A big data collection method and system of a smart livestock system

Abstract

This invention relates to the field of data acquisition technology, specifically to a big data acquisition method and system for a smart livestock system. The method includes: acquiring data on each animal in a livestock farm across different dimensions within the current time period; determining abnormal time periods for each animal in each dimension based on the distribution of data across different dimensions; filtering abnormal animals based on the overlap of abnormal time periods and data distribution across pairs of dimensions for the same animal; obtaining the degree of disease infection manifestation in each dimension for each region based on the number and severity of abnormal time periods for abnormal animals in the same dimension within each region; determining the data acquisition frequency for each dimension for each region based on the overlap of abnormal time periods and the degree of disease infection manifestation for abnormal animals in each dimension across other dimensions within each region, and then proceeding with subsequent data acquisition. This invention can adaptively adjust the data acquisition frequency for different dimensions in different regions within a livestock farm.

Description

Technical Field

[0001] This invention relates to the field of data acquisition technology, specifically to a big data acquisition method and system for a smart livestock system. Background Technology

[0002] A smart livestock system refers to the deep integration of advanced information technology, the Internet of Things (IoT), big data, and artificial intelligence (AI) to provide comprehensive data monitoring and analysis, thereby improving the management efficiency, animal health monitoring, and disease prevention and control capabilities of the livestock industry. Utilizing advanced IoT technology and sensor networks, the smart livestock system can monitor the health status, environmental conditions, and behavioral changes of livestock and poultry in real time. This real-time data collection can promptly detect abnormal signals, such as elevated animal body temperature and decreased activity levels, which may be early signs of disease. By collecting and analyzing data in real time, the system can provide early warnings to managers, prompting intervention measures before the disease spreads.

[0003] In the process of disease prevention and control in animal husbandry, existing technologies typically use the same collection frequency to collect data on all dimensions of each animal in the livestock farm during the big data collection process of smart livestock systems. However, since livestock farms are large in area and different regions may face different disease risks, and the data types affected by different diseases are also different, using a uniform frequency to collect all data can easily lead to data redundancy if the collection frequency is too high, while if the collection frequency is too low, it may not be able to effectively capture the subtle data change characteristics, resulting in failure to detect problems in time and reducing the overall prevention and control effect. Summary of the Invention

[0004] To address the problem that existing methods cannot adaptively adjust the data collection frequency for different regions and dimensions of livestock farms, the present invention aims to provide a big data collection method and system for a smart livestock system. The specific technical solution adopted is as follows:

[0005] In a first aspect, the present invention provides a big data collection method and system for an intelligent livestock system, the method comprising the following steps:

[0006] Obtain different dimensions of data for each animal in the livestock farm within the current time period;

[0007] The abnormal time periods for each animal in each dimension are determined based on the distribution of data from different dimensions for each animal; the severity of the abnormality for each animal is evaluated based on the overlap of abnormal time periods and the data distribution of the same animal in each pair of dimensions; and abnormal animals are screened based on the severity of the abnormality.

[0008] Based on the number and severity of abnormal time periods of abnormal animals in each region of the livestock farm under the same dimension, the degree of disease infection manifestation in each dimension of the data for each region is obtained; based on the overlap of abnormal time periods of abnormal animals in each dimension with other dimensions in each region, and the degree of disease infection manifestation, the data collection frequency for each dimension of each region is determined.

[0009] The aforementioned acquisition frequency was used to collect data on animals in different regions across different dimensions.

[0010] Preferably, determining the abnormal time period for each animal in each dimension based on the distribution of data from different dimensions for each animal includes:

[0011] For any dimension:

[0012] Calculate the data difference of any dimension of the candidate animal's data at each moment within the current time period and its next adjacent moment; if the data difference is greater than a preset difference threshold, then the corresponding moment is determined as an abnormal moment;

[0013] The time period consisting of consecutive adjacent abnormal moments is used as the candidate time period for the candidate animal in any of the dimensions.

[0014] By combining the duration, data range, and maximum data value of each candidate time period under any dimension of the candidate animals, the disease infection factor of the candidate animals under any dimension for each candidate time period is obtained. The duration, data range, and maximum data value are all positively correlated with the disease infection factor.

[0015] Candidate time periods where the infectious agent of a disease exceeds a preset infection threshold are considered abnormal time periods;

[0016] The candidate animal is any animal found in the livestock farm.

[0017] Preferably, the evaluation of the severity of abnormality for each animal based on the overlap of abnormal time periods and data distribution in pairwise dimensions of the same animal includes:

[0018] For any two dimensions: based on the overlap of abnormal time periods of candidate animals in any two dimensions, the degree of association of candidate animals in any two dimensions is obtained;

[0019] Count the number of times each moment is an anomalous moment across all dimensions within the current time period, and record the moment when the first count is greater than a preset value as the target moment; consecutive adjacent target moments constitute the target time period.

