Intelligent building pig farm feed control management method and system based on internet of things

Abstract

The present application relates to the technical field of animal husbandry, and discloses an intelligent building pig farm feed control management method and system based on the Internet of Things, which comprises the following steps: extracting pig group behavior trajectory data and pig group body posture data from pig group activity behavior data, respectively analyzing the pig group activity and abnormal behavior of the corresponding pig group in the building pig farm; analyzing the health status of the pig group in the building pig farm, the environment-foraging correlation effect of the building pig farm, and the feed demand trend of the building pig farm; recording the remaining amount data of the building pig farm, calculating the corresponding feed consumption rate of the building pig farm based on the remaining amount data, setting the hierarchical feeding rules of the building pig farm; calculating the pig group body fat dispersion of the building pig farm, determining the feed demand quota of the building pig farm, and using the Internet of Things control equipment to perform feed control management on the building pig farm to obtain the management result. The present application can improve the accuracy of intelligent building pig farm feed control management.

Description

Technical Field

[0001] This invention relates to a smart multi-story pig farm feed control and management method and system based on the Internet of Things, belonging to the field of animal husbandry technology. Background Technology

[0002] The Internet of Things (IoT) smart multi-story pig farm refers to a specific method that uses IoT technology to intelligently control and manage the feed of multi-story pig farms. A multi-story pig farm is a breeding model in which pig houses are built into multi-story buildings. This model can effectively save land resources and increase breeding density. In order to improve the breeding efficiency of pigs in multi-story pig farms, it is necessary to control and manage the feed.

[0003] Current intelligent multi-story pig farm feed control and management adopts a fixed feeding method based on preset growth stages. This method pre-sets fixed daily feeding amounts, feeding times, and feed formulas according to different growth stages of the pig herd (such as nursery, growing, and fattening). It uses sensors to monitor the remaining amount in the feed tower to achieve "threshold-triggered" feeding. Environmental parameters (such as temperature) are only used as a reference for manually adjusting the formula. However, this method relies on standardized processes rather than real-time dynamic data-driven approaches. It does not take into account the relevant dynamic activities of the pig herd or the growth and physical characteristics of the pig herd, which leads to low accuracy in intelligent multi-story pig farm feed control and management. Summary of the Invention

[0004] This invention provides a method and system for controlling and managing feed in smart multi-story pig farms based on the Internet of Things (IoT), the main purpose of which is to improve the accuracy of feed control and management in smart multi-story pig farms.

[0005] To achieve the above objectives, the present invention provides an intelligent multi-story pig farm feed control and management method based on the Internet of Things, comprising:

[0006] Collect pig activity behavior data and environmental sensor data from multi-story pig farms, extract pig behavior trajectory data and pig state data from the pig activity behavior data, and combine the pig behavior trajectory data and pig state data to analyze the pig activity level and abnormal pig behavior corresponding to the multi-story pig farms.

[0007] By combining the activity level and abnormal behavior of the pig herd, the health status of the pigs in the multi-story pig farm is analyzed. Based on the environmental sensor data, the environment-feeding association effect of the multi-story pig farm is analyzed. By combining the health status of the pig herd and the environment-feeding association effect, the feed demand trend of the multi-story pig farm is analyzed.

[0008] Record the feed tower remaining data of the multi-story pig farm, calculate the feed consumption rate corresponding to the multi-story pig farm based on the feed tower remaining data, and set the stratified feeding rules of the multi-story pig farm in combination with the feed demand trend and the feed consumption rate.

[0009] The biological information of the pig herd in the multi-story pig farm is queried, the fat dispersion of the pig herd in the multi-story pig farm is calculated, and the feed requirement quota of the multi-story pig farm is determined by combining the biological information of the pig herd and the fat dispersion of the pig herd. The corresponding IoT control device of the multi-story pig farm is obtained, and the feed control management of the multi-story pig farm is performed by using the IoT control device in combination with the stratified feeding rules and the feed requirement quota to obtain the management results.

[0010] Optionally, the step of combining the pig herd behavior trajectory data and the pig herd morphology data to analyze the pig herd activity and abnormal pig behavior corresponding to the multi-story pig farm includes:

[0011] Extract behavioral trajectory association information from the pig herd behavioral trajectory data, and identify the behavioral trajectory identifier corresponding to the behavioral trajectory association information;

[0012] Based on the behavioral trajectory identifiers, the pig herd behavior patterns in the pig herd behavioral trajectory data are analyzed.

[0013] Extract behavioral trajectory features from the behavioral trajectory association information, and calculate the behavioral activity entropy corresponding to the pig herd behavior pattern based on the behavioral trajectory features;

[0014] Based on the behavioral activity entropy, the activity level of the pig herd corresponding to the multi-story pig farm is analyzed;

[0015] Extract the morphological features of the pig population from the morphological data, and calculate the abnormal body posture index corresponding to the morphological features of the pig population.

[0016] Based on the abnormal body posture index, the abnormal behavior of the pig herd corresponding to the multi-story pig farm is analyzed.

[0017] Optionally, calculating the behavioral activity entropy corresponding to the pig herd behavior pattern based on the behavioral trajectory features includes:

[0018] Calculate the feature Gini value corresponding to the behavioral trajectory feature, and based on the feature Gini value, filter out the key trajectory features in the behavioral trajectory feature;

[0019] The key trajectory features are subjected to feature quantization processing to obtain trajectory feature values;

[0020] The trajectory feature values ​​are normalized to obtain normalized feature values;

[0021] Based on the normalized eigenvalues, calculate the probability matrix of the behavioral states corresponding to the pig herd behavior patterns;

[0022] Based on the behavioral state probability matrix, the behavioral activity entropy corresponding to the pig herd behavior pattern is calculated.

[0023] Optionally, the step of analyzing the health status of the pig herd in the multi-story pig farm by combining the pig herd activity level and the abnormal pig herd behavior includes:

[0024] Based on the pig herd activity level, a baseline health curve for the pig herd in the multi-story pig farm is plotted.

[0025] The activity deviation of the multi-story pig farm is calculated by combining the preset historical baseline curve and the pig herd health baseline curve.

[0026] Based on the activity deviation, the abnormal activity range of the multi-story pig farm is identified;

[0027] Analyze the abnormal correlation between the abnormally active regions and the abnormal behaviors of the pig herd;

[0028] By combining the abnormal correlations and the abnormal activity ranges, the health status of the pig herd in the multi-story pig farm is analyzed.

[0029] Optionally, the step of analyzing the environment-feed association effect of the multi-story pig farm based on the environmental sensor data includes:

[0030] The environmental sensing data is cleaned to obtain the target environmental data;

[0031] Extract environmental factors from the target environmental data and query the pig herd feed intake data corresponding to the environmental factors;

[0032] Based on the pig herd feeding data, the feeding patterns of the pig herd in the multi-story pig farm were analyzed.

[0033] The feeding patterns of the pig herd and the environmental factors are time-stamped to obtain the environment-feeding synchronization matrix.

