts:

[0077] (1) Weight adjustment of path selection: Dynamically adjust the weights of each path under different environmental conditions according to the selection probability and connection strength of path points.

[0078] (2) Balance between global and local: The model ensures the optimal layout of the path network in different environments by balancing the total length of the global path layout and the stability and economy of local paths.

[0079] (3) Connectivity optimization of the path network: The comprehensive prediction model ensures a high degree of connectivity of the path network at both the global and local levels, reducing the risk of breaks.

[0080] (4) Multiple iteration optimization: Based on simulation experiments under different conditions, the model repeatedly adjusts the path layout to achieve an optimal comprehensive path network.

[0081] The final model generates an optimized path network structure, where the selection, connection, and construction order of each path point are optimized to ensure the best application effect in complex terrains and dynamic environments. As Figure 5 shown, the figure shows a schematic diagram of the path network structure, including three levels: global layout, local optimization, and micro-adjustment. The blue lines represent the main path connections of the global layout, ensuring the coherence and efficiency of the path network. The green lines indicate the locally optimized paths, reflecting the optimization of path smoothness and stability in specific areas. The red lines show the micro-adjustments, which are used to handle minor corrections in the path, such as bypassing obstacles or making small-scale path adjustments to cope with emergencies and ensure the efficient operation of the network.

[0082] S13. Termite environmental adaptability: The environmental adaptability characteristics reflect the flexibility and adaptability of termites in the face of complex terrains and dynamic environments. By simulating the reactions of termites under different environmental conditions, the algorithm can be made to have the ability of dynamic adjustment, improving its adaptability in complex terrains.

[0083] First, different environmental conditions (such as temperature, humidity, light intensity, etc.) are set in the experimental environment to simulate the complex terrains and dynamic environments that termites may encounter, and the path selection, nest-building behavior, and adaptability reactions of termites under these conditions are recorded in real time through sensors and camera devices. In the present invention, an adaptability score formula is used to simulate different terrains and dynamic environments:

[0084]

[0085] Among them, A 环境 represents the environmental adaptability score, T k represents the temperature, H k represents the humidity, L k represents the light intensity, w T , w H , w L are the weight coefficients of these factors respectively.

[0086] Then, data analysis is carried out, and the path selection probability formula is used to study how termites adjust their paths when the environmental conditions change:

[0087]

[0088] Among them, P 选择 represents the selection probability of a path point, d represents the distance between the path point and the current nest or other path points, C represents the pheromone concentration of the path point, E represents environmental characteristics (such as temperature, humidity, etc.), ΔE represents the impact factor of environmental change (for example, the change rate of environmental characteristics), and α, β, γ, δ are weight coefficients used to balance the influence of these factors. This path selection probability formula introduces the impact factor ΔE of environmental change, which represents the degree of change in environmental conditions. Such changes may include sharp changes in temperature or fluctuations in humidity. By introducing this factor, the formula can more accurately reflect the path selection behavior of termites in a dynamic environment, ensuring that the algorithm can adapt to and respond to real-time changes in the environment.

[0089] Next, based on the responses of termites to different environmental conditions, a dynamic adjustment mechanism is designed using a dynamic optimization formula, enabling the algorithm to respond to environmental changes in real time:

[0090]

[0091] Among them, F 动态 represents the path score after dynamic optimization, A 环境 represents the environmental adaptability score, L represents the path length, S represents the path smoothness, θ represents the average angle between path points, w 1 , w 2 , w 3 are the weight coefficients of each factor.

[0092] Finally, the dynamic adjustment mechanism is tested in a simulated complex environment, and the environmental adaptability network optimization formula evaluates its improvement effect on the stability and efficiency of the path network:

[0093]

[0094] Among them, F 网络优化 represents the optimization score F of the entire network 动态i represents the dynamic optimization score of the i-th path point, R i represents the robustness of the path network (i.e., the stability of the path under environmental changes), w 动态 and w 鲁棒性 are the importance weight coefficients of dynamic adjustment and robustness respectively.

[0095] The adaptability and robustness scores of different path networks in a dynamic environment are calculated through the above formulas, as shown in Table 3:

[0096] Table 3

[0097]

[0098] As can be seen from the table, Network 1 has the highest optimization score of 0.815, indicating that it has the best comprehensive performance in terms of adaptability and robustness in a dynamic environment. Therefore, this network is more likely to maintain efficient path selection and a stable network structure in practical applications. Although Network 3 has the highest dynamic optimization score, its robustness score is slightly lower, so the comprehensive score is slightly lower than that of Network 1. Although Network 2 performs well in terms of robustness, its dynamic optimization score is relatively low, resulting in a lower comprehensive rating. Therefore, Network 1 is the optimal path network selection, especially in complex environments where both dynamic adaptability and stability need to be considered.

[0099] S14. Abnormal Behavior and Error Correction of Termites: Abnormal behavior and error correction features help the algorithm handle unexpected problems that may occur in complex terrains. By collecting the correction behaviors of termites after errors occur during path construction, observe how to adjust the path and nest structure to cope with unexpected situations.

[0100] First, set obstacles or introduce environmental changes in the experimental environment to induce errors or abnormal behaviors in termites during path construction, and record these behaviors through sensors and camera devices. Then, analyze the collected data and classify the abnormal behaviors according to the reactions of termites. For example, when a termite deviates from the predetermined path, it is classified as "path deviation"; if the pheromone distribution is abnormal due to environmental factors, it is classified as "abnormal pheromone distribution"; and when an obstacle appears in the path and the termite cannot continue to move forward, it is classified as "path blockage". Based on these classifications, observe the correction behaviors of termites in different situations. For example, in the case of path deviation, termites may strengthen the pheromone distribution or return to the origin to reselect the path; in the case of abnormal pheromone, termites will redistribute the pheromone to ensure the coherence of the path; in the case of path blockage, termites will try to bypass the obstacle or find an alternative path.

