Metapenaeus ensis breeding management method and system based on data analysis

Abstract

The invention discloses a metapenaeus ensis breeding management method and system based on data analysis, and the method comprises the steps: constructing a water quality anomaly detection model based on a generative adversarial network, obtaining environment monitoring information, and carrying out the abnormal water quality detection of a breeding region; evaluating the breeding quality influence degree under the current water quality according to the water quality anomaly detection information, and performing regulation and control early warning; performing disease monitoring on the metapenaeus ensis in the target breeding area to obtain disease monitoring information, performing disease prediction in combination with the water quality anomaly detection information, and analyzing the cause of the disease; noise identification analysis and noise tracing are carried out on the breeding area according to the environment monitoring information, and noise abnormity early warning is carried out; according to the disease type and the disease condition of the metapenaeus ensis in the target breeding pool, breeding feed optimization is carried out. The breeding management comprehensiveness of the metapenaeus ensis is improved, meanwhile, breeding feed optimization is carried out for the breeding ponds with diseases so as to enhance the disease resistance of other shrimps in the target breeding ponds, and the breeding quality is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of the breeding of Metapenaeus ensis, and particularly to a breeding management method and system for Metapenaeus ensis based on data analysis. Background Art

[0002] With the rapid development of the global aquaculture industry, shrimp farming has become one of the important economic activities. However, there are many challenges in the process of shrimp farming, such as changes in environmental conditions and frequent diseases, which seriously affect the breeding efficiency and survival rate of shrimp. Traditional shrimp breeding management methods mainly rely on experience and simple monitoring means, and it is difficult to comprehensively and real-time monitor and manage the complex factors in the breeding process, resulting in unstable breeding effects.

[0003] In recent years, with the continuous development of the Internet of Things technology, it has brought great changes to the field of aquaculture. By using the Internet of Things technology to monitor the breeding area of Metapenaeus ensis and obtain corresponding monitoring data, breeding control and management are carried out based on these monitoring data. The current existing technologies only manage the breeding living environment of shrimp, and cannot achieve comprehensive breeding control. Real-time monitoring and comprehensive management cannot be carried out for disease monitoring, noise environment and feed used in breeding. Therefore, how to manage the breeding of Metapenaeus ensis more comprehensively and improve the breeding quality of Metapenaeus ensis is an important issue. Summary of the Invention

[0004] The present invention overcomes the defects of the prior art and provides a breeding management method and system for Metapenaeus ensis based on data analysis, and its important purpose is to improve the comprehensiveness of the breeding management of Metapenaeus ensis and the breeding quality of Metapenaeus ensis.

[0005] To achieve the above purpose, the first aspect of the present invention provides a breeding management method for Metapenaeus ensis based on data analysis, including:

[0006] Constructing a water quality anomaly detection model based on a generative adversarial network, obtaining environmental monitoring information, detecting abnormal water quality in the breeding area, and obtaining water quality anomaly detection information;

[0007] Evaluating the influence degree of the breeding quality under the current water quality according to the water quality anomaly detection information, and performing regulation and warning;

[0008] Monitoring diseases of Metapenaeus ensis in the target breeding area to obtain disease monitoring information, combining the water quality anomaly detection information to predict diseases and analyze the causes of diseases;

[0009] Performing noise identification analysis and noise source tracing on the breeding area according to the environmental monitoring information, and performing noise anomaly warning;

[0010] Optimize the breeding feed according to the disease types and conditions of Metapenaeus ensis in the target breeding ponds.

[0011] In this solution, the water quality anomaly detection model is constructed based on the generative adversarial network, environmental monitoring information is obtained, and the breeding area is detected for abnormal water quality to obtain water quality anomaly detection information, which specifically includes:

[0012] Construct a water quality anomaly detection model based on the generative adversarial network, obtain the water quality benchmark information for the breeding of Metapenaeus ensis and the water quality change information of the shrimp breeding, and construct an instance dataset. Based on the instance dataset, construct a training dataset for model training;

[0013] Introduce a Markov chain for model optimization. Based on the instance dataset, obtain the water quality change information of normal breeding to construct a water quality change curve. Calculate the change amount of each water quality index in each time period according to the water quality benchmark information, and mark the water quality change curve;

[0014] Construct a water quality change state space based on the marked water quality change curve, calculate the state transition probability of each water quality index in the water quality change state space, and construct a state transition matrix through the calculated state transition probability;

[0015] Construct a water quality change threshold based on the change amount of each water quality index in each time period calculated, as a state transition penalty term, combine it with the state transition matrix to form a state transition constraint condition, and punish the generated sequence according to the state transition constraint condition;

[0016] Set the sequence acceptance rule of the discriminator based on the state transition matrix, obtain the state transition score of the target generated sequence based on the state transition matrix for the generated sequence, judge whether to accept the target generated sequence, and obtain an accepted sequence;

[0017] Calculate the reconstruction error between the accepted sequence and the original sequence, judge it with a preset threshold and generate an anomaly detection result to obtain a water quality anomaly detection model that meets expectations;

[0018] Obtain environmental monitoring information, where the environmental monitoring information includes water quality environmental monitoring information and noise environmental monitoring information, and input the environmental monitoring information into the water quality anomaly detection model for detection to obtain water quality anomaly detection information.

[0019] In this solution, the impact degree of the current water quality on the breeding quality is evaluated according to the water quality anomaly detection information, and regulation and warning are carried out, which specifically includes:

[0020] Obtain the water quality anomaly detection information, extract the abnormal water quality indexes and detection values according to the water quality anomaly detection information, and perform time series processing to form an abnormal water quality change curve;

[0021] Preset the normal deviation ranges of each water quality index, calculate the deviation values of each abnormal water quality index in combination with the abnormal water quality change curve, and merge the deviation values of each abnormal water quality index into the abnormal water quality change curve;

[0022] Obtain the historical water quality monitoring information of different breeding qualities, extract the abnormal water quality characteristics of each breeding quality, associate the breeding quality levels with the abnormal water quality indexes, and form the water quality index level intervals of each breeding level according to the detection values of each abnormal water quality index at different breeding quality levels;

[0023] Based on the constructed water quality index level intervals, conduct level evaluation on each abnormal water quality index in combination with the water quality abnormality detection information to obtain the level evaluation result information;

[0024] Calculate the mutual information values between each water quality index and the breeding quality and perform normalization processing, use the mutual information values of each water quality index after normalization processing as weights, and perform weighted calculation in combination with the level evaluation results;

[0025] Analyze the influence degree of each abnormal water quality index on the target breeding pond according to the weighted calculation results to obtain the breeding influence degree analysis information, and generate an abnormal water quality detection report for regulation and warning.

