A multi-pass coordinated field harvesting dynamic domain control method and harvester

Abstract

This invention provides a multi-channel collaborative small-area harvesting dynamic domain control operation method and harvester. The method predicts the grain movement trajectory under current pneumatic parameters and adjusts the direction, angle, and speed of the airflow on the header based on the trajectory deviation rate, causing the grains to converge on the target conveying area of ​​the header. Stop line information is fused based on dynamic weight allocation and a multi-level fault-tolerance strategy to trigger a stop and seed-clearing command. In the harvester, two hidden partitions are symmetrically arranged at the front end of the seed-clearing header's bottom plate. Each hidden partition has a seed-clearing air chamber on its lower surface, and the end of the hidden partition has an air-hole guide plate. The bottom of the conveying auger feed inlet is a debris collection bin, with a telescopic plate above it. During seed clearing, the telescopic plate opens, allowing debris on the bottom plate to enter the debris collection bin through the opening. Side seed-clearing air chambers are arranged on the side plates of the seed-clearing header. This invention effectively avoids the problem of mixed seeding during breeding and harvesting, improving the efficiency of breeding work.

Description

Technical Field

[0001] This application relates to the field of small-area crop harvesting, specifically to a multi-channel collaborative small-area harvesting dynamic domain control operation method and harvester. Background Technology

[0002] Small-scale breeding experiments are a crucial part of agricultural breeding, characterized by long operating cycles, high precision requirements, and high labor intensity.

[0003] In related technologies, unlike field harvesting, small plot crop harvesting faces two major challenges: first, the plot area is small; second, different varieties are planted in adjacent plots, and the seeds inside the machine need to be thoroughly cleaned during harvesting to prevent mixing.

[0004] Currently, there is a lack of mature combined harvesters for small plots on the market, and most harvesting still relies on manual labor, which is inefficient, results in significant losses, and can easily delay the planting season. Ordinary machine harvesters also lack seed cleaning functions, leaving a large number of seeds behind after harvesting one type of seed. If another type of seed is harvested at this time, it will cause a large amount of seed cross-contamination, requiring manual seed cleaning, which seriously affects the breeding process.

[0005] To address the issue of mixed cropping during harvesting, this invention proposes a dynamic domain control operation method for a multi-channel collaborative plot self-cleaning harvesting system. By integrating an operational environment sensing system and intelligent control technology, it achieves precise self-cleaning control of the plot crop harvesting process. Summary of the Invention

[0006] The purpose of this application is to provide a multi-channel collaborative small-cell harvesting dynamic domain control operation method and harvester, so as to at least solve the technical problem of easy mixed planting in the small-cell crop harvesting process in related technologies.

[0007] To achieve the above objectives, the embodiments of this application provide the following technical solutions.

[0008] According to one embodiment of this application, a method for dynamic domain control operation of multi-channel cooperative cell harvesting is provided;

[0009] Includes the following steps:

[0010] Based on the collected grain moisture content, the initial pneumatic parameters are estimated using a matching model that matches the relationship between moisture content and pneumatic parameters.

[0011] A dynamic model of grains in the airflow field is established based on the DEM-CFD coupling method to predict the trajectory of grains under the current pneumatic parameters. Based on the actual trajectory deviation rate, the direction angle and velocity of the airflow of the header are adjusted so that the grains converge on the target pushing area of ​​the header.

[0012] Stop line information is fused based on dynamic weight allocation and multi-level fault tolerance strategy to trigger a stop and clear command.

[0013] During header cleaning, the header area is planned, and cameras arranged in different areas monitor the grain residue in real time. The gradient field of harvested residue is dynamically compared with the state space model of the self-cleaning zone, and the control system parameters are adjusted in real time. Specifically, based on the size, distance, and coordinate position of the grains in the area, the control valves and corresponding hidden air plates are opened in different areas to adjust the pneumatic parameters and blow the grains towards the middle of the bottom of the header screw conveyor. The grain mixture enters the bottom funnel hopper, and the concentrated pneumatic conveyor transports the grain mixture into the threshing device. The striped conveyor belt of the conveyor rotates back and forth to clean the conveyor device. The grain mixture enters the threshing chamber through pipes, and after threshing, it falls into the collection device via conveyor belt, completing the overall cleaning of the machine.

[0014] After the seed cleaning is completed, the clamping rod of the collection device is triggered to open, a new collection bag is inserted, and the system automatically clamps the bag opening to complete the harvesting of the current plot.

[0015] Preferably, the step of estimating the initial pneumatic parameters includes:

[0016] Receive signals collected by the GMS grain moisture sensor ;

[0017] Based on the preset matching model of the relationship between moisture content and pneumatic parameters, the initial pneumatic parameter adjustment value is quickly estimated, and the fan speed, valve opening, outlet cross-sectional area and air outlet angle are monitored and controlled in real time.

[0018] In the matching model of the relationship between moisture content and pneumatic parameters:

[0019] The forces acting on the grains during pneumatic seed cleaning are represented as follows:

[0020]

[0021]

[0022] in, This represents the force acting on the grain during the pneumatic seed cleaning process, where G represents the weight of the grain itself. This indicates the drag force of airflow on the grains. This indicates the lift generated by the rotation of the grain. The frictional force between the grain and the header surface is represented by m, which represents the mass of a single wheat seed; g is the acceleration due to gravity. Indicates the drag coefficient; This represents the lift coefficient, which is related to the particle's rotational angular velocity and relative velocity. It is the projected area of ​​the grain perpendicular to the direction of the combined velocity; It is a constant term for air density; It is the relative velocity between the airflow and the grain; μ is the coefficient of friction;

[0023] Stress after considering the effects of moisture content, concentration, and stomatal distance Represented as:

[0024]

[0025] in, Where is the outlet angle, n is the fan speed, K1 is the moisture content expansion coefficient, k2 is the fan proportionality law, k3 is the concentration blocking coefficient, and k L Airflow attenuation coefficient, x1 is grain moisture content; x3 is grain concentration; d0 is the characteristic diameter of dry grains, k w This is the rotation scaling factor.

[0026] Preferably, the step of establishing a dynamic model of grains in the airflow field based on the DEM-CFD coupling method and predicting the trajectory of grains under the current aerodynamic parameters includes: establishing a dynamic model of grains in the airflow field based on the DEM-CFD coupling method; using DEM to simulate the discrete motion of grains; combining CFD to simulate the continuous characteristics of the airflow field; realizing the dynamic simulation of fluid-structure interaction through coupling calculation; simulating the trajectory of grains, ears of grains, and straw in the complex airflow field of the header; and generating a trajectory route database.

