Common animal disease early warning method and system based on breeding environment detection data

Abstract

The application discloses a common animal disease early warning method and system based on breeding environment detection data, and relates to the technical field of animal disease early warning. The method comprises the following steps: interacting with a target breeding environment, obtaining breeding environment detection data and neighborhood environment data; determining breeding object information of the target breeding environment, calling a pre-constructed disease classifier for qualitative classification analysis, and outputting a result as a potential disease set; determining a risk accumulation rate set of M environment indexes in the breeding environment detection data for the potential disease set according to historical disease data; generating predicted environment sequence data of the target breeding environment according to the neighborhood environment data and the breeding environment detection data, performing risk accumulation calculation on the potential disease set, and obtaining an accumulated disease risk of each potential disease; generating an animal disease early warning report based on the multiple accumulated disease risks, and performing early warning response. The application solves the technical problem of inaccurate animal disease early warning in the prior art.

Description

Technical Field

[0001] This invention relates to the field of animal disease early warning technology, and specifically to common animal disease early warning methods and systems based on monitoring data of the breeding environment. Background Technology

[0002] In modern livestock, poultry, and aquaculture industries, environmental factors play a crucial role in the health and disease occurrence of common animals. Abnormal fluctuations in environmental indicators such as temperature, humidity, light, oxygen concentration, and carbon dioxide concentration are often direct causes or key risk factors that induce or exacerbate specific animal diseases.

[0003] However, existing technologies can only trigger a general alarm when a single health indicator of a common farmed animal exceeds a preset safety range. They cannot accurately predict diseases in common animals, making it difficult to avoid disease risks in the farming of common animals, resulting in high farming risks and costs. Summary of the Invention

[0004] This application provides a method and system for early warning of common animal diseases based on aquaculture environment monitoring data, which is used to address the technical problem that the existing technology cannot combine aquaculture environment data for early warning, resulting in poor early warning effect for common animal diseases.

[0005] In view of the above problems, this application provides a method and system for early warning of common animal diseases based on monitoring data of the breeding environment.

[0006] Firstly, this application provides a method for early warning of common animal diseases based on breeding environment monitoring data, the method comprising:

[0007] Interact with the target aquaculture environment to obtain aquaculture environment monitoring data and neighborhood environment data.

[0008] The information of the aquaculture objects in the target aquaculture environment is determined, and combined with the aquaculture object information and the detection data of the aquaculture environment, a pre-constructed disease classifier is called to perform qualitative classification analysis, and the qualitative classification analysis results are output as a potential disease set.

[0009] Based on historical disease data, determine the risk accumulation rate set of M environmental indicators in the aquaculture environment monitoring data for the potential disease set, where M is greater than or equal to 2.

[0010] Based on the neighborhood environmental data and the aquaculture environment monitoring data, predictive environmental sequence data of the target aquaculture environment is generated, and risk accumulation calculation is performed on the potential disease set based on the aquaculture environment monitoring data and the risk accumulation rate set to obtain the cumulative disease risk of each potential disease.

[0011] Animal disease early warning reports are generated based on multiple cumulative disease risks, and early warning responses are executed.

[0012] Secondly, this application provides a common animal disease early warning system based on breeding environment monitoring data, including:

[0013] The environmental information acquisition module is used to interact with the target aquaculture environment and acquire aquaculture environment monitoring data and neighboring environment data.

[0014] The qualitative classification analysis module is used to determine the information of the aquaculture objects in the target aquaculture environment, and combine the aquaculture object information with the aquaculture environment detection data to call a pre-built disease classifier to perform qualitative classification analysis, and output the qualitative classification analysis results as a potential disease set.

[0015] The environmental risk acquisition module is used to determine the risk accumulation rate set of M environmental indicators in the aquaculture environment monitoring data toward the potential disease set based on historical disease data, where M is greater than or equal to 2.

[0016] The risk accumulation calculation module is used to generate predicted environmental sequence data of the target aquaculture environment based on the neighborhood environmental data and the aquaculture environment detection data, and to perform risk accumulation calculation on the potential disease set based on the aquaculture environment detection data and the risk accumulation rate set to obtain the cumulative disease risk of each potential disease.

[0017] The early warning response module is used to generate animal disease early warning reports based on multiple cumulative disease risks and to execute early warning responses.

[0018] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0019] This application proposes a method and system for early warning of common animal diseases based on aquaculture environment monitoring data. By integrating real-time environmental monitoring data, neighborhood environmental information, characteristics of farmed animals, and historical disease occurrence patterns, a dynamic and accurate early warning system for common animal diseases is constructed. This achieves the technical effect of improving the accuracy of early warning for common animal diseases, which is conducive to risk control for potential diseases when raising common animals, and guides farmers to take targeted disease prevention and control measures in advance to reduce the risk of disease outbreaks. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0021] Figure 1 This is a flowchart illustrating a common animal disease early warning method based on aquaculture environment monitoring data, provided in an embodiment of this application.

