Optimal wind pressure decision method, system and storage medium for pneumatic seeding operation

Abstract

The present invention discloses a method, system and storage medium for making an optimal wind pressure decision for pneumatic seeding operations. The method includes collecting corresponding seeding data under different wind pressures and different combinations of working condition parameters based on an electric drive voltage regulation seeding system and a sensor; the working condition parameters include speed, crop type and plant spacing; the seeding data is collected by the sensor; the collected data is processed to obtain processed data; the processed data is used to train a preset machine learning model to obtain a decision model; the optimal wind pressure is calculated based on the verified effective decision model and the working condition parameters actually collected, and the fan operation is controlled based on the optimal wind pressure. The present invention obtains an effective decision model by collecting experimental data, screening and processing data, and training based on data. It can make an optimal wind pressure decision based on the working condition parameters collected in real time when the subsequent seeding device performs seeding operations, thereby reducing the occurrence of multiple suction and missed suction.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural machinery control, and in particular to a method, system and storage medium for determining optimal wind pressure for pneumatic seeding operations. Background Art

[0002] Seeders based on suction holes have been widely used in the seeding operations of corn, soybeans, peanuts and other seeds. For example, patent CN106576518A provides a typical air suction seeder, whose working principle is to create an air pressure difference on both sides of the suction tray, so that the suction holes on the suction tray generate suction, so that the suction holes can suck up the seeds and transport the seeds to the seed outlet for discharge.

[0003] In this type of air-suction seed metering device, when the suction force of the suction hole is too small, the seeds cannot be sucked up or the sucked seeds are easy to fall. When the suction force is too large, one suction hole will suck up several seeds. Therefore, the air-suction seed metering device has the phenomenon of repeated sowing or missed sowing. In the prior art, in order to ensure the accuracy of sowing, a method of high wind pressure + seed cleaning mechanism is generally used to improve the sowing accuracy, that is, the wind pressure generated by the fan is adjusted to a value that greatly exceeds the actual required wind pressure, and the seeds sucked up by the suction hole are cleaned up by the seed cleaning mechanism. In this way, the sowing accuracy can be improved. In this method, on the one hand, the power consumption of the fan will increase, and on the other hand, it is still difficult to avoid multiple seeds being sucked up. For example, two seeds sucked up by the same suction hole are arranged front and back, which can avoid the working area of ​​the seed cleaning mechanism, resulting in multiple seeding.

[0004] In the prior art, patent CN 216567063 U discloses an electric-driven sowing monitoring and control device, which includes a seeder travel speed measurement unit, a seeding device speed measurement unit, a wind pressure sensor and a photoelectric sensor. Although this system can monitor the working parameters of the machine, the wind pressure parameters and other operating parameters need to be manually set before the operation begins, specifically setting the corresponding operating width, ground wheel diameter, seed spacing, maximum and minimum thresholds of DC motor current, wind pressure value and seeding device seed hole, etc., and transmitting these parameters to the lower computer, which calculates the operating area and other data according to the set parameters, and then transmits the calculated data to the upper computer for display and judgment. The system uses a method of setting a seeder travel speed measurement unit on the seeder and a seed meter speed measurement unit on the DC motor. At the same time, the function model relationship between the seeder travel speed and the seed meter speed set in the main control unit is used to enable the main control unit to adjust the speed of the DC motor in real time according to the seeder travel speed, ensuring that the seed distance is maintained within a specific distance, thereby achieving uniformity in sowing. However, once the wind pressure value is set, it will always be operated according to this value, which cannot meet the changing working conditions in a timely manner, and the sowing quality cannot be guaranteed.

[0005] Patent CN 113875361 A discloses the use of wind pressure, wind speed sensors and wind pressure control modules to monitor and dynamically adjust the wind pressure and flow in the pipeline in real time. It can monitor the air volume, wind speed and other parameters in the pipeline in real time, so that the wind pressure of the seed meter always remains consistent, and can achieve stable wind pressure control. However, it is still based on the preset wind pressure for pressure stabilization, which solves the problem of unstable wind pressure when the machine is idling and turning, and still cannot dynamically and intelligently adjust the wind pressure as needed. Summary of the invention

[0006] Purpose of the invention: In order to overcome the deficiencies in the prior art, the present invention provides a method, system and storage medium for making an optimal wind pressure decision for pneumatic seeding operations that can fully combine various operating parameters to regulate wind pressure and fully avoid over-suction and missed seeds.

