Method and system for dynamically monitoring and optimizing incubation of monopterus albus fries in indoor environment

Abstract

The invention relates to the technical field of data processing, and discloses a method and a system for dynamically monitoring and optimizing incubation of monopterus albus fries in an indoor environment. The method comprises the following steps: performing staged nutritional requirement analysis on fries of ricefield eels to obtain a multi-dimensional nutritional supplement scheme; carrying out microbial community structure analysis on the hatching water body to obtain a microbial regulation and control scheme; performing multi-parameter real-time monitoring on the hatching water body to obtain a water quality dynamic adjusting strategy; performing environment optimization on the hatching water body, performing multi-dimensional data acquisition and analysis on the monopterus albus fries, obtaining a behavior characteristic analysis result, performing multi-source stress factor exposure experiment scheme analysis, generating a multi-source stress factor exposure experiment scheme, and performing stress experiment on the monopterus albus fries through the multi-source stress factor exposure experiment scheme. Obtaining a stress response database; and generating a stress management scheme according to the stress response database. According to the invention, the efficiency and accuracy of dynamic monitoring and optimization of the incubation of the monopterus albus fries in the indoor environment are improved.

Description

Technical Field

[0001] The present application relates to the technical field of data processing, and particularly to a method and system for dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment. Background Art

[0002] At present, certain progress has been made in the indoor eel larva hatching technology. Existing technologies usually use equipment such as constant temperature water tanks, dissolved oxygen equipment, and pH adjustment devices to control the hatching environment parameters, and monitor the growth status of the larvae through manual observation and regular sampling. In terms of nutrient supply, the nutritional needs of the larvae are mainly met by relying on empirical formulas and fixed feeding plans. Water quality management mainly relies on regular water changes and the use of biological agents to maintain.

[0003] However, these traditional methods have some limitations. First, the control of environmental parameters is often static and single, and it is difficult to adapt to the dynamic demand changes during the growth process of the larvae. Second, the methods of manual observation and regular sampling are difficult to comprehensively and real-time grasp the growth status and behavioral characteristics of the larvae. Moreover, the fixed nutrient supply scheme lacks pertinence and flexibility, and it is difficult to meet the precise nutritional needs of the larvae at different growth stages. In addition, the traditional water quality management methods have limited ability to regulate the water body microecosystem, and it is difficult to effectively prevent and respond to various potential stress factors. Summary of the Invention

[0004] The present application provides a method and system for dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment, which is used to improve the efficiency and accuracy of dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment.

[0005] In a first aspect, the present application provides a method for dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment, and the method for dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment includes: performing multi-factor collaborative optimization processing on the pre-collected hatching environment parameters to obtain a dynamic regulation parameter set;

[0006] Based on the dynamic regulation parameter set, performing a phased nutritional requirement analysis on the eel larvae to obtain a multi-dimensional nutritional supplement plan;

[0007] According to the dynamic regulation parameter set and the multi-dimensional nutritional supplement plan, performing a microbial community structure analysis on the hatching water body to obtain a microbial regulation plan;

[0008] Based on the dynamic regulation parameter set, the multi-dimensional nutritional supplement plan, and the microbial regulation plan, performing multi-parameter real-time monitoring on the hatching water body to obtain a water quality dynamic regulation strategy, wherein the water quality dynamic regulation strategy includes: a hierarchical early warning threshold and an adjustment measure selection matrix;

[0009] Based on the water quality dynamic regulation strategy, optimize the environment of the hatching water body, and in the optimized hatching water body, collect and analyze multi-dimensional data of the eel seedlings to obtain the analysis results of behavioral characteristics;

[0010] Analyze the experimental scheme of multi-source stress factor exposure for the analysis results of the behavioral characteristics, generate an experimental scheme for multi-source stress factor exposure, and conduct a stress experiment on the eel seedlings through the experimental scheme for multi-source stress factor exposure to obtain a stress response database;

[0011] Generate a stress management plan according to the stress response database, wherein the stress management plan includes: stress source classification data, warning indicators, and a set of mitigation measures.

[0012] Combined with the first aspect, in the first implementation manner of the first aspect of the present application, the hatching environment parameters include: water temperature, dissolved oxygen, pH value, light intensity, salinity time series change data, hardness time series change data, and the correlation coefficient matrix between each parameter. The multi-factor collaborative optimization process is performed on the pre-collected hatching environment parameters to obtain a dynamic regulation parameter set, including:

[0013] Perform time series decomposition on the water temperature, dissolved oxygen, pH value, light intensity, salinity time series change data, and hardness historical data of the pre-collected hatching environment parameters to obtain the trend term, seasonal term, and random term of each parameter;

[0014] Perform polynomial fitting on the trend term of each parameter to obtain the long-term change trend function of each parameter;

[0015] Perform Fourier analysis on the seasonal term of each parameter to obtain the periodic change characteristics of each parameter;