[0026] In this solution, for the Metapenaeus ensis in the target breeding area, disease monitoring is carried out to obtain disease monitoring information, and disease prediction and analysis of the cause of the disease are carried out in combination with the water quality abnormality detection information, specifically including:

[0027] Based on large-scale retrieval, obtain the disease instances in the breeding process of each Metapenaeus ensis, obtain various disease categories according to the obtained disease instances, extract the characteristics of each disease category, and extract the water quality characteristics and Metapenaeus ensis disease characteristics of each disease category to obtain the disease category characteristic information;

[0028] Based on the disease instances, disease categories and disease category characteristic information, form an entity triple, and construct a disease knowledge graph according to the entity triple;

[0029] Obtain the disease monitoring information, calculate the similarity between the disease monitoring information and the disease knowledge graph to obtain the similarity value, judge the calculated similarity value with the preset threshold, conduct disease prediction, and obtain the disease prediction information;

[0030] According to the disease prediction information, obtain the inducements of the predicted diseases, calculate the Pearson correlation coefficients between each disease inducement and the corresponding disease, and conduct main inducement analysis to obtain the main inducement analysis information;

[0031] Obtain water quality anomaly detection information, extract the number of diseased Metapenaeus ensis in the target breeding pond according to the disease monitoring information, conduct an assessment of the disease severity, and analyze the cause of the disease by combining the main inducement analysis information and the water quality anomaly detection information to obtain the disease cause analysis information.

[0032] In this solution, the noise identification analysis and noise source tracing are carried out on the breeding area according to the environmental monitoring information, and noise anomaly early warning is carried out, which specifically includes:

[0033] Obtain environmental monitoring information, obtain noise environmental monitoring information according to the environmental monitoring information, and perform data preprocessing on the noise environmental monitoring information;

[0034] Divide the breeding area into multiple sub-areas, combine the noise environmental monitoring information to obtain the noise environmental monitoring information corresponding to each sub-area, and extract the features of the noise environmental monitoring information of each sub-area to obtain the noise environmental feature information;

[0035] Construct a noise time distribution map for each sub-area according to the noise environment assessment information, evaluate the noise level of each sub-area, and judge the noise time period of each sub-area to obtain the noise level assessment information;

[0036] Construct a noise identification model based on a recurrent neural network, obtain various noise data through big data retrieval and perform noise category annotation, construct a noise database and perform model training, and input the noise environmental feature information into the noise identification model for noise identification;

[0037] Generate a feature sequence according to the input noise environmental feature information, forget the background noise of the generated feature sequence through a forgetting gate to generate a feature processing sequence, and perform noise identification analysis according to the feature processing sequence;

[0038] Construct a noise type probability distribution according to the noise prediction types at each time step, extract the probabilities of each noise type according to the noise type probability distribution, sort them and generate a prediction result according to the sorting result to obtain the noise identification analysis information;

[0039] Combine the noise identification analysis information and the noise level assessment information to evaluate the noise level of each sub-area, and generate a noise category traceability report according to the noise identification analysis information for early warning prompts.

[0040] In this solution, the breeding feed is optimized according to the disease types and conditions of Metapenaeus ensis in the target breeding pond, which specifically includes:

[0041] Based on big data retrieval, obtain the nutritional indicators for Metapenaeus ensis to resist diseases and the disease types caused by the lack of corresponding nutritional indicators, and associate the disease types with the corresponding nutritional indicators;

[0042] Based on the nutritional indicators of Metapenaeus ensis against diseases, breeding feeds corresponding to each nutritional indicator are obtained, feature extraction is carried out, and the component features, content features and nutritional features of each breeding feed are obtained to obtain breeding feed feature information;

[0043] According to the breeding feed feature information, quantitative analysis of component efficacy is carried out, the component features, content features and nutritional indicators are correlated, a quantitative standard is set, and the equivalent relationship between each component and the corresponding nutritional indicator per unit weight is analyzed to obtain the first analysis information;

[0044] According to the equivalent relationship features between each component and the corresponding nutritional indicator under the same nutritional indicator in the first analysis information, component equivalent relationship analysis is carried out to judge the equivalent substitution ability between each component, and component equivalent substitution analysis information is obtained;

[0045] A breeding feed optimization model is constructed, and according to the first analysis information and the component equivalent substitution analysis information, model training is carried out to obtain a breeding feed optimization model that meets the expectations;

[0046] Obtain information on the cause of the disease, judge whether the cause of the disease of the diseased Metapenaeus ensis is insufficient disease resistance ability, and if it is insufficient disease resistance ability, optimize the breeding feed;

[0047] Obtain the preparation information of the current breeding feed for Metapenaeus ensis, and input it into the breeding feed optimization model in combination with the information on the cause of the disease for breeding feed optimization to obtain breeding feed optimization information.

[0048] The second aspect of the present invention provides a Metapenaeus ensis breeding management system based on data analysis, which includes: a memory and a processor. The memory contains a Metapenaeus ensis breeding management method program based on data analysis. When the Metapenaeus ensis breeding management method program based on data analysis is executed by the processor, the following steps are realized:

[0049] Based on a generative adversarial network, a water quality anomaly detection model is constructed, environmental monitoring information is obtained, and abnormal water quality detection is carried out on the breeding area to obtain water quality anomaly detection information;

[0050] According to the water quality anomaly detection information, evaluate the influence degree of the current water quality on the breeding quality and carry out regulation and warning;

[0051] Carry out disease monitoring on Metapenaeus ensis in the target breeding area to obtain disease monitoring information, and combine the water quality anomaly detection information to carry out disease prediction and analyze the cause of the disease;

[0052] According to the environmental monitoring information, carry out noise identification analysis and noise source tracing on the breeding area, and carry out noise anomaly warning;

[0053] Optimize the breeding feed according to the disease type and condition of Metapenaeus ensis in the target breeding pond.