[0034] Based on the environment-feeding synchronization matrix, the environment-feeding correlation effect of the multi-story pig farm was analyzed.

[0035] Optionally, calculating the feed consumption rate corresponding to the multi-story pig farm based on the feed tower's remaining data includes:

[0036] Identify the remaining material quantity and timestamp sequence in the remaining material quantity data of the material tower, and sort the remaining material quantity of the material tower based on the timestamp sequence to obtain the sequence remaining material quantity;

[0037] Query the intrinsic parameters of the feed tower corresponding to the multi-story pig farm, and calculate the feed tower correction coefficient of the multi-story pig farm based on the intrinsic parameters of the feed tower;

[0038] Calculate the time interval between the timestamp sequences, and combine the remaining feed amount in the sequence, the time interval, and the feed tower correction coefficient to calculate the feed consumption rate corresponding to the multi-story pig farm using the following formula:

[0039]

[0040] Where A represents the feed consumption rate corresponding to multi-story pig farms. This represents the amount of remaining material in the sequence at time a. This represents the remaining material quantity in the sequence at time a+1. This represents the feed tower correction factor. Indicates time interval, This represents the time correction factor.

[0041] Optionally, calculating the feed tower correction factor for the multi-story pig farm based on the intrinsic parameters of the feed tower includes:

[0042] The intrinsic parameters of the feed tower are vectorized to obtain parameter feature vectors;

[0043] Query the residual feed amount corresponding to each parameter in the intrinsic parameters of the feed tower;

[0044] Based on the amount of residual feed, calculate the feed residue rate corresponding to each parameter in the intrinsic parameters of the feed tower;

[0045] Based on the feed residue rate, assign parameter feature weights to the parameter feature vectors;

[0046] By combining the parameter feature vector and the parameter feature weight, the parameter correction factor corresponding to the intrinsic parameters of the silo is calculated;

[0047] Based on the parameter correction factor, the feed tower correction parameters for the multi-story pig farm are calculated.

[0048] Optionally, the step of calculating the parameter correction factor corresponding to the intrinsic parameters of the silo by combining the parameter feature vector and the parameter feature weight includes:

[0049]

[0050] Where F represents the parameter correction factor corresponding to the intrinsic parameters of the feed tower. This represents the feature weight of the b-th parameter in the feature vector. This represents the b-th vector in the parameter eigenvectors. This represents the ideal baseline value of the parameter corresponding to the b-th vector in the parameter feature vector, where b represents the sequence number of the parameter feature vector.

[0051] Optionally, calculating the fat dispersion of the pig population in the multi-story pig farm includes:

[0052] Collect image data of pigs in the multi-story pig farm, and perform noise reduction processing on the pig image data to obtain noise-reduced pig images;

[0053] The denoised pig herd image is subjected to image segmentation processing to obtain the main image of the pig herd;

[0054] Identify the pig outlines corresponding to the main image of the pig herd, and smooth the pig outlines to obtain smooth pig outlines.

[0055] Extract the contour feature parameters corresponding to the smoothed pig body outline, and determine the physiological age of each pig in the main image of the pig herd;

[0056] The standard profile feature parameters corresponding to the physiological age of the pigs are queried. The fat dispersion of the pig population in the multi-story pig farm is calculated by combining the profile feature parameters and the standard profile feature parameters.

[0057] To address the above problems, the present invention also provides an intelligent multi-story pig farm feed control and management system based on the Internet of Things, the system comprising:

[0058] The pig herd data analysis module is used to collect pig activity behavior data and environmental sensor data from multi-story pig farms, extract pig behavior trajectory data and pig herd status data from the pig activity behavior data, and combine the pig behavior trajectory data and pig herd status data to analyze the pig activity level and abnormal pig behavior corresponding to the multi-story pig farm.

[0059] The feed demand analysis module is used to analyze the health status of pigs in the multi-story pig farm by combining the activity level and abnormal behavior of the pigs, analyze the environment-feeding association effect of the multi-story pig farm based on the environmental sensor data, and analyze the feed demand trend of the multi-story pig farm by combining the health status of the pigs and the environment-feeding association effect.

[0060] The feeding rule setting module is used to record the feed tower remaining data of the multi-story pig farm, calculate the feed consumption rate corresponding to the multi-story pig farm based on the feed tower remaining data, and set the stratified feeding rules of the multi-story pig farm in combination with the feed demand trend and the feed consumption rate.

[0061] The feed control and management module is used to query the pig herd biological information of the multi-story pig farm, calculate the herd fat dispersion of the pig herd, determine the feed requirement quota of the multi-story pig farm by combining the pig herd biological information and the herd fat dispersion, obtain the corresponding IoT control device of the multi-story pig farm, and use the IoT control device to perform feed control and management of the multi-story pig farm by combining the stratified feeding rules and the feed requirement quota, and obtain the management results.

[0062] Compared to the problems described in the background art, this invention, by combining the pig herd behavior trajectory data and pig herd morphology data, analyzes the activity level and abnormal behavior of the pigs in the multi-story pig farm, thereby understanding the dynamic characteristics of the pig herd and providing a basis for subsequent analysis of the pig herd's health status. Furthermore, by combining the pig herd activity level and the abnormal behavior, this invention analyzes the health status of the pigs in the multi-story pig farm, revealing their physiological health level and providing a basis for subsequent analysis of feed demand trends in the multi-story pig farm. With data support, this invention calculates the feed consumption rate of the multi-story pig farm based on the feed tower's remaining data, thus understanding the feed utilization efficiency of the pig herd. This facilitates the subsequent setting of stratified feeding rules. Furthermore, by querying the biological information of the pig herd, this invention can understand the basic growth status of the pigs, such as breed, age, and weight distribution; and by calculating the fat dispersion of the pig herd, it can obtain a quantitative assessment of the uniformity of fat deposition among individual pigs, facilitating the determination of feed demand quotas for the multi-story pig farm. Therefore, the IoT-based intelligent multi-story pig farm feed control and management method and system provided in this invention can improve the accuracy of feed control and management in intelligent multi-story pig farms. Attached Figure Description

[0063] Figure 1 A flowchart illustrating an Internet of Things-based intelligent multi-story pig farm feed control and management method according to an embodiment of the present invention;

[0064] Figure 2 A schematic diagram of feed control management in the adaptive method of the intelligent lighting explosion-proof photosensitive controller provided by the present invention;

[0065] Figure 3 This is a schematic diagram of the modules for implementing the IoT-based intelligent multi-story pig farm feed control and management method according to an embodiment of the present invention.

[0066] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0067] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0068] This application provides an IoT-based intelligent multi-story pig farm feed control and management method. The executing entity of this IoT-based intelligent multi-story pig farm feed control and management method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application embodiment: a server, a terminal, etc. In other words, the IoT-based intelligent multi-story pig farm feed control and management method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0069] Example 1:

[0070] Reference Figure 1 The diagram shown is a flowchart illustrating an IoT-based intelligent multi-story pig farm feed control and management method according to an embodiment of the present invention. In this embodiment, the IoT-based intelligent multi-story pig farm feed control and management method includes:

[0071] S1. Collect pig activity behavior data and environmental sensor data from the multi-story pig farm, extract pig behavior trajectory data and pig morphology data from the pig activity behavior data, and analyze the pig activity level and abnormal pig behavior corresponding to the multi-story pig farm by combining the pig behavior trajectory data and the pig morphology data.

