Economic forest planting method and device by single row planting on the outside

Abstract

The application discloses an economic forest outside single line planting machine building garden method and device, comprising: step S1, obtaining tree growth data; step S2, according to the biological characteristics and growth data of trees, the operation mode, the layout of the garden road and the planting block is determined; step S3, according to the layout of the garden road and the planting block, the planting block is divided to form an ecological composite planting mode; step S4, according to the characteristics of the planting block, the position of the tree is identified through machine vision, the angle of the water and fertilizer integrated facility nozzle is automatically adjusted, and the pesticide and fertilizer are accurately sprayed around the tree crown; step S5, the maturity and yield of mature fruits are estimated. The technical scheme of the application realizes the full-process intelligentization of economic forest from planting layout, growth management to harvesting prediction.

Description

Technical Field

[0001] This invention belongs to the field of information technology, and in particular relates to a mechanized orchard establishment method and apparatus for single-row planting of economic forests on the outer side. Background Technology

[0002] In the process of establishing economic forests, how to ensure the healthy growth of trees while achieving mechanization and intelligent management of the orchard, effectively improving labor efficiency, and promoting early and high yields is a pressing technical challenge. Traditional economic forests often employ random planting methods, with small spacing between trees and no clear access for machinery, making it difficult for machinery to enter and operate, resulting in low management efficiency and continuously rising labor costs. At the same time, overly dense planting is also detrimental to ventilation and light penetration, affecting yield and quality.

[0003] To address this issue, it is necessary to explore a new model for establishing economic forests. Through rational planning and design, trees should be planted in single rows on the outer quarter of the horizontal planting strip within the park, with appropriate spacing between trees. A comprehensive network of interconnected roads should be established to cover the entire area, creating conditions for mechanized farming and management, thereby improving labor efficiency and reducing costs. However, this layout also presents new challenges: how to maximize planting density and yield within a limited land area while meeting the needs of mechanized operations; how to optimize tree varieties and planting techniques to ensure vigorous growth and early, high yields under reasonable spacing conditions; and how to design and improve machinery suitable for horizontal planting strips in mountainous areas, enabling flexible and efficient operations within the single-row planting strips. These are key technical challenges that need to be overcome in promoting the mechanization of economic forest management. Collaborative efforts from multiple disciplines, including farmland management, agricultural machinery, and agronomy, are required to achieve the mechanization-friendly, efficient, sustainable, and healthy development of economic forests under the single-row planting model on the outer side. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a mechanized orchard establishment method and apparatus for single-row planting of economic forests on the outer side.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A method for establishing an orchard suitable for mechanization, using single-row planting on the outer side of economic forests, includes:

[0007] Step S1: Obtain tree growth data;

[0008] Step S2: Determine the layout of park roads and planting areas based on the biological characteristics and growth data of trees and the management model;

[0009] Step S3: Based on the layout of the park roads and planting areas, divide the planting areas to form an ecological composite planting model;

[0010] Step S4: Based on the results of the planting area division, the location of the trees is identified by machine vision, and the angle of the spray nozzle of the integrated water and fertilizer facility is automatically adjusted to accurately spray the pesticide and fertilizer around the tree canopy.

[0011] Step S5: Identify the maturity of mature fruits and predict their yield.

[0012] As a preferred option, in step S1, experimental areas with different planting densities are set up according to the species and growth characteristics of economic forest trees and management models, tree growth data are collected, and the optimal planting parameters for each tree species are obtained.

[0013] As a preferred option, in step S2, the mountainous mechanization-friendly construction layout optimization system technology using UAV mapping is used to conduct a detailed survey of the park, establish a three-dimensional digital map model, and combine the optimal planting parameters with the three-dimensional model to determine the layout of park roads and planting blocks.

[0014] As a preferred option, in step S3, meteorological, hydrological, and soil information in and around the park is analyzed to determine the impact on tree growth. The analysis results are then combined with the growth characteristics of each tree species to divide the planting areas and form an ecological composite planting model.

