Intelligent precise feeding method and system based on pig farm unit, corollary equipment and monitoring and business use terminal
By combining aisle scales and environmental control systems with inspection robots in pigsty units, the nutritional requirements are dynamically calculated and feeding strategies are generated, solving the problem of mismatch between nutritional supply and demand in the pig farm feeding system, and achieving precise feeding and improved growth performance.
Patent Information
- Application Number
- CN202610196707.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-11
- Publication Date
- 2026-05-15
AI Technical Summary
Existing pig farm feeding systems lack the ability to perceive and dynamically coordinate with individual differences in pigs and changes in the environment in real time, resulting in a mismatch between nutrient supply and actual needs, leading to feed waste and unstable growth performance.
By setting up aisle scales in pigsty units to collect pig weights, and combining data from the environmental control system and health indicators from inspection robots, nutritional requirements are dynamically calculated. Based on user feed types and models, precise feeding strategies are generated, and closed-loop control is achieved using intelligent mixing equipment.
This achieves precise matching of nutrient supply and demand for pigs, reduces feed waste, and improves growth performance and feed utilization.
Smart Images

Figure CN122030330A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent precision feeding for animals, and in particular to an intelligent precision feeding method, system, supporting equipment, monitoring and business terminal based on a pig farm unit. Background Technology
[0002] Currently, my country's pig farming industry is undergoing a profound transformation from traditional scale expansion to a quality- and efficiency-oriented model. With feed costs accounting for 60%-70% of total farming costs, achieving cost reduction, efficiency improvement, and environmental protection through precise nutritional supply has become a core competitive advantage for the industry. While the traditional "staged feeding model" represents a significant improvement over single-feed systems, a significant mismatch remains between its inherent "constant formula within each stage" and the "daily changing nutritional needs" of fattening pigs. To ensure the needs of the fastest-growing individuals within a stage, the formula is often designed with a high nutritional level. However, for pigs at other growth rates within that stage (especially in later stages), the feed inevitably contains excess nutrients, resulting in waste and environmental pollution through excrement. Fixed formulas cannot respond to the daily fluctuations in nutritional needs caused by factors such as health and the environment, thus hindering the optimal optimization of production performance.
[0003] Against this backdrop, a "daily dynamic precision feeding model" has become both possible and necessary. This model provides feed formulations that meet the optimal nutritional requirements of pigs on a daily basis, taking into account changes in their daily weight, environmental changes, and health status. This achieves a perfect match between the daily nutritional needs and the nutritional supply for each pig. The core of this model lies in refining the feeding perspective from the "group stage" to the "individual daily stage," dynamically adjusting the diet formulation to ensure that the nutritional supply closely approximates the actual demand curve of the pigs.
[0004] Furthermore, in the current evolution of pig farm feeding management from automation to intelligence, existing pig farm feeding parameter settings are relatively simple, often only considering pig weight or age. They often fail to consider the differences and correlations between different growth stages, weight structure, environmental conditions, feeding behavior, and health indicators, or lack sufficiently mature calculation models. This results in a mismatch between nutrient supply and actual needs, leading to problems such as feed waste, decreased feed utilization, and unstable pig growth performance. For example, environmental factors such as temperature, humidity, and air quality in the pigsty have a significant impact on pig feed intake and nutritional requirements. Traditional feeding systems typically lack real-time sensing and dynamic linkage capabilities for environmental indicators, making it impossible to adjust feeding strategies promptly according to environmental changes. In addition, most existing systems only achieve quantitative feed delivery, lacking a continuous evaluation and optimization mechanism based on leftover feed feedback, making it difficult to form a closed-loop control and limiting further improvements in precision feeding. Therefore, there is an urgent need for an intelligent precision feeding method, system, supporting equipment, monitoring, and operational terminals based on pig farm units. Summary of the Invention
[0005] The purpose of this invention is to address the deficiencies in the existing technology by proposing an intelligent and precise feeding method, system, supporting equipment, monitoring and business terminal based on a pig farm unit.
[0006] To achieve the above objectives, the present invention adopts the following technical solution:
[0007] A smart precision feeding method based on pig farm units, the specific steps of which are as follows:
[0008] I. Accurately collect daily weight data of pigs using aisle scales set up in pigsty units;
[0009] II. Regularly acquire environmental indicators within the pigsty based on the data interface of the existing environmental control system;
[0010] III. Based on the remaining feed volume of each unit's sub-feed tower in the previous 5 days and combined with the health abnormality events collected by the inspection robot, assess the health indicators of the pigs.
[0011] IV. Based on the daily body weight, real-time environmental indicators, health indicators obtained from I, II, and III, and existing nutritional models, accurately calculate the daily nutritional requirements of pigs;
[0012] V. Based on the actual production situation, obtain the type of feed currently used by the user;
[0013] VI. Calculate the dosage and ratio of various feeds based on the obtained nutritional requirements, the type of feed currently used by the user, and the existing feed formulation model;
[0014] VII. Automatically create execution strategies for daily material feeding based on user-defined feeding strategies;
[0015] VIII. Monitor the operation of the mixing equipment in real time according to the dynamically generated feeding execution strategy.
