Agricultural mechanical fertilizer application device with high-efficiency mixing function and method thereof

Abstract

The application discloses an agricultural mechanical fertilization device with high-efficiency mixing function and a method thereof. Through the transmission mode of the first motor, the first belt pulley and the second belt pulley, the first rotating rod and the second rotating rod can be driven to rotate simultaneously, the first rotating rod and the second rotating rod drive the stirring rod to generate strong vortex, the moist and caked fertilizer is uniformly dispersed, the uniformity of subsequent mixing treatment of the fertilizer is ensured, balanced nutrient supply is provided for crops, the growth quality and yield of crops are helped to improve, the multi-bin design of the storage mechanism and the application of the flow regulating valve can flexibly adjust the mixing ratio of the fertilizer according to different fertilizer types and crop demands, and precise multi-component mixed fertilization is realized. According to the demands of different crop growth stages, the application adopts a dynamic adjustment mechanism, can better adapt to the nutrient demands of crops in different growth stages, promotes the healthy growth of crops, and improves the yield and quality of crops.

Description

Technical Field

[0001] This invention belongs to the field of agricultural machinery technology, and relates to an agricultural machinery fertilizer application device and method with efficient mixing function. Background Technology

[0002] In modern agricultural production, precision fertilization is crucial for increasing crop yields, improving the quality of agricultural products, and protecting the soil ecological environment. However, existing agricultural machinery fertilization devices have many shortcomings and are unable to meet the precision and efficiency requirements of modern agriculture.

[0003] Traditional fertilization devices mostly use simple stirring methods, which cannot achieve full mixing of different fertilizers. Different fertilizers have different particle sizes, densities and other physical properties, which can easily lead to stratification and agglomeration during the mixing process, resulting in uneven fertilizer mixing. This makes it impossible for crops to obtain a balanced supply of nutrients during growth, affecting the uniformity of their growth and development and overall yield.

[0004] Most fertilization devices lack multi-channel mixing capabilities, making it difficult to flexibly adjust the fertilizer mixing ratio according to different fertilizer types and crop needs at different growth stages. Farmers often have to rely on experience to fertilize, failing to achieve precise multi-component mixing, resulting in fertilizer waste and reduced fertilization effectiveness. Taking common field crops as an example, the required ratios of nitrogen, phosphorus, and potassium fertilizers vary at different growth stages. However, existing fertilization devices cannot precisely adjust to these changes, either using a single fertilizer mixing ratio throughout the entire growth cycle or only allowing for rough manual adjustments. This falls far short of meeting the precise nutrient requirements of crops, wasting fertilizer resources and failing to achieve optimal fertilization results.

[0005] To address these issues, we provide an agricultural machinery fertilization device and method with efficient mixing capabilities. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide an agricultural machinery fertilization device and method with efficient mixing function. By cooperating with the storage mechanism and the feeding mechanism, it solves the problems that most existing fertilization devices adopt a simple stirring method, which cannot achieve full mixing of different fertilizers, and most fertilization devices lack multi-channel mixing function.

[0007] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution.

[0008] In a first aspect, the present invention provides an agricultural mechanical fertilizer applicator with a high-efficiency mixing function, comprising a frame, a material storage mechanism fixedly connected to one side of the top of the frame, a material feeding mechanism fixedly connected to the inner cavity of the frame, a fixing frame fixedly connected to one side of the frame, a sprinkler mechanism fixedly connected to the inner cavity of the fixing frame, auxiliary wheel assemblies provided at the front and rear ends of the bottom side of the fixing frame, the material storage mechanism comprising multiple material storage bins, the surface of the material storage bins fixedly connected to the frame, a material feeding pipe connected to the bottom of the material storage bins, a flow solenoid valve provided on one side of the material feeding pipe, a first motor fixedly connected to one side of the frame, and an output terminal of the first motor fixedly connected to... A first rotating rod has a first pulley fixedly connected to its surface. A second rotating rod is movably connected to both the front and rear ends of one side of the frame. A second pulley is connected to the surface of the second rotating rod, which is driven by the first pulley via a belt. One side of both the first and second rotating rods extends into the inner cavity of the storage hopper. A stirring rod is fixedly connected to the surfaces of both the first and second rotating rods. The feeding mechanism includes a mixing hopper. The top of the mixing hopper communicates with a feeding pipe. A transmission rod extends through the inner cavity of the mixing hopper. Both ends of the transmission rod are fixedly connected to wheels. A third pulley is fixedly connected to one side of the transmission rod. The bottom of the mixing hopper is connected to… The device includes a distribution bin, the inner cavity of which is movably connected to a movable rod via bearings. A fourth pulley is fixedly connected to one side of the movable rod, which is driven by a belt to a third pulley. Spiral blades are fixedly connected to the surface of the drive rod, and a material-pushing block is fixedly connected to the surface of the movable rod. The material-pushing block is located within the inner cavity of the distribution bin. The frame serves as the supporting structure for the entire device, ensuring the overall stability of the fertilization device. With the cooperation of the traveling wheels and auxiliary wheels, stable operation is achieved, improving the usability of the fertilization device. There are three storage bins, each for storing different types of fertilizer. Each storage bin has a feed inlet at the top for easy fertilizer addition. The unit is connected to a discharge pipe equipped with a flow solenoid valve to control the fertilizer flow rate. Each storage bin is also equipped with a level sensor to monitor the fertilizer level in real time. The mixing bin is located below the storage bins and is connected via the discharge pipe. Fertilizer from the storage bins enters the mixing bin through the discharge pipe. Driven by the wheels, the drive rod and movable rod rotate to achieve uniform mixing of the fertilizer inside the mixing bin. Spiral blades are distributed on the surface of the drive rod, which can form a complex flow field during the mixing process, allowing the fertilizer to tumble and collide fully in the mixing bin for efficient mixing. The inner wall of the mixing bin is also equipped with baffles to further enhance the mixing effect of the fertilizer.