[0015] This invention also provides a mechanized orchard establishment device for single-row planting of economic forests on the outer side, comprising:

[0016] The acquisition module is used to acquire tree growth data;

[0017] The determination module is used to determine the layout of park roads and planting blocks based on tree growth data, biological characteristics, and management models;

[0018] The division module is used to divide the planting areas into ecological composite planting models based on the layout of park roads and planting blocks;

[0019] The first processing module is used to identify the location of trees through machine vision based on the planting block division results, dig and bury water and fertilizer integrated pipes outside the horizontal planting strip, and automatically adjust the spraying angle of the water and fertilizer integrated facility to accurately spray the pesticides and fertilizers around the tree canopy.

[0020] The second processing module is used to identify the maturity of mature fruits and predict their yield.

[0021] As a preferred option, the acquisition module sets up experimental areas with different planting densities based on the species, growth characteristics, and management models of economic forest trees, collects tree growth data, and obtains the optimal planting parameters for each tree species.

[0022] As a preferred option, the module uses the UAV-based mountainous mechanization-friendly construction layout optimization system technology to conduct detailed surveying of the park, establish a three-dimensional digital map model, and combine the optimal planting parameters with the three-dimensional model to determine the layout of park roads and planting blocks.

[0023] As a preferred approach, the module analyzes meteorological, hydrological, and soil information in and around the park to determine its impact on tree growth. The analysis results are then combined with the growth characteristics of each tree species to divide the planting areas and form an ecological composite planting model.

[0024] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0025] This invention collects growth data from experimental zones with varying planting densities and establishes a 3D digital map of the park using a drone-based mountainous mechanization optimization system. This map determines the optimal planting area division and infrastructure layout, including roads and water features. Big data analysis of meteorological, hydrological, and soil information is used to delineate ecological composite planting zones. Integrated water and fertilizer application systems with targeted spraying and fertilization are developed, enabling precise application through machine vision. Intelligent harvesting robots are employed, transmitting harvesting data in real time. A yield prediction model is established, combining canopy analysis and historical yield data to predict the current season's yield, and radio frequency identification (RFID) technology is used for traceability management. This invention achieves intelligent management of the entire process of economic forestry, from orchard establishment and planting layout to growth management and harvest prediction. It improves the planting efficiency and management level of economic forests, providing technical support for the mechanization, standardization, and intelligent production of economic forests. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. The drawings in the following description are merely embodiments of the present invention.

[0027] Figure 1 This is a flowchart of the mechanized orchard establishment method for single-row planting of economic forests on the outer side according to the present invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0029] Example 1

[0030] like Figure 1 As shown, an embodiment of the present invention provides a method for establishing an orchard using mechanized planting in single rows on the outer side of an economic forest, which may specifically include:

[0031] Step S1: Set up experimental areas with different planting densities according to the species and growth characteristics of economic forest trees, collect tree growth data, and obtain the optimal planting parameters for each tree species.

[0032] Obtain the growth characteristic parameters of the target tree species and set up at least two experimental areas with different planting densities; plant the target tree species in the experimental areas and collect the growth data of the target tree species; input the growth data into a computer with a support vector machine algorithm to obtain the tree growth curves in the experimental areas; compare the tree growth curves with the preset optimal growth curves to obtain the similarity; determine the experimental area with the highest similarity, obtain the planting density corresponding to the experimental area, and take it as the optimal planting density of the target tree species.

