Variable rate fertilization control method and system based on a knapsack spreader

Abstract

The present application relates to a kind of variable fertilization control method and system based on knapsack fertilizer distributor, belong to agricultural machinery control field.The system is composed of main controller, speed measurement module, satellite navigation module, motor control module and execution structure.By satellite navigation, the position information of the operator is determined, the field is divided into grid and its fertility is analyzed, combined with fertilizer density, walking speed etc., the discharge model is established, intelligent variable fertilization is realized, the fertilization accuracy is improved, and fertilizer waste and environmental pollution are avoided.Main controller connects each module to process data and send instructions, 4G chip transmits operation information to cloud platform to realize remote supervision, and the operator can remotely manage and adjust operation according to the information.The fertilizer device of execution structure is driven by motor to discharge.The system provides automatic and manual mode, has wide applicability and low cost, is suitable for farmers to use, and is of great significance to improve agricultural production efficiency and quality and promote sustainable development of agriculture.

Description

Technical Field

[0001] This invention relates to the field of agricultural machinery control, and in particular to a variable fertilizer application control method and system based on a backpack fertilizer spreader. Background Technology

[0002] Fertilization is a crucial aspect of agriculture, as the correct application of fertilizers directly impacts crop yield.

[0003] Due to the high cost of large agricultural machinery, research indicates that backpack fertilizer spreaders are the most widely used fertilizer application method. Many existing patents have made structural improvements to backpack fertilizer spreaders, enabling precise and uniform application. However, their functionality is relatively limited, relying primarily on manual operation and lacking intelligent control. Operators typically rely on experience for fertilization, making it difficult to accurately determine the amount and timing of application. This not only leads to low fertilization precision and uneven application, affecting crop growth and development, but also potentially wastes fertilizer and increases agricultural production costs. Furthermore, traditional backpack fertilizer spreaders cannot account for the impact of operating speed on fertilization effectiveness. In actual operation, changes in operating speed directly affect the amount of fertilizer applied per unit time. Failure to adjust the fertilizer application rate according to the operating speed can result in over- or under-fertilization, impacting crop yield and quality.

[0004] In conclusion, traditional fertilization methods and related systems can no longer meet the needs of modern agricultural production. There is an urgent need for a variable-rate fertilization control system based on a backpack fertilizer spreader, capable of intelligent fertilization, taking into account operating speed and plot differences, and possessing remote monitoring capabilities. This system should also be cost-effective and widely applicable to improve agricultural production efficiency and quality, reduce fertilizer waste, and protect the ecological environment. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention aims to provide a variable fertilization control method and system based on a backpack fertilizer spreader. This method solves the problems existing in the prior art. The invention utilizes a satellite navigation module to accurately determine the operator's location information, and combines this with the plot shape, area, and historical fertilization data to adaptively divide the plot into multiple grids. Through complex shape irregularity calculations and fertility distribution analysis (such as constructing a fertility distribution surface using Kriging interpolation), the soil fertility status of different grids is determined. Based on this, the grid size is determined according to shape and fertility change thresholds, ensuring accurate understanding of plot information and laying the foundation for precision fertilization. For example, smaller grids are used in areas with complex shapes and large fertility differences to more precisely identify local fertility changes; while larger grids are used in regular areas with uniform fertility, optimizing computational resource utilization while maintaining accuracy. The invention can dynamically calculate the motor speed based on actual conditions, thereby precisely controlling the discharge speed. This allows the fertilizer application rate to be adjusted in real time according to the needs of different plots and operating conditions, avoiding uneven fertilization and waste caused by experience-based fertilization, improving fertilizer utilization, and reducing environmental pollution risks.

[0007] (II) Technical Solution

[0008] To achieve the above objectives, the present invention provides the following technical solution: a variable fertilizer application control method based on a backpack fertilizer spreader, comprising the following steps:

[0009] Adaptive grid division: Obtain the boundary coordinate information of the plot through the satellite navigation module, determine the shape and area of ​​the plot, calculate the irregularity index of the plot shape, analyze the changes in the fertility distribution of the plot using the spatial interpolation algorithm, generate the fertility distribution surface, set the threshold for the irregularity of the shape and the fertility change, divide the grid according to the preset direction, and process the edge grid to make it fit the plot boundary.

[0010] Soil fertility assessment: Extract multi-dimensional feature data on soil fertility, topography, and crop planting history for each grid cell, and organize them into feature vectors; calculate the average fertilizer requirement index for each grid cell using a clustering algorithm, and formulate fertilization strategies;

[0011] Information acquisition and transmission: The satellite navigation module determines the operator's time, location, and speed information and inputs it to the main controller; the speed measurement module collects the current discharge speed of the equipment and inputs it to the main controller.

[0012] Information processing: The main controller receives GPS data from the satellite navigation module and parses the information to obtain the corresponding fertilizer requirements and integrates the requirements with walking speed and fertilizer density information.

[0013] Discharge rate calculation: Based on the selected fertilizer type, the software combines different fertilizer densities and the worker's walking speed to establish a discharge model;

[0014] Motor control and discharge: The main controller sends the calculated theoretical speed to the motor control module. The PWM signal generator in the motor control module receives and responds to the command signal from the main controller and converts it into a PWM wave. The motor driver is connected to the PWM signal generator and drives the motor to rotate. The motor converts the received voltage signal into the corresponding speed to control the discharge speed.

[0015] As a preferred embodiment of the present invention, the adaptive grid division step, which uses a spatial interpolation algorithm to analyze the changes in land fertility distribution, includes:

[0016] Data preparation: Collect historical fertilization data and existing fertility monitoring data within the plot;

[0017] Constructing the covariance matrix: Calculate the covariance between any two data points and construct the covariance matrix;

[0018] Solving for the weighting coefficients: Solving for the weighting coefficients of each known data point for the fertility estimate at the unknown location using a system of linear equations;

[0019] Calculate the fertility estimate: For any unknown location within the plot, multiply the fertility values ​​of the surrounding known data points by the corresponding weighting coefficients and sum them to obtain the fertility estimate for that unknown location;

[0020] Constructing a fertility distribution surface: Corresponding the calculated fertility estimation matrix to the coordinate system of the plots, using geographic information system tools, the fertility estimation value of each location is mapped to the corresponding spatial location to construct a fertility distribution surface.

[0021] As a preferred technical solution of the present invention, in the soil fertility calculation step, the feature vector is [mean nitrogen content, standard deviation of nitrogen content, mean phosphorus content, standard deviation of phosphorus content, mean potassium content, standard deviation of potassium content, average altitude, standard deviation of altitude, average slope, standard deviation of slope, crop type code, planting frequency, average yield, standard deviation of yield].