[0101] Then, based on these observations, establish correction models for each abnormal behavior. For example, for path deviation, it can be modeled as strengthening the pheromone concentration or adjusting the path selection strategy to correct the deviated path; for abnormal pheromone, it is modeled as a strategy for redistributing pheromone; for path blockage, it is modeled as a strategy for path replanning. And these correction strategies are quantified into formulas, and the effectiveness of each strategy is evaluated through the correction effect score formula F 修正 , specifically, the correction effect score formula is as follows:

[0102]

[0103] Among them, F 修正 represents the effect score of error correction, T l represents the type of abnormal behavior, S l represents the severity of the error, Rl Represents the time required for correction, w 类型 , w 严重性 , w 修正时间 are the weight coefficients of these factors respectively.

[0104] Next, these correction models are integrated into the algorithm so that it can automatically handle and correct error behaviors in actual applications. The path correction effect is evaluated and adjusted in real time through the correction formula to ensure that the path network can quickly recover after an accident:

[0105] R 修正 = α * F 修正 + β * A 环境

[0106] where R 修正 represents the path effect after correction, F 修正 is the correction effect score, A 环境 is the environmental adaptability score, and α and β are the weight coefficients of these factors respectively. This formula helps to optimize the path correction effect in real time to ensure that the path network can quickly recover after an accident.

[0107] Finally, test the performance of the algorithm in a complex environment, especially the effect of the correction mechanism in actual applications, with the aim of verifying whether these correction strategies are effective in different types of abnormal situations. In this process, the comprehensive correction optimization formula is used:

[0108]

[0109] where F 综合修正 represents the final score of the comprehensive correction, R 修正m represents the correction effect of the m-th path point, T m represents the time required for correction, w 修正 and w 时间 are the importance weight coefficients of the correction effect and time respectively.

[0110] Through these steps, the algorithm can quickly and effectively perform path correction when facing abnormal behaviors in complex terrains, ensuring the stability and continuity of the path network.

[0111] In this embodiment, S2 includes the following sub-steps:

[0112] S21. Mapping processing of data features: First, perform mapping processing on the termite behavior feature data collected in S1 to convert complex behavior features into numerical forms that can be processed by the model. These data include:

[0113] (1) Path selection: Represents the path decision-making characteristics of termites in different environments. Mapped to the numerical value 1, mainly focusing on indicators such as path length and pheromone concentration.

[0114] (2) Construction process: Record the construction sequence and connectivity of each path point during the nest building process of termites. Map it to the value 2, covering the selection basis, construction sequence, and connection strength of path nodes.

[0115] (3) Environmental impact: Reflect the adaptability of termites under different environmental conditions, including environmental parameters such as temperature and humidity. Map it to the value 3 for analyzing the impact of environmental changes on path selection and construction process.

[0116] (4) Error correction behavior: Record the correction behavior of termites after errors occur during path construction. Map it to the value 4, involving path deviation correction, pheromone redistribution, etc.

[0117] Through this behavior feature mapping operation, complex termite behavior data is converted into numerical features that can be directly used for algorithm processing. The behavior feature mapping is defined as follows:

[0118] Behavior feature mapping = {Path selection: 1, Construction process: 2, Environmental impact: 3, Error correction behavior: 4}

[0119] S22. Data cleaning and normalization processing: After completing the mapping process, clean and normalize the data to ensure the quality and consistency of the data. Specifically, it includes:

[0120] Data cleaning: Remove outliers, missing values, and merge and correct duplicate data to ensure the integrity and accuracy of the data.

[0121] Normalization processing: Perform normalization processing on the data for each feature dimension, map all data to the same numerical range (usually [0, 1]) to eliminate the scale differences between different feature dimensions and ensure the consistency of model processing.

[0122] S23. Data integration and storage: After completing data mapping, cleaning, and normalization, integrate the processed data into a unified feature database for subsequent model calling and training.

[0123] In this embodiment, S3 includes the following sub-steps:

[0124] S31. Multi-objective optimization; Based on the termite behavior feature data integrated in the previous steps, multiple optimization objectives are set to comprehensively optimize all aspects of the path network. The specific objectives include:

[0125] (1). Shortest path: Make the distance of the path from the nest to the target point the shortest to improve the overall efficiency.

[0126] (2). Minimum construction cost: Consider path smoothness and resource minimization to reduce the construction and maintenance costs.

[0127] (3) Maximum pheromone coverage rate: Ensure the uniform distribution of pheromones on the path to enhance the traceability and stability of the path.

[0128] (4) Highest structural stability: Maintain the overall connectivity of the path network and reduce the risk of breakage or failure.

[0129] Then, based on these goals, a multi-objective optimization formula was introduced to evaluate and quantify the comprehensive performance of each path:

[0130]

[0131] where: F 多目标 represents the comprehensive optimization score of the path, used to evaluate the comprehensive performance of each path on various optimization goals. L i represents the length of the i-th path. The optimization goal is to make the path as short as possible, so L i is used to represent the score of the path length. C i represents the pheromone concentration of the i-th path. The higher the pheromone concentration, the more preferred the path is in termite behavior. S i represents the structural stability or construction cost of the i-th path. The higher the structural stability, the better the path and the higher the score. w L , w C , w S are the weight coefficients of path length, pheromone concentration, and structural stability respectively, used to balance the importance of various optimization goals.