[0033] Specifically, firstly, growth characteristic parameters of the target tree species are extracted from an economic forest tree species database, such as the average annual height growth of a certain fruit tree being 5 meters and the crown width growth rate being 20%. Next, two experimental zones with different planting densities are set up for this tree species, for example, planting 500 trees per hectare and 1000 trees per hectare. The target tree species is planted in each experimental zone, and tree growth data, including parameters such as height, crown width, and diameter, are collected in real time through a sensor network. Then, data processing software is used to analyze the collected data and generate growth curves for the trees in each experimental zone. Next, a support vector machine algorithm is applied to perform a similarity analysis between the growth curve of each experimental zone and a preset optimal growth curve, assuming the optimal growth curve has the characteristics of an average annual height growth of 8 meters and a crown width growth rate of 25%. By calculating the similarity score, the experimental zone with the highest similarity is determined; for example, the experimental zone with 500 trees per hectare has a similarity score of 95, while the experimental zone with 1000 trees per hectare has a score of 85. Therefore, 500 trees per hectare is determined as the optimal planting density parameter for this tree species. Finally, the target tree species and its optimal planting density parameters are stored in the tree species planting database to guide the large-scale planting of the tree species.

[0034] Step S2: Use the UAV-based mountainous mechanization-friendly construction layout optimization system technology to conduct detailed surveying of the park, establish a three-dimensional digital map model, and combine the optimal planting parameters with the three-dimensional model to determine the layout of park roads and planting blocks.

[0035] Acquire high-definition image data of the park taken by a drone equipped with a high-resolution camera;

[0036] For the image data, GPS positioning technology is used for spatial correction to obtain corrected image data with high geographical location accuracy;

[0037] Based on the corrected image data, a high-precision three-dimensional point cloud model of the park is generated using stereo mapping technology.

[0038] The three-dimensional point cloud model is filtered, denoised, and simplified to obtain a three-dimensional digital map model of the park.

[0039] Based on the terrain features and planting requirements of the park, a clustering algorithm was used to divide the three-dimensional digital map model into different planting blocks.

[0040] For each planting block, the optimal planting parameters for that planting block are determined using a decision tree algorithm based on crop growth requirements and environmental conditions.

[0041] The optimal planting parameters are fused with the three-dimensional digital map model to generate a refined planting layout plan for the park. This refined planting layout plan is used to guide the park's road planning and planting area division, thereby achieving refined management of the park.

[0042] Specifically, firstly, a drone equipped with a high-resolution camera was used to take aerial photographs of the park, acquiring high-definition image data with a resolution of 1 meter. Then, GPS positioning technology was used to spatially correct the image data, and a feature point matching algorithm was used to register the image data with ground control points, achieving centimeter-level accuracy in the geographic location of the corrected image data. Next, stereo mapping technology was used to reconstruct the 3D image data. A dense point cloud was generated through multi-view image matching, and then a high-precision 3D point cloud model of the park was generated using a Poisson reconstruction algorithm, achieving a point cloud density of 100 points per square meter. Afterward, statistical filtering was applied to the 3D point cloud model to remove noise, and voxelization was used to simplify the point cloud, resulting in a 3D digital map model of the park. For this model, the K-means clustering algorithm was used to divide it into 5 planting blocks based on terrain features such as elevation and slope. For each planting block, based on the crop's requirements for environmental conditions such as temperature, humidity, and light, the C5 decision tree algorithm analyzes the environmental data within the block to obtain the optimal planting parameters for crop growth. For example, a certain block is suitable for planting crops at a temperature of 22℃, humidity of 60%, and 10 hours of light per day. Finally, the optimal planting parameters are assigned to the corresponding block's 3D model to generate a refined planting layout plan for the park. When planning park roads, it is considered that they match the boundaries of the planting blocks. This plan is used to guide park planning and planting management, achieving precision agriculture.

[0043] Step S3: Utilize big data analysis of meteorological information in the park and surrounding areas to determine the impact on tree growth. Combine the analysis results with the growth characteristics of each tree species to divide planting areas and form an ecological composite planting model.

[0044] Acquire meteorological information data inside and outside the park, including multi-dimensional data such as temperature, humidity, light intensity, and wind speed;

[0045] The meteorological information data is preprocessed to remove outliers and missing values, and the data is standardized.

