Garden automatic inspection method and system based on path planning

Abstract

The present invention relates to the technical field of intelligent inspection and path planning, and in particular to a garden automatic inspection method and system based on path planning, which integrates self-organizing mapping, fractal geometry analysis, quantum heuristic optimization, hybrid fuzzy neural network and adaptive genetic algorithm, and realizes intelligent planning and real-time adjustment of garden inspection paths; the system generates a three-dimensional model by collecting terrain, vegetation and environmental data of the garden, and identifies functional areas and sets inspection parameters based on the model; generates an initial optimized path through an intelligent algorithm, monitors environmental changes in real time during the inspection process, and adjusts the path; dynamically allocates and executes inspection tasks, and continuously optimizes the path planning strategy through feedback data, which significantly improves inspection efficiency and resource utilization; the present invention can flexibly respond to environmental changes in a complex garden environment and realize efficient automatic inspection task execution.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent inspection and path planning, and in particular to a garden automatic inspection method and system based on path planning. Background Art

[0002] As gardens expand in size and complexity, traditional manual inspection methods can no longer meet the needs of efficient and accurate inspections. Automated inspection technology can achieve efficient resource management and maintenance in large-scale gardens through intelligent path planning and automatic execution. The key to automated inspection technology is how to design an intelligent path planning system that can respond to environmental changes in real time and efficiently complete inspection tasks.

[0003] The existing automatic inspection technology (Chinese invention patent, publication number: CN116862092B, name: An adaptive path planning module and method for an automatic inspection device) usually adopts a path planning method based on distance priority, and forms an inspection route by calculating the shortest distance between inspection items. This method has several major defects in practical applications: first, the technology relies heavily on pre-set inspection routes and lacks the ability to adjust dynamically in real time, resulting in slow response when facing emergencies or environmental changes; second, the existing technology mostly adopts a simple distance calculation method, ignoring the influence of complex terrain and environmental factors, which easily leads to low inspection efficiency; finally, the allocation of inspection tasks is mainly based on a fixed average time, lacking consideration of equipment status and real-time environment, resulting in uneven resource allocation and increased energy consumption. Summary of the invention

[0004] In view of the many problems existing in the above-mentioned prior art, the present invention provides an automatic garden inspection method and system based on path planning. The present invention realizes intelligent planning and real-time adjustment of garden inspection paths by integrating multiple intelligent algorithms, including self-organizing mapping, fractal geometry analysis, quantum heuristic optimization, hybrid fuzzy neural network and adaptive genetic algorithm. The system can dynamically generate the optimal inspection path according to the three-dimensional model of the garden and environmental change information, and make adjustments in emergency situations. Through intelligent task allocation and feedback mechanism, the system can not only efficiently complete the inspection tasks, but also continuously optimize its own path planning strategy over time, thereby improving inspection efficiency and resource utilization.

[0005] A garden automatic inspection method based on path planning includes the following steps:

[0006] The terrain data and vegetation distribution data of the garden are collected through lidar, high-definition cameras and global positioning system equipment to generate a three-dimensional model of the garden. The real-time environmental condition data is collected through the environmental monitoring sensor system to generate comprehensive environmental data for path planning.

[0007] Analyze the comprehensive environmental data and the three-dimensional model, identify the functional areas of the garden through a machine learning algorithm, set inspection priorities and inspection parameters for each functional area, and generate functional area data and inspection parameter data for path planning;

[0008] Using a self-organizing map algorithm to determine path nodes based on the three-dimensional model and functional area data, optimizing the path morphology in combination with fractal geometry analysis, and globally optimizing the path through a quantum heuristic optimization algorithm to generate initial optimized path data for initial path planning;

[0009] During the inspection process, the initial optimized path data is usually used for inspection, and the dynamic monitoring sensor system is used to obtain environmental change information in real time. When an emergency situation or path obstruction is detected, the path decision is adjusted through a hybrid fuzzy neural network, and an adaptive genetic algorithm is used to optimize the path to generate the final adjusted path data for inspection path adjustment;

[0010] According to the initial optimized path data or the final adjusted path data generated in an emergency situation and the device status data, the inspection tasks are dynamically allocated and executed, and the inspection task execution feedback data is collected to optimize the path planning and inspection task allocation strategy.

[0011] Preferably, the environmental monitoring sensor system includes a temperature sensor, a humidity sensor and a light sensor, which are used to detect the environmental conditions of the garden and provide static environmental information to generate comprehensive environmental data.

[0012] Preferably, the machine learning algorithm includes a support vector machine algorithm and a random forest algorithm, which identifies the functional areas in the garden and determines the inspection priority and inspection parameters of each area by analyzing the features and three-dimensional models in the comprehensive environmental data.

[0013] Preferably, the self-organizing map algorithm is used to generate path node data based on the three-dimensional model and functional area data, and determine the node layout of the inspection path by calculating the topological relationship between the path nodes.

[0014] Preferably, the fractal geometry analysis is used to evaluate the complexity and curvature of the path, to adjust the path shape to obtain the optimal path shape, and to generate optimized path data.

[0015] Preferably, the quantum heuristic optimization algorithm uses quantum annealing technology to perform global optimization on the optimization path data to find the optimal solution of the path and generate initial optimization path data, wherein the optimization process is represented by the following formula:

[0016]

[0017] in, represents the total optimization cost of the path; Indicates The weight of each path; Indicates The optimization coefficient of each path; Indicates the number of paths.