[0027] Preferably, the step of fusing stop line information based on dynamic weight allocation and multi-level fault tolerance strategy is as follows:

[0028] The relative position and speed information of the vehicle are obtained based on the BeiDou positioning system and matched with the relative coordinate system of the sowing plan map; among them, the relative position of the vehicle is obtained based on the BeiDou positioning system ( The latitude and longitude coordinates (B, L, h) output by BeiDou are converted into relative coordinates of the experimental field by projecting geographic coordinates onto a plane coordinate system. Then, it is matched with the relative coordinate system (x, y) of the sowing plan, and represented as:

[0029]

[0030] in, =6378137m, (B0) L0) is the benchmark point for the field;

[0031] Wheat point cloud data at the front of the header is acquired by solid-state radar. The ground point cloud is segmented based on the RANSAC algorithm. When no wheat point cloud is detected in the area in front, it is preliminarily determined that the stop line has been reached. Specifically, wheat point cloud data at the front of the header is acquired by solid-state radar installed at the front. The ground point cloud is segmented based on the RANSAC algorithm. When no wheat point cloud is detected in the area 0.2~0.4m in front for 3 consecutive frames, and the length of the no-point-cloud area is ≥1.5m, the stop line has been reached.

[0032] In the step of fusion of stop line information:

[0033] Dynamic weight allocation is represented as:

[0034] w = 40% × (L / 1.5) + 30% × point cloud density coefficient + 20% × (1 - e) 2 -(0.5|v|))+10%×Environmental coefficient;

[0035] Where w represents the weight, v represents the vehicle speed obtained based on the BeiDou positioning system; 40% weight is the detection length confidence score, which represents the ratio of the detected length of the no-point-cloud region to the threshold; 30% weight is the point cloud density confidence score, which represents the effective point cloud quantity density within the detection area; 20% weight is the speed influence factor, which represents the impact of the harvester's travel speed v on the detection accuracy; and 10% weight is the environmental compensation factor, which is dynamically adjusted according to weather and crop status.

[0036] The multi-level fault tolerance strategy is a three-level fault tolerance mechanism, including: checking the quality of the current frame point cloud and performing frame-level verification; activating spatiotemporal compensation when two consecutive frames are abnormal, compensating for vehicle motion based on Beidou positioning data, reconstructing the spatial continuity of the point cloud in the most recent three frames, and verifying the trajectory consistency of the area without point cloud; and verifying the location in conjunction with the seeding planning map when five consecutive frames are abnormal.

[0037] Fusion Decision and Triggering Conditions: When the system continuously detects three frames with weights not less than 0.6 on the main path, and the length of the detected wheat-free point cloud region is not less than 1.5 meters, it enters the secondary compensation stage. In the secondary compensation stage, the data from the first two frames is used for verification. In the verification of the first two frames, the system detects that the length of the wheat-free point cloud region is not less than 1.5 meters, which serves as further confirmation. If it is still not possible to determine whether the stop line has been reached after the secondary compensation, it enters the tertiary arbitration stage. In the tertiary arbitration stage, the location information of the sowing plan is combined for final verification to confirm the existence of the wheat-free point cloud region. When the length of the wheat-free point cloud region meets the requirements, a stop and clearing command is triggered.

[0038] Preferably, the step of cleaning and seeding the cutting platform includes:

[0039] Upon receiving a shutdown and cleaning signal, extract information on the residual grain status on the worktable;

[0040] To identify grain cluster characteristics, the coordinates of the grain cluster at the maximum and minimum values ​​along the x and y axes are extracted, and represented as follows:

[0041]

[0042] Where A1~A4 are the four vertices of the grain cluster edge outline; x1 is the coordinate value of vertex A1 in the x-axis direction; y1 is the coordinate value of vertex A1 in the y-axis direction; x4 is the coordinate value of vertex A4 in the x-axis direction; y4 is the coordinate value of vertex A4 in the y-axis direction.

[0043] Determine the location of the grain cluster, start the fan, adjust the fan speed n, adjust the opening of the control valve, and simultaneously drive the motor to open the telescopic plate. Based on the motion trajectory model, adjust the angle and the size of the end opening. Based on the real-time feedback of the wind speed sensor to detect the airflow speed, concentrate the grains or ears of grain towards the grain target area. The grains are then transported to the threshing chamber entrance through the header impurity collection bin.

[0044] Once no seeds fall within 2 seconds as detected by the high-speed industrial camera, all airflow valves and fans are shut off, and the cleaning process is complete.

[0045] Preferably, after the seed cleaning is completed, the control collection device automatically releases the bag opening, and the coding area is detected by the photoelectric sensor to output a command signal by detecting the light reflected or blocked by the object. After receiving the command signal, the nozzle sprays the date and variety information, thus completing the harvesting work of the current plot of crops.

[0046] After the new bag is placed in the bag, the photoelectric sensor detects the target woven bag, triggering the automatic clamping device to continue harvesting in the next area.

[0047] According to another embodiment of this application, a harvester is provided for implementing the multi-channel cooperative cell harvesting dynamic domain control operation method provided in the above embodiments. The harvester includes:

[0048] Power chassis;

[0049] A multi-source information detection system for collecting and processing multi-source information includes solid-state radar, BeiDou positioning system, GMS grain moisture content sensor and laser ranging sensor.

[0050] The seed cleaning and cutting platform has a first high-speed industrial camera mounted on its upper baffle. An airflow seed cleaning system is also installed on the platform, comprising two concealed baffles symmetrically positioned at the front end of the platform's base plate. Each concealed baffle has a bottom seed cleaning air chamber on its lower surface, and its end has an air guide plate to discharge gas from the baffle, creating a uniform airflow. A second high-speed industrial camera and a third high-speed industrial camera are also respectively mounted on the two concealed baffles.

[0051] The rear center of the bottom plate of the seed cleaning cutting platform is a debris collection chamber. A telescopic plate is installed above the debris collection chamber, and debris on the bottom plate can enter the debris collection chamber by opening the telescopic plate. Side seed cleaning air chambers are provided on the side plates of the seed cleaning cutting platform.

[0052] The airflow distribution device has an air inlet connected to a centrifugal fan. Multiple air outlets of the airflow distribution device are connected to the bottom seed cleaning chamber, the side seed cleaning chamber, and the bottom chamber of the debris collection bin via corresponding air pipes. Each air pipe is equipped with a control valve.

[0053] Preferably, the solid-state radar is fixedly installed at the front end of the divider of the seed cleaning and harvesting platform to detect the height of the bottom of the ear of the plant above the ground;

[0054] The Beidou positioning system is installed at the front left of the power chassis and is used to obtain feedback on the harvester's position and speed.

[0055] The laser rangefinder sensor is installed at the bottom of the seed cleaning and cutting platform to detect the height of the cutting platform above the ground in real time;

[0056] The GMS grain moisture content sensor is installed on the side of the grain collection device to obtain the moisture content of the harvested grains.

[0057] Preferably, it further includes:

[0058] The conveying device includes a striped conveyor belt, one end of which is installed at the outlet of the seed cleaning and cutting table, and the other end is installed at the inlet of the threshing device.

[0059] The threshing device, mounted on a power chassis, is used to thresh the ears of crops.

[0060] The grain collecting device is connected to the discharge port of the threshing device and is used to bag crop grains.