[0022] As a preferred embodiment of the present invention, in the soil fertility calculation step, the clustering algorithm calculates the grid average fertilizer requirement index, including:

[0023] Distance calculation and cluster assignment: For each grid cell, calculate the distance between its feature vector and each cluster center;

[0024] Cluster center update: After completing a round of grid allocation, the center of each cluster is recalculated. For each cluster, the feature vectors of all grids belonging to that cluster are averaged to obtain the new cluster center feature vector.

[0025] Iterative loop: Repeat the above steps of distance calculation, cluster assignment, and cluster center update to form an iterative loop until the set maximum number of iterations is reached.

[0026] As a preferred embodiment of the present invention, the theoretical speed calculation formula for fertilization in the grid during the discharge speed calculation step is: V1=(F1 / A1)*W*V / K, where V1 represents the theoretical motor speed of grid 1 (r / s), F1 represents the amount of fertilizer required for grid 1 (kg), and A1 represents the area of ​​grid 1 (m²). 2 W represents the width (m), V represents the walking speed (m / s), and K represents the coefficient relationship between the motor speed (r / s) and the output amount (kg / s).

[0027] A variable fertilizer application control system based on a backpack fertilizer spreader includes a control system and an actuator.

[0028] As a preferred embodiment of the present invention, the control system includes:

[0029] 1. The main controller, as the core control unit of the entire system, coordinates and manages the various modules. It receives and processes data from the speed measurement module, the satellite navigation module, and the operating parameters input by the user via buttons. Based on this information, it calculates the required amount of fertilizer and the discharge rate for the plot, and sends instructions to the motor control module.

[0030] The main controller includes an LCD screen, buttons, a 4G chip, a main chip, and a serial communication unit, wherein:

[0031] a. LCD screen: Used to display operational information, allowing users to intuitively understand the current operating status of the system. The displayed content includes the operational mode (such as automatic mode, manual mode, etc.), operational speed (including travel speed and discharge speed, etc.), operational status (operating, paused, faulty, etc.), and set parameters (such as fertilizer type, fertilizer application rate settings, etc.), providing users with real-time feedback so that they can monitor the fertilizer application operation at any time and make adjustments as needed.

[0032] b. Buttons: These allow users to set operating parameters and personalize system operation. They enable settings for discharge acceleration / deceleration, adjusting the fertilizer discharge speed according to actual needs. Fertilizer type settings are also available; different fertilizers may require different application rates and discharge speed control strategies, which the system can adjust accordingly. The application rate setting directly determines the amount of fertilizer applied each time, meeting the diverse fertilizer requirements of different plots.

[0033] c. 4G Chip: Responsible for sending operation information to the cloud platform, enabling remote monitoring and operation management. Utilizing the MQTT protocol, it sends current operation information, such as operation location, speed, walking speed, amount of fertilizer applied, required amount of fertilizer, and fertilizer type, to the cloud platform. The cloud platform can store, process, and display this information, allowing operators to check the operation status anytime, anywhere. Simultaneously, the cloud platform can use the received operation location information to calculate operation mileage and area, providing operators with more comprehensive operation management functions.

[0034] d. Main Chip: As the core processing unit of the main controller, it undertakes the crucial tasks of data reception, processing, storage, and transmission. It receives data from various modules, including positioning, time, latitude and longitude, and speed information from the satellite navigation module, as well as discharge speed information and operational parameters input via buttons from the speed measurement module. It processes this data, such as resolving various information from the GPS data from the satellite navigation module and calculating plot identification and discharge speed using specific algorithms. For example, when dividing a large plot into multiple smaller grids, the main chip determines its current grid location based on positioning information and obtains the required fertilizer amount for that grid. It stores important data for subsequent querying and analysis. Simultaneously, based on the processing results, it sends instructions to the motor control module to control the discharge speed, achieving precise fertilization.

[0035] e. Serial Communication Unit: This unit serves as a connection and communication link between the main controller and the satellite navigation module and speed measurement module. It communicates with the satellite navigation module to obtain geographic location coordinates and speed information, ensuring the main controller can accurately determine the operator's position and operating speed. Connecting to the speed measurement module, it processes the received speed information and sends it to the LCD screen and cloud platform for display, allowing users and the remote monitoring platform to understand the current material output speed and operation progress.

[0036] 2. Speed ​​Measurement Module: Primarily used to measure the discharge speed of the fertilizer application equipment. Through a Hall effect sensor working in conjunction with a magnet mounted on the rotating shaft, it senses changes in the magnetic field and outputs different level signals, thereby calculating the current fertilization speed. This information is then transmitted to the main controller, which adjusts the fertilization operation based on the discharge speed and other factors.

[0037] 3. Satellite Navigation Module: This module determines the operator's location, time, and speed. The GPS receiver uses TTL data to calculate the operator's latitude and longitude coordinates every 100 milliseconds, obtaining the operator's current location and calculating the operator's average speed. This information is input into the main controller, which combines the location information with a pre-defined plot grid to determine the required amount of fertilizer for the plot where the operator is located, while also considering factors such as operating speed for fertilizer application control. The GPS antenna receives GPS signals to ensure the satellite navigation module functions properly.

[0038] 4. Motor Control Module: This module includes a PWM signal generator, a motor driver, and the motor itself. The motor control module controls the motor's operation based on instructions from the main controller. The PWM signal generator receives the instruction signals from the main controller and converts them into PWM waves. The motor driver drives the motor to rotate according to the PWM waves. The motor converts the received voltage signals into corresponding speeds, thereby achieving precise control of the fertilizer discharge rate.

[0039] As a preferred embodiment of the present invention, the execution structure includes a fertilizer applicator. The fertilizer applicator includes a power supply compartment, a power switch mounted on the right side of the power supply compartment, and a built-in battery fixed inside the power supply compartment. The built-in battery is connected to a first motor via wires. The first motor is connected to a discharge channel via a rotating shaft. The discharge channel is connected to a discharge hopper via a discharge hose. The discharge hopper includes a second motor connected to a turntable. A material bucket is mounted on the upper part of the power supply compartment, and shoulder straps are symmetrically fixed on both sides of the material bucket. The material bucket is used to load solid granular fertilizer. The built-in battery inside the power supply compartment powers all components of the system. Under the command of the control system, the first motor drives the fertilizer in the material bucket through the discharge channel. The discharge channel connects the material bucket and the discharge hose, conveying the fertilizer to the discharge hopper. The discharge hose connects and conveys the fertilizer. The discharge hopper includes a second motor and a turntable. The second motor drives the turntable to evenly spread the incoming fertilizer onto the ground at a set width. The shoulder straps facilitate the operator to carry the fertilizer applicator on their back for operation.

[0040] As a preferred technical solution of the present invention, the built-in battery is connected to the speed sensor via a wire. The speed sensor is installed on the right side of the discharge channel and is used in conjunction with a magnet installed on the rotating shaft. The speed sensor is connected to the main chip via a sensor interface, and the main chip calculates the motor speed.