[0054] The present invention discloses a method and system for breeding management of Metapenaeus ensis based on data analysis, including: constructing a water quality anomaly detection model based on a generative adversarial network, obtaining environmental monitoring information, and detecting abnormal water quality in the breeding area; evaluating the impact degree of breeding quality under the current water quality according to the water quality anomaly detection information, and performing regulation and warning; monitoring diseases of Metapenaeus ensis in the target breeding area to obtain disease monitoring information, combining the water quality anomaly detection information to predict diseases and analyze the causes of diseases; performing noise identification analysis and noise source tracing on the breeding area according to the environmental monitoring information, and performing noise anomaly warning; optimizing the breeding feed according to the disease type and condition of Metapenaeus ensis in the target breeding pond. Improve the comprehensiveness of the breeding management of Metapenaeus ensis. At the same time, optimize the breeding feed for the breeding pond with diseases to enhance the disease resistance of the remaining shrimps in the target breeding pond and improve the breeding quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments or exemplary examples of the present invention, the following will briefly introduce the drawings required for use in the description of the embodiments or exemplary examples. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained according to the drawings shown without creative efforts.

[0056] Figure 1 It is a flowchart of a method for breeding management of Metapenaeus ensis based on data analysis provided by an embodiment of the present invention;

[0057] Figure 2 It is a flowchart of an environmental noise monitoring and management method provided by an embodiment of the present invention;

[0058] Figure 3 It is a block diagram of a system for breeding management of Metapenaeus ensis based on data analysis provided by an embodiment of the present invention;

[0059] The realization, functional characteristics and advantages of the object of the present invention will be further described with reference to the embodiments and the drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] In order to more clearly understand the above objects, features and advantages of the present invention, the present invention will be further described in detail below with reference to the drawings and specific embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0061] In the following description, many specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways than those specifically described herein. Therefore, the scope of protection of the present invention is not limited by the specific embodiments disclosed below.

[0062] Figure 1 A flowchart of a method for breeding management of Metapenaeus ensis based on data analysis provided for an embodiment of the present invention;

[0063] As Figure 1 shown, the present invention provides a flowchart of a method for breeding management of Metapenaeus ensis based on data analysis, including:

[0064] Construct a water quality anomaly detection model based on a generative adversarial network, obtain environmental monitoring information, perform abnormal water quality detection on the breeding area, and obtain water quality anomaly detection information;

[0065] Evaluate the impact degree of breeding quality under the current water quality according to the water quality anomaly detection information, and conduct regulation and early warning;

[0066] Conduct disease monitoring on Metapenaeus ensis in the target breeding area to obtain disease monitoring information, and combine the water quality anomaly detection information to conduct disease prediction and analyze the causes of the disease;

[0067] Conduct noise identification analysis and noise source tracing on the breeding area according to the environmental monitoring information, and conduct noise anomaly early warning;

[0068] Optimize the breeding feed according to the disease types and conditions of Metapenaeus ensis in the target breeding pond.

[0069] It should be noted that the present invention provides a method and system for breeding management of Metapenaeus ensis based on data analysis. By conducting water quality management, disease management, noise management, and breeding feed management on the target breeding area, the comprehensiveness of the breeding management of Metapenaeus ensis is improved. A water quality anomaly detection model is constructed based on a generative adversarial network, and a Markov chain is used for model optimization to improve the accuracy of water quality anomaly detection. By conducting disease monitoring on the breeding area, diseases that may occur are predicted based on the individual information of Metapenaeus ensis in the monitored breeding pond and the water quality anomaly detection information. At the same time, noise monitoring is conducted on the breeding area to avoid excessive noise in certain areas, which may cause Metapenaeus ensis in the breeding pond to become overly excited, resulting in injuries and other situations. Based on the disease prediction results, the causes of the disease are judged, and the breeding feed is optimized to improve the disease resistance of Metapenaeus ensis and the breeding quality of Metapenaeus ensis.

[0070] Further, in a preferred embodiment of the present invention, the water quality anomaly detection model is constructed based on a generative adversarial network, environmental monitoring information is obtained, and abnormal water quality in the breeding area is detected to obtain water quality anomaly detection information, which specifically includes:

[0071] Construct a water quality anomaly detection model based on a generative adversarial network, obtain the water quality benchmark information for the breeding of Metapenaeus ensis and the water quality change information of Metapenaeus ensis breeding, construct an instance dataset, and construct a training dataset based on the instance dataset for model training;

[0072] Introduce a Markov chain for model optimization, obtain the water quality change information of normal breeding based on the instance dataset to construct a water quality change curve, calculate the change amount of each water quality index in each time period according to the water quality benchmark information, and mark the water quality change curve;

[0073] Construct a water quality change state space based on the marked water quality change curve, calculate the state transition probability of each water quality index in the water quality change state space, and construct a state transition matrix through the calculated state transition probability;

[0074] Construct a water quality change threshold based on the change amount of each water quality index in each time period obtained by calculation as a state transition penalty term, combine it with the state transition matrix to form a state transition constraint condition, and punish the generated sequence according to the state transition constraint condition;

[0075] Set the sequence acceptance rule of the discriminator based on the state transition matrix, obtain the state transition score of the target generated sequence based on the state transition matrix for the generated sequence, and judge whether to accept the target generated sequence to obtain an accepted sequence;

[0076] Calculate the reconstruction error between the accepted sequence and the original sequence, judge it with a preset threshold and generate an anomaly detection result to obtain a water quality anomaly detection model that meets the expectations;

[0077] Obtain environmental monitoring information, where the environmental monitoring information includes water quality environmental monitoring information and noise environmental monitoring information, and input the environmental monitoring information into the water quality anomaly detection model for detection to obtain water quality anomaly detection information.

