Fresh sheep manure intelligent fermentation treatment device based on ai visual recognition

Abstract

The application discloses an AI vision recognition-based fresh sheep manure intelligent fermentation treatment device and relates to the technical field of sheep manure intelligent fermentation appliances.The device comprises a fan;the fermentation activity comprehensive index is calculated by means of weighted summation of a fermentation determination module;the fermentation process is divided into four stages, i.e., fermentation standby, temperature rise fermentation, high-temperature composting and fermentation completion, in combination with the weight coefficient optimized by logic regression and the temperature change trend;the fermentation bin plane is segmented by means of a grid by means of an AI vision camera of a cloth distribution control module;the material accumulation height in each grid is identified in real time;the full-bin material layer average height and the material layer uniformity index are calculated;the speed and the rotation angle of an electric rotary cloth distributor are dynamically adjusted, so that the material height directly below the discharge port and the full-bin average height are kept in dynamic balance, the discharge direction is real-timely pointed to the center of the uncovered blind area, the area covered by the cloth distributor in the bin is calculated by means of integration, and the adjustment is automatically stopped when the covered area reaches the effective area of the bin body.

Description

Technical Field

[0001] This invention relates to the field of intelligent fermentation equipment for sheep manure, and more particularly to an intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition. Background Technology

[0002] In the livestock and poultry farming industry, fresh sheep manure is the main waste generated during the farming process. If it is not treated in a timely and effective manner, it will not only occupy the farming space, but may also breed bacteria, emit foul odors, and pollute the surrounding soil, water sources and air environment. Fresh sheep manure is rich in nutrients such as nitrogen, phosphorus and potassium. After fermentation, it can be transformed into high-quality organic fertilizer and realize resource utilization. Therefore, the fermentation treatment of fresh sheep manure has become a key link in the livestock and poultry farming industry chain. Currently, the fermentation of fresh sheep manure mainly relies on manual operation. The specific process is as follows: First, farmers need to enter the sheepfold and collect the fresh sheep manure scattered on the ground by sweeping and shoveling. Then, the collected manure is manually transported to the designated fermentation site. Second, after the sheep manure is piled into a fermentation heap, a designated person needs to regularly turn the heap manually. Turning the heap allows for ventilation and air exchange inside, regulating the temperature and oxygen concentration of the heap to ensure the smooth progress of the fermentation process. Finally, after fermentation is completed, the organic fertilizer still needs to be manually transported from the fermentation site to storage or transportation equipment. However, this traditional manual processing method has shortcomings. The collection and transportation of fresh sheep manure rely entirely on manual labor, which is not only extremely labor-intensive but also requires a large investment of manpower. Manual transportation is inefficient and cannot meet the manure treatment needs of large-scale farming scenarios, nor can it meet the actual need to reduce labor costs. At the same time, the uniformity of manual turning is poor, and it is difficult to accurately control the ventilation and temperature conditions of the fermentation pile, resulting in low fermentation efficiency and a fermentation cycle of 30-60 days. Furthermore, untimely or uneven turning can easily lead to incomplete fermentation, affecting the quality of organic fertilizer. It is impossible to improve fermentation efficiency and shorten the fermentation cycle, which seriously restricts the large-scale development of fresh sheep manure resource utilization. Therefore, the above problems need to be improved. Summary of the Invention