[0041] As a preferred embodiment of the present invention, the built-in battery is connected to the GPS receiver via a wire, the GPS receiver is connected to the GPS antenna via a GPS antenna interface, and the GPS receiver is connected to the main chip via a sensor interface. The main chip acquires GPS data and calculates positioning information, location information, time information, and speed information.

[0042] As a preferred embodiment of the present invention, the built-in battery is connected to the motor drive module via wires, and the motor drive module is connected to the main chip via a motor drive interface. The main chip calculates the theoretical rotation speed by comprehensively considering the acquired position information, speed information, and land information, and sends it to the motor drive module, which then executes the corresponding commands.

[0043] As a preferred embodiment of the present invention, the built-in battery is connected to the main controller via a power interface to provide power to the main controller.

[0044] (III) Beneficial Effects

[0045] The purpose of this invention is to provide a variable fertilization control method and system based on a backpack fertilizer spreader, which has the following beneficial effects: Regarding fertilization accuracy, satellite navigation is used to determine the operation location. Combined with plot grid division and fertility analysis, a discharge model is established based on multiple factors such as fertilizer type and walking speed. This allows for precise control of the fertilizer application amount according to the actual needs of different plots, effectively avoiding the uneven fertilization problem caused by traditional experience-based fertilization. This significantly improves fertilizer utilization, reduces fertilizer waste, and lowers environmental pollution to soil, water, and air caused by fertilizer runoff. It provides a more scientific nutrient supply for crop growth, promoting crop yield and quality improvement. Regarding remote monitoring, the main controller's 4G chip transmits operation information to a cloud platform. Operators can remotely view information such as operation location, speed, and fertilizer application amount in real time. They can also optimize the operation process and adjust fertilization strategies using the operation mileage and area calculated by the cloud platform. Managers can also centrally monitor multiple devices, greatly improving operation management efficiency and transparency. Based on the design of the backpack fertilizer spreader and its automatic and manual dual-mode operation, it has wide applicability and can be used in farmland of different sizes and in various terrain conditions. It meets the needs of different operators and scenarios, and effectively promotes agricultural fertilization towards precision, intelligence, efficiency and sustainability. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the structure of the present invention.

[0047] Figure 2 This is a front sectional view of the material bucket structure of the present invention.

[0048] Figure 3 This is a schematic diagram of the rear view of the material bucket structure of the present invention.

[0049] Figure 4 This is a schematic diagram of the main controller of the present invention.

[0050] Figure 5 This is a schematic diagram of the system of the present invention.

[0051] Figure 6 This is a flowchart illustrating the software functions of the present invention.

[0052] In the diagram: 1-Material bucket, 2-Shoulder strap, 3-Spindle, 4-Power supply compartment, 5-First motor, 6-Speed ​​sensor, 7-Built-in battery, 8-Motor drive module, 9-GPS receiver, 10-Power switch, 11-Discharge channel, 12-Discharge hose, 13-Handle, 14-Discharge bin, 15-Second motor, 16-Turntable, 17-GPS antenna, 18-Sensor interface, 19-Motor drive interface, 20-Power interface, 21-GPS antenna interface, 22-Button, 23-Main controller, 24-LCD screen. Detailed Implementation

[0053] The following will refer to the appendix in the examples of this invention. Figures 1-6 The technical solutions in the embodiments of this invention are clearly and completely described. Obviously, the described embodiments are only some embodiments of this invention, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0054] This invention provides a technical solution:

[0055] A variable fertilizer application control method based on a backpack fertilizer spreader includes the following steps:

[0056] I. Adaptive Raster Division:

[0057] 1. Land parcel data acquisition and preprocessing:

[0058] 1.1 The satellite navigation module is activated. Its GPS receiver uses TTL data to accurately calculate the latitude and longitude coordinates of each vertex of the land parcel boundary. Multiple boundary point coordinates are recorded sequentially in a clockwise or counterclockwise direction, such as point A (longitude 110.12°, latitude 30.56°) and point B (longitude 110.15°, latitude 30.53°). The approximate shape of the land parcel is determined by these coordinate points, and the area of ​​the land parcel is calculated using geometric calculation methods.

[0059] 1.2 Historical Fertilization Data Processing: Historical fertilization data records for this plot are retrieved from the system storage unit. This data includes information on fertilizer application amounts, fertilizer types, and corresponding crop yields at different times and locations within the plot. A comprehensive review of this data is conducted. If a data record shows a fertilizer application amount significantly exceeding the normal range (e.g., more than five times the average fertilizer application amount for similar plots) without a reasonable explanation, this data is marked as abnormal. Abnormal data is further verified for authenticity. If it is determined to be erroneous, data correction or deletion methods are employed to ensure the accuracy and reliability of the data, providing a valid basis for subsequent analysis.

[0060] 2. Analysis of land parcel shape and fertility:

[0061] 2.1 Calculation of irregularity of shape: Based on the obtained coordinate points of the land parcel boundary, the geometric algorithm of polygon analysis is used to calculate the irregularity index of the land parcel shape.

[0062] This includes calculating the concavity and convexity of the boundary by analyzing the changes in the angles between adjacent boundary segments. If there are many alternating acute and obtuse angles, it indicates that the shape of the land parcel is relatively complex. These factors are combined and weighted to obtain a numerical index representing the degree of irregularity of the shape, with a value between 0 and 1. 0 indicates a very regular shape (such as a rectangle), and 1 indicates an extremely complex shape.

[0063] 2.2 Fertility Distribution Analysis: Using historical fertilization data, a spatial interpolation algorithm (Kriging interpolation) was employed to analyze the distribution and changes in fertility within the plot. This included the following steps:

[0064] 2.2.1 Data preparation: First, collect historical fertilization data and existing fertility monitoring data within the plot. These data include the location coordinates (such as latitude and longitude) of each data point and the corresponding fertility values, such as the content of nutrients such as nitrogen, phosphorus, and potassium.

[0065] 2.2.2 Determining the Variance Model: Based on the characteristics of the data and the actual conditions of the plots, a Gaussian model was selected to describe the spatial correlation of fertility data. Through data analysis and fitting, the most suitable variability function and its parameters were determined to accurately characterize the spatial variation of fertility data.

[0066] 2.2.3 Constructing the Covariance Matrix: Based on the Gaussian model and the location information of the data points, calculate the covariance between any two data points and construct the covariance matrix. The covariance matrix reflects the degree of spatial correlation between data points.

[0067] 2.2.4 Solving for weighting coefficients: Based on the covariance matrix and the fertility values ​​of the known data points, the weighting coefficients of each known data point for the fertility estimate at the unknown location are solved through a system of linear equations.