[0046] Clustering algorithms were used to analyze the preprocessed meteorological information data to obtain the division results of the areas inside and outside the park;

[0047] Obtain data on the growth characteristics of common tree species in the park, including suitable temperature, humidity, and light conditions;

[0048] The regional division results are matched with the tree species growth characteristic data to determine the suitable tree species for planting in each region;

[0049] Based on the regional division results and the results of the suitable tree species judgment, a combined optimization algorithm is used to obtain the ecological composite planting scheme for each region;

[0050] The planting plan will be visualized.

[0051] The planting plan is dynamically adjusted according to changes in meteorological conditions to form an intelligent tree management model for the park.

[0052] Specifically, the process begins by deploying meteorological sensors both inside and outside the park to collect multi-dimensional meteorological data, including temperature, humidity, light intensity, and wind speed, every 10 minutes. Then, the Tukey algorithm is used to detect outliers in the collected meteorological data, and missing values ​​are interpolated using the most recent data point. Next, a min-max normalization method is used to map each dimension of the data to the 0-1 interval for standardization. Following this, K-means clustering is used to perform cluster analysis on the preprocessed meteorological data. The optimal number of clusters, determined by silhouette coefficient evaluation, is 4, dividing the park into four meteorological zones. Then, growth characteristic data for 10 common tree species within the park, such as ginkgo and Chinese scholar trees, are collected, including suitable temperature ranges, humidity ranges, and light intensity. A decision tree algorithm is used to analyze the matching degree between the meteorological characteristics of each meteorological zone and the growth characteristics of the tree species, determining the suitable tree species combinations for each zone. For example, zone A is suitable for planting ginkgo and Chinese scholar trees. Based on this, a simulated annealing algorithm was used to optimize tree species combinations, aiming to maximize carbon sequestration. The optimal scheme for ecological composite planting in each area was calculated, and a three-dimensional visualization map of tree species distribution in the park was generated. Finally, changes in meteorological conditions in the park were monitored in real time. When significant changes occurred in meteorological characteristics, a tree species suitability analysis process was triggered to dynamically adjust the planting plan, thereby achieving intelligent management of trees in the park.

[0053] By analyzing the preprocessed meteorological information data using clustering algorithms, regional division results under different meteorological conditions were obtained.

[0054] The meteorological data is acquired and preprocessed by removing outliers and missing values ​​and performing standardization to obtain preprocessed meteorological data. The preprocessed meteorological data is then input into at least one clustering algorithm for cluster analysis to obtain clustering results under different meteorological conditions. The clustering results are then visualized using scatter plots and radar charts to show the regional division under different meteorological conditions.

[0055] Specifically, meteorological data from both inside and outside the park is acquired, including data on multiple dimensions such as temperature, humidity, light intensity, and wind speed. The acquired meteorological data undergoes preprocessing. By setting thresholds, outliers exceeding the normal range are removed; for example, outliers in the temperature data such as -50℃ or 100℃ are identified as abnormal and removed. For missing data, data interpolation is used to fill in the missing values. If temperature data for a certain time point is missing, the temperature values ​​from the preceding and following time points are averaged to obtain the filled data. Finally, a min-max standardization method is used to normalize the data of each dimension according to their value range, resulting in standardized data between 0 and 1. The preprocessed meteorological data was input into the K-means clustering algorithm, with the number of clusters k set to 4. Through iterative optimization, four cluster centers were obtained, each representing different meteorological conditions. For example, the center point of the first cluster is (8,6,9,5), representing meteorological conditions of high temperature, moderate humidity, strong sunlight, and moderate wind speed. The clustering results were visualized, showing the distribution of different sample points in the four clusters in the form of a scatter plot. Combined with a radar chart, the meteorological conditions represented by each cluster center were intuitively displayed, providing a meteorological analysis basis for subsequent tree species planning.