[0018] Preferably, the hybrid fuzzy neural network performs path decision adjustment according to the environmental change information obtained by the dynamic monitoring sensor system during the inspection process to cope with emergencies or path obstructions and generate path adjustment decision data.

[0019] Preferably, the adaptive genetic algorithm optimizes the path adjustment decision data through dynamic crossover rate and mutation rate, and the formula is as follows:

[0020]

[0021]

[0022] in, and are the crossover rate and mutation rate, respectively. and are the minimum and maximum values ​​of the crossover rate, and are the minimum and maximum values ​​of the mutation rate, is the current fitness, and are the maximum fitness and average fitness in the current generation respectively.

[0023] Preferably, the final adjustment path data for inspection path adjustment is generated by dynamically adjusting the crossover rate and the mutation rate.

[0024] A system for implementing the automatic garden inspection method based on path planning, comprising:

[0025] A data acquisition module, including a laser radar, a high-definition camera and a global positioning system device, for collecting terrain data and vegetation distribution data of the garden to generate a three-dimensional model of the garden; and an environmental monitoring sensor system, for collecting real-time environmental condition data to generate comprehensive environmental data, the environmental monitoring sensor system including a temperature sensor, a humidity sensor and a light sensor;

[0026] A data analysis module is used to analyze the comprehensive environmental data and the three-dimensional model, identify the functional areas of the garden through a support vector machine algorithm and a random forest algorithm, set the inspection priority and inspection parameters of each functional area, and generate functional area data and inspection parameter data;

[0027] A path planning module, including a self-organizing map algorithm module, a fractal geometry analysis module and a quantum heuristic optimization algorithm module, for generating initial optimized path data based on the three-dimensional model and the functional area data;

[0028] A dynamic path adjustment module, including a dynamic monitoring sensor system, is used to obtain environmental change information in real time, and uses a hybrid fuzzy neural network module to adjust the path decision, and generates the final adjusted path data through an adaptive genetic algorithm module;

[0029] The task allocation and execution module is used to dynamically allocate and execute inspection tasks according to the initial optimized path data or the final adjusted path data, and collect inspection task execution feedback data to optimize path planning and inspection task allocation strategies.

[0030] Compared with the prior art, the advantages and beneficial effects of the present invention are:

[0031] The present invention realizes accurate modeling of complex garden terrain and environment through self-organizing mapping algorithm and fractal geometry analysis, and optimizes the inspection path morphology;

[0032] The present invention realizes the global optimal solution search of the path through the quantum heuristic optimization algorithm, thereby improving the efficiency of the inspection path;

[0033] The present invention realizes real-time response and path adjustment to emergencies by mixing fuzzy neural network and adaptive genetic algorithm, and enhances the flexibility and adaptability of the system.

[0034] The present invention realizes the reasonable configuration of inspection resources and the continuous optimization of path planning through dynamic task allocation and feedback mechanism, thereby improving the overall efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of the process of the present invention;

[0036] Figure 2 It is a flow chart of data collection and modeling in the present invention;

[0037] Figure 3 This is a flow chart of intelligent path planning in the present invention;

[0038] Figure 4 This is a real-time path adjustment flow chart of the present invention;

[0039] Figure 5 A flowchart for task allocation and feedback optimization in the present invention;

[0040] Figure 6 It is a structural block diagram of the system of the present invention. DETAILED DESCRIPTION

[0041] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is obvious that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0042] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0043] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0044] like Figure 1-2 As shown, a garden automatic inspection method based on path planning includes the following steps:

[0045] The terrain data and vegetation distribution data of the garden are collected through lidar, high-definition cameras and global positioning system equipment to generate a three-dimensional model of the garden. The real-time environmental condition data is collected through the environmental monitoring sensor system to generate comprehensive environmental data for path planning.

[0046] Preferably, the environmental monitoring sensor system includes a temperature sensor, a humidity sensor and a light sensor, which are used to detect the environmental conditions of the garden and provide static environmental information to generate comprehensive environmental data.

[0047] In the present invention, data collection is the basis for the operation of the entire system. Through effective data collection, necessary basic information can be provided for subsequent regional identification, path planning and dynamic adjustment. The system first collects the terrain and vegetation distribution of the garden through laser radar, high-definition camera and global positioning system equipment. The laser radar uses the principle of laser ranging to measure the relative distance between objects and terrain in the garden by emitting and receiving laser beams, thereby constructing a detailed three-dimensional model. This three-dimensional model provides an accurate geometric basis for identifying terrain features, evaluating vegetation coverage, and detecting potential inspection obstacles. In addition, the high-definition camera captures the growth status of plants and the overall environment of the garden by taking high-resolution images, providing intuitive visual information for inspection. The global positioning system equipment is used to obtain accurate geographic location data to ensure that the path planning has accurate geographic reference.

[0048] Through the environmental monitoring sensor system, the system can collect real-time environmental condition data and generate comprehensive environmental data for path planning. The system preferably includes temperature sensors, humidity sensors and light sensors. The temperature sensor monitors temperature changes in the garden and can be used to analyze the changing trend of the plant growth environment; the humidity sensor monitors air humidity, which is critical for the assessment of plant health status; the light sensor records light conditions and helps identify light distribution and changes in the garden. These sensors together generate comprehensive environmental data, which can provide the inspection system with detailed information on the current environment and ensure the adaptability and accuracy of the inspection process.