[0068] 2.2.5 Calculating Fertility Estimates: For any unknown location within the plot, multiply the fertility values ​​of the surrounding known data points by their respective weighting coefficients and sum them to obtain the fertility estimate for that unknown location. By performing this calculation on all unknown locations within the plot, the fertility estimate matrix for the entire plot can be obtained.

[0069] 2.2.6 Constructing a Fertility Distribution Surface: The calculated fertility estimation matrix is ​​mapped to the coordinate system of the plots. Using Geographic Information System (GIS) tools, the fertility estimation value at each location is mapped to the corresponding spatial location, thus constructing a continuous fertility distribution surface. This surface visually displays the distribution and changes in fertility within the plots, including high and low fertility areas and gradient changes, providing an important basis for subsequent raster division and fertilization strategy formulation.

[0070] For example, for a certain coordinate point (longitude 110.13°, latitude 30.54°), the nitrogen content is calculated to be 15 mg / kg and the phosphorus content is 8 mg / kg, thus providing an intuitive understanding of the variation trend of fertility in different areas of the plot and the gradient distribution of fertility.

[0071] 3. Determine the grid size:

[0072] 3.1 Threshold Setting: Based on extensive experimental data and practical experience, thresholds are set for the degree of shape irregularity and fertility change. For example, the threshold for shape irregularity is set to 0.5. If the calculated irregularity index of the plot exceeds 0.5, the plot is considered to have a complex shape. Thresholds for fertility change are also set, such as a fertility standard deviation threshold of 10. If the standard deviation of fertility data within the plot is greater than 10, it indicates a large change in fertility.

[0073] 3.2 Size Determination and Resource Considerations: The calculated indicators of plot shape irregularity and fertility variation are compared with the set thresholds. If the plot shape is highly irregular and the fertility variation is large (both exceeding the threshold), it is identified as a complex area, and a smaller grid size is initially set, such as a square grid with a side length of 2-5 meters. Conversely, if the plot is relatively regular and the fertility is relatively uniform (both below the threshold), a larger grid size is used, such as a square grid with a side length of 10-20 meters.

[0074] 4. Grid division:

[0075] 4.1 Starting Point and Direction Selection: Determine the starting point for dividing the grid, starting from one corner of the plot, selecting the top left corner as the starting point. Simultaneously determine the dividing direction, such as dividing sequentially from left to right and from top to bottom.

[0076] 4.2 Conventional Grid Division: Based on the determined grid size, grids are divided sequentially along a preset direction, starting from the initial point. For example, if the determined grid side length is 5 meters, a grid boundary point is determined every 5 meters horizontally, starting from the initial point; similarly, a boundary point is determined every 5 meters vertically, thus forming a series of square grids. During the division process, the number of each grid and its corresponding boundary coordinate range are recorded to ensure accurate identification and positioning of each grid later.

[0077] 4.3 Edge Grid Processing: Due to the irregularity of the plot boundaries, some grids may extend beyond the plot area or fail to completely cover the plot edge. For portions extending beyond the plot area, a trimming method is used to remove the excess portion, ensuring the grid is entirely within the plot. For example, if the upper right corner of a grid extends beyond the plot boundary, a trimming line is determined based on the plot boundary curve, and the excess portion is removed. For cases where the grid cannot completely cover the plot edge, the grid shape is adjusted or merged based on the edge shape and remaining space. For instance, multiple small edge grids can be merged into a larger grid with a shape that better fits the edge; or the boundary of an edge grid can be adjusted to coincide with the plot boundary. Changing the grid shape (e.g., to a trapezoid or triangle) better adapts to the irregularities of the plot edge, ensuring each plot area is reasonably divided into the corresponding grid, thus improving the accuracy and effectiveness of grid division.

[0078] II. Soil fertility measurement

[0079] 1. Feature extraction and data preparation:

[0080] 1.1 Soil fertility feature extraction: For each grid cell, the content information of major nutrient elements such as nitrogen, phosphorus, and potassium was obtained from the soil fertility database or previous detection data. The mean and standard deviation of these nutrient contents were calculated to reflect the average level and degree of variation of soil fertility within the grid cell.

[0081] 1.2 Topographic Feature Extraction: Topographic information within a grid is obtained using topographic survey data or satellite remote sensing data. The average elevation and standard deviation of elevation within the grid are calculated to understand the topographic relief. For example, if the average elevation within a grid is 50 meters and the standard deviation is 10 meters, it indicates that the terrain in that area has some undulation. Furthermore, the slope information of the grid is calculated by analyzing the ratio of the elevation difference to the horizontal distance at different locations within the grid to determine the average slope and the range of slope variation. For example, the average slope is 5° and the maximum slope is 10°.

[0082] 1.3 Crop Planting History Feature Extraction: The crop planting database is queried to obtain information on the types of crops previously planted in this raster. Different crop types are coded, for example, wheat is coded as 1, corn as 2, etc. The frequency of each crop planted in the past is calculated, such as wheat planted 3 times and corn planted 2 times in the past 5 years. Simultaneously, crop yield data for each planting is obtained, and the average yield and standard deviation of yield are calculated to assess the productivity of this raster under different crop planting conditions. This data related to crop planting history constitutes another part of the feature vector.

[0083] 1.4 Feature Vector Processing: The extracted multi-dimensional feature data, such as soil fertility, topography, and crop planting history, are combined in a specific order to form the feature vector corresponding to each grid cell. The feature vector of a grid cell is [mean nitrogen content, standard deviation of nitrogen content, mean phosphorus content, standard deviation of phosphorus content, mean potassium content, standard deviation of potassium content, average altitude, standard deviation of altitude, average slope, standard deviation of slope, crop type code, planting frequency, average yield, standard deviation of yield].

[0084] 2. Clustering Algorithm Selection and Initialization:

[0085] Clustering algorithm selection: The K-Means clustering algorithm is relatively simple and efficient when processing large-scale data, so this algorithm was chosen to cluster the raster.

[0086] Cluster size K estimation and setting: The cluster size K is estimated based on the historical planting history, topography, and approximate distribution of soil fertility of the plot. For example, if the plot has historically been planted with three main types of crops, and the topography and fertility have certain regional characteristics, K can be initially set to 3-5. Simultaneously, some empirical rules can be referenced, such as the elbow rule. By calculating the clustering error index (such as the sum of squares within clusters) under different K values, a curve showing the relationship between K value and the error index can be plotted. The K value corresponding to the elbow position of the curve is usually a more suitable cluster size selection. However, the final K value still needs to be adjusted and determined based on the actual situation and expert experience.