[0078] It should be noted that, first, the water quality baseline information of the breeding of Neopenaeus lanceolatus and the change information of water quality during the breeding process are obtained, and an example data set is constructed accordingly. The Markov chain is introduced to improve the accuracy and stability of the detection model. Based on the example data set, the water quality change information under normal breeding conditions is obtained, and the water quality change curve is constructed accordingly. The change amount of each water quality index in different time periods is determined by the water quality baseline information, which is used to mark the water quality change curve. Next, the water quality change state space is constructed according to the marked water quality change curve, and the state transition probability of each water quality index in the state space is calculated. The calculation result of the state transition probability is used to construct the state transition matrix. On this basis, the water quality change threshold is further constructed as a penalty term for state transition. The generated sequence is penalized according to the constraint conditions to ensure the rationality of the generated sequence. The sequence acceptance rule of the discriminator is set based on the state transition matrix, and the state transition score of the generated sequence is calculated to determine whether the target generated sequence is accepted, thereby obtaining the final accepted sequence. Subsequently, the reconstruction error between the accepted sequence and the original sequence is calculated, and the error is compared with the preset threshold to generate anomaly detection results and obtain a model that meets the expectations. Finally, environmental monitoring information is obtained, including water quality environmental monitoring information and noise environmental monitoring information. This environmental monitoring information is input into the trained water quality anomaly detection model for detection, and finally water quality anomaly detection information is obtained. This allows real-time monitoring of water quality changes, and early warning of potential water quality anomalies to ensure environmental safety for the breeding of Penaeus lanceolatus.

[0079] Furthermore, in a preferred embodiment of the present invention, the evaluation of the degree of influence on breeding quality under the current water quality according to the water quality abnormality detection information and the regulation and early warning are specifically included:

[0080] Acquire abnormal water quality detection information, extract abnormal water quality indicators and detection values ​​according to the abnormal water quality detection information, and perform time series processing to form an abnormal water quality change curve;

[0081] Preset a normal deviation range for each water quality indicator, calculate the deviation value of each abnormal water quality indicator in combination with the abnormal water quality change curve, and merge the deviation value of each abnormal water quality indicator into the abnormal water quality change curve;

[0082] Obtain historical water quality monitoring information of different breeding qualities, extract abnormal water quality characteristics of each breeding quality, associate the breeding quality level with the abnormal water quality index, and form the water quality index level range of each breeding level according to the detection values ​​of each abnormal water quality index at different breeding quality levels;

[0083] Based on the constructed water quality index grade interval, each abnormal water quality index is graded and evaluated in combination with the water quality anomaly detection information to obtain grade evaluation result information;

[0084] Calculate the mutual information value between each water quality index and the breeding quality and perform normalization processing, and use the mutual information value of each water quality index after normalization processing as the weight to perform weighted calculation in combination with the grade evaluation result;

[0085] Analyze the influence degree of each abnormal water quality index on the target breeding pond according to the weighted calculation result to obtain breeding influence degree analysis information, and generate an abnormal water quality detection report for regulation and warning.

[0086] It should be noted that first, obtain the water quality abnormality detection information. Extract the abnormal water quality index and its detection value, and perform time series processing to form an abnormal water quality change curve. Preset the normal deviation range of each water quality index, and combine it with the abnormal water quality change curve to calculate the deviation value of each abnormal water quality index, and merge these deviation values into the abnormal water quality change curve. Next, obtain the historical water quality monitoring information of different breeding qualities, extract the abnormal water quality characteristics of each breeding quality, and associate the breeding quality grade with the abnormal water quality index. According to the detection values of each abnormal water quality index at different breeding quality grades, form the water quality index grade interval of each breeding grade. Based on the constructed water quality index grade interval, combine the current water quality abnormality detection information to perform grade evaluation on each abnormal water quality index to obtain grade evaluation result information. Subsequently, calculate the mutual information value between each water quality index and the breeding quality, and perform normalization processing on these mutual information values. Use the mutual information value of each water quality index after normalization processing as the weight to perform weighted calculation in combination with the grade evaluation result. According to the weighted calculation result, analyze the influence degree of each abnormal water quality index on the target breeding pond to obtain breeding influence degree analysis information. Finally, generate an abnormal water quality detection report and provide regulation and warning suggestions so as to take measures in time to ensure the health of the breeding environment of the penaeid shrimp.

[0087] Further, in a preferred embodiment of the present invention, disease monitoring is performed on the Metapenaeus ensis in the target breeding area to obtain disease monitoring information, and disease prediction and analysis of the cause of the disease are performed in combination with the water quality abnormality detection information, specifically including:

[0088] Based on large number retrieval, obtain disease instances in the breeding process of each Metapenaeus ensis, obtain various disease categories according to the obtained disease instances, perform feature extraction on each disease category, and extract the water quality characteristics and Metapenaeus ensis disease characteristics of each disease category to obtain disease category feature information;

[0089] Based on the disease instance, disease category and disease category feature information, form an entity triple, and construct a disease knowledge graph according to the entity triple;

[0090] Obtain disease monitoring information, calculate the similarity between the disease monitoring information and the disease knowledge graph to obtain a similarity value, compare the calculated similarity value with a preset threshold to make a disease prediction, and obtain disease prediction information;

[0091] Obtain the inducing factors of the predicted disease according to the disease prediction information, calculate the Pearson correlation coefficient between each disease inducing factor and the corresponding disease, and conduct a main inducing factor analysis to obtain main inducing factor analysis information;

[0092] Obtain water quality anomaly detection information, extract the number of diseased Metapenaeus ensis in the target breeding pond according to the disease monitoring information, evaluate the degree of disease, and combine the main inducing factor analysis information and the water quality anomaly detection information to conduct an analysis of the cause of disease to obtain cause-of-disease analysis information.

[0093] It should be noted that, first, various disease instances in the process of penaeid shrimp breeding are obtained based on big data retrieval. Various disease categories are obtained through the disease instances, and feature extraction is performed on each disease category, including water quality features and Metapenaeus ensis disease features, to obtain disease category feature information. Next, entity triples are formed using the disease instances, disease categories, and disease category feature information. A disease knowledge graph is constructed based on the entity triples. The diseases, their features, and their relationships are systematically represented, thus providing strong support for disease prediction. During the disease monitoring process, the obtained disease monitoring information is calculated for similarity with the constructed disease knowledge graph to obtain a similarity value. The calculated similarity value is compared with a preset threshold to make a disease prediction and obtain corresponding disease prediction information. According to the disease prediction information, the inducing factors of the predicted disease are further obtained, and the Pearson correlation coefficient between each disease inducing factor and the corresponding disease is calculated for main inducing factor analysis. At the same time, water quality anomaly detection information is obtained, and the number of diseased Metapenaeus ensis in the target breeding pond is extracted in combination with the disease monitoring information to evaluate the degree of disease. Combining the main inducing factor analysis information and the water quality anomaly detection information, further analysis of the cause of disease is conducted to obtain cause-of-disease analysis information, including water quality reasons, resistance ability reasons, etc. Thus, diseases can be accurately predicted, the inducing factors of diseases can be deeply analyzed, and scientific basis and guidance can be provided for the health management of penaeid shrimp.