[0049] Through this multi-level data collection method, the system can create a detailed and dynamic environmental model, providing reliable data support for subsequent path planning and dynamic adjustment. For example, the three-dimensional model generated by the lidar can help identify path obstacles in the garden and provide a geometric basis for the selection of path nodes. The real-time environmental conditions provided by the environmental monitoring sensor system can monitor the dynamic changes of the environment during the inspection process, so that the path adjustment module can respond to sudden changes in the external environment in a timely manner, thereby improving the efficiency and accuracy of the inspection. In general, the combination of these collection methods enables the garden automatic inspection system to operate flexibly in a complex garden environment, ensuring comprehensive coverage and efficient execution of inspection tasks. The present invention not only improves the accuracy and responsiveness of the inspection, but also optimizes the utilization of resources, making garden management more intelligent and efficient.

[0050] Analyze the comprehensive environmental data and the three-dimensional model, identify the functional areas of the garden through a machine learning algorithm, set inspection priorities and inspection parameters for each functional area, and generate functional area data and inspection parameter data for path planning;

[0051] Preferably, the machine learning algorithm includes a support vector machine algorithm and a random forest algorithm, which identifies the functional areas in the garden and determines the inspection priority and inspection parameters of each area by analyzing the features and three-dimensional models in the comprehensive environmental data.

[0052] In the present invention, data analysis is one of the core functions of the system, which directly affects the planning of inspection paths and the execution effect of tasks. Through the analysis of comprehensive environmental data and three-dimensional models, the system can effectively identify the functional areas in the garden and set corresponding inspection priorities and parameters for each area. This process relies on advanced machine learning algorithms, such as support vector machine algorithms and random forest algorithms, which are used to extract valuable information from large amounts of data. The support vector machine algorithm divides data into different categories by establishing hyperplanes, which is suitable for identifying the boundaries of different functional areas in the garden, for example, distinguishing lawns, flower beds, woodlands and other areas. The random forest algorithm improves the accuracy and robustness of recognition by creating multiple decision trees and integrating their output results, especially when dealing with complex and changeable garden environments.

[0053] By analyzing the temperature, humidity, light intensity and other characteristics in the comprehensive environmental data and the three-dimensional model of the garden, the system can identify the characteristics and needs of each functional area. For example, when identifying a flower bed area, the system can determine the health status of the area and the required inspection frequency by detecting the lighting conditions, soil moisture and terrain shape of the area, combined with historical data. In addition, the system can assign different inspection priorities to different functional areas based on the importance and specificity of the area. For example, low-lying areas that are prone to waterlogging or areas with valuable vegetation may require a higher inspection priority to ensure that these areas receive timely attention under severe weather conditions.

[0054] Through this data analysis process, the system generates functional area data and inspection parameter data for path planning. These data provide the basis for inspection path planning, ensuring that the path can cover all important areas and take into account the impact of environmental changes. In effect, this not only improves the efficiency and accuracy of inspections, but also makes garden management more proactive and refined, thereby reducing resource waste and optimizing the allocation of inspection resources. This intelligent analysis and planning capability enables the garden automatic inspection system to complete tasks in an efficient and reliable manner in a complex and changing garden environment, ensuring the health and safety of the garden.

[0055] Using a self-organizing map algorithm to determine path nodes based on the three-dimensional model and functional area data, optimizing the path morphology in combination with fractal geometry analysis, and globally optimizing the path through a quantum heuristic optimization algorithm to generate initial optimized path data for initial path planning;

[0056] like Figure 3As shown, in the present invention, path planning is an important step for the system to achieve efficient inspection. This step comprehensively optimizes the inspection path by combining the self-organizing map algorithm, fractal geometry analysis and quantum heuristic optimization algorithm, thereby generating initial optimized path data for initial path planning. First, the self-organizing map algorithm (SOM) is an unsupervised neural network algorithm suitable for dimensionality reduction and cluster analysis of high-dimensional data. In the present invention, SOM is used to analyze the three-dimensional model and functional area data of the garden to determine the path nodes. These path nodes are the basis of the inspection route and represent important geographical locations or checkpoints that need to be covered during the inspection process.

[0057] After determining the path nodes, the system optimizes the path shape through fractal geometry analysis. Fractal geometry analysis is a mathematical tool for studying complex morphological structures, which can identify and optimize the natural complexity and curvature of the path. In garden inspections, this analysis helps ensure that the path can effectively avoid obstacles and find the best route in complex terrain. For example, when the path needs to pass through a dense forest, fractal geometry analysis can help plan a short and safe path without unnecessary twists and turns.

[0058] In order to optimize the path globally, the system uses a quantum-inspired optimization algorithm for final adjustments. Quantum-inspired optimization uses the principles of quantum computing to find the optimal solution by simulating the quantum annealing process. This method can quickly traverse all possible path combinations and select the inspection route with the lowest cost. In a complex garden environment, the application of the quantum-inspired optimization algorithm enables the inspection path planning to quickly converge to the optimal solution, minimizing travel time and energy consumption. For example, for a complex inspection task in a large garden, the system can dynamically adjust the order and route of the path according to the priority and inspection parameters of each functional area to ensure the comprehensiveness and efficiency of the inspection.

[0059] Through this series of algorithm combinations, the generated initial optimized path data not only ensures the scientificity and rationality of path planning, but also improves the adaptability and stability of the system under different environmental conditions. Overall, this method significantly improves the execution efficiency and accuracy of the garden automatic inspection system through intelligent path planning and optimization, reduces resource waste, and improves the intelligent level of garden management.

[0060] Preferably, the self-organizing map algorithm is used to generate path node data based on the three-dimensional model and functional area data, and determine the node layout of the inspection path by calculating the topological relationship between the path nodes.