[0061] The grain collection device includes:

[0062] The cyclone separator has its inlet connected to the grain collection port via a grain conveying pipe. The grain collection port is connected to a grain collection fan, which blows the grain from the collection port into the grain conveying pipe and then into the cyclone separator. A fourth industrial high-speed camera is installed on the outer wall of the grain outlet of the cyclone separator to detect the completion status of the shutdown and cleaning operation.

[0063] The grain collection box is equipped with a bag clamping mechanism and a multi-link system to drive the clamping mechanism. The multi-link system is connected to a hydraulic cylinder on the grain collection box. The extension and retraction of the hydraulic cylinder drives the multi-link system to clamp the woven bags. The grain collection box is also equipped with a photoelectric sensor, a coding and marking device, and a grain collection controller. The photoelectric sensor detects the woven bag information in real time, triggering the grain collection controller to retrieve the crop variety information of the current plot, and drives the coding and marking device to spray the corresponding label on the surface of the woven bag to mark the packaged grain. A GMS grain moisture content sensor is installed on the grain collection box to detect the moisture content data of the harvested grain in the current plot.

[0064] Compared with the prior art, the beneficial effects of the operating method and harvester of the embodiments of this application are:

[0065] First, after the harvester completes the harvesting operation of the current plot, the present invention uses a seed cleaning system to clean the harvester's header. This involves using high-speed industrial cameras arranged in different zones to monitor the residue on the header surface and in the conveying channels in real time. Control valves are opened in different zones, and the airflow is adjusted according to the size and distance of the affected area, blowing the grains towards the center of the bottom of the header's spiral conveyor. Simultaneously, a concealed air plate is opened, allowing the grain mixture to enter the bottom funnel. Other air ducts are closed, and concentrated airflow is used to transport the mixture into the threshing device. After seed cleaning is completed, the harvesting operation for the next plot begins.

[0066] Secondly, this invention plans the entire harvesting platform into zones, monitors seed movement and air pressure changes in real time, and dynamically adjusts the pneumatic parameters of the seed cleaning device to adapt to the needs of different seed residues. At the same time, when a single plot is harvested, it performs precise control based on the characteristics of the seeds in each area, carries out comprehensive cleaning, and implements real-time detection and seed cleaning self-cleaning zone control. Based on the moisture content, it adjusts the threshing drum speed and seed cleaning pneumatic parameters in real time. The collected seeds are then marked and bagged.

[0067] In summary, this invention effectively avoids the problem of mixed breeding and harvesting, reduces damage rate, and improves the efficiency of breeding work. Attached Figure Description

[0068] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention.

[0069] Figure 1 This is a schematic diagram of the structure of a harvester provided in an embodiment of the present invention;

[0070] Figure 2 This is a schematic diagram of the air path connection between the seed cleaning and cutting platform and the airflow distribution device in a harvester provided in an embodiment of the present invention;

[0071] Figure 3 This is a schematic diagram illustrating the cooperation between the debris collection bin and the threshing device in a harvester provided in an embodiment of the present invention;

[0072] Figure 4 This is a schematic diagram of the grain collection device in a harvester provided in an embodiment of the present invention;

[0073] Figure 5 This is a schematic diagram of the structure of the concealed partition provided in an embodiment of the present invention;

[0074] Figure 6 A schematic diagram of the seed cleaning operation structure and transport direction of the seed cleaning and cutting platform provided by the present invention;

[0075] Figure 7 The system architecture diagram of the multi-channel cooperative cell harvesting dynamic domain control operation method provided by the present invention is shown below;

[0076] Figure 8 This is a flowchart illustrating the implementation of the multi-channel collaborative cell harvesting dynamic domain control operation method of the present invention.

[0077] The attached figures are labeled as follows:

[0078] 1. Power chassis;

[0079] 2. Multi-source information detection system; 201. Solid-state radar; 202. Beidou positioning system; 203. GMS grain moisture content sensor; 204. Laser rangefinder sensor;

[0080] 3. Electrical control box; 301. First high-speed industrial camera; 302. Second high-speed industrial camera; 303. Third high-speed industrial camera;

[0081] 4. Clean the planting platform;

[0082] 5. Conveying device; 501. Striped conveyor belt;

[0083] 6. Threshing device;

[0084] 7. Grain collecting device; 701. Grain collecting inlet; 702. Grain collecting fan; 703. Grain conveying pipeline; 704. Cyclone separator; 705. Fourth industrial high-speed camera; 706. Grain outlet; 707. Bag clamping mechanism; 708. Photoelectric sensor; 709. Inkjet marking device; 710. Grain collecting controller; 711. Hydraulic cylinder; 712. Multi-link;

[0085] 8. Airflow seed cleaning system; 801. Centrifugal fan; 802. Airflow distribution device; 803. Control valve; 804. Concealed partition; 805. Side seed cleaning air chamber; 806. Debris collection bin; 807. Threshing bin inlet; 808. Air pipe; 809. Telescopic plate; 810. Air hole guide plate; 811. Wind speed sensor. Detailed Implementation

[0086] 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.

[0087] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0088] Please refer to Figures 1-5 In one embodiment of the present invention, a harvester is provided, which includes a power chassis 1 with a tracked mobile structure, which facilitates the movement of the harvester in the harvesting area when harvesting crops.

[0089] The harvester provided in this embodiment of the present disclosure also includes a multi-source information detection system 2 for collecting and processing multi-source information;

[0090] Among them, the multi-source information detection system 2 includes a solid-state radar 201, a Beidou positioning system 202, a GMS grain moisture content sensor 203, and a laser rangefinder 204.

[0091] Among them, the solid-state radar 201 is fixedly installed at the front end of the divider of the seed cleaning and cutting platform 4, and is used to detect the height of the bottom of the ear of the plant from the ground;

[0092] The Beidou positioning system 202 is installed at the left front end of the power chassis 1 to obtain feedback information on the harvester's position and speed.

[0093] Both the GMS grain moisture content sensor 203 and the laser rangefinder 204 are installed at the bottom of the seed cleaning and cutting platform 4.

[0094] The laser rangefinder 204 is used to detect the height of the header above the ground in real time and feed it back to the control system. Based on the height information of the bottom of the ear above the ground, the height of the header is adjusted in real time to feed all the ears into the header.

[0095] For further details, please refer to Figure 1 and Figure 2The harvester provided in this embodiment of the invention also includes a seed cleaning cutter 4. A first high-speed industrial camera 301 is installed on the upper baffle of the seed cleaning cutter 4. An airflow seed cleaning system 8 is provided on the seed cleaning cutter 4. The airflow seed cleaning system 8 includes two hidden partitions 804 symmetrically arranged at the front end of the bottom plate of the seed cleaning cutter 4.

[0096] like Figure 5 As shown, in one implementation of the concealed partition 804 provided in this disclosure, a bottom clearing air chamber is provided on the lower surface of each concealed partition 804, and an air hole guide plate 810 is provided at the end of the concealed partition 804 to discharge the gas in the concealed partition 804 and form a uniform airflow.