[0022] Figure 2 This is a schematic diagram of the structure of a common animal disease early warning system based on aquaculture environment monitoring data provided in an embodiment of this application.

[0023] The components represented by each number in the attached diagram are explained below:

[0024] The system includes an environmental information acquisition module (100), a qualitative classification analysis module (200), an environmental risk acquisition module (300), a risk accumulation calculation module (400), and an early warning response module (500). Detailed Implementation

[0025] This application provides a method and system for early warning of common animal diseases based on aquaculture environment monitoring data, which addresses the technical problem that the existing technology cannot combine aquaculture environment data for early warning, resulting in poor early warning effects for common animal diseases.

[0026] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0027] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0028] Example 1, as Figure 1 As shown, this application provides a method for early warning of common animal diseases based on breeding environment monitoring data, wherein the method for early warning of common animal diseases based on breeding environment monitoring data includes:

[0029] S10: Interact with the target aquaculture environment to obtain aquaculture environment monitoring data and neighborhood environment data.

[0030] Current monitoring of aquaculture environments is mainly limited to real-time data collection within a single aquaculture environment. It lacks sufficient consideration of the spatial correlation and dynamic propagation of environmental factors, resulting in the isolation of environmental factors and making it difficult to accurately characterize changes.

[0031] Step S10 in the method provided in this application embodiment includes:

[0032] It interacts with multi-source sensors deployed in the target aquaculture environment to collect real-time detection data from the target aquaculture environment.

[0033] Based on a preset time window and preset geographical boundaries, it interacts with external data sources to obtain neighborhood environmental data.

[0034] Environmental information about the breeding environment is a crucial factor affecting the health of farmed animals. In this embodiment, information is exchanged with multi-source sensors deployed in the target breeding environment to collect real-time detection data such as temperature, humidity, and oxygen concentration.

[0035] Based on a preset time window and a preset geographical boundary, neighboring environmental data is acquired through interactive external data sources. The preset time window and geographical boundary serve as constraints on the neighboring environment of the target aquaculture environment; a larger time window and a farther geographical boundary result in a larger neighboring environment. By integrating this data with extended neighboring environmental data, such as weather and air quality, the overall conditions of the aquaculture environment can be more accurately assessed. Extended neighboring environmental data has transitivity and predictive significance; for example, changes in external weather conditions may impact the aquaculture environment, thereby affecting animal health. However, an excessively large time window and a too-far-reaching geographical boundary may introduce overly broad and distant neighboring environmental data, diluting the transitivity of changes in the aquaculture environment. Therefore, for example, the preset time window is set to 3 days, and the preset geographical boundary is set to 2 km outward from the boundary of the aquaculture environment.

[0036] By simultaneously acquiring real-time monitoring data of the target aquaculture environment itself and neighboring environmental data within a preset geographical boundary, the limitations of traditional single-environment monitoring are significantly overcome, laying a complete data foundation for subsequent analysis.

[0037] S20: Determine the information of the aquaculture objects in the target aquaculture environment, and combine the aquaculture object information with the aquaculture environment detection data, call the pre-built disease classifier to perform qualitative classification analysis, and output the qualitative classification analysis results as a potential disease set.

[0038] Existing technologies generally suffer from insufficient specificity when processing environmental data and linking it to diseases. They fail to fully consider the specific characteristics of farmed species, resulting in overly general and broad early warning information and risk assessments that remain at a broad and imprecise level.

[0039] Step S20 in the method provided in this application embodiment includes:

[0040] Interactively obtain information about the aquaculture objects in the target aquaculture environment.

[0041] Based on the information of the aquaculture objects, feature extraction is performed to obtain category features, variety features, and stage features, and the output is the aquaculture object features.

[0042] Based on the characteristics of the aquaculture objects, the corresponding qualitative classification channel is activated in the pre-constructed disease classifier, and the aquaculture object information and the aquaculture environment detection data are input for qualitative classification analysis to obtain the qualitative classification analysis results.

[0043] This includes determining the information of the aquaculture objects in the target aquaculture environment, and combining the aquaculture object information with the aquaculture environment monitoring data, calling a pre-constructed disease classifier for qualitative classification analysis. Prior to this, the process includes:

[0044] Based on the historical disease data, multiple qualitative classification branches are trained corresponding to various types of aquaculture objects. Each qualitative classification branch includes multiple qualitative classification branch channels, and each qualitative classification branch channel corresponds to a certain aquaculture period of the aquaculture object corresponding to the qualitative classification branch.

[0045] By integrating multiple qualitative classification branches through the classification layer of farmed objects, the output becomes the disease classifier.

[0046] The output qualitative classification analysis results are a set of potential diseases, including:

[0047] Obtain the category probability corresponding to each potential disease in the qualitative classification analysis results, and output it as a category probability set.

[0048] Based on the preset significance constraints, the significance is initially screened by traversing the probability set of the categories.