[0003] The purpose of this invention is to address the shortcomings of existing technologies by proposing an intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition, comprising a multi-layer sheepfold located on the ground, a fermentation chamber located underground, and a fan installed on the top of the multi-layer sheepfold. Each layer of the multi-layer sheepfold has a slatted floor at the bottom for sheep manure to fall through. The multi-layer sheepfold is equipped with a collection mechanism for centralized treatment of sheep manure. The fermentation chamber is equipped with a sensing mechanism for detecting the fermentation status of sheep manure. The fermentation chamber is equipped with a conveying mechanism for conveying manure inside and below the sensing mechanism. The rear end of the fan is connected to an odor treatment component. The fermentation treatment device is equipped with a fermentation determination module and a material distribution control module at its terminal. The fermentation determination module performs linear normalization on real-time temperature, oxygen concentration, humidity, and pH value; calculates the comprehensive fermentation activity index through weighted summation; and divides the fermentation state into four stages based on the FAI value: waiting to ferment, rising temperature fermentation, high temperature composting, and fermentation completed. The material distribution control module uses a vision system to grid the fermentation chamber, calculate the material height in each grid and the average height of the entire chamber, and then obtain the material layer uniformity index; it identifies the material height in the area directly below the material distributor outlet and the center coordinates of the uncovered blind spots; it dynamically adjusts the rotation speed and angle of the material distributor; it calculates the area of ​​the chamber covered by the material distributor in real time, and when the covered area reaches the effective area of ​​the chamber, it determines that there are no blind spots in the entire chamber and stops angle adjustment.

[0005] Preferably, the data analysis steps for the fermentation determination module are as follows: M1: Continuous data acquisition Historical data from a complete fermentation cycle were used to determine the minimum and maximum values ​​of each parameter using the 1st and 99th percentiles, respectively. Real-time valid data were then linearly normalized. A comprehensive fermentation activity index was calculated through weighted summation. ; M2: Determine the fermentation status based on the FAI value: when The fermentation material is determined to be in a state of unfermentation; when The fermentation process is determined to be in a state of rising temperature fermentation; when The fermentation process is judged to be in a state of high-temperature composting; when and The mixture begins to descend, indicating that fermentation is complete.

[0006] Preferably, the data analysis steps for the fabric control module are as follows: Q1: Install an electric rotary material feeder at the outlet of the sewage pipe, and use an AI vision camera to identify the height of sheep manure accumulation in the fermentation chamber in real time. The fermentation chamber was divided into a grid pattern. Calculate the average height of the material layer in the entire bin using a grid. and material layer uniformity , For the first Real-time material height within each grid; simultaneously identifies the material height in the area directly below the material distributor outlet. and the center coordinates of the uncovered blind spots ; Q2: Dynamically adjust the feeder parameters based on the visual recognition results: Adjust the feeder speed to... ,in Based on the base speed, This is the speed adjustment coefficient; adjusting the fabric feeder rotation angle is... ,in The coordinates of the material distributor's center are used; the area S covered by the material distributor in the bin is calculated in real time through integration. ,in and These are the starting and ending angles of the current rotation of the fabric feeder. An angle for AI visual recognition The maximum radius of sheep manure accumulation in the direction; when Once the effective area of ​​the storage compartment is reached, it is determined that there are no blind spots in the entire compartment, and angle adjustment is stopped.

[0007] Preferably, the collection mechanism includes a collection funnel installed on each layer of the multi-layer sheepfold and located below the manure slats, and a sewage pipe connected at one end to a blower and extending to the interior of the fermentation chamber at the other end. The sewage pipe passes through the manure slats and collection funnels on each layer. Multiple sets of sewage holes for the discharge of manure and odor are equidistantly provided on the sewage pipe, and the height of the sewage holes is 40 cm.

[0008] Preferably, the drain hole has an inner opening to prevent sheep manure from clogging the hole.

[0009] Preferably, the sensing mechanism includes two insulated fermentation chambers disposed in the fermentation bottom chamber, the insulated fermentation chambers being connected to the sewage pipe, a sensor group being installed on the four side walls of the insulated fermentation chambers at a height of half the distance from the top of the chamber, a vision camera being installed on the inner wall of the fermentation bottom chamber at an angle downwards, and an aeration pipe being provided inside the insulated fermentation chamber, the air inlet end of the aeration pipe extending through the insulated fermentation chamber to the ground.

[0010] Preferably, the sensor group includes a temperature sensor, a humidity sensor, a pH meter, and an oxygen concentration sensor.

