Intelligent epidemic prevention system for animal husbandry and veterinary in remote areas

Abstract

The invention relates to the technical field of data processing, in particular to an intelligent epidemic prevention system for animal husbandry and veterinary in remote areas, which comprises a segmentation module, a harmful gas influence degree acquisition module, a correlation coefficient acquisition module and a data weight acquisition module, the historical time period is segmented according to the change of the total number of sick animals every day in the historical time period of the farm, and the harmful gas influence degree of each segment is obtained according to the abnormal condition of the harmful gas concentration of each segment of the farm and the disease development severity of each segment. Obtaining a change relation between the daily harmful gas concentration in each segment and the number of newly increased sick animals to obtain a correlation coefficient of each segment, fusing the harmful gas influence degree and the correlation coefficient of each segment to obtain a data weight of each harmful gas concentration in each segment, and ensuring reasonable distribution of the data weights; and the prediction accuracy of the farm environment data is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an intelligent animal husbandry and veterinary epidemic prevention system for remote areas. Background Art

[0002] There are certain defects in the infrastructure, personnel and technology of animal husbandry epidemic prevention in remote areas. The epidemic prevention facilities are simple and lack basic epidemic prevention facilities, such as standard isolation houses and disinfection channels, which make it difficult to effectively prevent and control and isolate diseases. Once a disease occurs, it is easy to spread rapidly. Due to the lack of advanced disease detection equipment and laboratories, animal diseases cannot be accurately diagnosed and monitored in a timely manner. Judgments can only be made based on experience, which may lead to misdiagnosis or missed diagnosis and delay epidemic prevention work.

[0003] Livestock epidemic prevention in remote areas can be achieved through real-time monitoring and collection of animal disease information and on-farm environmental data, as well as analysis and prediction of this data. This predicted environmental data can be used to proactively identify potential diseases, identify potential risks promptly, and enable proactive response to mitigate losses. Currently, the Auto-Regressive Integrated Moving Average (ARIMA) model is commonly used to predict environmental data. When using the ARIMA model for prediction, appropriate data weighting can help the model more accurately capture various data characteristics and reduce prediction errors. However, when using the ARIMA model to predict on-farm environmental data based on animal disease information and on-farm environmental data, the weighting of each data point is assigned indiscriminately, meaning all data points are given equal weight. This can lead to irrational data weighting. For example, data with minimal impact on the predicted on-farm environmental data may be overweighted, while important data may be underweighted. This can lead to biased data fit in the ARIMA model, affecting the accuracy of the on-farm environmental data prediction and, consequently, livestock epidemic prevention. Summary of the Invention

[0004] In order to solve the technical problem that the data weight distribution method of the ARIMA model is unreasonable, which affects the prediction accuracy of farm environmental data, the purpose of the present invention is to provide an intelligent animal husbandry and veterinary epidemic prevention system for remote areas. The technical solutions adopted are as follows: The present invention provides an intelligent animal husbandry and veterinary epidemic prevention system for remote areas, comprising: A segmentation module, for segmenting the historical time period according to the change in the total number of sick animals per day in the historical time period of the farm; The harmful gas impact degree acquisition module is used to obtain the harmful gas impact degree of each segment based on the abnormal concentration of harmful gases in the farm and the severity of the disease in each segment; The correlation coefficient acquisition module is used to obtain the changing relationship between the daily concentration of harmful gases and the number of newly sick animals in each segment, and obtain the correlation coefficient of each segment; The data weight acquisition module is used to fuse the harmful gas impact degree and correlation coefficient of each segment to obtain the data weight of each harmful gas concentration in each segment.

[0005] In an exemplary embodiment, before the correlation coefficient acquisition module, the system further includes a harmful gas impact degree correction module, which is used to: Each segment is clustered according to the total number of sick animals in the segment to obtain multiple clusters; According to the length of each segment and the length of the longest segment in the same cluster, the segment length feature of each segment is obtained; According to the harmful gas impact degree of each segment and the maximum harmful gas impact degree in the same cluster, the harmful gas impact degree characteristics of each segment are obtained; Based on the correction coefficient of each segment, the harmful gas impact degree of each segment is corrected; the correction coefficient is obtained by fusing the segment length characteristics and the harmful gas impact degree characteristics.

[0006] In an exemplary embodiment, the correlation coefficient acquisition module is specifically configured to: Determine the abnormal degree of harmful gas in each segment every day, and obtain a segmented abnormal degree sequence of harmful gas; The number of newly added sick animals in each segment is used to obtain the sequence of the number of newly added sick animals in each segment on a daily basis; Obtain the correlation coefficient between the segmented harmful gas abnormality degree series and the new number of sick animals series.

[0007] In an exemplary embodiment, the process of obtaining the abnormal degree of harmful gas includes: Obtaining the number of abnormal harmful gas concentrations per day in the segment, the maximum abnormal degree per day, and the length of the longest abnormal segment per day; the abnormal harmful gas concentration is a harmful gas concentration greater than a preset normal harmful gas concentration threshold, and the abnormal degree is the degree to which the harmful gas concentration exceeds the preset normal harmful gas concentration threshold; the abnormal segment is obtained from continuous abnormal harmful gas concentrations; The daily abnormal degree of harmful gas is obtained according to the number of abnormal harmful gas concentrations per day, the maximum abnormal degree per day, and the length of the longest abnormal section per day.

