Method and system for controlling a facility for cultivating daphnia magna in the field

Abstract

The present application relates to the technical field of aquaculture equipment control, and particularly relates to a control method and system for a large field culture facility of a chironomid. A chironomid growth activity anomaly probability prediction model is constructed according to historical culture environment parameters and in combination with a Bayesian network. Real-time culture environment parameters in a preset culture time period in the large field culture facility of the chironomid are obtained, and the activity state of the chironomid in a preset future time period is predicted and analyzed according to the chironomid growth activity anomaly probability prediction model and in combination with the real-time culture environment parameters. If the result is a second analysis result, the real-time culture environment parameters are imported into a previously constructed semantic database for matching, and a recommended prevention and control scheme is obtained by searching in a prevention and control log according to the abnormal type of the chironomid growth activity anomaly that will occur in the large field culture facility of the chironomid in the preset future time period. The present application can achieve precise prevention and control, and improve chironomid culture efficiency and benefits.

Description

Technical Field

[0001] This invention relates to the field of aquaculture equipment control technology, and in particular to a control method and system for a field breeding facility for rice worms. Background Technology

[0002] Rice worm large-scale farming facilities refer to facilities for large-scale rice worm farming using paddy field environments. These facilities typically include paddy fields, environmental monitoring equipment, environmental control equipment, feeding equipment, disease control equipment, harvesting equipment, and a data management system. Traditional rice worm large-scale farming relies mainly on experience and manual monitoring, making it difficult to respond promptly and effectively to abnormal growth and activity caused by changes in the farming environment. With the development of IoT and AI technologies, it has become possible to monitor and control the farming environment using data-driven methods. However, most existing farming facility control methods are based on simple threshold judgments and cannot fully utilize historical data and model predictions, resulting in a lack of targeted and effective regulation of the farming environment. Summary of the Invention

[0003] This invention overcomes the shortcomings of the prior art and provides a control method and system for rice worm field breeding facilities.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0005] This invention discloses a control method for a field breeding facility for rice worms, comprising the following steps:

[0006] Obtain the breeding record event book of the field breeding facility for rice worms, and analyze the breeding environment parameter deviation events based on the breeding record event book to obtain the related breeding environment parameter deviation events or the unrelated breeding environment parameter deviation events.

[0007] Obtain the historical aquaculture environment parameters corresponding to the field aquaculture facilities of *Sinocyclocheilus scoparia* when a correlated or uncorrelated aquaculture environment parameter shift event occurs. Construct a probability prediction model for abnormal growth and activity of *Sinocyclocheilus scoparia* based on the historical aquaculture environment parameters and in conjunction with a Bayesian network.

[0008] The real-time breeding environment parameters of the rice worm in the field breeding facility within a preset breeding time period are obtained, and the activity status of the rice worm in the preset future time period is predicted and analyzed based on the rice worm growth activity abnormality probability prediction model and the real-time breeding environment parameters.

[0009] If the result is the first analysis result, no adjustment or treatment will be applied to the field breeding facility for rice worms; if the result is the second analysis result, the real-time breeding environment parameters will be imported into a pre-built semantic database for matching, and the abnormal type of rice worms in the field breeding facility that will exhibit abnormal growth and activity within a preset future time period will be matched.

[0010] Based on the abnormal types of growth and activity abnormalities that are expected to occur in the field breeding facilities of rice worms within a preset future time period, recommended control plans are retrieved from the control log, and the field breeding facilities of rice worms are regulated and treated according to the recommended control plans.

[0011] Preferably, the event log of the rice worm field breeding facility is obtained, and the breeding environment parameter deviation events are analyzed based on the event log to obtain correlated or uncorrelated breeding environment parameter deviation events, specifically:

[0012] Obtain the breeding record event book of the field breeding facility for rice worms, and extract the breeding environment parameter offset events of the field breeding facility for rice worms in the historical working process based on the breeding record event book;

[0013] Based on the aforementioned breeding record event book, the number of abnormal growth and activity events of rice worms caused by various breeding environment parameter deviation events in the rice worm field breeding facility was further counted.

[0014] The abnormal probability value of the abnormal growth and activity of the rice worm caused by the occurrence of various deviation events of breeding environmental parameters in the rice worm field breeding facility is calculated.

[0015] Events with aquaculture environment parameter offsets whose abnormal probability values ​​exceed a preset threshold are marked as related aquaculture environment parameter offset events;

[0016] Events with aquaculture environment parameter offsets whose abnormal probability values ​​are not greater than a preset threshold are marked as non-associated aquaculture environment parameter offset events.

[0017] Preferably, the historical aquaculture environment parameters corresponding to the field aquaculture facilities of *Sinocypripedium spp.* at the time of the occurrence of a correlated or uncorrelated aquaculture environment parameter shift event are obtained. Based on the historical aquaculture environment parameters and combined with a Bayesian network, a probability prediction model for abnormal growth and activity of *Sinocypripedium spp.* is constructed, specifically as follows:

[0018] Obtain the breeding record event book of the field breeding facility for rice worms; based on the breeding record event book, obtain the historical breeding environment parameters corresponding to the field breeding facility for rice worms when a related breeding environment parameter deviation event occurs; and obtain the historical breeding environment parameters corresponding to the field breeding facility for rice worms when a non-related breeding environment parameter deviation event occurs.

[0019] The related aquaculture environment parameter shift event is used as the first condition node, and the historical aquaculture environment parameters corresponding to the field aquaculture facility of the rice worm at the time of the related aquaculture environment parameter shift event are used as the first random variable node; and each first condition node and the first random variable node are connected in a directed manner to form a first topology diagram.

[0020] The non-correlated aquaculture environment parameter shift event is used as the second condition node, and the historical aquaculture environment parameters corresponding to the field aquaculture facility of the rice worm at the time of the non-correlated aquaculture environment parameter shift event are used as the second random variable node; and each second condition node and the second random variable node are connected in a directed manner to form a second topology diagram.

[0021] Based on the first topological structure diagram, a first conditional probability matrix is ​​obtained for abnormal growth and activity events of worms caused by a correlated aquaculture environment parameter deviation event, and a second conditional probability matrix is ​​obtained for abnormal growth and activity events of worms caused by a non-correlated aquaculture environment parameter deviation event.

[0022] A Bayesian network is constructed, and the first and second conditional probability matrices are imported into the Bayesian network for parameter learning. The posterior probability of the Bayesian network after each learning is obtained until the posterior probability converges to a preset probability value. The final learning parameters are then saved to obtain the abnormal probability prediction model of the growth activity of the rice worm.