[0020] The severity of anomalies in candidate animals is determined by the number of times each moment is identified as a target moment across all dimensions, the duration of the target time period, and the degree of correlation between all pairs of dimensions.

[0021] Preferably, the determination of the correlation between candidate animals in any two dimensions includes:

[0022] For candidate animals:

[0023] Obtain the intersection time period between the abnormal time periods of the candidate animals in any two dimensions, and obtain the union time period corresponding to each intersection time period; calculate the correlation between the data values ​​of the candidate animals in each intersection time period in any two dimensions; calculate the first ratio between the duration of each intersection time period of the candidate animals in any two dimensions and the duration of the corresponding union time period.

[0024] The degree of association of candidate animals in any two dimensions is determined based on the correlation and the first ratio.

[0025] Preferably, the step of screening abnormal animals based on the severity of the abnormality includes: classifying animals whose severity of abnormality is greater than a preset severity threshold as abnormal animals.

[0026] Preferably, the step of obtaining the degree of disease infection manifestation in each dimension of data for each region based on the number and severity of abnormal animals in the same dimension within each region of the livestock farm includes:

[0027] For any area of ​​the livestock farm:

[0028] Calculate the first product of the number of abnormal time periods of each abnormal animal in any region under the dimension to be analyzed and the corresponding severity of the abnormality, and the first ratio between the number of abnormal animals that have had abnormal time periods under the dimension to be analyzed and the number of abnormal animals in any region.

[0029] By combining the first ratio and the first product, the degree of disease infection manifestation in any dimension of data to be analyzed in the region is obtained;

[0030] The dimension to be analyzed can be any dimension.

[0031] Preferably, obtaining the degree of disease infection manifestation of the data in any region to be analyzed by combining the first ratio and the first product includes:

[0032] Calculate the sum of the first products of all abnormal animals in any given region under the dimension to be analyzed;

[0033] The product of the summation and the first ratio is determined as the degree of disease infection manifestation in any dimension of data to be analyzed in the region.

[0034] Preferably, determining the data collection frequency for each dimension in each region based on the overlap of abnormal time periods in each dimension with other dimensions of abnormal animals within each region, and the degree of disease infection manifestation, includes:

[0035] For any area of ​​the livestock farm:

[0036] Based on the distribution of abnormal animals in any region during abnormal time periods under different dimensions, the similarity values ​​of other abnormal dimensions in any region, excluding the dimension to be analyzed, are obtained.

[0037] Calculate the normalized result of the ratio between the degree of disease infection manifestation and the similarity value of the data in any dimension to be analyzed in the region;

[0038] Calculate the sum of constant 1 and the normalization result, and take the floor of the product of the sum and the initial data acquisition frequency as the acquisition frequency of the data of the dimension to be analyzed in any region.

[0039] Preferably, obtaining similarity values ​​for any region in other abnormal dimensions besides the dimension to be analyzed, based on the distribution of abnormal animals in different dimensions during abnormal time periods in any region, includes:

[0040] Any abnormal animal in any region that has an abnormal time period under the dimension to be analyzed is recorded as the animal to be processed. All dimensions other than the dimension to be analyzed in the dimensions where the animal to be processed has had an abnormal time period constitute the reference dimension set of the animal to be processed.

[0041] The number of dimensions in the intersection of the reference dimension sets of all abnormal animals in any region that have abnormal time periods under the dimension to be analyzed is taken as the similarity value of other abnormal dimensions in any region besides the dimension to be analyzed.

[0042] Secondly, the present invention provides a big data acquisition system for a smart livestock system, the system being used to implement the above-mentioned method, the system comprising:

[0043] The data acquisition module is used to acquire different dimensions of data for each animal in the livestock farm within the current time period;

[0044] The filtering module is used to determine the abnormal time period of each animal in each dimension based on the distribution of data in different dimensions for each animal; to evaluate the severity of abnormality for each animal based on the overlap of abnormal time periods in pairs of dimensions and the data distribution; and to filter abnormal animals based on the severity of abnormality.

[0045] The frequency determination module is used to obtain the degree of disease infection manifestation in each dimension of data for each region based on the number and severity of abnormal time periods of abnormal animals in the same dimension in each region of the livestock farm; and to determine the collection frequency of data for each dimension of each region based on the overlap of abnormal time periods of abnormal animals in each dimension with other dimensions in each region, as well as the degree of disease infection manifestation.

[0046] The data acquisition module is used to collect data on animals in different regions and dimensions using the acquisition frequency.

