Variable spray method and system based on crop distribution perception

Abstract

The application discloses a kind of variable spray method and system based on crop distribution perception, the method includes: construct the fog drop distribution model of nozzle under different operation height, operation parameter;Perceive field target crop information, determine actual operation height and theoretical demand liquid distribution model;Optimal fog drop distribution model and corresponding operation parameter are screened by vector matching;Solve global optimal spraying time, realize accurate spraying.The application solves the problems of traditional spraying liquid waste, uneven spraying by accurate perception of crop distribution and dynamic matching of fog drop distribution model, improves the pertinence and resource utilization rate of spraying, and is suitable for various crop field accurate pesticide application scene.

Description

Technical Field

[0001] This invention relates to the field of agricultural plant protection spraying technology, specifically to a variable spraying method and system based on crop distribution sensing, which is particularly suitable for precision pesticide application scenarios in field crops. Background Technology

[0002] In agricultural production, spraying is a crucial step in controlling crop diseases and pests and ensuring crop yield. Traditional spraying techniques often employ fixed operating parameters (such as fixed pressure, fixed nozzle angle, and fixed spraying time) for uniform spraying across the entire area, without considering differences in crop distribution, plant height, and the severity of pests and diseases in the field. This extensive spraying method has several problems: Firstly, in areas where crops are densely packed or pests and diseases are severe, the supply of pesticide solution is insufficient, failing to achieve effective control; secondly, in areas where crops are sparse or absent, excessive spraying leads to pesticide waste and may also cause soil and water pollution; furthermore, the fixed operating height does not match the actual height of the crops, easily resulting in uneven droplet deposition, further reducing the spraying effect.

[0003] While existing variable-rate spraying technologies attempt to achieve precise application by adjusting spraying parameters, they often rely on single-crop information (such as crop location or disease severity) for parameter adjustments. They lack a dynamic matching mechanism between droplet distribution and crop needs, and fail to coordinate optimization of operating height, spraying parameters, and spraying time. This results in insufficient spraying accuracy, pesticide waste, and poor control effects. Therefore, there is an urgent need for a variable-rate spraying technology that can accurately sense crop distribution and needs, dynamically match droplet distribution models, and collaboratively optimize operating parameters and spraying time to address the shortcomings of existing technologies. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] In a first aspect, the present invention provides a variable spraying method based on crop distribution sensing, comprising:

[0006] S1: Construct a droplet distribution model for the nozzle under different operating heights and different operating parameters;

[0007] S2: Sensing and acquiring information about the target crop in the field, and determining the actual operating height of the nozzle and the theoretical required amount of pesticide solution distribution model for the target crop based on the target crop information;

[0008] S3: Calculate the matching degree between several droplet distribution models of the nozzle at the corresponding actual working height and the theoretical required pesticide volume distribution model of the target crop, and determine the droplet distribution model with the highest matching degree, which is denoted as the target droplet distribution model;

[0009] S4: The optimal operating parameters are the combination of operating parameters under the target droplet distribution model.

[0010] S5: Based on the target droplet distribution model, solve for the global optimal spraying time of the nozzle, and control the nozzle to perform spraying operations according to the optimal operating parameters and the global optimal spraying time.

[0011] Furthermore, S1 specifically includes:

[0012] S101: The experiment simulates the spray deposition of the nozzle on the target plane under different operating heights and various combinations of operating parameters; the size of the target plane is preset according to the size of the single target operating area of ​​the nozzle;

[0013] S102: Collect the droplet deposition distribution of each grid cell of the nozzle on the target plane under each working condition;

[0014] S103: Calculate and label the droplet deposition per unit time for each grid cell in the target plane, and generate a droplet distribution model of the nozzle under this condition;

[0015] S104: Associate and map the droplet distribution model with the corresponding combination of operating parameters and operating height, and store the data.