[0016] As a further aspect of the present invention, the specific steps for accurately collecting the weight of pigs using the aisle scale set up in step I of the pigsty unit are as follows:
[0017] S1.1: Within the pigsty unit, each aisle is designated as a feeding area and a drinking area. The two areas are connected by a one-way door for pigs to pass through in one direction. That is, after a pig finishes eating, it can only go from the feeding area to the drinking area to drink water through the aisle, and vice versa. When a pig needs to eat, it can only enter the feeding area from the drinking area through the one-way door located symmetrically in the aisle of the pen.
[0018] S1.2: While the weight of the pigs is obtained through the aisle scale, the pig's ear tag is read by the built-in RFID identification device of the aisle scale to identify the pig's identity, thereby enabling targeted tracking and monitoring of the feeding situation of individual pigs;
[0019] S1.3: The cloud center synchronously uploads the pig weight information obtained from the aisle scale, and after cleaning, filtering and clustering, obtains the accurate weight and number of pigs in each unit of the pig house.
[0020] S1.4: The average weight of pigs is obtained by taking the K-Means average weight of the pigs and then verifying the reasonableness of this average weight value with yesterday's average weight. Unreasonable values are removed, and the next batch of aisle weighing data is waited for to be processed again. If the average weight of pigs cannot be calculated after processing all batches of aisle weighing data on the same day, that is, the weight obtained from each batch is invalid, then traceability processing is performed.
[0021] S1.5: The cloud-based system processes weighing data from the aisle scales to obtain pig weights and simultaneously identifies individual pigs to determine the number of pigs that have consumed feed that day. This data is then compared with known production parameters in the system. When the difference exceeds a threshold, the system alerts management personnel to intervene manually and calculates nutritional requirements based on the number identified the previous day. This avoids calculation errors caused by pig ear tags falling off, being damaged, or becoming invalid.
[0022] As a further aspect of the present invention, the specific steps of the traceability processing in step S1.4 are as follows:
[0023] S2.1: During the traceability process, based on the time series set of the weight values of each pig on the day, compare them in reverse order with the weight of the pig yesterday until a reasonable value is found. After the traceability process of all pigs is completed, perform K-Means mean processing on all values again to obtain the average weight of pigs in the unit on the day.
[0024] S2.2: If the traceability process still fails and the average weight of the unit on the day cannot be obtained, the system will send an alarm message to prompt the management personnel to intervene manually. On the other hand, it will fit the average weight of the day as the parameter required for the calculation of nutritional requirements based on the average value of yesterday and historical data using a univariate polynomial formula.
[0025] S2.3: When tracing and processing the weighing data of the aisle scale, it is also a process of secondary integration of the weight of each pig. The above-mentioned invalid data is also specially identified. If a pig is found to have a weight decrease relative to the previous day or a weight gain lower than the preset change value, the system will automatically alarm to prompt the management personnel to intervene manually. At the same time, the pathological condition or health factors of the pig will be analyzed, and the existing nutritional requirement model will be dynamically adjusted.
[0026] As a further aspect of the present invention, the specific steps for evaluating pig health indicators based on the remaining feed volume of each unit's sub-feed tower in the previous 5 days and in conjunction with health abnormality events collected by the inspection robot in step III are as follows:
[0027] S3.1: The presence of leftover feed indicates a positive deviation in the calculation of nutritional requirements. Excluding cases where a large number of pigs are moved to different pens due to production needs, the presence of leftover feed is due to adverse health conditions in a group or multiple pigs. The system will automatically send a warning message to the management personnel and will also automatically analyze and model based on the feeding, consumption, leftover feed and existing leftover feed models over the past 5 days.
[0028] S3.2: In the feed surplus analysis model, the number of abnormal pigs is estimated based on the amount of feed surplus and the daily nutritional requirements of a single pig; and based on the feed surplus situation over 5 days, including the quantity, numerical change curve, time period of feed surplus occurrence, duration, and frequency of occurrence, it is determined whether it is a valid abnormality and the possible abnormality type is assessed.
[0029] S3.3: Compare the leftover feed situation of other pig house units using the same or highly similar types of feed to assess whether there are feed source problems, and automatically alarm to prompt the existence of feed sources;
[0030] S3.3: The assessment of pig health indicators can be optimized based on abnormal pig health events obtained by the inspection robot, including collecting pig body temperature through infrared cameras and collecting abnormal sounds from pigs through sound collectors, thereby determining whether there are potential disease risks such as fever, cough, and asthma.
[0031] S3.4: Based on the above assessment, information on the number of pigs with adverse health conditions and the types of pathological conditions is obtained, and the nutritional requirements and nutritional ratios are dynamically adjusted in a targeted manner.
[0032] As a further aspect of the present invention, the specific steps in step VI of obtaining the amount and ratio of each type of feed for the user based on the obtained nutritional requirements, combined with the type of feed currently used by the user and the existing nutritional formula model, are as follows:
[0033] S4.1: Based on the typical bulk raw material types in the feed tower corresponding to the feeding unit, the corresponding energy, protein, amino acids, macro and micro elements, and vitamin values are adapted from the existing nutritional model, and the previously calculated nutritional requirements are used as the ratio target to obtain the optimal ratio combination.