[0072] This invention combines pig herd behavior trajectory data and pig herd morphology data to analyze the activity level and abnormal behavior of pigs in multi-story pig farms. This allows for an understanding of the dynamic characteristics of the pig herd in these farms, providing a basis for subsequent analysis of pig health status. Specifically, the pig activity behavior data consists of daily activity information collected by sensors in the multi-story pig farm; the environmental sensor data consists of environmental parameters within the pigsty obtained using various environmental sensors; the pig herd behavior trajectory data is the sequence information of pig movement paths from the pig activity behavior data; and the pig herd morphology data... The data refers to characteristic data reflecting the physical condition of pigs in the pig herd activity behavior data; the pig herd activity level is an indicator of the frequency of daily activities of pigs in the multi-story pig farm; the abnormal pig herd behavior is an abnormal manifestation of pigs deviating from the normal behavior pattern in the multi-story pig farm. Furthermore, the data can be collected by deploying cameras, RFID readers, and sensors such as temperature and humidity in the pigsty using Internet of Things (IoT) technology; the data can also be extracted by using extraction functions, which are compiled from scripting languages.

[0073] As an embodiment of the present invention, the step of combining the pig herd behavior trajectory data and the pig herd morphology data to analyze the pig herd activity and abnormal pig herd behavior corresponding to the multi-story pig farm includes:

[0074] Extract behavioral trajectory association information from the pig herd behavioral trajectory data, and identify the behavioral trajectory identifier corresponding to the behavioral trajectory association information;

[0075] Based on the behavioral trajectory identifiers, the pig herd behavior patterns in the pig herd behavioral trajectory data are analyzed.

[0076] Extract behavioral trajectory features from the behavioral trajectory association information, and calculate the behavioral activity entropy corresponding to the pig herd behavior pattern based on the behavioral trajectory features;

[0077] Based on the behavioral activity entropy, the activity level of the pig herd corresponding to the multi-story pig farm is analyzed;

[0078] Extract the morphological features of the pig population from the morphological data, and calculate the abnormal body posture index corresponding to the morphological features of the pig population.

[0079] Based on the abnormal body posture index, the abnormal behavior of the pig herd corresponding to the multi-story pig farm is analyzed.

[0080] The behavioral trajectory association information comprises specific information such as location, speed, and time contained in the pig herd behavioral trajectory data; the behavioral trajectory identifier is a unique marker corresponding to the behavioral trajectory association information; the pig herd behavioral pattern is the regular activity characteristic presented by the pig herd behavioral trajectory data; the behavioral trajectory feature is a key parameter reflecting the movement pattern in the behavioral trajectory association information (such as movement distance, area switching frequency, etc.); the behavioral activity entropy is a quantitative index of disorder / complexity corresponding to the pig herd behavioral pattern calculated based on the behavioral trajectory features; the pig herd physiology feature is a feature reflecting the body state in the pig herd physiology data (such as body size, posture, body surface health, etc.); and the physiology abnormality index is a comprehensive quantitative value of deviation from the normal state corresponding to the pig herd physiology feature.

[0081] Furthermore, behavioral trajectory association information in the pig herd behavior trajectory data can be extracted using data parsing algorithms, such as regular expression parsing algorithms; behavioral trajectory identifiers corresponding to the behavioral trajectory association information can be identified using unique ID generation algorithms; based on the behavioral trajectory identifiers, cluster sequence pattern mining algorithms can be used to analyze the pig herd behavior patterns in the pig herd behavior trajectory data; behavioral trajectory features in the behavioral trajectory association information can be extracted using feature engineering methods; based on the behavioral activity entropy, the activity level of the pig herd corresponding to the multi-story pig farm can be analyzed by comparing the historical entropy value distribution range with a set threshold, such as using the KS test to determine whether the current entropy value exceeds the 95% confidence interval; pig herd morphology features in the pig herd morphology data can be extracted using deep learning models, such as the ResNet feature extraction network; the body morphology abnormality index corresponding to the pig herd morphology features can be calculated using multi-index weighted calculation, such as weighted summation based on coat gloss (weight 40%), skin redness and swelling area (30%), and gait angle deviation (30%); based on the body morphology abnormality index, abnormal behavior of the pig herd corresponding to the multi-story pig farm can be analyzed, such as triggering a pathological behavior warning when the index > 85.

[0082] Furthermore, as an optional embodiment of the present invention, calculating the behavioral activity entropy corresponding to the pig herd behavior pattern based on the behavioral trajectory features includes:

[0083] Calculate the feature Gini value corresponding to the behavioral trajectory feature, and based on the feature Gini value, filter out the key trajectory features in the behavioral trajectory feature;

[0084] The key trajectory features are subjected to feature quantization processing to obtain trajectory feature values;

[0085] The trajectory feature values ​​are normalized to obtain normalized feature values;

[0086] Based on the normalized eigenvalues, calculate the probability matrix of the behavioral states corresponding to the pig herd behavior patterns;

[0087] Based on the behavioral state probability matrix, the behavioral activity entropy corresponding to the pig herd behavior pattern is calculated.

[0088] Wherein, the feature Gini value is a classification ability evaluation index corresponding to the behavioral trajectory feature, the key trajectory feature is the feature with the most discriminative power for classifying behavioral patterns selected from the behavioral trajectory features based on the feature Gini value; the trajectory feature value is the quantitative numerical representation of the key trajectory feature; the normalized feature value is the value of the trajectory feature value mapped to the interval [0, 1] after normalization processing; and the behavioral state probability matrix is ​​the distribution matrix of the frequency of occurrence of each behavioral state corresponding to the pig herd behavioral pattern.

[0089] Furthermore, the feature Gini value corresponding to the behavioral trajectory feature can be calculated using the Gini impurity formula of the CART decision tree algorithm; based on the feature Gini value, the key trajectory features in the behavioral trajectory features are selected by sorting the Gini values ​​from smallest to largest and setting a threshold; the key trajectory features can be quantized using an encoding algorithm to obtain trajectory feature values; the trajectory feature values ​​can be normalized using a Min-Max normalization algorithm to obtain normalized feature values; based on the normalized feature values, the time proportion of each behavioral state is statistically analyzed and a matrix is ​​constructed to calculate the behavioral state probability matrix corresponding to the pig herd behavior pattern; based on the behavioral state probability matrix, the behavioral activity entropy corresponding to the pig herd behavior pattern is calculated using the Shannon entropy formula.

[0090] S2. Combining the pig herd activity and abnormal pig behavior, analyze the health status of the pigs in the multi-story pig farm. Based on the environmental sensor data, analyze the environment-feeding correlation effect of the multi-story pig farm. Combining the pig herd health status and the environment-feeding correlation effect, analyze the feed demand trend of the multi-story pig farm.