[0016] Furthermore, S2 specifically includes:

[0017] S201: Sensing and acquiring field target crop information within the area to be operated by the nozzle, the target crop information including target crop image and target crop height data;

[0018] S202: Determine the actual operating height of the sprinkler head based on the sprinkler head position and the target crop height data;

[0019] S203: Construct a target crop distribution model for the target crop image, wherein the size and grid cells of the target crop distribution model are the same as those of the fog droplet distribution model;

[0020] S204: Image recognition determines the number of target crops and disease information of each target crop in each grid cell of the target crop distribution model, and determines the spraying demand of each target crop based on the disease information of each target crop and the preset crop disease spraying volume model;

[0021] S205: Mark the spraying demand of each target crop in each grid cell of the target crop distribution model, and generate a theoretical pesticide distribution model for the target crop.

[0022] Furthermore, S3 specifically includes:

[0023] S301: Based on the theoretical pesticide spray volume distribution model of the target crop, calculate the total pesticide spray volume distribution model of the target crop in each grid cell, and vectorize the total pesticide spray volume distribution model of the target crop in each grid cell to generate the theoretical pesticide spray volume distribution vector.

[0024] S302: Determine the droplet distribution model of the nozzle at its actual operating height, and vectorize the droplet deposition amount per unit time of each grid cell of the droplet distribution model to generate a droplet distribution model vector;

[0025] S303: Calculate the matching degree between the theoretical required liquid volume distribution vector and the droplet distribution model vector, and determine the droplet distribution model with the highest matching degree with the theoretical required liquid volume distribution model, which is denoted as the target droplet distribution model; the matching degree calculation method includes the cosine similarity method or the root mean square error method.

[0026] Furthermore, S5 specifically includes:

[0027] S501: Obtain the droplet deposition rate per unit time for each grid cell in the target droplet distribution model. ;

[0028] S502: Obtain the total spraying demand of the target crop in each grid cell of the theoretical pesticide distribution model. ;

[0029] S503: Solve for the optimal spraying time in each grid cell of the theoretical pesticide distribution model for the target crop. , ;

[0030] S504: Optimal spraying time for all grid cells The average is calculated to obtain the globally optimal spraying time T, and the nozzles are controlled to perform precise spraying operations according to the optimal operating parameters and the globally optimal spraying time T.

[0031] Secondly, a variable spraying system based on crop distribution perception is provided, including: a droplet distribution model establishment module, an intelligent sensing module, a determination module, a pesticide volume distribution model generation module, a model matching module, a time solution module, a control module, and a variable spraying execution module;

[0032] The droplet distribution model building module is used to construct droplet distribution models of the nozzle under different operating heights and different combinations of operating parameters.

[0033] The intelligent sensing module is used to sense and acquire information about target crops in the field.

[0034] The determining module is used to determine the actual operating height of the nozzle based on the field target crop information;

[0035] The pesticide solution distribution model generation module is used to generate a theoretical pesticide solution distribution model for the target crop based on the field target crop information.

[0036] The model matching module is used to calculate the matching degree between the droplet distribution model of the nozzle at the actual working height and the theoretical required liquid volume distribution model, and to determine the droplet distribution model with the highest matching degree, which is denoted as the target droplet distribution model.

[0037] The time calculation module is used to solve for the globally optimal spraying time based on the target droplet distribution model;

[0038] The control module is used to control the nozzles of the variable spray execution module to perform variable spray operations according to the optimal operating parameters corresponding to the target droplet distribution model and the global optimal spraying time.

[0039] Furthermore, the intelligent sensing module is a depth camera or a LiDAR.

[0040] Furthermore, the nozzle is a PWM variable nozzle, a hydraulic nozzle, or a centrifugal nozzle.

[0041] The beneficial effects of this invention are as follows:

[0042] 1. This invention uses depth cameras or lidar to sense crop distribution, height, and pest and disease information, and constructs a refined theoretical demand model for pesticide liquid distribution to accurately depict crop needs; by establishing a droplet distribution model library for nozzles, and using vector matching algorithms to select the droplet distribution model and operating parameters that best match crop needs, it solves the problems of fixed parameters and poor adaptability in traditional spraying.