[0034] S4.2: Based on the different concentrations of complete feed types in the feed tower, and using the calculated nutritional requirements as the target ratio, obtain the optimal ratio of complete feed at each concentration while meeting the nutritional needs for energy, protein, amino acids, macro and micro elements, and vitamins.
[0035] As a further aspect of the present invention, the actual feeding strategy of the user described in step VII is specifically a user-defined, simple and direct feed feeding method requirement.
[0036] The execution strategy is specifically based on the user-defined actual feeding strategy. The system automatically creates an equipment operation strategy based on factors such as the feed input required by each target pig house unit, the corresponding equipment feeding capacity, mixing efficiency, conveying capacity, equipment operation and maintenance requirements, and the distance between the feeding equipment and each pig house unit. This generates the optimal execution configuration for equipment operation and the daily feeding needs of the pig house units. This includes automatically optimizing the feeding sequence of each feeding pig house and the configuration of each batch of feed.
[0037] As a further aspect of the present invention, the specific steps of step VIII, which involve batch feeding of the material tower, mixing, and unit conveying using the asynchronous parallel strategy of the motors in each part of the mixing equipment system, are as follows:
[0038] S5.1: The feeding motor can be divided into three stages for a single batch feeding process: motor start-up stage, stable operation stage, and remaining material dropping after shutdown. The remaining time for the motor to stop running in this batch is dynamically calculated in real time using a calculation formula composed of the amount of material to be fed, the current weighing value, the remaining material dropping after shutdown, and the feeding amount per unit time.
[0039] S5.2: When the feeding motor starts feeding each batch, it adopts a two-start method. The first start is used to obtain the remaining material dropping amount after the shutdown in the calculation formula. The second start is completed by the immediate weighing after the shutdown. When the motor is detected to be running smoothly, the motor stops running and the remaining material dropping amount after the shutdown is obtained based on the weighing value measured at the time of shutdown and the lag measurement.
[0040] S5.3: Real-time calculation and update of current weighing value and feed rate per unit time. At the same time, based on three parameters, namely the current feed line feeding capacity, the amount of feed to be fed in this batch, and the feeding accuracy that the system needs to achieve, the sampling period for real-time weighing is set.
[0041] S5.4: The feeding, mixing and conveying operations are performed asynchronously and in parallel. When the feeding is completed and the material is discharged to the mixing unit, the discharge valve can be closed without waiting for the mixing to be completed, and feeding can continue. The conveying unit only needs its built-in level sensor to detect the presence of feed to perform the conveying action.
[0042] A smart precision feeding system based on a pig farm unit includes a pig weighing data synchronization module, an environmental control system interface module, a cloud center residual feed monitoring module, a data persistence module, and a motor drive algorithm module.
[0043] The pig weighing data synchronization module of the aisle scale uses the built-in RFID identification device of the aisle scale to identify the pig's ID and the weighing sensor to obtain the weighing value, and synchronizes it to the cloud center through the Internet of Things gateway.
[0044] The environmental control system interface module is used to synchronize environmental indicators with the environmental control system at regular intervals, including temperature, humidity, and information on carbon dioxide, ammonia, hydrogen sulfide, PM2.5, and PM10.
[0045] The cloud-based residual feed monitoring module is used to store and analyze the residual feed values of each feeding synchronized by the weighing sensor of the pigsty feed tower, and to provide feedback to improve the nutrient requirement model based on the residual feed analysis model.
[0046] The data persistence module is used to store the established nutrition model, feed formulation model and user-defined feeding strategy for a long time, and to provide model and strategy support when calculating various feed indicators;
[0047] The motor drive algorithm module is used to accurately and efficiently implement material feeding and control the motor running time based on real-time calculations of the feeding situation.
[0048] A set of intelligent precision feeding equipment based on a pig farm unit includes: a feeding unit, a medication unit, a mixing unit, and a conveying unit;
[0049] The feeding unit equipment includes one or more feed towers for feeding various types and levels of feed components required for feeding, as well as a motor, a matching auger, and a weighing system for controlling the feeding.
[0050] The drug feeding unit includes a drug storage bin, a weighing bin, and a discharge valve, which are used to accurately dispense drugs or micronutrients into the mixing bin according to the drug-material ratio.
[0051] The mixing unit equipment includes a high-power mixing motor, a reducer bearing, a paddle agitator, and a discharge valve, which are used to complete dynamic and efficient mixing.
[0052] The feeding unit equipment is used to accurately deliver the correctly mixed feed to the corresponding pigsty unit, and includes a feeding motor with a feed disc feed line, control valves for feeding each sub-feed tower, and a weighing system.
[0053] A smart precision feeding monitoring and business terminal based on a pig farm unit, wherein the monitoring and business terminal specifically includes a desktop terminal, a mobile terminal and a central display screen;
[0054] The desktop terminal is used by daily supervisors to monitor the current operation of the pig farm in real time, including feed delivery, pig growth, and faults and alarms that occur during operation.
[0055] The mobile terminal is used by on-site technicians to make real-time adjustments to operating equipment and various strategies or to handle fault repairs based on the actual needs of the site, using handheld devices or mobile phones.