[0091] This invention analyzes the health status of pigs in multi-story pig farms by combining the activity level and abnormal behavior of the pig herd. This allows for an understanding of the physiological health level of the pigs in the multi-story pig farm, providing data support for subsequent analysis of feed demand trends in the multi-story pig farm. The health status of the pig herd is a comprehensive reflection of whether the physiological functions of the pigs in the multi-story pig farm are normal, whether their behavior is abnormal, and their potential health risks.

[0092] As an embodiment of the present invention, the step of analyzing the health status of the pig herd in the multi-story pig farm by combining the pig herd activity level and the abnormal pig herd behavior includes:

[0093] Based on the pig herd activity level, a baseline health curve for the pig herd in the multi-story pig farm is plotted.

[0094] The activity deviation of the multi-story pig farm is calculated by combining the preset historical baseline curve and the pig herd health baseline curve.

[0095] Based on the activity deviation, the abnormal activity range of the multi-story pig farm is identified;

[0096] Analyze the abnormal correlation between the abnormally active regions and the abnormal behaviors of the pig herd;

[0097] By combining the abnormal correlations and the abnormal activity ranges, the health status of the pig herd in the multi-story pig farm is analyzed.

[0098] The pig herd health baseline curve is a dynamic baseline (e.g., daily / weekly activity fluctuation trend) plotted based on the pig herd activity level in the multi-story pig farm. The preset historical baseline curve is a normal distribution reference curve formed by long-term activity data of healthy pig herds of the same breed and age. The activity deviation indicates the degree of deviation between the current pig herd activity level and the historical health level in the multi-story pig farm (the larger the value, the more significant the deviation). The abnormal activity interval is a time period or region in the multi-story pig farm where the activity level deviates significantly from the historical baseline, based on the activity deviation. The abnormal correlation is whether there is a causal relationship or a regular relationship of synchronous occurrence between the abnormal activity interval and the abnormal behavior of the pig herd.

[0099] Furthermore, based on the pig herd activity level, a health baseline curve for the pigs in the multi-story pig farm can be plotted using a drawing tool, such as an Excel charting tool. Combining the preset historical baseline curve and the pig herd health baseline curve, an activity deviation corresponding to the multi-story pig farm is calculated using a curve comparison algorithm (such as root mean square error). Based on the activity deviation, a threshold is set to filter significantly deviating intervals and identify abnormal activity intervals in the multi-story pig farm; for example, when the activity deviation exceeds ±2σ of the historical baseline curve, it is marked as an abnormal interval. The abnormal correlation between the abnormal activity intervals and the abnormal behavior of the pigs can be analyzed using timestamp correlation and statistical tests (such as chi-square tests). Combining the abnormal correlation and the abnormal activity intervals, a comprehensive assessment of the coupling risk between abnormal activity and abnormal behavior is conducted to analyze the health status of the pigs in the multi-story pig farm, such as determining whether the abnormal activity level has triggered disease or stress responses in the pigs.

[0100] This invention analyzes the environment-feeding correlation effect of the multi-story pig farm based on the environmental sensing data. It can understand the intrinsic relationship between environmental factors such as temperature, humidity, and air quality and the feed intake of pigs, providing a basis for subsequent analysis of feed demand trends in the multi-story pig farm. The environment-feeding correlation effect refers to the influence and interaction relationship of environmental factors (such as temperature, humidity, light, and air quality) on the feeding behavior (feed intake, feeding frequency, etc.) of pigs in the multi-story pig farm.

[0101] As an embodiment of the present invention, the analysis of the environment-feed association effect of the multi-story pig farm based on the environmental sensing data includes:

[0102] The environmental sensing data is cleaned to obtain the target environmental data;

[0103] Extract environmental factors from the target environmental data and query the pig herd feed intake data corresponding to the environmental factors;

[0104] Based on the pig herd feeding data, the feeding patterns of the pig herd in the multi-story pig farm were analyzed.

[0105] The feeding patterns of the pig herd and the environmental factors are time-stamped to obtain the environment-feeding synchronization matrix.

[0106] Based on the environment-feeding synchronization matrix, the environment-feeding correlation effect of the multi-story pig farm was analyzed.

[0107] The target environmental data refers to the effective data after cleaning and screening of environmental sensor data. The environmental factors are specific monitoring indicators in the target environmental data (such as temperature, humidity, ammonia concentration, etc.). The pig herd feeding data is the pig herd feeding behavior data (such as feed intake, feeding time, etc.) corresponding to the environmental factors at the time. The pig herd feeding pattern is the periodic characteristics or regular pattern of pig herd feeding in multi-story pig farms. The environment-feeding synchronization matrix is ​​a time-series associated structured data formed after aligning the pig herd feeding pattern and environmental factors with timestamps.

[0108] Furthermore, the environmental sensor data can be cleaned using box plots to obtain target environmental data; environmental factors can be extracted from the target environmental data using data field extraction tools (such as Python's Pandas library); the pig herd feeding data corresponding to the environmental factors can be queried using database join queries (such as SQL JOIN statements); based on the pig herd feeding data, statistical analysis methods (such as periodogram analysis and time series decomposition) can be used to analyze the feeding patterns of the pig herd in the multi-story pig farm; the pig herd feeding patterns and environmental factors can be timestamped using time series alignment algorithms (such as Dynamic Time Warping (DTW)) to obtain an environment-feeding synchronization matrix; based on the environment-feeding synchronization matrix, the environment-feeding association effect of the multi-story pig farm can be analyzed through correlation modeling and causal analysis (such as Pearson correlation coefficient and Granger causality test).

[0109] This invention analyzes the feed demand trends of multi-story pig farms by combining the health status of the pig herd with the environment-feeding correlation effect. It comprehensively considers the physiological health, behavioral characteristics, and environmental influences of the pig herd to understand feed consumption needs at different times. The feed demand trend refers to the changing patterns and predicted directions of feed consumption and nutrient requirements of the pig herd at different time periods. Furthermore, by combining the pig herd's health status with the environment-feeding correlation effect, the invention analyzes the feed demand trends of multi-story pig farms. For example, based on the pig herd's health status, its digestive capacity and nutritional needs can be assessed. If the pig herd's health is poor, the feed amount can be appropriately reduced and the nutrient ratio adjusted. Considering the environment-feeding correlation effect, in high-temperature and high-humidity environments, the pig herd's feed intake decreases, so the feed supply can be reduced and palatable components increased. Based on the pig herd's activity level and feeding patterns, sufficient and nutritionally adequate feed can be prepared in advance before the peak feeding period.

[0110] S3. Record the feed tower remaining data of the multi-story pig farm. Based on the feed tower remaining data, calculate the feed consumption rate corresponding to the multi-story pig farm. Combine the feed demand trend and the feed consumption rate to set the stratified feeding rules for the multi-story pig farm.