[0007] Technical solution: To achieve the above-mentioned purpose, the optimal wind pressure decision method for pneumatic seeding operation of the present invention comprises:

[0008] Based on the electric drive voltage regulation seeding system and the sensor, the seeding data corresponding to different wind pressures and different working condition parameter combinations are collected; the working condition parameters include speed, crop type and plant spacing; the seeding data is collected by the sensor;

[0009] Processing the collected data to obtain processed data;

[0010] Using the processed data to train a preset machine learning model to obtain a decision model;

[0011] The optimal wind pressure is calculated based on the verified effective decision model and the actually collected operating parameters, and the operation of the fan is controlled based on the optimal wind pressure.

[0012] Furthermore, the electric drive voltage regulation sowing system and the sensor are used to collect sowing data corresponding to different wind pressures and different working condition parameter combinations, including:

[0013] Determine the operating condition parameters as variables;

[0014] When the variable is equal to a plurality of corresponding typical values, the wind pressure output by the fan is controlled to gradually change from the minimum wind pressure to the maximum wind pressure, and the data generated by the sensor is recorded;

[0015] Based on the data generated by the sensor, the optimal wind pressure corresponding to each test condition is determined.

[0016] Furthermore, the processing of the collected data to obtain processed data includes:

[0017] Clean, transform and pre-process the collected data to obtain data that meets the requirements;

[0018] The data meeting the requirements are normalized to obtain the processed data.

[0019] Furthermore, the decision model is expressed as:

[0020] wp=α m a x +β m b x +γ m c x ; where wp is the optimal wind pressure; a x is the wind pressure value determined based on the speed of the machine; α m is the first influence coefficient; b x is the wind pressure value determined based on the crop type; β m is the second influence coefficient; c x is the wind pressure value determined based on the plant spacing; γ m is the third influence coefficient.

[0021] Furthermore, the optimal wind pressure is calculated based on the verified effective decision model and the actually collected working condition parameters, including:

[0022] Get the specific values ​​of each working condition parameter;

[0023] Determine the preferred wind pressure corresponding to each operating parameter based on the specific value of each operating parameter;

[0024] Substitute the preferred wind pressure corresponding to each operating parameter into the decision model to obtain the optimal wind pressure.

[0025] Furthermore, the processing data is used to train a preset machine learning model to obtain a decision model, including:

[0026] Using 85% of the processed data as a training data set, training the machine learning model, determining model parameters, and obtaining a decision model;

[0027] Using 10% of the processed data as a validation data set to validate the obtained decision model;

[0028] Use 5% of the processed data as a test data set to test the final effect of the decision-making model.

[0029] The optimal wind pressure decision system for pneumatic seeding operation includes:

[0030] A collection module, which is used to collect corresponding sowing data under different wind pressures and different combinations of working parameters based on the electric drive voltage regulation sowing system and the sensor; the working parameters include speed, crop type and plant spacing; the sowing data is collected by the sensor;

[0031] A processing module, which is used to process the collected data to obtain processed data;

[0032] A training module, which is used to train a preset machine learning model using the processed data to obtain a decision model;

[0033] A decision module is used to calculate the optimal wind pressure based on the verified effective decision model and the actually collected working condition parameters, and control the operation of the fan of the electric drive pressure-regulated seeding system based on the optimal wind pressure.

[0034] A storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned optimal wind pressure decision method for pneumatic seeding operations.

[0035] Beneficial effects: The optimal wind pressure decision method, system and storage medium for pneumatic seeding operation of the present invention have the following beneficial effects:

[0036] (1) An effective decision-making model is obtained by collecting experimental data, filtering and processing data, and training the machine learning model based on the collected data. When the subsequent seeding device performs seeding operations, the optimal wind pressure decision can be made based on the real-time collected working parameters, thereby reducing the occurrence of over-suction and missed suction, ensuring sowing accuracy, and improving sowing quality.

[0037] (2) By rotating the combination of multiple operating parameters and wind pressure data, the sensor data of any operating parameter under any test wind pressure value is obtained. The machine learning model is trained based on the obtained data to obtain an effective decision-making model.

[0038] (3) The optimal wind pressure is obtained by integrating various working condition parameters based on the influence coefficient. The optimal wind pressure value obtained fully considers various influencing factors and can enable the seed meter to have the best seeding effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a structural schematic diagram of the electric drive pressure regulating sowing system of the pneumatic seed metering device;

[0040] Figure 2 It is a structural schematic diagram of the electric drive and voltage regulating sowing system of the air suction type seed metering device;

[0041] Figure 3 It is a flowchart of the optimal wind pressure decision method for pneumatic seeding operation;

[0042] Figure 4 This is a schematic diagram of the optimal wind pressure decision-making system for pneumatic seeding operations.