[0056] The central display screen is used to macroscopically display the comprehensive operating data of the entire pig farm or group user for the year, including various statistical reports, views, and trend analyses.
[0057] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0058] This invention accurately collects basic indicators such as the weight and number of pigs in each unit of a pig farm daily, as well as real-time temperature, humidity, and environmental indicators such as carbon dioxide, ammonia, hydrogen sulfide, PM2.5, and PM10 in the pigsty. It also comprehensively assesses pig health indicators based on the remaining feed in the sub-feed towers for five consecutive days and health anomalies collected by the inspection robot. A nutritional model is used to obtain the precise daily nutritional requirements for each unit. Furthermore, based on the type of feed currently used in the user's production, a feed model is used to obtain the specific feed ratios. Based on the feed ratios, nutritional requirements, and feeding strategies for each unit obtained from the feed model, specific execution strategies can be generated for different pigsties on a daily basis. The execution strategy includes the execution details of each stage of the intelligent mixing equipment's batch feed tower feeding, mixing, and unit feeding. By obtaining the remaining feed in each unit's sub-feed tower, it determines whether to adjust the nutritional model calculation parameters and adjust the feeding strategy in real time. Through the above method, the present invention provides a closed-loop response to the daily dynamic and differentiated fine and precise feeding needs of each pig house, and greatly improves the matching degree between the feeding amount and the nutritional requirements of individual pig growth. Attached Figure Description
[0059] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.
[0060] Figure 1 This is a flowchart illustrating an intelligent precision feeding method, system, supporting equipment, monitoring, and business terminal based on a pig farm unit proposed in this invention.
[0061] Figure 2 This is a schematic diagram of the feeding process of an intelligent precision feeding method, system, supporting equipment, monitoring and business terminal based on a pig farm unit proposed in this invention;
[0062] Figure 3 This is a schematic diagram illustrating the calculation of nutritional requirements for an intelligent and precise feeding method, system, supporting equipment, monitoring, and business terminal based on a pig farm unit, as proposed in this invention.
[0063] Figure 4 This is a schematic diagram illustrating the calculation of leftover feed and nutritional requirements for an intelligent precision feeding method, system, supporting equipment, monitoring, and business terminal based on a pig farm unit proposed in this invention.
[0064] Figure 5This is a schematic diagram of the feeding / medication and mixing process of an intelligent precision feeding method, system, supporting equipment, monitoring and business terminal based on a pig farm unit proposed in this invention;
[0065] Figure 6 This is a schematic diagram of the feeding process of an intelligent precision feeding method, system, supporting equipment, monitoring and business terminal based on a pig farm unit proposed in this invention;
[0066] Figure 7 This is a flowchart illustrating the feeding accuracy control process of an intelligent precision feeding method, system, supporting equipment, monitoring, and business terminal based on a pig farm unit proposed in this invention.
[0067] Figure 8 This is a diagram illustrating the operational scenario of an intelligent precision feeding method, system, supporting equipment, monitoring, and business terminal based on a pig farm unit proposed in this invention. Detailed Implementation
[0068] 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.
[0069] Example 1, referring to Figure 1-8 This embodiment discloses an intelligent precision feeding method based on a pig farm unit, and its specific implementation is described below:
[0070] [1], reference Figure 1-2 As can be seen, the system uses aisle scales to measure and collect pig weights. The measured weights and individual identification codes of the pigs are synchronized to the cloud center via a message queue, where they are then cleaned and analyzed. Table 1 shows a sample of the collected data, including identification codes, collection timestamps, and weighing information. The cloud center groups data according to the identification codes and sorts it by timestamp, while also identifying abnormal data. Abnormal data refers to weighing values that fluctuate above or below a reference value exceeding a set threshold (usually greater than 2 kg). Abnormal data is directly filtered out. When determining the reasonable range of the weighing fluctuation interpolation, the reference value can be selected in several ways.
[0071] Specifically, if the individual's weight from the previous day is used as a reference, and after clustering the collected values for the day, most sample values are within a reasonable threshold, then the collected values for the day are considered normal. If, again, the individual's weight from the previous day is used as a reference, and after clustering the collected values for the day, no or only a small portion of sample values are within a reasonable threshold, then the collected values for the day are considered abnormal, and an abnormal weighing analysis is required. This may be due to pathological conditions or feed-related abnormalities in collective feed intake, and is usually accompanied by abnormal leftover feed events. If it is the first day of weighing, there is no reference value from the previous day. In this case, after collecting a sufficient number of weighing values, the batch of numbers can be clustered, and the mean K-Means of the largest selected cluster can be used as the reference value for weighing on the day.
[0072] It should be further noted that the sample data collected is shown in the table below:
[0073] Table 1
[0074]
[0075] Table 1 above specifically shows the weighing data collection dataset.