[0056] Based on the regional division results and the judgment results of suitable tree species, a combined optimization algorithm is used to determine the ecological composite planting scheme for each region.

[0057] Obtain the regional division results and suitable tree species judgment results, and construct an ecological composite planting combination optimization model; use a heuristic algorithm to solve the combination optimization model to obtain the optimal tree species combination scheme for the region; evaluate the tree species combination scheme to determine whether the tree species combination scheme meets the regional characteristics and ecological requirements. If it meets the requirements, the tree species combination scheme is determined as the final scheme; if it does not meet the requirements, return to the combination optimization model to solve it again.

[0058] Specifically, after obtaining the regional division results and suitable tree species determination results, the first step is to construct a combined optimization model for ecological intercropping. Assume a region is divided into three different meteorological condition zones, and each zone requires the selection of three suitable tree species for planting. By analyzing historical meteorological data and soil conditions, the growth adaptability score of each tree species in different zones is determined, ranging from 0 to 10. Next, a heuristic algorithm, such as a genetic algorithm, is used to solve the combined optimization model. The genetic algorithm optimizes the tree species combination scheme step by step through selection, crossover, and mutation operations. The initial population size is set to 100, the number of iterations is 50, the crossover probability is 8, and the mutation probability is 1. In each iteration, the fitness value of each combination scheme is calculated, which is a weighted average of the tree species' growth adaptability score and ecological benefit score. After multiple iterations, the tree species combination scheme with the highest fitness value is obtained. Then, the scheme is evaluated to determine whether it meets the regional characteristics and ecological requirements. Evaluation criteria include biodiversity index, soil retention capacity, and water conservation capacity. If the evaluation results show that the scheme meets all requirements, it is determined as the final scheme. If the conditions are not met, the solution is returned to the combinatorial optimization model for re-solving, adjusting parameters or introducing new tree species selection strategies to improve the adaptability and ecological benefits of the scheme. This method, supported by information technology, can automatically provide optimal tree species combinations for regions under different meteorological conditions.

[0059] The planting plan is dynamically adjusted based on changes in meteorological conditions to determine the feasibility of forming an intelligent tree management model for the park.

[0060] Acquire meteorological data collected by meteorological monitoring equipment, wherein the meteorological data includes at least temperature, humidity and rainfall;

[0061] Based on the variation patterns of the meteorological data, a model for adjusting vegetation layout was established;

[0062] The real-time meteorological data is input into the vegetation layout adjustment model to obtain a vegetation layout scheme adapted to the current meteorological conditions.

[0063] Specifically, firstly, meteorological data, including temperature, humidity, and rainfall, is acquired through meteorological monitoring equipment. For example, in a certain region, temperature data might record between 15°C and 30°C, humidity fluctuating between 40% and 80%, and rainfall between 50 mm and 200 mm per month. Next, based on the patterns of these meteorological data fluctuations, a model for adjusting vegetation layout is established using time series analysis. Specifically, the ARIMA model can be used to predict temperature, humidity, and rainfall, identifying the periodic and trend changes in meteorological data. Assuming that analysis indicates the temperature will rise to an average of 25°C, humidity will remain around 60%, and rainfall will decrease to 70 mm per month over the next month, then the real-time meteorological data is input into the vegetation layout adjustment model. The model simulates the response of different vegetation layouts to meteorological conditions and uses a genetic algorithm to optimize the vegetation layout scheme to adapt to current meteorological conditions. For example, in the case of rising temperatures and decreasing rainfall, the model might suggest increasing the proportion of drought-tolerant plants, such as cacti and succulents, while reducing plants with higher water requirements, such as ferns. Ultimately, the vegetation layout scheme output by the model can maximize ecological benefits under current meteorological conditions, ensuring healthy vegetation growth and ecosystem stability.