[0132] This formula can balance the relationship between path length, pheromone concentration, and structural stability during the optimization process by adjusting the weight coefficients w L , w C , w S . For example, in an environment with limited resources, the weight w S of construction cost can be increased to reduce resource consumption; in an environment with high path complexity, the weight w L of path length can be increased to ensure that the path is shorter and more efficient.

[0133] Then, after setting the optimization goals and formula, the optimization effect is verified under different environmental conditions through multiple simulation runs. Specifically, the simulation experiment will set different environmental parameters, such as terrain complexity, resource availability, environmental interference, etc. In each simulation run, the algorithm generates different path networks according to these conditions and calculates the comprehensive score of each path according to the multi-objective optimization formula:

[0134] F 路径 = w L * L1 +w C *C + w S *S

[0135] Where: F 路径 represents the comprehensive score of the path. L represents the path length, and the optimization goal is to make it the shortest, so L is used in the formula 1 . C represents the pheromone concentration of the path. The higher the concentration, the better the path. S represents the structural stability or construction cost of the path. The higher the structural stability, the higher the score. w L 、w C 、w S are the weight coefficients of path length, pheromone concentration, and structural stability respectively.

[0136] The simulation results are shown in Table 4:

[0137] Table 4

[0138]

[0139] According to the table data, the comprehensive score of Path 1 is the highest (0.029), so it is selected as the optimal path in this simulation. Although the pheromone concentration of Path 2 is the highest and the structural stability of Path 3 is the best, after comprehensively considering path length, pheromone concentration, and structural stability, Path 1 provides the most balanced optimization effect. Therefore, Path 1 is selected as the optimal solution by the algorithm for further path planning and implementation.

[0140] The results after multiple simulations show that the algorithm can flexibly adjust the path planning scheme according to changes in environmental conditions. As Figure 6 shown, in the simulation with tight resources, the algorithm tends to select paths with lower construction costs; in the simulation with higher environmental complexity, the algorithm pays more attention to the stability of the path and the distribution of pheromones. This not only improves the practicality of the algorithm but also ensures that resource conservation, construction cost reduction, and long-term stable operation of facilities can be achieved in practical applications.

[0141] S32. Hierarchical architecture: In S31, multiple key optimization goals are set through multi-objective optimization, and the preliminary direction of path planning is verified to ensure that the path can achieve a balance among multiple optimization goals. However, the diversity of complex terrains and environments requires further in-depth optimization. Therefore, in S32, a hierarchical genetic algorithm architecture is introduced, and the path planning task is decomposed into global, local, and microscopic three-level optimizations. Each level processes specific optimization goals to ensure that the path network can achieve the best results at different levels.

[0142] First, in the top - level optimization, the focus is on the planning of the global layout, aiming to ensure the coherence and effectiveness of the entire path network at the macroscopic level. To achieve this goal, a global optimization formula is used:

[0143]

[0144] Among them, F 全局 represents the comprehensive score of global optimization, which is used to measure the performance of the path network at the global level. L 总 represents the total length of the path network. The optimization goal is to make it as short as possible, so is used to represent the score of the path length. E represents the adaptability of the path to environmental changes. This factor measures the effectiveness of the path network under different environmental conditions. w G1 and w G2 are the weight coefficients of the path length and environmental adaptability respectively, which are used to balance the importance of these two factors in global optimization. Through this formula, the algorithm can find a network layout that adapts to complex environments and keeps the overall path length the shortest globally. This is equivalent to ensuring the rationality of the overall location of the termite nest and the planning of the main paths in the behavior of termites.

[0145] Then, after the global path layout is determined, the algorithm enters the middle - level optimization stage. This level focuses on the fine - tuning of local paths. The goal of local optimization is to ensure the smoothness and stability of the path in complex areas while minimizing construction costs. During this process, a local optimization formula is used:

[0146]

[0147] Among them, F 局部 represents the comprehensive score of local optimization, which is used to measure the performance of the path in the local area. S 局部 represents the stability of the local path. The higher the stability, the better the performance of the path in this area. C 局部 represents the construction cost of the local path. The lower the cost, the better. So is used to represent the score of the construction cost. w L1 and w L2 are the weight coefficients of path stability and construction cost respectively, which are used to balance the importance of these two factors in local optimization. This formula helps the algorithm optimize the path layout in each area, just like termites finely constructing secondary paths connecting the nest and resource points after building the main nest, ensuring that the paths in the local area are economical and stable.

[0148] Next, after the middle-level optimization, the algorithm performs micro-optimization, which is the most meticulous layer. Micro-optimization aims to correct minor issues in the path to ensure that the path can flexibly adapt to environmental changes during actual construction. The micro-optimization formula is used:

[0149]

[0150] Among them, F 微观 represents the comprehensive score of micro-optimization, which is used to measure the performance of the path at the micro level. ΔP represents the error correction amount between path points. The smaller the error, the better. Therefore, is used to represent the score of error correction. ΔT represents the time or frequency of path adjustment. The lower the adjustment time and frequency, the better. Therefore, is used to represent the score of adjustment time and frequency. w M1 and w M2 are the weight coefficients of the error correction amount and the adjustment time respectively, which are used to balance the importance of these two factors in micro-optimization. This process is similar to the adjustment of the existing path by termites when facing obstacles or emergencies, ensuring that the path remains efficient and stable at the micro level.

[0151] Finally, through this hierarchical architecture, the optimization process of path planning is effectively decomposed into three levels: global, local, and micro. As Figure 7 shown, the figure shows the comprehensive optimization scores of different levels. Specifically: The global optimization score is 0.036, indicating that the overall performance of the path network layout at the global level is good, the path length is short, and the environmental adaptability is strong. Local optimization: The score is 0.028, indicating that after the path smoothness and construction cost optimization in the local area, the stability of the path has been effectively improved. Micro-optimization: The score is 0.095, which is the highest among the three levels. This shows that through the detail adjustment at the micro level, such as error correction and time optimization, the path network has reached the highest accuracy and flexibility.