[0008] In an exemplary embodiment, the data weight acquisition module is specifically configured to: The importance of harmful gases in each segment is obtained by integrating the harmful gas impact degree and correlation coefficient of each segment; According to the importance of harmful gases in each segment and the number of abnormal harmful gas concentrations, the data weight of each abnormal harmful gas concentration in each segment is obtained; the abnormal harmful gas concentration is a harmful gas concentration that is greater than a preset harmful gas normal concentration threshold; the data weights of the abnormal harmful gas concentrations in the same segment are the same.

[0009] In an exemplary embodiment, the data weight of each harmful gas concentration other than the abnormal harmful gas concentration in each segment is a preset constant, and the preset constant is smaller than the data weight of the abnormal harmful gas concentration in each segment.

[0010] In an exemplary embodiment, the process of obtaining the abnormal situation of the harmful gas concentration includes: Acquire the abnormal harmful gas concentration in the segment, and obtain the abnormal segment from the continuous abnormal harmful gas concentration; the abnormal harmful gas concentration is the harmful gas concentration greater than the preset harmful gas normal concentration threshold; The abnormal continuity degree of the segment is obtained according to the number of abnormal segments in the segment and the length of the longest abnormal segment in the segment; the abnormal continuity degree is proportional to the number of abnormal segments and the length of the longest abnormal segment.

[0011] In an exemplary embodiment, the degree of influence of the harmful gas is obtained by the abnormal continuity and the severity of the disease progression, and the degree of influence of the harmful gas is proportional to both the abnormal continuity and the severity of the disease progression.

[0012] In an exemplary embodiment, the process of obtaining the severity of the disease progression includes: Obtaining a change trend in the number of newly added sick animals in each segment, wherein the change trend represents a change pattern in the number of newly added sick animals on each day in the segment; Obtaining a segmented number characteristic of sick animals, wherein the number characteristic of sick animals is the average of the total number of sick animals on each day in the segment; By integrating the change trend and the number characteristics of sick animals, the severity of the disease progression in segments is obtained.

[0013] In an exemplary embodiment, the segmentation module is specifically configured to segment the historical time period using a mean segmentation algorithm according to changes in the total number of sick animals per day within the historical time period of the farm.

[0014] The present invention has the following beneficial effects: by segmenting historical time periods and grouping days with similar numbers of sick animals into one segment, the accuracy of subsequent data analysis can be improved. The importance of the harmful gas concentration in each segment can be determined by the degree of harmful gas impact in each segment and the correlation coefficient of each segment, thereby obtaining the data weight of each harmful gas concentration in each segment. The data weight of each harmful gas concentration in each segment in the ARIMA model can be determined based on the importance of the harmful gas concentration in each segment, thereby allocating different weights to harmful gas concentration data of different importance, ensuring that the data weight distribution is reasonable, and thus helping the ARIMA model to accurately capture the changing trend of the data and improve the prediction accuracy of farm environmental data. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a schematic diagram of the structure of an intelligent animal husbandry and veterinary epidemic prevention system for remote areas provided by one embodiment of the present invention; Figure 2 This is a flowchart of the steps corresponding to various modules of the intelligent animal husbandry and veterinary epidemic prevention system for remote areas provided by one embodiment of the present invention; Figure 3 This is a flow chart for obtaining abnormal conditions of harmful gas concentrations provided by one embodiment of the present invention; Figure 4 This is a flow chart for obtaining the severity of a disease progression provided by one embodiment of the present invention; Figure 5 This is a specific implementation flow chart of a harmful gas impact degree correction module provided by an embodiment of the present invention; Figure 6 This is a specific implementation flow chart of a correlation coefficient acquisition module provided by an embodiment of the present invention; Figure 7 This is a flow chart for obtaining the abnormal degree of harmful gas provided by one embodiment of the present invention; Figure 8 This is a specific implementation flow chart of a data weight acquisition module provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0016] To further illustrate the technical means and effects employed by the present invention to achieve its intended objectives, the following detailed description of the specific embodiments, structures, features, and effects of the present invention is provided in conjunction with the accompanying drawings and preferred embodiments. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the art to which this invention pertains. The data and information collected in this application were obtained with full consent and authorization.

[0018] This embodiment provides an intelligent animal husbandry and veterinary epidemic prevention system for remote areas, such as Figure 1 As shown, it includes a segmentation module, a harmful gas impact degree acquisition module, a correlation coefficient acquisition module and a data weight acquisition module. Each module can be a software module, which is essentially a corresponding method step; it can also be a hardware module, and the executed method steps are configured in the hardware module so that the hardware module realizes the corresponding function. Accordingly, the intelligent epidemic prevention system for animal husbandry and veterinary medicine in remote areas can be a software system, configured in relevant processors, computer hosts, and relevant veterinary supervision platforms; it can also be a hardware system, such as a server, computer host, etc. This embodiment does not limit the specific configuration methods of each module and the intelligent epidemic prevention system for animal husbandry and veterinary medicine in remote areas.