[0049] Based on the initial significance screening results, the qualitative classification analysis results are cleaned, and the cleaned qualitative classification analysis results are output as the potential disease set, wherein the potential disease set includes N potential diseases, where N is greater than or equal to 1.

[0050] In this embodiment of the application, information about the aquaculture objects in the target aquaculture environment is obtained through aquaculture logs.

[0051] Based on the information of the breeding objects, feature extraction is performed to obtain category features such as cattle and sheep, breed features such as Chinese Holstein dairy cattle and small-tailed Han sheep, and stage features such as juvenile stage and fattening stage, and the output is the breeding object features.

[0052] Based on historical aquaculture logs, historical disease data of the aquaculture species were collected and integrated.

[0053] Based on historical disease data, multiple qualitative classification branches are trained for various types of farmed animals. Each qualitative classification branch includes multiple qualitative classification channels, and each channel corresponds to a specific farming period for that particular animal. For example, a decision tree is used, with an independent decision tree classifier trained for each farming period of each animal type. For instance, a qualitative classification channel is constructed for young Chinese Holstein dairy cows using a single-layer decision tree with a maximum depth of 5 and the Gini coefficient as the splitting criterion. The input information includes environmental data and farmed animal information, and the output information includes multiple potential diseases and their probabilities.

[0054] By using a classification layer based on the livestock species, multiple qualitative classification branches are integrated. Specifically, these branches are categorized and integrated according to the livestock species. For example, the qualitative classification branches related to Chinese Holstein dairy cows are integrated into a Chinese Holstein dairy cow disease classifier. The output is the integrated disease classifier.

[0055] Based on the characteristics of the farmed animals, the corresponding qualitative classification channels are activated in a pre-constructed disease classifier. For example, for Chinese Holstein dairy cows, the Chinese Holstein dairy cow disease classifier is activated. Inputting farmed animal information and environmental monitoring data, qualitative classification analysis is performed to obtain the results. The qualitative classification analysis results include various potential diseases of Chinese Holstein dairy cows and their probabilities.

[0056] Obtain the category probability corresponding to each potential disease in the qualitative classification analysis results, and output it as a category probability set.

[0057] Based on preset significance constraints, a preliminary significance screening is performed by traversing the category probability set. The significance constraints can be preset by statistically testing potential diseases with high occurrence probabilities in historical data, ensuring that the selected disease types have a high probability of occurrence. For example, the preset significance constraint is set to 30% by statistically analyzing historical data. The preliminary significance screening is then performed by traversing the category probability set to identify disease types with probabilities greater than or equal to the significance constraint.

[0058] Based on the initial significance screening results, the qualitative classification analysis results are cleaned. Disease types with probabilities less than the significance constraint are removed. The cleaned qualitative classification analysis results are then output as potential diseases. The potential disease set includes N potential diseases, where N is greater than or equal to 1.

[0059] By combining the characteristic information of the aquaculture species and activating the corresponding channels of the disease classifier for qualitative analysis, the pertinence and specificity of the early warning were significantly improved, laying a core data foundation for subsequent early warning.

[0060] S30: Based on historical disease data, determine the risk accumulation rate set of M environmental indicators in the aquaculture environment monitoring data for the potential disease set, where M is greater than or equal to 2.

[0061] Existing technologies cannot consider the complex correlation between multiple environmental indicators and multiple disease types, lack dynamic quantification capabilities, and cannot address the cumulative effects of different environmental conditions on the risk of specific diseases over time, resulting in risk assessment remaining at the level of static threshold alarms.

[0062] Step S30 in the method provided in this application embodiment includes:

[0063] Based on the characteristics of the aquaculture species, the original historical disease data and homologous historical disease data of the target aquaculture environment are collected and output as the historical disease data.

[0064] Using the set of potential diseases as an index, the historical disease data is traversed and classified and filtered to obtain historical data of potential diseases, wherein the historical data of potential diseases includes N historical data clusters corresponding to N potential diseases.

[0065] M environmental indicators from the aquaculture environment monitoring data are extracted as the analysis object set, and risk accumulation modeling is performed on N historical data clusters respectively to obtain the risk accumulation rate model of the M environmental indicators, and the output is the risk accumulation rate set.

[0066] In this embodiment of the application, based on the characteristics of the farmed species, original historical disease data and homologous historical disease data of the target farming environment are collected and integrated to output historical disease data. The original historical disease data refers to historical disease data from the historical data of this farming environment, while the homologous historical disease data refers to historical disease data from other farming environments with the same farmed species characteristics, such as other farming disease databases.

[0067] Using the potential disease set as an index, the historical disease data is traversed, classified and filtered to obtain the historical data of potential diseases. The historical data of potential diseases includes N historical data clusters corresponding to N potential diseases. Each historical data cluster contains one potential disease type and multiple aquaculture environment indicators.