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

[0048] This invention first analyzes the distribution of data across different dimensions for each animal in a livestock farm, identifying the abnormal time periods for each animal in each dimension. Then, based on the overlap of abnormal time periods for each pair of animals in different dimensions and the distribution of dimensional data, a preliminary assessment of the severity of abnormalities in each animal is made. Based on this assessment, multiple abnormal animals in the livestock farm are selected. Furthermore, by analyzing whether the data of each dimension of a single abnormal animal in the farm can better reflect the degree of abnormality, and the data correlation between multiple abnormal dimensions of the same animal, the importance of each dimension of data in each area of ​​the livestock farm is obtained. Subsequently, the collection frequency of each dimension of data in each area of ​​the livestock farm is adaptively adjusted, avoiding data redundancy while ensuring that the collected data can more accurately reflect the abnormalities of different dimensions of animal data in the livestock farm, thus improving the overall prevention and control effect. Attached Figure Description

[0049] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 A flowchart illustrating a big data collection method for an intelligent livestock system provided in an embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of the intersection and union time periods provided in an embodiment of the present invention;

[0052] Figure 3 This is a structural block diagram of a big data acquisition system for a smart livestock system provided in an embodiment of the present invention. Detailed Implementation

[0053] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following detailed description of a big data collection method and system for a smart livestock system based on the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0055] The following describes in detail, with reference to the accompanying drawings, an embodiment of the big data collection method and system for the intelligent livestock system provided by the present invention:

[0056] The specific scenario addressed in this embodiment is as follows: Most existing livestock farms are large-scale intensive farms with a large number of animals, resulting in high density in each area and more frequent contact between animals, which increases the chance of disease transmission. Therefore, when abnormal group behavior occurs in a certain area of ​​the farm, it is necessary to increase the frequency of data collection in the corresponding dimension to monitor early signs of disease transmission in a timely manner and prevent the spread of the epidemic.

[0057] This embodiment proposes a big data collection method for a smart livestock system, such as... Figure 1 As shown, a big data collection method for a smart livestock system in this embodiment includes the following steps:

[0058] Step S1: Obtain different dimensions of data for each animal in the livestock farm during the current time period.

[0059] Livestock farms are typically large, so they are divided into different areas based on factors such as the animals' growth stage, sex, and health status. Farms use fences, cages, or enclosures to separate the animals. These enclosures not only help prevent the spread of infectious diseases but also allow for space allocation, ensuring each animal lives in a suitable environment and enabling better control over the diet, health, and growth of each group.

[0060] Wearable devices, such as smart collars, are fitted to each animal in the livestock farm to collect multi-dimensional data on each animal, including body temperature, heart rate, respiratory rate, and movement speed. These smart collars are specifically designed for livestock; unlike pet smart collars which are primarily used for pet tracking and health monitoring, livestock smart collars focus on improving breeding efficiency and scientific management. The collars typically integrate multiple sensors, such as: a temperature sensor for real-time monitoring of the animal's body temperature; a heart rate sensor for monitoring heart rate using photoplethysmography (PPG) or electrode sensors; an accelerometer for measuring the animal's movement speed; an airflow sensor for analyzing respiratory rate by detecting changes in airflow or air pressure; and a GPS module for providing location information and calculating the animal's distance traveled. Data on each dimension of each animal in the livestock farm is collected within the current time period, which is the set of all historical moments with a time interval less than or equal to a preset duration. In this embodiment, the initial data collection frequency is set to 12 times / minute, and the preset duration is 1 hour. In specific applications, the implementer can adjust these settings according to specific circumstances.

[0061] Thus, this embodiment has collected multi-dimensional data for each animal in the livestock farm at each moment within the current time period. This embodiment has performed decimal scaling standardization on the collected multi-dimensional data to give them similar scales and distributions for better comparison. Decimal scaling standardization is a well-known technique and will not be elaborated upon here. It should be noted that all dimensional data mentioned subsequently are data after decimal scaling standardization.

[0062] Step S2: Determine the abnormal time period for each animal in each dimension based on the distribution of data in different dimensions for each animal; evaluate the severity of abnormality for each animal based on the overlap of abnormal time periods in pairs of dimensions and the data distribution; and screen abnormal animals based on the severity of abnormality.

[0063] Under normal circumstances, an animal's body temperature, heart rate, and respiratory rate change relatively little over time. This is mainly because organisms possess a sophisticated set of regulatory mechanisms that help them maintain physiological stability in the face of changes in the internal and external environment. Moreover, the space within a livestock farm is smaller than that of a pasture, and livestock do not engage in a lot of strenuous exercise. Therefore, when an animal's data in a certain dimension shows unstable fluctuations, it indicates a greater likelihood that the animal is experiencing an abnormality.

[0064] The following example uses an animal from a livestock farm for illustration. The method provided in this example can be used to process other animals in the livestock farm.