[0019] The intelligent animal husbandry and veterinary epidemic prevention system for remote areas provided in this embodiment is applied to farms, specifically indoor farms. The animals raised in the farms are not limited. Environmental factors such as air quality in the farms are closely related to the survival and spread of pathogens. The concentration of harmful gases in the farms determines the air quality. Harmful gases are common gases in farms that are harmful to animal breeding, such as ammonia, hydrogen sulfide, etc. If the concentration of harmful gases is too high, it will irritate the respiratory mucosa of animals, destroy the defense barrier of the respiratory tract, and make animals more susceptible to respiratory diseases. Accordingly, the diseases suffered by animals in the farms of this embodiment are specifically respiratory diseases. This embodiment monitors and collects the concentration of harmful gases in the farms in real time. In an exemplary embodiment, the harmful gas in this embodiment is specifically ammonia, and the concentration of the harmful gas is the concentration of ammonia.

[0020] Since ammonia is less dense than air, it tends to accumulate on top of the farm. Ammonia is also produced in animal activity areas. Therefore, multiple ammonia monitoring points are set up on top of the farm and in animal activity areas, such as the top of the farm and at the same height as the animal activity level. Furthermore, ammonia monitoring points should avoid locations such as vents and exhaust fans to avoid affecting the measurement results. The number of ammonia monitoring points is determined based on actual needs. An ammonia concentration sensor is provided at each ammonia monitoring point to detect the ammonia concentration at the corresponding ammonia monitoring point. The detection frequency of the ammonia concentration sensor is determined based on actual needs, such as once every ten minutes. The sampling frequencies of the various ammonia concentration sensors are the same, and data is collected synchronously. During each test, the average ammonia concentration of all ammonia monitoring points is used as the ammonia concentration data for that test.

[0021] This embodiment predicts future ammonia concentrations in a farm based on data from a historical time period, thereby enabling early detection of abnormalities in the farm environment based on the predicted ammonia concentrations. The length of the historical time period is set based on actual needs, for example, 30 days. The last day of the historical time period is the current day. Therefore, the historical time period includes multiple days, each of which includes multiple ammonia sampling times.

[0022] This embodiment also counts the total number of sick animals in the farm every day during the farm's historical time period, as well as the number of newly sick animals in the farm every day. The newly sick animals refer to the number of animals that were not sick the day before but became sick on the day of the previous day.

[0023] like Figure 2 As shown, the method steps corresponding to each module are as follows: A segmentation module is used to segment the historical time period according to the change in the total number of sick animals per day in the historical time period of the farm; The harmful gas impact degree acquisition module is used to obtain the harmful gas impact degree of each segment based on the abnormal concentration of harmful gases in the farm and the severity of the disease in each segment; The correlation coefficient acquisition module is used to obtain the changing relationship between the daily concentration of harmful gases and the number of newly sick animals in each segment, and obtain the correlation coefficient of each segment; The data weight acquisition module is used to fuse the harmful gas impact degree and correlation coefficient of each segment to obtain the data weight of each harmful gas concentration in each segment.

[0024] The specific implementation process of each module is described below with reference to the accompanying drawings.

[0025] The segmentation module is used to segment the historical time period according to the change in the total number of sick animals per day in the historical time period of the farm.

[0026] The changing trend in the total number of sick animals on a farm over a historical period can provide a direct indication of the prevalence of the disease within the farm, reflecting issues with the farming environment and the effectiveness of disease prevention measures. If the total number of sick animals continues to rise daily over a historical period, it indicates that the disease is spreading rapidly and the overall health of the farm is deteriorating. A gradual decrease in the number of sick animals may indicate that the disease is under effective control and the farm's health is improving. Therefore, the first step is to determine the disease's progression on the farm based on the daily changes in the total number of sick animals over the historical period. Since the farm environment is not fixed over the historical period, and the total number of sick animals is not exactly the same each day, it is necessary to first determine the daily changes in the total number of sick animals on the farm over the historical period.

[0027] Based on the total number of sick animals per day in the historical time period of the farm, a total number of sick animals sequence is obtained, and then the mean segmentation algorithm is used to segment the total number of sick animals sequence to achieve segmentation of the historical time period and obtain multiple segments. The mean segmentation algorithm determines the segmentation boundaries based on the mean of the total number of sick animals sequence, so that the data in each segment has relatively consistent characteristics, that is, the mean of the data in each segment is relatively stable. It should be understood that the length of each segment, that is, the number of days included, is not exactly the same. For any segment, the average value of the total number of sick animals on each day in the segment is also obtained to characterize the overall level of the number of sick animals in the segment.

[0028] The harmful gas impact degree acquisition module is used to obtain the harmful gas impact degree of each segment based on the abnormal situation of the harmful gas concentration in the farm in each segment and the severity of the disease development in each segment.

[0029] Taking each segment as the analysis object, the degree of harmful gas impact in each segment is obtained based on the abnormal concentration of harmful gases in the farms of each segment and the severity of the disease development in each segment.

[0030] Environmental factors are the primary cause of the rapid spread and severe development of diseases on farms. Poor farm environments, such as poor ventilation leading to high concentrations of harmful gases like ammonia, can reduce respiratory defenses and increase the risk of respiratory infections. Environmental factors significantly impact pathogen transmission and animal physiology. Good air quality ensures the proper function of the animal's respiratory mucosa and reduces pathogen growth. Therefore, it's crucial to first analyze whether the severity of respiratory diseases on farms is due to poor air quality.