[0023] Preferably, real-time breeding environment parameters are obtained in the field breeding facility for a preset breeding period. Based on the predicted probability model of abnormal growth activity of the rice worms and combined with the real-time breeding environment parameters, the activity status of the rice worms in the preset future period is predicted and analyzed. Specifically:

[0024] The real-time breeding environment parameters of the field breeding facility for the rice worm are obtained within a preset breeding time period, and the real-time breeding environment parameters of the field breeding facility for the preset breeding time period are imported into the rice worm growth activity abnormality probability prediction model for prediction.

[0025] Obtain the prediction results of the abnormal growth activity probability prediction model of rice worm, and obtain the probability value of the occurrence of abnormal growth activity events of rice worm in the field breeding facility under the current real-time breeding environment parameter conditions based on the prediction results.

[0026] If the probability of an abnormal growth and activity event of rice worms in the field breeding facility caused by the current real-time breeding environment parameters is not greater than the preset probability value, it means that rice worms will not exhibit abnormal growth and activity within the preset future time period, and the first analysis result is generated.

[0027] If the probability of an abnormal growth and activity event of rice worms in the field breeding facility under the current real-time breeding environment parameters is greater than the preset probability value, it indicates that the rice worms will experience abnormal growth and activity within the preset future time period, and a second analysis result will be generated.

[0028] Preferably, if the result is the second analysis result, the real-time aquaculture environment parameters are imported into a pre-constructed semantic database for matching, and the abnormality type that indicates abnormal growth and activity of the rice worms in the rice worm field aquaculture facility within a preset future time period is identified, specifically:

[0029] Obtain the breeding record event book of the field breeding facility for rice worms, and extract the abnormal growth and activity events of rice worms that occurred in the field breeding facility during the historical operation based on the breeding record event book;

[0030] Obtain the abnormality type of each abnormal event in the growth activity of the rice worm, and obtain the historical breeding environment parameters of the rice worm field breeding facility within a preset time period before each abnormal event in the growth activity of the rice worm occurred.

[0031] A semantic network is constructed, and several cutting nodes are determined based on the abnormality type of each abnormal event in the growth and activity of the rice worm. The semantic network is then cut according to each cutting node to obtain several branch nodes. Each branch node is assigned a corresponding abnormality type index according to the abnormality type corresponding to each cutting node.

[0032] The historical farming environment parameters of the field farming facilities for each type of abnormal growth activity of rice worms within a preset time period before the occurrence of each abnormal event are stored on the branch nodes after being assigned the corresponding abnormality type index, thus obtaining a semantic database.

[0033] If the result is the second analysis result, then the real-time breeding environment parameters within the preset breeding time period in the field breeding facility of rice worm are obtained, and the real-time breeding environment parameters are imported into the semantic database;

[0034] The Euclidean distance algorithm is introduced, and the similarity between the real-time aquaculture environment parameters and the historical aquaculture environment parameters on each branch node in the semantic database is calculated based on the Euclidean distance algorithm to obtain several similarity scores.

[0035] Sort out the maximum similarity among several similarities, obtain the branch node corresponding to the maximum similarity, and extract the anomaly type index assigned to the branch node corresponding to the maximum similarity.

[0036] Based on the extracted anomaly type index, the anomaly types that indicate abnormal growth and activity of rice worms in field farming facilities within a preset future time period are determined.

[0037] Preferably, a recommended control plan is retrieved from the control log based on the type of abnormal growth and activity of rice worms in the rice worm field breeding facility within a preset future time period. The rice worm field breeding facility is then regulated and controlled based on the recommended control plan, specifically as follows:

[0038] Search terms are generated based on the types of abnormalities that indicate abnormal growth and activity of rice worms in field farming facilities within a preset future time period.

[0039] Obtain the prevention and control logs of the field breeding facility for rice worms, and search the prevention and control logs based on the search terms to find the corresponding prevention and control measures for abnormal growth and activity of rice worms during the breeding process.

[0040] And obtain the success rate of various prevention and control measures obtained from the search, and screen out prevention and control measures whose success rate is not greater than the preset success rate;

[0041] Obtain the energy consumption value of the remaining control measures, select the control measure with the lowest energy consumption value as the recommended control plan, and send the recommended control plan to the control terminal of the field breeding facility for rice worms;

[0042] Based on the recommended prevention and control plan, the field breeding facilities for rice worms were regulated and treated to prevent the rice worms from experiencing abnormal growth and activity.

[0043] The prevention and control measures include ventilation and heat dissipation, shed insulation, baiting, drug administration, and drainage.

[0044] The aquaculture environment parameters include pH value, dissolved oxygen concentration, ammonia nitrogen concentration, nitrite concentration, temperature, humidity, and light intensity.

[0045] The aforementioned aquaculture environment parameter offset events are one or more combinations of pH value parameter offset, dissolved oxygen concentration parameter offset, ammonia nitrogen concentration parameter offset, nitrite concentration parameter offset, temperature parameter offset, humidity parameter offset, and light intensity parameter offset.

[0046] The abnormal growth and activity events of the rice worm include abnormal growth rate, abnormal mortality rate, abnormal activity, abnormal morphology, abnormal reproductive capacity, and abnormal behavior.

[0047] The aforementioned field breeding facilities for rice worms include:

[0048] The main structure of the greenhouse is built on a paddy field;

[0049] The water supply and drainage system module is used to regulate the water flow speed and water level in paddy fields;

[0050] The feeding system module is used to feed the rice worms at regular intervals and in fixed quantities.

[0051] The environmental control module is used to control the parameters of the aquaculture environment;

[0052] The data acquisition and analysis module collects various types of data and performs analysis and processing.

[0053] The automatic control module connects all the modules to achieve automated control of the breeding process.

[0054] This invention addresses the technical deficiencies in the prior art and offers the following advantages: It constructs a probability prediction model for abnormal growth and activity of *Clerodendrum trichotomum* based on historical aquaculture environmental parameters and a Bayesian network; it obtains real-time aquaculture environmental parameters within a preset aquaculture time period in a field aquaculture facility, and predicts and analyzes the activity status of *Clerodendrum trichotomum* within a preset future time period based on the probability prediction model and real-time aquaculture environmental parameters; if the result is the first analysis result, no control measures are taken for the field aquaculture facility; if the result is the second analysis result, the real-time aquaculture environmental parameters are imported into a pre-constructed semantic database for matching, identifying the abnormal types of growth and activity abnormalities that will occur in the field aquaculture facility within the preset future time period; based on the abnormal types of growth and activity abnormalities that will occur in the field aquaculture facility within the preset future time period, a recommended control plan is retrieved from the control log, and the field aquaculture facility is controlled based on the recommended control plan. This invention can predict the future growth and activity status of *Clerodendrum trichotomum* through a probability prediction model and real-time environmental parameters, thereby achieving precise control and improving aquaculture efficiency and effectiveness. Attached Figure Description

[0055] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained from these drawings without creative effort.