[0076] [2] When measuring the weight of pigs, the identification code of the pig can be obtained at the same time. When the number of pigs in the relevant unit of the production system cannot be obtained, this information can be used as a substitute for determining the number of pigs. The system records the effective weight value of each pig on a daily basis according to the identification code. This record can be used to track the feeding and growth of each pig. When performing weighing anomaly analysis in [1], it can be traced back, that is, it can be judged whether the pig has a pathological condition through historical weighing data. At the same time, an alarm is issued to the relevant management personnel. For the unidentified pigs, it is usually due to ear tag damage, loss, collection equipment failure or data abnormality. The number of unidentified weighing values on the day, the sampling time period, the distribution, etc. can be used to determine whether it is due to abnormal sampling data, abnormal communication link, collection equipment failure, invalid individual ear tag, etc., so that targeted anomaly analysis and alarm prompts can be carried out, which is convenient for management personnel to carry out subsequent tracking or manual intervention.
[0077] It should be further noted that the effective weight values of each pig after daily recording are shown in the table below:
[0078] Table 2
[0079]
[0080] Table 2 above shows the daily weighing processing results.
[0081] [3] When the identification code described in [2] is invalid, the weighing data without identification code can generally be clustered without a fixed number of clusters. The K-means value of each cluster result is combined with the pig weighing records described in [2] (i.e., Table 2) to perform a univariate polynomial equation fitting, thereby identifying and compensating the value without identification code. The results are then verified from the subsequent data collected on the same day, and finally a highly reliable identification code fitting result can be obtained. If the identification code cannot be automatically matched based on the value collected on the same day, the system will issue an alarm to prompt the management personnel to intervene.
[0082] [4] When the effective weight and number of pigs cannot be obtained through the methods [1] to [3], while alerting the management personnel to intervene, the number, weight and health abnormalities of pigs can be obtained by the inspection robot. This includes behavior recognition to identify long periods of stillness, frequent movement, and abnormal drinking / feeding frequency to determine potential health risks; voiceprint recognition to identify possible disease risks such as coughing and asthma in pigs; calculation of the number of pigs based on group density, spatial distribution, and distance radius; and acquisition of the three-dimensional dimensions of pigs based on depth cameras, importing them into models to obtain estimated weights of pigs.
[0083] [5] By inputting the calculated average weight and number of pigs into the existing nutritional model, the basic daily nutritional requirements of each pig house can be obtained. In actual production, environmental factors are also important conditions affecting the feed intake of pigs. These factors include temperature, humidity, and environmental indicators such as carbon dioxide, ammonia, hydrogen sulfide, PM2.5, and PM10. These can be retrieved periodically through the interface of the existing environmental control system and input into the nutritional model to obtain the nutritional requirements of each pen that are more suitable for the current actual pig farming environment.
[0084] [6] When incorporating environmental factors into the nutritional model calculation, the model is adapted to the pig's growth stage based on the average weight (or age) of the pigs, including nursery, piglets, medium-sized pigs, and large pigs, and the influence weights of each environmental factor indicator are set accordingly. Usually, the influence weights of each environmental factor indicator will gradually decrease according to the growth stage of the pigs from small to large.
[0085] [7] Among environmental factors, the influence weight of temperature and humidity will be reduced when they are within a certain threshold range. This threshold is called the comfort zone, that is, the temperature and humidity suitable threshold. Conversely, when outside the threshold, the influence weight of environmental factors will be greater. For example, high temperature and high humidity will cause heat stress in pigs and reduce their appetite.
[0086] [8] When harmful gases such as ammonia and hydrogen sulfide in the air quality index exceed the concentration threshold, they have a significant negative impact on the feed intake of pigs.
[0087] [9] In addition to supporting basic calculations of pig weight, the nutrition model also supports data modeling based on multiple variables such as pig growth stage, temperature, humidity, and harmful gases. When environmental factors exceed the threshold, an alarm will be automatically triggered to prompt managers to intervene in a timely manner.
[0088]
[10] The residual feed analysis described in a smart precision feeding method based on pig farm units (buildings) is an optimized feedback condition for the precise calculation of nutritional requirements. That is, when there is an abnormality in residual feed in a pig farm unit, the residual feed analysis model determines whether there is a pathological abnormality, feed abnormality, or positive deviation in the model calculation, and then adjusts the nutritional requirements and duration accordingly. Among them, pathological abnormalities can be obtained through the variation analysis of weighing values in [1]; feed abnormalities can be obtained by horizontal comparison with other pig farm units with the same feed type; the rest can be attributed to positive deviation in the calculation model.
[0089]
[11] Based on the user's actual feed type (such as corn, soybean meal, wheat bran, premix or various complete feeds with different nutrient concentrations) and the calculated nutrient requirements, the optimal formula model can be established from the aspects of energy, protein, amino acids, macro and micro elements, vitamins and other nutrients. That is, under the premise of meeting the healthy growth and development of pigs, the amount of feed is controlled, the feeding cost is reduced and feed waste is avoided. As shown in Table 3, it is a feed formula table based on the pig conditions, environmental conditions and two different concentrations of complete feed.
[0090] It should be further noted that the effective weight values of each pig after daily recording are shown in the table below:
[0091] Table 3
[0092]
[0093] Table 3 above shows the results of the nutrient requirements calculation.