[0011] Preferably, the conveying mechanism includes a first spiral feeder disposed at the bottom of the fermentation chamber, the bottom end of the heat-insulating fermentation chamber being connected to the interior of the first spiral feeder, one end of the first spiral feeder being connected to a second spiral feeder, and the end of the second spiral feeder being provided with a drain outlet.

[0012] Preferably, each floor of the multi-story sheepfold is equipped with a fence, and an elevator connected to each floor is installed on one side of the multi-story sheepfold.

[0013] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention utilizes the interconnected design of a multi-layered sheep pen with slatted floor, collection funnel, and sewage pipe, combined with the negative pressure airflow of a blower and the assistance of air inlet holes, to facilitate the automatic centralized transfer of fresh sheep manure to the insulated fermentation chamber, eliminating the need for manual collection and handling. Furthermore, the connection between the first and second screw feeders facilitates the automatic discharge of the fermented product, improving the convenience of transfer and thus achieving fully automated operation throughout the entire process. 2. By coordinating the control of the insulated fermentation chamber and aeration pipes, and combining the dual monitoring of the sensor group and the vision camera, it is easy to dynamically adjust the fermentation environment parameters, replacing the traditional manual turning of the pile, and improving the uniformity and stability of fermentation; furthermore, through the heat exchange coordination of the insulation structure and the heat recovery system, it is easy to preheat the fresh sheep manure with waste heat, reduce additional energy consumption, and thus achieve precise control of fermentation and energy saving. 3. Through the sealed connection and cooperation of the blower, sewage pipe and odor treatment components, it is easy to capture the malodorous gases in the sheepfold and fermentation chamber. Then, through the multi-step cooperation of physical dissolution, chemical reaction and microbial decomposition inside the odor treatment components, it is easy to convert harmful gases into harmless substances or recover methane, improve the thoroughness of purification, and thus achieve the functions of zero emission of pollutants and resource reuse. 4. By using the fermentation judgment module to calculate the comprehensive index of fermentation activity through weighted summation, and combining the weight coefficients optimized by logistic regression with the temperature change trend, the fermentation process is accurately divided into four stages: waiting to ferment, heating fermentation, high-temperature composting, and fermentation completion. Compared with the traditional method that relies on human experience to judge, this module can reflect the true activity level of sheep manure fermentation in real time and objectively, avoiding the problems of incomplete fermentation or quality decline caused by improper timing of manual turning or misjudgment of fermentation status. It significantly improves the accuracy and stability of fermentation process control, which is conducive to shortening the fermentation cycle and improving the quality of organic fertilizer products. 5. The material distribution control module utilizes an AI vision camera to divide the fermentation chamber into grids, identifying the material accumulation height within each grid in real time and calculating the average height and uniformity index of the entire chamber's material layer. It dynamically adjusts the rotation speed and angle of the electric rotary material distributor to maintain a dynamic balance between the material height directly below the discharge port and the average height of the entire chamber, while simultaneously directing the discharge direction towards the center of any uncovered blind spots. Furthermore, it calculates the area covered by the material distributor through integration, automatically stopping adjustment when the covered area reaches the effective area of ​​the chamber. This mechanism effectively solves common problems in traditional manual or fixed material distribution methods, such as localized accumulation, uneven thickness, and edge blind spots. It ensures a more uniform distribution of sheep manure within the fermentation chamber, providing a uniform material base for subsequent aeration, ventilation, and aerobic microbial fermentation. This further improves the uniformity and efficiency of fermentation, preventing fermentation stagnation or incomplete decomposition due to localized oxygen deficiency or uneven temperature, while also reducing the labor intensity of manually leveling the material layer. 6. The material distribution control module ensures the spatial uniformity of the fermentation substrate, providing a more representative monitoring environment for the sensor group. This allows the data acquired by the fermentation judgment module to more accurately reflect the overall fermentation status. The interim conclusions output by the fermentation judgment module can serve as a higher-level reference for adjusting the rotation speed and angle of the material distributor. The collaborative work of the two modules achieves a closed-loop control of "uniform material distribution → precise monitoring → intelligent judgment → dynamic adjustment," completely replacing traditional manual turning and experience-based judgment. This significantly reduces the frequency of manual intervention and labor costs, and improves the automation and intelligence level of the fermentation process, making it suitable for large-scale intensive sheep manure treatment scenarios. Attached Figure Description