[0065] Specifically, any animal in the livestock farm is designated as a candidate animal.

[0066] For any dimension:

[0067] The absolute value of the difference between the candidate animal's data at each moment within the current time period and its adjacent next moment in that dimension is calculated, and this absolute value is taken as the data difference corresponding to each moment within the candidate animal's current time period. If the data difference is greater than a preset difference threshold, the corresponding moment is identified as an abnormal moment. It should be noted that since multi-dimensional data of the moment following the last moment in the current time period is not collected, this embodiment uses the data difference corresponding to the second-to-last moment in the current time period as the data difference corresponding to the last moment in the current time period. The time period consisting of consecutively adjacent abnormal moments is taken as the candidate time period for the candidate animal in that dimension. In this embodiment, the preset difference threshold is 0.6. In specific applications, the implementer can set it according to the specific situation.

[0068] Next, by combining the duration, data range, and maximum data value of each candidate time period of the candidate animal in this dimension, the disease infection factor of the candidate animal in each candidate time period in this dimension is obtained. The duration, the data range, and the maximum data value are all positively correlated with the disease infection factor.

[0069] Among them, a positive correlation means that the dependent variable increases as the independent variable increases, and the dependent variable decreases as the independent variable decreases. It can be an additive relationship, a multiplicative relationship, etc., which is determined by the actual application.

[0070] In this embodiment, a specific calculation formula for disease infection factors is given. The disease infection factor of the a-th animal in the w-th dimension and the s-th candidate time period can be expressed as:

[0071]

[0072] Among them, Y a,w,s t represents the disease infectious agent of animal a in the w-th dimension at the s-th candidate time period. a,w,s This represents the duration of the s-th candidate time period for the 'a'-th animal in the 'w'-th dimension. J represents the maximum data value of the a-th animal in the w-th dimension at the s-th candidate time period. a,w,s This represents the data range of the a-th animal in the w-th dimension at the s-th candidate time period, and norm() represents the normalization function.

[0073] The larger the maximum data value of animal a in the w-th dimension at the s-th candidate time period, the greater the likelihood of disease infection in the s-th candidate time period. The larger the data range of animal a in the w-th dimension at the s-th candidate time period, the greater the fluctuation range of the data in the w-th candidate time period, and the greater the likelihood of disease infection in the animal.

[0074] Using the above method, the disease infection factors of candidate animals in each dimension and each candidate time period can be obtained. Candidate time periods where the disease infection factor exceeds a preset infection threshold are designated as abnormal time periods. In this embodiment, the preset infection threshold is 0.7; in specific applications, the implementer may set this threshold according to the specific circumstances.

[0075] Abnormal health in animals is usually caused by multiple factors affecting simultaneously. For example, when an animal contracts a disease and develops a fever, its thermoregulatory center is affected, leading to an increase in body temperature. At the same time, to meet the increased metabolic demands, the heart needs to beat faster to deliver more oxygen and nutrients, resulting in a faster heart rate; the respiratory rate also increases accordingly to obtain more oxygen and expel carbon dioxide. Furthermore, the animal's discomfort can cause it to be unable to move normally, resulting in slower movement and shorter travel distances.

[0076] For any two dimensions: For a candidate animal, obtain the intersection time period between the abnormal time periods of the candidate animal in these two dimensions, and obtain the union time period corresponding to each intersection time period. For example... Figure 2 As shown in the figure, taking the w-th and v-th dimensions as examples, the figure illustrates the results of obtaining the intersection and union time periods. The Pearson correlation coefficient is calculated between the data values ​​of the candidate animals in each intersection time period under these two dimensions, and this Pearson correlation coefficient is used as the correlation between the data values ​​of the candidate animals in each intersection time period under these two dimensions. The ratio between the duration of each intersection time period and the duration of the corresponding union time period for the candidate animals under these two dimensions is calculated, and this ratio is recorded as the first ratio. The degree of association between the candidate animals in these two dimensions is determined based on the correlation and the first ratio.

[0077] In this embodiment, a specific formula for calculating the degree of association is given. The degree of association between the a-th animal and the w-th and v-th dimensions can be expressed as:

[0078]

[0079] Among them, C a,w,v Let t represent the degree of association between the a-th animal and the w-th and v-th dimensions, J represent the number of time intervals where the a-th animal intersects with the w-th and v-th dimensions, and t represent the degree of association between the a-th animal and the w-th and v-th dimensions. a,w,v,j This represents the duration of the j-th intersection time segment of the a-th animal in the w-th and v-th dimensions. r represents the duration of the j-th union time interval of the a-th animal in the w-th and v-th dimensions. a,w,v,j This represents the correlation between the data values ​​of the a-th animal at the j-th intersection time period under the w-th and v-th dimensions.