[0016] Perform autocorrelation analysis on the random term of each parameter to obtain the short-term fluctuation characteristics of each parameter;

[0017] Based on the long-term change trend function of each parameter, the periodic change characteristics of each parameter, and the short-term fluctuation characteristics of each parameter, construct a time series prediction model for each parameter to obtain the predicted time series data of each parameter;

[0018] Perform correlation analysis on the predicted time series data of each parameter to obtain the correlation coefficient matrix between parameters;

[0019] According to the correlation coefficient matrix between the parameters, perform principal component analysis on each parameter to obtain the main influencing factors and their weights, and based on the main influencing factors and their weights, perform weighted combination on the predicted time series data of each parameter to obtain a comprehensive environment index;

[0020] According to the comprehensive environment index, perform collaborative optimization on each parameter through a dynamic programming algorithm to obtain a dynamic regulation parameter set.

[0021] Combined with the first aspect, in the second implementation manner of the first aspect of the present application, the multi-dimensional nutritional supplement plan includes: a correspondence table of growth stage, bait type, feeding frequency, proportion of nutritional components, and addition amount of trace elements, and the influence weight of each nutritional component on the growth index of seedlings. Based on the dynamic regulation parameter set, the phased nutritional requirements of rice field eel seedlings are analyzed to obtain a multi-dimensional nutritional supplement plan, including:

[0022] Segment the water temperature, dissolved oxygen, and pH value data in the hatching environment parameters to obtain the growth stage division standard of rice field eel seedlings;

[0023] According to the growth stage division standard, segment the growth cycle of rice field eel seedlings to obtain the time range of each growth stage, and conduct morphological measurements on the rice field eel seedlings in each growth stage to obtain data on body length, body weight, and development degree of feeding organs;

[0024] According to the data on body length, body weight, and development degree of feeding organs, conduct bait palatability analysis on the rice field eel seedlings in each growth stage to obtain the target bait type for each stage;

[0025] Conduct nutritional component analysis on the target bait type for each stage to obtain the trace element content data of each bait type, and calculate the proportion of nutritional components in the bait formula according to the trace element content data of each bait type and the nutritional requirements of rice field eel seedlings in each growth stage to obtain the proportion of nutritional components in each growth stage;

[0026] Collect the feeding behavior of rice field eel seedlings in each growth stage to obtain the target feeding frequency and single feeding amount, and calculate the addition amount of trace elements according to the proportion of nutritional components and single feeding amount in each growth stage to obtain the addition amount of trace elements in each growth stage;

[0027] Measure the growth indexes of rice field eel seedlings in each growth stage to obtain data on body length growth rate, body weight growth rate, and survival rate, and conduct correlation analysis between the data on body length growth rate, body weight growth rate, and survival rate and nutritional components through multiple regression analysis to obtain the influence weight of each nutritional component on the growth index of seedlings, and generate a multi-dimensional nutritional supplement plan according to the influence weight of each nutritional component on the growth index of seedlings and the addition amount of trace elements in each growth stage.

[0028] Combined with the first aspect, in the third implementation manner of the first aspect of the present application, according to the dynamic regulation parameter set and the multi-dimensional nutritional supplement plan, conduct microbial community structure analysis on the hatching water body to obtain a microbial regulation plan, including:

[0029] Sample the hatching water body, extract and amplify microbial DNA from the water sample through high-throughput sequencing technology to obtain the original sequence data of the microbial community, and perform quality control and chimera removal on the original sequence data to obtain the effective sequence data;

[0030] Perform operational taxonomic unit clustering analysis on the effective sequence data to obtain the species composition and abundance information of the microbial community;

[0031] According to the species composition and abundance information, calculate the α-diversity index and β-diversity index to obtain the diversity evaluation result of the microbial community, and perform redundancy analysis (RDA) on the microbial community structure and environmental factors according to the diversity evaluation result and the dynamic regulation parameter set to obtain the influence degree of key environmental factors on the microbial community structure;

[0032] According to the influence degree and the multi-dimensional nutrient supplement plan, screen out the microbial species beneficial to the growth of eel seedlings to obtain the target microbial flora combination;

[0033] Conduct a culture experiment on the target microbial flora combination, measure its growth curve and metabolites to obtain the target culture conditions and yield data, and design the formula and production process of the microbial preparation according to the target culture conditions and yield data to obtain the microbial preparation sample;

[0034] Conduct stability and activity tests on the microbial preparation sample to obtain the storage conditions and shelf life of the preparation, and design the delivery strategy of the microbial preparation according to the storage conditions and shelf life. The delivery strategy includes the delivery time series and dosage, and generate the microbial regulation plan through the delivery strategy.