[0014] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 This is a schematic diagram of the overall appearance structure of the device proposed in this invention; Figure 2 This is a schematic diagram of the overall appearance of the device proposed in this invention from another perspective. Figure 3 This is a schematic diagram of the overall cross-sectional structure of the device proposed in this invention; Figure 4 This is a schematic diagram of the multi-layer sheepfold structure proposed in this invention; Figure 5 This is a schematic diagram of the conveying mechanism and sensing mechanism proposed in this invention; Figure 6 This is a schematic diagram of the internal structure of the conveying mechanism and sensing mechanism proposed in this invention; Figure 7 The present invention proposes Figure 3 Enlarged schematic diagram of the structure at part A in the middle; Figure 8 This is a flowchart of the system proposed in this invention.

[0015] The following are the components listed in the diagram: 1. Multi-layer sheep pen; 2. Fermentation chamber; 3. Fan; 4. Slatted floor; 5. Collection funnel; 6. Sewage pipe; 7. Sewage outlet; 8. Inner corner; 9. Odor treatment component; 10. Insulated fermentation chamber; 11. Sensor group; 12. Aeration pipe; 13. Vision camera; 14. First screw feeder; 15. Second screw feeder; 16. Sewage outlet; 17. Elevator; 18. Fence. Detailed Implementation

[0016] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0017] Example 1: See Figures 1 to 7 The present invention discloses an AI-based intelligent fermentation treatment device for fresh sheep manure, comprising a multi-story sheep pen 1 located above ground, a fermentation chamber 2 located underground, and a fan 3 installed at the top of the multi-story sheep pen 1. Each level of the multi-story sheep pen 1 has a slatted floor 4 at its bottom for allowing sheep manure to fall naturally, preventing accumulation within the sheep pen and not affecting the sheep's standing and movement. The multi-story sheep pen 1 contains a collection mechanism for centralized processing of sheep manure, facilitating the rapid collection of scattered sheep manure from each level for centralized raw material for subsequent fermentation. The fermentation chamber 2 contains a sensing mechanism for detecting the fermentation status of the sheep manure, enabling real-time monitoring of key parameters during the fermentation process, thus providing intelligent... The fermentation conditions are controlled to provide data support; a conveying mechanism for transporting manure is located inside the fermentation chamber 2 and below the sensing mechanism. The conveying mechanism facilitates the rapid transfer of fermented sheep manure (or incompletely fermented raw materials) to achieve automated connection between fermentation and discharge; the rear end of the blower 3 is connected to the odor treatment component 9. The blower 3 draws negative pressure into the pipeline, and the odorous gas containing ammonia, hydrogen sulfide, small molecule sulfides, and methane enters the odor treatment component 9. After the ammonia dissolves in water, it forms ammonia water. The remaining hydrogen sulfide reacts with iron oxide to generate ferrous sulfide, elemental sulfur, and water, which are stored in the odor treatment component 9 for further treatment. The remaining small molecule sulfides are decomposed into harmless substances by microbial decomposition, and the remaining methane is collected for use as fuel.

[0018] Reference Figures 3 to 4As shown, the collection mechanism includes a collection funnel 5 located on each floor of the multi-layer sheepfold 1 and below the slatted floor 4, and a sewage pipe 6 connected at one end to a blower 3 and extending to the interior of the fermentation chamber 2 at the other end. The sewage pipe 6 passes through each layer of the slatted floor 4 and the collection funnel 5. The connection between the collection funnel 5 and the sewage pipe 6 facilitates the precise diversion of sheep manure falling from the slatted floor 4 into the sewage pipe 6, preventing it from scattering. The sewage pipe 6 passes through each layer of the slatted floor 4 and the collection funnel 5. Multiple sets of sewage holes 7 for the discharge of manure and odor are equidistantly provided on the sewage pipe 6. The height of the sewage holes 7 is 40 cm. The sewage holes 7 are provided with an inner tangent angle 8 to prevent sheep manure from clogging the holes. The inner tangent angle 8 helps to prevent sheep manure from clogging the air holes when it falls, ensuring that the ventilation channel is unobstructed.