[0080] This represents the first ratio corresponding to the j-th intersection time period of the a-th animal under the w-th and v-th dimensions. The larger the first ratio and the higher the correlation between the data values ​​of the a-th animal under the j-th intersection time period of the w-th and v-th dimensions, the higher the degree of correlation between the data of the two dimensions.

[0081] Using the above method, we can obtain the degree of correlation between each pair of animals in each dimension.

[0082] Furthermore, the number of times each moment within the current time period is an abnormal moment across all dimensions is counted, and this number is recorded as the first count. Each moment within the current time period has a corresponding first count. Moments with first counts greater than a preset value are recorded as target moments. Consecutive adjacent target moments constitute a target time period. In this embodiment, the preset value is... Where N is the number of all dimensions.

[0083] The following explanation will continue using candidate animals as an example. Based on the number of times each moment in all target time periods of the candidate animal is identified as the target moment in all dimensions, the duration of the target time period, and the degree of correlation between all pairs of dimensions, the severity of the anomaly of the candidate animal is obtained.

[0084] In this embodiment, a specific formula for calculating the severity of the abnormality is given. The severity of the abnormality in the a-th animal can be expressed as:

[0085]

[0086] Among them, Y a This indicates the severity of the abnormality in the a-th animal. represents the average degree of correlation between all pairs of dimensions for the a-th animal, and Ne represents the number of target time periods. T represents the average number of times that all moments in the e-th target time period of the a-th animal are identified as target moments across all dimensions. e This represents the duration of the e-th target time period for the a-th animal, and norm() represents the normalization function.

[0087] This is used to reflect the persistence of abnormalities in animal a. When abnormalities occur continuously, it means that the data of animal a in various dimensions continue to show abnormalities during that period. With the cumulative effect, the animal's physical condition will be affected to a greater extent, indicating that the symptoms of the disease will gradually worsen and the severity of the abnormality will be higher.

[0088] Using the above method, the severity of abnormalities for each animal in the livestock farm can be obtained, and animals with an abnormality severity greater than a preset severity threshold are identified as abnormal animals. In this embodiment, the preset severity threshold is 0.7; in specific applications, the implementer can set it according to the specific circumstances.

[0089] Thus far, this embodiment has identified several abnormal animals.

[0090] Step S3: Based on the number and severity of abnormal time periods of abnormal animals in each region of the livestock farm under the same dimension, obtain the degree of disease infection manifestation in each dimension of the data for each region; based on the overlap of abnormal time periods of abnormal animals in each dimension with other dimensions in each region, and the degree of disease infection manifestation, determine the data collection frequency for each dimension of each region.

[0091] Livestock farms are typically large, and animals in different areas may face varying environmental and feeding conditions. If sick animals within a region exhibit common abnormalities in certain shared dimensions—for example, multiple sick animals simultaneously displaying elevated body temperature and increased heart rate—it's highly likely that a common problem exists in the area. This could be due to pathogen transmission leading to mass animal infections, or contamination of feed or water sources, causing multiple related health conditions to be affected simultaneously. By analyzing these shared abnormalities, potential health threats within the region can be quickly identified, allowing for timely adjustments to data collection frequency to prevent further deterioration and spread of the problem.

[0092] For any area of ​​the livestock farm:

[0093] Taking one dimension as an example, let's denote any dimension as the dimension to be analyzed. Calculate the product of the number of abnormal periods for each abnormal animal in the region under the dimension to be analyzed and the corresponding severity of the abnormality; this product is denoted as the first product. Calculate the ratio between the number of abnormal animals that experienced abnormal periods under the dimension to be analyzed and the total number of abnormal animals in the region; this ratio is denoted as the first ratio. Calculate the sum of the first products corresponding to all abnormal animals in the region under the dimension to be analyzed. The product of this sum and the first ratio is determined as the degree of disease infection manifestation in the data of the dimension to be analyzed in that region.

[0094] In this embodiment, a specific formula for calculating the degree of disease infection manifestation is given. The degree of disease infection manifestation in the w-th dimension of the q-th region can be expressed as:

[0095]

[0096] Among them, Z q,w N represents the degree of disease infection manifestation in the q-th region and the w-th dimension of the data. q,wB represents the number of abnormal animals that occurred during abnormal periods in the w-th dimension. q P represents the number of abnormal animals in the q-th region. q,w,b Y represents the number of abnormal time periods for the b-th abnormal animal in the q-th region under the w-th dimension. q,b This indicates the severity of the abnormality of the b-th abnormal animal in the q-th region.

[0097] P q,w,b ×Y q,b This represents the first product corresponding to the b-th abnormal animal in the q-th region under the w-th dimension. This represents the sum of the first products of all abnormal animals in the q-th region in the w-th dimension. This represents the first ratio. The more animals that exhibited anomalies in the w-th dimension, the more pronounced the anomalies were in the data on the w-th dimension.