[0094] Further, in a preferred embodiment of the present invention, the noise identification analysis and noise source tracing are performed on the breeding area according to the environmental monitoring information, and noise anomaly early warning is carried out, which specifically includes:

[0095] Obtain environmental monitoring information, obtain noise environmental monitoring information according to the environmental monitoring information, and perform data preprocessing on the noise environmental monitoring information;

[0096] Divide the breeding area into multiple sub-areas, obtain the noise environment monitoring information corresponding to each sub-area in combination with the noise environment monitoring information, extract the characteristics of the noise environment monitoring information of each sub-area to obtain the noise environment characteristic information;

[0097] Construct a noise time distribution map for each sub-area according to the noise environment assessment information, evaluate the noise level of each sub-area, and judge the noise time period of each sub-area to obtain the noise level assessment information;

[0098] Construct a noise recognition model based on a recurrent neural network, obtain various noise data through big data retrieval and perform noise category annotation, construct a noise database and perform model training, and input the noise environment characteristic information into the noise recognition model for noise recognition;

[0099] Generate a feature sequence according to the input noise environment characteristic information, forget the background noise of the generated feature sequence through a forget gate to generate a feature processing sequence, and perform noise recognition analysis according to the feature processing sequence;

[0100] Construct a noise type probability distribution according to the noise prediction types at each time step, extract the probabilities of each noise type according to the noise type probability distribution, sort them and generate a prediction result according to the sorting result to obtain the noise recognition analysis information;

[0101] Combine the noise recognition analysis information and the noise level assessment information to evaluate the noise level of each sub-area, and generate a noise category traceability report according to the noise recognition analysis information for early warning prompts.

[0102] It should be noted that, first, noise environment monitoring information is obtained based on environmental monitoring information. Data preprocessing is performed on the noise environment monitoring information to ensure the accuracy and consistency of the data. Next, the breeding area is divided into multiple sub-areas, and the corresponding noise environment monitoring information for each sub-area is obtained in combination with the noise environment monitoring information. Feature extraction is performed on these monitoring information to obtain noise environment feature information. According to the noise environment feature information, a noise time distribution map of each sub-area is constructed, and the noise level of each sub-area is evaluated to determine the noise time period of each sub-area, thereby obtaining noise level evaluation information. Subsequently, a noise recognition model is constructed based on a recurrent neural network (RNN). Various noise data are obtained through big data retrieval and noise category labeling is performed to form a noise database. These data are used for model training, and the extracted noise environment feature information is input into the noise recognition model for noise recognition. During the noise recognition process, the input noise environment feature information generates a feature sequence, and the background noise is forgotten for the generated feature sequence through a forgetting gate to generate a feature processing sequence. Noise recognition analysis is performed according to the feature processing sequence. Further, a noise type probability distribution is formed based on the noise prediction types at each time step, the probabilities of each noise type are extracted from it, and sorted. A prediction result is generated according to the sorting result to obtain noise recognition analysis information. Combining the noise recognition analysis information and the noise level evaluation information, the noise level of each sub-area is rated. Finally, a noise category traceability report is generated according to the noise recognition analysis information and a warning prompt is given. Ensuring the comprehensive monitoring and analysis of the noise environment helps to timely discover and handle potential noise problems and avoid problems such as the injury of Metapenaeus ensis due to excessive stress.

[0103] Furthermore, in a preferred embodiment of the present invention, the optimization of the breeding feed according to the disease type and disease condition of Metapenaeus ensis in the target breeding pond specifically includes:

[0104] Based on big data retrieval, obtain the nutritional indicators for Metapenaeus ensis to resist diseases and the disease types caused by the lack of corresponding nutritional indicators, and associate the disease types with the corresponding nutritional indicators;

[0105] Based on the nutritional indicators for Metapenaeus ensis to resist diseases, obtain the corresponding breeding feeds for each nutritional indicator, perform feature extraction, and obtain the component features, content features, and nutritional features of each breeding feed to obtain breeding feed feature information;

[0106] According to the breeding feed feature information, perform quantitative analysis of component efficacy, associate the component features, content features, and nutritional indicators, set quantitative standards, and analyze the equivalent relationship between each component and the corresponding nutritional indicator per unit weight to obtain the first analysis information;

[0107] Extract the equivalent relationship characteristics between each component and the corresponding nutritional index under the same nutritional index according to the first analysis information, conduct component equivalent relationship analysis, judge the equivalent substitution ability between components, and obtain component equivalent substitution analysis information;

[0108] Construct a breeding feed optimization model, and perform model training according to the first analysis information and the component equivalent substitution analysis information to obtain a breeding feed optimization model that meets expectations;

[0109] Obtain the analysis information on the cause of the disease, and judge whether the cause of the disease of the diseased Metapenaeus ensis is insufficient disease resistance. If it is insufficient disease resistance, then optimize the breeding feed;

[0110] Obtain the preparation information of the currently used Metapenaeus ensis breeding feed, and input it into the breeding feed optimization model in combination with the analysis information on the cause of the disease for breeding feed optimization to obtain breeding feed optimization information.

[0111] It should be noted that for the disease situation of Metapenaeus ensis during breeding, it may be because the nutritional indicators in the feed of Metapenaeus ensis are not sufficient to enable Metapenaeus ensis to have better disease resistance, resulting in diseases. At this time, it is necessary to optimize the breeding feed of Metapenaeus ensis. First, based on big data retrieval, obtain the nutritional indicators related to disease resistance and the types of diseases caused by the lack of nutritional indicators, and associate the disease types with the corresponding nutritional indicators. Next, based on the nutritional indicators of Metapenaeus ensis for disease resistance, obtain the breeding feeds corresponding to each nutritional indicator, extract the characteristics of these breeding feeds, obtain their composition characteristics, content characteristics, and nutritional characteristics, and obtain the breeding feed characteristic information. Conduct a quantitative analysis of the component efficacy according to the breeding feed characteristic information, associate the composition characteristics and content characteristics with the nutritional indicators, set quantitative standards, analyze the equivalent relationship between each component and the corresponding nutritional indicator per unit weight, and obtain the first analysis information. Extract the equivalent relationship characteristics between each component and the corresponding nutritional indicator under the same nutritional indicator according to the first analysis information, conduct an analysis of the component equivalent relationship, and judge the equivalent substitution ability between each component to obtain the component equivalent substitution analysis information. On this basis, construct a breeding feed optimization model, and train the model according to the first analysis information and the component equivalent substitution analysis information to obtain a breeding feed optimization model that meets the expectations. Finally, obtain the analysis information on the cause of the disease, and judge whether the cause of the disease of the diseased Metapenaeus ensis is insufficient disease resistance. If it is determined that the disease resistance is insufficient, it is necessary to optimize the breeding feed. Obtain the preparation information of the currently used breeding feed of Metapenaeus ensis, combine the analysis information on the cause of the disease, input this information into the breeding feed optimization model for breeding feed optimization, and finally obtain the breeding feed optimization information. Improve the breeding feed formula specifically to improve the disease resistance of Metapenaeus ensis, thereby reducing the disease incidence rate, improving the breeding survival rate and health level, and realizing the refined management and efficient production of shrimp breeding.