[0152] The optimization of each level addresses specific problems to ensure that the path network can achieve the optimal effect at different levels. This hierarchical optimization method enables the algorithm to provide an efficient and stable solution when facing complex terrains and dynamic environments, ultimately ensuring the long-term stability and efficient operation of the water conservancy facility layout.

[0153] In this embodiment, S4 is:

[0154] S41. Consolidate the application of the hierarchical optimization architecture

[0155] First, S41 will continuously optimize the total length and environmental adaptability of the path network using the global optimization formula to ensure that the global path layout always remains in an optimal state. At the same time, to maintain the optimal state of the global path layout, a dynamic adjustment mechanism is proposed to adjust the weights in the global optimization formula according to real-time environmental data. The specific implementation steps are as follows:

[0156] Calculate the score of the current path network using the global optimization formula, and compare the current score with the historical best score. If it is found that the score has decreased, adjust the weights in the following ways:

[0157] (1) Increase the weight w G1 to shorten the path length L 总 .

[0158] (2) Increase the weight w G2 to enhance the environmental adaptability E.

[0159] (3) After adjusting the weights, recalculate the optimal path and regenerate the path network layout.

[0160] Suppose in a situation where a heavy rain causes a change in water flow, the total length L of the path network 总 increases from 120 meters to 140 meters, and the environmental adaptability E decreases from 0.9 to 0.8. By adjusting the weights w G1 and w G2 , recalculate the path so that the total length of the new path is reduced to 130 meters and the environmental adaptability is increased to 0.85.

[0161] As Figure 8 shown in the graph of the change trends of the total path length and environmental adaptability, the graph shows the changes in the total path length and environmental adaptability before and after adjustment. Before adjustment, the total path length was 120 meters and the environmental adaptability was 0.90. However, after an environmental change (such as a heavy rain), the total path length increased to 140 meters and the environmental adaptability decreased to 0.80. Through the systematic dynamic adjustment, the global optimization algorithm reallocated the weights, resulting in the total path length being shortened to 130 meters after adjustment, while the environmental adaptability recovered to 0.85. This result shows that the system can quickly adjust the path layout after an environmental change to ensure that the overall efficiency and stability of the path network are maintained.

[0162] Then, in local path optimization, by balancing the smoothness and construction cost of the path, the path layout is optimized in terms of economy and stability. The implementation steps are as follows:

[0163] (1) Calculate the smoothness S 局部 and construction cost C 局部 of the local path using the local optimization formula.

[0164] (2) Adjust the weights w L1 and w L2 to ensure a smooth path and the lowest construction cost.

[0165] (3) Among multiple alternative paths, select the path with the highest comprehensive score.

[0166] Suppose there are two alternative paths in a local area:

[0167] Path A: Smoothness S 局部 = 0.85S_local = 0.85, Construction cost C 局部 = 100

[0168] Path B: Smoothness S 局部 = 0.90, Construction cost C 局部 = 120

[0169] By adjusting the weights w L1 and w L2 , the score of Path A is 0.034, while the score of Path B is 0.032. Then, select Path A as the optimal path.

[0170] As Figure 9 shown in the score comparison chart of Path A and Path B, the chart shows the difference in the comprehensive scores of the two paths. The comprehensive score of Path A is 0.034, slightly higher than 0.032 of Path B. Although Path B may perform excellently in some aspects, Path A performs more balancedly in terms of comprehensive factors such as smoothness and construction cost. Therefore, it is selected as the preferred path by the algorithm. This indicates that in the multi-objective optimization process, the path with a higher comprehensive score can better meet the overall requirements of the system and is finally selected as the actual construction path.

[0171] Next, at the micro-optimization level, S41 will ensure that the path can be corrected in real time at the micro level during actual construction. The implementation steps are as follows:

[0172] (1) Real-time error monitoring: Through positioning devices such as GPS, monitor the path deviation ΔP and adjustment time ΔT in real time.

[0173] (2) Calculate the correction score: Use the micro-optimization formula to calculate the correction score of the current path.

[0174] (3) Path correction: If the correction score drops, adjust the path to avoid construction errors and maintain the accuracy of the path network.

[0175] Through these corrections, S41 ensures that the path network can remain efficient and stable at the micro level, especially being able to flexibly respond to emergencies in actual operations. For example, during the construction of a certain section of the path, a path deviation of ΔP == 0.10 was monitored, and the adjustment time was ΔT == 5 seconds. After adjustment, the deviation of the new path was reduced to ΔP = 0.05, and the adjustment time was reduced to ΔT = 3 seconds, and the path score was significantly improved.

[0176] As Figure 10 shown in the trend chart of path deviation and adjustment time, the chart shows the changes before, during, and after the path correction by the system. It can be seen that before the correction, the path deviation was large and the adjustment time was long. As the adjustment progresses, the path deviation gradually decreases, and the adjustment time is also significantly shortened. Finally, after the adjustment, the path deviation almost disappears, and the adjustment time also drops to the minimum value. This indicates that through the real-time correction algorithm, the system can quickly reduce the path deviation and optimize the adjustment time, thus maintaining the efficiency and accuracy of the path network.

[0177] S42: Introduce real-time data feedback and dynamic optimization; the goal of S42 is to ensure that the path optimization at the global, local, and micro levels can flexibly respond to the dynamic changes of the environment through the introduction of a real-time data feedback mechanism, so as to maintain efficient and stable operation in actual applications. This feedback mechanism will make the optimization process form a closed loop, and it can continuously self-adjust and optimize during the construction and operation process.