[0035] Combined with the first aspect, in the fourth implementation manner of the first aspect of the present application, based on the dynamic regulation parameter set, the multi-dimensional nutrient supplement plan and the microbial regulation plan, perform multi-parameter real-time monitoring on the hatching water body to obtain a water quality dynamic regulation strategy, where the water quality dynamic regulation strategy includes: hierarchical warning thresholds and a regulation measure selection matrix, including:

[0036] Continuously sample and detect the dissolved oxygen, pH value, ammonia nitrogen, nitrite and nitrate contents in the hatching water body to obtain the time series data of water quality parameters, and perform trend analysis on the time series data of water quality parameters to obtain the change trends and fluctuation ranges of each water quality parameter;

[0037] According to the changing trends and fluctuation ranges of the various parameters, multi-level warning thresholds are set for each water quality parameter through the water quality index data in the dynamic regulation parameter set, a hierarchical warning threshold table is obtained, and correlation analysis is performed on the thresholds at each level in the hierarchical warning threshold table and the multi-dimensional nutrition supplement plan to obtain the interaction relationship between each water quality parameter and bait feeding;

[0038] According to the interaction relationship between each water quality parameter and bait feeding and the microbial regulation plan, the adjustment measures under different water quality conditions are classified and sorted to obtain a preliminary adjustment measure library, and an effect evaluation experiment is carried out on each measure in the preliminary adjustment measure library to obtain the response time and influence degree data of each adjustment measure;

[0039] According to the response time and influence degree data of each adjustment measure, the adjustment measures are sorted by priority and combined and optimized to obtain an adjustment measure selection matrix for different water quality conditions, and the hierarchical warning threshold table and the adjustment measure selection matrix are fused with data to obtain the water quality dynamic regulation strategy.

[0040] Combined with the first aspect, in the fifth implementation manner of the first aspect of the present application, based on the water quality dynamic regulation strategy, the environment of the hatching water body is optimized, and in the optimized hatching water body, multi-dimensional data of the eel seedlings are collected and analyzed to obtain the analysis results of behavioral characteristics, including:

[0041] The water quality parameters of the hatching water body are adjusted to obtain a target water body that meets the requirements of the hierarchical warning threshold, and in the target water body, high-definition video recording of the eel seedlings is carried out to obtain the original data of the seedling swimming trajectories;

[0042] Image processing is performed on the original data of the seedling swimming trajectories to obtain the motion parameters of the seedling individuals, including swimming speed, turning frequency, and activity range, and statistical analysis is performed on the motion parameters to obtain the group behavior characteristics, including aggregation degree, distribution uniformity, and overall activity;

[0043] The feeding behavior of the eel seedlings in the target water body is collected and analyzed to obtain the feeding behavior data, and time series analysis is performed on the feeding behavior data to obtain the feeding rhythm and satiation degree change curves;

[0044] The sound signals of the eel seedlings are collected through an underwater microphone, the sound signals are subjected to spectral analysis to obtain an acoustic characteristic spectrum, and multi-dimensional characteristic analysis is performed on the motion parameters, group behavior characteristics, feeding behavior data, and acoustic characteristic spectrum to obtain a multi-dimensional behavior characteristic data set;

[0045] Perform cluster analysis on the multi-dimensional behavioral feature dataset, classify normal, sub-healthy, and disease states to obtain the behavioral state classification result, and perform correlation analysis between the behavioral state classification result and physiological indicators to obtain the behavioral feature analysis result.

[0046] Combined with the first aspect, in the sixth implementation manner of the first aspect of the present application, perform analysis on the multi-source stress factor exposure experimental plan for the behavioral feature analysis result, generate a multi-source stress factor exposure experimental plan, and perform a stress experiment on the eel seedlings through the multi-source stress factor exposure experimental plan to obtain a stress response database, including:

[0047] Perform cluster analysis on the behavioral feature analysis result to obtain the typical behavioral patterns of eel seedlings, and screen potential stress factors according to the typical behavioral patterns to obtain a list of candidate stress factors;

[0048] Classify the factors in the list of candidate stress factors to obtain three types of stress sources: physical, chemical, and biological, and design intensity gradients for the three types of physical, chemical, and biological stress sources to obtain a multi-level stress intensity plan;

[0049] According to the multi-level stress intensity plan, construct an exposure experimental matrix of single-factor and multi-factor combinations to obtain a stress factor exposure experimental plan, and through the stress factor exposure experimental plan, conduct grouped experiments on eel seedlings, record the behavioral changes, physiological indicators, and survival rates of the seedlings to obtain the original stress response data;

[0050] Perform statistical analysis on the original stress response data to obtain the influence degree and threshold of each stress factor on the seedlings, and establish a quantitative relationship between the stress factors and the seedling growth parameters according to the influence degree and threshold of each stress factor on the seedlings to obtain a stress response mapping table;

[0051] Perform simulation mapping analysis on the stress response mapping table to obtain the stress response database.