[0064] Step S4: Based on the planting area division results, develop a directional spraying and fertilizing machine suitable for a single-row planting pattern on the outer side of the horizontal planting strip, using machine vision to identify the tree positions. A water and fertilizer integrated pipeline is dug and buried on the outer side of the horizontal planting strip, automatically adjusting the nozzle angle to precisely spray pesticides and fertilizers around the tree canopy.

[0065] Geographic information data of the planting area is acquired to obtain the specific location and boundary information of the planting area; based on the geographic information data, the location of trees within the planting area is identified; for each tree location, it is determined whether the initial angle setting of the fertilization machine's nozzle meets the preset spraying requirements; if the initial angle setting of the nozzle does not meet the preset spraying requirements, the nozzle angle is adjusted to meet the preset spraying requirements; based on the preset spraying parameters, combined with the specific location and boundary information of the planting area and the tree location, the spraying path of the fertilization machine is automatically generated; the fertilization machine is controlled to move according to the spraying path, and during the movement, the angle of the nozzle is controlled to perform targeted fertilization on the trees.

[0066] Specifically, the process involves acquiring geographic information data for the planting areas. A Geographic Information System (GIS) is used to pinpoint the location and boundaries of these areas to the centimeter level. For example, data obtained using RTK differential GPS technology indicates that area A is located at 35.678 degrees North latitude and 124.567 degrees East longitude, with a rectangular area measuring 100 meters east-west and 50 meters north-south. Next, LiDAR scanning technology is used to identify the locations of trees within the area, ensuring that the positional error of each tree is within 10 centimeters. For example, the coordinates of a tree located at the center of the area are determined to be (50, 25). Subsequently, computer vision algorithms are used to analyze the initial angle of the fertilizer spraying machine's nozzles. If the horizontal distance between the nozzle and the tree is found to be 5 meters, while the preset spraying requirement is a horizontal distance of 3 meters, the nozzle angle is adjusted using a stepper motor to meet the preset requirement. Based on preset spraying parameters, such as a nozzle flow rate of 10 liters per minute and a spraying range with a radius of 2 meters, combined with GIS data and tree locations, the spraying path of the fertilizer spraying machine is automatically planned. For example, the path planning algorithm calculates that starting from the block entrance, the machine moves east-west at a speed of 1 meter per second, automatically adjusting the nozzle direction to fertilize each tree upon arrival. In actual operation, the fertilizing machine moves along the planned path, adjusting the nozzle angle in real time using angle sensors mounted on the nozzles to ensure spraying accuracy. Throughout the process, the fertilizing machine's automatic navigation system utilizes pre-loaded block maps and real-time sensor data to precisely control its movement trajectory and spraying actions, achieving precise fertilization of the trees.

[0067] Step S5: Use a fruit-harvesting robot to harvest mature fruits. The ripeness of the fruits is identified by machine vision, and the flexible robotic arm gently picks up the fruits. The harvesting data is transmitted to the yield prediction model in real time.

[0068] Acquire fruit images, identify the maturity of the fruits in the images, and obtain maturity data; use a preset yield prediction model to analyze the maturity data and obtain a yield prediction result; if the prediction result does not match the preset threshold, adjust the accuracy parameter of the maturity identification, re-identify the maturity of the fruits, until the prediction result meets the preset threshold, and output the yield prediction result that meets the preset threshold.

[0069] Specifically, when acquiring fruit images, high-resolution cameras can be used to photograph the fruit in the orchard, ensuring image clarity and detail integrity. Image processing algorithms, such as convolutional neural networks (CNNs), are then used to analyze the fruit images, identifying the fruit's color, shape, and texture features to determine ripeness. Ripeness data can be determined by comparing the RGB values ​​of the fruit's color with the standard color values ​​of ripe fruit; for example, if the red component of the fruit reaches 200 or higher, it is considered ripe. Next, a pre-set yield prediction model is used. This model, based on a linear regression algorithm, takes ripeness data and historical yield data as input and outputs a yield prediction for the current orchard. If the predicted result is lower than a pre-set yield threshold, such as below 1000 kg, the accuracy parameters for ripeness identification need to be adjusted, such as increasing the image acquisition frequency or improving image resolution, and the fruit ripeness identification is repeated. By iteratively adjusting the parameters until the yield prediction result reaches the pre-set threshold, the accuracy of the model is ensured. Finally, the yield prediction results that meet the pre-set threshold are output for fruit growers to make production decisions.