[0061] In the present invention, the self-organizing map algorithm (SOM) is a key tool for generating path node data. SOM is an unsupervised learning algorithm that is good at extracting features from high-dimensional data and performing dimensionality reduction processing. In the present invention, SOM generates node data of the inspection path based on the three-dimensional model of the garden and the functional area data. First, the three-dimensional model provides detailed terrain information of the garden, including ups and downs, vegetation distribution, and path obstacles, while the functional area data describes the characteristics and inspection priorities of different areas in the garden. By integrating these two types of data, SOM can identify important nodes that need to be covered during the inspection process.

[0062] In actual applications, SOM automatically identifies and classifies path nodes through a competitive learning mechanism. This process is similar to the working principle of artificial neural networks: the algorithm will cluster similar data together based on the characteristics of the input data, thereby forming several "node clusters" in a high-dimensional space. These node clusters represent key points that need to be visited during the inspection process, such as irrigation equipment, plant communities, and areas with large terrain changes. After identifying these nodes, SOM further calculates the topological relationship between the nodes to determine the layout of the inspection path. This process ensures that the path planning can cover all important areas while maintaining the coherence and efficiency of the route.

[0063] In this way, SOM not only effectively simplifies the complexity of data processing, but also ensures the scientific distribution of path nodes. For example, in a more complex garden area, SOM can design a route for the inspection path that can avoid natural obstacles and efficiently cover all important areas by identifying and classifying different terrain features. This path layout is crucial for automatic garden inspections because it determines the travel efficiency and working range of the inspection equipment.

[0064] In general, the application of the self-organizing map algorithm in the present invention not only improves the accuracy of path node identification, but also significantly improves the working efficiency and accuracy of the garden automatic inspection system through intelligent path layout planning. This method ensures comprehensive coverage and optimization of the inspection process, making garden management more intelligent and efficient.

[0065] Preferably, the fractal geometry analysis is used to evaluate the complexity and curvature of the path, to adjust the path shape to obtain the optimal path shape, and to generate optimized path data.

[0066] In the present invention, fractal geometry analysis is used as a powerful mathematical tool to evaluate and optimize the complexity and curvature of the inspection path. Fractal geometry is a branch of mathematics that studies irregular shapes and complex structures, and is suitable for analyzing and describing common complex forms in nature. In the application of automatic garden inspection, fractal geometry analysis can help identify the natural characteristics of the path and make fine adjustments to it to obtain the optimal path shape.

[0067] The core of fractal geometry analysis lies in its ability to evaluate path complexity. In garden inspections, paths usually need to pass through various complex terrains and vegetation. The complexity and curvature of the path directly affect the efficiency and energy consumption of the inspection. By using mathematical indicators such as fractal dimension and path curvature, fractal geometry analysis can quantify the complexity of the path. These indicators help identify unnecessary twists and turns in the path and provide data support for optimizing the path.

[0068] By analyzing the fractal characteristics of the path, the system can automatically adjust the path shape to reduce unnecessary complexity and curvature. This adjustment process includes smoothing and simplifying the path to ensure that the inspection equipment can move more efficiently on the path. For example, in a dense forest, the original path may cause inspection speed to decrease and energy consumption to increase due to too many turns and detours. Fractal geometry analysis can identify these overly complex areas and generate alternative paths to make the inspection route more intuitive and efficient.

[0069] Through fractal geometry analysis, the generated optimized path data can significantly improve the smoothness and efficiency of the inspection process. This process ensures that the inspection path can effectively avoid obstacles and maintain an efficient travel speed in complex terrain. Applied to the garden automatic inspection system, this path optimization strategy not only reduces the inspection time, but also reduces the energy consumption of the equipment and improves the overall inspection efficiency. In summary, fractal geometry analysis provides strong support for path optimization through its unique perspective and method, ensuring that garden inspections can operate flexibly and efficiently complete tasks in complex environments.

[0070] Preferably, the quantum heuristic optimization algorithm uses quantum annealing technology to perform global optimization on the optimization path data to find the optimal solution of the path and generate initial optimization path data, wherein the optimization process is represented by the following formula:

[0071]

[0072] in, represents the total optimization cost of the path; Indicates The weight of each path; Indicates The optimization coefficient of each path; Indicates the number of paths.

[0073] In the present invention, the quantum heuristic optimization algorithm is an innovative method for path optimization, the core of which is to use quantum annealing technology to achieve global optimization. Quantum annealing is a bionic algorithm that imitates the annealing process of quantum systems and finds the global optimal solution by probabilistic sampling of the problem solution space. In the automatic garden inspection, the application of the quantum heuristic optimization algorithm enables the system to quickly find the optimal inspection route in a complex path combination, thereby improving the inspection efficiency and accuracy.

[0074] The quantum heuristic optimization algorithm in this invention evaluates the advantages and disadvantages of different paths by quantifying the total optimization cost of the path. Formula This process is described in represents the total optimization cost of the path, It is The weight of each path represents the relative importance or priority of the path. It is The optimization coefficient of each path reflects the performance of the path during the optimization process. By adjusting these weights and optimization coefficients, the quantum heuristic optimization algorithm can quickly identify the optimal solution among a large number of path choices.

[0075] In practice, the algorithm performs a global search for paths in the garden, calculates the total cost of each path under different weights and optimization coefficients by simulating the transcendence and superposition of quantum states, and selects the path with the lowest cost as the optimal path. This method is particularly suitable for dealing with complex path combinations in the garden. For example, when it is necessary to avoid obstacles, optimize inspection efficiency, and reduce energy consumption, the quantum heuristic optimization algorithm can find an inspection route that is optimal in terms of time and resources by adjusting the path nodes and path shapes.