[0019] Example 2: The technical solution is basically the same as that of Example 1, except that, as Figure 5 , Figure 6 As shown, the sensing mechanism includes two insulated fermentation chambers 10 located within the fermentation chamber 2. These chambers provide a stable temperature environment for sheep manure fermentation, reducing the impact of external temperature fluctuations on the fermentation effect. The dual-chamber design allows for alternating fermentation or processing of different batches. Each insulated fermentation chamber 10 is connected to a drain pipe 6. Sensor arrays 11 are installed on the four side walls of each chamber 10 at a height halfway from the top. These evenly distributed sensor arrays 11 facilitate comprehensive monitoring of the sheep manure status at different locations within the insulated fermentation chamber 10, ensuring accurate data monitoring. Accuracy and comprehensiveness: A visual camera 13 is installed on the inner wall of the fermentation chamber 2 at an angle downwards. The downward-angled visual camera 13 facilitates the real-time observation of the shape and accumulation of sheep manure in the insulated fermentation chamber 10 using AI visual recognition technology, assisting in judging the fermentation progress and forming a dual monitoring system with the sensor group 11. An aeration pipe 12 is installed inside the insulated fermentation chamber 10, which facilitates the introduction of air (or oxygen) into the fermenting sheep manure, adjusting the oxygen concentration of the fermentation environment and promoting efficient aerobic fermentation. The air inlet end of the aeration pipe 12 extends through the insulated fermentation chamber 10 to the ground.

[0020] Reference Figure 5 As shown, the sensor group 11 includes a temperature sensor, a humidity sensor, a pH meter, and an oxygen concentration sensor. The combination of multiple parameters such as temperature, humidity, pH, and oxygen concentration facilitates the comprehensive capture of the core environmental indicators of sheep manure fermentation, providing accurate basis for intelligent control of the fermentation process (such as aeration intensity and heat preservation temperature adjustment).

[0021] Example 3: The technical solution is basically the same as that of Example 1, except that, as Figure 5 , Figure 6As shown, the conveying mechanism includes a first spiral feeder 14 located at the bottom of the fermentation chamber 2. The bottom end of the insulated fermentation chamber 10 is connected to the interior of the first spiral feeder 14. One end of the first spiral feeder 14 is connected to a second spiral feeder 15. The end of the second spiral feeder 15 is provided with a drain outlet 16. The connection and cooperation between the first spiral feeder 14 and the second spiral feeder 15 facilitates the centralized conveying of fermentation products from multiple insulated fermentation chambers 10 to the drain outlet 16, thereby achieving automated discharge, reducing manual handling costs, and the spiral structure can prevent sheep manure from sticking and clogging during the conveying process.

[0022] Reference Figures 2 to 4 As shown, each floor of the multi-story sheepfold 1 is equipped with a fence 18. The fence 18 facilitates the division of the breeding areas of each floor, prevents sheep from falling or getting into chaos across areas, and ensures breeding safety. An elevator 17 connected to each floor is installed on one side of the multi-story sheepfold 1. The elevator 17 facilitates the up and down transportation of breeding personnel, feed or equipment, improves the convenience of breeding operations in the multi-story sheepfold 1, and reduces management difficulty.