[0052] Combined with the first aspect, in the seventh implementation manner of the first aspect of the present application, generate a stress management plan according to the stress response database, where the stress management plan includes: stress source classification data, warning indicators, and a set of mitigation measures, including:

[0053] Perform hierarchical cluster analysis on the stress factors in the stress response database to obtain stress source classification data, and extract characteristic indicators of each type of stress source according to the stress source classification data to obtain a stress source identification feature set;

[0054] Perform principal component analysis on the stressor recognition feature set, screen out key early warning indicators, obtain an early warning indicator system, and design multi-level early warning thresholds according to the early warning indicator system to obtain the early warning indicators;

[0055] Conduct effectiveness evaluation on the stress response database to obtain a list of effective mitigation measures for various stressors, and construct coping strategies for different stressors and intensities according to the list of effective mitigation measures to obtain a mitigation measure matrix;

[0056] Perform similarity matching of mitigation measures on the mitigation measure matrix to obtain the set of mitigation measures.

[0057] In a second aspect, the present application provides a dynamic monitoring and optimization system for the hatching of eel larvae in an indoor environment. The dynamic monitoring and optimization system for the hatching of eel larvae in an indoor environment includes:

[0058] An optimization processing module for performing multi-factor collaborative optimization processing on the pre-collected hatching environment parameters to obtain a set of dynamic regulation parameters;

[0059] A requirement analysis module for performing phased nutritional requirement analysis on eel larvae based on the set of dynamic regulation parameters to obtain a multi-dimensional nutritional supplement plan;

[0060] A structure analysis module for performing microbial community structure analysis on the hatching water body according to the set of dynamic regulation parameters and the multi-dimensional nutritional supplement plan to obtain a microbial regulation plan;

[0061] A real-time monitoring module for performing multi-parameter real-time monitoring on the hatching water body based on the set of dynamic regulation parameters, the multi-dimensional nutritional supplement plan, and the microbial regulation plan to obtain a water quality dynamic regulation strategy, where the water quality dynamic regulation strategy includes: hierarchical early warning thresholds and a regulation measure selection matrix;

[0062] An environment optimization module for optimizing the hatching water body based on the water quality dynamic regulation strategy, and performing multi-dimensional data collection and analysis on eel larvae in the optimized hatching water body to obtain a behavioral feature analysis result;

[0063] A scheme analysis module for performing multi-source stress factor exposure experiment scheme analysis on the behavioral feature analysis result, generating a multi-source stress factor exposure experiment scheme, and performing a stress experiment on eel larvae through the multi-source stress factor exposure experiment scheme to obtain a stress response database;

[0064] A scheme generation module for generating a stress management scheme according to the stress response database, where the stress management scheme includes: stressor classification data, early warning indicators, and a set of mitigation measures.

[0065] In the technical solution provided by this application, through multi-factor collaborative optimization processing of the pre-collected hatching environment parameters, a dynamic regulation parameter set is obtained, realizing the precise control and dynamic adjustment of the hatching environment, effectively avoiding the problem of mismatched environmental parameters caused by traditional static control methods, and providing a more stable and suitable growth environment for yellow eel seedlings; based on the dynamic regulation parameter set, a phased nutritional requirement analysis is carried out on yellow eel seedlings, and a multi-dimensional nutritional supplement plan is obtained, making the nutritional supply more accurate and personalized, effectively improving the bait utilization rate, and promoting the healthy growth of seedlings; through the analysis of the microbial community structure of the hatching water body, a microbial regulation plan is obtained, realizing the precise regulation of the water body microecosystem, improving the water quality stability and self-purification ability, and creating a better microenvironment for the growth of seedlings; based on the dynamic regulation parameter set, the multi-dimensional nutritional supplement plan and the microbial regulation plan, multi-parameter real-time monitoring of the hatching water body is carried out, and a water quality dynamic regulation strategy is obtained, including a hierarchical early warning threshold and a regulation measure selection matrix, making the water quality management more proactive and precise, being able to timely prevent and respond to various water quality problems, and ensuring the stability of the seedling growth environment; through multi-dimensional data collection and analysis of yellow eel seedlings in the optimized hatching water body, an analysis result of behavioral characteristics is obtained, realizing the comprehensive and real-time monitoring of the growth status and behavioral characteristics of seedlings, and providing reliable data support for subsequent management decisions; an analysis of a multi-source stress factor exposure experimental plan is carried out on the analysis result of behavioral characteristics, and a stress response database is obtained through experiments, systematically revealing the influence mechanism of various potential stress factors on yellow eel seedlings, and providing a scientific basis for stress prevention and control; finally, a stress management plan formulated according to the stress response database, including stress source classification data, early warning indicators and a set of mitigation measures, realizes the precise identification, timely early warning and effective mitigation of various stress factors, greatly improving the stress resistance and survival rate of yellow eel seedlings. This hatching management method combining systematicness, dynamics and precision not only significantly improves the hatching survival rate and growth quality of yellow eel seedlings, but also reduces the labor intensity and management cost, and improves the hatching efficiency. The method provided by the present invention realizes the comprehensive optimization and precise control of the yellow eel seedling hatching process by combining modern information technology, biotechnology and refined management, and improves the efficiency and accuracy of dynamic monitoring and optimization of yellow eel seedling hatching in the indoor environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0067] Figure 1Schematic diagram of an embodiment of the method for dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment in the embodiments of the present application;