[0076] For example, in a large garden, the system can set path weights according to the importance of different areas and set optimization coefficients according to the complexity of the terrain, and then calculate an optimal path that can cover all key areas and minimize the path length through the quantum annealing process. The initial optimized path data generated by the path provides an efficient and reliable basis for subsequent inspections.

[0077] In summary, the application of the quantum heuristic optimization algorithm in the present invention greatly improves the intelligence and accuracy of path planning. Through its powerful global optimization capability, the system can quickly adapt to the complex and changing garden environment and ensure the efficient execution of inspection tasks. This algorithm not only optimizes the inspection path, but also significantly reduces the inspection time and resource consumption, making garden management more efficient and intelligent.

[0078] During the inspection process, the initial optimized path data is usually used for inspection, and the dynamic monitoring sensor system is used to obtain environmental change information in real time. When an emergency situation or path obstruction is detected, the path decision is adjusted through a hybrid fuzzy neural network, and an adaptive genetic algorithm is used to optimize the path to generate the final adjusted path data for inspection path adjustment;

[0079] like Figure 4 As shown, in the present invention, the path adjustment during the inspection process is a key link to ensure the flexibility and adaptability of the system. In most cases, the inspection equipment uses the initial optimized path data generated by the quantum heuristic optimization algorithm for inspection. However, due to the dynamic characteristics of the garden environment, such as weather changes, emergencies (such as fallen trees), tourist flow, etc., the original path may be blocked or no longer the best choice. To this end, the system integrates a dynamic monitoring sensor system that can obtain environmental change information in real time and trigger the path adjustment mechanism when an emergency is detected.

[0080] During the route adjustment process, the system first analyzes the collected environmental change information through a hybrid fuzzy neural network. The hybrid fuzzy neural network combines the advantages of fuzzy logic and neural networks and can handle uncertainty and fuzzy information. Its principle is to use fuzzy logic to fuzzify environmental change information (such as the degree of route blockage and the impact of weather changes), and then use neural networks to perform pattern recognition and decision output on the fuzzy information. In this way, the system can quickly determine whether the current route needs to be adjusted and generate route adjustment decision data.

[0081] Once it is determined that the path needs to be adjusted, the system starts the adaptive genetic algorithm to optimize the path. The adaptive genetic algorithm is a bionic optimization algorithm that searches for the optimal solution by simulating natural selection and genetic mechanisms. Its adaptability is reflected in the ability to dynamically adjust the crossover rate and mutation rate of the genetic algorithm according to environmental changes, thereby improving search efficiency and adaptability. In the inspection path adjustment, the adaptive genetic algorithm generates the final adjustment path data for the inspection path adjustment by performing a global search and optimization of the path adjustment decision data. This process ensures that even when the environment changes drastically, the system can respond quickly and adjust the inspection route to maintain an efficient inspection process.

[0082] For example, during the inspection process, if the sensor detects that the path ahead is blocked by trees, the system will immediately evaluate the impact of the obstacle through a hybrid fuzzy neural network and replan the path through an adaptive genetic algorithm to select a new path that bypasses the obstacle and still effectively covers the inspection target area. This intelligent path adjustment mechanism not only ensures the continuity and efficiency of the inspection task, but also improves the response speed and reliability of the system in a dynamic environment.

[0083] In summary, the path adjustment mechanism based on dynamic sensing and intelligent algorithms during the inspection process enables the garden automatic inspection system to flexibly respond to environmental changes and ensure the smooth completion of the inspection task. This real-time adjustment capability greatly improves the intelligence level and practical value of the system.

[0084] Preferably, the hybrid fuzzy neural network performs path decision adjustment according to the environmental change information obtained by the dynamic monitoring sensor system during the inspection process to cope with emergencies or path obstructions and generate path adjustment decision data.

[0085] In the present invention, the hybrid fuzzy neural network plays a key role in the inspection process, responsible for processing the environmental change information obtained by the dynamic monitoring sensor system and generating path adjustment decision data. The hybrid fuzzy neural network combines the advantages of fuzzy logic system and neural network, and can effectively cope with complex and uncertain environmental changes. Its core principle is that fuzzy logic can handle imprecise data and fuzzy information, while neural networks are good at pattern recognition and decision generation.

[0086] During the inspection process, the dynamic monitoring sensor system collects real-time information about environmental changes, including temperature, humidity, light intensity, and whether the path is blocked by obstacles. When the system detects environmental changes, the hybrid fuzzy neural network begins to analyze this information. First, fuzzy logic fuzzifies these input data and converts them into different fuzzy sets. For example, light intensity can be fuzzified into three states: "low", "medium", and "high", and the degree of path blockage can be fuzzified into "unblocked", "partially blocked", and "fully blocked". This fuzzification process can effectively deal with the uncertainty and noise in sensor data.

[0087] Next, the neural network uses the fuzzified data for pattern recognition and path decision adjustment. By learning the impact of environmental changes on the inspection path, the neural network can quickly determine whether the current path needs to be adjusted and generate corresponding path adjustment decision data. For example, when the sensor detects that the path ahead is blocked by a tree branch, the hybrid fuzzy neural network will determine the severity of the obstacle through fuzzy logic and quickly evaluate whether the path needs to be adjusted through the neural network.