[0023] Working Principle: In the use of this invention, firstly, the sheep flock moves on the slatted floors 4 of each layer of the multi-layer sheepfold 1. The fresh sheep manure produced falls naturally through the gaps in the slatted floors 4 into the collection funnel 5 below. Subsequently, under the action of the fan 3, a negative pressure environment is formed in the sewage pipe 6. The sheep manure in the collection funnel 5 smoothly enters the sewage pipe 6 through the sewage hole 7 with an inner tangent angle 8, and finally enters the heat-insulated fermentation chamber 10 in the fermentation bottom chamber 2. During the sheep manure fermentation process, the sensor group 11 (including temperature sensor, humidity sensor, pH meter and oxygen concentration sensor) installed on the four-way side walls of the heat-insulated fermentation chamber 10 at half the height from the top of the chamber monitors the key fermentation parameters such as temperature, humidity, pH value and oxygen concentration in the chamber in real time. The visual camera 13 installed at an angle downward on the inner wall of the fermentation bottom chamber 2 uses AI visual recognition technology to dynamically observe the shape, color and accumulation state of the sheep manure in the heat-insulated fermentation chamber 10, and assists in judging the fermentation progress. The data fed back by the visual camera 13 can be used to adjust the oxygen concentration in the heat-insulating fermentation chamber 10 by controlling the air intake of the aeration pipe 12, so as to optimize the fermentation environment and promote the efficient fermentation of sheep manure. After the sheep manure fermentation is completed, the fermentation products enter the first screw feeder 14 from the bottom of the heat-insulating fermentation chamber 10, and are transported by the first screw feeder 14 to the second screw feeder 15, and finally discharged through the sewage outlet 16 at the end of the second screw feeder 15. In addition, the fence 18 set on each layer of the multi-layer sheep house 1 can effectively divide the breeding area and ensure the safety of the sheep flock; the elevator 17 installed on one side facilitates the up and down transportation of breeding personnel, feed and equipment, and improves management efficiency. The odor generated in the sheep house is drawn into the odor treatment component 9 by the fan 3, and after a series of treatments (such as ammonia dissolving in water to form ammonia water, hydrogen sulfide reacting with iron oxide, small molecule sulfides being decomposed by microorganisms, and methane being collected for use as fuel), the emissions meet the standards or are reused.

[0024] Example 4: See Figure 8 The fermentation treatment device is equipped with a fermentation determination module and a material distribution control module at its terminal. The fermentation determination module performs linear normalization on real-time temperature, oxygen concentration, humidity, and pH value; calculates the comprehensive fermentation activity index through weighted summation; and divides the fermentation state into four stages based on the FAI value: waiting to ferment, rising temperature fermentation, high temperature composting, and fermentation completed. The material distribution control module uses a vision system to grid the fermentation chamber, calculate the material height in each grid and the average height of the entire chamber, and then obtain the material layer uniformity index; it identifies the material height in the area directly below the material distributor outlet and the center coordinates of the uncovered blind spots; it dynamically adjusts the rotation speed and angle of the material distributor; it calculates the area of ​​the chamber covered by the material distributor in real time, and when the covered area reaches the effective area of ​​the chamber, it determines that there are no blind spots in the entire chamber and stops angle adjustment; The sensors in sensor group 11 are all connected to themselves Data is collected by each probe, through Multiple probes enable a single sensor to acquire data simultaneously. Each corresponding data item is preprocessed; The collected data was sorted according to the collection time, and corresponding items collected at the same time were sorted. averaging the data and standard deviation The calculation, and the mean obtained from the calculation. and standard deviation Collect data fluctuation range for corresponding items The system is configured to compare the collected data for a given item with its fluctuation range, mark items outside the fluctuation range as outliers, and record the number of outliers. ,like If the collected data is abnormal, the data will be re-tested; if If outliers are removed, the mean of the remaining corresponding test data after outlier removal is calculated. The calculation, and the mean obtained from the calculation. As the corresponding data detected at the corresponding time; Based on the national standard for organic fertilizers (NY525-2021) and the aerobic fermentation process manual, the safety thresholds, process ranges, and optimal values ​​of each parameter were determined as the initial values ​​for normalized optimal values; continuous data were collected from the on-site fermentation system. Historical sensor data for a complete fermentation cycle were collected, outliers were removed, and statistical quantiles were calculated for each parameter. The 1st quantile of the data was the minimum value, and the 99th quantile of the data was the maximum value. Temperature normalization: , For real-time temperature data, and These represent the maximum and minimum values ​​of the temperature data; oxygen normalization: , For real-time oxygen concentration data, and These represent the maximum and minimum values ​​of oxygen concentration data, respectively; humidity normalization: , For real-time oxygen concentration data, and These represent the maximum and minimum values ​​of the humidity data; pH normalization: , For real-time valid pH data, and These represent the maximum and minimum values ​​of the pH data, respectively. Fermentation Activity Index , - The weight coefficients of the corresponding terms are equal to one; when The fermentation material is determined to be in a state of unfermentation; when The fermentation process is determined to be in a state of rising temperature fermentation; when The fermentation process is judged to be in a state of high-temperature composting; when and The material begins to descend, indicating that fermentation is complete. Collection site Sensor data from a complete fermentation cycle are used to label the fermentation status at each time point, forming a labeled dataset. Normalized sensor data is used as input, and labeled fermentation status is used as labels. Weights are optimized through logistic regression, and the regression coefficients directly correspond to the weights. After normalization, the specific values ​​of the corresponding weight coefficients are obtained.