[0112] Figure 2 Flowchart of the environmental noise monitoring and management method provided by an embodiment of the present invention;

[0113] As Figure 2 shown, the present invention provides a flowchart of an environmental noise monitoring and management method, including:

[0114] S202, obtain environmental monitoring information, extract noise environmental monitoring information according to the environmental monitoring information, and perform data preprocessing;

[0115] S204, construct a noise time distribution map of each sub-region according to the noise environment assessment information, evaluate the noise level of each sub-region, and obtain the noise level assessment information;

[0116] S206, input the noise environment monitoring information into a noise recognition model to obtain noise recognition analysis information;

[0117] S208, conduct noise level assessment for each sub-region by combining the noise recognition analysis information and the noise degree assessment information, and generate a noise category traceability report based on the noise recognition analysis information for regional noise anomaly warning.

[0118] It should be noted that by monitoring the noise in the target breeding area, evaluating the noise degree based on the obtained noise data, judging the noise levels of each sub-region, and thus understanding the noise areas exceeding the preset level. Then, identify the types of noise through a noise model to analyze the categories of noise generation. Finally, obtain the identified noise types based on the noise recognition analysis information, conduct noise traceability by combining the noise degree assessment information, lock the areas with high noise, and generate corresponding noise recognition results, further providing a solid foundation for noise control and ensuring the stability of the breeding environment of Metapenaeus ensis.

[0119] Figure 3 A Metapenaeus ensis breeding management system 3 based on data analysis provided by an embodiment of the present invention, the system includes: a memory 31, a processor 32, the memory 31 contains a Metapenaeus ensis breeding management method program based on data analysis, and when the Metapenaeus ensis breeding management method program based on data analysis is executed by the processor 32, the following steps are implemented:

[0120] Construct a water quality anomaly detection model based on a generative adversarial network, obtain environmental monitoring information, conduct abnormal water quality detection on the breeding area to obtain water quality anomaly detection information;

[0121] Evaluate the influence degree of the current water quality on the breeding quality according to the water quality anomaly detection information, and conduct regulation and warning;

[0122] Conduct disease monitoring on Metapenaeus ensis in the target breeding area to obtain disease monitoring information, and combine the water quality anomaly detection information to conduct disease prediction and analyze the causes of diseases;

[0123] Conduct noise recognition analysis and noise traceability on the breeding area according to the environmental monitoring information, and conduct noise anomaly warning;

[0124] Optimize the breeding feed according to the disease types and conditions of Metapenaeus ensis in the target breeding pond.

[0125] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces. The indirect coupling or communication connection of devices or units can be electrical, mechanical, or other forms.

[0126] The units described above as separate components may or may not be physically separated. The components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0127] In addition, each functional unit in the embodiments of the present invention can be all integrated in a processing unit, or each unit can be separately used as a unit, or two or more units can be integrated in a unit. The above-mentioned integrated units can be implemented in the form of hardware, or in the form of a combination of hardware and software functional units.

[0128] Those of ordinary skill in the art can understand that all or part of the steps to implement the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps including the above method embodiments. The foregoing storage medium includes: removable storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks and other various media that can store program codes.

[0129] Alternatively, if the above-mentioned integrated units of the present invention are implemented in the form of software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiments of the present invention, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: removable storage devices, ROM, RAM, magnetic disks, or optical disks and other various media that can store program codes.

[0130] As described above, it is only the specific implementation manner of the present invention. However, the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the said claims.

Claims

1. A method for breeding and managing Penaeus scutellariae based on data analysis, characterized in that: include: Based on the generative adversarial network, a water quality anomaly detection model is constructed to obtain environmental monitoring information, conduct abnormal water quality detection in the breeding area, and obtain water quality anomaly detection information; According to the abnormal water quality detection information, the impact on breeding quality under the current water quality is evaluated, and regulation and early warning are carried out; Conduct disease monitoring on Penaeus lanceolatus in the target breeding area to obtain disease monitoring information, and perform disease prediction and analyze the causes of disease in combination with the abnormal water quality detection information; Conduct noise identification and analysis and noise source tracing in the breeding area based on the environmental monitoring information, and issue an early warning of abnormal noise; Breeding feed optimization is carried out according to the disease type and disease condition of Penaeus lanceolatus in the target breeding pond.

2. A method for breeding and managing Penaeus lanceolatus based on data analysis according to claim 1, characterized in that: The water quality anomaly detection model is constructed based on the generative adversarial network, environmental monitoring information is obtained, abnormal water quality detection is performed on the breeding area, and water quality anomaly detection information is obtained, specifically including: Construct a water quality anomaly detection model based on a generative adversarial network, obtain water quality benchmark information for breeding Penaeus scutellaria and information on changes in water quality for breeding Penaeus scutellaria, construct an example data set, and construct a training data set based on the example data set to perform model training; The Markov chain is introduced to optimize the model, and the water quality change information of normal breeding is obtained based on the example data set to construct a water quality change curve. The change amount of each water quality index in each time period is calculated according to the water quality benchmark information, and the water quality change curve is marked; Construct a water quality change state space according to the marked water quality change curve, calculate the state transition probability of each water quality indicator in the water quality change state space, and construct a state transfer matrix through the calculated state transition probability; Based on the calculated changes in each water quality index in each time period, a water quality change threshold is constructed as a state transfer penalty item, and a state transfer constraint condition is formed in combination with the state transfer matrix, and a generated sequence is penalized according to the state transfer constraint condition; Based on the state transition matrix, the sequence acceptance rule of the discriminator is set, and the state transition score of the target generation sequence is calculated based on the state transition matrix to obtain the generated sequence, and it is determined whether the target generation sequence is acceptable to obtain the accepted sequence; Calculate the reconstruction error between the received sequence and the original sequence, make a judgment with the preset threshold and generate anomaly detection results to obtain a water quality anomaly detection model that meets the expectations; Environmental monitoring information is obtained, wherein the environmental monitoring information includes water quality environmental monitoring information and noise environmental monitoring information, and the environmental monitoring information is input into a water quality anomaly detection model for detection to obtain water quality anomaly detection information.