[0098] In livestock farms, the same disease usually does not cause anomalies in one dimension of data. For example, an epidemic can cause an increase in body temperature and heart rate. Therefore, it is also necessary to analyze the similarity of each dimension of abnormal animals in each region with other abnormal dimensions.

[0099] For any area of ​​the livestock farm:

[0100] Based on the distribution of abnormal animal periods across different dimensions in the region, similarity values ​​for other abnormal dimensions (excluding the dimension to be analyzed) are obtained. Specifically, any abnormal animal in the region that exhibits an abnormal period under the dimension to be analyzed is designated as a "to-be-treated animal." All dimensions other than the dimension to be analyzed within the dimensions where the to-be-treated animal exhibited an abnormal period constitute a reference dimension set for the to-be-treated animal. Each abnormal animal in the region that exhibits an abnormal period under the dimension to be analyzed has a corresponding reference dimension set. The number of dimensions in the intersection of the reference dimension sets of all abnormal animals in the region that exhibit abnormal periods under the dimension to be analyzed is used as the similarity value for other abnormal dimensions in the region (excluding the dimension to be analyzed). The greater the number of dimensions in the intersection of the reference dimension sets of abnormal animals in the region that exhibit abnormal periods, the higher the similarity value for other abnormal dimensions in the region (excluding the dimension to be analyzed). This indicates a higher likelihood of abnormalities caused by the same disease, meaning a stronger correlation between the dimension to be analyzed and other dimensions, and a lower complexity of the cause of the abnormality. Therefore, the data collection frequency for the dimension to be analyzed should be set lower to avoid collecting a large amount of data that is redundant or strongly correlated with other dimensions, thus reducing the data processing burden.

[0101] Based on the above characteristics, the normalized result of the ratio between the degree of disease infection manifestation and the similarity value of the data to be analyzed in the region is calculated; the sum of constant 1 and the normalized result is calculated, and the product of the sum and the initial data collection frequency is rounded up as the collection frequency of the data to be analyzed in any region.

[0102] In this embodiment, a specific formula for calculating the collection frequency is given. The collection frequency of the data in the w-th dimension of the q-th region of the livestock farm can be expressed as:

[0103]

[0104] Among them, f q,w Z represents the data collection frequency of the q-th region and w-th dimension of the livestock farm, f0 represents the initial data collection frequency, and Z represents the frequency of data collection. q,w X represents the degree of disease infection manifestation in the q-th region and the w-th dimension of the data. q,w This represents the similarity value of the w-th dimension of the q-th region to other dimensions, and norm() represents the normalization function. The symbol indicates rounding up.

[0105] In this embodiment, the norm function is used to... The value is normalized to [-0.6, 0.6]. The greater the degree of disease infection manifestation in the w-th dimension of the q-th region and the smaller the similarity value between the w-th dimension of the q-th region and other dimensions, the more abnormal the data in the w-th dimension of the q-th region is, and the higher the data collection frequency of the dimension to be analyzed should be set.

[0106] Using the above method, the data collection frequency for each region and dimension of the livestock farm can be determined.

[0107] Step S4: Collect data on animals in different regions and dimensions using the acquisition frequency.

[0108] After determining the data collection frequency for each region and dimension of the livestock farm, this frequency will be used for subsequent data collection.

[0109] Analyzing the group activities in each area helps reveal the overall health status and behavioral trends of the herd. For example, if most animals suddenly experience a rise in body temperature or a drop in heart rate, this could be a signal of disease transmission. Therefore, increasing the frequency of data collection for body temperature and heart rate can more accurately capture early signs of these diseases or health problems and is less susceptible to interference from accidental changes in individual animals. The microcontroller is the core component of the collar. It controls the data collection process according to a preset initial data collection frequency program. After obtaining the collection frequency for each dimension of data in each area, it uses this frequency as the collection frequency for all animals in that area for subsequent data collection. The microcontroller triggers the corresponding sensors to collect data at regular intervals according to its own collection frequency settings. For example, if the body temperature collection frequency is set to once every 10 seconds, the microcontroller will send a collection command to the body temperature sensor every 10 seconds. Data for each dimension is collected in the next time period in the same manner. In this embodiment, a time period is one hour. The process iterates continuously in the above manner, adjusting the data collection frequency every hour to complete the big data collection for the smart livestock system.

[0110] Thus, the big data collection for the intelligent livestock system has been completed using the method provided in this embodiment.