[0043] In the figure: 1- fan; 11- first motor; 2- seeding device; 21- housing; 22- seeding plate; 2a- suction hole; 23- second motor; 24- seeding tube; 25- release mechanism; 3- sensor; 4- shaking mechanism. " DETAILED DESCRIPTION

[0044] The present invention will be further described below in conjunction with the accompanying drawings.

[0045] The optimal wind pressure decision method for pneumatic seeding operation of the present invention is implemented based on an electric drive pressure regulating seeding system. Figure 1 The figure shows an electrically driven pressure-regulated sowing system based on a pneumatic seed meter, which is composed of a fan 1 and a seed meter 2. The seed meter 2 includes a housing 21 and a seed tray 22. The seed tray 22 has suction holes 2a arranged in a circular array. The fan blades of the fan 1 are driven by a first motor 11, and the seed tray 22 of the seed meter 2 is driven by a second motor 23. The first motor 11 and the second motor 23 are respectively connected to a control unit through a driver. The control unit can control the rotation speed of the first motor 11 and the second motor 23 to regulate the wind pressure generated by the fan 1 and the rotation speed of the seed tray 22. The housing 21 is connected to the seeding tube 24. A release mechanism 25 for eliminating the pressure difference on both sides of the suction hole 2a is installed in the housing 21. When the suction hole 2a moves to the position of the release mechanism 25, the release mechanism 25 eliminates the pressure difference on both sides of the suction hole 2a, so that the seeds sucked by the suction hole 2a can fall into the seeding tube 24 and be discharged from the seeding tube 24. Generally, the release mechanism 25 is a roller placed on the back side of the seeding disc 22, and the back side of the suction hole 2a can be blocked by the roller to eliminate the pressure difference. A sensor 3 is installed at the seeding tube 24, and the sensor 3 can collect the number of dropped seeds.

[0046] Figure 2 It is an electric drive pressure regulating sowing system based on pneumatic seed metering device. Figure 1 The difference is that the fan 1 sucks the air on the back side of the seed disc 22 to make the air pressure on the back side of the seed disc 22 lower than the air pressure on the front side, so that a pressure difference is formed on both sides of the suction hole 2a, so that the seeds sucked by the suction hole 2a can fall into the seed tube 24 and be discharged from the seed tube 24.

[0047] like Figure 3 The optimal wind pressure decision method for pneumatic seeding operation shown includes the following steps S101-S104:

[0048] Step S101, based on the electric drive voltage regulation sowing system and the sensor, corresponding sowing data under different wind pressures and different working condition parameter combinations are collected; the working condition parameters include speed, crop type and plant spacing; the sowing data is collected by the sensor;

[0049] In this step, the sensor is a photoelectric sensor installed at the seeding tube of the electric drive pressure regulating seeding system, which can collect the number of seeds dropped. In addition, the wind pressure data is collected by the wind pressure sensor, and the speed of the machine and the sowing spacing can be converted into the rotation speed of the seeding disk. The rotation speed of the seeding disk is controlled by the motor, and the control unit can accurately control the rotation speed of the seeding disk based on the speed data feedback elements such as encoders.

[0050] Step S102, processing the collected data to obtain processed data; the processed data exists in the form of a data group, which includes not only the data generated by the sensor, but also the specific values ​​of the working condition parameters and the specific values ​​of the wind pressure;

[0051] In this step, the processing of data mainly includes removing useless data, normalizing the data to facilitate the establishment of associations between different working condition parameters, etc. Useless data includes the situation where the seed suction hole sucks multiple seeds (one seed suction hole sucks up multiple seeds) and missed seeds (the seed suction hole does not suck up seeds). The situation where the seed suction hole sucks multiple seeds or the seed suction hole misses seeds reflects that the data collected by the sensor is different from the correct data. When the seed meter is normally sowing, it should be one seed per hole, corresponding to the seeding amount collected by the sensor of 1; when it is not sowing normally, there will be a single seeding amount of 2-4 (multiple suction) or a single seeding amount of 0 (missing suction).

[0052] Step S103, using the processed data to train a preset machine learning model to obtain a decision model; the decision model reflects the relationship between the wind pressure and the operating condition parameters;

[0053] Step S104, calculating the optimal wind pressure based on the verified effective decision model and the actually collected operating parameters, and controlling the operation of the fan based on the optimal wind pressure.