3. The method for breeding and managing Penaeus lanceolatus based on data analysis according to claim 1, characterized in that: The evaluation of the impact on breeding quality under the current water quality according to the water quality abnormality detection information and the regulation and early warning specifically include: Acquire abnormal water quality detection information, extract abnormal water quality indicators and detection values ​​according to the abnormal water quality detection information, and perform time series processing to form an abnormal water quality change curve; Preset a normal deviation range for each water quality indicator, calculate the deviation value of each abnormal water quality indicator in combination with the abnormal water quality change curve, and merge the deviation value of each abnormal water quality indicator into the abnormal water quality change curve; Obtain historical water quality monitoring information of different breeding qualities, extract abnormal water quality characteristics of each breeding quality, associate the breeding quality level with the abnormal water quality index, and form the water quality index level range of each breeding level according to the detection values ​​of each abnormal water quality index at different breeding quality levels; Based on the constructed water quality index grade interval, each abnormal water quality index is graded and evaluated in combination with the water quality anomaly detection information to obtain grade evaluation result information; Calculating the mutual information value between each water quality index and the breeding quality and performing normalization processing, using the mutual information value of each water quality index after normalization processing as a weight, and performing weighted calculation in combination with the grade evaluation result; According to the weighted calculation results, the impact of each abnormal water quality index on the target breeding pond is analyzed to obtain the breeding impact analysis information, and an abnormal water quality detection report is generated for regulation and early warning.

4. The method for breeding and managing Penaeus lanceolatus based on data analysis according to claim 1, characterized in that: The disease monitoring of Penaeus lanceolatus in the target breeding area is performed to obtain disease monitoring information, and the disease prediction and disease cause analysis are performed in combination with the abnormal water quality detection information, specifically including: Based on big data retrieval, disease examples in the breeding process of Penaeus lanceolatus were obtained, and various disease categories were obtained according to the obtained disease examples. The features of each disease category were extracted, and the water quality characteristics and disease characteristics of Penaeus lanceolatus were extracted to obtain the disease category feature information. Based on the disease instance, the disease category and the disease category feature information, an entity triple is formed, and a disease knowledge graph is constructed according to the entity triple; Obtain disease monitoring information, calculate similarity between the disease monitoring information and the disease knowledge graph, obtain a similarity value, judge the calculated similarity value with a preset threshold, perform disease prediction, and obtain disease prediction information; Acquire the cause of the predicted disease according to the disease prediction information, calculate the Pearson correlation coefficient between each disease cause and the corresponding disease, perform main cause analysis, and obtain main cause analysis information; Acquire abnormal water quality detection information, extract the number of diseased Penaeus lanceolatus in the target breeding pond according to the disease monitoring information, evaluate the degree of disease, analyze the cause of disease in combination with the main cause analysis information and the abnormal water quality detection information, and obtain disease cause analysis information.

5. The method for breeding and managing Penaeus scutellariae based on data analysis according to claim 1, characterized in that: The noise identification and analysis and noise source tracing of the breeding area according to the environmental monitoring information, and the noise abnormality warning are specifically included: Acquire environmental monitoring information, acquire noise environment monitoring information according to the environmental monitoring information, and perform data preprocessing on the noise environment monitoring information; Divide the breeding area into multiple sub-areas, obtain the noise environment monitoring information corresponding to each sub-area in combination with the noise environment monitoring information, extract features of the noise environment monitoring information of each sub-area, and obtain noise environment feature information; Constructing a noise time distribution map of each sub-area according to the noise environment assessment information, conducting a noise level assessment on each sub-area, and determining a noise time period of each sub-area to obtain noise level assessment information; A noise recognition model is constructed based on a recurrent neural network, various noise data are obtained through big data retrieval and noise category annotation is performed, a noise database is constructed and model training is performed, and the noise environment characteristic information is input into the noise recognition model for noise recognition; Generate a feature sequence according to input noise environment feature information, forget background noise on the generated feature sequence through a forget gate, generate a feature processing sequence, and perform noise recognition analysis according to the feature processing sequence; Constructing a noise type probability distribution according to the noise prediction type at each time step, extracting the probability of each noise type according to the noise type probability distribution, sorting them, and generating a prediction result according to the sorting result to obtain noise recognition analysis information; The noise level of each sub-area is assessed in combination with the noise identification and analysis information and the noise level assessment information, and a noise category tracing report is generated based on the noise identification and analysis information to provide early warning prompts.

6. The method for breeding and managing Penaeus lanceolatus based on data analysis according to claim 1, characterized in that: The optimization of breeding feed according to the disease type and disease condition of Penaeus lanceolatus in the target breeding pond specifically includes: Based on big data retrieval, we can obtain the nutritional indicators of Penaeus lanceolatus to resist diseases and the types of diseases caused by the lack of corresponding nutritional indicators, and then associate the disease types with the corresponding nutritional indicators. Based on the nutritional indexes of Penaeus lanceolatus to resist diseases, the breeding feeds corresponding to the nutritional indexes are obtained, and feature extraction is performed to obtain the component characteristics, content characteristics and nutritional characteristics of each breeding feed, and obtain the breeding feed feature information; Performing a quantitative analysis of the efficacy of the ingredients according to the breeding feed characteristic information, correlating the ingredient characteristics, content characteristics and nutritional indicators, setting a quantitative standard, analyzing the equivalent relationship between each ingredient and the corresponding nutritional indicator under unit weight, and obtaining first analysis information; Extracting the equivalent relationship characteristics of each component under the same nutritional index and the corresponding nutritional index according to the first analysis information, performing component equivalent relationship analysis, judging the equivalent substitution ability between the components, and obtaining component equivalent substitution analysis information; Constructing a breeding feed optimization model, and performing model training according to the first analysis information and the component equivalent substitution analysis information to obtain a breeding feed optimization model that meets expectations; Obtain disease cause analysis information to determine whether the disease cause of the diseased Penaeus lanceolatus is insufficient disease resistance. If so, optimize the breeding feed; The currently used breeding feed preparation information of Penaeus scabra is obtained, and the information is combined with the disease cause analysis information and input into the breeding feed optimization model to optimize the breeding feed, thereby obtaining the breeding feed optimization information.