[0068] M environmental indicators from aquaculture environmental monitoring data are extracted as the analysis object set, and risk accumulation modeling is performed on N historical data clusters. For example, PCA processing is first performed on the historical environmental monitoring data to identify and analyze the main characteristic changes of key environmental patterns, i.e., the trend of environmental pattern changes that coexist before the occurrence of multiple diseases. For instance, oxygen concentration data for two consecutive weeks before a disease outbreak is collected, and the disease type and outbreak time are labeled as sample environmental monitoring data and sample risk accumulation rate. Machine learning methods such as neural networks are used for risk accumulation modeling. The model adopts a three-layer structure: the input layer receives the environmental monitoring data, the hidden layer has 64 nodes activated by the ReLU function, and the output layer outputs the risk accumulation rate. The model is trained in a supervised manner using sample environmental monitoring data and sample risk accumulation rate, and the parameters are continuously adjusted to achieve model convergence. If the accuracy of the output risk accumulation rate is above 90% when inputting environmental monitoring data, the risk accumulation rate model training is considered complete. The risk accumulation rate models of the M environmental indicators are obtained, and the integrated output is a risk accumulation rate set.

[0069] This application not only considers historical disease data in the current aquaculture environment, but also introduces homologous historical disease data for more accurate risk assessment. Through risk accumulation modeling, it realizes the quantification process of risk indicators under different environmental factors, providing accurate data support for subsequent risk analysis.

[0070] S40: Based on the neighborhood environment data and the aquaculture environment monitoring data, generate predicted environmental sequence data for the target aquaculture environment, and perform risk accumulation calculation on the potential disease set based on the aquaculture environment monitoring data and the risk accumulation rate set to obtain the cumulative disease risk of each potential disease.

[0071] Traditional risk assessment methods are often static risk assessments, which have inherent limitations due to insufficient dynamic foresight. They rely solely on the current or historical environment for judgment and cannot simulate the risk development trends caused by future environmental evolution.

[0072] Step S40 in the method provided in this application embodiment includes:

[0073] Starting with the aquaculture environment monitoring data, based on the sliding window method, step-by-step environment prediction is performed according to the neighborhood environment data to obtain the predicted environment sequence data, wherein the predicted environment sequence data includes M predicted environment indicator sequences.

[0074] A first predicted environment indicator sequence is randomly extracted from the predicted environment sequence data. Based on the risk accumulation rate set, the first predicted environment indicator sequence is mapped to N potential disease-predicted risk accumulation rate sequences and fitted accordingly to generate N potential disease-predicted risk accumulation rate curves.

[0075] Traverse the predicted environment sequence data to obtain M sets of cumulative rate curves corresponding to M predicted environment indicator sequences.

[0076] Based on N potential diseases, the cumulative rate curves of the M groups are integrated in groups, and the grouped integration results are summed to obtain the cumulative disease risk of the N potential diseases.

[0077] Specifically, based on N potential diseases, the cumulative rate curves of the M groups are grouped and integrated, and the grouped integrated results are summed to obtain the cumulative disease risk of the N potential diseases, including:

[0078] Using N potential diseases as indices, perform grouping reconstruction on the M groups of cumulative rate curves to obtain N groups of potential disease-cumulative rate curves.

[0079] By iterating through N sets of potential disease-cumulative rate curves and performing integral calculations, the potential disease-cumulative disease risk of N sets can be obtained.

[0080] The cumulative disease risk of each of the N potential diseases is obtained by summing the risks within each group.

[0081] In this embodiment, starting with aquaculture environment monitoring data, a sliding window method is used to perform step-by-step environmental prediction based on neighboring environmental data to obtain predicted environmental sequence data. For example, the step size is set to 1 hour. Using aquaculture environment monitoring data as the starting data, a linear regression model is employed to perform step-by-step environmental prediction based on neighboring environmental data. The model uses a single-layer structure, inputting aquaculture environment monitoring data and outputting predicted environmental data for the next step, integrating these to obtain the predicted environmental sequence data. The sliding window method is used to filter out environmental sequence data with trends consistent with those of the neighboring environmental data, integrating these to obtain the predicted environmental sequence data. The predicted environmental sequence data includes M predicted environmental indicator sequences, such as temperature, humidity, and oxygen concentration sequences. Comprehensive environmental indicator predictions are more beneficial for subsequent data analysis.

[0082] A first predicted environmental indicator sequence is randomly extracted from the predicted environmental sequence data. Based on the risk accumulation rate set, the first predicted environmental indicator sequence is mapped to N potential disease-predicted risk accumulation rate sequences and fitted accordingly to generate N potential disease-predicted risk accumulation rate curves. For example, an environmental indicator such as temperature from the first predicted environmental indicator sequence is retrieved from the risk accumulation rate set, and corresponding potential disease and risk accumulation rate data are extracted. Regression analysis is then used to fit these corresponding potential disease and risk accumulation rate data to generate a potential disease-predicted risk accumulation rate curve for the temperature indicator. Further, other environmental indicators such as humidity and oxygen concentration at each time point in the first predicted environmental indicator sequence are retrieved, extracted, and fitted using the same steps described above to generate multiple potential disease-predicted risk accumulation rate curves. These potential disease-predicted risk accumulation rate curves are then integrated to obtain N potential disease-predicted risk accumulation rate curves.