[0009] The invention is further configured such that the irrigation mechanism includes a housing, one side of which is fixedly connected to a fixed frame, a pump body is fixedly connected to one side of the top of the housing, one side of the pump body is connected to the housing via a connecting pipe, and a flexible hose is connected to one side of the pump body. A pipe body is movably connected to one side of the fixed frame, one side of which penetrates the inner cavity of the fixed frame and extends to the outside of the fixed frame. A sector gear is fixedly connected to the surface of the pipe body, and a second motor is fixedly connected to one side of the surface of the fixed frame. A circular gear is fixedly connected to the output end of the second motor, and the circular gear meshes with the sector gear. By activating the pump body, the pump body draws liquid fertilizer from inside the housing through the connecting pipe and delivers it to the inside of the pipe body through the flexible hose, where it is sprayed out from the nozzle on the surface of the pipe body, thereby achieving the effect of fertilizing with liquid fertilizer. By activating the second motor, the second motor drives the circular gear to rotate, which in turn drives the sector gear to rotate, which in turn drives the pipe body to rotate, allowing adjustment of the irrigation angle of the pipe body and improving the functionality of the irrigation mechanism.

[0010] The invention is further configured such that the auxiliary wheel assembly includes a support frame, the top of the support frame is fixedly connected to a fixed frame, a spring damper is fixedly connected to the inner cavity of the fixed frame, a connecting block is fixedly connected to the bottom of the spring damper, a movable wheel is fixedly connected to one side of the connecting block, and a spring damper is provided on the top of the movable wheel, which can ensure that the movable wheel has an efficient shock absorption effect during travel, thus ensuring the convenience and stability of the agricultural machinery fertilization device.

[0011] The present invention is further configured such that a reinforcing plate is fixedly connected to the surface of the second motor, one side of the reinforcing plate is fixedly connected to the fixing frame, and the other side is fixed to the second motor, which enables the installation of the second motor and ensures the stability of the installation of the second motor.

[0012] Secondly, the present invention provides an adaptive multivariate intelligent fertilization method using an agricultural mechanical fertilization device with efficient mixing function, comprising the following steps:

[0013] Step S1: Data Acquisition and Preprocessing

[0014] One week before sowing, multiple soil sensor nodes are evenly deployed in the gramineous crop fields; the soil sensors are used to collect soil condition data in real time.

[0015] Step S2: Use a pre-trained multivariate nonlinear regression model to periodically predict the growth index y of gramineous crops under different fertilizer application rates during the sowing period; adjust the flow solenoid valve according to the optimal fertilizer application rate under the optimal gramineous crop growth index y, thereby realizing the amount of fertilizer dispensed from the corresponding storage bin, and thus fertilizing gramineous crops during the sowing period.

[0016] Step S3: Use fuzzy logic to adjust the fertilizer application rate during the sowing period for gramineous crops that have entered the jointing stage. Then, use the adjusted fertilizer application rate to control the flow solenoid valve, thereby realizing the amount of fertilizer dispensed from the corresponding storage bin, and thus fertilizing the gramineous crops during the jointing stage.

[0017] Step S4: For gramineous crops that have entered the heading stage, a reinforcement learning optimization strategy is used to adjust the fertilizer application rate during the jointing stage. Then, the adjusted fertilizer application rate is used to control the flow solenoid valve, thereby realizing the amount of fertilizer dispensed from the corresponding storage bin, thus fertilizing the gramineous crops during the heading stage.

[0018] Step S5, Adaptive Weight Update during Grouting Period:

[0019] The weights of each component in the fertilizer during the heading stage are adjusted using Bayesian optimization. Then, the flow solenoid valve is controlled by the adjusted fertilizer application scheme to control the amount of fertilizer dispensed from the corresponding storage bin, thereby fertilizing gramineous crops during the grain-filling stage.

[0020] The present invention is further configured such that the soil condition data includes soil moisture, soil temperature, and soil electrical conductivity.

[0021] The present invention is further configured such that, after the soil sensor collects soil condition data in real time, it also performs noise reduction filtering on the soil condition data to obtain noise-reduced soil condition data. ; ;in This indicates the types of soil condition data. This represents the i-th soil condition state data after noise reduction and filtering.

[0022] The present invention is further configured such that, in step S2, the multivariate nonlinear regression model adopts the Gaussian process regression (GPR) model for prediction; and the kernel function adopts the radial basis function (RBF).

[0023] Gaussian process regression formula:

[0024]

[0025]

[0026] in Indicate the model hyperparameters; f represents the Gaussian noise term; f() represents the Gaussian process regression function; Indicates similarity in fertilizer formulation; Indicates the application amount of different fertilizers i and j; Represents the radial basis function (RBF) kernel; Indicates the length scale parameter;

[0027] In step S3, a fuzzy rule is designed based on at least one indicator in the soil condition state data. The rule is: "If the soil condition state data..." If the amount is below the threshold, increase the application rate of the i-th fertilizer component. ", i∈[1,n];

[0028] Let membership function For an S-shaped curve, when <threshold It approaches 1, and conversely, it approaches 0;

[0029] According to regular detection Update the application rate of the i-th fertilizer component. :

[0030]

[0031] in This represents the updated dosage of the i-th fertilizer component. This represents the maximum permissible dosage of the i-th fertilizer component;

[0032] The reinforcement learning optimization strategy described in step S4 employs the double-Q learning algorithm;

[0033] Define state s t Given the soil conditions and gramineous crop growth indicators at current time t, action a t Adjust the fertilizer ratio for the current time t, and reward r. t This represents the change in the health status of grass crops at the current time t.