[0111] This embodiment first analyzes the distribution of data across different dimensions for each animal in the livestock farm, identifying the abnormal time periods for each animal in each dimension. Then, based on the overlap of abnormal time periods for each pair of animals in different dimensions and the distribution of dimensional data, a preliminary assessment of the severity of abnormalities in each animal is made. Based on this assessment, multiple abnormal animals in the livestock farm are selected. Furthermore, by analyzing whether the data of each dimension of a single abnormal animal in the farm can better reflect the degree of abnormality, and the data correlation between multiple abnormal dimensions of the same animal, the importance of each dimension of data in each area of ​​the livestock farm is obtained. Subsequently, the collection frequency of each dimension of data in each area of ​​the livestock farm is adaptively adjusted, avoiding data redundancy while ensuring that the collected data can more accurately reflect the abnormalities of different dimensions of animal data in the livestock farm, thus improving the overall prevention and control effect.

[0112] An example of a big data acquisition system for a smart livestock system:

[0113] See Figure 3 The diagram shows a structural block diagram of a big data acquisition system for a smart livestock system provided in an embodiment of the present invention. The system may include a data acquisition module, a filtering module, a frequency determination module, and a data collection module.

[0114] The data acquisition module is used to acquire different dimensions of data for each animal in the livestock farm during the current time period.

[0115] The filtering module is used to determine the abnormal time period of each animal in each dimension based on the distribution of data in different dimensions for each animal; to evaluate the severity of abnormality for each animal based on the overlap of abnormal time periods in pairs of dimensions and the data distribution; and to filter abnormal animals based on the severity of abnormality.

[0116] The frequency determination module is used to obtain the degree of disease infection manifestation in each dimension of data for each region based on the number and severity of abnormal time periods of abnormal animals in the same dimension in each region of the livestock farm; and to determine the collection frequency of data for each dimension of each region based on the overlap of abnormal time periods of abnormal animals in each dimension with other dimensions in each region, as well as the degree of disease infection manifestation.

[0117] The data acquisition module is used to collect data on animals in different regions and dimensions using the acquisition frequency.

[0118] It should be understood that Figure 2 The structural block diagram and modules of the big data acquisition system for a smart livestock system shown can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented using hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated hardware. Those skilled in the art will understand that the above-described methods and systems can be implemented using computer-executable instructions and / or included in processor control code, for example, on a media such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The systems and modules of this specification can be implemented not only with hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips and transistors, or programmable hardware devices such as field-programmable gate arrays and programmable logic devices, but also with software executed by various types of processors, or with a combination of the above-described hardware circuits and software (e.g., firmware).

[0119] For more details about the above modules, please refer to other parts of this manual; they will not be repeated here.

[0120] In other embodiments, a medium is also provided, the medium storing at least one computer-executable program, which, when executed by a computer, causes the computer to perform the steps in the big data collection method of the smart livestock system described above, the medium being a computer-readable storage medium.

[0121] The system and media provided are used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can be referred to the beneficial effects of the corresponding methods provided above, and will not be repeated here.

[0122] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A big data collection method of a smart livestock system, characterized by, The method includes the following steps: Obtain different dimensions of data for each animal in the livestock farm within the current time period; The abnormal time periods for each animal in each dimension are determined based on the distribution of data from different dimensions for each animal; the severity of the abnormality for each animal is evaluated based on the overlap of abnormal time periods and the data distribution of the same animal in each pair of dimensions; and abnormal animals are screened based on the severity of the abnormality. Based on the number and severity of abnormal time periods of abnormal animals in each region of the livestock farm under the same dimension, the degree of disease infection manifestation in each dimension of the data for each region is obtained; based on the overlap of abnormal time periods of abnormal animals in each dimension with other dimensions in each region, and the degree of disease infection manifestation, the data collection frequency for each dimension of each region is determined. Data on animals in different regions and from different dimensions are collected using the aforementioned collection frequency; The process of determining the abnormal time periods for each animal in each dimension based on the distribution of data from different dimensions for each animal includes: For any dimension: Calculate the data difference of any dimension of the candidate animal's data at each moment within the current time period and its next adjacent moment; if the data difference is greater than a preset difference threshold, then the corresponding moment is determined as an abnormal moment; The time period consisting of consecutive adjacent abnormal moments is used as the candidate time period for the candidate animal in any of the dimensions. By combining the duration, data range, and maximum data value of each candidate time period under any dimension of the candidate animals, the disease infection factor of the candidate animals under any dimension for each candidate time period is obtained. The duration, data range, and maximum data value are all positively correlated with the disease infection factor. Candidate time periods where the infectious agent of a disease exceeds a preset infection threshold are considered abnormal time periods; The candidate animal can be any animal found in the livestock farm; The severity of abnormalities in each animal is evaluated based on the overlap of abnormal time periods and data distribution across pairwise dimensions within the same animal, including: For any two dimensions: based on the overlap of abnormal time periods of candidate animals in any two dimensions, the degree of association of candidate animals in any two dimensions is obtained; Count the number of times each moment is an anomalous moment across all dimensions within the current time period, and record the moment when the first count is greater than a preset value as the target moment; consecutive adjacent target moments constitute the target time period. The severity of anomalies in candidate animals is determined by the number of times each moment is identified as a target moment across all dimensions, the duration of the target time period, and the degree of correlation between all pairs of dimensions. The method of determining the data collection frequency for each dimension in each region based on the overlap of abnormal time periods in each dimension with other dimensions of abnormal animals in each region, and the degree of disease infection manifestations, includes: For any area of ​​the livestock farm: Based on the distribution of abnormal animals in any region during abnormal time periods under different dimensions, the similarity values ​​of other abnormal dimensions in any region, excluding the dimension to be analyzed, are obtained. Calculate the normalized result of the ratio between the degree of disease infection manifestation and the similarity value of the data in any dimension to be analyzed in the region; Calculate the sum of constant 1 and the normalization result, and take the floor of the product of the sum and the initial data acquisition frequency as the acquisition frequency of the data of any dimension to be analyzed in any region.