[0070] Step S6: Establish a yield prediction model. Analyze parameters such as tree canopy size and fruit quantity using fruit tree image recognition technology. Combine historical yields, meteorological data, and the real-time harvesting data to predict the fruit yield for the current season. Use radio frequency identification technology to identify and trace the fruit.

[0071] The process involves: acquiring the size of the fruit tree canopy and the number of fruits, using these as input parameters; retrieving historical yield data and corresponding meteorological data from a pre-established database, and combining this data with the real-time acquired canopy size and fruit number; analyzing the combined data using machine learning algorithms to establish a yield prediction model and predict the current season's fruit yield; comparing the prediction result with a preset threshold; if the predicted yield is higher than the threshold, proceeding to the next step; if the predicted yield is lower than the threshold, adjusting the parameters of the yield prediction model and re-predicting; using radio frequency identification (RFID) technology to identify the fruit and associating the fruit information with the yield prediction result; transmitting the identified fruit information to the database for traceability management and updating the fruit tree status information; and continuously optimizing the yield prediction model based on real-time harvest data to improve prediction accuracy.

[0072] Specifically, LiDAR scanning technology can be used to obtain the size of the fruit tree canopy and the number of fruits. This involves creating a 3D model of the fruit tree, extracting parameters such as canopy volume, and simultaneously using computer vision algorithms to identify the fruits and determine their quantity. Historical yield data and corresponding meteorological data, including average yield per tree, average temperature, and precipitation over the past five years, are retrieved from a pre-established database and fused with real-time canopy volume and fruit quantity parameters. A support vector machine (SVM) regression algorithm is then used to train the combined data to establish a yield prediction model. The current fruit tree parameters are input into the model to predict the current season's fruit yield. If the predicted yield is lower than a preset threshold (e.g., 5000 kg), the kernel function parameters of the SVM, such as the width of the Gaussian kernel, are adjusted, and the model is retrained and predicted until the predicted yield meets the requirements. RFID electronic tags are used to identify harvested fruit, recording information such as harvest time, tree number, and quality. This information is then linked to the yield prediction results and uploaded to the database. The database uses a distributed architecture, supporting petabyte-level large-scale data storage and real-time querying, facilitating fruit traceability management and fruit tree status updates. Meanwhile, real-time harvesting data from the harvesting robot, such as harvesting time, fruit size, and maturity, is transmitted to the database to enable online learning and optimization of the yield prediction model. The gradient descent algorithm is used to update the model parameters in real time, continuously improving the prediction accuracy.

[0073] Example 2:

[0074] This invention also provides a mechanized orchard establishment device for single-row planting of economic forests on the outer side, comprising:

[0075] The acquisition module is used to acquire tree growth data;

[0076] The determination module is used to determine the layout of park roads and planting blocks based on tree growth data, biological characteristics, and management models;

[0077] The division module is used to divide the planting areas into ecological composite planting models based on the layout of park roads and planting blocks;

[0078] The first processing module is used to identify the location of trees through machine vision based on the planting block division results, dig and bury water and fertilizer integrated pipes outside the horizontal planting strip, and automatically adjust the spray angle of the water and fertilizer integrated facility to accurately spray the pesticides and fertilizers around the tree canopy.

[0079] The second processing module is used to identify the maturity of mature fruits and predict their yield.

[0080] As one embodiment of the present invention, the acquisition module sets up experimental areas with different planting densities according to the species and growth characteristics of economic forest trees and management models, collects tree growth data, and obtains the optimal planting parameters for each tree species.