[0094]
[12] The feeding strategy described in a smart precision feeding method based on pig farm units (buildings) refers to the calculation of the feeding order and corresponding feeding time of each pig house during each feeding round. The above method is completed by a set of equipment, which corresponds to the feeding tasks of multiple sets of feeding pig house units. The distance of each pig house from the equipment, the amount of feed to be fed, and the feeding line capacity determine the equipment feeding time of each pig house unit, and the specific calculation is as follows:
[0095]
[0096] The feed rate per second (Qs) is obtained by multiplying the feed line velocity per second (v), feed density (y), feed line cross-sectional area (πr²), and feed filling coefficient (K).
[0097]
[0098] The feeding time (Ts) of each feeding unit is obtained by dividing the required feed amount (A) by the time (Qs) of the feed amount per second, plus the idle time of the feed when it arrives at the feeding unit (l / v, where l is the distance between the feeding unit and the mixing equipment).
[0099] The system defaults to sorting the units in ascending order based on the runtime obtained from the above indicators, and setting the material feeding order of each unit in ascending order of runtime.
[0100] Meanwhile, in a smart precision feeding method based on pig farm units (buildings), the feeding accuracy is achieved by a feeding motor drive algorithm. Specifically, based on the unit time feeding volume of the feed line, the batch feeding volume (maximum being the feed hopper capacity), and the accuracy setting value (e.g., 0.3%), the weighing sampling period is dynamically set, and the current feeding weight and the feeding volume (feeding volume per unit time) of each sampling period are obtained in real time. Finally, based on the real-time obtained current weight and periodic feeding volume, the motor running time is calculated and adjusted in real time. For example:
[0101] When using a 125 feed line, the theoretical feed rate per unit time = (4200kg / 3600s) * 10 = 11.6g / 10ms. Based on accuracy requirements, we can set the sampling time for the feed weighing data to 10ms, and the feed motor running time is:
[0102]
[0103] The motor stop time (Ts) is calculated by subtracting the amount of feed remaining after the motor stops (Wr) and the current amount of feed already fed from the amount of feed to be fed (A), and then dividing by the amount of feed fed per unit sampling time (W). The current amount of feed already fed is the difference between the real-time feed weighing value (Wn) and the weighing value before the motor starts (Ws).
[0104]
[13] , reference Figure 1-4 It is understood that a smart precision feeding method based on pig farm units (buildings) includes feed feeding, mixing, and conveying processes. Furthermore, based on the user's actual pig farming needs, the pig units will be subject to periodic or irregular vaccination or micronutrient supplementation, which can be achieved through vaccination. For example... Figure 4 (The flowchart for feeding / medication and mixing is shown). The number of feeding rounds shown in the flowchart refers to the total amount of feed to be fed to a pig house unit divided by the capacity of the medicine storage bin, that is, how many batches (bins) are needed to complete the total amount of feed supplied to each pig house in this cycle.
[0105]
[14] , Reference Figure 1-2As can be seen from 6, in the feeding process, the running time of the feeding motor needs to take into account the calculation of the delayed shutdown time. This is because after the feeding stops, there is still a distance between the mixing equipment and the feeding pig house. During this distance, there is feed that has not yet been delivered to the required pig house unit feed tower.
[0106] Example 2, refer to Figure 5-7 This embodiment discloses an intelligent precision feeding system based on a pig farm unit. The system applies the aforementioned intelligent precision feeding method based on a pig farm unit, including a pig weighing data synchronization module, an environmental control system interface module, a cloud center residual feed monitoring module, a data persistence module, and a motor drive algorithm module.
[0107] [1] As described above, we know that the pig weighing data synchronization module and the environmental control system interface module are used for calculating nutritional requirements and provide model calculation parameters, such as... Figure 3 As shown, the system performs actual calculations through an interactive interface. Users can adjust the calculation parameters in multiple dimensions according to their actual needs, including adjustments to trace elements, feed ratios, and health factors.
[0108] [2] As described above, the calculation of nutritional requirements can be optimized and fed back through residual material analysis, such as... Figure 7 As shown, the system provides an interactive interface where users can visually view the results of feed waste analysis and automatically adjust the feed ratio accordingly to suit actual production conditions and avoid feed waste.
[0109] [3] Based on the feeding accuracy control described above, the feeding system provides a corresponding motor drive operation monitoring and interactive operation interface. The specific operation status of the feeding process can be viewed, mainly including the operation status of the motors in the feeding, mixing, and conveying stages, as well as the feeding status of the pigsty unit currently being fed (including feed quantity, amount to be fed, and duration). The motor operating parameters can also be adjusted according to actual production needs, such as conveying time and feed compensation adjustment. For details, please refer to [link / reference]. Figure 7 As shown.
Claims
1. A smart precision feeding method based on a pig farm unit, characterized in that, The specific steps for this feeding method are as follows: Ⅰ. Accurately collect pig weights using aisle scales set up in pigsty units; II. Obtain environmental indicators within the pigsty based on the data interface of the existing environmental control system; III. Based on the remaining feed volume of each unit's sub-feed tower in the previous 5 days and combined with the health abnormality events collected by the inspection robot, assess the health indicators of the pigs. IV. Based on the body weight, environmental indicators, health indicators obtained from I, II, and III, and existing nutritional models, accurately calculate the daily nutritional requirements of pigs; V. Based on the actual production situation, obtain the type of feed currently used by the user; VI. Calculate the dosage and ratio of various feeds based on the obtained nutritional requirements, the type of feed currently used by the user, and the existing feed formulation model; VII. Automatically create execution strategies for daily material feeding based on user-defined feeding strategies; VIII. Monitor the operation of the mixing equipment in real time according to the dynamically generated feeding execution strategy.