[0088] The generated route adjustment decision data provides key input information for subsequent route optimization. These data include recommended route adjustment plans and priorities, ensuring that inspection equipment can quickly respond to environmental changes and adjust inspection routes to keep inspection tasks going smoothly.

[0089] Through this intelligent path decision mechanism, the hybrid fuzzy neural network can significantly improve the adaptability and flexibility of the garden automatic inspection system, enabling it to flexibly respond to various emergencies and path blockages in a complex garden environment. This not only ensures the continuity and efficiency of the inspection task, but also improves the intelligence level of the system, making garden management more refined and efficient. In short, the application of the hybrid fuzzy neural network greatly enhances the intelligent ability of garden inspection and provides reliable technical support for the realization of efficient automatic inspection.

[0090] Preferably, the adaptive genetic algorithm optimizes the path adjustment decision data through dynamic crossover rate and mutation rate, and the formula is as follows:

[0091]

[0092]

[0093] in, and are the crossover rate and mutation rate, respectively. and are the minimum and maximum values ​​of the crossover rate, and are the minimum and maximum values ​​of the mutation rate, is the current fitness, and are the maximum fitness and average fitness in the current generation respectively.

[0094] In the present invention, an adaptive genetic algorithm (AGA) is used to optimize the path adjustment decision data to improve the adaptability and efficiency of the inspection path. The adaptive genetic algorithm is an evolutionary computing technology that imitates the process of natural selection and genetic evolution to find the optimal solution to the problem through operations such as crossover and mutation. Unlike traditional genetic algorithms, AGA has the ability to adaptively adjust parameters and can dynamically adjust the crossover rate and mutation rate according to the fitness of the current population, thereby improving the convergence speed and optimization performance of the algorithm.

[0095] In adaptive genetic algorithms, crossover rate and mutation rate are two key parameters that determine the diversity and search ability of the new generation population.

[0096] During the path adjustment process, the dynamic adjustment of the crossover rate and mutation rate enables the adaptive genetic algorithm to flexibly change its search strategy according to the fitness of the population. When the fitness of the population varies greatly, the crossover rate and mutation rate are increased to expand the search range and avoid falling into the local optimum; when the population gradually converges, the crossover rate and mutation rate are reduced to maintain the stability and accuracy of the search. This mechanism ensures that the algorithm can quickly converge to the global optimal solution and adapt to the dynamic changes of the environment.

[0097] For example, during a garden inspection, if a path is detected to be blocked by a temporary obstacle, the adaptive genetic algorithm can quickly find an alternative path to bypass the obstacle by adjusting the crossover rate and mutation rate while maintaining inspection efficiency and path optimality. Through this process, the generated path adjustment data can significantly improve the response speed and decision-making quality of the inspection system in complex environments.

[0098] In general, the application of adaptive genetic algorithm in the present invention not only enhances the intelligent degree of path adjustment, but also significantly improves the flexibility and adaptability of the system in a dynamic environment. This real-time optimization capability ensures the efficient execution of garden automatic inspection tasks and improves the reliability and practical value of the inspection system.

[0099] Preferably, the final adjustment path data for inspection path adjustment is generated by dynamically adjusting the crossover rate and the mutation rate.

[0100] In the present invention, the process of dynamically adjusting the crossover rate and mutation rate is the key to the adaptive genetic algorithm (AGA) to generate the final adjustment path data. The adaptive genetic algorithm optimizes the path by imitating the process of natural selection and genetic evolution, using operations such as crossover and mutation. In the algorithm, the two parameters of crossover rate and mutation rate determine the diversity and search ability of the new generation population, thereby affecting the final effect of path adjustment.

[0101] The crossover rate refers to the frequency of gene combinations between individuals in a population, while the mutation rate refers to the frequency of gene mutations in a population. Dynamic adjustment of these two parameters can help the algorithm to explore and develop more effectively at different stages of evolution. When the algorithm detects that the fitness of the population is very different (i.e., the diversity of path selection is insufficient), increasing the crossover rate and mutation rate can expand the search range and increase the diversity of the population, thereby avoiding falling into the local optimal solution, promoting the algorithm to jump out of the dilemma and find a better path. When the population gradually converges (i.e., the path selection is close to the optimal solution), reducing the crossover rate and mutation rate will help maintain the stability and accuracy of the solution and ensure the accuracy of the search.

[0102] During the inspection process, the system monitors the changes in the garden environment in real time, and dynamically adjusts the crossover rate and mutation rate based on the current fitness information. This mechanism allows the algorithm to perform adaptive path optimization according to environmental changes. For example, when a sudden obstacle is detected blocking the inspection path, the adaptive genetic algorithm can generate new path options by increasing the mutation rate, and combine the advantages of the existing paths through crossover operations to quickly generate an optimized path that bypasses the obstacle. This adaptive adjustment ensures the consistency and efficiency of the inspection process.

[0103] The final adjusted path data generated not only meets the needs of the current inspection task, but also provides strong support for the system to maintain the best path planning in a dynamic environment. By adaptively adjusting the path, the system can maintain a high level of inspection efficiency and accuracy in the face of complex and changing garden environments. This intelligent path adjustment capability makes the garden automatic inspection system more flexible and adaptable in a dynamic environment, greatly improving the reliability and efficiency of the inspection task, and providing advanced technical means for intelligent garden management.

[0104] According to the initial optimized path data or the final adjusted path data generated in an emergency situation and the device status data, the inspection tasks are dynamically allocated and executed, and the inspection task execution feedback data is collected to optimize the path planning and inspection task allocation strategy.