[0034] Update the Q value according to the following formula:

[0035]

[0036] Learning rate Control the update step size; This represents the Q-value of the current state-action pair; This is the current state; This is the current action; It's an instant reward; It is a discount factor; Indicates the next state One action from the set of all possible actions; Indicates the next state In, all possible actions The maximum value among the Q values;

[0037] In step S5, the optimal weight parameters are found. Maximize the posterior probability of the following objective function:

[0038]

[0039] Historical data ,in This represents the application rate of all fertilizers. These are the corresponding growth indicators for all grass crops; It is the posterior probability, representing the parameter given historical data D. The probability distribution; It is the likelihood function, which measures the likelihood of a given set of parameters. At that time, the probability of observing historical data D; It is the prior distribution, representing the distribution of... The initial assumptions; This indicates the initial weight of each component in the fertilizer; , Let m represent the growth index of the m-th gramineous crop, where m ∈ [1, M].

[0040] The present invention is further configured such that the method includes real-time monitoring and anomaly detection, specifically: during the entire growth period of the gramineous crop, a first anomaly score and a second anomaly score are calculated respectively using soil condition data and gramineous crop growth indicators; if at least one of the first anomaly score and the second anomaly score is greater than a threshold, an alarm will be triggered and the fertilization strategy will be adjusted according to the anomaly point.

[0041] The present invention is further configured such that the first abnormal score The calculation is as follows:

[0042]

[0043] in Represents soil condition data Average path length in an isolated tree; This represents the factor used to adjust the path length; t represents the current time t; n represents the expected value of the path length of the isolated tree;

[0044] Second abnormal score The calculation is as follows:

[0045]

[0046] in Indicators of growth of grass crops Average path length in an isolated tree.

[0047] The present invention has the following beneficial effects:

[0048] 1. This invention, through the transmission of a first motor, a first pulley, and a second pulley, can simultaneously drive a first rotating rod and a second rotating rod to rotate. The first and second rotating rods drive the stirring rod to generate a strong vortex, which enables the fertilizer inside each storage bin to be fully dispersed in a short time, uniformly dispersing the moist and lumpy fertilizer, ensuring the uniformity of subsequent fertilizer mixing and processing, providing a balanced nutrient supply to crops, and helping to improve the growth quality and yield of crops. The multi-bin design of the storage mechanism and the application of flow regulating valves enable the device to flexibly adjust the fertilizer mixing ratio according to different fertilizer types and crop needs, achieving precise multi-component mixed fertilization, and meeting the diversified fertilization needs of modern agriculture.

[0049] 2. This invention utilizes a transmission assembly consisting of a traveling wheel, a transmission rod, a third pulley, and a fourth pulley to drive the spiral blades and the feeding block to rotate. The spiral blades thoroughly mix the fertilizer inside the mixing chamber, while the feeding block, driven by the traveling wheel, evenly distributes the fertilizer, achieving continuous feeding during operation and preventing feeding from continuing even after the equipment stops. This highly efficient mixing structure ensures the uniform distribution of various nutrients in the fertilizer, providing a balanced nutrient supply to crops. This helps crops receive sufficient and appropriate nutrition in all parts during growth, thereby improving crop growth uniformity. More uniform crop growth facilitates unified management and harvesting, reduces yield losses due to growth differences, and contributes to improving the overall yield and quality of crops.

[0050] 3. This invention dynamically adjusts fertilization strategies based on the needs of different crop growth stages, employing methods such as fuzzy logic, reinforcement learning, and Bayesian optimization. This dynamic adjustment mechanism better adapts to the nutrient requirements of crops at different growth stages, promoting healthy crop growth and improving crop yield and quality. Furthermore, this invention monitors soil conditions and crop growth indicators in real time throughout the entire crop growth period, calculates anomaly scores, and triggers alarms promptly. This method can quickly identify problems such as abnormal soil fertility or hindered crop growth and adjust fertilization strategies accordingly. In summary, through precise fertilization and dynamic adjustment strategies, this invention significantly reduces soil pollution and water eutrophication caused by over-fertilization, while avoiding poor crop growth due to insufficient fertilization. This eco-friendly fertilization method not only improves agricultural production efficiency but also promotes sustainable agricultural development, aligning with the development concept of modern green agriculture. Attached Figure Description

[0051] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 This is a three-dimensional diagram of an agricultural mechanical fertilizer application device with a highly efficient mixing function.

[0053] Figure 2 This is a three-dimensional view of the fixed frame in an agricultural machinery fertilizer application device with efficient mixing function.

[0054] Figure 3 This is a perspective view of the first motor and its connection structure in an agricultural mechanical fertilizer applicator with efficient mixing function.

[0055] Figure 4 This is a three-dimensional view of the mixing chamber in an agricultural machinery fertilizer application device with a high-efficiency mixing function.

[0056] Figure 5 This is a three-dimensional view of the storage bin in an agricultural machinery fertilizer application device with efficient mixing function.

[0057] Figure 6 This is a three-dimensional diagram of the connection structure between the third and fourth pulleys in an agricultural machinery fertilizer application device with efficient mixing function.

[0058] Figure 7 This is a perspective view of the auxiliary wheel assembly in an agricultural machinery fertilizer application device with efficient mixing function.

[0059] Figure 8 This is a three-dimensional view of the second motor and its connection structure in an agricultural mechanical fertilizer applicator with efficient mixing function.

[0060] Figure 9 This is a flowchart of an adaptive multivariate intelligent fertilization method.

[0061] In the attached diagram: 1. Frame; 2. Storage mechanism; 201. Storage bin; 202. Feed pipe; 203. Flow solenoid valve; 204. First motor; 205. First rotating rod; 206. First pulley; 207. Second rotating rod; 208. Second pulley; 209. Mixing rod; 3. Feeding mechanism; 301. Mixing bin; 302. Transmission rod; 303. Traveling wheel; 304. Third pulley; 305. Distributing bin; 306. Movable rod; 307. Fourth pulley. 308. Pulley; 309. Spiral blade; 4. Material feeding block; 5. Fixing frame; 6. Sprinkler mechanism; 501. Housing; 502. Pump body; 503. Hose; 504. Pipe body; 505. Sector gear; 506. Second motor; 507. Circular gear; 6. Auxiliary wheel assembly; 601. Support frame; 602. Spring damper; 603. Connecting block; 604. Moving wheel; 7. Reinforcing plate; 8. Connecting lug; 9. Reinforcing rod; 10. Cover plate; 11. Reinforcing base. Detailed Implementation

[0062] The technical solutions of the present invention will be described below with reference to the accompanying drawings. The described embodiments are only some embodiments of the present invention, and not all embodiments.