[0068] Figure 2 Schematic diagram of an embodiment of the system for dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment in the embodiments of the present application. Detailed implementation manners

[0069] The embodiments of the present application provide a method and a system for dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment. Terms such as "first", "second", "third", "fourth", etc. (if any) in the specification, claims and the above-mentioned drawings of the present application are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order other than those illustrated or described here. In addition, the term "including" or "having" and any variation thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0070] For ease of understanding, the specific process of the embodiments of the present application will be described below. Please refer to Figure 1 An embodiment of the method for dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment in the embodiments of the present application includes:

[0071] Step S101: Perform multi-factor collaborative optimization processing on the pre-collected hatching environment parameters to obtain a set of dynamic regulation parameters;

[0072] It can be understood that the execution subject of the present application can be a system for dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment, or a terminal or a server. Specifically, it is not limited here. The embodiments of the present application will be described by taking the server as the execution subject as an example.

[0073] Specifically, key parameters in the hatching environment of rice field eels are comprehensively collected. These parameters usually include water temperature, dissolved oxygen, pH value, light intensity, salinity, hardness, etc. During the collection process, a high-precision sensor network is used to ensure the accuracy and real-time nature of the data. Then, time series decomposition is performed on the collected historical data, and each parameter is decomposed into a trend term, a seasonal term, and a random term to deeply analyze the variation laws of each parameter. Next, polynomial fitting is performed on the trend term to obtain the long-term variation trend function of each parameter; Fourier analysis is performed on the seasonal term to reveal the periodic variation characteristics of each parameter; autocorrelation analysis is performed on the random term to understand the short-term fluctuation characteristics of each parameter. Based on these analysis results, a time series prediction model for each parameter is constructed to generate predicted values of the parameters for a future period of time.

[0074] To achieve the collaborative optimization of multiple factors, the mutual relationships between the parameters are deeply studied. Through correlation analysis, the correlation coefficient matrix between the parameters is calculated to reveal the degree of mutual influence between the parameters. Subsequently, using the principal component analysis method, the main factors that have the most significant impact on the hatching environment are selected from numerous parameters, and their weights are determined. This helps to simplify the subsequent optimization process and improve the calculation efficiency. After obtaining the main influencing factors and their weights, the predicted time series data of each parameter are weighted and combined to form a comprehensive environmental index, which can comprehensively reflect the overall situation of the hatching environment. Finally, based on this comprehensive environmental index, the dynamic programming algorithm is used to perform collaborative optimization on each parameter to obtain a set of dynamic regulation parameters. This parameter set not only considers the individual impacts of each parameter but also takes into account the interactions between the parameters, and can ensure the overall optimality of the hatching environment while meeting the constraint conditions of each parameter.

[0075] For example, assume that during the optimization process, it is found that there is a significant negative correlation between water temperature and dissolved oxygen, that is, when the water temperature rises, the dissolved oxygen content will decrease. Through collaborative optimization, when the water temperature rises, the aeration intensity is appropriately increased to maintain the dissolved oxygen within the appropriate range. At the same time, considering that the pH value is jointly affected by water temperature and dissolved oxygen, the addition amount of the pH regulator will also be adjusted accordingly to maintain the overall stability of the water quality.

[0076] Step S102: Based on the set of dynamic regulation parameters, conduct a phased nutritional requirement analysis on the rice field eel seedlings to obtain a multi-dimensional nutritional supplementation plan;

[0077] Specifically, by using key data such as water temperature, dissolved oxygen, and pH value in the dynamic regulation parameter set, and through data analysis and biological knowledge, the key growth stages of rice field eel seedlings are determined. This usually includes the newly hatched stage, the first feeding stage, the metamorphosis stage, and the seedling stage, etc. Each stage has its specific physiological characteristics and nutritional requirements. Then, detailed morphological measurements are carried out on the rice field eel seedlings at each growth stage, including body length, body weight, and the degree of development of feeding organs, etc. These data provide a basis for subsequent nutritional requirement analysis. Based on the morphological data and the characteristics of each growth stage, the palatability analysis of bait is carried out. This is crucial because the requirements for bait size and type of seedlings at different growth stages are different. For example, in the newly hatched stage, it mainly relies on the yolk sac for energy supply. In the first feeding stage, it may need tiny baits such as rotifers, and in the later stage, it can gradually transition to nauplii of Artemia salina and artificial compound feeds. Through palatability analysis, the most suitable bait type can be determined for each growth stage.