[0087] Initial Cluster Center Selection: After determining the number of clusters K, K grid cells are randomly selected from all grid cells as initial cluster centers. The feature vectors of these initial cluster centers will serve as the initial representative points for each cluster, and subsequent grid cells will be clustered based on their distance from these centers. For example, the first randomly selected cluster center grid cell may have the feature vector [10mg / kg, 2mg / kg, 8mg / kg, 1mg / kg, 15mg / kg, 2mg / kg, 40m, 5m, 3°, 2°, 1, 0.6, 400kg / mu, 50kg / mu], which will represent the central features of a cluster in the initial stage.

[0088] 3. Clustering Iterative Calculation:

[0089] 3.1 Distance Calculation and Cluster Assignment: For each grid cell, calculate the distance between its feature vector and each cluster center. The Euclidean distance formula is used for calculation.

[0090]

[0091] Where X i C is the feature vector of the i-th grid. jLet be the feature vector of the j-th cluster center, and n be the dimension of the feature vector. Calculate the distance between them. Assign each grid cell to the nearest cluster. After assigning all grid cells, a preliminary clustering result is formed.

[0092] 3.2 Cluster Center Update: After completing one round of grid allocation, the center of each cluster is recalculated. For each cluster, the feature vectors of all grid cells belonging to that cluster are averaged to obtain the new cluster center feature vector. For example, if a cluster has 10 grid cells, the average nitrogen content of these 10 grid cells is added together and divided by 10 to obtain the new average nitrogen content of the cluster center. This process is repeated to update the average values ​​of all feature dimensions, forming new cluster centers.

[0093] 3.3 Iterative Loop: Repeat the above steps of distance calculation, cluster assignment, and cluster center update to form an iterative loop. Each iteration optimizes the clustering results until a stopping condition is met. The stopping condition is a maximum number of iterations; when the maximum number of iterations is reached, the iteration process stops regardless of whether the cluster centers have changed.

[0094] 4. Fertilization strategy formulation:

[0095] 4.1 Calculation of Statistical Indicators within Clusters: After clustering is completed, for each cluster, statistical indicators such as the average fertilizer requirement and average soil fertility of its internal cells are calculated. For example, the average nitrogen content and average phosphorus content of all cells within a cluster are calculated as the average soil fertility indicator for that cluster. Based on historical fertilization data and crop yield data, the average amount of fertilizer required per unit area within the cluster is analyzed and calculated, such as the grams of nitrogen, phosphorus, and potassium fertilizer required per square meter. These data will serve as an important basis for formulating fertilization strategies.

[0096] 4.2 Determination of Preliminary Fertilization Strategy: Based on statistical indicators such as average soil fertility and average fertilizer requirement of each cluster, a preliminary fertilization strategy is formulated for each cluster. If a cluster has low average soil fertility and high average fertilizer requirement, the amount of fertilizer applied can be appropriately increased, and the fertilizer formula can be adjusted according to the main elements lacking in soil fertility, increasing the proportion of the corresponding element fertilizer. For example, for clusters with low nitrogen content, the amount of nitrogen fertilizer applied can be increased, and phosphorus and potassium fertilizers can be reasonably combined to determine a preliminary fertilization range and fertilizer formula combination.

[0097] 4.3 Fine-tuning of Special Grids: For individual special grids within a cluster, further analysis is conducted to fine-tune the fertilization strategy based on their unique circumstances. For example, if a grid belongs to a cluster but has particularly low soil fertility (e.g., nitrogen content far below the cluster average) or special terrain (e.g., located in a low-lying, waterlogged area that may lead to fertilizer loss), the fertilizer application rate for that grid is appropriately increased or the fertilizer formula is adjusted based on the initial fertilization strategy to better suit the grid's specific conditions. This ensures that each grid receives precise and reasonable fertilization, improving fertilization effectiveness and crop yield.

[0098] III. Information Collection and Transmission: The satellite navigation module determines the operator's time, location, and speed information and inputs it to the main controller; the speed measurement module collects the current discharge speed of the equipment and inputs it to the main controller.

[0099] IV. Information Processing: The serial communication unit of the main controller receives GPS data from the satellite navigation module and parses out relevant information to obtain the corresponding fertilizer requirements and integrate the requirements with walking speed and fertilizer density information.

[0100] Specifically, the following steps are included:

[0101] 1. GPS Data Reception and Preliminary Analysis: The serial communication unit in the main controller receives GPS data from the satellite navigation module. The received GPS data is preliminarily analyzed to extract raw positioning information, time information, latitude and longitude information, and speed information.

[0102] 2. Fertilizer requirement determination: Obtain the amount of fertilizer required for the current grid based on the information from soil fertilizer calculation.

[0103] 3. Combine with other information for comprehensive processing: Combine the fertilizer demand information of the current grid with the walking speed information of the operator parsed from GPS data, and the fertilizer density determined according to the fertilizer type in automatic mode, and other information for comprehensive processing.

[0104] V. Discharge Rate Calculation: Based on the selected fertilizer type, the software combines different fertilizer densities and operator walking speed to establish a discharge model, including the following steps:

[0105] 1. Determine fertilizer characteristic parameters: Based on the type of fertilizer selected by the operator, query the pre-stored characteristic parameters of that fertilizer, such as density, in the system. Different fertilizer types may have significantly different densities, which will directly affect the calculation of the output.

[0106] 2. Obtain walking speed information: Extract the operator's walking speed from the information transmitted from the satellite navigation module to the main controller. Walking speed is a dynamically changing value, and the system needs to obtain it in real time to ensure the accuracy of the material output speed.

[0107] 3. Establish a mathematical model: Taking into account fertilizer characteristic parameters and walking speed, construct a discharge model. Calculate on a grid-by-grid basis, first determining the relevant parameters of each small grid, such as grid area and required fertilizer amount.

[0108] For a specific small grid, the theoretical velocity calculation formula is:

[0109] V1=(F1 / A1)*W*V / K

[0110] Where V1 represents the theoretical motor speed of grid 1 (r / s), F1 represents the amount of fertilizer required for grid 1 (kg), and A1 represents the area of ​​grid 1 (m²). 2 W represents the width (m), V represents the walking speed (m / s), and K represents the coefficient relationship between the motor speed (r / s) and the output amount (kg / s).

[0111] 4. Calculate the discharge speed: Substitute the specific parameter values ​​into the calculation formula of the discharge model, perform mathematical calculations, and obtain the theoretical motor speed. Based on the correspondence between motor speed and discharge speed, determine the discharge speed. This speed will serve as an important basis for controlling motor operation to achieve precise fertilization.

[0112] VI. Motor control and material discharge:

[0113] 1. Main controller command transmission: Based on the theoretical rotational speed calculated in the previous steps, the main controller determines the required motor speed under the current operating condition to achieve accurate material discharge.

[0114] The main controller sends instructions containing theoretical speed information to the motor control module to ensure that the motor can run at the expected speed.