[0083] Traverse the predicted environmental sequence data to obtain M sets of cumulative rate curves corresponding to M predicted environmental indicator sequences. Each set of cumulative rate curves contains N potential disease-predicted risk cumulative rate curves.

[0084] Using N potential diseases as indices, a grouping reconstruction is performed on M groups of cumulative rate curves to obtain N groups of potential disease-cumulative rate curves. Each group of potential disease-cumulative rate curves contains the same potential disease; for example, all potential diseases in a group of potential disease-cumulative rate curves are bovine brucellosis. The reconstructed curve groups better characterize the risk of the same potential disease, which is beneficial for analyzing the risk of the same potential disease.

[0085] By iterating through N sets of potential disease-cumulative rate curves and performing integral calculations, the potential disease-cumulative disease risk of N sets can be obtained. The integral calculation can yield the area of ​​the potential disease-cumulative rate curve, which can represent the cumulative situation of the potential disease's cumulative rate, thus obtaining the potential disease-cumulative disease risk data.

[0086] The cumulative disease risk of N potential diseases is obtained by summing the cumulative disease risks within each group. Intra-group summation ensures that the cumulative disease risk of potential diseases contributes to the calculation of the cumulative disease risk under each predicted environmental sequence data condition; that is, the case of each predicted environmental sequence data condition is considered.

[0087] By integrating neighboring environmental data with real-time monitoring data within the farm, this application predicts and generates future environmental sequence data for the target aquaculture environment. This predicted environmental sequence data is then mapped and fitted to a cumulative risk rate curve corresponding to each potential disease. Through grouped integration of these curves, the cumulative disease risk that each potential disease may reach under the predicted environmental conditions is dynamically quantified. This application combines environmental prediction with a risk accumulation quantification model, achieving a leap from static risk assessment to dynamic predictive risk accumulation quantification, providing a core basis for issuing timely and accurate early warnings.

[0088] S50: Generate animal disease early warning reports based on multiple cumulative disease risks and execute early warning responses.

[0089] Existing animal disease early warning systems are often too general and cannot guide farmers to take accurate risk intervention measures.

[0090] For example, an animal disease early warning report is generated based on the cumulative calculation results of disease risk and in conjunction with preset graded early warning rules. For instance, the risk levels can be preset to be divided into low risk (cumulative disease risk less than 10%), medium risk (cumulative disease risk greater than or equal to 10% and less than 50%), and high risk (cumulative disease risk greater than or equal to 50%).

[0091] Early warning reports may include the following:

[0092] Disease Type: Lists all disease types in the potential disease set and displays the predicted probability and corresponding cumulative risk value for each disease.

[0093] Disease risk level: Based on the cumulative risk value of each disease, it is classified into low risk, medium risk or high risk to help farm managers understand the current risk situation.

[0094] Early warning recommendations: Based on the risk level of each disease type, corresponding prevention and control recommendations are provided. For example, if the risk level of a certain disease is high, it is recommended to immediately strengthen environmental monitoring, adjust breeding conditions, and increase animal health checks.

[0095] Accurate and effective early warning information can provide more precise guidance for aquaculture, enabling farmers to intervene in advance based on the warning information and reduce the risk of disease outbreaks.

[0096] Example 2, as Figure 2 As shown, based on the same inventive concept as the common animal disease early warning method based on breeding environment monitoring data provided in Embodiment 1, this embodiment of the invention also provides a common animal disease early warning system based on breeding environment monitoring data, including:

[0097] The environmental information acquisition module 100 is used to interact with the target aquaculture environment and acquire aquaculture environment monitoring data and neighborhood environment data.

[0098] The qualitative classification analysis module 200 is used to determine the information of the aquaculture objects in the target aquaculture environment, and combine the aquaculture object information with the aquaculture environment detection data to call a pre-built disease classifier to perform qualitative classification analysis, and output the qualitative classification analysis results as a potential disease set.

[0099] The environmental risk acquisition module 300 is used to determine the risk accumulation rate set of M environmental indicators in the aquaculture environment monitoring data toward the potential disease set based on historical disease data, wherein M is greater than or equal to 2.

[0100] The risk accumulation calculation module 400 is used to generate predicted environmental sequence data of the target aquaculture environment based on the neighborhood environmental data and the aquaculture environment detection data, and to perform risk accumulation calculation on the potential disease set based on the aquaculture environment detection data and the risk accumulation rate set to obtain the cumulative disease risk of each potential disease.

[0101] The early warning response module 500 is used to generate animal disease early warning reports based on multiple cumulative disease risks and to execute early warning responses.

[0102] In one embodiment, the environmental information acquisition module 100 is further configured to:

[0103] It interacts with multi-source sensors deployed in the target aquaculture environment to collect real-time detection data from the target aquaculture environment.