[0078] After determining the bait type, detailed nutritional composition analysis is carried out on various baits, including the contents of protein, fat, carbohydrates, and various trace elements. These data are combined with the nutritional requirements of rice field eel seedlings at each growth stage, and through precise ratio calculation, the optimal nutritional composition ratio for each stage is obtained. At the same time, by observing and analyzing the feeding behavior of seedlings, the optimal feeding frequency and single feeding amount are determined. This can not only meet the nutritional requirements of seedlings but also avoid water quality pollution problems caused by overfeeding. According to the previously obtained nutritional composition ratio and feeding amount, the addition amount of trace elements required for each growth stage is accurately calculated. Considering the interactions between different trace elements and their existence forms and bioavailability in the water body. In order to verify and optimize the nutritional supplement plan, continuous growth index measurements are carried out on the rice field eel seedlings at each growth stage, including body length growth rate, body weight growth rate, and survival rate, etc. Through multiple regression analysis, the influence weights of various nutritional components on the seedling growth index can be determined, which provides a scientific basis for further optimizing the nutritional supplement plan.

[0079] For example, assume that during the analysis, it is found that the eel larvae in the metamorphosis stage have a particularly high demand for protein and a relatively low demand for fat. Based on this finding, the protein content in the bait can be appropriately increased and the fat content can be reduced during this stage. In addition, if it is found that a certain trace element (such as zinc) has a significant impact on the immune function of the larvae, the addition amount of this element can be appropriately increased at a specific stage. Through this refined nutritional management, not only can the specific needs of the larvae at different growth stages be met, but also the utilization efficiency of the bait can be improved, reducing waste and water pollution. By integrating the bait types, nutrient composition ratios, feeding strategies, and trace element addition schemes at each growth stage, a comprehensive multi-dimensional nutritional supplement plan is formed. This plan is not static, but can be dynamically adjusted according to the growth status of the larvae and the changes in environmental parameters during the actual breeding process, so as to achieve precise and personalized nutritional management and provide the best support for the healthy growth of eel larvae.

[0080] Step S103: According to the dynamic regulation parameter set and the multi-dimensional nutritional supplement plan, analyze the microbial community structure of the hatching water body to obtain a microbial regulation plan;

[0081] Specifically, scientifically sample the hatching water body to ensure that the samples can represent the microbial community structure of the entire water body. Different levels and regions of the water body need to be considered during sampling to obtain a comprehensive microbial distribution. Then, use high-throughput sequencing technology, such as the Illumina sequencing platform, to extract and amplify the microbial DNA in the water sample. Usually, it is targeted at the 16S rRNA gene of bacteria or the ITS region of fungi to obtain the original sequence data of the microbial community. After obtaining the original sequence data, strict bioinformatics processing is required, including quality control and chimera removal. Quality control can filter out low-quality sequences, and chimera removal can eliminate the artificial sequences generated during the PCR amplification process, thus ensuring the accuracy of subsequent analysis. After these treatments, the obtained valid sequence data will be used for subsequent microbial community structure analysis.

[0082] The next step is to perform operational taxonomic unit (OTU) clustering analysis on the valid sequence data. This process clusters sequences with high similarity into one category, and each category represents a possible species or genus. In this way, information on the species composition and relative abundance of the microbial community can be obtained. Based on this information, α-diversity indices (such as Shannon index, Simpson index, etc.) and β-diversity indices are calculated to evaluate the diversity level of the microbial community and the differences between communities. Then, the diversity assessment results of the microbial community are combined with the environmental factor data in the dynamic regulation parameter set for redundancy analysis (RDA) or canonical correspondence analysis (CCA). These analytical methods can reveal the degree of influence of key environmental factors (such as water temperature, pH, dissolved oxygen, etc.) on the microbial community structure, helping us understand how environmental changes affect the water body microecosystem.

[0083] Combined with the multi-dimensional nutritional supplement plan and the results of microbial community analysis, beneficial microbial species for the growth of Monopterus albus seedlings are screened out. These beneficial microorganisms may include strains that can decompose organic matter, regulate water quality, inhibit pathogenic bacteria, or promote seedling digestion. Through literature research and experimental verification, the optimal culture conditions and yield data of these microorganisms are determined, laying a foundation for the subsequent development of microbial agents. Next, according to the screened beneficial microorganisms, the formulation and production process of the microbial agent are designed. This process needs to consider the interactions between microorganisms, their survival ability in the water body, and their compatibility with Monopterus albus seedlings. The prepared microbial agent samples need to be subjected to strict stability and activity tests to determine their optimal storage conditions and shelf life. Finally, based on the characteristics of the microbial agent and the actual situation of the hatching water body, the delivery strategy of the microbial agent is designed. This includes determining the optimal delivery time sequence and dosage, and may also need to consider the use of slow-release technology to maintain a stable concentration of beneficial microorganisms in the water body.