[0178] First of all, the global path layout will be continuously evaluated based on real-time data feedback. Real-time monitoring will collect environmental data including climate change, geological conditions, water flow velocity, etc., and feed this data back into the global optimization model. The following global dynamic optimization formula is used:

[0179]

[0180] Among them, represents the dynamic comprehensive score of global optimization, which combines the influence of real-time environmental data. L 总 represents the total length of the path network, and the goal is to keep the path as short as possible. E represents the adaptability of the path to environmental changes. ΔE represents the amplitude of the current environmental change, which helps to quickly respond to drastic environmental changes. w G1 、w G2 、w G3 are the weight coefficients of different optimization goals, which determine the importance of these factors in global dynamic optimization.

[0181] Then, in local path optimization and micro path optimization, the real-time data feedback mechanism will further refine the optimization process. For example, in the dynamic adjustment of the local path, the following local dynamic optimization formula will be used:

[0182]

[0183] Among them, represents the dynamic optimization score of the local path. S 局部 represents the stability of the local path, the higher the better. C 局部 represents the construction cost of the local path, and the goal is to reduce the construction cost. ΔS represents the fluctuation of the path stability with time or construction environment changes. w L1 、w L2 、w L3 are the weight coefficients used to balance different optimization objectives.

[0184] At the micro-path level, it will ensure that any path deviation or unexpected situation occurring during the construction process can be quickly adjusted through real-time feedback. The dynamic optimization formula for the micro-path is as follows:

[0185]

[0186] Among them, represents the dynamic optimization score of the micro-path. ΔP represents the error correction amount between path points, the smaller the better. ΔT represents the time or frequency of path adjustment, and the goal is to reduce the adjustment time and frequency. Δ 偏差 represents the real-time path deviation caused by construction or environmental changes. w M1 、w M2 、w M3 are the weight coefficients for balancing the optimization objectives at the micro level.

[0187] Finally, with the integration of multiple iterations and real-time feedback, S42 will construct an adaptive closed-loop optimization. In this, the global, local, and micro-path optimizations can work in coordination, form an effective feedback mechanism, and continuously optimize the path layout in practical applications to ensure that the entire network always remains efficient and stable.

[0188] Such as Figure 11 shown is the diagram of the adaptive closed-loop optimization system. The figure shows the mutual relationship between global optimization, local optimization, and micro-optimization, as well as their connections with the real-time data feedback and dynamic adjustment mechanism. Global optimization is responsible for the planning of the overall path network, local optimization conducts refined adjustments for specific areas, and micro-optimization fine-tunes the path during actual operation. The real-time data feedback mechanism plays a key role in these three levels, ensuring that the system can dynamically adjust the path layout according to real-time data, thus forming an adaptive closed-loop optimization system that continuously optimizes and maintains the efficiency and stability of the path network.

[0189] S43: Simulation and Field Test

[0190] Conduct tests in a simulation environment. Based on the termite behavior feature data extracted in S1 (such as path selection, nest building and path construction, environmental adaptability, and abnormal behavior correction), and using the data integration and mapping completed in S2, verify the application effect of these feature data in the optimization algorithm. Through the multi-objective optimization and hierarchical genetic algorithm constructed in S3, optimize and evaluate the paths under different simulation conditions to ensure the accuracy and operability of the data. During the simulation process, use the hierarchical optimization architecture in S41 to ensure that global optimization, local optimization, and microscopic optimization can operate effectively under conditions of different complexities. Finally, through the real-time data feedback mechanism introduced in S42, dynamically adjust the path layout to ensure that the path network is always in an optimal state.

[0191] The results of the simulation test are shown in Table 5 as follows:

[0192] Table 5

[0193]

[0194]

[0195] Observing the comprehensive scores of each path in the table under different complexity conditions, it can be seen that as the environmental complexity increases, the comprehensive scores of the paths decrease. This indicates that under high-complexity conditions, the increase in path length and the decrease in pheromone concentration have a greater impact on the optimization effect of the path network. Generally speaking, in a low-complexity environment, the comprehensive score of Path 1 is the highest, which is 0.029. And under medium-complexity and high-complexity conditions, Path 1 still maintains a relatively high comprehensive score, although its advantage has weakened. This shows that under different environmental complexities, the performance of Path 1 is relatively stable, and it is frequently selected as the optimal path in the simulation test.

[0196] After being satisfied with the simulation test results, S43 will enter the field test stage. In this stage, it will be deployed in the real water conservancy facility environment to test the performance of the optimization algorithm in actual operation. The field test will verify the following key points:

[0197] (1) Verification of path selection features: The field test will focus on verifying the path selection features extracted in S1, including path length and pheromone concentration. By measuring the distribution of the actual path length and pheromone concentration, verify whether the optimized path conforms to the pattern of the termite's natural selection path.

[0198] (2) Verification of nest building and path construction features: Evaluate the path construction features extracted in S1, verify the connectivity and structural stability of the path network, and ensure that the optimized path can meet the expected stability and construction cost requirements in the field environment.

[0199] (3) Verification of environmental adaptability features: Field tests will verify the environmental adaptability features extracted in S1, evaluate the response ability and adaptability of the path network under different environmental conditions, and ensure that the path can be dynamically adjusted according to real-time environmental data.

[0200] (4) Verification of abnormal behavior and error correction features: Field tests will also evaluate the abnormal behavior and error correction features extracted in S1, and verify whether the algorithm can adjust and correct the path in a timely manner when encountering unforeseen environmental changes, maintaining stability and effectiveness.