[0111] This invention calculates the feed consumption rate of a multi-story pig farm based on the feed tower remaining data, thereby understanding the feed utilization efficiency of the pig herd and facilitating the setting and processing of subsequent stratified feeding rules. The feed tower remaining data is the weight or volume of remaining feed in each feed tower, collected in real time by sensors or vision devices. The feed consumption rate represents the amount of feed consumed by the pig herd per unit time, reflecting the real-time intensity and dynamic changes in feed intake. Furthermore, the feed tower remaining data of the multi-story pig farm can be recorded using pressure sensors at the bottom of the feed towers.

[0112] As an embodiment of the present invention, the step of calculating the feed consumption rate corresponding to the multi-story pig farm based on the feed tower remaining data includes:

[0113] Identify the remaining material quantity and timestamp sequence in the remaining material quantity data of the material tower, and sort the remaining material quantity of the material tower based on the timestamp sequence to obtain the sequence remaining material quantity;

[0114] Query the intrinsic parameters of the feed tower corresponding to the multi-story pig farm, and calculate the feed tower correction coefficient of the multi-story pig farm based on the intrinsic parameters of the feed tower;

[0115] Calculate the time interval between the timestamp sequences, and combine the remaining feed amount in the sequence, the time interval, and the feed tower correction coefficient to calculate the feed consumption rate corresponding to the multi-story pig farm using the following formula:

[0116]

[0117] Where A represents the feed consumption rate corresponding to multi-story pig farms. This represents the amount of remaining material in the sequence at time a. This represents the remaining material quantity in the sequence at time a+1. This represents the feed tower correction factor. Indicates time interval, This represents the time correction factor.

[0118] The remaining feed volume in the feed tower and the timestamp sequence are the basic components of the feed tower remaining data. The sequenced remaining feed volume is a time-series data set after sorting the remaining feed volume in the feed tower based on the timestamp sequence. The intrinsic parameters of the feed tower are the inherent physical property parameters of the feed tower corresponding to the multi-story pig farm (such as volume, shape, and outlet size). Based on the intrinsic parameters of the feed tower, the feed tower correction coefficient is a compensation parameter for the impact of differences in the physical characteristics of the feed tower on the feed consumption calculation in the multi-story pig farm. The time interval is the difference between the timestamp sequences. The time correction coefficient is a compensation parameter reflecting the impact of different time periods (such as day and night, season, and pig herd physiological cycle) on the feed consumption rate. It is obtained by analyzing the periodic characteristics (such as daily feed intake peak and seasonal feed intake fluctuation) in historical time series data and combining Fourier transform, LSTM neural network and other algorithms for modeling.

[0119] Furthermore, the remaining feed volume and timestamp sequence in the feed tower balance data can be identified through a JSON parser; based on the timestamp sequence, the remaining feed volume in the feed tower can be sorted using a time series sorting algorithm to obtain the sequenced remaining feed volume; the intrinsic parameters of the feed tower corresponding to the multi-story pig farm can be queried through the equipment parameter database query interface.

[0120] Furthermore, as an optional embodiment of the present invention, calculating the feed tower correction coefficient for the multi-story pig farm based on the intrinsic parameters of the feed tower includes:

[0121] The intrinsic parameters of the feed tower are vectorized to obtain parameter feature vectors;

[0122] Query the residual feed amount corresponding to each parameter in the intrinsic parameters of the feed tower;

[0123] Based on the amount of residual feed, calculate the feed residue rate corresponding to each parameter in the intrinsic parameters of the feed tower;

[0124] Based on the feed residue rate, assign parameter feature weights to the parameter feature vectors;

[0125] By combining the parameter feature vector and the parameter feature weight, the parameter correction factor corresponding to the intrinsic parameters of the silo is calculated;

[0126] Based on the parameter correction factor, the feed tower correction parameters for the multi-story pig farm are calculated.

[0127] The parameter feature vector is a multi-dimensional numerical representation of the intrinsic parameters of the feed tower after vectorization, mapping the physical parameters to Euclidean space coordinates for algorithm processing; the residual feed amount is a quantitative statistical value of the remaining feed after the end of the historical feeding cycle corresponding to each parameter in the intrinsic parameters of the feed tower, reflecting the residual feed situation in actual use of the feed tower; the feed residue rate is the ratio of the residual feed amount corresponding to each parameter in the intrinsic parameters of the feed tower to the total amount fed, quantifying the impact of feed tower design on feed waste; the parameter feature weight is the importance coefficient corresponding to the parameter feature vector, and the contribution of each intrinsic parameter to feed consumption prediction is determined by the algorithm; the parameter correction factor is the single parameter correction value corresponding to the intrinsic parameters of the feed tower.

[0128] Furthermore, the intrinsic parameters of the feed tower can be vectorized using feature engineering algorithms (such as one-heat coding, standardization / normalization) to obtain parameter feature vectors; the residual feed amount corresponding to each parameter in the intrinsic parameters of the feed tower can be queried through historical feeding database query interfaces (such as SQL retrieval, API calls); based on the residual feed amount, the feed residue rate corresponding to each parameter in the intrinsic parameters of the feed tower is calculated using the ratio calculation formula (residual amount / feed amount × 100%); based on the feed residue rate, the parameter feature weights corresponding to the parameter feature vectors are assigned using a weight allocation algorithm (such as AHP);

[0129] The average value of the parameter correction factor is calculated to obtain the feed tower correction parameter for the multi-story pig farm.

[0130] Furthermore, as an optional embodiment of the present invention, the step of calculating the parameter correction factor corresponding to the intrinsic parameters of the silo by combining the parameter feature vector and the parameter feature weight includes:

[0131]

[0132] Where F represents the parameter correction factor corresponding to the intrinsic parameters of the feed tower. This represents the feature weight of the b-th parameter in the feature vector. This represents the b-th vector in the parameter eigenvectors. This represents the ideal baseline value of the parameter corresponding to the b-th vector in the parameter feature vector, where b represents the sequence number of the parameter feature vector.

[0133] The ideal benchmark value of the parameters refers to the optimal combination of feed tower parameters determined through scientific verification or long-term practical optimization for specific breeding scenarios (such as the fattening period and gestation period of multi-story pig farms). It can be obtained through experimental verification. Under the same breeding environment, different parameter combinations are tested to determine indicators such as feed residue rate and feeding efficiency, and the optimal value is selected.

[0134] This invention establishes tiered feeding rules for multi-story pig farms by combining feed demand trends and feed consumption rates. This facilitates dynamic and precise matching of feed supply to pigs on different floors, reducing feed waste and improving breeding efficiency. The tiered feeding rules are a differentiated feed supply strategy for different floors of the multi-story pig farm. Targeted feeding amounts, frequencies, and feed formulations are developed based on the growth stage, weight distribution, or health status of pigs on different floors. Furthermore, the tiered feeding rules of the multi-story pig farm are established by combining feed demand trends and feed consumption rates. The feeding rules are set up as follows: combining the feed demand trend (such as the average daily feed intake curve of pigs at different growth stages) and the feed consumption rate (such as the real-time feed intake intensity of each floor), first set the basic feeding ratio of each floor according to the demand trend, and then dynamically adjust the real-time feeding frequency based on the consumption rate (for example, if the consumption rate of a certain floor is higher than expected by 15%, a supplementary feeding mechanism is triggered; if it is lower than the threshold for 2 consecutive hours, the amount of feed given in the next meal is automatically reduced), and finally form a dynamic stratified feeding strategy that covers the time dimension (such as circadian rhythm) and the spatial dimension (such as the density of pigs on each floor).