[0043] 2. Based on the correspondence between crop demand and droplet deposition, this invention solves the globally optimal spraying time, avoiding overspraying or underspraying, reducing pesticide waste, and effectively reducing environmental pollution; it achieves coordinated optimization of operating height, spraying parameters, and spraying time, ensuring that droplets are uniformly deposited in the target crop area, thereby improving the effect of pest and disease control.

[0044] 3. The variable spraying method of this invention supports multiple nozzle types such as PWM variable nozzles and hydraulic nozzles, and can be adapted to field operation scenarios of various crops such as corn, wheat, and vegetables, with strong versatility. Attached Figure Description

[0045] Figure 1 This is a flowchart of the variable spraying method based on crop distribution sensing of the present invention.

[0046] Figure 2This is a schematic diagram illustrating the structural principle of a variable spraying system based on crop distribution sensing, according to an embodiment of the present invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Example

[0048] This invention provides a variable spraying method based on crop distribution sensing, such as... Figure 1 As shown, the implementation steps are as follows:

[0049] S1: Construct droplet distribution models for the nozzle under different operating heights and parameters; specifically including:

[0050] S101: The experiment simulates the spray deposition of the nozzle on a target plane under different operating heights and various combinations of operating parameters. The size of the target plane is preset according to the size of the single target operating area of ​​the nozzle. Specifically, the size of the single target operating area of ​​the nozzle is set according to the coverage range of the nozzle during normal operation. For example, the single target operating area of ​​the nozzle is set to a rectangular area with a horizontal width of 2m and a forward direction coverage of 1.5m. In actual operation, the nozzle may produce droplets outside the target operating area, which is ignored in this invention. The nozzle in the embodiment of this invention can be a PWM variable nozzle, a hydraulic nozzle, or a centrifugal nozzle, or other types of nozzles. The corresponding operating parameters are different when the nozzle type is different. For example, when the nozzle is a PWM variable nozzle, its operating parameters include spray pressure, nozzle angle, and PWM duty cycle. When a hydraulic nozzle is used, the operating parameters include spray pressure, nozzle angle, and hydraulic valve opening. If a centrifugal nozzle is used, the core operating parameters include spray pressure, nozzle angle, and centrifugal speed. In this embodiment of the invention, the operating height refers to the height of the nozzle above the target plane. The target plane refers to the plane formed by the average position of the top of the target crop within the operating area, used to simulate the main operating interface in real-world spraying operations. The operating height simulated in the experiment can be selected according to the height range of the nozzle from the target crop during actual plant protection machinery spraying operations, such as operating height gradients of 10cm, 20cm, 30cm, 40cm, and 50cm. In specific experiments, the above-mentioned target plane and nozzle installation mechanism can be built on an indoor experimental platform. The nozzle installation mechanism needs to have three-dimensional adjustment functions (horizontal position, vertical height, and angle adjustment). The experimental conditions are set according to the preset operating height and operating parameter combination. The nozzle is started to spray, and the spraying duration is adjusted according to the nozzle type (e.g., 30s for a PWM variable nozzle and 20s for a centrifugal nozzle) to ensure that stable droplet deposition is received in all areas of the target plane.

[0051] S102: Collect the droplet deposition distribution of the nozzle on each grid cell of the target plane under each operating condition; in this embodiment of the invention, the target plane is pre-divided into grid cells. A droplet collection card or an electronic droplet sensor can be used to collect the droplet deposition distribution on the target plane. If a droplet collection card is used for droplet collection, the size of the collection card is set according to the size of the grid cell to ensure that one collection card corresponds to the center of each grid cell; after spraying, the collection card is scanned by an image scanner, and software is used to analyze the number of droplets and the droplet diameter to calculate the total droplet deposition of a single collection card.