[0084] For example, assume that through the analysis of the microbial community structure, it is found that under specific water temperature and pH conditions, the abundance of a certain photosynthetic bacterium (such as Rhodospirillum) increases significantly, and this bacterium can effectively degrade organic matter in the water body and improve water quality. Based on this finding, this photosynthetic bacterium can be incorporated into the formulation of the microbial agent. At the same time, if it is found that a certain lactic acid bacterium has a positive impact on the intestinal health of Monopterus albus seedlings, it can also be added to the agent. When designing the delivery strategy, the different delivery frequencies and dosages can be determined according to the growth characteristics of the photosynthetic bacterium and the lactic acid bacterium. For example, it may be necessary to deliver the photosynthetic bacterium twice a day, morning and evening, to maintain water quality stability, while the lactic acid bacterium may only need to be added before and after feeding.

[0085] Step S104: Based on the dynamic regulation parameter set, the multi-dimensional nutrition supplementation plan, and the microbial regulation plan, conduct real-time multi-parameter monitoring on the hatching water body to obtain a water quality dynamic regulation strategy. Among them, the water quality dynamic regulation strategy includes: hierarchical warning thresholds and a regulation measure selection matrix;

[0086] Specifically, deploy a high-precision multi-parameter sensor network in the hatching water body. These sensors can monitor key water quality parameters in real time, such as dissolved oxygen, pH value, ammonia nitrogen, nitrite, and nitrate content, etc. These sensors are connected to the central control system through Internet of Things technology to ensure real-time data transmission and processing. The collected real-time data is then subjected to time series analysis. Through statistical methods such as moving average and exponential smoothing techniques, analyze the change trends and fluctuation ranges of each parameter. The purpose is to identify the short-term fluctuations and long-term trends of water quality parameters, providing a basis for the subsequent warning system and regulation strategy. At the same time, combining the historical data and target values in the dynamic regulation parameter set, the system can more accurately judge the normality and potential risks of the current water quality state.

[0087] Based on the analysis results and the water quality index requirements in the dynamic regulation parameter set, set multi-level warning thresholds for each water quality parameter. These thresholds usually include multiple levels such as normal, attention, warning, and danger. Each level corresponds to different water quality conditions and risk levels. The setting of the thresholds needs to consider the tolerance of the juvenile eels at different growth stages, as well as the mutual influence between parameters. For example, the dissolved oxygen threshold under high-temperature conditions may need to be adjusted upward accordingly because high temperature will reduce the oxygen content in the water body. Conduct correlation analysis between these hierarchical warning thresholds and the bait feeding strategy in the multi-dimensional nutrition supplementation plan. The purpose is to understand the impact of bait feeding on water quality parameters, so as to consider the feeding factor in the warning system. For example, adjust the warning thresholds of certain parameters in a short period before and after feeding to adapt to the temporary water quality fluctuations caused by feeding. Classify and organize the regulation measures under different situations based on the microbial regulation plan and the water quality conditions to form a preliminary regulation measure library. This measure library contains various possible water quality regulation methods, such as increasing aeration, partial water change, adding buffer agents, and putting microbial agents, etc. Each measure will be marked with its applicable water quality conditions, expected effects, and possible side effects. In order to optimize the selection of regulation measures, conduct effect evaluation experiments on the measures in the measure library. These experiments will test the response time and influence degree of each measure, providing data support for subsequent decision-making. Based on the experimental results, rank the regulation measures by priority and conduct combined optimization to form a regulation measure selection matrix for different water quality conditions. This matrix will consider multiple factors such as the effect, cost, and operation difficulty of the measures, and recommend the optimal regulation plan for each water quality condition.

[0088] Finally, integrate the hierarchical warning thresholds and the adjustment measure selection matrix to form a complete dynamic water quality adjustment strategy. This strategy is not static but can be dynamically adjusted according to real-time monitoring data and the growth status of the seedlings. Continuously learn and optimize by analyzing historical data and adjustment effects through machine learning algorithms, and continuously improve the warning thresholds and adjustment measures.

[0089] For example, assume that a downward trend in the dissolved oxygen level is detected but has not reached the warning threshold. Based on the warning system, it will first issue a "caution" - level reminder. If this occurs before the scheduled feeding time, it is recommended to postpone or reduce the feeding amount to avoid further burdening the water quality. At the same time, it will select appropriate measures from the adjustment measure selection matrix, such as increasing the aeration intensity or starting the standby oxygenation equipment. If the situation continues to deteriorate to the "warning" level, it is recommended to partially change the water and add microbial agents that help improve the dissolved oxygen. Throughout the process, continuously monitor the changes in various parameters and adjust the strategy in a timely manner according to the feedback.