[0031] For the sake of convenience, any segment is set as the target segment. First, based on the concentration of harmful gases in the target segment, the abnormality of the concentration of harmful gases in the target segment is obtained. In an exemplary embodiment, Figure 3 As shown, a specific process for obtaining abnormal conditions of harmful gas concentration is given below: Step 2-1: Obtain abnormal harmful gas concentrations in the segment, and obtain abnormal segments from continuous abnormal harmful gas concentrations.

[0032] Ammonia concentration, a key environmental indicator for farms, significantly impacts animal health and production performance. Generally speaking, ammonia concentrations in farms should be controlled below 20 ppm (milligrams per cubic meter), considered a relatively safe range that ensures the normal growth and health of farmed animals and reduces the incidence of respiratory diseases and other problems. Accordingly, a threshold for normal hazardous gas concentration is preset. This threshold serves as the upper limit of the safe range, for example, 20 ppm (milligrams per cubic meter).

[0033] The harmful gas concentrations at each moment in the target segment are arranged in chronological order and compared with a preset normal harmful gas concentration threshold. Thus, the harmful gas concentrations in the target segment are classified into two categories: one category is a harmful gas concentration greater than the preset normal harmful gas concentration threshold, and the other category is a harmful gas concentration less than or equal to the preset normal harmful gas concentration threshold. A harmful gas concentration greater than the preset normal harmful gas concentration threshold is defined as an abnormal harmful gas concentration, and a harmful gas concentration less than or equal to the preset normal harmful gas concentration threshold is defined as a normal harmful gas concentration.

[0034] Obtain the degree of abnormality of each abnormal harmful gas concentration. The degree of abnormality is the degree to which the abnormal harmful gas concentration exceeds a preset normal harmful gas concentration threshold. Specifically, the difference between the abnormal harmful gas concentration and the preset normal harmful gas concentration threshold is calculated, and the difference obtained is the degree of abnormality of the abnormal harmful gas concentration. The higher the degree to which the abnormal harmful gas concentration exceeds the preset normal harmful gas concentration threshold, that is, the greater the difference between the abnormal harmful gas concentration and the preset normal harmful gas concentration threshold, the higher the degree of abnormality of the abnormal harmful gas concentration. It should be understood that for a normal harmful gas concentration, since it is less than or equal to the preset normal harmful gas concentration threshold, the degree of abnormality of the normal harmful gas concentration is set to 0.

[0035] Therefore, each harmful gas concentration in the target segment is divided into abnormal harmful gas concentration and normal harmful gas concentration. Abnormal harmful gas concentration may appear individually and sporadically, that is, an occasional abnormality, or multiple continuous abnormal harmful gas concentrations may appear, that is, the harmful gas concentration is continuously abnormal for a certain period of time. At least two abnormal harmful gas concentrations that are consecutive in time in the target segment are combined into an abnormal segment, thereby obtaining several abnormal segments in the target segment. It should be understood that if the harmful gas concentrations in the target segment are all normal harmful gas concentrations, indicating that there is no abnormal harmful gas concentration in the target segment, then the subsequent processing process will no longer be performed, but the data weights of the various harmful gas concentrations in the segment that meets this situation will be directly set to the preset constants described later. Moreover, if there is no abnormal segment in the target segment, then the subsequent processing process will no longer be performed, but the data weights of the various harmful gas concentrations in the segment that meets this situation will be directly set to the preset constants described later.

[0036] Step 2-2: According to the number of abnormal segments in the segment and the length of the longest abnormal segment in the segment, the abnormal continuity degree of the segment is obtained.

[0037] The number of abnormal segments in the target segment and the length of each abnormal segment in the target segment (length refers to the number of abnormal harmful gas concentrations contained in each segment) are obtained. The longest abnormal segment is then determined from the lengths of the abnormal segments in the target segment, thereby obtaining the length of the longest abnormal segment in the target segment. The greater the number of abnormal segments in the target segment, the more obvious the abnormal continuity of the target segment, that is, the higher the abnormal continuity degree. The longer the longest abnormal segment in the target segment, the more obvious the abnormal continuity of the target segment, that is, the higher the abnormal continuity degree. Therefore, the abnormal continuity degree is proportional to both the number of abnormal segments and the length of the longest abnormal segment.

[0038] In an exemplary embodiment, a specific quantitative method for the abnormal continuity degree is given as follows: ; in, Indicates the abnormal continuity degree of the xth segment, represents the number of abnormal segments in the xth segment, Indicates the length of the longest abnormal segment in the xth segment, Indicates the total length of the xth segment.

[0039] It represents the ratio of the length of the longest abnormal segment in the x-th segment to the total length of the x-th segment. The larger the ratio, the more obvious the abnormal continuity of the x-th segment, that is, the higher the abnormal continuity degree of the x-th segment.

[0040] Express Normalization of , the normalization method can be: ,in, It is an exponential function with the natural constant e as its base.

[0041] Then, the severity of the disease progression of the target segment is obtained. In an exemplary embodiment, Figure 4 As shown, a specific acquisition process is given as follows: Step 2-3: Obtain the changing trend of the number of newly sick animals in each segment.