[0052] S103: Calculate and mark the droplet deposition amount per unit time for each grid cell in the target plane, and generate the droplet distribution model of the nozzle under this condition; for each grid cell, calculate the droplet deposition amount per unit time as the ratio of the droplet deposition distribution amount to the experimental duration based on the collected droplet deposition distribution amount and the experimental duration.

[0053] S104: Associate and map the droplet distribution model with the corresponding combination of operating parameters and operating height, and store the data. A droplet distribution model database can be established, with database fields including: nozzle type (e.g., PWM variable nozzle), operating height (cm), operating parameter 1 (e.g., spray pressure), operating parameter 2 (e.g., nozzle angle), and operating parameter 3 (PWM duty cycle). It supports quick retrieval of the corresponding droplet distribution model by operating height.

[0054] In actual sprinkler operations, the distribution of spray droplets varies depending on the sprinkler's operating parameters and operating height. Constructing a sprinkler droplet distribution model can effectively simulate the actual droplet distribution during sprinkler operation and match it with the distribution of crops to achieve precise pesticide application.

[0055] S2: Sensing and acquiring field target crop information, and determining the actual operating height of the sprayer and the theoretical pesticide solution distribution model required by the target crop based on the target crop information; specifically including:

[0056] S201: Sensing and acquiring field target crop information within the area to be operated by the nozzle, the target crop information including target crop images and target crop height data; specifically, the target crop information in the area to be operated can be sensed by a depth camera or lidar.

[0057] S202: Determine the actual operating height of the sprinkler head based on the sprinkler head position and the target crop height data; In this embodiment of the invention, the initial height position of the sprinkler head is fixed and known, but it can also be dynamically adjusted to determine the sprinkler head height; Since the area to be operated by the sprinkler head includes multiple target crops, the average height of the target crops can be obtained by averaging the height data of each target crop, and the actual operating height of the sprinkler head can be obtained by subtracting the sprinkler head height from the average height of the target crops.

[0058] S203: Construct a target crop distribution model for the target crop image. The size and grid cells of the target crop distribution model are the same as those of the droplet distribution model. The area of ​​the target crop distribution model is defined from the target crop image based on the nozzle position. The droplets during nozzle operation can be assumed to be entirely within the area of ​​the target crop distribution model. This embodiment achieves adaptive matching between droplet distribution and target crop distribution by constructing a target crop distribution model with the same size and grid cells as the droplet distribution model.

[0059] S204: Image recognition determines the number of target crops and disease information of each target crop in each grid cell of the target crop distribution model. Based on the disease information of each target crop and the preset crop disease spray volume model, the spray demand of each target crop is determined. Specifically, through image recognition technology, the number of crops in each grid cell of the target crop distribution model and the disease type, severity, and other disease information of each crop are determined. The variable spray system of this invention has a preset crop disease spray volume model, which can characterize the different spray demand corresponding to different disease types and severity. Different disease types correspond to different spray demand, and the severity can be classified as mild, moderate, and severe. For example, mild disease requires 3 mL of pesticide solution per plant, moderate disease requires 5 mL, and severe disease requires 7 mL. The spray demand of each target crop is calculated based on the crop disease spray volume model.

[0060] S205: Mark the spraying demand of each target crop in each grid cell of the target crop distribution model, and generate a theoretical pesticide distribution model for the target crop.

[0061] S3: Calculate the matching degree between several droplet distribution models at the corresponding actual operating height and the theoretical pesticide volume distribution model required by the target crop, and determine the droplet distribution model with the highest matching degree, denoted as the target droplet distribution model; specifically including:

[0062] S301: Based on the theoretical pesticide spray volume distribution model of the target crop, the total spray volume of the target crop in each grid cell is calculated by summing the theoretical pesticide spray volume distribution model of the target crop, and the total spray volume of the target crop in each grid cell is vectorized to generate the theoretical pesticide spray volume distribution vector.