[0201] As Figure 12 shown is the heat map of the path target achievement rate under field test conditions. The figure shows the target achievement rates of different paths under two test conditions: sunny and rainy days. It can be seen from the figure that under sunny conditions, the target achievement rate of Path 1 is the highest, reaching 95%, followed by Path 2 and Path 3. Under rainy conditions, the target achievement rates of all paths have decreased. Among them, Path 1 still performs the best, with an achievement rate of 90%. This indicates that Path 1 has high adaptability and stability under different weather conditions, while the performance of Path 3 is relatively poor, especially under rainy conditions, with an achievement rate of only 85%.

[0202] In summary, in the present invention, data on the path selection, construction process, environmental adaptability, and abnormal behavior correction features of termites under different environmental conditions are collected and integrated; the collected data is mapped, cleaned, and normalized to ensure the quality and consistency of the data; a multi-objective optimization and hierarchical genetic algorithm is constructed using the integrated data to ensure the optimization effect of the path network at the global, local, and microscopic levels; a real-time data feedback mechanism is introduced to construct an adaptive closed-loop optimization system. Through simulation and field tests, the effectiveness and reliability of the algorithm are verified, ensuring that it can be flexibly adjusted and the continuous optimization of the path network can be achieved in practical applications.

[0203] Finally, it is verified that the method provided by the present invention effectively improves the efficiency and stability of the layout of water conservancy facilities, providing a new solution for practical engineering applications in complex environments.

[0204] The beneficial effects of the present invention are as follows:

[0205] Strong dynamic adaptability: By combining multi-dimensional feature data of termite behavior with a multi-objective optimization algorithm, the present invention can dynamically adapt to complex terrains and environmental changes, ensuring the optimality and stability of the layout of water conservancy facilities under different conditions.

[0206] Balance between global and local optimization: Through a hierarchical optimization architecture, both the integrity of the global layout and the refined adjustment of local areas are achieved, solving the problem that traditional methods are difficult to balance global and local optimization.

[0207] Efficient path network construction: The present invention utilizes the termite behavior simulation technology in the bio-inspired algorithm to efficiently construct a path network with strong adaptability and good connectivity, significantly improving the efficiency and effect of the layout of water conservancy facilities.

[0208] Adaptive closed-loop optimization system: By introducing a real-time data feedback mechanism to form a closed-loop optimization system, it can adjust the path layout in real time according to the dynamic changes of the environment, ensuring the continuous optimization and stability of water conservancy facilities during long-term operation.

[0209] According to the water conservancy facility layout and termite control method of the embodiments of the present invention, by combining the natural behavior of termites and multi-objective optimization and hierarchical genetic algorithm technology, a more efficient and flexible path optimization layout is achieved, providing an innovative solution for the long-term stable operation of water conservancy facilities and termite control.

[0210] To implement the above embodiments, as Figure 13 shown, in this embodiment, a water conservancy facility layout and termite control system 10 is further provided, including:

[0211] A feature data acquisition module 100, configured to construct a multi-dimensional comprehensive data set based on termite path selection features, nest building and path construction data, environmental adaptability data, and abnormal behavior and error correction data;

[0212] A data preprocessing module 200, configured to perform mapping processing and preprocessing on the multi-dimensional comprehensive data set to obtain preprocessing data for model training;

[0213] A target optimization and hierarchical module 300, configured to set multiple optimization targets based on the multi-dimensional comprehensive data set to obtain a multi-objective optimization formula, and calculate the comprehensive score of each path in the path network according to the multi-objective optimization formula for preliminary optimization of the path; training a hierarchical genetic algorithm model using the preprocessing data to decompose the path planning task into multiple levels of optimization to obtain an optimization result;

[0214] A multi-level optimization and feedback module 400, configured to optimize the total length and environmental adaptability of the path network using a global optimization formula, and adjust the weights in the global optimization formula according to real-time environmental data to perform path correction at multiple levels according to the optimization result.

[0215] According to the water conservancy facility layout and termite control system of the embodiments of the present invention, by combining the natural behavior of termites and multi-objective optimization and hierarchical genetic algorithm technology, a more efficient and flexible path optimization layout is achieved, providing an innovative solution for the long-term stable operation of water conservancy facilities and termite control.

[0216] In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0217] In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of these features. In the description of the present invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise specifically defined.

Claims

1. A method for water conservancy facility layout and termite control, characterized in that: include: Construct a multi-dimensional comprehensive data set based on termite path selection characteristics, nesting and path construction data, environmental adaptability data, abnormal behavior and error correction data; Performing mapping processing and preprocessing on the multi-dimensional comprehensive data set to obtain preprocessed data for model training; Based on the multi-dimensional comprehensive data set, multiple optimization objectives are set to obtain a multi-objective optimization formula, and the comprehensive score of each path in the path network is calculated according to the multi-objective optimization formula to perform preliminary optimization of the path; the hierarchical genetic algorithm model is trained using the preprocessed data to decompose the path planning task into multiple levels of optimization to obtain an optimization result; The total length and environmental adaptability of the path network are optimized using a global optimization formula, and the weights in the global optimization formula are adjusted according to real-time environmental data to perform multiple levels of path correction according to the optimization results.