[0056] Figure 1 This is the first method flowchart of the control method for field breeding facilities of this rice worm;

[0057] Figure 2 This is the second method flowchart of the field breeding facility control method for rice worms;

[0058] Figure 3 This is a simplified structural diagram of the field breeding facility for the rice worm. Detailed Implementation

[0059] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0060] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0061] This invention discloses a control method for a field breeding facility for rice worms, such as... Figure 1 As shown, it includes the following steps:

[0062] S102: Obtain the breeding record event book of the field breeding facility for rice worms, and analyze the breeding environment parameter deviation events based on the breeding record event book to obtain the related breeding environment parameter deviation events or the unrelated breeding environment parameter deviation events.

[0063] S104: Obtain the historical aquaculture environment parameters corresponding to the field aquaculture facilities of the worm when a related or unrelated aquaculture environment parameter shift event occurs, and construct a worm growth activity abnormality probability prediction model based on the historical aquaculture environment parameters and combined with a Bayesian network.

[0064] S106: Obtain real-time breeding environment parameters in the field breeding facility for a preset breeding time period, and predict and analyze the activity status of the worms in the preset future time period based on the worm growth activity abnormality probability prediction model and the real-time breeding environment parameters.

[0065] S108: If the result is the first analysis result, no adjustment or treatment will be made to the field breeding facility for rice worms; if the result is the second analysis result, the real-time breeding environment parameters will be imported into the pre-built semantic database for matching, and the abnormal type of rice worms in the field breeding facility will show abnormal growth and activity in the preset future time period will be matched.

[0066] S110: Based on the abnormality type of the abnormal growth and activity of the rice worms in the rice worm field breeding facility within a preset future time period, a recommended prevention and control plan is obtained from the prevention and control log, and the rice worm field breeding facility is regulated and treated based on the recommended prevention and control plan.

[0067] The aquaculture environment parameters include pH value, dissolved oxygen concentration, ammonia nitrogen concentration, nitrite concentration, temperature, humidity, and light intensity. Deviation events in these parameters include one or more combinations of pH value deviation, dissolved oxygen concentration deviation, ammonia nitrogen concentration deviation, nitrite concentration deviation, temperature deviation, humidity deviation, and light intensity deviation. Abnormal events related to the growth and activity of the rice worms include abnormal growth rate, abnormal mortality rate, abnormal activity level, abnormal morphology, abnormal reproductive capacity, and abnormal behavior.

[0068] This invention proposes a control method for field breeding facilities of rice worms based on Bayesian networks and semantic databases. The aim is to identify related / unrelated breeding environment parameter deviation events by analyzing breeding record event books, construct a probability prediction model for abnormal growth activity, and combine it with real-time environmental parameters to predict future growth activity status, thereby achieving precise prevention and control and improving breeding efficiency and benefits.

[0069] Preferably, the event log of the rice worm field breeding facility is obtained, and the breeding environment parameter deviation events are analyzed based on the event log to obtain correlated or uncorrelated breeding environment parameter deviation events, such as... Figure 2 As shown, specifically:

[0070] S202: Obtain the breeding record event book of the field breeding facility for rice worms, and extract the breeding environment parameter offset events of the field breeding facility for rice worms in the historical working process based on the breeding record event book;

[0071] S204: Based on the aforementioned breeding record event book, further count the number of abnormal growth and activity events of rice worms caused by various breeding environment parameter deviation events in the rice worm field breeding facility;

[0072] S206: Calculate the abnormal probability value of the abnormal growth activity of the rice worm caused by the occurrence of various breeding environment parameter deviation events in the rice worm field breeding facility.

[0073] S208: Mark the aquaculture environment parameter offset event corresponding to the abnormal probability value being greater than the preset threshold as a related aquaculture environment parameter offset event;

[0074] S210: Mark the aquaculture environment parameter offset events whose abnormal probability values ​​are not greater than the preset threshold as non-correlated aquaculture environment parameter offset events.

[0075] It should be noted that, firstly, all relevant data from the historical operation of the field breeding facility for rice worms are collected and organized to form a comprehensive breeding record event book. This data typically includes, but is not limited to, key environmental parameters such as temperature, humidity, light intensity, water pH, and nutrient supply, and their changes. Through in-depth analysis of the breeding record event book, abnormal changes in environmental parameters at specific points in time are identified, i.e., breeding environmental parameter deviation events. These events may be caused by natural factors, human operational errors, or other unforeseen factors. Furthermore, it is analyzed whether each environmental parameter deviation event is accompanied by abnormal phenomena in the growth and activity of rice worms (such as slower growth rate, increased mortality, etc.). By calculating the proportion of abnormal growth and activity events caused by specific environmental parameter deviation events, an abnormality probability value is obtained. Based on a pre-set threshold, environmental parameter deviation events with an abnormality probability value exceeding the threshold are marked as "related breeding environmental parameter deviation events," meaning that these events have a direct causal relationship with the abnormal growth and activity of rice worms; while those with an abnormality probability values ​​below or equal to the threshold are marked as "unrelated breeding environmental parameter deviation events." By accurately identifying changes in environmental parameters that negatively impact the growth and activity of *Gnaphalium affine*, and selecting highly reliable data, a prediction model for the abnormal growth and activity of *Gnaphalium affine* can be constructed, which can effectively improve the model's prediction accuracy.

[0076] Preferably, the historical aquaculture environment parameters corresponding to the field aquaculture facilities of *Sinocypripedium spp.* at the time of the occurrence of a correlated or uncorrelated aquaculture environment parameter shift event are obtained. Based on the historical aquaculture environment parameters and combined with a Bayesian network, a probability prediction model for abnormal growth and activity of *Sinocypripedium spp.* is constructed, specifically as follows:

[0077] Obtain the breeding record event book of the field breeding facility for rice worms; based on the breeding record event book, obtain the historical breeding environment parameters corresponding to the field breeding facility for rice worms when a related breeding environment parameter deviation event occurs; and obtain the historical breeding environment parameters corresponding to the field breeding facility for rice worms when a non-related breeding environment parameter deviation event occurs.