[0063] The theoretical required drug volume distribution model includes: Each grid cell The corresponding total demand for spraying the target crop is Calculate the total target crop spraying requirements for all grid cells, prioritizing them by row or column. The values ​​are arranged sequentially to form a set with The column vector or row vector of each element is denoted as the distribution vector of the theoretical required liquid volume. .

[0064] S302: Determine the droplet distribution model of the nozzle at its actual operating height, and vectorize the droplet deposition per unit time of each grid cell in the droplet distribution model to generate a droplet distribution model vector; each droplet distribution model also includes Each grid cell Corresponding droplet deposition rate per unit time , adopt and The same order, will Arranged into droplet distribution model vectors .

[0065] in, .

[0066] S303: Calculate the matching degree between the theoretical required drug volume distribution vector and the droplet distribution model vector, and determine the droplet distribution model with the highest matching degree to the theoretical required drug volume distribution model, denoted as the target droplet distribution model; the matching degree calculation method includes the cosine similarity method or the root mean square error method. When using the cosine similarity method, the cosine value of the angle between the theoretical required drug volume distribution vector and the droplet distribution model vector in the direction can be calculated as the similarity value calculation formula, as follows:

[0067]

[0068] Iterate through all candidate droplet distribution models in the model library and calculate the similarity value for each model. The model with the highest similarity value is selected as the target droplet distribution model. The root mean square error method for calculating similarity is a mature existing method, and will not be elaborated upon in this invention.

[0069] S4: The optimal operating parameters are the combination of operating parameters under the target droplet distribution model.

[0070] S5: Based on the target droplet distribution model, solve for the globally optimal spraying time of the nozzle, and control the nozzle to perform spraying operations according to the optimal operating parameters and the globally optimal spraying time. Specifically, this includes:

[0071] S501: Obtain the droplet deposition rate per unit time for each grid cell in the target droplet distribution model. ;

[0072] S502: Obtain the total spraying demand of the target crop in each grid cell of the theoretical pesticide distribution model. ;

[0073] S503: Solve for the optimal spraying time in each grid cell of the theoretical pesticide distribution model for the target crop. , ;

[0074] S504: Optimal spraying time for all grid cells The average is calculated to obtain the globally optimal spraying time T. The nozzles are then controlled to perform precision spraying operations according to the optimal operating parameters and the globally optimal spraying time T. After the nozzles complete the spraying of the target crop in the work area at a certain point using the variable spraying method based on crop distribution perception in this embodiment, they can move to the next point under the support of the walking unit, and repeat the variable spraying method based on crop distribution perception in this embodiment to spray the target crop at the next point, and so on in a cyclical manner.

[0075] This invention constructs a refined theoretical model of pesticide distribution based on crop distribution, height, and pest and disease information, enabling precise characterization of crop needs. By establishing a droplet distribution model library for nozzles and using a vector matching algorithm to select the droplet distribution model and operating parameters that best match crop needs, it solves the problems of fixed spray parameters and poor adaptability in traditional spraying. Furthermore, it solves for the globally optimal spraying time, avoiding over-spraying or under-spraying, reducing pesticide waste, and effectively lowering environmental pollution. It achieves coordinated optimization of operating height, spraying parameters, and spraying time, ensuring uniform droplet deposition in the target crop area and effectively improving pest and disease control. Example