[0054] In the above steps, an effective decision-making model is obtained by collecting experimental data, screening and processing data, and training the machine learning model based on the collected data. When the subsequent seeding device performs seeding operations, the optimal wind pressure decision can be made based on the working conditions parameters collected in real time, thereby reducing the occurrence of over-suction and missed suction, ensuring sowing accuracy, and improving sowing quality.

[0055] Preferably, the step S101 described above, in which the electric drive voltage regulation sowing system and the sensor are used to collect sowing data corresponding to different wind pressures and different working condition parameter combinations, includes the following steps S201-S203:

[0056] Step S201, determining an operating condition parameter as a variable;

[0057] Step S202, when the variable is equal to a plurality of corresponding typical values, respectively controlling the wind pressure output by the fan to gradually change from the minimum wind pressure to the maximum wind pressure, and recording the data generated by the sensor;

[0058] Step S203, based on the data generated by the sensor, determine the optimal wind pressure corresponding to each test condition. Here, each test condition corresponds to a determined machine travel speed, crop type and sowing spacing.

[0059] For example: In the above steps, when the speed parameter is used as a variable, the sensor generates data in the process of controlling the wind pressure output by the fan from the minimum wind pressure to the maximum wind pressure under multiple speed values ​​(i.e. the above typical values) for testing. The above multiple speed values ​​can be selected as typical values ​​for testing at equal steps within the maximum operating speed range of the machine. In addition, the wind pressure generated by the fan can also be taken at a specific step distance between the minimum wind pressure and the maximum wind pressure to obtain multiple test wind pressure values ​​for testing.

[0060] When the plant spacing is used as a variable and multiple typical plant spacing conditions are tested, the data generated by the sensor during the process of controlling the wind pressure output by the fan to gradually change from the minimum wind pressure to the maximum wind pressure.

[0061] For each crop type (such as corn, soybeans, peanuts, etc., where the crop type itself is also a variable), the above-mentioned data collection process with speed as a variable and plant spacing as a variable is carried out. In this way, the three operating parameters and four data of wind pressure can be rotated and combined to obtain the sensor data of any operating parameter under any test wind pressure value.

[0062] When the sensor collects data, under each set of operating parameter combinations, the seeding motor that drives the seeding disc drives the seeding disc to rotate for a preset N cycles. If there are t suction holes on the seeding disc, the number of seeding times when the seeding disc rotates for N cycles is N·t, and the sensor collects data N·t times accordingly.

[0063] Preferably, the step S203 described above of determining the optimal wind pressure corresponding to each test condition based on the data generated by the sensor includes the following steps S301-S302:

[0064] Step S301, collecting statistics on the abnormal broadcast rate of the test condition according to the data generated by the sensor; here, the abnormal broadcast rate includes the replay rate and the missed broadcast rate.

[0065] Step S302, determining the optimal wind pressure based on the data of the two dimensions of the different broadcasting rate and the wind pressure.

[0066] In this step, the optimal wind pressure is determined based on the seeding rate, and the result is relatively accurate, because the seeding rate itself is obtained by statistics based on N·t sensor data, which can eliminate the inaccurate data caused by chance. In the process of determining the optimal wind pressure, the wind pressure can be used as the horizontal coordinate and the seeding rate as the vertical coordinate to examine the trend of the fitting curve. Generally, when the wind pressure value is small, the suction force generated by the suction hole is small, and the suction force of the suction hole is not enough to suck the seeds, or the sucked seeds are separated from the suction hole due to the centrifugal force of the seeding disk. At this time, the missed seeding rate is high. As the wind pressure increases, the missed seeding rate gradually decreases. As the wind pressure continues to increase, the suction force generated by the suction hole is too large, and one suction hole may suck up multiple seeds, resulting in multiple seeds being discharged, resulting in a gradual increase in the reseeding rate.

[0067] In actual implementation, when calculating the abnormal broadcast rate, the replay rate and the missed broadcast rate can be determined separately, and the trend curve of the missed broadcast rate and the trend curve of the replay rate with increasing wind pressure can be determined separately. The wind pressure corresponding to the intersection of the two trend curves is used as the optimal wind pressure.

[0068] Preferably, the step S102 described above of processing the collected data to obtain processed data includes the following steps S401-S402:

[0069] Step S401, cleaning, converting and preprocessing the collected data to remove abnormal data and obtain data that meets the requirements;

[0070] In this step, abnormal data refers to data that does not generate sensor data at all, or has a high rate of abnormal transmission. These data will increase unnecessary computing power waste in subsequent calculations, and because the data jumps too much, it will have an adverse effect on the subsequent fitting data trend. Therefore, first excluding them can save computing power while preventing abnormal data from having an adverse effect on subsequent data processing.