[0104] Based on a preset time window and preset geographical boundaries, it interacts with external data sources to obtain neighborhood environmental data.

[0105] In one embodiment, the qualitative classification analysis module 200 is further configured to:

[0106] Interactively obtain information about the aquaculture objects in the target aquaculture environment.

[0107] Based on the information of the aquaculture objects, feature extraction is performed to obtain category features, variety features, and stage features, and the output is the aquaculture object features.

[0108] Based on the characteristics of the aquaculture objects, the corresponding qualitative classification channel is activated in the pre-constructed disease classifier, and the aquaculture object information and the aquaculture environment detection data are input for qualitative classification analysis to obtain the qualitative classification analysis results.

[0109] This includes determining the information of the aquaculture objects in the target aquaculture environment, and combining the aquaculture object information with the aquaculture environment monitoring data, calling a pre-constructed disease classifier for qualitative classification analysis. Prior to this, the process includes:

[0110] Based on the historical disease data, multiple qualitative classification branches are trained corresponding to various types of aquaculture objects. Each qualitative classification branch includes multiple qualitative classification branch channels, and each qualitative classification branch channel corresponds to a certain aquaculture period of the aquaculture object corresponding to the qualitative classification branch.

[0111] By integrating multiple qualitative classification branches through the classification layer of farmed objects, the output becomes the disease classifier.

[0112] The output qualitative classification analysis results are a set of potential diseases, including:

[0113] Obtain the category probability corresponding to each potential disease in the qualitative classification analysis results, and output it as a category probability set.

[0114] Based on the preset significance constraints, the significance is initially screened by traversing the probability set of the categories.

[0115] Based on the initial significance screening results, the qualitative classification analysis results are cleaned, and the cleaned qualitative classification analysis results are output as the potential disease set, wherein the potential disease set includes N potential diseases, where N is greater than or equal to 1.

[0116] In one embodiment, the environmental risk acquisition module 300 is further configured to:

[0117] Based on the characteristics of the aquaculture species, the original historical disease data and homologous historical disease data of the target aquaculture environment are collected and output as the historical disease data.

[0118] Using the set of potential diseases as an index, the historical disease data is traversed and classified and filtered to obtain historical data of potential diseases, wherein the historical data of potential diseases includes N historical data clusters corresponding to N potential diseases.

[0119] M environmental indicators from the aquaculture environment monitoring data are extracted as the analysis object set, and risk accumulation modeling is performed on N historical data clusters respectively to obtain the risk accumulation rate model of the M environmental indicators, and the output is the risk accumulation rate set.

[0120] In one embodiment, the risk accumulation calculation module 400 is further configured to:

[0121] Starting with the aquaculture environment monitoring data, based on the sliding window method, step-by-step environment prediction is performed according to the neighborhood environment data to obtain the predicted environment sequence data, wherein the predicted environment sequence data includes M predicted environment indicator sequences.

[0122] A first predicted environment indicator sequence is randomly extracted from the predicted environment sequence data. Based on the risk accumulation rate set, the first predicted environment indicator sequence is mapped to N potential disease-predicted risk accumulation rate sequences and fitted accordingly to generate N potential disease-predicted risk accumulation rate curves.

[0123] Traverse the predicted environment sequence data to obtain M sets of cumulative rate curves corresponding to M predicted environment indicator sequences.

[0124] Based on N potential diseases, the cumulative rate curves of the M groups are integrated in groups, and the grouped integration results are summed to obtain the cumulative disease risk of the N potential diseases.

[0125] Specifically, based on N potential diseases, the cumulative rate curves of the M groups are grouped and integrated, and the grouped integrated results are summed to obtain the cumulative disease risk of the N potential diseases, including:

[0126] Using N potential diseases as indices, perform grouping reconstruction on the M groups of cumulative rate curves to obtain N groups of potential disease-cumulative rate curves.

[0127] By iterating through N sets of potential disease-cumulative rate curves and performing integral calculations, the potential disease-cumulative disease risk of N sets can be obtained.

[0128] The cumulative disease risk of each of the N potential diseases is obtained by summing the risks within each group.

[0129] In summary, the embodiments of this application have at least the following technical effects:

[0130] This application proposes a method and system for early warning of common animal diseases based on aquaculture environment monitoring data. By deeply integrating real-time environmental monitoring data, surrounding environmental information, characteristics of specific farmed animals, and historical disease occurrence patterns, a dynamic, accurate, and predictable quantitative assessment and early warning system for disease risk is constructed. This significantly improves the accuracy of early identification of potential animal disease risks in farms, the ability to distinguish types, and the predictability of risk accumulation trends. Compared with traditional methods that mainly rely on single environmental indicator thresholds for alarms, the technical solution provided in this application significantly overcomes the fundamental shortcomings of weak correlation between environmental anomalies and specific disease types, difficulty in quantifying risk accumulation effects, and lack of targeted and forward-looking early warnings. Specifically, the process involves several steps. First, collecting diverse environmental data provides robust support for subsequent data analysis. Second, a pre-built multi-channel disease classifier, tailored to the characteristics of farmed organisms, performs qualitative classification analysis on the environmental data, intelligently identifying the most likely set of potential diseases under specific farming conditions. Third, combining neighboring environmental data generates a predicted environmental sequence to simulate future environmental trends. Fourth, risk accumulation rate modeling for potential disease sets is introduced, quantifying the dynamic cumulative contribution of multiple environmental indicators to the risk of each potential disease based on historical data. By grouping and integrating the risk accumulation rate curves of each potential disease under the predicted environmental sequence, a precise quantitative assessment of the future cumulative risk of each potential disease is achieved. Finally, accurate and instructive early warning content is implemented. This series of collaborative mechanisms ensures that the early warning report not only clearly identifies the specific disease types that may occur but also clearly presents the risk accumulation level and development trend of each disease under current and predicted environmental conditions.

[0131] This application proposes a method and system for early warning of common animal diseases based on aquaculture environment monitoring data. By integrating real-time environmental monitoring data, neighborhood environmental information, characteristics of farmed animals, and historical disease occurrence patterns, a dynamic and accurate early warning system for common animal diseases is constructed. This achieves the technical effect of improving the accuracy of early warning for common animal diseases, which is conducive to risk control for potential diseases when raising common animals, and guides farmers to take targeted disease prevention and control measures in advance to reduce the risk of disease outbreaks.

[0132] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0133] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0134] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A common animal disease early warning method based on farmed environment detection data, characterized in that, Comprise: An interactive target breeding environment, obtaining breeding environment detection data and neighborhood environment data; Determine the breeding object information of the target breeding environment, and combine the breeding object information with the breeding environment detection data to call the pre-constructed disease classifier for qualitative classification analysis, and output the qualitative classification analysis result as a potential disease set, including: Interactively obtain the breeding object information of the target breeding environment; Based on the breeding object information, feature extraction is performed to obtain category features, breed features and stage features, and the output is breeding object features; Based on the breeding object features, activate the corresponding qualitative classification channel in the pre-constructed disease classifier, and input the breeding object information and the breeding environment detection data for qualitative classification analysis to obtain the qualitative classification analysis result; Wherein, before determining the breeding object information of the target breeding environment, and combining the breeding object information with the breeding environment detection data, calling the pre-constructed disease classifier for qualitative classification analysis, including: Based on historical disease data, train multiple qualitative classification branches corresponding to multiple types of breeding objects, wherein the qualitative classification branch includes multiple qualitative classification branch channels, and each qualitative classification branch channel corresponds to a breeding period of the breeding object corresponding to the qualitative classification branch. Through the breeding object classification layer, integrate multiple qualitative classification branches, and output the disease classifier; Obtain the category probability corresponding to each potential disease in the qualitative classification analysis result, and output the category probability set; Based on the preset significance constraint, traverse the category probability set for significance preliminary screening; According to the significance preliminary screening result, data cleaning is performed on the qualitative classification analysis result, and the data cleaned qualitative classification analysis result is output as the potential disease set, wherein the potential disease set includes N potential diseases, and N is greater than or equal to 1; According to the historical disease data, determine the risk accumulation rate set of M environment indicators in the breeding environment detection data for the potential disease set, wherein M is greater than or equal to 2, including: According to the breeding object features, collect the original historical disease data and homologous historical disease data of the target breeding environment, and output the historical disease data; Take the potential disease set as an index, traverse the historical disease data for classification screening to obtain potential disease historical data, wherein the potential disease historical data includes N historical data clusters corresponding to N potential diseases; Extract M environment indicators in the breeding environment detection data as an analysis object set, and traverse N historical data clusters respectively to perform risk accumulation modeling to obtain risk accumulation rate models of M environment indicators, and output the risk accumulation rate set; According to the neighborhood environment data and the breeding environment detection data, generate the prediction environment sequence data of the target breeding environment, and according to the breeding environment detection data and the risk accumulation rate set, perform risk accumulation calculation on the potential disease set to obtain the cumulative disease risk of each potential disease, including: Taking the farming environment detection data as a starting state, performing step-by-step environment prediction based on the sliding window method and the neighborhood environment data to obtain prediction environment sequence data, wherein the prediction environment sequence data comprises M prediction environment index sequences; Randomly extracting a first prediction environment index sequence from the prediction environment sequence data, and mapping the first prediction environment index sequence to N potential disease-prediction risk accumulation rate sequences based on the risk accumulation rate set and corresponding fitting to generate N potential disease-prediction risk accumulation rate curves; Traversing the prediction environment sequence data to obtain M sets of accumulation rate curve groups corresponding to M prediction environment index sequences; Grouping and integrating the M sets of accumulation rate curve groups based on N potential diseases, and grouping and adding the grouping integration results to obtain the cumulative disease risks of the N potential diseases; Generating an animal disease early warning report based on the multiple cumulative disease risks and performing early warning response. 2.The common animal disease early warning method based on the aquaculture environment detection data according to claim 1, wherein, Interacting with a target farming environment to obtain farming environment detection data and neighborhood environment data, comprising: Interacting with a plurality of source sensors arranged in the target farming environment to collect real-time detection data in the target farming environment; Based on a preset time window and a preset geographical boundary, interacting with an external data source to obtain neighborhood environment data. 3.The common animal disease early warning method based on the aquaculture environment detection data according to claim 2, wherein, Grouping and integrating the M sets of accumulation rate curve groups based on N potential diseases, and grouping and adding the grouping integration results to obtain the cumulative disease risks of the N potential diseases, comprising: Grouping and reconstructing M sets of accumulation rate curve groups with N potential diseases as indexes to obtain N sets of potential disease-accumulation rate curve groups; Traversing the N sets of potential disease-accumulation rate curve groups to perform integration calculation to obtain N sets of potential disease-cumulative disease risks; Respectively grouping and adding the N sets of potential disease-cumulative disease risks to obtain the cumulative disease risks of the N potential diseases.