[0076] An embodiment of the present invention provides a variable spraying system based on crop distribution sensing, used to implement the variable spraying method in Embodiment 1, comprising: a droplet distribution model building module 4, an intelligent sensing module 1, a determination module 2, a pesticide volume distribution model generation module 3, a model matching module 5, a time solution module 6, a control module 7, and a variable spraying execution module 8; the droplet distribution model building module 4 is used to construct a droplet distribution model of the nozzle under different operating heights and different combinations of operating parameters; the intelligent sensing module 1 is used to sense and acquire field target crop information, and the intelligent sensing module 1 is a depth camera or lidar; the determination module 2 is used to determine the actual spraying position of the nozzle based on the field target crop information. The operating height; the pesticide distribution model generation module 3 is used to generate a theoretical pesticide distribution model for the target crop based on the field target crop information; the model matching module 5 is used to calculate the matching degree between the droplet distribution model of each nozzle at the actual operating height and the theoretical pesticide distribution model, and determine the droplet distribution model with the highest matching degree, which is denoted as the target droplet distribution model; the time solving module 6 is used to solve for the globally optimal spraying time based on the target droplet distribution model; the control module 7 is used to control the nozzle of the variable spray execution module 8 to perform variable spraying operations according to the optimal operating parameters corresponding to the target droplet distribution model and the globally optimal spraying time. In this embodiment of the invention, the variable spray execution module 8 includes a controller, a liquid supply tank, a liquid pump, a multi-channel relay, a flow meter, a spray pipe, a solenoid valve, and a nozzle. The liquid pump is located at the outlet of the liquid tank, and the outlet of the liquid pump is connected to the nozzle through the spray pipe. The solenoid valve and the flow meter are both located on the spray pipe. The controller is a Raspberry Pi, which is communicatively connected to the control module 7. The Raspberry Pi receives control commands from the control module and then controls the multi-channel relay to send corresponding signals to the solenoid valve and the liquid pump based on the control commands, so as to control the opening and closing degree of the solenoid valve, the opening and closing of the liquid pump, and the spray pressure, so as to realize the dynamic adjustment of spray flow rate, time, etc.

[0077] In this embodiment of the invention, the number of nozzles may be one or more. When there is only one nozzle, after the nozzle has completed spraying the target crop in a certain work area, it immediately enters the next work area to perform the variable spraying method of Embodiment 1. When the variable spraying execution module 8 contains multiple nozzles, all multiple nozzles must complete the variable spraying operation according to their respective optimal operating parameters and the global optimal spraying time before they can enter the next work area for repeated operation.

[0078] In this embodiment of the invention, the nozzle is a PWM variable nozzle, a hydraulic nozzle, or a centrifugal nozzle. Different nozzle types correspond to different operating parameters. This invention is applicable to various nozzle types and can be adapted to field operation scenarios for various crops such as corn, wheat, vegetables, and rice, demonstrating strong versatility.

[0079] A variable spraying system based on crop distribution sensing according to an embodiment of the present invention can be applied to a sprayer. The sprayer includes a walking unit, a control unit, and the variable spraying system based on crop distribution sensing according to the present invention. The control unit is connected to the walking unit and the variable spraying system based on crop distribution sensing according to the present invention, and can control the walking unit and the variable spraying system to work together.

[0080] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A variable spraying method based on crop distribution sensing, characterized in that, include: S1: Construct a droplet distribution model for the nozzle under different operating heights and different operating parameters; S2: Sensing and acquiring information about the target crop in the field, and determining the actual operating height of the nozzle and the theoretical required amount of pesticide solution distribution model for the target crop based on the target crop information; S3: Calculate the matching degree between several droplet distribution models at the corresponding actual operating height and the theoretical pesticide volume distribution model required by the target crop, and determine the droplet distribution model with the highest matching degree, denoted as the target droplet distribution model; specifically including: S301: Based on the theoretical pesticide spray volume distribution model of the target crop, calculate the total pesticide spray volume distribution model of the target crop in each grid cell, and vectorize the total pesticide spray volume distribution model of the target crop in each grid cell to generate the theoretical pesticide spray volume distribution vector. S302: Determine the droplet distribution model of the nozzle at its actual operating height, and vectorize the droplet deposition amount per unit time of each grid cell of the droplet distribution model to generate a droplet distribution model vector; S303: Calculate the matching degree between the theoretical required liquid volume distribution vector and the droplet distribution model vector, and determine the droplet distribution model with the highest matching degree with the theoretical required liquid volume distribution model, which is denoted as the target droplet distribution model; the matching degree calculation method includes the cosine similarity method or the root mean square error method; S4: The optimal operating parameters are the combination of operating parameters under the target droplet distribution model. S5: Based on the target droplet distribution model, solve for the globally optimal spraying time of the nozzle, and control the nozzle to perform spraying operations according to the optimal operating parameters and the globally optimal spraying time; specifically including: S501: Obtain the droplet deposition rate per unit time for each grid cell in the target droplet distribution model. ; S502: Obtain the total spraying demand of the target crop in each grid cell of the theoretical pesticide distribution model. ; S503: Solve for the optimal spraying time in each grid cell of the theoretical pesticide distribution model for the target crop. , ; S504: Optimal spraying time for all grid cells The average is calculated to obtain the globally optimal spraying time T, and the nozzles are controlled to perform precise spraying operations according to the optimal operating parameters and the globally optimal spraying time T.