[0078] The related aquaculture environment parameter shift event is used as the first condition node, and the historical aquaculture environment parameters corresponding to the field aquaculture facility of the rice worm at the time of the related aquaculture environment parameter shift event are used as the first random variable node; and each first condition node and the first random variable node are connected in a directed manner to form a first topology diagram.

[0079] The non-correlated aquaculture environment parameter shift event is used as the second condition node, and the historical aquaculture environment parameters corresponding to the field aquaculture facility of the rice worm at the time of the non-correlated aquaculture environment parameter shift event are used as the second random variable node; and each second condition node and the second random variable node are connected in a directed manner to form a second topology diagram.

[0080] Based on the first topological structure diagram, a first conditional probability matrix is ​​obtained for abnormal growth and activity events of worms caused by a correlated aquaculture environment parameter deviation event, and a second conditional probability matrix is ​​obtained for abnormal growth and activity events of worms caused by a non-correlated aquaculture environment parameter deviation event.

[0081] In this system, related events are designated as the first type of nodes, and unrelated events as the second type. Each event represents a conditional state. Edges represent the influence relationships between events. Each event node is connected to its corresponding historical environmental parameter node, indicating the change of a specific environmental parameter under that event. Initial probability values ​​are initialized for each event node and environmental parameter node; these values ​​can be set based on the frequency of historical data or prior knowledge. For each related event node, the joint probability distribution of all its environmental parameter nodes is calculated. Using Bayes' theorem, the probability of the event occurring is multiplied by the probability of abnormal growth activity after the environmental parameter change to obtain the probability of abnormal growth activity under a specific event. These probability values ​​are organized into a matrix to obtain the first conditional probability matrix. Similarly, for each unrelated event node, the joint probability distribution of all its environmental parameter nodes is calculated, and the probability of abnormal growth activity is calculated using Bayes' theorem and organized into a matrix to obtain the second conditional probability matrix.

[0082] A Bayesian network is constructed, and the first and second conditional probability matrices are imported into the Bayesian network for parameter learning. The posterior probability of the Bayesian network after each learning is obtained until the posterior probability converges to a preset probability value. The final learning parameters are then saved to obtain the abnormal probability prediction model of the growth activity of the rice worm.

[0083] It should be noted that historical data was collected from the field breeding facilities for *Sinocyclocheilus gramineus* during both correlated and uncorrelated environmental parameter shifts. This included specific environmental parameter values ​​such as temperature, humidity, and light intensity, and whether changes in these parameters triggered abnormal growth and activity in *Sinocyclocheilus gramineus*. Based on the collected data, correlated events were designated as the first type of nodes, and uncorrelated events as the second type. The environmental parameters of each event served as connection information between nodes, forming two independent Bayesian network topologies. Based on the first and second topologies, the probability distributions of abnormal growth and activity in *Sinocyclocheilus gramineus* during correlated and uncorrelated events were calculated, forming a first conditional probability matrix and a second conditional probability matrix. The first and second conditional probability matrices were integrated into the Bayesian network model, and the model parameters were adjusted through iterative learning until the posterior probability of the model stabilized at a preset value. This process essentially trains the model using historical data, enabling it to predict the impact of similar future events on the growth and activity of *Sinocyclocheilus gramineus*. By constructing a Bayesian network model, the impact of changes in different environmental parameters on the growth and activity of *Sinocyclocheilus gramineus* can be predicted, helping to take preventative measures to prevent abnormal growth and activity. The model's predictions provide a scientific basis for aquaculture decisions, such as adjusting environmental parameters to optimize the growth environment of the sandworm or intervening promptly when potential risks are identified. In summary, the Bayesian network-based prediction model provides an advanced risk management tool for large-scale sandworm farming, not only improving farming efficiency and product quality but also providing technical support for achieving eco-friendly aquaculture.

[0084] Preferably, real-time breeding environment parameters are obtained in the field breeding facility for a preset breeding period. Based on the predicted probability model of abnormal growth activity of the rice worms and combined with the real-time breeding environment parameters, the activity status of the rice worms in the preset future period is predicted and analyzed. Specifically:

[0085] The real-time breeding environment parameters of the field breeding facility for the rice worm are obtained within a preset breeding time period, and the real-time breeding environment parameters of the field breeding facility for the preset breeding time period are imported into the rice worm growth activity abnormality probability prediction model for prediction.

[0086] Obtain the prediction results of the abnormal growth activity probability prediction model of rice worm, and obtain the probability value of the occurrence of abnormal growth activity events of rice worm in the field breeding facility under the current real-time breeding environment parameter conditions based on the prediction results.

[0087] If the probability of an abnormal growth and activity event of rice worms in the field breeding facility caused by the current real-time breeding environment parameters is not greater than the preset probability value, it means that rice worms will not exhibit abnormal growth and activity within the preset future time period, and the first analysis result is generated.

[0088] If the probability of an abnormal growth and activity event of rice worms in the field breeding facility under the current real-time breeding environment parameters is greater than the preset probability value, it indicates that the rice worms will experience abnormal growth and activity within the preset future time period, and a second analysis result will be generated.

[0089] It should be noted that real-time environmental parameters of the field breeding facility for *Sinonovacula contorta*, such as temperature, humidity, light intensity, and pH, are continuously collected within a preset time period. These parameters directly affect the growth and activity of *Sinonovacula contorta*. The collected real-time environmental parameter data is input into a pre-trained prediction model for the probability of abnormal growth and activity in *Sinonovacula contorta*. This model learns the relationship between environmental parameters and abnormal growth and activity in historical data, and can predict the probability of abnormal growth and activity in *Sinonovacula contorta* under given environmental conditions. The model outputs the prediction result, providing the probability value of abnormal growth and activity events in *Sinonovacula contorta* within a preset future time period under the current real-time environmental parameters. Based on the prediction result, the system generates corresponding analysis results: if the prediction result indicates that the probability of abnormal growth and activity events in *Sinonovacula contorta* within the future time period is not greater than the preset probability value, the system generates a first analysis result, indicating that the current environmental conditions are suitable and abnormal growth and activity are not expected. Conversely, if the prediction result shows that the probability value is higher than the preset value, the system generates a second analysis result, warning of the potential risk of abnormal growth and activity in the future, requiring corresponding measures. Through the learning and prediction capabilities of the machine learning model, the growth and activity status of *Sinonovacula contorta* can be predicted more accurately, reducing uncertainty. The system can issue early warnings before problems occur, allowing for timely prevention and adjustments to avoid losses.