2. The method according to claim 1, characterized in that The termite path selection characteristics include: The foraging behavior of termites in the natural environment is simulated and the path length is calculated. The calculation formula of the path length is: Where L represents the total length of the path chosen by the termite, d(x i ,x i+1 ) represents the adjacent point x on the path i and x i+1 The Euclidean distance between The pheromone concentration released by termites on the path is measured, and the calculation formula of pheromone concentration C is as follows: Among them, C represents the sum of pheromone concentrations at all points on the path, c(x i ) represents the point x on the path i The local pheromone concentration; To evaluate the smoothness of the path, the calculation formula of the path smoothness S is as follows: Among them, S represents the smoothness of the path, θ i-1,i,i+1 Represents the midpoint x of the path i-1 ,x i+1 ,x i The angle between them, dist(x i-1 ,x i+1 ) represents x i-1 and x i+1 The distance between Integrate all collected path lengths L, pheromone concentrations C, and path smoothness S into the evaluation function to comprehensively evaluate the optimization effect of the path. The evaluation function is expressed as: Among them, w1 and w2 are the importance weight coefficients that measure the balance between pheromone concentration and path length and path smoothness, respectively.

3. The method according to claim 1, characterized in that The nesting and path building data include: In the experimental environment, multiple nest locations and connecting paths are set. In this process, the path selection function is used to calculate the selection probability of each path point: Among them, P 选择 represents the probability of selecting a path point, d represents the distance between the path point and the current nest or other paths, C represents the pheromone concentration of the path point, E represents the environmental characteristics, and α, β, and γ are the weight coefficients of the balance factors; The order of building the path for tracking termites is selected based on the path points. By observing the behavior of termites, the construction time and order of each path point are recorded in turn. Based on the connectivity between termites’ path points, the path connectivity formula is used: Among them, F 连接 represents the optimization score of path connection, P 选择 represents the probability of selecting a path point, l 连接强度 It is an indicator to measure the closeness of the connection between path points; Integrate all collected data to build a comprehensive prediction model; wherein the comprehensive prediction model includes weight adjustment of path selection, global and local balance, connectivity optimization of path network and multiple iterative optimization; Generate an optimized path network structure based on a comprehensive prediction model.

4. The method according to claim 1, characterized in that: The environmental adaptability data include: Different environmental conditions were set in the experimental environment to simulate the complex terrain and dynamic environment that termites may encounter. The path selection, nesting behavior and adaptive response of termites under these conditions were recorded in real time. The adaptive score formula was used to simulate different terrain and dynamic environments: Among them, A 环境 represents the environmental adaptability score, T k Represents temperature, H k Represents humidity, L k Represents the light intensity, w T ,w H ,w L are the weight coefficients of these factors respectively; Use the path selection probability formula to calculate the termites' ability to adjust their paths when environmental conditions change: Among them, P 选择 represents the probability of selecting a path point, d represents the distance between the path point and the current nest or other path points, C represents the pheromone concentration of the path point, E represents the environmental characteristics, ΔE represents the influencing factor of environmental changes, and α, β, γ, δ are weight coefficients; Based on the termites' responses to different environmental conditions, a dynamic adjustment mechanism is designed using a dynamic optimization formula: Among them, F 动态 Represents the path score after dynamic optimization, A 环境 represents the environmental adaptability score, L represents the path length, S represents the path smoothness, θ represents the average angle between path points, and w1, w2, and w3 are the weight coefficients of each factor; The dynamic adjustment mechanism is tested in a simulated complex environment to evaluate the improvement effect on path network stability and efficiency through the environmental adaptive network optimization formula: Among them, F 网络优化 Represents the optimization score F of the entire network 动态i represents the dynamic optimization score of the i-th path point, R i represents the robustness of the path network, w 动态 and w 鲁棒性 are the importance weight coefficients of dynamic adjustment and robustness, respectively.

5. The method according to claim 1, characterized in that: The abnormal behavior and error correction data include: Setting up obstacles or introducing environmental changes in the experimental environment to induce termites to make errors or abnormal actions during path building, and classifying abnormal behaviors based on the termites’ responses; For path deviation, the model is to strengthen the pheromone concentration or adjust the path selection strategy to correct the deviated path; for pheromone anomalies, it is modeled as a strategy to redistribute pheromones; and the correction strategy is quantified as the formula through the correction effect score formula F 修正 To evaluate the effectiveness of each strategy; the modified effect score formula is as follows: Among them, F 修正 Represents the effect score of error correction, T l Represents the type of abnormal behavior, S l Represents the severity of the error, R l represents the time required for correction, w 类型 ,w 严重性 ,w 修正时间 are the weight coefficients of the factors respectively; Evaluate and adjust the path correction effect in real time through the correction formula: R 修正 =α*F 修正 +β*A 环境 Among them, R 修正 represents the corrected path effect, F 修正 is the correction effect score, A 环境 is the environmental adaptability score, α and β are the weight coefficients of these factors; The performance of the algorithm in a complex environment is tested based on a comprehensive modified optimization formula: Among them, F 综合修正 Represents the final score of comprehensive correction, R 修正m represents the correction effect of the mth path point, T m represents the time required for correction, w 修正 and w 时间 are the importance weight coefficients of correction effect and time respectively.

6. The method according to claim 1, characterized in that Mapping and preprocessing the multi-dimensional comprehensive data set to obtain preprocessed data for model training, including: Mapping multi-dimensional comprehensive data sets to transform complex behavioral characteristics into numerical forms for model processing, wherein the data processed by the mapping include path selection, construction process, environmental impact, and error correction behavior; After the mapping process is completed, the data is cleaned and normalized; After completing data mapping, cleaning and normalization, the processed data are integrated into a unified feature database.