[0063] Example 1

[0064] Please see Figure 1-8 The present invention is an agricultural mechanical fertilizer application device with efficient mixing function, including a frame 1, a material storage mechanism 2 fixedly connected to one side of the top of the frame 1, a material feeding mechanism 3 fixedly connected to the inner cavity of the frame 1, a fixing frame 4 fixedly connected to one side of the frame 1, a sprinkler irrigation mechanism 5 fixedly connected to the inner cavity of the fixing frame 4, and auxiliary wheel assemblies 6 provided at the front and rear ends of the bottom side of the fixing frame 4. The storage mechanism 2 includes a storage bin 201. The number of storage bins 201 can be set according to actual needs. The surface of the storage bin 201 is fixedly connected to the frame 1. The bottom of the storage bin 201 is connected to a discharge pipe 202. A flow solenoid valve 203 is provided on one side of the discharge pipe 202. A first motor 204 is fixedly connected to one side of the frame 1. A first rotating rod 205 is fixedly connected to the output end of the first motor 204. A first pulley 206 is fixedly connected to the surface of the first rotating rod 205. A second rotating rod 207 is movably connected to the front and rear ends of one side of the frame 1 through bearings. A second pulley 208 is fixedly connected to the surface of the second rotating rod 207. The second pulley 208 is connected to the first pulley 206 through a belt. One side of the first rotating rod 205 and one side of the second rotating rod 207 both penetrate into the inner cavity of the storage bin 201. A stirring rod 209 is fixedly connected to the surface of the first rotating rod 205 and the surface of the second rotating rod 207.

[0065] Specifically: The frame 1 serves as the supporting structure for the entire device, ensuring the stability of the overall structure of the fertilization device. With the cooperation of the traveling wheels 303 and the auxiliary wheel assembly 6, it achieves stable operation and improves the effectiveness of the fertilization device structure. In this embodiment, there are three storage bins 201, which are used to store different types of fertilizers. The top of the storage bin 201 is equipped with a feed inlet for easy addition of fertilizer, and the bottom is connected to a discharge pipe 202. The discharge pipe 202 is equipped with a flow solenoid valve 203, which controls the outflow of fertilizer. Each storage bin is also equipped with a liquid level sensor for real-time monitoring of the fertilizer storage amount.

[0066] Example 2

[0067] Please see Figure 1-8 Based on Embodiment 1, the feeding mechanism 3 includes a mixing chamber 301. The top of the mixing chamber 301 is connected to the feeding pipe 202. A transmission rod 302 is installed through the inner cavity of the mixing chamber 301. Both ends of the transmission rod 302 are fixedly connected to a traveling wheel 303. A third pulley 304 is fixedly connected to one side of the surface of the transmission rod 302. The bottom of the mixing chamber 301 is connected to a distribution chamber 305. A movable rod 306 is movably connected to the inner cavity of the distribution chamber 305 through a bearing. A fourth pulley 307 is fixedly connected to one side of the movable rod 306. The fourth pulley 307 is connected to the third pulley 304 through a belt. A spiral blade 308 is fixedly connected to the surface of the transmission rod 302. A material-pushing block 309 is fixedly connected to the surface of the movable rod 306. The material-pushing block 309 is disposed in the inner cavity of the distribution chamber 305.

[0068] Specifically: The mixing chamber 301 is located below the storage chamber 201 and is connected by the discharge pipe 202. The fertilizer inside the storage chamber 201 enters the mixing chamber 301 through the discharge pipe 202. Under the action of the traveling wheel 303, the transmission rod 302 and the movable rod 306 are rotated to achieve the purpose of uniformly mixing the fertilizer inside the mixing chamber 301. The spiral blades 308 are distributed on the surface of the transmission rod 302, which can form a complex flow field during the mixing process, so that the fertilizer can be fully rolled and collided in the mixing chamber 301 to achieve the purpose of efficient mixing. The inner wall of the mixing chamber 301 is also equipped with a baffle to further enhance the mixing effect of the fertilizer.

[0069] Example 3

[0070] Please see Figure 1-8Based on Embodiment 1, the sprinkler mechanism 5 includes a housing 501, one side of which is fixedly connected to a fixing frame 4. A pump body 502 is fixedly connected to one side of the top of the housing 501. One side of the pump body 502 is connected to the housing 501 via a connecting pipe, and a hose 503 is connected to one side of the pump body 502. A pipe body 504 is movably connected to one side of the fixing frame 4. One side of the pipe body 504 penetrates the inner cavity of the fixing frame 4 and extends to the outside of the fixing frame 4. A sector gear 505 is fixedly connected to the surface of the pipe body 504. A second motor 506 is fixedly connected to one side of the surface of the fixing frame 4. A circular gear 507 is fixedly connected to the output end of the second motor 506. The circular gear 507 meshes with the sector gear 505. The auxiliary wheel assembly 6 includes a support frame 601, the top of which is fixedly connected to the fixing frame 4. A spring damper 602 is fixedly connected to the inner cavity of the fixing frame 4, and the bottom of the spring damper 602 is fixed. A connecting block 603 is connected, and a moving wheel 604 is fixedly connected to one side of the connecting block 603. A reinforcing plate 7 is fixedly connected to the surface of the second motor 506, and one side of the reinforcing plate 7 is fixedly connected to the fixed frame 4. The number of material distribution bins 305 can be set according to actual needs. In this embodiment, there are five, which are evenly distributed at the bottom of the mixing bin 301. The number of material feeding blocks 309 is the same as that of the material distribution bins 305. A connecting ear 8 is fixedly connected to one side of the top of the fixed frame 4. The bottom of the connecting ear 8 is threadedly connected to the fixed frame 4 by bolts. A reinforcing rod 9 is fixedly connected to the top of one side of the fixed frame 4. One side of the reinforcing rod 9 is fixedly connected to the frame 1. A cover plate 10 is provided on the top of the storage bin 201. A handle is fixedly connected to the top of the cover plate 10. A reinforcing seat 11 is fixedly connected to the bottom of the first motor 204. One side of the reinforcing seat 11 is fixedly connected to the frame 1. A soil sensor is built into the surface of the walking wheel 303. The soil sensor monitors the soil fertility parameters in real time.