[0090] Preferably, if the result is the second analysis result, the real-time aquaculture environment parameters are imported into a pre-constructed semantic database for matching, and the abnormality type that indicates abnormal growth and activity of the rice worms in the rice worm field aquaculture facility within a preset future time period is identified, specifically:

[0091] Obtain the breeding record event book of the field breeding facility for rice worms, and extract the abnormal growth and activity events of rice worms that occurred in the field breeding facility during the historical operation based on the breeding record event book;

[0092] Obtain the abnormality type of each abnormal event in the growth activity of the rice worm, and obtain the historical breeding environment parameters of the rice worm field breeding facility within a preset time period before the occurrence of each abnormal event in the growth activity of the rice worm.

[0093] A semantic network is constructed, and several cutting nodes are determined based on the abnormality type of each abnormal event in the growth and activity of the rice worm. The semantic network is then cut according to each cutting node to obtain several branch nodes. Each branch node is assigned a corresponding abnormality type index according to the abnormality type corresponding to each cutting node.

[0094] The historical farming environment parameters of the field farming facilities for each type of abnormal growth activity of rice worms within a preset time period before the occurrence of each abnormal event are stored on the branch nodes after being assigned the corresponding abnormality type index, thus obtaining a semantic database.

[0095] If the result is the second analysis result, then the real-time breeding environment parameters within the preset breeding time period in the field breeding facility of rice worm are obtained, and the real-time breeding environment parameters are imported into the semantic database;

[0096] The Euclidean distance algorithm is introduced, and the similarity between the real-time aquaculture environment parameters and the historical aquaculture environment parameters on each branch node in the semantic database is calculated based on the Euclidean distance algorithm to obtain several similarity scores.

[0097] Sort out the maximum similarity among several similarities, obtain the branch node corresponding to the maximum similarity, and extract the anomaly type index assigned to the branch node corresponding to the maximum similarity.

[0098] Based on the extracted anomaly type index, the anomaly types that indicate abnormal growth and activity of rice worms in the field breeding facility within a preset future time period are determined.

[0099] It should be noted that, based on the aquaculture record event book, all abnormal events related to the growth and activity of *Sinocyclocheilus* (a type of worm) in the field aquaculture facility in the past are extracted. Then, an anomaly type is assigned to each abnormal event, and aquaculture environmental parameters for a period of time prior to the occurrence of the abnormal event are collected. A semantic network is constructed based on this information, where nodes represent anomaly types and edges represent relationships between anomaly types. The network is segmented by determining cutting nodes to form branch nodes, and a corresponding anomaly type index is assigned to each branch node. When the system detects a possible abnormal growth and activity situation (i.e., generates the second analysis result), it acquires the current real-time aquaculture environmental parameters. These real-time parameters are imported into the constructed semantic database. Then, the similarity between the real-time aquaculture environmental parameters and the historical environmental parameters recorded on each branch node in the database is calculated, and the branch node containing the historical environmental parameter most similar to the real-time parameter is identified. Based on the branch node containing the historical environmental parameter most similar to the real-time parameter, its corresponding anomaly type index is extracted. Finally, the index is used to determine the possible types of abnormal growth and activity of *Sinocyclocheilus* in the future within a preset future time period. By matching real-time data with historical data, the system can accurately predict possible future anomaly types, improving the accuracy and reliability of predictions. Automated prediction and response mechanisms reduce the need for manual monitoring and improve the level of intelligence in aquaculture management. Comprehensive analysis based on historical data and real-time monitoring provides a scientific basis for decision-making, optimizing aquaculture strategies and environmental control. In summary, by combining historical data mining, real-time data matching, and similarity calculation, efficient prediction of abnormal growth and activity in rice worms is achieved, providing crucial technical and decision support for field aquaculture.

[0100] Preferably, a recommended control plan is retrieved from the control log based on the type of abnormal growth and activity of rice worms in the rice worm field breeding facility within a preset future time period. The rice worm field breeding facility is then regulated and treated based on the recommended control plan, specifically as follows:

[0101] Search terms are generated based on the types of abnormalities that indicate abnormal growth and activity of rice worms in field farming facilities within a preset future time period.

[0102] Obtain the prevention and control logs of the field breeding facility for rice worms, and search the prevention and control logs based on the search terms to find the prevention and control measures corresponding to the abnormal growth and activity of rice worms during the breeding process.

[0103] And obtain the success rate of various prevention and control measures obtained from the search, and screen out prevention and control measures whose success rate is not greater than the preset success rate;

[0104] Obtain the energy consumption value of the remaining control measures, select the control measure with the lowest energy consumption value as the recommended control plan, and send the recommended control plan to the control terminal of the field breeding facility for rice worms;

[0105] Based on the recommended prevention and control plan, the field breeding facilities for rice worms were regulated and treated to prevent the rice worms from experiencing abnormal growth and activity.

[0106] The prevention and control measures include ventilation and heat dissipation, shed insulation, baiting, drug administration, and drainage.

[0107] It should be noted that, based on the predicted anomaly type, the system generates specific search keywords and then searches the prevention and control logs of the rice worm field breeding facility. The prevention and control logs record various prevention and control measures implemented during the past breeding process, including ventilation and heat dissipation, shed closure and insulation, feed distribution, pesticide application, and drainage. The system filters out effective prevention and control measures for specific anomaly types from the search results. This step also includes evaluating the success rate of these measures, eliminating measures with success rates below a preset standard to ensure the recommended prevention and control plan has high effectiveness. Among the remaining prevention and control measures, the system selects the measure with the lowest energy consumption as the recommended plan. Low energy consumption means lower implementation costs and reduced environmental impact, reflecting the concept of green agriculture. Finally, the system sends the recommended prevention and control plan to the control terminal of the rice worm field breeding facility, enabling the facility to adjust and process according to the recommended plan to prevent abnormal growth and activity of rice worms. By effectively preventing abnormal growth and activity, the impact of diseases and adverse environmental conditions on rice worms can be reduced, thereby improving yield and product quality. By predicting, screening, and optimizing prevention and control measures, intelligent management of the field breeding process of rice worms has been achieved, effectively preventing abnormal growth and activity, and improving the efficiency and sustainability of agricultural production.