[0105] like Figure 5 As shown, in the present invention, dynamic allocation and execution of inspection tasks are key steps to ensure efficient operation of the system and smooth completion of tasks. This process relies on the initial optimized path data and the final adjusted path data generated in emergency situations, as well as the device status data, to achieve intelligent allocation and execution of inspection tasks. The initial optimized path data is the optimal inspection path pre-planned by the quantum heuristic optimization algorithm. This path can efficiently cover all important garden areas under ideal conditions. However, due to the dynamic nature of the garden environment, such as weather changes, the appearance of obstacles, etc., the system may need to generate final adjusted path data to adapt to these emergencies. The device status data includes information such as the power, load and working status of the inspection equipment to ensure that the equipment performs tasks in the best state.

[0106] During the execution of the inspection task, the system first dynamically assigns the inspection task based on the current path data (whether the initial path or the adjusted path) and the device status. Specifically, the system analyzes the current status of the device and the complexity of the path, and assigns the most appropriate task to each device. For example, if a device has a low battery, the system may assign it a shorter inspection segment to ensure that it can successfully complete the task and return to the charging station. At the same time, the system will optimize the execution order of the tasks based on the urgency and priority of the inspection tasks to ensure that important areas are inspected first.

[0107] During the execution of the inspection task, the device will collect task execution feedback data in real time, including actual inspection coverage, task completion time, equipment energy consumption and other information. By analyzing these feedback data, the system can effectively evaluate the accuracy of path planning and the rationality of task allocation strategy. For example, if the feedback data shows that a certain path takes too long or consumes too much energy, the system can adjust the path or reallocate the task in the next task.

[0108] By dynamically adjusting the allocation and execution strategies of inspection tasks, the system realizes an adaptive inspection process, ensuring that inspection tasks can be completed efficiently in a complex garden environment. This feedback-driven optimization mechanism not only improves the accuracy and efficiency of inspections, but also continuously improves the system's path planning and task allocation strategies over time, thereby improving the intelligence level and operational efficiency of the entire inspection system. Ultimately, the system achieves refined and efficient garden management through intelligent inspection management, providing strong support for the construction of intelligent gardens.

[0109] like Figure 6 As shown, a system for implementing the automatic garden inspection method based on path planning comprises:

[0110] The data acquisition module includes a laser radar, a high-definition camera and a global positioning system device, which is used to collect the terrain data and vegetation distribution data of the garden to generate a three-dimensional model of the garden; and an environmental monitoring sensor system, which is used to collect real-time environmental condition data to generate comprehensive environmental data. The environmental monitoring sensor system includes a temperature sensor, a humidity sensor and a light sensor; the data acquisition module is the basis of the system, and uses a laser radar, a high-definition camera and a global positioning system device to collect the terrain data and vegetation distribution data of the garden. The laser radar generates three-dimensional point cloud data of the terrain by emitting and receiving laser beams, measuring distances and angles, while the high-definition camera captures detailed vegetation images, and the global positioning system device provides accurate geographic location data. These data are combined to generate a three-dimensional model of the garden, providing comprehensive basic information for subsequent analysis and path planning. The environmental monitoring sensor system includes temperature, humidity and light sensors, which collect data on the garden's environmental conditions in real time and generate comprehensive environmental data. This information is used to dynamically evaluate the environmental status of the garden and help adjust inspection strategies and paths.

[0111] The data analysis module is used to analyze the comprehensive environmental data and the three-dimensional model, identify the functional areas of the garden through the support vector machine algorithm and the random forest algorithm, set the inspection priority and inspection parameters of each functional area, and generate functional area data and inspection parameter data; in the data analysis module, the system uses the support vector machine algorithm and the random forest algorithm to analyze the comprehensive environmental data and the three-dimensional model. The support vector machine is a supervised learning model that can identify patterns and structures in the data, and the random forest provides powerful classification and regression capabilities through the integration of multiple decision trees. Through these algorithms, the system can identify functional areas in the garden, such as flower beds, lawns, and bushes, and set different inspection priorities and parameters for these areas. The generation of functional area data and inspection parameter data enables the system to formulate personalized inspection plans for different garden areas.

[0112] The path planning module includes a self-organizing map algorithm module, a fractal geometry analysis module and a quantum heuristic optimization algorithm module, which are used to generate initial optimized path data based on the three-dimensional model and functional area data; the path planning module generates initial optimized path data through the self-organizing map algorithm module, the fractal geometry analysis module and the quantum heuristic optimization algorithm module. The self-organizing map algorithm is an unsupervised neural network used to map high-dimensional data to low-dimensional space, help identify path nodes and optimize path morphology. Fractal geometry analysis optimizes the path morphology by evaluating the complexity and self-similarity of the path to make it more in line with the natural structure of the garden. The quantum heuristic optimization algorithm globally optimizes the path by simulating the annealing process in quantum mechanics to find the optimal inspection route. This process not only improves the efficiency of path planning, but also ensures the global optimality of the inspection path.

[0113] The dynamic path adjustment module includes a dynamic monitoring sensor system, which is used to obtain environmental change information in real time, and uses a hybrid fuzzy neural network module to adjust the path decision, and generates the final adjusted path data through an adaptive genetic algorithm module; the dynamic path adjustment module obtains environmental change information in real time through a dynamic monitoring sensor system. When the path is blocked or the environment mutates during the inspection process, the hybrid fuzzy neural network module adjusts the path decision in time. This module combines fuzzy logic and neural networks, and can handle uncertainty and fuzzy information, and adaptively adjust the inspection path. The adaptive genetic algorithm module further optimizes the adjustment path data, improves the adaptability and optimization performance of the algorithm through dynamic crossover and mutation, and ensures that the generated final adjustment path data can effectively respond to emergencies.