[0115] 2. The function of the PWM signal generator: The PWM signal generator in the motor control module receives command signals from the main controller. It processes the command signals and converts them into PWM waves with a specific duty cycle. The duty cycle of the PWM wave determines the average voltage of the motor, thereby controlling the motor speed. By adjusting the duty cycle of the PWM wave, precise control of the motor speed can be achieved, making it as close as possible to the theoretical speed.

[0116] 3. Function of the motor driver: The motor driver is connected to the PWM signal generator and receives PWM wave signals.

[0117] The motor driver adjusts and amplifies the input power supply voltage based on the signal characteristics of the PWM wave to provide the motor with appropriate drive voltage and current. The motor driver ensures stable and reliable motor operation and can quickly respond to commands from the PWM signal generator to achieve changes in motor speed.

[0118] 4. Motor Operation and Discharge Control: The motor is connected to a motor driver and receives voltage signals from the driver. The motor converts the received voltage signals into corresponding rotational speeds and adjusts its own rotational speed according to instructions. The motor's rotation drives the discharge device, controlling the discharge speed. For example, in a backpack fertilizer spreader, the motor may drive the fertilizer conveying device through a transmission mechanism, causing the fertilizer to be discharged from the outlet at a specific speed. By precisely controlling the motor speed, accurate adjustment of the discharge speed can be achieved, ensuring that the fertilizer application meets the needs of the current work area and improving the accuracy and efficiency of fertilization.

[0119] VII. Information Transmission and Remote Monitoring:

[0120] 1. Information Collection and Preparation: Various modules in the main controller continuously collect key information during the operation. This includes the operation location information obtained through the satellite navigation module, accurate to specific latitude and longitude coordinates; the operation speed and walking speed obtained from the speed measurement module and the satellite navigation module; the amount of fertilizer spread based on the motor control and material discharge process statistics; the required amount of fertilizer determined from the information processing steps; and the fertilizer type set by the operator in the system.

[0121] 2.4G Chip Information Transmission: The 4G chip in the main controller plays a crucial role, using the MQTT (Message Queuing Telemetry Transport) protocol to send collected job information to the cloud platform. MQTT is a lightweight messaging protocol suitable for resource-constrained devices and unstable network environments, ensuring efficient and reliable information transmission. The transmitted job information includes job location, speed, walking speed, amount of fertilizer applied, required fertilizer amount, and fertilizer type, comprehensively reflecting the real-time status and key parameters of the job.

[0122] 3. Cloud Platform Reception and Processing: After receiving the operation information from the main controller, the cloud platform stores and processes this information. First, the operation information is displayed so that operators and relevant managers can easily check the progress of the operation. Then, further calculations are performed using the received operation location information. By analyzing the operation location at different time points, the operation mileage, i.e., the total distance traveled by the operator during the operation, can be calculated. Simultaneously, based on information such as the operation location and operation width, the operation area can be calculated to understand the land area covered by the operation.

[0123] 4. Remote Monitoring and Decision Support: Operators can access the cloud platform via the internet to remotely monitor the work process. Regardless of their location, they can stay informed about the status and progress of the work. The operational information and calculation results provided by the cloud platform offer decision support to operators. For example, based on the area covered and the amount of fertilizer applied, it can be determined whether the fertilization strategy needs adjustment; based on the work speed and walking speed, the work process can be optimized to improve efficiency. Managers can also use the cloud platform to centrally monitor the operation of multiple devices, enabling rational resource allocation and unified management of operations.

[0124] A variable fertilizer application control system based on a backpack fertilizer spreader includes a control system and an actuator.

[0125] The control system includes:

[0126] 1. The main controller, as the core control unit of the entire system, coordinates and manages the various modules. It receives and processes data from the speed measurement module, the satellite navigation module, and the operating parameters input by the user via buttons. Based on this information, it calculates the required amount of fertilizer and the discharge rate for the plot, and sends instructions to the motor control module.

[0127] The main controller includes an LCD screen, buttons, a 4G chip, a main chip, and a serial communication unit, wherein:

[0128] a. LCD screen: Used to display operational information, allowing users to intuitively understand the current operating status of the system. The displayed content includes the operational mode (such as automatic mode, manual mode, etc.), operational speed (including travel speed and discharge speed, etc.), operational status (operating, paused, faulty, etc.), and set parameters (such as fertilizer type, fertilizer application rate settings, etc.), providing users with real-time feedback so that they can monitor the fertilizer application operation at any time and make adjustments as needed.

[0129] b. Buttons: These allow users to set operating parameters and personalize system operation. They enable settings for discharge acceleration / deceleration, adjusting the fertilizer discharge speed according to actual needs. Fertilizer type settings are also available; different fertilizers may require different application rates and discharge speed control strategies, which the system can adjust accordingly. The application rate setting directly determines the amount of fertilizer applied each time, meeting the diverse fertilizer requirements of different plots.

[0130] c. 4G Chip: Responsible for sending operation information to the cloud platform, enabling remote monitoring and operation management. Utilizing the MQTT protocol, it sends current operation information, such as operation location, speed, walking speed, amount of fertilizer applied, required amount of fertilizer, and fertilizer type, to the cloud platform. The cloud platform can store, process, and display this information, allowing operators to check the operation status anytime, anywhere. Simultaneously, the cloud platform can use the received operation location information to calculate operation mileage and area, providing operators with more comprehensive operation management functions.

[0131] d. Main Chip: As the core processing unit of the main controller, it undertakes the crucial tasks of data reception, processing, storage, and transmission. It receives data from various modules, including positioning, time, latitude and longitude, and speed information from the satellite navigation module, as well as discharge speed information and operational parameters input via buttons from the speed measurement module. It processes this data, such as resolving various information from the GPS data from the satellite navigation module and calculating plot identification and discharge speed using specific algorithms. For example, when dividing a large plot into multiple smaller grids, the main chip determines its current grid location based on positioning information and obtains the required fertilizer amount for that grid. It stores important data for subsequent querying and analysis. Simultaneously, based on the processing results, it sends instructions to the motor control module to control the discharge speed, achieving precise fertilization.

[0132] e. Serial Communication Unit: This unit serves as a connection and communication link between the main controller and the satellite navigation module and speed measurement module. It communicates with the satellite navigation module to obtain geographic location coordinates and speed information, ensuring the main controller can accurately determine the operator's position and operating speed. Connecting to the speed measurement module, it processes the received speed information and sends it to the LCD screen and cloud platform for display, allowing users and the remote monitoring platform to understand the current material output speed and operation progress.

[0133] 2. Speed ​​Measurement Module: Primarily used to measure the discharge speed of the fertilizer application equipment. Through a Hall effect sensor working in conjunction with a magnet mounted on the rotating shaft, it senses changes in the magnetic field and outputs different level signals, thereby calculating the current fertilization speed. This information is then transmitted to the main controller, which adjusts the fertilization operation based on the discharge speed and other factors.