[0097] A second high-speed industrial camera 302 and a third high-speed industrial camera 303 are respectively installed on the two hidden partitions 804;

[0098] like Figures 1-3 As shown, the seed cleaning and cutting platform 4 is equipped with a conveying auger, which has two sections with opposite rotation directions. The middle of the rear end of the bottom plate of the seed cleaning and cutting platform 4 is a debris collection bin 806. The debris collection bin 806 is located at the bottom of the feeding inlet of the conveying auger. When the conveying auger rotates, the two sections with opposite rotation directions allow the harvest residue on the seed cleaning and cutting platform 4 to enter the debris collection bin 806.

[0099] A telescopic plate 809 is suspended above the debris collection bin 806. When the telescopic plate 809 is installed above the debris collection bin 806, both ends of the debris collection bin 806 are open structures, so that debris on the bottom plate of the seed cleaning and cutting platform 4 can enter the debris collection bin 806 through the openings on both sides of the telescopic plate 809.

[0100] The side plate of the seed cleaning and cutting platform 4 is provided with a side seed cleaning air chamber 805;

[0101] Please continue to refer to Figure 1 and Figure 2 The harvester in this embodiment of the application also includes an airflow distribution device 802. The air inlet of the airflow distribution device 802 is connected to a centrifugal fan 801. The multiple air outlets of the airflow distribution device 802 are respectively connected to the bottom seed cleaning air chamber, the side seed cleaning air chamber 805 and the bottom cavity of the debris collection bin 806 through corresponding air pipes 808. Each air pipe 808 is provided with a control valve 803.

[0102] Furthermore, the harvester provided in this embodiment of the invention also includes a conveying device 5 and a threshing device 6;

[0103] The conveying device 5 includes a striped conveyor belt 501, one end of which is installed at the outlet of the seed cleaning and cutting table 4, and the other end is installed at the inlet of the threshing device 6.

[0104] The threshing device 6 is mounted on the power chassis 1 and is used to thresh the crop ears.

[0105] Furthermore, such as Figure 1 and Figure 4 As shown, the harvester provided in this embodiment of the invention further includes a grain collecting device 7, which is connected to the discharge port of the threshing device 6 and is used to bag crop grains; wherein, the grain collecting device 7 includes:

[0106] Cyclone separator 704, the inlet of which is connected to grain collection port 701 via grain conveying pipe 703, and grain collection port 701 is connected to grain collection fan 702, which is used to blow grain from grain collection port 701 into grain conveying pipe 703 and then into cyclone separator 704; a fourth industrial high-speed camera 705 is installed on the outer wall of grain outlet 706 of cyclone separator 704, which is used to detect the completion status of shutdown and cleaning operation;

[0107] The grain collection box is equipped with a bag clamping mechanism 707 and a multi-link 712 for driving the bag clamping mechanism 707. The multi-link 712 is connected to a hydraulic cylinder 711 on the grain collection box. The extension and retraction of the hydraulic cylinder 711 causes the multi-link 712 to drive the bag clamping mechanism 707 to clamp the woven bag. The grain collection box is also equipped with a photoelectric sensor 708, a coding and marking device 709, and a grain collection controller 710. The photoelectric sensor 708 detects the woven bag information in real time, triggering the grain collection controller 710 to call up the crop variety information of the current plot, and driving the coding and marking device 709 to spray the corresponding label on the surface of the woven bag to mark the packaged grain.

[0108] Preferably, in this embodiment of the invention, the GMS grain moisture content sensor 203 is installed on the grain collection box to detect the moisture content data of the grains harvested in the current plot.

[0109] Before the grain collection operation begins, the woven bag is fixedly placed over the grain outlet 706. The bag opening clamping mechanism 707, under the combined action of the hydraulic cylinder 711 and the multi-link 712, fixes the opening of the woven bag. The photoelectric sensor 708 detects the information of the woven bag and feeds it back to the grain collection controller 710, which drives the inkjet printer 709 and the grain collection fan 702 to work. The inkjet marking device 709 sprays the corresponding label on the surface of the woven bag.

[0110] The grain collecting fan 702 is normally open, blowing the grains or impurities collected at the grain collecting port 701 through the grain conveying pipe 703 to the cyclone separator 704. Straw in the impurities flies out from the upper end of the cyclone separator 704, while the grains fall to the grain outlet 706 under gravity. The fourth high-speed industrial camera 705 is fixedly installed on the lower pipe of the cyclone separator 704 to detect the completion status of the shutdown and seed cleaning operation. By detecting whether any grains fall within 2 seconds, it determines whether the seed cleaning is complete, and then begins the harvesting operation of the next plot.

[0111] like Figures 7-8 In one embodiment of the present invention, a multi-channel collaborative cell harvesting dynamic domain control operation method is provided to at least solve the technical problem of easy intercropping during cell harvesting in related technologies;

[0112] This work method includes the following steps:

[0113] S1. Based on the collected grain moisture content, the initial pneumatic parameters are estimated using a matching model that matches the relationship between moisture content and pneumatic parameters.

[0114] S2. Based on the DEM-CFD coupling method, establish a dynamic model of the grain in the airflow field, predict the trajectory of the grain under the current pneumatic parameters, and adjust the direction angle and speed of the airflow of the header according to the trajectory offset rate so that the grain converges on the target pushing area of ​​the header.

[0115] S3. Based on dynamic weight allocation and multi-level fault tolerance strategy, the stop line information of the two is fused to trigger the shutdown and seed clearing command;

[0116] During header cleaning, the header area is planned, and cameras arranged in different areas monitor the grain residue in real time. The gradient field of harvested residue is dynamically compared with the state space model of the self-cleaning zone, and the parameters of the control system are adjusted in real time. Specifically, according to the size, distance, and coordinate position of the grain clusters in the area, the control valves are opened in different areas to adjust the pneumatic parameters and blow the grains towards the middle of the bottom of the header screw conveyor. At the same time, its hidden air plate is opened, and the grain mixture enters the bottom funnel chamber. The concentrated pneumatic conveyor transports the grain mixture into the threshing device. Meanwhile, the striped conveyor belt of the conveyor device rotates back and forth for 2 seconds to ensure that the conveyor device is cleaned. The grain mixture enters the threshing chamber through pipes, and after threshing, it falls into the collection device via conveyor belt, completing the overall cleaning of the machine.

[0117] S4. After the seed cleaning is completed, the clamping rod of the collection device is triggered to open, a new collection bag is inserted, and the system automatically clamps the bag opening to complete the harvesting of the current plot.

[0118] In step S2 of this application embodiment, the step of establishing a dynamic model of grains in the airflow field based on the DEM-CFD coupling method and predicting the trajectory of grains under the current aerodynamic parameters includes: establishing a dynamic model of grains in the airflow field based on the DEM-CFD coupling method, using DEM to simulate the discrete motion of grains, combining CFD to simulate the continuous characteristics of the airflow field, realizing the dynamic simulation of fluid-structure interaction through coupling calculation, simulating the trajectory of grains, ears of grains and straw in the complex airflow field of the header, and generating a trajectory route database.