[0108] In practical applications, this control method also includes the following steps:

[0109] The video frame information of the behavioral characteristics of the rice worm when it is in various abnormal growth and activity states is obtained through big data network, as well as the video frame information of the behavioral characteristics of the rice worm when it is in a healthy growth state.

[0110] A fuzzy evaluation algorithm is introduced, and the behavioral characteristics video frame information of the rice worm in a healthy growth state and the behavioral characteristics video frame information of the rice worm in various abnormal growth activity states are evaluated based on the fuzzy evaluation algorithm to obtain the membership degree of the behavioral characteristics of the rice worm in a healthy growth state and in each abnormal growth activity state.

[0111] Based on the membership degree of the behavioral characteristics of the rice worm in the healthy growth state and various abnormal growth activity states, a membership degree matrix is ​​constructed. A conditional random field model is introduced, and the membership degree matrix is ​​imported into the conditional random field model to obtain the conditional transition probability of the rice worm when it transitions from the healthy growth state to various abnormal growth activity states.

[0112] A conditional transition probability matrix is ​​constructed based on the conditional transition probabilities of the rice worm when it transitions from a healthy growth state to various abnormal growth activity states.

[0113] A state transition probability prediction model is constructed based on a deep learning network. The conditional transition probability matrix is ​​imported into the state transition probability prediction model for cross-learning training until the loss function value is less than a preset value. Then, the state transition probability prediction model for the growth of rice worms is output.

[0114] The real-time behavioral characteristics video frame information of the rice worms inside the rice worm breeding facility is obtained. The real-time behavioral characteristics video frame information is evaluated based on the fuzzy evaluation algorithm to obtain the membership degree of the real-time behavioral characteristics of the rice worms at the current timestamp.

[0115] The membership degree of the real-time behavioral features of the rice worm at the current timestamp is imported into the rice worm growth state transition probability prediction model for prediction, so as to obtain the transition probability of the rice worm to the abnormal growth activity state at the current timestamp.

[0116] If the probability of a sandworm migrating to an abnormal growth state at the current timestamp is not greater than a preset probability value, then a first analysis result is generated; if the probability of a sandworm migrating to an abnormal growth state at the current timestamp is greater than a preset probability value, then a second analysis result is generated.

[0117] It should be noted that, firstly, a large amount of video frame information on the behavioral characteristics of sandworms in different growth states (healthy and abnormal) is acquired through a big data network. This data includes, but is not limited to, movement frequency, activity range, and interaction patterns with companions, providing a foundation for subsequent analysis. A fuzzy evaluation algorithm is introduced to process the collected video frame information, quantifying the behavioral characteristics of sandworms in healthy and abnormal states. The algorithm calculates the membership degree of behavioral features, i.e., the degree of behavioral features in healthy and abnormal states, forming a membership degree matrix. A conditional random field model is used to further analyze the membership degree matrix, estimating the probability of sandworms transitioning from a healthy state to different abnormal states. This model considers time series characteristics and can capture the dependencies between states, providing a more accurate probability estimate for prediction. A state transition probability prediction model is constructed based on a deep learning network. Through cross-learning training, the model parameters are optimized until the loss function value meets the preset standard. This model can predict the probability of sandworms transitioning from a healthy state to a specific abnormal state at any given time. The system receives video frame information on the behavioral characteristics of sandworms inside the breeding facility in real time and quickly calculates the membership degree of the sandworms' behavioral features at the current moment using the fuzzy evaluation algorithm. The membership degree is input into the prediction model to predict the probability of the rice worm transitioning to an abnormal state at the current moment. Based on the prediction results, the system generates different analysis results. If the transition probability is lower than a preset threshold, it indicates that the current state is safe, and the first analysis result is generated, prompting the farmer to continue maintaining the existing management strategy. Conversely, if the transition probability is higher than the preset threshold, the system considers there to be a high risk, and the second analysis result is generated, suggesting preventive measures or adjustments to the management strategy to avoid or mitigate possible abnormal growth activity. Overall, by integrating modern technologies such as big data, deep learning, and fuzzy logic, intelligent monitoring and prediction of the rice worm's growth status are achieved. The aim is to monitor and predict the changes in the rice worm's health status during its growth process in real time, especially predicting its transition to an abnormal growth activity state, enabling refined analysis and prediction of rice worm behavior patterns, thereby assisting in aquaculture management decisions and improving aquaculture efficiency and biological health management levels.

[0118] It should be noted that fuzzy evaluation algorithms can transform video frame information describing the behavior of rice worms into fuzzy sets. For example, descriptions such as "fast movement speed," "straight movement trajectory," and "bright body color" can be converted into fuzzy linguistic variables and assigned corresponding membership degrees. By comparing the fuzzy sets in healthy and abnormal growth states, the differences in the behavioral characteristics of rice worms in different growth states can be identified. For example, in a healthy growth state, rice worms move quickly and have a straight trajectory, while in an abnormal state, they move slowly and have a curved trajectory.

[0119] In practical applications, this control method also includes the following steps:

[0120] Data on the dissolution rate of worm bait under various environmental parameters is obtained through a big data network. A graph neural network is then introduced, with each environmental parameter serving as the first graph node and the data on the dissolution rate of worm bait under each environmental parameter serving as the second graph node.

[0121] Calculate the Manhattan distance between the first graph node and the second graph node, connect two graph nodes whose Manhattan distance is less than a threshold to form an edge, and obtain the topology graph of environmental parameters-bait dissolution rate; where the nodes of the topology graph represent environmental parameters and dissolution rate data, and the edges represent the relationship between them;

[0122] Obtain the adjacency matrix of the topology graph, construct a prediction model for the dissolution rate of worm bait based on a deep learning network, import the adjacency matrix into the prediction model for the dissolution rate of worm bait for learning and training, and obtain the trained prediction model for the dissolution rate of worm bait after training.

[0123] The feeding location area of ​​the rice worm field breeding facility is obtained, and the feeding location area is divided into several sub-areas. Before feeding each sub-area, the real-time environmental parameters of each sub-area are obtained through the rice worm field breeding facility.

[0124] The real-time environmental parameters of each sub-region are imported into the trained rice worm bait dissolution rate prediction model for prediction, and the predicted dissolution rate of rice worm bait in each sub-region is obtained.

[0125] If the predicted dissolution rate of the worm bait in a certain sub-region is greater than the preset dissolution rate, the amount of bait fed in that sub-region will be adjusted based on the predicted dissolution rate of the worm bait in that sub-region.