[0134] 3. Satellite Navigation Module: This module determines the operator's location, time, and speed. The GPS receiver uses TTL data to calculate the operator's latitude and longitude coordinates every 100 milliseconds, obtaining the operator's current location and calculating the operator's average speed. This information is input into the main controller, which combines the location information with a pre-defined plot grid to determine the required amount of fertilizer for the plot where the operator is located, while also considering factors such as operating speed for fertilizer application control. The GPS antenna receives GPS signals to ensure the satellite navigation module functions properly.

[0135] 4. Motor Control Module: This module includes a PWM signal generator, a motor driver, and the motor itself. The motor control module controls the motor's operation based on instructions from the main controller. The PWM signal generator receives the instruction signals from the main controller and converts them into PWM waves. The motor driver drives the motor to rotate according to the PWM waves. The motor converts the received voltage signals into corresponding speeds, thereby achieving precise control of the fertilizer discharge rate.

[0136] The system includes a fertilizer applicator, which comprises a power supply compartment with a power switch on its right side. An internal battery is fixed inside the power supply compartment and connected to a first motor via wires. The first motor is connected to a discharge channel via a rotating shaft. The discharge channel is connected to a discharge hopper via a feeding hose. The discharge hopper includes a second motor connected to a turntable. A material hopper is mounted on top of the power supply compartment, with symmetrical shoulder straps fixed to its left and right sides. The material hopper is used to hold solid granular fertilizer. The internal battery in the power supply compartment powers all components of the system. Under control, the first motor drives the fertilizer in the hopper through the discharge channel. The discharge channel connects the hopper and the feeding hose, transporting the fertilizer to the discharge hopper. The feeding hose connects and transports the fertilizer. The discharge hopper includes a second motor and a turntable. The second motor drives the turntable to evenly spread the fertilizer onto the ground at a predetermined width. The shoulder straps facilitate operation by the operator carrying the fertilizer applicator on their back.

[0137] The built-in battery is connected to the speed sensor via wires. The speed sensor is installed on the right side of the discharge channel and works in conjunction with a magnet installed on the rotating shaft. The speed sensor is connected to the main chip via a sensor interface, and the main chip calculates the motor speed.

[0138] The built-in battery is connected to the GPS receiver via wires. The GPS receiver is connected to the GPS antenna via the GPS antenna interface. The GPS receiver is connected to the main chip via the sensor interface. The main chip acquires GPS data and calculates the positioning information, location information, time information, and speed information.

[0139] The built-in battery is connected to the motor drive module via wires. The motor drive module is connected to the main chip via a motor drive interface. The main chip calculates the theoretical rotation speed by combining the obtained position information, speed information, and land information, and sends it to the motor drive module, which then executes the corresponding commands.

[0140] The built-in battery is connected to the main controller via a power interface to provide power to the main controller.

[0141] The workflow of this embodiment is as follows:

[0142] I. Working principle of automatic mode

[0143] 1. Plot Information Processing: The satellite navigation module is activated to accurately acquire the latitude and longitude coordinates of the plot boundary vertices, calculate the plot shape and area, and read historical fertilization data for anomaly processing and organization. Based on the plot boundary coordinates, the degree of shape irregularity is analyzed. Simultaneously, Kriging interpolation is used to analyze fertility distribution using historical fertilization data and fertility monitoring point data, constructing a fertility distribution surface. The grid size is determined based on the shape and fertility status, and grid division is completed, including starting point selection, regular division, and edge processing. Multi-dimensional feature data is extracted from each grid to form a feature vector. The K-Means clustering algorithm is selected, the number of clusters K is estimated and set, and K grids are randomly selected as initial cluster centers to complete the preparatory work for soil fertility measurement.

[0144] 2. Starting the fertilization operation: Fill the bucket with solid granular fertilizer, and the operator carries the fertilizer spreader on their back using the shoulder strap. Turn on the power switch on the right side. At this time, the main controller starts and the GPS module starts working. If the LCD screen on the main controller displays the current walking speed of the operator, it indicates that the GPS module has successfully located the target.

[0145] 3. Information Acquisition and Processing: The satellite navigation module collects the operator's time, location, and speed information via a GPS receiver and antenna, and inputs this information to the main controller. The serial communication unit in the main controller receives the GPS data from the satellite navigation module and calculates the positioning, time, latitude, longitude, and speed information. A large plot of land is divided into multiple smaller grids. A ray is drawn from the positioning point in any direction, and the number of intersections between the ray and each side of the polygon is calculated to determine the current grid cell, thereby obtaining the required amount of fertilizer for that grid cell. This information is then combined with other data, such as fertilizer density determined based on fertilizer type in automatic mode, for comprehensive processing.

[0146] 4. Discharge Speed ​​Calculation: Based on the selected fertilizer type, the software combines different fertilizer densities and the operator's walking speed to establish a discharge model. For example, when the operator is in the first small grid, the theoretical speed of grid 1 is calculated as follows: V1=(F1 / A1)WV / K, where V1 represents the theoretical motor speed of grid 1, F1 represents the amount of fertilizer required for grid 1, A1 represents the area of ​​grid 1, W represents the width, V represents the walking speed, and K represents the coefficient relationship between motor speed (r / s) and discharge amount (kg / s).

[0147] 5. Motor Control and Discharge: The main controller sends the calculated theoretical rotational speed to the motor control module. The PWM signal generator in the motor control module receives and responds to the command signal from the main controller, converting it into a PWM wave. The motor driver is connected to the PWM signal generator, driving the motor to rotate. The motor converts the received voltage signal into a corresponding rotational speed, controlling the discharge speed. The fertilizer in the hopper, driven by the first motor, passes through the discharge channel, which is connected to the feeding hose. The fertilizer flows along the feeding hose into the discharge hopper. The discharge hopper is equipped with a turntable, driven by a second motor, which evenly spreads the incoming fertilizer onto the ground at a set width.

[0148] 6. Information Transmission and Remote Monitoring: The 4G chip in the main controller uses MQTT to send current operation information, including operation location, operation speed, walking speed, amount of fertilizer applied, required amount of fertilizer, and fertilizer type, to the cloud platform. The cloud platform displays the operation information and uses the received operation location information to calculate the operation mileage and operation area, facilitating remote operation management by the operator.

[0149] II. Working principle of manual mode

[0150] 1. System startup and preparation: Same as in automatic mode, turn on the power switch, the main controller starts, and waits for the operator to operate.

[0151] 2. Parameter Setting and Operation: The operator sets operating parameters via buttons on the main controller, including adjusting discharge acceleration / deceleration, fertilizer type, and fertilizer application rate. In manual mode, the system does not automatically calculate the discharge speed based on plot information and walking speed; instead, the operator must manually adjust it according to the actual situation.