[0119] Through the above-mentioned DEM-CFD bidirectional coupling process, the motion trajectory of the grain group under specific airflow parameters is virtually reproduced; the position coordinate sequence (x(t), y(t), z(t)) of each grain in the simulation time period is extracted to obtain the motion trajectory, and then the trajectory data is generated.

[0120] This application embodiment analyzes the trajectory endpoints of a large number of grains (such as the position where they fall onto the bottom plate of the header) to statistically determine the spatial distribution of the grain group and obtain the trajectory deviation rate (for example, what percentage falls into the target transport area and what percentage deviates). In calculating the trajectory deviation rate, the trajectory deviation rate is a quantitative index calculated based on the predicted distribution of the grain movement trajectory under the predicted current pneumatic parameters, and is expressed as: Deviation rate = (number of grains falling outside the target area) / (total number of grains) * 100%.

[0121] Furthermore, in step S1, the step of estimating the initial pneumatic parameters includes:

[0122] Receive signals collected by the GMS grain moisture sensor ;

[0123] Based on the preset matching model of the relationship between moisture content and pneumatic parameters, the initial pneumatic parameter adjustment value is quickly estimated, and the fan speed, valve opening, outlet cross-sectional area and air outlet angle are monitored and controlled in real time.

[0124] In the matching model of the relationship between moisture content and pneumatic parameters:

[0125] The forces acting on the grains during pneumatic seed cleaning are represented as follows:

[0126]

[0127]

[0128] in, This represents the force acting on the grain during the pneumatic seed cleaning process, where G represents the weight of the grain itself. This indicates the drag force of airflow on the grains. This indicates the lift generated by the rotation of the grain. The frictional force between the grain and the header surface is represented by m, which represents the mass of a single wheat seed; g is the acceleration due to gravity. Indicates the drag coefficient; This represents the lift coefficient, which is related to the particle's rotational angular velocity and relative velocity. It is the projected area of ​​the grain perpendicular to the direction of the combined velocity; It is a constant term for air density; It is the relative velocity between the airflow and the grain; μ is the coefficient of friction;

[0129] Stress after considering the effects of moisture content, concentration, and stomatal distance Represented as:

[0130]

[0131] in, Where is the outlet angle, n is the fan speed, K1 is the moisture content expansion coefficient, k2 is the fan proportionality law, k3 is the concentration blocking coefficient, and k L Airflow attenuation coefficient, x1 is grain moisture content; x3 is grain concentration; d0 is the characteristic diameter of dry grains, k w This is the rotation scaling factor.

[0132] In step S3, the process of fusing stop line information based on dynamic weight allocation and a multi-level fault tolerance strategy is as follows:

[0133] The relative position and speed information of the vehicle are obtained based on the BeiDou positioning system and matched with the relative coordinate system of the sowing plan map; among them, the relative position of the vehicle is obtained based on the BeiDou positioning system ( The latitude and longitude coordinates (B, L, h) output by BeiDou are converted into relative coordinates of the experimental field by projecting geographic coordinates onto a plane coordinate system. Then, it is matched with the relative coordinate system (x, y) of the sowing plan, and represented as: ;in, =6378137m, ( , () serves as the benchmark point for the field;

[0134] Wheat point cloud data at the front of the header is acquired by solid-state radar. The ground point cloud is segmented based on the RANSAC algorithm. When no wheat point cloud is detected in the area in front, it is preliminarily determined that the stop line has been reached. Specifically, wheat point cloud data at the front of the header is acquired by solid-state radar installed at the front. The ground point cloud is segmented based on the RANSAC algorithm. When no wheat point cloud is detected in the area 0.2~0.4m in front (cell isolation zone) for 3 consecutive frames, and the length of the no-point-cloud area is ≥1.5m (corresponding to typical cell planting agronomy), the stop line has been reached.

[0135] In the step of fusion of stop line information:

[0136] Dynamic weight allocation is represented as:

[0137] w = 40% × (L / 1.5) + 30% × point cloud density coefficient + 20% × (1 - e) 2 -(0.5|v|))+10%×Environmental coefficient;

[0138] Where v represents the vehicle speed obtained based on the BeiDou positioning system;

[0139] 40% of the weight is the detection length confidence score, which represents the proportional weight of the detected length (L) of the point cloud-free region to the threshold (1.5m);

[0140] The 30% weight is the point cloud density confidence score, which represents the effective point cloud quantity density within the detection area;

[0141] The 20% weighting is for the speed influence factor, which represents the impact of the harvester's travel speed v on the detection accuracy;

[0142] 10% of the weight is for the environmental compensation factor, which is dynamically adjusted based on weather (such as sunny, rainy, or foggy days) and crop condition (such as crop lodging).

[0143] Three-level fault tolerance mechanism: Check the quality of the current frame point cloud and perform frame-level verification (single frame anomaly handling); when two consecutive frames are abnormal, activate spatiotemporal compensation (continuous anomaly handling), compensate for vehicle motion based on Beidou positioning data, reconstruct the spatial continuity of the most recent three frame point clouds, and verify the trajectory consistency of areas without point clouds; when five consecutive frames are abnormal, verify the location in conjunction with the seeding planning map.

[0144] Integration of decision-making and triggering conditions:

[0145] When the system continuously detects three frames with a weight of not less than 0.6 and a detected wheat-free point cloud region length of not less than 1.5 meters via the main path, it enters the secondary compensation stage. In the secondary compensation stage, the data from the first two frames is used for verification. In the verification of the first two frames, the system detects that the length of the wheat-free point cloud region is not less than 1.5 meters, which serves as further confirmation. If it is still not possible to determine whether the stop line has been reached after the secondary compensation, it enters the tertiary arbitration stage. In the tertiary arbitration stage, the location information of the sowing plan is combined for final verification to confirm the existence of the wheat-free point cloud region. When the length of the wheat-free point cloud region meets the requirements, a stop and clearing command is triggered, and the system is reset.

[0146] Specifically, the fusion decision and triggering conditions are divided into three conditions:

[0147] Condition ①: When the system continuously detects 3 frames with weights of 0.6 or higher through the main path, and the length of the detected wheat-free point cloud region reaches or exceeds 1.5 meters, the system will enter the secondary compensation stage.

[0148] Condition ②: In the secondary compensation stage, the system will use the data from the first two frames for verification, i.e. "2-frame confirmation". If the system also detects that the length of the wheat point cloud region reaches or exceeds 1.5 meters in these two frames, this will serve as further confirmation and enhance the system's confidence in judging the stop line.

[0149] Condition ③: If the system still cannot determine whether the stop line has been reached after the second-level compensation, it will enter the third-level arbitration stage. In this stage, the system will combine the location information of the sowing plan map for final verification, namely "1-frame detection". If the system can confirm the existence of no wheat point cloud area in this frame and the length meets the requirements, then the shutdown and clearing command will be triggered.