2. The big data collection method of the intelligent livestock system according to claim 1, characterized in that, The determination of the correlation between candidate animals in any two of the stated dimensions includes: For candidate animals: Obtain the intersection time period between the abnormal time periods of the candidate animals in any two dimensions, and obtain the union time period corresponding to each intersection time period; calculate the correlation between the data values ​​of the candidate animals in each intersection time period in any two dimensions; calculate the first ratio between the duration of each intersection time period of the candidate animals in any two dimensions and the duration of the corresponding union time period. The degree of association of candidate animals in any two dimensions is determined based on the correlation and the first ratio.

3. The big data collection method of the intelligent livestock system according to claim 1, characterized in that, The method of screening abnormal animals based on the severity of the abnormality includes: classifying animals whose severity of abnormality is greater than a preset severity threshold as abnormal animals.

4. The big data collection method of the intelligent livestock system according to claim 1, characterized in that, The degree of disease infection manifestation in each dimension of data for each region is obtained based on the number and severity of abnormal animals in each region of the livestock farm during the same time period, including: For any area of ​​the livestock farm: Calculate the first product of the number of abnormal time periods of each abnormal animal in any region under the dimension to be analyzed and the corresponding severity of the abnormality, and the first ratio between the number of abnormal animals that have had abnormal time periods under the dimension to be analyzed and the number of abnormal animals in any region. By combining the first ratio and the first product, the degree of disease infection manifestation in any dimension of data to be analyzed in any region is obtained.

5. The big data collection method of the intelligent livestock system according to claim 4, characterized in that, The process of combining the first ratio and the first product to obtain the degree of disease infection manifestation in any region of the data to be analyzed includes: Calculate the sum of the first products of all abnormal animals in any given region under the dimension to be analyzed; The product of the summation and the first ratio is determined as the degree of disease infection manifestation in any dimension of data to be analyzed in the region.

6. The big data collection method of the intelligent livestock system according to claim 1, characterized in that, The step of obtaining similarity values ​​for abnormal dimensions other than the dimension to be analyzed in any region based on the distribution of abnormal animals in different dimensions during abnormal time periods in any region includes: Any abnormal animal in any region that has an abnormal time period under the dimension to be analyzed is recorded as the animal to be processed. All dimensions other than the dimension to be analyzed in the dimensions where the animal to be processed has had an abnormal time period constitute the reference dimension set of the animal to be processed. The number of dimensions in the intersection of the reference dimension sets of all abnormal animals in any region that have abnormal time periods under the dimension to be analyzed is taken as the similarity value of other abnormal dimensions in any region besides the dimension to be analyzed.

7. A big data acquisition system for an intelligent livestock system, the system being used to implement the method of claim 1, characterized in that, The system includes: The data acquisition module is used to acquire different dimensions of data for each animal in the livestock farm within the current time period; The filtering module is used to determine the abnormal time period of each animal in each dimension based on the distribution of data in different dimensions for each animal; to evaluate the severity of abnormality for each animal based on the overlap of abnormal time periods in pairs of dimensions and the data distribution; and to filter abnormal animals based on the severity of abnormality. The frequency determination module is used to obtain the degree of disease infection manifestation in each dimension of data for each region based on the number and severity of abnormal time periods of abnormal animals in the same dimension in each region of the livestock farm; and to determine the collection frequency of data for each dimension of each region based on the overlap of abnormal time periods of abnormal animals in each dimension with other dimensions in each region, as well as the degree of disease infection manifestation. The data acquisition module is used to collect data on animals in different regions and dimensions using the acquisition frequency.

Images