[0071] Specifically: By activating pump 502, the pump 502 draws liquid fertilizer from inside tank 501 through connecting pipe and delivers it to pipe 504 through hose 503. The fertilizer is then sprayed out from nozzles on the surface of pipe 504, achieving the effect of fertilizing with liquid fertilizer. Activating the second motor 506 drives circular gear 507 to rotate, which in turn drives sector gear 505 to rotate. Sector gear 505 then rotates pipe 504, allowing adjustment of the spray angle and improving efficiency. The sprinkler mechanism 5 is functionally designed with spring dampers 602 on the top of the moving wheels 604, ensuring efficient shock absorption during movement and guaranteeing the convenience and stability of the agricultural machinery fertilization device. One side of the reinforcing plate 7 is connected to the fixing frame 4, and the other side is fixed to the second motor 506, enabling the installation of the second motor 506 and ensuring its stability. The distribution bins 305 are evenly spaced at the bottom of the mixing bins 301, allowing for efficient use of the agricultural machinery fertilization device. 305 ensures uniform fertilizer application, guaranteeing the fertilization effect of the device and providing a balanced nutrient supply to crops. Connecting lug 8 connects to external traction equipment for easy towing and movement of the device, achieving rapid fertilization. Reinforcing rod 9 connects the fixed frame 4 and the vehicle frame 1, ensuring the stability of the connection between the fixed frame 4 and the vehicle frame 1 and improving the overall structural stability of the device. Cover plate 10 and handle seal the storage bin 201, ensuring its airtightness and preventing external impurities from entering. Reinforcing base 11 mounts the first motor 204, ensuring its stable installation. During movement, the soil sensors on the surface of the traveling wheels 303 contact the soil, enabling real-time monitoring of soil fertility. This allows for precise control of fertilizer application based on crop needs and soil fertility, preventing overuse, reducing waste, lowering agricultural production costs, and preventing environmental problems such as soil pollution and eutrophication caused by excessive fertilization, thus protecting the ecological environment.

[0072] The working principle of this invention is as follows: Workers place various fertilizers into the storage silo 201. An external controller starts the first motor 204, which drives the first rotating rod 205 to rotate. The first rotating rod 205 drives the first pulley 206 to rotate. The first pulley 206 drives the second pulley 208 to rotate via a belt. The second pulley 208 drives the second rotating rod 207 to rotate. Simultaneously, the second rotating rod 207 and the first rotating rod 205 drive the stirring rod 209 to rotate. The stirring rod 209 disperses the fertilizer inside the storage silo 201, preventing clumps of fertilizer from clogging the discharge pipe 202 and affecting the subsequent mixing effect. By opening the flow solenoid valve 203, the fertilizer enters the mixing chamber 301 through the discharge pipe 202. The device is connected to a traction device via the connecting lug 8. Under the drive of the traction device, the device moves. During movement, the traveling wheel 303 drives the transmission rod 302 to rotate, which in turn drives the spiral blade 308. The fertilizer inside the mixing chamber 301 is thoroughly mixed. The transmission rod 302 drives the third pulley 304 to rotate, which in turn drives the fourth pulley 307 via a belt. The fourth pulley 307 drives the movable rod 306 to rotate, which in turn drives the material-distributing block 309 to rotate. The material-distributing block 309 agitates the fertilizer inside the distribution chamber 305, achieving uniform fertilization during this agitation. The pump body 502 is then activated, and it extracts fertilizer from the tank via a connecting pipe. Liquid fertilizer is stored inside body 501 and transported through hose 503 to the inside of pipe body 504, where it is sprayed out from nozzles on the surface of pipe body 504, thus achieving the effect of fertilizing with liquid fertilizer. By starting the second motor 506, the second motor 506 drives the circular gear 507 to rotate, the circular gear 507 drives the sector gear 505 to rotate, and the sector gear 505 drives the pipe body 504 to rotate, which can adjust the spray angle of pipe body 504 and improve the functionality of the sprinkler irrigation mechanism 5.

[0073] Based on the agricultural machinery fertilization device provided in the above embodiments, the present invention also provides an adaptive multivariate intelligent fertilization method. For example... Figure 9 As shown, the method includes the following steps:

[0074] Step S1: Data Acquisition and Preprocessing

[0075] One week before sowing, multiple soil sensor nodes are evenly deployed in fields cultivated for gramineous crops (wheat is used as an example in this embodiment); the soil sensors are used to collect soil condition data in real time.

[0076] The soil condition data includes soil moisture H, soil temperature T, and soil electrical conductivity EC.

[0077] These soil sensors automatically record data every 30 minutes and transmit it to the central control system via a wireless network.

[0078] Example data:

[0079] - Average soil moisture H = 0.35 (volume water content)

[0080] - Average soil temperature T = 15°C

[0081] - Average conductivity EC = 0.4 dS / m

[0082] The central control system uses a Kalman filter to denoise and filter the soil condition data, resulting in... ; ;in This indicates the types of soil condition data. This represents the i-th soil condition state data after noise reduction and filtering.

[0083] The formula for the Kalman filter is:

[0084]

[0085] in F is the estimated value of the i-th soil condition state data after the k-th iteration; F is the state transition matrix, representing the natural trend of the i-th soil condition state data. is the Kalman gain, used to balance the weights of the estimated and measured values ​​of the i-th soil condition state data; H is the observation matrix, representing the linear relationship between the sensor estimates and the measured values; It is the measured value of the i-th soil condition state data in the k-th measurement;

[0086] For the soil moisture in this example, we set the initial estimate of soil moisture as follows: Prediction error covariance The measurement noise covariance R = 0.005, and the system noise covariance Q = 0.001. After multiple iterations, a more accurate estimate of soil moisture was obtained.