[0150] If condition ① is met, then condition ② is verified before executing the shutdown and cleanup command. If condition ② is not met, the shutdown will not be performed. If condition ② is met, then condition ③ is verified before executing the shutdown. If condition ③ is also met, the shutdown and cleanup command can be confirmed. If condition ③ is not met, the shutdown will not be performed until condition ③ is met. After the shutdown, the system will be reset.

[0151] The system triggers a shutdown and seed clearing command, and after resetting the detection system, it prepares to proceed to the next step of the operation.

[0152] Furthermore, during the cleaning of the header in step S3, the header surface area is planned, such as... Figure 6 As shown, the cutting platform is divided into five zones: zone 1, zone 2, zone 3, zone 4, and zone 5. During seed cleaning, the harvested residues in zones 4 and 5 are blown towards zones 1 and 3 respectively through the air guide plates 810 on the corresponding hidden partitions 804. Furthermore, the harvested residues in zones 1 and 3 can be pushed towards zone 2 by a conveying auger. The harvested residues in zones 1 and 3 enter the debris collection bin 806 (i.e., zone 2) through the openings on both sides of the corresponding telescopic plates 809. At the same time, the side seed cleaning air chambers 805 are also used to blow the harvested residues in zones 1 and 3 towards zone 2, thereby achieving seed cleaning.

[0153] When performing seed cleaning on the cutting platform:

[0154] Upon receiving a shutdown and cleaning signal, extract information on the residual grain status on the worktable;

[0155] To identify grain cluster characteristics, the coordinates of the grain cluster at the maximum and minimum values ​​along the x and y axes are extracted, and represented as follows: Where A1~A4 are the four vertices of the grain cluster edge outline; x1 is the coordinate value of vertex A1 in the x-axis direction; y1 is the coordinate value of vertex A1 in the y-axis direction; x4 is the coordinate value of vertex A4 in the x-axis direction; y4 is the coordinate value of vertex A4 in the y-axis direction.

[0156] Determine the location of the grain cluster, start the fan, adjust the fan speed n, adjust the opening of control valve 803, and simultaneously drive the motor to open the telescopic plate 804. Adjust the angle and the size of the end opening according to the motion trajectory model.

[0157] Based on the real-time feedback detection of airflow speed by wind speed sensor 811, the grains or ears of grain are concentrated in the grain target transport area, and the grains are transported to the threshing chamber inlet 807 through header debris collection bin 806.

[0158] Once no seeds fall within 2 seconds as detected by the high-speed industrial camera 705, all airflow valves and fans are shut off, and the cleaning process is complete.

[0159] Furthermore, this invention extracts the coordinate lines of the termination points of each sowing plot in the planning map based on the Beidou positioning system. The speed sensor will detect that the forward speed of the harvester is zero and the radar point cloud data will identify and determine that the vehicle has reached the stop line. It is determined that the vehicle has reached the stop line of the plot, and the harvester will start the seed cleaning system to carry out the stop seed cleaning operation and enter the whole machine cleaning stage.

[0160] Therefore, after the harvester completes the harvesting operation in the current plot, the seed cleaning system is used to clean the harvester's header.

[0161] In this embodiment of the invention, high-speed industrial cameras arranged in different areas monitor the surface of the cutting table and the residue in the conveying channel in real time. Control valves are opened in different areas, and the wind force is adjusted according to the size and distance of the area to blow the grains towards the middle of the bottom of the screw conveyor of the cutting table. At the same time, the hidden air plate is opened, and the grain mixture enters the bottom funnel chamber. Other air ducts are closed, and the air force is concentrated to convey the mixture into the threshing device.

[0162] After the seed cleaning is completed, the harvesting operation for the next plot begins.

[0163] Therefore, in this embodiment of the invention, by planning the entire cutting platform into regions, monitoring the seed transport status and air pressure changes in real time, and dynamically adjusting the pneumatic parameters of the seed cleaning device, the machine can adapt to the needs of different seed residues. At the same time, when a single plot is harvested, precise control is carried out based on the characteristics of the seeds in each area to perform a comprehensive cleaning operation.

[0164] In this embodiment of the invention, in the step of adjusting the component parameters of the harvester header based on the processing results, the processing results include the harvester forward speed, grain cluster edge information, ear bottom height and grain moisture content.

[0165] In the process of adjusting the airflow direction, angle and speed by triggering the pneumatic parameter adjustment signal of the header, the airflow speed is corrected by combining the grain moisture content data. Specifically, for every 1% increase in grain moisture content, the speed compensation increases by 0.5 m / s, achieving millimeter-level monitoring of grain movement and dynamic matching of pneumatic parameters. This ensures seed cleaning efficiency under different crops (wheat, rice) and different working conditions (lodged crops, high humidity environment) while reducing energy consumption.

[0166] In addition, it is understandable that for the initial parameter settings of the air path system, the pre-trained pneumatic parameter mapping table (a decision tree model based on historical operation data) is called. For different residual scenarios (such as residue at the outlet of the threshing drum and accumulation at the edge of the screen), the zoned air path control is triggered (only the control valve of the corresponding area is opened, the airflow is concentrated to remove local residue, and the overall energy consumption is not wasted).

[0167] Furthermore, in this embodiment of the invention, when harvesting the previous cell, the method further includes the following steps:

[0168] After the seed cleaning is completed, the control collection device automatically releases the bag opening, and the coding area is detected by photoelectric sensors. By detecting the light reflected or blocked by objects, the output command signal is generated. After receiving the command, the nozzle sprays the date and variety information, completing the steps of harvesting crops in the current plot.

[0169] After the new bag is placed in the bag, the photoelectric sensor detects the target woven bag, triggering the automatic clamping device to continue harvesting in the next area.

[0170] Specifically, in this embodiment, when the industrial camera detects that no seeds are falling in the cyclone separator area, it seals the bag tightly and starts the conveyor belt to transport the woven bag to the coding area; the coding area is detected by a photoelectric sensor to output a command signal by detecting the light reflected or blocked by the object, and the printhead sprays the date and variety information after receiving the command.

[0171] This invention introduces a sensor monitoring and feedback adjustment mechanism to monitor the flow status of grains and ears in real time and dynamically adjust pneumatic parameters to adapt to the needs of different crops and operating environments.

[0172] In summary, after the harvester completes the harvesting operation of the current plot, the present invention utilizes a seed cleaning system to clean the harvester's header. Specifically, high-speed industrial cameras arranged in different zones monitor the residue on the header surface and within the conveying channels in real time. Control valves are opened in different zones, and the airflow is adjusted according to the size and distance of the zone, blowing the grains towards the center of the bottom of the header's screw conveyor. Simultaneously, a concealed air plate is opened, allowing grain mixtures to enter the bottom funnel chamber. Other air ducts are closed, concentrating airflow to transport the mixtures into the threshing device. After seed cleaning is completed, the harvesting operation for the next plot begins.