[0042] The number of newly added sick animals per day in the target segment is arranged in chronological order to obtain a sequence of newly added sick animals. A changing trend of the sequence of newly added sick animals in the target segment is determined, and the changing trend represents the changing pattern of the number of newly added sick animals on each day in the target segment. In an exemplary embodiment, a two-dimensional coordinate system is constructed with time as the horizontal axis and the number of newly added sick animals per day as the vertical axis. The number of newly added sick animals on each day in the target segment is mapped to the two-dimensional coordinate system, and a straight line is fitted to the number of newly added sick animals on each day in the two-dimensional coordinate system to obtain the slope of the fitted line. When the slope is positive, it indicates that the number of newly added sick animals in the target segment is increasing, and the larger the slope, the worsening of the disease progression and the higher the severity of the disease progression in the target segment. When the slope is 0, it indicates that the number of newly added sick animals in the target segment is stable. When the slope is negative, it indicates that the number of newly added sick animals in the target segment is decreasing. The slope of the fitted line of the target segment is used as the changing trend of the number of newly added sick animals in the target segment.

[0043] It should be understood that when the slope of the fitted line is 0 or negative, the target segment has not progressed and there is no severity level. In this embodiment, the segment with a slope of 0 or negative is determined to be a normal segment. The normal segment is not further processed. Instead, the data weights of each harmful gas concentration in the normal segment are directly set to the preset constants described below. The segments participating in the data analysis in subsequent steps are all segments with positive slopes.

[0044] Step 2-4: Obtain segmented diseased animal quantity features.

[0045] The number characteristic of sick animals in the target segment is obtained, where the number characteristic of sick animals in the target segment is the average of the total number of sick animals on each day in the target segment.

[0046] Steps 2-5: Integrate the change trend and the number of sick animals to obtain the segmented severity of the disease.

[0047] The higher the trend of change in the number of newly diagnosed sick animals in the target segment, that is, the greater the slope of the fitted straight line of the target segment, the more severe the disease progression and the higher the severity of the disease progression in the target segment; the higher the characteristic of the number of diagnosed sick animals in the target segment, the higher the overall level of the number of diagnosed sick animals in the target segment and the higher the severity of the disease progression in the target segment. Therefore, the trend of change in the target segment and the characteristic of the number of diagnosed sick animals in the target segment are combined to obtain the result as the severity of the disease progression in the target segment. In an exemplary embodiment, a specific quantification process for the severity of the disease progression is given as follows: ; in, Indicates the severity of the disease progression in the xth segment, represents the number of sick animals in the xth segment, Represents the changing trend of the number of newly sick animals in the xth segment.

[0048] Express Normalization of , the normalization method can be: .

[0049] The higher the abnormal continuity of the target segment and the higher the severity of the disease progression, the greater the impact of the harmful gas concentration on the disease progression in the target segment, and the higher the harmful gas impact level in the target segment. Therefore, based on the abnormal continuity and disease progression severity of the target segment, the harmful gas impact level of the target segment is obtained, and the harmful gas impact level is proportional to both the abnormal continuity and disease progression severity. In an exemplary embodiment, the calculation formula for the harmful gas impact level is as follows: ; in, Indicates the degree of harmful gas impact in the xth segment. The normalization here is: Get the The maximum and minimum values ​​in the , and then the maximum and minimum values ​​are normalized to the xth segment Perform normalization.

[0050] The above steps are mainly to determine the degree of harmful gas impact based on the distribution characteristics of harmful gas concentrations within the segments. For segments with similar average values ​​of the total number of sick animals, there will be certain differences in the degree of harmful gas impact. Therefore, it is necessary to further correct the degree of harmful gas impact of each segment based on the differences between different segments to obtain a more accurate degree of harmful gas impact after correction. Then, after obtaining the degree of harmful gas impact and before obtaining the correlation coefficient, that is, before the correlation coefficient acquisition module, the system also includes a harmful gas impact degree correction module. Figure 5As shown in the figure, the specific implementation process of the harmful gas impact degree correction module is as follows: Step 2-6: Cluster each segment according to the total number of sick animals in the segment to obtain multiple clusters.

[0051] The average value of the total number of sick animals in each segment is obtained, and then all segments are clustered based on the average value of the total number of sick animals in each segment to obtain multiple clusters. A K-means clustering algorithm can be used, and the value of the number of clusters K is set according to actual needs. As other implementations, other clustering methods can also be used, such as setting multiple different ranges of total number of sick animals and classifying segments within the same range into one cluster.

[0052] Step 2-7: Based on the length of each segment and the length of the longest segment in the same cluster, obtain the segment length feature of each segment.

[0053] For the target segment, the length of the target segment (i.e., the number of harmful gas concentrations contained therein) is obtained. The cluster in which the target segment is located is determined, the lengths of the segments in the cluster in which the target segment is located are obtained, and the longest segment among them is obtained, thereby obtaining the length of the longest segment.

[0054] According to the length of the target segment and the length of the longest segment in the cluster where the target segment is located, the segment length feature of the target segment is obtained. In an exemplary embodiment, the segment length feature of the target segment is the ratio of the length of the target segment to the length of the longest segment in the cluster where the target segment is located.

[0055] Step 2-8: Obtain the harmful gas impact level characteristics of each segment based on the harmful gas impact level of each segment and the maximum harmful gas impact level in the same cluster.

[0056] Obtain the harmful gas impact levels of each segment in the cluster in which the target segment is located, and then obtain the maximum harmful gas impact level therefrom. Based on the harmful gas impact level of the target segment and the maximum harmful gas impact level in the cluster in which the target segment is located, obtain a harmful gas impact level characteristic of the target segment. In an exemplary embodiment, the harmful gas impact level characteristic of the target segment is the ratio of the harmful gas impact level of the target segment to the maximum harmful gas impact level in the cluster in which the target segment is located.