[0135] S4. Query the pig herd biological information of the multi-story pig farm, calculate the herd fat dispersion of the multi-story pig farm, combine the pig herd biological information and the herd fat dispersion to determine the feed requirement quota of the multi-story pig farm, obtain the IoT control device corresponding to the multi-story pig farm, combine the stratified feeding rules and the feed requirement quota, and use the IoT control device to perform feed control management of the multi-story pig farm to obtain management results.

[0136] This invention allows for the understanding of basic growth status of pigs, such as breed, age, and weight distribution, by querying the biological information of the pig herd in the multi-story pig farm. It also allows for the calculation of the herd's fat dispersion to obtain a quantitative assessment of the uniformity of fat deposition among individual pigs, facilitating the determination of feed requirement quotas for the multi-story pig farm. The pig herd's biological information is a set of basic attribute data for the pig herd in the multi-story pig farm (including information that can quantify the growth characteristics of the pig herd, such as breed, age, weight, and health status). The herd's fat dispersion is a quantitative indicator of the degree of difference in fat deposition among individual pigs in the multi-story pig farm (reflecting the balance of feed nutrition and the effectiveness of breeding management through statistical analysis of the standard deviation or coefficient of variation of body fat percentage within the herd). Furthermore, the herd's biological information can be queried through the pig farm management information system database.

[0137] As an embodiment of the present invention, the calculation of the fat dispersion of the pig population in the multi-story pig farm includes:

[0138] Collect image data of pigs in the multi-story pig farm, and perform noise reduction processing on the pig image data to obtain noise-reduced pig images;

[0139] The denoised pig herd image is subjected to image segmentation processing to obtain the main image of the pig herd;

[0140] Identify the pig outlines corresponding to the main image of the pig herd, and smooth the pig outlines to obtain smooth pig outlines.

[0141] Extract the contour feature parameters corresponding to the smoothed pig body outline, and determine the physiological age of each pig in the main image of the pig herd;

[0142] The standard profile feature parameters corresponding to the physiological age of the pigs are queried. The fat dispersion of the pig population in the multi-story pig farm is calculated by combining the profile feature parameters and the standard profile feature parameters.

[0143] Wherein, the pig herd image data is real-time or historical image data of the pig herd in the multi-story pig farm; the denoised pig herd image is an image of the pig herd image data processed by a denoising algorithm (such as median filtering, Gaussian filtering) to reduce noise interference; the main pig herd image is an image containing only the pig body region extracted from the denoised pig herd image by an image segmentation algorithm (such as semantic segmentation, threshold segmentation); the pig body contour is the boundary curve of the pig body shape extracted by an edge detection algorithm (such as Canny, Sobel) corresponding to the main pig herd image; the smoothed pig body contour is... The pig's body contour is processed by a smoothing algorithm (such as Gaussian smoothing or B-spline interpolation) to eliminate jagged edges and form a continuous curve. The contour feature parameters are the quantified geometric attributes (perimeter, area, curvature, etc.) and derived indices (body index, etc.) corresponding to the smoothed pig's body contour. The pig's physiological age is the age attribute (e.g., 28-day-old weaned piglets) corresponding to each pig in the main image of the pig herd, based on the growth stage (e.g., age in days, physiological maturity). The standard contour feature parameters are the average contour feature reference values ​​of healthy pigs at the same physiological stage corresponding to the pig's physiological age.

[0144] Furthermore, image data of the pigs in the multi-story pig farm can be collected using devices such as cameras and drones; contour feature parameters corresponding to the smooth body contours of the pigs can be extracted using computer vision algorithms (such as contour geometric feature calculation functions and curvature analysis algorithms); the physiological age of the pigs can be determined by the ear tag corresponding to each pig in the main image of the pig herd, and the growth data of the pigs can be queried by the ear tag to determine the physiological age of the pigs; the standard contour feature parameters corresponding to the physiological age of the pigs can be queried by the farm management system database, the contour standard deviation between the contour feature parameters and the standard contour feature parameters can be calculated, and the fat dispersion of the pig herd in the multi-story pig farm can be obtained based on the contour standard deviation.

[0145] This invention determines the feed requirement quota for multi-story pig farms by combining the pig herd's biological information and the pig herd's lipid dispersion, thereby achieving precise matching of feed supply with the pig herd's growth stage and individual differences, reducing feed waste rate and improving herd growth uniformity. The feed requirement quota is a refined feed supply standard for the multi-story pig farm, segmented by stage and group. Furthermore, by combining the pig herd's biological information and the pig herd's lipid dispersion, the feed requirement quota for the multi-story pig farm is determined. First, the pig herd's growth stages are divided and a basic feed amount is set based on the pig herd's biological information. Then, based on the pig herd's lipid dispersion, a pig herd growth curve is plotted. Based on the pig herd growth curve, the nutritional ratio of individuals or the entire herd is adjusted (e.g., increasing protein supplementation for groups with high dispersion), ultimately determining the feed requirement quota for the multi-story pig farm.

[0146] This invention improves the accuracy of feed control management in multi-story pig farms by combining the tiered feeding rules and feed requirement quotas with an IoT control device. The IoT control device is a smart terminal system corresponding to the multi-story pig farm, integrating sensors, actuators, and communication modules to achieve precise control over feed delivery amount, delivery time, and feeding area, and providing real-time data feedback to support dynamic management decisions. Furthermore, by combining the tiered feeding rules and feed requirement quotas with the IoT control device to perform feed control management, the tiered feeding rules and feed requirement quotas are programmed and input into the IoT control device to execute feed control management, ultimately yielding the management results. For a more intuitive understanding of the lighting processing flow of the IoT-based intelligent multi-story pig farm feed control management method in this application, please refer to [reference needed]. Figure 2 The diagram shown is a schematic representation of feed control management in the IoT-based intelligent multi-story pig farm feed control management method provided by this invention. It should be noted that in this invention... Figure 2 The feed control and management diagram presented is only for the feed control and management process of the IoT-based smart multi-story pig farm feed control and management method, and is not limited to the feed control and management of the IoT-based smart multi-story pig farm feed control and management method in different actual application scenarios.