[0173] This invention plans the entire harvesting platform into zones, monitors seed transport status and air pressure changes in real time, and dynamically adjusts the pneumatic parameters of the seed cleaning device to adapt to the needs of different seed residues. At the same time, when a single plot is harvested, it performs precise control based on the characteristics of the seeds in each area, carries out comprehensive cleaning work, and implements real-time detection and seed cleaning self-cleaning zone control. Based on the moisture content, it adjusts the threshing drum speed and seed cleaning pneumatic parameters in real time. The collected seeds are then marked and bagged.

[0174] Therefore, this invention effectively avoids the problem of mixed breeding and harvesting, reduces the damage rate, and improves the efficiency of breeding work.

[0175] The above solutions are merely illustrative examples of preferred embodiments and are not limited thereto. When implementing this invention, appropriate substitutions and / or modifications can be made according to user needs.

[0176] The number of devices and processing scale described herein are for the purpose of simplifying the description of the invention. Applications, modifications, and variations of the invention will be readily apparent to those skilled in the art.

[0177] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. It can be applied to various fields suitable for the present invention. Other modifications can be readily made by those skilled in the art. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and examples shown and described herein.

Claims

1. A method for dynamic domain control operation of multi-channel collaborative cell harvesting, characterized in that, Includes the following steps: Based on the collected grain moisture content, the initial pneumatic parameters are estimated using a matching model that matches the relationship between moisture content and pneumatic parameters. The steps for estimating the initial pneumatic parameters include: receiving signals acquired by the GMS grain moisture sensor. Based on a preset matching model of the relationship between moisture content and pneumatic parameters, the initial adjustment values ​​of the pneumatic parameters are estimated, and the fan speed, valve opening, outlet cross-sectional area, and outlet angle are monitored and controlled in real time. In the matching model, the forces acting on the grains during the pneumatic seed cleaning process are expressed as follows: ； ； in, This represents the force acting on the grain during the pneumatic seed cleaning process, where G represents the weight of the grain itself. This indicates the drag force of airflow on the grains. This indicates the lift generated by the rotation of the grain. The frictional force between the grain and the header surface is represented by m, which represents the mass of a single wheat seed; g is the acceleration due to gravity. Indicates the drag coefficient; This represents the lift coefficient, which is related to the particle's rotational angular velocity and relative velocity. It is the projected area of ​​the grain perpendicular to the direction of the combined velocity; It is a constant term for air density; It is the relative velocity between the airflow and the grain; μ is the coefficient of friction; Stress after considering the effects of moisture content, concentration, and stomatal distance Represented as: ； in, Where n is the outlet angle and n is the fan speed. The coefficient of expansion is the water content. k is the fan proportionality coefficient, k3 is the concentration blocking coefficient, and k L Airflow attenuation coefficient, x1 is grain moisture content; x3 is grain concentration; d0 is the characteristic diameter of dry grains. This is the rotation scaling factor; A dynamic model of grains in the airflow field is established based on the DEM-CFD coupling method to predict the trajectory of grains under the current pneumatic parameters. Based on the actual trajectory deviation rate, the direction angle and velocity of the airflow of the header are adjusted so that the grains converge on the target pushing area of ​​the header. Stop line information is fused based on dynamic weight allocation and multi-level fault tolerance strategy to trigger a stop and clear command. During header cleaning, the header area is planned, and cameras arranged in different areas monitor the grain residue in real time. The gradient field of harvested residue is dynamically compared with the state space model of the self-cleaning zone, and the control system parameters are adjusted in real time. Specifically, based on the size, distance, and coordinate position of the grain clusters in the area, control valves and corresponding hidden partitions are opened in different areas, and the pneumatic parameters are adjusted to blow the grains towards the middle of the bottom of the header screw conveyor. The grain mixture enters the bottom funnel hopper, and the concentrated pneumatic conveyor transports the grain mixture into the threshing device. The striped conveyor belt of the conveyor rotates back and forth to clean the conveyor device. The grain mixture enters the threshing chamber through pipes, and after threshing, it falls into the collection device via conveyor belt, completing the overall cleaning of the machine. After the seed cleaning is completed, the clamping rod of the collection device is triggered to open, a new collection bag is inserted, and the system automatically clamps the bag opening to complete the harvesting of the current plot.

2. The method for dynamic domain control operation of multi-channel cooperative cell harvesting according to claim 1, characterized in that, The steps for establishing a dynamic model of grains in an airflow field based on the DEM-CFD coupling method and predicting the trajectory of grains under the current aerodynamic parameters include: using DEM to simulate the discrete motion of grains, combining CFD to simulate the continuous characteristics of the airflow field, realizing the dynamic simulation of fluid-structure interaction through coupling calculation, simulating the trajectory of grains, ears, and straw in the complex airflow field of the header, and generating a trajectory route database.

3. The method for dynamic domain control operation of multi-channel cooperative cell harvesting according to claim 2, characterized in that, The steps for fusion of stop line information based on dynamic weight allocation and multi-level fault tolerance strategy include: The relative position and speed information of the vehicle are obtained based on the BeiDou positioning system and matched with the relative coordinate system of the sowing plan map; among them, the relative position of the vehicle is obtained based on the BeiDou positioning system ( The latitude and longitude coordinates (B, L, h) output by BeiDou are converted into relative coordinates of the experimental field by projecting geographic coordinates onto a plane coordinate system. Then, it is matched with the relative coordinate system (x, y) of the sowing plan, and represented as: ;in, =6378137m, (B0, L0) are the field reference points; Wheat point cloud data at the front of the header is acquired by solid-state radar. The ground point cloud is segmented based on the RANSAC algorithm. When no wheat point cloud is detected in the area in front, it is preliminarily determined that the stop line has been reached. Specifically, wheat point cloud data at the front of the header is acquired by solid-state radar installed at the front. The ground point cloud is segmented based on the RANSAC algorithm. When no wheat point cloud is detected in the area 0.2~0.4m in front for 3 consecutive frames, and the length of the no-point-cloud area is ≥1.5m, the stop line has been reached. In the step of stopping line information fusion: Dynamic weight allocation is represented as: w = 40% × (L / 1.5) + 30% × point cloud density coefficient + 20% × (1 - e 2 -(0.5|v|) + 10% × environmental coefficient; Where w represents the weight, v represents the vehicle speed obtained based on the BeiDou positioning system; 40% weight is the detection length confidence score, which represents the ratio of the detected length of the no-point-cloud region to the threshold; 30% weight is the point cloud density confidence score, which represents the effective point cloud quantity density within the detection area; 20% weight is the speed influence factor, which represents the impact of the harvester's travel speed v on the detection accuracy; and 10% weight is the environmental compensation factor, which is dynamically adjusted according to weather and crop status. The multi-level fault tolerance strategy is a three-level fault tolerance mechanism, including: checking the quality of the current frame point cloud and performing frame-level verification; activating spatiotemporal compensation when two consecutive frames are abnormal, compensating for vehicle motion based on Beidou positioning data, reconstructing the spatial continuity of the point cloud in the most recent three frames, and verifying the trajectory consistency of the area without point cloud; and verifying the location in conjunction with the seeding planning map when five consecutive frames are abnormal. The fusion decision-making and triggering conditions include: when the system continuously detects three frames with weights not less than 0.6 and the length of the detected wheat-free point cloud region is not less than 1.5 meters, it enters the secondary compensation stage; in the secondary compensation stage, the data from the first two frames is used for verification. In the verification of the first two frames, the system detects that the length of the wheat-free point cloud region is not less than 1.5 meters, which serves as further confirmation; if it is still not possible to determine whether the stop line has been reached after the secondary compensation, it enters the tertiary arbitration stage. In the tertiary arbitration stage, the location information of the sowing plan is combined for final verification to confirm the existence of the wheat-free point cloud region. When the length of the wheat-free point cloud region meets the requirements, a stop and clearing command is triggered.