[0081] As one embodiment of the present invention, the determination module uses the UAV-based mountainous mechanization construction layout optimization system technology to conduct detailed surveying of the park, establish a three-dimensional digital map model, and combine the optimal planting parameters with the three-dimensional model to determine the layout of park roads and planting blocks.

[0082] As one embodiment of the present invention, the division module analyzes the meteorological, hydrological and soil information of the park and surrounding areas to determine the impact on tree growth. The analysis results are combined with the growth characteristics of each tree species to divide the planting blocks and form an ecological composite planting model.

[0083] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for establishing an orchard suitable for mechanization, characterized by single-row planting on the outer side of economic forests. include: Step S1: Obtain tree growth data; Step S2: Determine the layout of park roads and planting areas based on the biological characteristics and growth data of trees and the management model; Step S3: Based on the layout of the park roads and planting areas, divide the planting areas to form an ecological composite planting model; Step S4: Based on the results of the planting area division, the location of the trees is identified by machine vision, and the angle of the spray nozzle of the integrated water and fertilizer facility is automatically adjusted to accurately spray the pesticide and fertilizer around the tree canopy. Step S5: Identify the maturity of mature fruits and predict their yield.

2. The method for establishing an orchard using mechanized planting in single rows on the outer side of economic forests as described in claim 1, characterized in that, In step S1, experimental areas with different planting densities are set up according to the species and growth characteristics of economic forest trees and management models, tree growth data are collected, and the optimal planting parameters for each tree species are obtained.

3. The method for establishing an orchard using mechanized planting in single rows on the outer side of economic forests as described in claim 2, characterized in that, In step S2, the mountainous mechanization-friendly construction layout optimization system technology using UAV mapping is used to conduct detailed mapping of the park, establish a three-dimensional digital map model, and combine the optimal planting parameters with the three-dimensional model to determine the layout of park roads and planting blocks.

4. The method for establishing an orchard using mechanized planting in single rows on the outer side of economic forests as described in claim 3, characterized in that, In step S3, meteorological, hydrological, and soil information in and around the park is analyzed to determine the impact on tree growth. The analysis results are combined with the growth characteristics of each tree species to divide the planting area and form an ecological composite planting model.

5. A mechanized orchard establishment device for single-row planting of economic forests on the outer side, characterized in that, include: The acquisition module is used to acquire tree growth data; The determination module is used to determine the layout of park roads and planting blocks based on tree growth data, biological characteristics, and management models; The division module is used to divide the planting areas into ecological composite planting models based on the layout of park roads and planting blocks; The first processing module is used to identify the location of trees through machine vision based on the planting block division results, dig and bury water and fertilizer integrated pipes outside the horizontal planting strip, and automatically adjust the spraying angle of the water and fertilizer integrated facility to accurately spray the pesticides and fertilizers around the tree canopy. The second processing module is used to identify the maturity of mature fruits and predict their yield.

6. The mechanized orchard establishment device for single-row planting of economic forests on the outer side as described in claim 5, characterized in that, The acquisition module sets up experimental areas with different planting densities based on the species, growth characteristics, and management models of economic forest trees, collects tree growth data, and obtains the optimal planting parameters for each tree species.

7. The mechanized orchard establishment device for single-row planting of economic forests on the outer side as described in claim 6, characterized in that, The module uses UAV mapping technology to conduct detailed mapping of the park, establish a three-dimensional digital map model, and combine the optimal planting parameters with the three-dimensional model to determine the layout of park roads and planting blocks.

8. The mechanized orchard establishment device for single-row planting of economic forests on the outer side as described in claim 7, characterized in that, The module analyzes meteorological, hydrological, and soil information in and around the park to determine the impact on tree growth. The analysis results are combined with the growth characteristics of each tree species to divide the planting area and form an ecological composite planting model.

Images