[0025] An electric rotary spreader was installed at the outlet of sewage pipe 6 to spread out the sheep manure that was locally accumulated and of varying thickness in the fermentation chamber; AI vision camera 13 was used to identify the height of the sheep manure accumulation in the fermentation chamber in real time. Distribution uniformity Blind spots ; Material layer uniformity ,in The average height of the material layer in the entire bin. The total number of grids in the internal grid division (the AI ​​vision system divides the fermentation chamber plane into grids). For the first Real-time material height within each grid; fabric feeder speed ,in The basic rotational speed of the feeder. This is the speed adjustment coefficient. AI vision is used to identify the material height in the area directly below the material outlet of the feeder in real time; rotation angle ,in The coordinates of the center of the feeder are: The coordinates of the center of the blind zone; The area of ​​the warehouse already covered by the fabric distributor ,in and These are the starting and ending angles of the current rotation of the fabric feeder, respectively. For AI visual recognition, angle The maximum radius of sheep manure accumulation in the direction; when When the effective area of ​​the storage compartment is greater than or equal to the area of ​​the storage compartment, the system determines that there are no blind spots in the entire compartment and stops angle adjustment.

[0026] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition, comprising a multi-story sheep pen (1) located above ground, a fermentation chamber (2) located underground, and a fan (3) installed at the top of the multi-story sheep pen (1), characterized in that: Each layer of the multi-story sheepfold (1) is equipped with a slatted floorboard (4) at the bottom for sheep manure to fall. The multi-story sheepfold (1) is equipped with a collection mechanism for centralized processing of sheep manure. The fermentation chamber (2) is equipped with a sensing mechanism for detecting the fermentation status of sheep manure. The fermentation chamber (2) is equipped with a conveying mechanism for conveying manure inside and below the sensing mechanism. The fan (3) is connected to an odor treatment component (9) at the rear end. The fermentation treatment device is equipped with a fermentation determination module and a material distribution control module at its terminal. The fermentation determination module performs linear normalization on real-time temperature, oxygen concentration, humidity, and pH value; calculates the comprehensive fermentation activity index through weighted summation; and divides the fermentation state into four stages based on the FAI value: waiting to ferment, rising temperature fermentation, high temperature composting, and fermentation completed. The material control module uses a vision system to grid the fermentation chamber, calculate the material height in each grid and the average height of the entire chamber, and then obtain the material layer uniformity index. Identify the material height in the area directly below the material distributor outlet and the center coordinates of the uncovered blind spot; dynamically adjust the material distributor's speed and rotation angle; The area covered by the material distributor in the bin is calculated in real time. When the coverage area reaches the effective area of ​​the bin, it is determined that there are no blind spots in the entire bin, and the angle adjustment is stopped.