[0071] Step S402, normalizing the data that meets the requirements to obtain the processed data.

[0072] In this step, after the data is normalized, it is convenient to examine the impact of various operating parameters on the optimal wind pressure under the same system, which facilitates data fitting and machine learning.

[0073] Preferably, the decision model is expressed as:

[0074] wp=α m a x +β m b x +γ m c x ; where wp is the optimal wind pressure; a x is the wind pressure value determined based on the speed of the machine; α mis the first influence coefficient; b x is the wind pressure value determined based on the crop type; β m is the second influence coefficient; c x is the wind pressure value determined based on the plant spacing; γ m is the third influence coefficient. m and x are serial numbers.

[0075] The above-mentioned first influence coefficient, second influence coefficient and third influence data are model parameters obtained through training based on machine learning models and training data sets. Through machine learning, the degree of influence of each operating condition parameter on the optimal wind pressure can be obtained, and the influence coefficient can be obtained. In this way, when making subsequent decisions on the optimal wind pressure, the optimal wind pressure can be obtained based on the influence coefficient and comprehensive consideration of various operating condition parameters. The obtained optimal wind pressure value fully considers various influencing factors and can enable the seed meter to have the best seeding effect.

[0076] In addition, the seed meter 2 is installed on the shaking mechanism 4, which can drive the seed meter 2 to shake up and down as a whole. The shaking amplitude and frequency of the shaking mechanism 4 can be adjusted to simulate the ground bumps when sowing in the field. Based on this, the working condition parameters in step S101 also include the shaking frequency and shaking amplitude. The preferred wind pressure model constructed on this basis is: wp = (ζk + ηs) (α m a x +β m b x +γ m c x ); (ζk+ηs) is the fourth influence coefficient determined based on the shaking frequency f and the shaking amplitude s. The fourth influence coefficient is a value greater than or equal to 1, which determines the influence of the turbulence generated by the machine during field operation on the optimal wind pressure. ζ is the fourth influence coefficient corresponding to the shaking frequency f, and η is the fifth influence coefficient corresponding to the shaking amplitude s. ζ and η normalize and coefficientize the shaking frequency f and the shaking amplitude s to uniformly describe the influence of the two on the wind pressure decision.

[0077] Preferably, the above step S105, calculating the optimal wind pressure based on the verified effective decision model and the actually collected working condition parameters, includes the following steps S501-S503:

[0078] Step S501, obtaining specific values ​​of various operating parameters;

[0079] Step S502, determining the preferred wind pressure corresponding to each operating condition parameter based on the specific value of each operating condition parameter;

[0080] Step S503: Substitute the preferred wind pressure corresponding to each operating condition parameter into the decision model to obtain the optimal wind pressure.

[0081] Preferably, the step S103 described above of using the processed data to train a preset machine learning model to obtain a decision model includes the following steps S601-S603:

[0082] Step S601, using 85% of the processed data as a training data set, training the machine learning model, determining model parameters, and obtaining a decision model;

[0083] Step S602, using 10% of the processed data as a verification data set to verify the obtained decision model;

[0084] Step S603, using 5% of the processed data as a test data set to test the final effect of the decision model.

[0085] In this way, the collected data can be fully utilized to complete the training, verification and testing of the decision model to ensure the reliability of the obtained decision model.

[0086] The present invention also provides an optimal wind pressure decision system for pneumatic seeding operation (hereinafter referred to as: decision system), which may include or be divided into one or more program modules, one or more program modules are stored in a storage medium and executed by one or more processors to complete the present invention and implement the above-mentioned optimal wind pressure decision method for pneumatic seeding operation. The program module referred to in the embodiment of the present invention refers to a series of computer program instruction segments that can complete specific functions, and is more suitable for describing the execution process of the optimal wind pressure decision method for pneumatic seeding operation in a storage medium than the program itself.

[0087] like Figure 4 As shown in the figure, the optimal wind pressure decision system for pneumatic seeding operation includes:

[0088] The acquisition module 701 is used to collect the corresponding sowing data under different wind pressures and different working condition parameter combinations based on the electric drive voltage regulation sowing system and the sensor; the working condition parameters include speed, crop type and plant spacing; the sowing data is collected by the sensor;

[0089] A processing module 702 is used to process the collected data to obtain processed data;

[0090] A training module 703 is used to train a preset machine learning model using the processed data to obtain a decision model;

[0091] The decision module 704 is used to calculate the optimal wind pressure based on the verified effective decision model and the actually collected working condition parameters, and control the operation of the fan of the electric drive pressure regulation seeding system based on the optimal wind pressure.