[0126] It should be noted that data on the dissolution rate of sandworm bait under different environmental parameters was collected through a big data network. Then, a graph neural network was used to construct a topology graph, where environmental parameters served as the first-layer nodes, and the sandworm bait dissolution rate data under each environmental parameter condition served as the second-layer nodes. By calculating the Manhattan distance between these nodes, nodes with a distance less than a certain threshold were connected, forming a topology graph relating environmental parameters and dissolution rate. This topology graph helps reveal how environmental parameters affect the dissolution rate of sandworm bait. Next, by obtaining the adjacency matrix of the topology graph, a deep learning network model was constructed to predict the dissolution rate of sandworm bait. This model was trained and learned by studying the relationship between environmental parameters and dissolution rate in the adjacency matrix. After training, the model was able to predict the dissolution rate of sandworm bait under given environmental parameter conditions.

[0127] In practical applications, the system acquires real-time environmental parameters from different areas within the field breeding facility for *Sinocypriniformes* (a type of worm) and inputs them into a trained prediction model. The model then predicts the feed dissolution rate for each sub-region. If the predicted dissolution rate for a sub-region exceeds a pre-set threshold, the system adjusts the feed amount based on that region's predicted dissolution rate. This dynamic adjustment effectively prevents nutrient loss due to rapid feed dissolution, improves feed utilization, and optimizes the growth environment and breeding efficiency of *Sinocypriniformes*. Overall, this method, utilizing graph neural networks and deep learning techniques to predict environmental parameters and feed dissolution rate in *Sinocypriniformes*, aims to optimize feed management during field breeding of *Sinocypriniformes* through accurate prediction and dynamic adjustment, thereby improving breeding efficiency and the health of the organisms.

[0128] like Figure 3 As shown, the field breeding facility for rice worms includes:

[0129] The main structure of the greenhouse, number 1101, is built on a paddy field.

[0130] The water supply and drainage system module 1102 is used to regulate the water flow speed and water level in the paddy field;

[0131] Feeding system module 1103 is used to feed rice worms at regular intervals and in fixed quantities;

[0132] Environmental control module 1104 is used to control aquaculture environmental parameters;

[0133] The data acquisition and analysis module collects various types of data and performs analysis and processing.

[0134] The automatic control module connects all the modules to achieve automated control of the breeding process.

[0135] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for controlling the rearing of rice worms in a field, characterized in that, Includes the following steps: Obtain the breeding record event book of the field breeding facility for rice worms, and analyze the breeding environment parameter deviation events based on the breeding record event book to obtain the related breeding environment parameter deviation events or the unrelated breeding environment parameter deviation events. Obtain the historical aquaculture environment parameters corresponding to the field aquaculture facilities of *Sinocyclocheilus scoparia* when a correlated or uncorrelated aquaculture environment parameter shift event occurs. Construct a probability prediction model for abnormal growth and activity of *Sinocyclocheilus scoparia* based on the historical aquaculture environment parameters and in conjunction with a Bayesian network. The real-time breeding environment parameters of the rice worm in the field breeding facility within a preset breeding time period are obtained, and the activity status of the rice worm in the preset future time period is predicted and analyzed based on the rice worm growth activity abnormality probability prediction model and the real-time breeding environment parameters. If the result is the first analysis result, no control measures will be taken for the field breeding facilities of rice worms; If the result is the second analysis result, the real-time breeding environment parameters are imported into the pre-built semantic database for matching, and the abnormal type of abnormal growth and activity of the rice worms in the field breeding facility will be matched within a preset future time period. Based on the abnormal types of growth and activity abnormalities that are expected to occur in the field breeding facilities of rice worms within a preset future time period, recommended control plans are retrieved from the control log, and the field breeding facilities of rice worms are regulated and treated according to the recommended control plans.

2. The control method for a field breeding facility for rice worms according to claim 1, characterized in that, Obtain the breeding record event book of the rice worm field breeding facility, and analyze the breeding environment parameter deviation events based on the breeding record event book to obtain the correlated breeding environment parameter deviation events or the uncorrelated breeding environment parameter deviation events, specifically: Obtain the breeding record event book of the field breeding facility for rice worms, and extract the breeding environment parameter offset events of the field breeding facility for rice worms in the historical working process based on the breeding record event book; Based on the aforementioned breeding record event book, the number of abnormal growth and activity events of rice worms caused by various breeding environment parameter deviation events in the rice worm field breeding facility was further counted. The abnormal probability value of the abnormal growth and activity of the rice worm caused by the occurrence of various deviation events of breeding environmental parameters in the rice worm field breeding facility is calculated. Events with aquaculture environment parameter offsets whose abnormal probability values ​​exceed a preset threshold are marked as related aquaculture environment parameter offset events; Events with aquaculture environment parameter offsets whose abnormal probability values ​​are not greater than a preset threshold are marked as non-associated aquaculture environment parameter offset events.

3. The control method for a field breeding facility for rice worms according to claim 1, characterized in that, The historical aquaculture environment parameters corresponding to the field aquaculture facilities of *Sinocypripedium spp.* at the time of occurrence of correlated or uncorrelated aquaculture environment parameter shift events are obtained. Based on these historical aquaculture environment parameters and combined with a Bayesian network, a probability prediction model for abnormal growth and activity of *Sinocypripedium spp.* is constructed, specifically as follows: Obtain the breeding record event book of the field breeding facility for rice worms; based on the breeding record event book, obtain the historical breeding environment parameters corresponding to the field breeding facility for rice worms when a related breeding environment parameter deviation event occurs; and obtain the historical breeding environment parameters corresponding to the field breeding facility for rice worms when a non-related breeding environment parameter deviation event occurs. The related aquaculture environment parameter shift event is used as the first condition node, and the historical aquaculture environment parameters corresponding to the field aquaculture facility of the rice worm at the time of the related aquaculture environment parameter shift event are used as the first random variable node; and each first condition node and the first random variable node are connected in a directed manner to form a first topology diagram. The non-correlated aquaculture environment parameter shift event is used as the second condition node, and the historical aquaculture environment parameters corresponding to the field aquaculture facility of the rice worm at the time of the non-correlated aquaculture environment parameter shift event are used as the second random variable node; and each second condition node and the second random variable node are connected in a directed manner to form a second topology diagram. Based on the first topological structure diagram, a first conditional probability matrix is ​​obtained for abnormal growth and activity events of worms caused by a correlated aquaculture environment parameter deviation event, and a second conditional probability matrix is ​​obtained for abnormal growth and activity events of worms caused by a non-correlated aquaculture environment parameter deviation event. A Bayesian network is constructed, and the first and second conditional probability matrices are imported into the Bayesian network for parameter learning. The posterior probability of the Bayesian network after each learning is obtained until the posterior probability converges to a preset probability value. The final learning parameters are then saved to obtain the abnormal probability prediction model of the growth activity of the rice worm.