2. The intelligent precision feeding method based on a pig farm unit according to claim 1, characterized in that, The specific steps for accurately collecting the weight of pigs using the aisle scale set up in step I are as follows: S1.1: Within the pigsty unit, each aisle is designated as a feeding area and a drinking area. The two areas are connected by a one-way door for pigs to pass through in one direction. That is, pigs can only go from the feeding area to the drinking area to drink water after finishing feeding, and vice versa, they can only enter the feeding area from the drinking area to feed through the one-way door at the symmetrical position. S1.2: While the weight of the pigs is obtained through the aisle scale, the pig's ear tag is read by the built-in RFID identification device of the aisle scale to identify the pig's identity, thereby enabling targeted tracking and monitoring of the feeding situation of individual pigs; S1.3: The cloud center cleans, filters, and clusters the pig weight information synchronously uploaded by the aisle scale to obtain the accurate weight and number of pigs in each unit of the pig house. S1.4: The average weight of pigs is obtained by taking the K-Means average weight of the pigs and then verifying the reasonableness of this average weight value with yesterday's average weight. Unreasonable values are removed, and the next batch of aisle weighing data is waited for to be processed again. If the average weight of pigs cannot be calculated after processing all batches of aisle weighing data on the same day, that is, the weight obtained from each batch is invalid, then traceability processing is performed. S1.5: The cloud-based system processes weighing data from the aisle scales to obtain pig weights and simultaneously identifies individual pigs to determine the number of pigs that have consumed feed that day. This data is then compared with known production parameters in the system. When the difference exceeds a threshold, the system alerts management personnel to intervene manually and calculates nutritional requirements based on the number identified the previous day. This avoids calculation errors caused by pig ear tags falling off, being damaged, or becoming invalid.
3. The intelligent precision feeding method based on a pig farm unit according to claim 2, characterized in that, The specific steps of the traceability processing described in step S1.4 are as follows: S2.1: During the traceability process, based on the time series set of the weight values of each pig on the day, compare them in reverse order with the weight of the pig yesterday until a reasonable value is found. After the traceability process of all pigs is completed, perform K-Means mean processing on all values again to obtain the average weight of pigs in the unit on the day. S2.2: If the traceability process still fails and the average weight of the unit on the day cannot be obtained, the system will send an alarm message to prompt the management personnel to intervene manually. On the other hand, it will fit the average weight of the day as the parameter required for the calculation of nutritional requirements based on the average value of yesterday and historical data using a univariate polynomial formula. S2.3: When tracing and processing the weighing data of the aisle scale, it is also a process of secondary integration of the weight of each pig. The above-mentioned invalid data is also specially identified. If a pig is found to have a weight decrease relative to the previous day or a weight gain lower than the preset change value, the system will automatically alarm to prompt the management personnel to intervene manually. At the same time, the pathological condition or health status of the pig will be analyzed, and the existing nutritional requirement model will be dynamically adjusted.
4. The intelligent precision feeding method based on a pig farm unit according to claim 1, characterized in that, The specific steps for assessing pig health indicators based on the remaining feed levels in each unit's feed tower over the previous 5 days and in conjunction with health anomaly events collected by the inspection robot, as described in Step III, are as follows: S3.1: The presence of leftover feed indicates a positive deviation in the calculation of nutritional requirements. Excluding cases where a large number of pigs are moved to different pens due to production needs, the presence of leftover feed is due to health problems in a group or multiple pigs. The system will automatically send a warning reminder to the management personnel and will also automatically analyze and model based on the feeding, consumption, leftover feed and existing leftover feed models over the past 5 days. S3.2: In the feed surplus analysis model, the number of abnormal pigs is estimated based on the amount of feed surplus and the daily nutritional requirements of a single pig; and based on the feed surplus situation over 5 days, including the quantity, numerical change curve, time period of feed surplus occurrence, duration, and frequency of occurrence, it is determined whether it is a valid abnormality and the possible abnormality type is assessed. S3.3: Compare the leftover feed situation of other pig house units using the same or highly similar types of feed to assess whether there are feed source problems, and automatically alarm to prompt the existence of feed sources; S3.4: Based on the above assessment, obtain information on the number of pigs with health problems and the types of pathological conditions, and dynamically adjust the nutritional requirements and nutritional ratios.
5. The intelligent precision feeding method based on a pig farm unit according to claim 1, characterized in that, Step VI involves calculating the dosage and ratio of various feeds based on the obtained nutritional requirements, the type of feed currently used by the user, and the existing feed formulation model. The specific details are as follows: S4.1: Based on the typical bulk raw material types in the feed tower corresponding to the feeding unit, the corresponding energy, protein, amino acids, macro and micro elements, and vitamin values are adapted from the existing nutritional model, and the previously calculated nutritional requirements are used as the ratio target to obtain the optimal ratio combination. S4.2: Based on the different concentrations of complete feed types in the feed tower, and using the calculated nutritional requirements as the target ratio, obtain the optimal ratio of complete feed at each concentration while meeting the nutritional needs for energy, protein, amino acids, macro and micro elements, and vitamins.