[0092] Other contents of realizing the above-mentioned optimal wind pressure decision method for pneumatic seeding operation based on the optimal wind pressure decision system for pneumatic seeding operation have been introduced in detail in the previous embodiments. Please refer to the corresponding contents in the previous embodiments and will not be repeated here.

[0093] The present invention also provides a computer-readable storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory (for example, an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a disk, an optical disk, a server, etc., on which a computer program is stored, and when the program is executed by a processor, the above-mentioned optimal wind pressure decision method for pneumatic seeding operations is implemented.

[0094] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. The optimal wind pressure decision method for pneumatic seeding operation is characterized by: The method comprises: Based on the electric drive voltage regulation seeding system and the sensor, the seeding data corresponding to different wind pressures and different working condition parameter combinations are collected; the working condition parameters include speed, crop type and plant spacing; the seeding data is collected by the sensor; Processing the collected data to obtain processed data; Using the processed data to train a preset machine learning model to obtain a decision model; The optimal wind pressure is calculated based on the verified effective decision model and the actually collected operating parameters, and the operation of the fan is controlled based on the optimal wind pressure.

2. The optimal wind pressure decision method for pneumatic seeding operation according to claim 1 is characterized in that: The electric drive voltage regulation seeding system and the sensor are used to collect seeding data corresponding to different wind pressures and different working condition parameter combinations, including: Determine the operating condition parameters as variables; When the variable is equal to a plurality of corresponding typical values, the wind pressure output by the fan is controlled to gradually change from the minimum wind pressure to the maximum wind pressure, and the data generated by the sensor is recorded; Based on the data generated by the sensor, the optimal wind pressure corresponding to each test condition is determined.

3. The optimal wind pressure decision method for pneumatic seeding operation according to claim 1 is characterized in that: The step of processing the collected data to obtain processed data includes: Clean, transform and pre-process the collected data to obtain data that meets the requirements; The data meeting the requirements are normalized to obtain the processed data.

4. The optimal wind pressure decision method for pneumatic seeding operation according to claim 1, characterized in that: The decision model is expressed as: wp=α m a x +β m b x +γ m c x ; where wp is the optimal wind pressure; a x is the wind pressure value determined based on the speed of the machine; α m is the first influence coefficient; b x is the wind pressure value determined based on the crop type; β m is the second influence coefficient; cx is the wind pressure value determined based on the plant spacing; γ m is the third influence coefficient.

5. The optimal wind pressure decision method for pneumatic seeding operation according to claim 4 is characterized in that: The calculating of the optimal wind pressure based on the verified effective decision model and the actually collected working condition parameters includes: Get the specific values ​​of each working condition parameter; Determine the preferred wind pressure corresponding to each operating parameter based on the specific value of each operating parameter; Substitute the preferred wind pressure corresponding to each operating parameter into the decision model to obtain the optimal wind pressure.

6. The optimal wind pressure decision method for pneumatic seeding operation according to claim 4, characterized in that: The method of using the processed data to train a preset machine learning model to obtain a decision model includes: Using 85% of the processed data as a training data set, training the machine learning model, determining model parameters, and obtaining a decision model; Using 10% of the processed data as a validation data set to validate the obtained decision model; Use 5% of the processed data as a test data set to test the final effect of the decision-making model.

7. The optimal wind pressure decision system for pneumatic seeding operation is characterized by: It includes: A collection module, which is used to collect corresponding sowing data under different wind pressures and different working condition parameter combinations based on the electric drive voltage regulation sowing system and sensors; The working condition parameters include speed, crop type and plant spacing; the sowing data is collected by the sensor; A processing module, which is used to process the collected data to obtain processed data; A training module, which is used to train a preset machine learning model using the processed data to obtain a decision model; A decision module is used to calculate the optimal wind pressure based on the verified effective decision model and the actually collected working condition parameters, and control the operation of the fan of the electric drive pressure-regulated seeding system based on the optimal wind pressure.

8. A storage medium, characterized in that A computer program is stored thereon, and when the program is executed by a processor, the optimal wind pressure decision method for pneumatic seeding operations described in any one of claims 1-6 is implemented.

Images