4. The control method for a field breeding facility for rice worms according to claim 1, characterized in that, Real-time breeding environment parameters are obtained in the field breeding facility for a preset breeding period of *Sinocyprinus spp.* (a type of worm). Based on the *Sinocyprinus spp.* growth activity abnormality probability prediction model and combined with the real-time breeding environment parameters, the activity status of *Sinocyprinus spp.* in the preset future time period is predicted and analyzed. Specifically: The real-time breeding environment parameters of the field breeding facility for the rice worm are obtained within a preset breeding time period, and the real-time breeding environment parameters of the field breeding facility for the preset breeding time period are imported into the rice worm growth activity abnormality probability prediction model for prediction. Obtain the prediction results of the abnormal growth activity probability prediction model of rice worm, and obtain the probability value of the occurrence of abnormal growth activity events of rice worm in the field breeding facility under the current real-time breeding environment parameter conditions based on the prediction results. If the probability of an abnormal growth and activity event of rice worms in the field breeding facility caused by the current real-time breeding environment parameters is not greater than the preset probability value, it means that rice worms will not exhibit abnormal growth and activity within the preset future time period, and the first analysis result is generated. If the probability of an abnormal growth and activity event of rice worms in the field breeding facility under the current real-time breeding environment parameters is greater than the preset probability value, it indicates that the rice worms will experience abnormal growth and activity within the preset future time period, and a second analysis result will be generated.

5. The control method for a field breeding facility for rice worms according to claim 1, characterized in that, If the result is the second analysis result, the real-time aquaculture environment parameters are imported into a pre-built semantic database for matching. The matching identifies the abnormal types in which the rice worms in the field aquaculture facility will exhibit abnormal growth and activity within a preset future time period, specifically: Obtain the breeding record event book of the field breeding facility for rice worms, and extract the abnormal growth and activity events of rice worms that occurred in the field breeding facility during the historical operation of the breeding facility based on the breeding record event book; Obtain the abnormality type of each abnormal event in the growth activity of the rice worm, and obtain the historical breeding environment parameters of the rice worm field breeding facility within a preset time period before the occurrence of each abnormal event in the growth activity of the rice worm. A semantic network is constructed, and several cutting nodes are determined based on the abnormality type of each abnormal event in the growth and activity of the rice worm. The semantic network is then cut according to each cutting node to obtain several branch nodes. Each branch node is assigned a corresponding abnormality type index according to the abnormality type corresponding to each cutting node. The historical farming environment parameters of the field farming facilities for each type of rice worm growth activity abnormality event within a preset time period before the occurrence of each rice worm growth activity abnormality event are stored on the branch nodes after being assigned the corresponding abnormality type index, thus obtaining a semantic database. If the result is the second analysis result, then the real-time breeding environment parameters within the preset breeding time period in the field breeding facility of rice worm are obtained, and the real-time breeding environment parameters are imported into the semantic database; The Euclidean distance algorithm is introduced, and the similarity between the real-time aquaculture environment parameters and the historical aquaculture environment parameters on each branch node in the semantic database is calculated based on the Euclidean distance algorithm to obtain several similarity scores. Sort out the maximum similarity among several similarities, obtain the branch node corresponding to the maximum similarity, and extract the anomaly type index assigned to the branch node corresponding to the maximum similarity. Based on the extracted anomaly type index, the anomaly types that indicate abnormal growth and activity of rice worms in the field breeding facility within a preset future time period are determined.

6. The control method for a field breeding facility for rice worms according to claim 1, characterized in that, Based on the types of abnormal growth and activity that are expected to occur in the field breeding facilities of rice worms within a preset future time period, recommended control measures are retrieved from the control log. Based on these recommended control measures, the field breeding facilities of rice worms are then regulated and treated, specifically as follows: Search terms are generated based on the types of abnormalities that indicate abnormal growth and activity of rice worms in field farming facilities within a preset future time period. Obtain the prevention and control logs of the field breeding facility for rice worms, and search the prevention and control logs based on the search terms to find the corresponding prevention and control measures for abnormal growth and activity of rice worms during the breeding process. And obtain the success rate of various prevention and control measures obtained from the search, and screen out prevention and control measures whose success rate is not greater than the preset success rate; Obtain the energy consumption value of the remaining control measures, select the control measure with the lowest energy consumption value as the recommended control plan, and send the recommended control plan to the control terminal of the field breeding facility for rice worms; Based on the recommended prevention and control plan, the field breeding facilities for rice worms were regulated and treated to prevent the rice worms from experiencing abnormal growth and activity. The prevention and control measures include ventilation and heat dissipation, shed insulation, baiting, drug administration, and drainage.

7. The control method for a field breeding facility for rice worms according to claim 1, characterized in that: The aquaculture environment parameters include pH value, dissolved oxygen concentration, ammonia nitrogen concentration, nitrite concentration, temperature, humidity, and light intensity.

8. The control method for a field breeding facility for rice worms according to claim 1, characterized in that: The aforementioned aquaculture environment parameter offset events are one or more combinations of pH value parameter offset, dissolved oxygen concentration parameter offset, ammonia nitrogen concentration parameter offset, nitrite concentration parameter offset, temperature parameter offset, humidity parameter offset, and light intensity parameter offset.

9. The control method for a field breeding facility for rice worms according to claim 1, characterized in that: The abnormal growth and activity events of the rice worm include abnormal growth rate, abnormal mortality rate, abnormal activity, abnormal morphology, abnormal reproductive capacity, and abnormal behavior.

10. A control method for a field breeding facility for rice worms as described in any one of claims 1 to 9, characterized in that, The field breeding facilities for rice worms include: The main structure of the greenhouse is built on a paddy field; The water supply and drainage system module is used to regulate the water flow speed and water level in paddy fields; The feeding system module is used to feed the rice worms at regular intervals and in fixed quantities. The environmental control module is used to control the parameters of the aquaculture environment; The data acquisition and analysis module collects various types of data and performs analysis and processing. The automatic control module connects all the modules to achieve automated control of the breeding process.

Images