[0147] Compared to the problems described in the background art, this invention, by combining the pig herd behavior trajectory data and pig herd morphology data, analyzes the activity level and abnormal behavior of the pigs in the multi-story pig farm, thereby understanding the dynamic characteristics of the pig herd and providing a basis for subsequent analysis of the pig herd's health status. Furthermore, by combining the pig herd activity level and the abnormal behavior, this invention analyzes the health status of the pigs in the multi-story pig farm, revealing their physiological health level and providing a basis for subsequent analysis of feed demand trends in the multi-story pig farm. With data support, this invention calculates the feed consumption rate of the multi-story pig farm based on the feed tower's remaining data, thus understanding the feed utilization efficiency of the pig herd. This facilitates the subsequent setting of stratified feeding rules. Furthermore, by querying the biological information of the pig herd, this invention can understand the basic growth status of the pigs, such as breed, age, and weight distribution; and by calculating the fat dispersion of the pig herd, it can obtain a quantitative assessment of the uniformity of fat deposition among individual pigs, facilitating the determination of feed demand quotas for the multi-story pig farm. Therefore, the IoT-based intelligent multi-story pig farm feed control and management method and system provided in this invention can improve the accuracy of feed control and management in intelligent multi-story pig farms.

[0148] Example 2:

[0149] like Figure 3 The diagram shown is a functional module diagram of an intelligent multi-story pig farm feed control and management system based on the Internet of Things according to the present invention.

[0150] The IoT-based intelligent multi-story pig farm feed control and management system 300 described in this invention can be installed in an electronic device. Depending on the functions implemented, the IoT-based intelligent multi-story pig farm feed control and management system may include a pig herd data analysis module 301, a feed demand analysis module 302, a feeding rule setting module 303, and a feed control and management module 304. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0151] In this embodiment of the invention, the functions of each module / unit are as follows:

[0152] The pig herd data analysis module 301 is used to collect pig herd activity behavior data and environmental sensor data in multi-story pig farms, extract pig herd behavior trajectory data and pig herd status data from the pig herd activity behavior data, and combine the pig herd behavior trajectory data and pig herd status data to analyze the pig herd activity level and abnormal pig herd behavior corresponding to the multi-story pig farm.

[0153] The feed demand analysis module 302 is used to analyze the health status of pigs in the multi-story pig farm by combining the pig activity level and the abnormal behavior of the pigs, analyze the environment-feeding association effect of the multi-story pig farm based on the environmental sensor data, and analyze the feed demand trend of the multi-story pig farm by combining the pig health status and the environment-feeding association effect.

[0154] The feeding rule setting module 303 is used to record the feed tower remaining data of the multi-story pig farm, calculate the feed consumption rate corresponding to the multi-story pig farm based on the feed tower remaining data, and set the stratified feeding rules of the multi-story pig farm in combination with the feed demand trend and the feed consumption rate.

[0155] The feed control and management module 304 is used to query the pig herd biological information of the multi-story pig farm, calculate the herd fat dispersion of the multi-story pig farm, determine the feed demand quota of the multi-story pig farm by combining the pig herd biological information and the herd fat dispersion, obtain the IoT control device corresponding to the multi-story pig farm, and use the IoT control device to perform feed control and management of the multi-story pig farm by combining the stratified feeding rules and the feed demand quota, and obtain the management result.

[0156] In detail, the modules in the IoT-based intelligent multi-story pig farm feed control and management system 300 described in this embodiment of the invention employ the same methods as described above. Figure 1 The method used is the same as the IoT-based smart multi-story pig farm feed control and management method described above, and can produce the same technical effect, so it will not be repeated here.

[0157] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0158] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for controlling and managing feed in an intelligent multi-story pig farm based on the Internet of Things, characterized in that, The method includes: Collect pig activity behavior data and environmental sensor data from multi-story pig farms, extract pig behavior trajectory data and pig state data from the pig activity behavior data, and combine the pig behavior trajectory data and pig state data to analyze the pig activity level and abnormal pig behavior corresponding to the multi-story pig farms. By combining the activity level and abnormal behavior of the pig herd, the health status of the pigs in the multi-story pig farm is analyzed. Based on the environmental sensor data, the environment-feeding association effect of the multi-story pig farm is analyzed. By combining the health status of the pig herd and the environment-feeding association effect, the feed demand trend of the multi-story pig farm is analyzed. The analysis of the environment-feed association effect in the multi-story pig farm based on the environmental sensor data includes: The environmental sensing data is cleaned to obtain the target environmental data; Extract environmental factors from the target environmental data and query the pig herd feed intake data corresponding to the environmental factors; Based on the pig herd feeding data, the feeding patterns of the pig herd in the multi-story pig farm were analyzed. The feeding patterns of the pig herd and the environmental factors are time-stamped to obtain the environment-feeding synchronization matrix. Based on the environment-feeding synchronization matrix, the environment-feeding correlation effect of the multi-story pig farm was analyzed; Record the feed tower remaining data of the multi-story pig farm, calculate the feed consumption rate corresponding to the multi-story pig farm based on the feed tower remaining data, and set the stratified feeding rules of the multi-story pig farm in combination with the feed demand trend and the feed consumption rate. The step of calculating the feed consumption rate corresponding to the multi-story pig farm based on the feed tower's remaining data includes: Identify the remaining material quantity and timestamp sequence in the remaining material quantity data of the material tower, and sort the remaining material quantity of the material tower based on the timestamp sequence to obtain the sequence remaining material quantity; Query the intrinsic parameters of the feed tower corresponding to the multi-story pig farm, and calculate the feed tower correction coefficient of the multi-story pig farm based on the intrinsic parameters of the feed tower; Calculate the time interval between the timestamp sequences, and combine the remaining feed amount in the sequence, the time interval, and the feed tower correction coefficient to calculate the feed consumption rate corresponding to the multi-story pig farm using the following formula: Where A represents the feed consumption rate corresponding to multi-story pig farms. This represents the amount of remaining material in the sequence at time a. This represents the remaining material quantity in the sequence at time a+1. This represents the feed tower correction factor. Indicates time interval, Indicates the time correction factor; The biological information of the pig herd in the multi-story pig farm is queried, the fat dispersion of the pig herd in the multi-story pig farm is calculated, and the feed requirement quota of the multi-story pig farm is determined by combining the biological information of the pig herd and the fat dispersion of the pig herd. The corresponding IoT control device of the multi-story pig farm is obtained, and the feed control management of the multi-story pig farm is performed by using the IoT control device in combination with the stratified feeding rules and the feed requirement quota to obtain the management results.

2. The IoT-based intelligent multi-story pig farm feed control and management method as described in claim 1, characterized in that, The analysis of pig activity and abnormal pig behavior in the multi-story pig farm, combining the pig herd behavior trajectory data and the pig herd morphology data, includes: Extract behavioral trajectory association information from the pig herd behavioral trajectory data, and identify the behavioral trajectory identifier corresponding to the behavioral trajectory association information; Based on the behavioral trajectory identifiers, the pig herd behavior patterns in the pig herd behavioral trajectory data are analyzed. Extract behavioral trajectory features from the behavioral trajectory association information, and calculate the behavioral activity entropy corresponding to the pig herd behavior pattern based on the behavioral trajectory features; Based on the behavioral activity entropy, the activity level of the pig herd corresponding to the multi-story pig farm is analyzed; Extract the morphological features of the pig population from the morphological data, and calculate the abnormal body posture index corresponding to the morphological features of the pig population. Based on the abnormal body posture index, the abnormal behavior of the pig herd corresponding to the multi-story pig farm is analyzed.