7. The method according to claim 1, characterized in that Based on the multi-dimensional comprehensive data set, multiple optimization objectives are set to obtain a multi-objective optimization formula, and the comprehensive score of each path in the path network is calculated according to the multi-objective optimization formula to perform preliminary optimization of the path; The preprocessed data is used to train a hierarchical genetic algorithm model to decompose the path planning task into multiple levels of optimization to obtain optimization results, including: S31, setting multiple optimization goals based on the multi-dimensional comprehensive data to comprehensively optimize various aspects of the path network; wherein the multiple optimization goals include the shortest path, the minimum construction cost, the maximum pheromone coverage rate and the highest structural stability; A multi-objective optimization formula is determined based on the multiple optimization objectives to evaluate and quantify the comprehensive performance of each path: Among them: F 多目标 represents the comprehensive optimization score of the path, L i represents the length of the i-th path, C i represents the pheromone concentration of the ith path, S i represents the structural stability or construction cost of the i-th path, w L ,w C ,w S are the weight coefficients for path length, pheromone concentration, and structural stability, respectively; Through multiple simulation runs, the optimization effect is verified under different environmental conditions. In each simulation run, different path networks are generated according to the preset conditions, and the comprehensive score of each path is calculated according to the multi-objective optimization formula: F 路径 =w L *L1+w C *C+w S *S Among them: F 路径 represents the comprehensive score of the path, L represents the path length, C represents the pheromone concentration of the path, S represents the structural stability or construction cost of the path, and w L 、w C 、w S are the weight coefficients for path length, pheromone concentration, and structural stability, respectively; S32, construct a hierarchical genetic algorithm model to decompose the path planning task into three levels of optimization: global, local, and micro. Each level processes a specific optimization goal, including: In the top-level optimization, the global layout is planned, and the global optimization formula is: Among them, F 全局 Represents the comprehensive score of global optimization, L 总 represents the total length of the path network, E represents the adaptability of the path to environmental changes, and w G1 and w G2 are the weight coefficients of path length and environmental adaptability respectively; After the global path layout is determined, the local optimization formula is performed: Among them, F 局部 Represents the comprehensive score of local optimization, S 局部 represents the stability of the local path, C 局部 represents the construction cost of the local path, w L1 and w L2 are the weight coefficients of path stability and construction cost, respectively; After the mid-level optimization, the micro-optimization formula is performed: Among them, F 微观 represents the comprehensive score of micro-optimization, ΔP represents the error correction between path points, ΔT represents the time or frequency of path adjustment, and w M1 and w M2 are the weight coefficients of error correction amount and adjustment time respectively.

8. The method according to claim 7, characterized in that The total length and environmental adaptability of the path network are optimized using a global optimization formula, and the weights in the global optimization formula are adjusted according to real-time environmental data to perform multiple levels of path correction according to the optimization results, including: S41, using a global optimization formula to continuously optimize the total length and environmental adaptability of the path network, and adjusting the weights in the global optimization formula according to real-time environmental data, including: The global optimization formula is used to calculate the score of the current path network, and the current score is compared with the best score in history; if the score is found to have dropped, the weight is adjusted in the following way: Increase the weight w G1 To shorten the path length L 总 ; Increase weight w G2 To enhance environmental adaptability E; after adjusting the weight, recalculate the optimal path and regenerate the path network layout; In local path optimization, the path layout is optimized in terms of economy and stability by balancing the smoothness of the path and the construction cost: the smoothness S of the local path is calculated using the local optimization formula 局部 and construction cost C 局部 ; Adjust the weight w according to the calculation results L1 and w L2 ; Among multiple alternative paths, select the path with the highest comprehensive score; At the micro-optimization level, ensure that the path is corrected in real time at the micro level during actual construction: monitor the path deviation ΔP and adjustment time ΔT in real time, and use the micro-optimization formula to calculate the correction score of the current path; S42, introduces a real-time data feedback mechanism to optimize the path at the global, local and micro levels, including: Real-time monitoring collects real-time data, including climate change, geological conditions, and water flow speed, and feeds the real-time data into the global optimization model; using the following global dynamic optimization formula: in, Represents the dynamic comprehensive score of global optimization, L 总 represents the total length of the path network, E represents the adaptability of the path to environmental changes, ΔE represents the magnitude of the current environmental change, and w G1 、w G2 、w G3 is the weight coefficient of different optimization objectives; In the dynamic adjustment of the local path, the local dynamic optimization formula is used: in, Represents the dynamic optimization score of the local path, S 局部 represents the stability of the local path, C 局部 represents the construction cost of the local path, ΔS represents the fluctuation of path stability over time or changes in the construction environment, and w L1 、w L2 、w L3 is the weight coefficient used to balance different optimization objectives; The dynamic optimization formula of the micro-path is as follows: in, represents the dynamic optimization score of the micro-path, ΔP represents the error correction between path points, ΔT represents the time or frequency of path adjustment, and Δ 偏差 represents the real-time path deviation caused by construction or environmental changes, w M1 、w M2 、w M3 is the weight coefficient for balancing the optimization objective at the micro level; S43, obtaining simulation test results, and performing field tests based on the simulation test results.

9. A water conservancy facility layout and termite control system, characterized in that: include: The feature data acquisition module is used to construct a multi-dimensional comprehensive data set based on termite path selection characteristics, nesting and path construction data, environmental adaptability data, abnormal behavior and error correction data; A data preprocessing module, used for mapping and preprocessing the multi-dimensional comprehensive data set to obtain preprocessed data for model training; The target optimization and hierarchical module is used to set multiple optimization targets based on the multi-dimensional comprehensive data set to obtain a multi-objective optimization formula, and calculate the comprehensive score of each path in the path network according to the multi-objective optimization formula to perform preliminary optimization of the path; use the pre-processed data to train a hierarchical genetic algorithm model to decompose the path planning task into multiple levels of optimization to obtain an optimization result; The multi-layer optimization and feedback module is used to optimize the total length and environmental adaptability of the path network using a global optimization formula, and adjust the weights in the global optimization formula according to real-time environmental data to perform multi-level path corrections according to the optimization results.

Images