[0057] Step 2-9: Based on the correction coefficient of each segment, correct the degree of harmful gas impact of each segment.

[0058] The segment length characteristics of the target segment and the harmful gas impact degree characteristics are integrated to obtain the correction coefficient of the target segment. Then, the harmful gas impact degree of the target segment is corrected according to the correction coefficient of the target segment.

[0059] In an exemplary embodiment, a specific quantitative method of correction is given as follows: ; in, Indicates the corrected harmful gas impact level of the xth segment, defined as the relative harmful gas impact level, Table Length of the xth segment; Indicates the length of the longest segment in the cluster where the x-th segment is located, Indicates the maximum harmful gas impact in the cluster where the x-th segment is located.

[0060] It represents the correction coefficient of the x-th segment. The closer the length of the x-th segment is to the length of the longest segment in the cluster where it is located, the greater the relative harmful gas impact of the x-th segment is among the segments with similar average values ​​of the total number of sick animals; the closer the harmful gas impact of the x-th segment is to the maximum harmful gas impact in the cluster where it is located, the greater the relative harmful gas impact of the x-th segment is among the segments with similar average values ​​of the total number of sick animals, that is, The larger the value of , the greater the relative harmful gas impact of the x-th segment.

[0061] The relative harmful gas impact levels of each segment are determined in this way. The harmful gas impact levels mentioned below are all relative harmful gas impact levels.

[0062] The correlation coefficient acquisition module is used to obtain the changing relationship between the daily harmful gas concentration and the number of newly added sick animals in each segment, and obtain the correlation coefficient of each segment.

[0063] For the target segment, the correlation coefficient of the target segment is obtained based on the relationship between the daily concentration of harmful gases and the daily number of newly sick animals in the target segment. The correlation coefficient represents the correlation between the daily concentration of harmful gases and the daily number of newly sick animals. In an exemplary embodiment, Figure 6 As shown, the correlation coefficient acquisition module is specifically used for: Step 3-1: Determine the abnormal degree of harmful gases on each day in the segment, and obtain a segmented abnormal degree sequence of harmful gases.

[0064] Determine the abnormal degree of harmful gases in the target segment each day, where the abnormal degree of harmful gases represents the abnormality of the concentration of harmful gases on that day. In an exemplary embodiment, Figure 7 As shown, a specific process for obtaining the abnormal degree of harmful gases includes: Step 3-1-1: Obtain the number of abnormal harmful gas concentrations per day in the segment, the maximum abnormal degree per day, and the length of the longest abnormal segment per day.

[0065] For any day in the target segment, define it as the target day. Obtain the number of abnormal hazardous gas concentrations on the target day. Obtain the abnormality levels of each abnormal hazardous gas concentration on the target day, and determine the maximum abnormality level. Obtain the abnormal segments on the target day, obtain the lengths of each abnormal segment on the target day, and determine the longest abnormal segment, thereby determining the length of the longest abnormal segment.

[0066] Step 3-1-2: According to the number of abnormal harmful gas concentrations per day, the maximum abnormal degree per day, and the length of the longest abnormal section per day, the abnormal degree of harmful gases per day is obtained.

[0067] The abnormal degree of harmful gas on the target day is obtained based on the number of abnormal harmful gas concentrations on the target day, the maximum abnormal degree on the target day, and the length of the longest abnormal segment on the target day. The greater the number of abnormal harmful gas concentrations, the higher the abnormal degree of harmful gas; the greater the maximum abnormal degree, the higher the abnormal degree of harmful gas; the longer the longest abnormal segment, the higher the abnormal degree of harmful gas. In an exemplary embodiment, a specific quantitative method for the abnormal degree of harmful gas is given as follows: ; in, Indicates the first segment in the xth segment The abnormal level of harmful gases in the day, Indicates the first segment in the xth segment The number of abnormal harmful gas concentrations per day; Indicates the first segment in the xth segment The maximum abnormal degree of the day; Indicates the first segment in the xth segment The length of the longest anomalous segment of the day.

[0068] Through the above process, the daily abnormal degree of harmful gases in the target segment is obtained, and then arranged in time sequence to obtain the abnormal degree sequence of harmful gases in the target segment. Furthermore, the maximum and minimum values ​​in the abnormal degree sequence of harmful gases in the target segment can be obtained, and then the abnormal degree sequence of harmful gases in the target segment is normalized using the maximum and minimum value normalization method.

[0069] Step 3-2: Obtain a sequence of the number of newly infected animals in each segment based on the number of newly infected animals on each day in the segment.

[0070] The number of newly diagnosed sick animals in the target segment is obtained each day and then arranged in chronological order to obtain a sequence of newly diagnosed sick animals in the target segment. Furthermore, the maximum and minimum values ​​in the sequence of newly diagnosed sick animals in the target segment can be obtained, and then the sequence of newly diagnosed sick animals in the target segment can be normalized using a maximum-minimum normalization method.

[0071] Step 3-3: Obtain the correlation coefficient between the segmented harmful gas abnormality degree series and the new number of sick animals series.