3. The IoT-based intelligent multi-story pig farm feed control and management method as described in claim 2, characterized in that, The step of calculating the behavioral activity entropy corresponding to the pig herd behavior pattern based on the behavioral trajectory features includes: Calculate the feature Gini value corresponding to the behavioral trajectory feature, and based on the feature Gini value, filter out the key trajectory features in the behavioral trajectory feature; The key trajectory features are subjected to feature quantization processing to obtain trajectory feature values; The trajectory feature values ​​are normalized to obtain normalized feature values; Based on the normalized eigenvalues, calculate the probability matrix of the behavioral states corresponding to the pig herd behavior patterns; Based on the behavioral state probability matrix, the behavioral activity entropy corresponding to the pig herd behavior pattern is calculated.

4. The IoT-based intelligent multi-story pig farm feed control and management method as described in claim 1, characterized in that, The analysis of the health status of the pigs in the multi-story pig farm, combining the pig herd activity level and the abnormal pig behavior, includes: Based on the pig herd activity level, a baseline health curve for the pig herd in the multi-story pig farm is plotted. The activity deviation of the multi-story pig farm is calculated by combining the preset historical baseline curve and the pig herd health baseline curve. Based on the activity deviation, the abnormal activity range of the multi-story pig farm is identified; Analyze the abnormal correlation between the abnormally active regions and the abnormal behaviors of the pig herd; By combining the abnormal correlations and the abnormal activity ranges, the health status of the pig herd in the multi-story pig farm is analyzed.

5. The IoT-based intelligent multi-story pig farm feed control and management method as described in claim 1, characterized in that, The calculation of the feed tower correction factor for the multi-story pig farm based on the intrinsic parameters of the feed tower includes: The intrinsic parameters of the feed tower are vectorized to obtain parameter feature vectors; Query the residual feed amount corresponding to each parameter in the intrinsic parameters of the feed tower; Based on the amount of residual feed, calculate the feed residue rate corresponding to each parameter in the intrinsic parameters of the feed tower; Based on the feed residue rate, assign parameter feature weights to the parameter feature vectors; By combining the parameter feature vector and the parameter feature weight, the parameter correction factor corresponding to the intrinsic parameters of the silo is calculated; Based on the parameter correction factor, the feed tower correction parameters for the multi-story pig farm are calculated.

6. The IoT-based intelligent multi-story pig farm feed control and management method as described in claim 5, characterized in that, The step of calculating the parameter correction factor corresponding to the intrinsic parameters of the silo by combining the parameter feature vector and the parameter feature weight includes: Where F represents the parameter correction factor corresponding to the intrinsic parameters of the feed tower. This represents the feature weight of the b-th parameter in the feature vector. This represents the b-th vector in the parameter eigenvectors. This represents the ideal baseline value of the parameter corresponding to the b-th vector in the parameter feature vector, where b represents the sequence number of the parameter feature vector.

7. The IoT-based intelligent multi-story pig farm feed control and management method as described in claim 1, characterized in that, The calculation of the fat dispersion of the pig population in the multi-story pig farm includes: Collect image data of pigs in the multi-story pig farm, and perform noise reduction processing on the pig image data to obtain noise-reduced pig images; The denoised pig herd image is subjected to image segmentation processing to obtain the main image of the pig herd; Identify the pig outlines corresponding to the main image of the pig herd, and smooth the pig outlines to obtain smooth pig outlines. Extract the contour feature parameters corresponding to the smoothed pig body outline, and determine the physiological age of each pig in the main image of the pig herd; The standard profile feature parameters corresponding to the physiological age of the pigs are queried. The fat dispersion of the pig population in the multi-story pig farm is calculated by combining the profile feature parameters and the standard profile feature parameters.

8. A smart multi-story pig farm feed control and management system based on the Internet of Things, characterized in that, The system includes: The pig herd data analysis module is used to collect pig activity behavior data and environmental sensor data from multi-story pig farms, extract pig behavior trajectory data and pig herd status data from the pig activity behavior data, and combine the pig behavior trajectory data and pig herd status data to analyze the pig activity level and abnormal pig behavior corresponding to the multi-story pig farm. The feed demand analysis module is used to analyze the health status of pigs in the multi-story pig farm by combining the activity level and abnormal behavior of the pigs, analyze the environment-feeding association effect of the multi-story pig farm based on the environmental sensor data, and analyze the feed demand trend of the multi-story pig farm by combining the health status of the pigs and the environment-feeding association effect. The analysis of the environment-feed association effect in the multi-story pig farm based on the environmental sensor data includes: The environmental sensing data is cleaned to obtain the target environmental data; Extract environmental factors from the target environmental data and query the pig herd feed intake data corresponding to the environmental factors; Based on the pig herd feeding data, the feeding patterns of the pig herd in the multi-story pig farm were analyzed. The feeding patterns of the pig herd and the environmental factors are time-stamped to obtain the environment-feeding synchronization matrix. Based on the environment-feeding synchronization matrix, the environment-feeding correlation effect of the multi-story pig farm was analyzed; The feeding rule setting module is used to record the feed tower remaining data of the multi-story pig farm, calculate the feed consumption rate corresponding to the multi-story pig farm based on the feed tower remaining data, and set the stratified feeding rules of the multi-story pig farm in combination with the feed demand trend and the feed consumption rate. The step of calculating the feed consumption rate corresponding to the multi-story pig farm based on the feed tower's remaining data includes: Identify the remaining material quantity and timestamp sequence in the remaining material quantity data of the material tower, and sort the remaining material quantity of the material tower based on the timestamp sequence to obtain the sequence remaining material quantity; Query the intrinsic parameters of the feed tower corresponding to the multi-story pig farm, and calculate the feed tower correction coefficient of the multi-story pig farm based on the intrinsic parameters of the feed tower; Calculate the time interval between the timestamp sequences, and combine the remaining feed amount in the sequence, the time interval, and the feed tower correction coefficient to calculate the feed consumption rate corresponding to the multi-story pig farm using the following formula: Where A represents the feed consumption rate corresponding to multi-story pig farms. This represents the amount of remaining material in the sequence at time a. This represents the remaining material quantity in the sequence at time a+1. This represents the feed tower correction factor. Indicates time interval, Indicates the time correction factor; The feed control and management module is used to query the pig herd biological information of the multi-story pig farm, calculate the herd fat dispersion of the pig herd, determine the feed requirement quota of the multi-story pig farm by combining the pig herd biological information and the herd fat dispersion, obtain the corresponding IoT control device of the multi-story pig farm, and use the IoT control device to perform feed control and management of the multi-story pig farm by combining the stratified feeding rules and the feed requirement quota, and obtain the management results.

Images