4. The method for dynamic domain control operation of multi-channel cooperative cell harvesting according to claim 3, characterized in that, The steps for cleaning and seeding the cutting platform include: Upon receiving a shutdown and seed clearing signal, extract information on the residual seed status on the worktable; To identify grain cluster characteristics, the coordinates of the grain cluster at the maximum and minimum values ​​along the x and y axes are extracted, and represented as follows: ； Where A1~A4 are the four vertices of the grain cluster edge outline; x1 is the coordinate value of vertex A1 in the x-axis direction; y1 is the coordinate value of vertex A1 in the y-axis direction; x4 is the coordinate value of vertex A4 in the x-axis direction; y4 is the coordinate value of vertex A4 in the y-axis direction. Determine the area where the grains are located, start the fan, adjust the fan speed n, adjust the opening of the control valve (803), and at the same time drive the motor to open the telescopic plate (809). According to the motion trajectory model, adjust the angle and the size of the end opening. Based on the real-time feedback of the wind speed sensor (811) to detect the airflow speed, concentrate the grains or ears of grains towards the grain target area, and transport the grains through the debris collection bin (806) to the threshing bin inlet (807). Once the fourth industrial high-speed camera (705) detects that no seeds have fallen within 2 seconds, all control valves and fans are shut off, and seed cleaning is completed.

5. The multi-channel cooperative cell harvesting dynamic domain control operation method according to claim 4, characterized in that, After the seed cleaning is completed, the control collection device automatically releases the bag opening, and the coding area is detected by photoelectric sensors. By detecting the light reflected or blocked by objects, the output command signal is generated. After receiving the command signal, the nozzle sprays the date and variety information to complete the harvesting of crops in the current plot. After the new bag is placed in the bag, the photoelectric sensor detects the target woven bag, triggering the automatic clamping device to continue harvesting in the next area.

6. A harvester, characterized in that, For implementing the multi-channel cooperative cell harvesting dynamic domain control operation method as described in any one of claims 1 to 5, the harvester includes: Power chassis (1); A multi-source information detection system (2) for collecting and processing multi-source information includes a solid-state radar (201), a Beidou positioning system (202), a GMS grain moisture content sensor (203), and a laser rangefinder (204). A seed cleaning and cutting platform (4) is provided. A first high-speed industrial camera (301) is installed on the upper baffle of the seed cleaning and cutting platform (4). An airflow seed cleaning system (8) is provided on the seed cleaning and cutting platform (4). The airflow seed cleaning system (8) includes two hidden baffles (804) symmetrically arranged at the front end of the bottom plate of the seed cleaning and cutting platform (4). The lower surface of each hidden baffle (804) is provided with a bottom seed cleaning air chamber. The end of the hidden baffle (804) has an air hole guide plate (810) for guiding the gas in the hidden baffle (804). Discharge to form a uniform airflow; a second high-speed industrial camera (302) and a third high-speed industrial camera (303) are respectively installed on the two hidden partitions (804); the middle of the rear end of the bottom plate of the seed cleaning cutting platform (4) is a debris collection chamber (806), and a telescopic plate (809) is installed above the debris collection chamber (806). The debris on the bottom plate enters the debris collection chamber (806) through the openings on both sides of the telescopic plate (809); a side seed cleaning air chamber (805) is installed on the side plate of the seed cleaning cutting platform (4). The airflow distribution device (802) has an air inlet end connected to a centrifugal fan (801). Multiple air outlets of the airflow distribution device (802) are connected to the bottom cleaning air chamber, the side cleaning air chamber (805) and the bottom cavity of the debris collection chamber (806) through corresponding air pipes. Each air pipe is equipped with a control valve (803).

7. A harvester according to claim 6, characterized in that, The solid-state radar (201) is fixedly installed at the front end of the divider of the seed cleaning and cutting platform (4) and is used to detect the height of the bottom of the ear of the plant from the ground. The Beidou positioning system (202) is installed at the left front end of the power chassis (1) and is used to obtain feedback information on the position and speed of the harvester; The laser rangefinder (204) is installed at the bottom of the seed cleaning and cutting platform (4) to detect the height of the cutting platform above the ground in real time; The GMS grain moisture sensor (203) is installed on the side of the grain collection device (7) to obtain the moisture content of the harvested grains.

8. A harvester according to claim 7, characterized in that, Also includes: The conveying device (5) includes a striped conveyor belt (501), one end of which is installed at the outlet of the seed cleaning table (4) and the other end is installed at the inlet of the threshing device (6). Threshing device (6), which is mounted on power chassis (1) and is used to thresh crop ears; Grain collecting device (7) is connected to the discharge port of threshing device (6) and is used to bag crop grains; The grain collection device (7) includes: Cyclone separator (704), the inlet of cyclone separator (704) is connected to grain collection port (701) through grain conveying pipe (703), and grain collection port (701) is connected to grain collection fan (702) to blow grain from collection port (701) into grain conveying pipe (703) and then into cyclone separator (704); a fourth industrial high-speed camera (705) is installed on the outer wall of grain outlet (706) of cyclone separator (704) to detect the completion status of shutdown and seed cleaning operation; The grain collection box is equipped with a bag clamping mechanism (707) and a multi-link (712) for driving the bag clamping mechanism (707). The multi-link (712) is connected to a hydraulic cylinder (711) on the grain collection box. The extension and retraction of the hydraulic cylinder (711) causes the multi-link (712) to drive the bag clamping mechanism (707) to clamp the woven bag. The grain collection box is also equipped with a photoelectric sensor (708), a coding and marking device (709), and a grain collection controller (710). The photoelectric sensor (708) detects the woven bag information in real time and triggers the grain collection controller (710) to call the crop variety information of the current plot and drive the coding and marking device (709) to spray the corresponding label on the surface of the woven bag to mark the packaged grain. The GMS grain moisture content sensor (203) is installed on the grain collection box to detect the moisture content data of the grain harvested in the current plot.

Images