[0087] Step S2: Use a pre-trained multivariate nonlinear regression model to periodically predict wheat growth index y under different fertilizer application rates during the sowing period; adjust the flow solenoid valve 203 according to the optimal fertilizer application rate under the optimal wheat growth index y, thereby realizing the amount of fertilizer given in the corresponding storage bin 201, so as to fertilize wheat during the sowing period.

[0088] The amount of fertilizer applied is as follows: Indicated, n represents the types of components in the fertilizer;

[0089] The multivariate nonlinear regression model uses a Gaussian process regression (GPR) model for prediction; the kernel function used is the radial basis function (RBF).

[0090] Gaussian process regression formula:

[0091]

[0092]

[0093] in Indicate the model hyperparameters; This represents the Gaussian noise term and the measurement error. ( ) represents the Gaussian process regression function; Represents Euclidean distance, measuring the similarity of fertilizer formulations; Indicates the application amount of different fertilizers i and j; This represents the radial basis function (RBF) kernel, used to calculate covariance; The length scale parameter controls the width of the RBF kernel and determines how quickly the similarity between two points changes with distance. By adjusting the value of l, the sensitivity of the model to differences in fertilizer ratios can be controlled, thereby optimizing the prediction effect.

[0094] This embodiment assumes that the fertilizer consists of nitrogen fertilizer, phosphorus fertilizer, and potassium fertilizer, and that their application rates are as follows: Crop growth indicators y (such as chlorophyll content or plant height) can be predicted using a Gaussian process regression (GPR) model. The kernel function is a radial basis function (RBF), and the length scale parameter l is determined to be 0.5 after cross-validation.

[0095] Based on current soil conditions and the initial target yield, the preliminary calculations suggest the following optimal fertilizer ratio:

[0096] Nitrogen fertilizer

[0097] Phosphate fertilizer

[0098] Potassium fertilizer

[0099] Step S3: Use fuzzy logic to adjust the fertilizer application rate during the sowing period for wheat that has entered the jointing stage, and then apply fertilizer using the adjusted fertilizer application rate.

[0100] Based on at least one indicator in the soil condition state data, design a fuzzy rule. The rule is: "If the soil condition state data..." If the amount is below the threshold, increase the application rate of the i-th fertilizer component. ", i∈[1,n];

[0101] Let membership function For an S-shaped curve, when <threshold It approaches 1, and conversely, it approaches 0;

[0102] According to regular detection Update the application rate of the i-th fertilizer component. :

[0103]

[0104] in This indicates the updated nitrogen fertilizer application rate. This represents the maximum permissible amount of nitrogen fertilizer.

[0105] This embodiment uses a fuzzy rule for soil moisture design. As the crop enters the jointing stage, temperatures rise and water requirements increase, so the proportion of nitrogen fertilizer should be appropriately increased.

[0106] The fuzzy rule can be expressed as: "If the soil moisture H is low, increase the amount of nitrogen fertilizer applied."

[0107] Let membership function The curve is S-shaped, gradually approaching 1 when H < 0.3, and conversely, tending towards 0 when H > 0. At this point, if the average soil moisture is detected to decrease to H = 0.25, then according to the formula:

[0108]

[0109] Example:

[0110] Assuming the maximum allowable nitrogen fertilizer application rate When the soil moisture content is 0.25, the following is calculated based on the S-type membership function: Therefore, the new nitrogen fertilizer application rate is:

[0111]

[0112] Step S4: For wheat that has entered the heading stage, use a reinforcement learning optimization strategy to adjust the fertilizer application rate during the jointing stage, and then apply fertilizer using the adjusted fertilizer application rate.

[0113] The reinforcement learning optimization strategy adopts the Double Q-learning (DQN) algorithm;

[0114] Define state s t Given the soil conditions and wheat growth indicators at time t, action a t Adjust the fertilizer ratio for the current time t, and reward r. t This represents the change in the health status of the wheat at the current time t.

[0115] Update the Q value according to the following formula:

[0116]

[0117] Learning rate Control the update step size; This represents the Q-value of the current state-action pair; It is the current state, representing the set of soil condition data and crop growth indicators at the current time t, including chlorophyll content, plant height, and health status. It represents the current action, indicating the current state. The fertilization actions taken include the amount of fertilizer applied, the rate of application of liquid fertilizer by sprinkler irrigation, and the adjustment of the sprinkler angle. It is an instant reward, indicating the state. Take action below The immediate reward obtained afterward; it reflects the direct impact of the current fertilization action on crop growth and soil fertility, for example:

[0118] If the crop's health improves after fertilization, a reward will be given. It is a positive value;

[0119] If soil fertility improves after fertilization, reward [the individual]. It is a positive value;

[0120] If crop growth indicators do not change significantly after fertilization, a reward will be given. Approaching zero;

[0121] If fertilizer burn or soil salinization occurs after fertilization, reward It is a negative value;

[0122] It is a discount factor, a hyperparameter used to measure the importance of future rewards; its value ranges from [0,1]. In wheat cultivation, if... The system is highly intelligent and can select a fertilization strategy. Although the crop growth indicators may not improve significantly in the short term, it can improve the yield and quality of wheat in the long term. Indicates the next state It is an action from the set of all possible actions; it is a dummy variable used to represent any action that can be taken in the next state; it helps the agent consider the optimal behavior in the next state when updating the Q value, and achieve the learning of the long-term optimal policy; Indicates the next state In, all possible actions The maximum value among the Q values.

[0123] Assume the agent is in state The following three actions can be taken:

[0124] Action 1: Increase the amount of nitrogen fertilizer applied;

[0125] Action 2: Increase the amount of phosphate fertilizer applied;

[0126] Action 3: Increase the amount of potassium fertilizer applied.