2. The variable spraying method based on crop distribution sensing according to claim 1, characterized in that, S1 specifically includes: S101: The experiment simulates the spray deposition of the nozzle on the target plane under different operating heights and various combinations of operating parameters; the size of the target plane is preset according to the size of the single target operating area of ​​the nozzle; S102: Collect the droplet deposition distribution of each grid cell of the nozzle on the target plane under each working condition; S103: Calculate and label the droplet deposition per unit time for each grid cell in the target plane, and generate a droplet distribution model of the nozzle under this condition; S104: Associate and map the droplet distribution model with the corresponding combination of operating parameters and operating height, and store the data.

3. The variable spraying method based on crop distribution sensing according to claim 1, characterized in that, S2 specifically includes: S201: Sensing and acquiring field target crop information within the area to be operated by the nozzle, the target crop information including target crop image and target crop height data; S202: Determine the actual operating height of the sprinkler head based on the sprinkler head position and the target crop height data; S203: Construct a target crop distribution model for the target crop image, wherein the size and grid cells of the target crop distribution model are the same as those of the fog droplet distribution model; S204: Image recognition determines the number of target crops and disease information of each target crop in each grid cell of the target crop distribution model, and determines the spraying demand of each target crop based on the disease information of each target crop and the preset crop disease spraying volume model; S205: Mark the spraying demand of each target crop in each grid cell of the target crop distribution model, and generate a theoretical pesticide distribution model for the target crop.

4. A variable spraying system based on crop distribution sensing, used to implement the variable spraying method based on crop distribution sensing as described in any one of claims 1-3, characterized in that, include: The system includes a droplet distribution model establishment module, an intelligent sensing module, a determination module, a liquid volume distribution model generation module, a model matching module, a time solution module, a control module, and a variable spray execution module. The droplet distribution model building module is used to construct droplet distribution models of the nozzle under different operating heights and different combinations of operating parameters. The intelligent sensing module is used to sense and acquire information about target crops in the field. The determining module is used to determine the actual operating height of the nozzle based on the field target crop information; The pesticide solution distribution model generation module is used to generate a theoretical pesticide solution distribution model for the target crop based on the field target crop information. The model matching module is used to calculate the matching degree between the droplet distribution model of the nozzle at the actual working height and the theoretical required liquid volume distribution model, and to determine the droplet distribution model with the highest matching degree, which is denoted as the target droplet distribution model. The time calculation module is used to solve for the globally optimal spraying time based on the target droplet distribution model; The control module is used to control the nozzles of the variable spray execution module to perform variable spray operations according to the optimal operating parameters corresponding to the target droplet distribution model and the global optimal spraying time.

5. A variable spraying system based on crop distribution sensing according to claim 4, characterized in that, The intelligent sensing module is a depth camera or a lidar.

6. A variable spraying system based on crop distribution sensing according to claim 4, characterized in that, The nozzle is a PWM variable nozzle, a hydraulic nozzle, or a centrifugal nozzle.

Images