[0152] 3. Motor Control and Discharge: When the operator presses the button to accelerate or decelerate the discharge, the main controller sends the corresponding command to the motor control module. The motor control module adjusts the duty cycle of the PWM wave output by the PWM signal generator according to the command, thereby changing the voltage and current supplied to the motor by the motor driver, thus changing the motor speed and controlling the discharge speed. Driven by the motor, the fertilizer enters the discharge hopper through the discharge channel and discharge hose, and is then evenly spread on the ground by the turntable.

[0153] 4. Information Display and Feedback: The main controller's LCD screen displays operational information, including the operating mode (manual mode), operating speed, operating status, and set parameters. The operator can understand the current operating status of the fertilizer spreader based on the information on the screen and adjust the operation accordingly.

[0154] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A variable rate application control method based on a knapsack spreader, characterized by, The method comprises the following steps: Adaptive grid division: obtaining the plot boundary coordinate information through the satellite navigation module, determining the plot shape and area, calculating the plot shape irregularity index, analyzing the plot fertility distribution change using a spatial interpolation algorithm, generating a fertility distribution surface, setting the shape irregularity degree and fertility change threshold, dividing the grid in the preset direction, and processing the edge grid to make it fit the plot boundary; Soil fertility calculation: extracting soil fertility, topography, crop planting history multi-dimensional feature data for each grid, and arranging them into feature vectors; calculating the average fertilizer demand index of the grid by clustering algorithm, and formulating the fertilization strategy; Information collection and transmission: the satellite navigation module determines the time information, position information and speed information of the operator and inputs them to the main controller; the speed measurement module collects the current device discharge speed and inputs it to the main controller; Information processing: the main controller receives the GPS data of the satellite navigation module and analyzes the information, obtains the corresponding fertilizer demand, and integrates the demand with the walking speed and fertilizer density information; Discharge speed calculation: according to the selected fertilizer type, the software combines different fertilizer densities and the walking speed of the operator to establish a discharge model; Motor control and discharge: the main controller sends the calculated theoretical speed to the motor control module, the PWM signal generator in the motor control module receives and responds to the instruction signal from the main controller, and converts it into a PWM wave; the motor driver is connected with the PWM signal generator to drive the motor to rotate, and the motor converts the received voltage signal into corresponding speed to control the discharge speed; In the adaptive grid division step, the spatial interpolation algorithm for analyzing the plot fertility distribution change comprises: Data preparation: collecting historical fertilization data and existing fertility detection point data in the plot; Constructing the covariance matrix: calculating the covariance between any two data points to construct the covariance matrix; Solving the weight coefficient: solving the weight coefficient of each known data point to the unknown position fertility estimation value through a linear equation set; Calculating the fertility estimation value: for any unknown position in the plot, multiply the fertility values of the known data points around it by the corresponding weight coefficients and sum them up to obtain the fertility estimation value of the unknown position; Constructing the fertility distribution surface: corresponding the calculated fertility estimation value matrix to the coordinate system of the plot, using geographic information system tools to map the fertility estimation value of each position to the corresponding spatial position, and constructing the fertility distribution surface; In the soil fertility calculation step, the feature vector is [nitrogen content mean, nitrogen content standard deviation, phosphorus content mean, phosphorus content standard deviation, potassium content mean, potassium content standard deviation, average elevation, elevation standard deviation, average slope, slope standard deviation, crop type code, planting frequency, average yield, yield standard deviation]; In the soil fertility calculation step, the clustering algorithm for calculating the average fertilizer demand index of the grid comprises: Distance calculation and clustering assignment: for each grid, calculate the distance between its feature vector and each cluster center; Cluster center updating: after completing a round of grid assignment, the center of each cluster is recalculated, and for each cluster, all grid feature vectors belonging to the cluster are averaged to obtain a new cluster center feature vector; Iteration loop: repeat the above distance calculation and cluster assignment, cluster center updating steps, form an iterative loop, until the maximum number of iterations is reached.

2. The variable rate application control method based on a knapsack spreader according to claim 1, wherein The formula for calculating the theoretical speed of fertilizer application in the grid in the discharge speed calculation step is: V1= (F1 / A1) *W*V / K, wherein V1 represents the theoretical motor speed of grid 1 (r / s), F1 represents the required fertilizer amount of grid 1 (kg), A1 represents the area of grid 1 (m²), W represents the width (m), V represents the walking speed (m / s), and K represents the coefficient relationship between the motor speed (r / s) and the discharge amount (kg / s).

3. A variable rate fertilization control system for implementing the method of any of claims 1-2, characterized by, The control system comprises a main controller, a speed measurement module, a satellite navigation module and a motor control module connected to the main controller.

4. The variable rate application control system based on a knapsack fertilizer applicator according to claim 3, wherein The main controller comprises an LCD screen, a key, a 4G chip, a main chip and a serial communication unit, the LCD screen is connected with the main chip, used for displaying operation information, the key is connected with the main chip, used for setting operation parameters, the 4G chip is used for sending operation information to the cloud platform, and the main chip is used for receiving, processing, storing and sending data of each module, and identifying plots and calculating discharge speed.

5. The variable rate application control system based on a knapsack fertilizer applicator according to claim 3, wherein The satellite navigation module comprises a GPS receiver and a GPS antenna, the GPS receiver is used for calculating specific longitude and latitude coordinates every 100 milliseconds by using TTL data, obtaining the average speed of the operator through relevant calculation, and inputting the obtained fertilizer application speed signal into the main controller; and the GPS antenna is used for receiving GPS signals.

6. The variable rate application control system based on a knapsack fertilizer applicator according to claim 4, wherein The serial communication unit is connected with the satellite navigation module to communicate and obtain geographic position coordinate information and speed information; and the serial communication unit is connected with the speed measurement sensor to process and send the received speed information to the screen and the cloud platform for display.

7. The variable rate application control system based on a knapsack fertilizer applicator according to claim 3, characterized in that: The execution mechanism comprises a fertilizer application device, the fertilizer application device comprises a barrel (1), the barrel (1) is symmetrically fixed with a shoulder strap (2) on the left and right sides, a power supply compartment (4) is arranged at the bottom of the barrel (1), a power switch (10) is mounted on the left side of the power supply compartment (4), a built-in battery (7) is fixed in the power supply compartment (4), the built-in battery (7) is connected with a first motor (5) through a wire, the first motor (5) is connected with a discharge channel (11) through a rotating shaft (3), the discharge channel (11) is connected with a discharge compartment (14) through a discharging hose (12), and the discharge compartment (14) comprises a second motor (15). The built-in battery (7) is connected with a speed measurement sensor (6) through a wire, the speed measurement sensor (6) is installed on the right side of the discharge channel (11) and cooperates with a magnet installed on the rotating shaft, and the speed measurement sensor (6) is connected with the main chip through a sensor interface (18).

Images