[0114] The task allocation and execution module is used to dynamically allocate and execute inspection tasks according to the initial optimized path data or the final adjusted path data, and collect inspection task execution feedback data to optimize path planning and inspection task allocation strategies. The task allocation and execution module is responsible for dynamically allocating and executing inspection tasks according to the initial optimized path data or the final adjusted path data. This module uses the status data of the inspection equipment to intelligently allocate inspection tasks to ensure the rational use of resources and maximize inspection efficiency. During the execution of the inspection task, the system will collect feedback data, which is used to optimize path planning and task allocation strategies. Through continuous feedback and optimization, the system can improve the accuracy and efficiency of inspections and realize the automation and intelligence of garden management.

[0115] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware.

[0116] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A garden automatic inspection method based on path planning, characterized in that: The following steps are involved: The terrain data and vegetation distribution data of the garden are collected through lidar, high-definition cameras and global positioning system equipment to generate a three-dimensional model of the garden. The real-time environmental condition data is collected through the environmental monitoring sensor system to generate comprehensive environmental data for path planning. Analyze the comprehensive environmental data and the three-dimensional model, identify the functional areas of the garden through a machine learning algorithm, set the inspection priority and inspection parameters of each functional area, and generate functional area data and inspection parameter data for path planning; the machine learning algorithm includes a support vector machine algorithm and a random forest algorithm, which identifies the functional areas in the garden and determines the inspection priority and inspection parameters of each area by analyzing the features in the comprehensive environmental data and the three-dimensional model; Using a self-organizing map algorithm to determine path nodes based on the three-dimensional model and functional area data, optimizing the path morphology in combination with fractal geometry analysis, and globally optimizing the path through a quantum heuristic optimization algorithm to generate initial optimized path data for initial path planning; the fractal geometry analysis is used to evaluate the complexity and curvature of the path, and the path morphology is adjusted to obtain the optimal path morphology, and to generate optimized path data; the quantum heuristic optimization algorithm uses quantum annealing technology to globally optimize the optimized path data to find the optimal solution for the path, and to generate initial optimized path data; During the inspection process, the initial optimized path data is usually used for inspection, and the dynamic monitoring sensor system is used to obtain environmental change information in real time. When an emergency situation or path obstruction is detected, the path decision is adjusted through a hybrid fuzzy neural network, and an adaptive genetic algorithm is used to optimize the path to generate the final adjusted path data for inspection path adjustment; According to the initial optimized path data or the final adjusted path data generated in an emergency situation and the device status data, the inspection tasks are dynamically allocated and executed, and the inspection task execution feedback data is collected to optimize the path planning and inspection task allocation strategy.

2. The method according to claim 1, characterized in that: The environmental monitoring sensor system includes a temperature sensor, a humidity sensor and a light sensor, which are used to detect the environmental conditions of the garden and provide static environmental information to generate comprehensive environmental data.

3. The method according to claim 1, characterized in that: The self-organizing map algorithm is used to generate path node data according to the three-dimensional model and the functional area data, and to determine the node layout of the inspection path by calculating the topological relationship between the path nodes.

4. The method according to claim 1, characterized in that The optimization process of the global optimization is expressed by the following formula: in, represents the total optimization cost of the path; Indicates The weight of each path; Indicates The optimization coefficient of each path; Indicates the number of paths.

5. The method according to claim 1, characterized in that: The hybrid fuzzy neural network makes path decision adjustments according to the environmental change information obtained by the dynamic monitoring sensor system during the inspection process to cope with emergencies or path obstructions and generate path adjustment decision data.

6. The method according to claim 1, characterized in that The adaptive genetic algorithm optimizes the path adjustment decision data through dynamic crossover rate and mutation rate. The formula is as follows: in, and are the crossover rate and mutation rate, respectively. and are the minimum and maximum values ​​of the crossover rate, and are the minimum and maximum values ​​of the mutation rate, is the current fitness, and are the maximum fitness and average fitness in the current generation respectively.

7. A system for implementing the path planning-based automatic garden inspection method according to any one of claims 1 to 6, characterized in that: include: A data acquisition module, including a laser radar, a high-definition camera and a global positioning system device, for collecting terrain data and vegetation distribution data of the garden to generate a three-dimensional model of the garden; and an environmental monitoring sensor system, for collecting real-time environmental condition data to generate comprehensive environmental data, the environmental monitoring sensor system including a temperature sensor, a humidity sensor and a light sensor; A data analysis module is used to analyze the comprehensive environmental data and the three-dimensional model, identify the functional areas of the garden through a support vector machine algorithm and a random forest algorithm, set the inspection priority and inspection parameters of each functional area, and generate functional area data and inspection parameter data; A path planning module, including a self-organizing map algorithm module, a fractal geometry analysis module and a quantum heuristic optimization algorithm module, for generating initial optimized path data based on the three-dimensional model and the functional area data; A dynamic path adjustment module, including a dynamic monitoring sensor system, is used to obtain environmental change information in real time, and uses a hybrid fuzzy neural network module to adjust the path decision, and generates the final adjusted path data through an adaptive genetic algorithm module; The task allocation and execution module is used to dynamically allocate and execute inspection tasks according to the initial optimized path data or the final adjusted path data, and collect inspection task execution feedback data to optimize path planning and inspection task allocation strategies.

Images