[0072] Obtain a correlation coefficient between a target segment's harmful gas anomaly degree sequence and a target segment's newly added sick animal number sequence. In an exemplary embodiment, calculate a Pearson correlation coefficient between the normalized target segment's harmful gas anomaly degree sequence and the normalized target segment's newly added sick animal number sequence. Then, normalize the Pearson correlation coefficient to obtain the correlation coefficient between the target segment's harmful gas anomaly degree sequence and the newly added sick animal number sequence. Since the Pearson correlation coefficient has a numerical range of [-1, 1], the Pearson correlation coefficient can be normalized by calculating the sum of the Pearson correlation coefficient and the value 1, and then dividing the sum by 2.

[0073] When the values ​​of the harmful gas abnormality series in the target segment gradually increase, it means that the harmful gas is becoming more abnormal. If the values ​​of the newly sick animal number series also gradually increase, it means that the number of newly sick animals is gradually increasing, which further indicates that respiratory diseases are mainly caused by abnormal harmful gas concentrations. Using the above method, the correlation coefficient of each segment is obtained.

[0074] The data weight acquisition module is used to fuse the harmful gas impact degree and correlation coefficient of each segment to obtain the data weight of each harmful gas concentration in each segment.

[0075] The above respectively obtains the harmful gas impact degree and correlation coefficient of each segment. These two types of data information affect the data weight of each harmful gas concentration in each segment. Then, by integrating these two types of data information, the data weight of each harmful gas concentration in each segment is obtained.

[0076] In an exemplary embodiment, Figure 8 As shown, the data weight acquisition module is specifically used to: Step 4-1: Integrate the harmful gas impact degree and correlation coefficient of each segment to obtain the importance of the harmful gas in each segment.

[0077] The higher the relative harmful gas impact of the target segment, the more important the harmful gas concentration of the target segment is in subsequent predictions for the target segment, and the higher the importance of the harmful gas in the target segment; the higher the correlation coefficient of the target segment, the closer the relationship between the harmful gas concentration in the target segment and the number of newly added sick animals, which means the more important the harmful gas concentration of the target segment is in subsequent predictions, and the higher the importance of the harmful gas in the target segment. The relative harmful gas impact and correlation coefficient of the target segment are fused to obtain the importance of the harmful gas in the target segment. In an exemplary embodiment, a specific fusion method is given as follows: ; in, Indicates the importance of harmful gases in the xth segment; represents the correlation coefficient of the xth segment.

[0078] Step 4-2: According to the importance of harmful gases in each segment and the number of abnormal harmful gas concentrations, the data weight of each abnormal harmful gas concentration in each segment is obtained.

[0079] Determine the importance of each segment's harmful gas concentration in the prediction based on step 4-1. Each segment contains both normal and abnormal harmful gas concentrations. When using the ARIMA algorithm for disease prediction, different data weights should be assigned to normal and abnormal harmful gas concentrations. Based on the importance of the harmful gas in the target segment and the number of abnormal harmful gas concentrations in the target segment, the data weight for each abnormal harmful gas concentration in the target segment is determined.

[0080] In an exemplary embodiment, the quantification method of the data weight of each abnormal harmful gas concentration is given as follows: ; in, Indicates the data weight of the yth abnormal harmful gas concentration in the xth segment, Indicates the number of abnormal harmful gas concentrations in the xth segment.

[0081] Indicates the importance of the data of the yth abnormal harmful gas concentration in the xth segment. The normalization method is: obtain the maximum and minimum values ​​of the number of abnormal harmful gas concentrations in each segment, and then use the maximum and minimum value normalization method to normalize the Normalization is performed. The numerical range of the data importance of each abnormal harmful gas concentration in the target segment is 0-1. By adding the value 1, the numerical range of the data weight of each abnormal harmful gas concentration is limited to the numerical range of 1-2. This results in the data weight of each abnormal harmful gas concentration in the target segment. From the above calculation, it can be seen that the data weight of each abnormal harmful gas concentration in the target segment is the same, that is, the data weight of each abnormal harmful gas concentration in the same segment is the same.

[0082] Then, the data weights for all hazardous gas concentrations in the target segment, excluding the abnormal hazardous gas concentrations (i.e., the normal hazardous gas concentrations in the target segment), are assigned a preset constant that is smaller than the data weight for the abnormal hazardous gas concentrations in the target segment. Because normal hazardous gas concentrations are less important than abnormal hazardous gas concentrations during the prediction process, the data weights for normal hazardous gas concentrations are also lower. In one exemplary embodiment, the preset constant is 1, meaning that the data weight for normal hazardous gas concentrations in the target segment is 1.

[0083] This results in the data weights of each harmful gas concentration in the target segment, and then the data weights of each harmful gas concentration in each segment. This determines the data weights of each harmful gas concentration within a historical time period. Based on each harmful gas concentration and its data weight, the autoregressive integrated moving average (ARIMA) model is used to predict future harmful gas concentrations in the farm. Based on the predicted harmful gas concentrations, anomalies in the farm environment can be detected in advance. When the predicted harmful gas concentration is too high (i.e., greater than the preset concentration threshold), it may indicate a deterioration in the farm environment and an increased risk of animal respiratory diseases and other diseases. Preemptive measures can be taken, such as improving ventilation, strengthening disinfection, increasing the frequency of manure cleaning, and adjusting stocking density, to create a good breeding environment, reduce the opportunities for pathogen breeding and spread at the source, and prevent and reduce the incidence of diseases.