4. A common animal disease early warning system based on farming environment detection data, characterized by, The system for implementing the common animal disease early warning method based on farming environment detection data according to any one of claims 1-3, comprising: An environment information acquisition module for interacting with a target farming environment to obtain farming environment detection data and neighborhood environment data; A qualitative classification analysis module for determining farming object information of the target farming environment, and combining the farming object information and the farming environment detection data to call a pre-constructed disease classifier for qualitative classification analysis and output a qualitative classification analysis result as a potential disease set, comprising: Interactively obtaining farming object information of the target farming environment; Based on the farming object information, performing feature extraction to obtain category features, breed features, and stage features, and outputting as farming object features; Based on the farming object features, activating corresponding qualitative classification channels in the pre-constructed disease classifier, and inputting the farming object information and the farming environment detection data for qualitative classification analysis to obtain a qualitative classification analysis result; Before determining the farming object information of the target farming environment and combining the farming object information and the farming environment detection data to call the pre-constructed disease classifier for qualitative classification analysis, comprising: Based on historical disease data, a plurality of qualitative classification branches corresponding to a plurality of cultured objects are trained, wherein each of the qualitative classification branches comprises a plurality of qualitative classification branch channels, and each of the qualitative classification branch channels corresponds to a cultured period of the cultured object corresponding to the qualitative classification branch; Through a cultured object classification layer, the plurality of qualitative classification branches are integrated, and an output is the disease classifier; A category probability corresponding to each potential disease in the qualitative classification analysis result is obtained, and an output is a category probability set; Based on a preset significance constraint, a significance preliminary screening is performed on the category probability set; According to the significance preliminary screening result, data cleaning is performed on the qualitative classification analysis result, and an output is the potential disease set after data cleaning, wherein the potential disease set comprises N potential diseases, and N is greater than or equal to 1; An environmental risk acquisition module is configured to determine, according to historical disease data, a risk accumulation rate set of M environmental indicators in the cultured environment detection data for the potential disease set, wherein M is greater than or equal to 2, and the environmental risk acquisition module comprises: According to the cultured object characteristics, original historical disease data and homologous historical disease data of a target cultured environment are collected, and an output is the historical disease data; Taking the potential disease set as an index, the historical disease data are traversed for classification screening, and potential disease historical data are obtained, wherein the potential disease historical data comprise N historical data clusters corresponding to N potential diseases; M environmental indicators in the cultured environment detection data are extracted as an analysis object set, and N historical data clusters are traversed for risk accumulation modeling, respectively, to obtain a risk accumulation rate model of the M environmental indicators, and an output is the risk accumulation rate set; A risk accumulation calculation module is configured to generate predicted environment sequence data of a target cultured environment according to the neighborhood environment data and the cultured environment detection data, and to calculate the risk accumulation of the potential disease set according to the cultured environment detection data and the risk accumulation rate set, to obtain the accumulated disease risk of each potential disease, and the risk accumulation calculation module comprises: Taking the cultured environment detection data as a starting state, a step-by-step environment prediction is performed according to the neighborhood environment data based on a sliding window method, to obtain the predicted environment sequence data, wherein the predicted environment sequence data comprises M predicted environment indicator sequences; A first predicted environment indicator sequence is randomly extracted from the predicted environment sequence data, and the first predicted environment indicator sequence is mapped to N potential disease-predicted risk accumulation rate sequences based on the risk accumulation rate set, and N potential disease-predicted risk accumulation rate curves are fitted and generated; M sets of accumulation rate curve groups corresponding to M predicted environment indicator sequences are obtained by traversing the predicted environment sequence data; N sets of accumulated disease risks of the potential diseases are obtained by grouping and integrating N potential diseases with M sets of the accumulation rate curve groups, and by grouping and adding the grouping integration results; An early warning response module is configured to generate an animal disease early warning report based on a plurality of accumulated disease risks, and to perform early warning response.

Images