2. The intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition according to claim 1, characterized in that: The data analysis steps for the fermentation determination module are as follows: M1: Continuous data acquisition Historical data from a complete fermentation cycle were used to determine the minimum and maximum values ​​of each parameter using the 1st and 99th percentiles, respectively. Real-time valid data were then linearly normalized. A comprehensive fermentation activity index was calculated through weighted summation. ; M2: Determine the fermentation status based on the FAI value: when The fermentation material is determined to be in a state of unfermentation; when The fermentation process is determined to be in a state of rising temperature fermentation; when The fermentation process is judged to be in a state of high-temperature composting; when and The mixture begins to descend, indicating that fermentation is complete.

3. The intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition according to claim 1, characterized in that: The data analysis steps for the fabric control module are as follows: Q1: Install an electric rotary material feeder at the outlet of the sewage pipe, and use an AI vision camera to identify the height of sheep manure accumulation in the fermentation chamber in real time. The fermentation chamber was divided into a grid pattern. Calculate the average height of the material layer in the entire bin using a grid. and material layer uniformity , For the first Real-time material height within each grid; simultaneously identifies the material height in the area directly below the material distributor outlet. and the center coordinates of the uncovered blind spots ; Q2: Dynamically adjust the feeder parameters based on the visual recognition results: Adjust the feeder speed to... ,in Based on the base speed, This is the speed adjustment coefficient; adjusting the fabric feeder rotation angle is... ,in The coordinates of the material distributor's center are used; the area S covered by the material distributor in the bin is calculated in real time through integration. ,in and These are the starting and ending angles of the current rotation of the fabric feeder. An angle for AI visual recognition The maximum radius of sheep manure accumulation in the direction; when Once the effective area of ​​the storage compartment is reached, it is determined that there are no blind spots in the entire compartment, and angle adjustment is stopped.

4. The intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition according to claim 1, characterized in that: The collection mechanism includes a collection funnel (5) set on each layer of the multi-layer sheepfold (1) and located below the manure slat (4), and a sewage pipe (6) with one end connected to a blower (3) and the other end extending into the fermentation chamber (2). The sewage pipe (6) passes through the manure slat (4) and the collection funnel (5) of each layer. Multiple sets of sewage holes (7) for the discharge of manure and odor are equidistantly opened on the sewage pipe (6). The sewage holes (7) are 40 cm high.

5. The intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition according to claim 4, characterized in that: The drain hole (7) is provided with an inner tangent (8) to prevent sheep manure from clogging the hole.

6. The intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition according to claim 1, characterized in that: The sensing mechanism includes two insulated fermentation chambers (10) set in the fermentation bottom chamber (2). The insulated fermentation chambers (10) are connected to the sewage pipe (6). Sensor groups (11) are installed on the four side walls of the insulated fermentation chambers (10) at a height of half the height from the top of the chamber. A vision camera (13) is installed on the inner wall of the fermentation bottom chamber (2) at an angle downward. An aeration pipe (12) is provided inside the insulated fermentation chamber (10). The air inlet end of the aeration pipe (12) extends through the insulated fermentation chamber (10) to the ground.

7. The intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition according to claim 6, characterized in that: The sensor group (11) includes a temperature sensor, a humidity sensor, a pH meter and an oxygen concentration sensor.

8. The intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition according to claim 6, characterized in that: The conveying mechanism includes a first spiral feeder (14) located at the bottom of the fermentation chamber (2). The bottom end of the heat-insulating fermentation chamber (10) is connected to the interior of the first spiral feeder (14). One end of the first spiral feeder (14) is connected to a second spiral feeder (15). The end of the second spiral feeder (15) is provided with a drain outlet (16).

9. The intelligent fermentation treatment device for fresh sheep manure based on AI visual recognition according to claim 1, characterized in that: Each floor of the multi-story sheepfold (1) is equipped with a fence (18), and an elevator (17) connected to each floor of the sheepfold is installed on one side of the multi-story sheepfold (1).

Images