[0127] Step S5: Adaptive Weight Update During Grouting Period

[0128] Considering the different nutritional requirements in different seasons, Bayesian optimization was used to adaptively adjust the weights of each component in the fertilizer during the heading stage, and then the adjusted fertilizer application scheme was used for fertilization.

[0129] By finding the optimal weight parameters Maximize the posterior probability of the following objective function:

[0130]

[0131] Historical data ,in This represents the application rate of all fertilizers. These are all the corresponding wheat growth indicators; It is the posterior probability, representing the parameter given historical data D. The probability distribution; It is the likelihood function, which measures the likelihood of a given set of parameters. At that time, the probability of observing historical data D; It is the prior distribution, representing the distribution of... The initial assumptions; This indicates the initial weight of each component in the fertilizer; , Let m represent the m-th wheat growth index, where m∈[1,M].

[0132] For example, during the grain-filling stage, it was found that potassium fertilizer had a particularly significant impact on grain plumpness. Therefore, by studying historical data, the weighting of potassium fertilizer was increased.

[0133] Assuming initial weights for After Bayesian optimization, the new weights Become This means that the proportion of potash fertilizer in the total nutrient supply has increased.

[0134] Step S6: Real-time monitoring and anomaly detection

[0135] Throughout the wheat growing season, the Isolation Forest algorithm is used to calculate the first and second anomaly scores. If at least one of the first and second anomaly scores is greater than a threshold (e.g., 0.7), an alarm will be triggered and the fertilization strategy will be adjusted based on the anomaly.

[0136] First abnormal score The calculation is as follows:

[0137]

[0138] in Represents soil condition data The average path length in an isolated tree, for example, 5; This represents the factor used to adjust the path length, for example, 10; t represents the current time t; n represents the expected (or average) path length of the isolated tree.

[0139] Second abnormal score The calculation is as follows:

[0140]

[0141] in Indicators of wheat growth Average path length in an isolated tree.

[0142] The specific adjustments to the fertilization strategy are as follows:

[0143] If the first abnormal score is greater than the threshold, if the soil moisture is abnormal, the irrigation strategy can be adjusted or the use of water-retaining agents can be increased or decreased; if the soil electrical conductivity is abnormal, the amount of chemical fertilizer used can be increased or decreased, and the use of organic fertilizer can be increased.

[0144] If the second abnormal score is greater than the threshold, the fertilizer ratio can be adjusted according to the specific growth indicators, such as increasing the amount of nitrogen fertilizer or potassium fertilizer.

[0145] The adaptive multivariate intelligent fertilization method described above significantly improved fertilizer utilization throughout the entire growing season, reducing unnecessary waste. Simultaneously, precise nutrient supply promoted healthy crop growth. Furthermore, this method possesses excellent scalability and adaptability, enabling customized configurations based on soil characteristics and crop varieties in different regions, providing strong technical support for modern agricultural production.

[0146] The preferred embodiments of the present invention disclosed above are only for the purpose of illustrating the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation described herein. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can better understand and utilize the present invention.

Claims

1. An agricultural mechanical fertilizer applicator with efficient mixing function, comprising a frame (1), characterized in that: A material storage mechanism (2) is fixedly connected to one side of the top of the frame (1), a material feeding mechanism (3) is fixedly connected to the inner cavity of the frame (1), a fixed frame (4) is fixedly connected to one side of the frame (1), a spraying mechanism (5) is fixedly connected to the inner cavity of the fixed frame (4), and auxiliary wheel assemblies (6) are provided at the front and rear ends of the bottom side of the fixed frame (4). The storage mechanism (2) includes multiple storage bins (201). The surface of each storage bin (201) is fixedly connected to the frame (1). The bottom of each storage bin (201) is connected to a discharge pipe (202). A flow solenoid valve (203) is installed on one side of the discharge pipe (202). A first motor (204) is fixedly connected to one side of the frame (1). A first rotating rod (205) is fixedly connected to the output end of the first motor (204). A surface of the first rotating rod (205) is fixedly connected to... The first pulley (206) is connected to the front and rear ends of one side of the frame (1) by a second rotating rod (207). The surface of the second rotating rod (207) is connected to a second pulley (208). The second pulley (208) is connected to the first pulley (206) by a belt. One side of the first rotating rod (205) and one side of the second rotating rod (207) both penetrate into the inner cavity of the storage bin (201). The surface of the first rotating rod (205) and the surface of the second rotating rod (207) are both fixedly connected to a stirring rod (209). The feeding mechanism (3) includes a mixing chamber (301), the top of which is connected to the feeding pipe (202). A transmission rod (302) is installed through the inner cavity of the mixing chamber (301). Both ends of the transmission rod (302) are fixedly connected to a traveling wheel (303). A third pulley (304) is fixedly connected to one side of the surface of the transmission rod (302). The bottom of the mixing chamber (301) is connected to a distribution chamber (305). 5) The inner cavity is movably connected to a movable rod (306) via a bearing. A fourth pulley (307) is fixedly connected to one side of the movable rod (306). The fourth pulley (307) is connected to a third pulley (304) via a belt. A spiral blade (308) is fixedly connected to the surface of the transmission rod (302). A material-pushing block (309) is fixedly connected to the surface of the movable rod (306). The material-pushing block (309) is located in the inner cavity of the material distribution bin (305).

2. The agricultural mechanical fertilizer applicator with high-efficiency mixing function according to claim 1, characterized in that: The sprinkler mechanism (5) includes a housing (501), one side of which is fixedly connected to a fixed frame (4). A pump body (502) is fixedly connected to one side of the top of the housing (501). One side of the pump body (502) is connected to the housing (501) through a connecting pipe. A hose (503) is connected to one side of the pump body (502). A pipe body (504) is movably connected to one side of the fixed frame (4). One side of the pipe body (504) penetrates the inner cavity of the fixed frame (4) and extends to the outside of the fixed frame (4). A sector gear (505) is fixedly connected to the surface of the pipe body (504). A second motor (506) is fixedly connected to one side of the surface of the fixed frame (4). A circular gear (507) is fixedly connected to the output end of the second motor (506). The circular gear (507) meshes with the sector gear (505).