7. A breeding management system for Penaeus serratus based on data analysis, characterized in that: The system comprises: a memory and a processor, wherein the memory contains a method program for breeding and managing Penaeus lanceolatus shrimp based on data analysis, and when the method program for breeding and managing Penaeus lanceolatus shrimp based on data analysis is executed by the processor, the following steps are implemented: Based on the generative adversarial network, a water quality anomaly detection model is constructed to obtain environmental monitoring information, conduct abnormal water quality detection in the breeding area, and obtain water quality anomaly detection information; According to the abnormal water quality detection information, the impact on breeding quality under the current water quality is evaluated, and regulation and early warning are carried out; Conduct disease monitoring on Penaeus lanceolatus in the target breeding area to obtain disease monitoring information, and perform disease prediction and analyze the causes of disease in combination with the abnormal water quality detection information; Conduct noise identification and analysis and noise source tracing in the breeding area based on the environmental monitoring information, and issue an early warning of abnormal noise; Breeding feed optimization is carried out according to the disease type and disease condition of Penaeus lanceolatus in the target breeding pond.

8. The data analysis-based breeding management system for Penaeus lanceolatus according to claim 7 is characterized in that: The water quality anomaly detection model is constructed based on the generative adversarial network, environmental monitoring information is obtained, abnormal water quality detection is performed on the breeding area, and water quality anomaly detection information is obtained, specifically including: Construct a water quality anomaly detection model based on a generative adversarial network, obtain water quality benchmark information for breeding Penaeus scutellaria and information on changes in water quality for breeding Penaeus scutellaria, construct an example data set, and construct a training data set based on the example data set to perform model training; The Markov chain is introduced to optimize the model, and the water quality change information of normal breeding is obtained based on the example data set to construct a water quality change curve. The change amount of each water quality index in each time period is calculated according to the water quality benchmark information, and the water quality change curve is marked; Construct a water quality change state space according to the marked water quality change curve, calculate the state transition probability of each water quality indicator in the water quality change state space, and construct a state transfer matrix through the calculated state transition probability; Based on the calculated changes in each water quality index in each time period, a water quality change threshold is constructed as a state transfer penalty item, and a state transfer constraint condition is formed in combination with the state transfer matrix, and a generated sequence is penalized according to the state transfer constraint condition; Based on the state transition matrix, the sequence acceptance rule of the discriminator is set, and the state transition score of the target generation sequence is calculated based on the state transition matrix to obtain the generated sequence, and it is determined whether the target generation sequence is acceptable to obtain the accepted sequence; Calculate the reconstruction error between the received sequence and the original sequence, make a judgment with the preset threshold and generate anomaly detection results to obtain a water quality anomaly detection model that meets the expectations; Environmental monitoring information is obtained, wherein the environmental monitoring information includes water quality environmental monitoring information and noise environmental monitoring information, and the environmental monitoring information is input into a water quality anomaly detection model for detection to obtain water quality anomaly detection information.

9. The data analysis-based breeding management system for Penaeus serratus according to claim 7 is characterized in that: The evaluation of the impact on breeding quality under the current water quality according to the water quality abnormality detection information and the regulation and early warning specifically include: Acquire abnormal water quality detection information, extract abnormal water quality indicators and detection values ​​according to the abnormal water quality detection information, and perform time series processing to form an abnormal water quality change curve; Preset a normal deviation range for each water quality indicator, calculate the deviation value of each abnormal water quality indicator in combination with the abnormal water quality change curve, and merge the deviation value of each abnormal water quality indicator into the abnormal water quality change curve; Obtain historical water quality monitoring information of different breeding qualities, extract abnormal water quality characteristics of each breeding quality, associate the breeding quality level with the abnormal water quality index, and form the water quality index level range of each breeding level according to the detection values ​​of each abnormal water quality index at different breeding quality levels; Based on the constructed water quality index grade interval, each abnormal water quality index is graded and evaluated in combination with the water quality anomaly detection information to obtain grade evaluation result information; Calculating the mutual information value between each water quality index and the breeding quality and performing normalization processing, using the mutual information value of each water quality index after normalization processing as a weight, and performing weighted calculation in combination with the grade evaluation result; According to the weighted calculation results, the impact of each abnormal water quality index on the target breeding pond is analyzed to obtain the breeding impact analysis information, and an abnormal water quality detection report is generated for regulation and early warning.

10. The data analysis-based breeding management system for Penaeus scutellariae according to claim 7 is characterized in that: The disease monitoring of Penaeus lanceolatus in the target breeding area is performed to obtain disease monitoring information, and the disease prediction and disease cause analysis are performed in combination with the abnormal water quality detection information, specifically including: Based on big data retrieval, disease examples in the breeding process of Penaeus lanceolatus were obtained, and various disease categories were obtained according to the obtained disease examples. The features of each disease category were extracted, and the water quality characteristics and disease characteristics of Penaeus lanceolatus were extracted to obtain the disease category feature information. Based on the disease instance, the disease category and the disease category feature information, an entity triple is formed, and a disease knowledge graph is constructed according to the entity triple; Obtain disease monitoring information, calculate similarity between the disease monitoring information and the disease knowledge graph, obtain a similarity value, judge the calculated similarity value with a preset threshold, perform disease prediction, and obtain disease prediction information; Acquire the cause of the predicted disease according to the disease prediction information, calculate the Pearson correlation coefficient between each disease cause and the corresponding disease, perform main cause analysis, and obtain main cause analysis information; Acquire abnormal water quality detection information, extract the number of diseased Penaeus lanceolatus in the target breeding pond according to the disease monitoring information, evaluate the degree of disease, analyze the cause of disease in combination with the main cause analysis information and the abnormal water quality detection information, and obtain disease cause analysis information.

Images