6. The intelligent precision feeding method based on a pig farm unit according to claim 1, characterized in that, Step VII, which describes the automatic creation of daily feeding execution strategy requirements based on user-defined feeding strategies, involves the system automatically creating equipment operation strategies based on the user-defined actual feeding frequency and time requirements. This is done by considering factors such as the feed quantity required by each target pig house unit, the corresponding equipment feeding capacity, mixing efficiency, conveying capacity, equipment operation and maintenance requirements, and the distance between the feeding equipment and each pig house unit. The goal is to generate the optimal execution configuration for equipment operation and the daily feeding needs of the pig house units. This includes automatically optimizing the feeding sequence and batch feeding quantity configurations for each feeding pig house.
7. The intelligent precision feeding method based on a pig farm unit according to claim 1, characterized in that, The specific steps for real-time monitoring of the mixing equipment operation based on the dynamically generated feeding execution strategy, as described in step VIII, are as follows: S5.1: The feeding motor can be divided into three stages for a single batch feeding process: motor start-up stage, stable operation stage, and remaining material dropping after shutdown. The remaining time for the motor to stop running in this batch is dynamically calculated in real time using a calculation formula composed of the amount of material to be fed, the current weighing value, the remaining material dropping after shutdown, and the feeding amount per unit time. S5.2: When the feeding motor starts feeding each batch, it adopts a two-start method. The first start is used to obtain the remaining material dropping amount after the shutdown in the calculation formula. The second start is completed by the immediate weighing after the shutdown. When the motor is detected to be running smoothly, the motor stops running and the remaining material dropping amount after the shutdown is obtained based on the weighing value measured at the time of shutdown and the lag measurement. S5.3: Real-time calculation and update of current weighing value and feed rate per unit time. At the same time, based on three parameters, namely the current feed line feeding capacity, the amount of feed to be fed in this batch, and the feeding accuracy that the system needs to achieve, the sampling period for real-time weighing is set. S5.4: The feeding, mixing and conveying operations are performed asynchronously and in parallel. When the feeding is completed and the material is discharged to the mixing unit, the discharge valve can be closed without waiting for the mixing to be completed, and feeding can continue. The conveying unit only needs its built-in level sensor to detect the presence of feed to perform the conveying action.
8. An intelligent precision feeding system based on a pig farm unit, used to achieve the content of any one of claims 1-7, characterized in that, It includes a pig weighing data synchronization module for the aisle, an environmental control system interface module, a cloud-based residual feed monitoring module, a data persistence module, and a motor drive algorithm module; The pig weighing data synchronization module of the aisle scale uses the built-in RFID identification device of the aisle scale to identify the pig's ID and the weighing sensor to obtain the weighing value, and synchronizes it to the cloud center through the Internet of Things gateway. The environmental control system interface module is used to synchronize environmental indicators with the environmental control system at regular intervals, including temperature, humidity, carbon dioxide, ammonia, hydrogen sulfide, PM2.5, and PM10. The cloud-based residual feed monitoring module is used to store and analyze the residual feed values of each feeding synchronized by the weighing sensor of the pigsty feed tower, and to provide feedback to improve the nutrient requirement model based on the residual feed analysis model. The data persistence module is used to store the established nutrition model, feed formulation model and user-defined feeding strategy for a long time, and to provide model and strategy support when calculating various feed indicators; The motor drive algorithm module is used to accurately and efficiently implement material feeding and control the motor running time based on real-time calculations of the feeding situation.
9. An intelligent precision feeding device based on a pig farm unit, used to achieve the content of any one of claims 1-8, characterized in that, include: Feeding unit equipment, drug feeding unit equipment, mixing unit equipment, conveying unit equipment; The feeding unit equipment includes one or more feed towers for feeding various types and levels of feed components required for feeding, as well as a motor, a matching auger, and a weighing system for controlling the feeding. The drug feeding unit includes a drug storage bin, a weighing bin, and a discharge valve, which are used to accurately dispense drugs or micronutrients into the mixing bin according to the drug-material ratio. The mixing unit equipment includes a high-power mixing motor, a reducer bearing, a paddle agitator, and a discharge valve, which are used to complete dynamic and efficient mixing. The feeding unit equipment is used to accurately deliver the correctly mixed feed to the corresponding pigsty unit, and includes a feeding motor with a feed disc feed line, control valves for feeding each sub-feed tower, and a weighing system.
10. A smart precision feeding monitoring and operational terminal based on a pig farm unit, used to implement any one of claims 1-8, characterized in that, The monitoring and business terminals specifically include desktop terminals, mobile terminals, and a central display screen. The desktop terminal is used by daily supervisors to monitor the current operation of the pig farm in real time, including feed delivery, pig growth, and faults and alarms that occur during operation. The mobile terminal is used by on-site technicians to make real-time adjustments to operating equipment and various strategies or to handle fault repairs based on the actual needs of the site, using handheld devices or mobile phones. The central display screen is used to macroscopically display the comprehensive operating data of the entire pig farm or group user for the year, including various statistical reports, views, and trend analyses.