3. The agricultural machinery fertilizer application device with high-efficiency mixing function according to claim 1, characterized in that: The auxiliary wheel assembly (6) includes a support frame (601), the top of which is fixedly connected to a fixed frame (4), a spring damper (602) is fixedly connected to the inner cavity of the fixed frame (4), a connecting block (603) is fixedly connected to the bottom of the spring damper (602), and a moving wheel (604) is fixedly connected to one side of the connecting block (603).

4. The agricultural machinery fertilizer application device with high-efficiency mixing function according to claim 1, characterized in that: The surface of the second motor (506) is fixedly connected to a reinforcing plate (7), and one side of the reinforcing plate (7) is fixedly connected to the fixing frame (4).

5. A method for adaptive multivariate intelligent fertilization using an agricultural machinery fertilization device with efficient mixing function as described in any one of claims 1-4, characterized in that, The method includes the following steps: Step S1: Data Acquisition and Preprocessing One week before sowing, multiple soil sensor nodes are evenly deployed in the gramineous crop fields; the soil sensors are used to collect soil condition data in real time. Step S2: Use a pre-trained multivariate nonlinear regression model to periodically predict the growth index y of gramineous crops under different fertilizer application rates during the sowing period; adjust the flow solenoid valve (203) according to the fertilizer application rate under the optimal gramineous crop growth index y, thereby realizing the amount of fertilizer given in the corresponding storage bin (201) and fertilizing the gramineous crops during the sowing period. Step S3: Use fuzzy logic to adjust the fertilizer application amount during the sowing period for gramineous crops that have entered the jointing stage, and then use the adjusted fertilizer application amount to control the flow solenoid valve (203) to realize the amount of fertilizer given in the corresponding storage bin (201), thereby fertilizing gramineous crops during the jointing stage. Step S4: For gramineous crops that have entered the heading stage, the fertilizer application rate during the jointing stage is adjusted using a reinforcement learning optimization strategy. Then, the adjusted fertilizer application rate is used to control the flow solenoid valve (203) to achieve the amount of fertilizer given in the corresponding storage bin (201), thereby fertilizing the gramineous crops during the heading stage. Step S5, Adaptive Weight Update during Grouting Period: The weights of each component in the fertilizer during the heading stage are adjusted using Bayesian optimization. Then, the flow solenoid valve (203) is controlled by the adjusted fertilizer application scheme to achieve the amount of fertilizer given in the corresponding storage bin (201), thereby fertilizing the gramineous crops during the grain filling stage.

6. The method according to claim 5, characterized in that, The soil condition data includes soil moisture, soil temperature, and soil electrical conductivity.

7. The method according to claim 5, characterized in that, After collecting soil condition data in real time, the soil sensor also performs noise reduction and filtering on the soil condition data to obtain noise-reduced soil condition data.

8. The method according to claim 5, characterized in that, In step S2, the multivariate nonlinear regression model uses the Gaussian process regression (GPR) model for prediction; the kernel function used is the radial basis function (RBF). Gaussian process regression formula: ； ； in Indicate the model hyperparameters; f represents the Gaussian noise term; f() represents the Gaussian process regression function; Indicates similarity in fertilizer formulation; Indicates the application amount of different fertilizers i and j; Represents the radial basis function (RBF) kernel; Indicates the length scale parameter; In step S3, a fuzzy rule is designed based on at least one index in the soil condition state data. The rule is: "If the i-th type of soil condition state data..." If the amount is below the threshold, increase the application rate of the i-th fertilizer component. ", i∈[1,n]; Let membership function For an S-shaped curve, when < threshold It approaches 1, and conversely, it approaches 0; According to regular detection Update the application rate of the i-th fertilizer component. : ； in This represents the updated dosage of the i-th fertilizer component. This represents the maximum permissible dosage of the i-th fertilizer component; The reinforcement learning optimization strategy described in step S4 employs the double-Q learning algorithm; Define state s t Given the soil conditions and gramineous crop growth indicators at current time t, action a t Adjust the fertilizer ratio for the current time t, and reward r. t This represents the change in the health status of grass crops at the current time t. Update the Q value according to the following formula: ； Learning rate Control the update step size; This represents the Q-value of the current state-action pair; This is the current state; This is the current action; It's an instant reward; It is a discount factor; Indicates the next state One action from the set of all possible actions; Indicates the next state In, all possible actions The maximum value among the Q values; In step S5, the optimal weight parameters are found. Maximize the posterior probability of the following objective function: ； Historical data ,in This represents the application rate of all fertilizers. These are the corresponding growth indicators for all grass crops; It is the posterior probability, representing the parameter given historical data D. The probability distribution; It is the likelihood function, which measures the likelihood of a given set of parameters. At that time, the probability of observing historical data D; It is the prior distribution, representing the distribution of... The initial assumptions; This indicates the initial weight of each component in the fertilizer; , Let m represent the growth index of the m-th gramineous crop, where m ∈ [1, M].

9. The method according to claim 5, characterized in that, The method includes real-time monitoring and anomaly detection. Specifically, during the entire growth period of gramineous crops, a first anomaly score and a second anomaly score are calculated using soil condition data and gramineous crop growth indicators, respectively. If at least one of the first anomaly score and the second anomaly score is greater than a threshold, an alarm will be triggered and the fertilization strategy will be adjusted according to the anomaly.

10. The method according to claim 9, characterized in that, First abnormal score The calculation is as follows: ； in Represents soil condition data Average path length in an isolated tree; This represents the factor used to adjust the path length; t represents the current time t; n represents the expected value of the path length of the isolated tree; Second abnormal score The calculation is as follows: ； in Indicators of growth of grass crops Average path length in an isolated tree.

Images