[0084] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0085] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. An intelligent animal husbandry and veterinary epidemic prevention system for remote areas, characterized by: include: A segmentation module, for segmenting the historical time period according to the change in the total number of sick animals per day in the historical time period of the farm; The harmful gas impact degree acquisition module is used to obtain the harmful gas impact degree of each segment based on the abnormal concentration of harmful gases in the farm and the severity of the disease in each segment; The correlation coefficient acquisition module is used to obtain the changing relationship between the daily concentration of harmful gases and the number of newly sick animals in each segment, and obtain the correlation coefficient of each segment; The data weight acquisition module is used to fuse the harmful gas impact degree and correlation coefficient of each segment to obtain the data weight of each harmful gas concentration in each segment.

2. The intelligent animal husbandry and veterinary epidemic prevention system for remote areas as claimed in claim 1 is characterized in that: Before the correlation coefficient acquisition module, the system further includes a harmful gas impact degree correction module, which is used to: Each segment is clustered according to the total number of sick animals in the segment to obtain multiple clusters; According to the length of each segment and the length of the longest segment in the same cluster, the segment length feature of each segment is obtained; According to the harmful gas impact degree of each segment and the maximum harmful gas impact degree in the same cluster, the harmful gas impact degree characteristics of each segment are obtained; Based on the correction coefficient of each segment, the harmful gas impact degree of each segment is corrected; the correction coefficient is obtained by fusing the segment length characteristics and the harmful gas impact degree characteristics.

3. The intelligent animal husbandry and veterinary epidemic prevention system for remote areas as claimed in claim 1 is characterized in that: The correlation coefficient acquisition module is specifically used for: Determine the abnormal degree of harmful gas in each segment every day, and obtain a segmented abnormal degree sequence of harmful gas; The number of newly added sick animals in each segment is used to obtain the sequence of the number of newly added sick animals in each segment on a daily basis; Obtain the correlation coefficient between the segmented harmful gas abnormality degree series and the new number of sick animals series.

4. The intelligent animal husbandry and veterinary epidemic prevention system for remote areas as claimed in claim 3 is characterized in that: The process of obtaining the abnormal degree of harmful gas includes: Obtaining the number of abnormal harmful gas concentrations per day in the segment, the maximum abnormal degree per day, and the length of the longest abnormal segment per day; the abnormal harmful gas concentration is a harmful gas concentration greater than a preset normal harmful gas concentration threshold, and the abnormal degree is the degree to which the harmful gas concentration exceeds the preset normal harmful gas concentration threshold; the abnormal segment is obtained from continuous abnormal harmful gas concentrations; The daily abnormal degree of harmful gas is obtained according to the number of abnormal harmful gas concentrations per day, the maximum abnormal degree per day, and the length of the longest abnormal section per day.

5. The intelligent animal husbandry and veterinary epidemic prevention system for remote areas as claimed in claim 1 is characterized in that: The data weight acquisition module is specifically used to: The importance of harmful gases in each segment is obtained by integrating the harmful gas impact degree and correlation coefficient of each segment; According to the importance of harmful gases in each segment and the number of abnormal harmful gas concentrations, the data weight of each abnormal harmful gas concentration in each segment is obtained; the abnormal harmful gas concentration is a harmful gas concentration that is greater than a preset harmful gas normal concentration threshold; the data weights of the abnormal harmful gas concentrations in the same segment are the same.

6. The intelligent animal husbandry and veterinary epidemic prevention system for remote areas as claimed in claim 5 is characterized in that: The data weight of each harmful gas concentration other than the abnormal harmful gas concentration in each segment is a preset constant, and the preset constant is smaller than the data weight of the abnormal harmful gas concentration in each segment.

7. The intelligent animal husbandry and veterinary epidemic prevention system for remote areas as claimed in claim 1 is characterized in that: The process of obtaining the abnormal situation of the harmful gas concentration includes: Acquire the abnormal harmful gas concentration in the segment, and obtain the abnormal segment from the continuous abnormal harmful gas concentration; the abnormal harmful gas concentration is the harmful gas concentration greater than the preset harmful gas normal concentration threshold; The abnormal continuity degree of the segment is obtained according to the number of abnormal segments in the segment and the length of the longest abnormal segment in the segment; the abnormal continuity degree is proportional to the number of abnormal segments and the length of the longest abnormal segment.

8. The intelligent animal husbandry and veterinary epidemic prevention system for remote areas as claimed in claim 7 is characterized in that: The degree of influence of the harmful gas is obtained from the abnormal continuity and the severity of the disease progression, and the degree of influence of the harmful gas is proportional to both the abnormal continuity and the severity of the disease progression.

9. The intelligent animal husbandry and veterinary epidemic prevention system for remote areas as claimed in claim 1 is characterized in that: The process of obtaining the severity of the disease progression includes: Obtaining a changing trend of the number of newly added sick animals in each segment, where the changing trend represents a changing pattern of the number of newly added sick animals on each day in the segment; Obtaining a segmented number characteristic of sick animals, wherein the number characteristic of sick animals is the average of the total number of sick animals on each day in the segment; By integrating the change trend and the number characteristics of sick animals, the severity of the disease progression in segments is obtained.

10. The intelligent animal husbandry and veterinary epidemic prevention system for remote areas as claimed in claim 1, characterized in that: The segmentation module is specifically used to segment the historical time period using a mean segmentation algorithm according to the change in the total number of sick animals per day in the historical time period of the farm.

Images