[0090] Step S105: Based on the dynamic water quality adjustment strategy, optimize the environment of the hatching water body, and in the optimized hatching water body, collect and analyze multi - dimensional data of the rice field eel seedlings to obtain the analysis results of behavioral characteristics;

[0091] Specifically, according to the hierarchical warning thresholds and the adjustment measure selection matrix in the dynamic water quality adjustment strategy, precisely control the hatching water body. This may include fine - tuning parameters such as water temperature, pH value, and dissolved oxygen, while considering the mutual influence between the parameters. For example, if the ammonia nitrogen content is detected to be close to the warning threshold, automatically increase the use of biological filters and simultaneously fine - tune the pH value to promote the conversion of ammonia. This control is not a one - time process but a continuous one, continuously optimizing the control strategy according to real - time feedback until the ideal water quality state is achieved. After the water quality is optimized, the next step is to conduct all - round data collection on the rice field eel seedlings. This usually includes multi - dimensional observations and records. First, use high - definition underwater cameras to continuously record videos of the seedlings, capturing their swimming trajectories and behavior patterns. These raw video data will then be processed through computer vision algorithms to extract the movement parameters of the seedlings, such as swimming speed, turning frequency, and activity range. At the same time, the group behavior characteristics of the seedlings will also be analyzed, including aggregation degree, distribution uniformity, and overall activity level. These data can reflect the health status of the seedlings and their adaptability to the environment.

[0092] In addition to visual data, the feeding behaviors of the seedlings are also recorded and analyzed. Through a precisely controlled feeding system and underwater cameras, the reaction speed, feeding frequency, and food intake of the seedlings during each feeding can be accurately recorded. After time series analysis of these data, the feeding rhythm and satiation change curves of the seedlings can be obtained, providing an important basis for optimizing the feeding strategy. The sound signals of the rice field eel seedlings are collected through an underwater microphone system. Although these sounds are usually very weak, through highly sensitive underwater microphones and advanced signal processing techniques, the subtle sounds generated by the activities of the seedlings can be captured. After spectral analysis of these sounds, a unique acoustic feature spectrum can be obtained, which potentially can reflect certain physiological states or behavior patterns of the seedlings.

[0093] All these multi-dimensional data, including motion parameters, group behavior characteristics, feeding behavior data, and acoustic feature spectra, are integrated into a comprehensive data analysis platform. This platform uses machine learning algorithms, especially deep learning networks, to process and analyze these massive data. Through multi-dimensional feature extraction and pattern recognition, the system can identify meaningful behavior patterns and physiological state indicators from the seemingly chaotic data. Finally, cluster analysis is performed on these multi-dimensional behavior feature data to classify the states of the seedlings into several categories such as normal, sub-healthy, and diseased. This classification is not a simple binary division but a continuous spectrum that can reflect the subtle changes in the states of the seedlings. At the same time, the system also conducts correlation analysis between these behavior characteristics and known physiological indicators to establish a behavior-physiology correspondence relationship, thereby realizing the inference of the physiological state of the seedlings through non-invasive behavior observation.

[0094] For example: Suppose the system detects a slight decrease in the average swimming speed of a group of seedlings and an increase in aggregation behavior on a certain day. Based on these data alone, it may not be sufficient to determine the health status of the seedlings. However, when the system further analyzes and finds that the feeding reaction speed of these seedlings also slows down and the frequency of the sounds they produce changes slightly, it will infer that this group of seedlings is in a mild stress state. Based on this judgment, it is recommended to slightly lower the water temperature, reduce the feeding amount, and add certain trace elements that help relieve stress. After implementing these measures, continuously monitor the behavior changes of the seedlings. If it is found that the swimming speed returns to normal, the aggregation behavior decreases, and the feeding activity increases, the effectiveness of the adjustment measures can be confirmed.

[0095] Step S106: Analyze the multi-source stress factor exposure experiment plan for the results of the behavior feature analysis, generate a multi-source stress factor exposure experiment plan, and conduct a stress experiment on the rice field eel seedlings through the multi-source stress factor exposure experiment plan to obtain a stress response database;

[0096] Specifically, conduct in-depth cluster analysis on the previously obtained behavioral feature analysis results. Using machine learning algorithms such as K-means or hierarchical clustering method, classify the behavioral patterns of rice field eel seedlings into several typical types. These typical behavioral patterns may include normal activity, overexcitement, slow response, etc. Each pattern represents the behavioral characteristics of the seedlings under specific states. Based on these typical behavioral patterns, conduct extensive literature research and expert consultations to identify potential stress factors that may cause these behavioral changes. These factors can generally be classified into categories such as physical factors (such as temperature change, light intensity, water flow velocity), chemical factors (such as pH value fluctuation, dissolved oxygen change, pollutant exposure), and biological factors (such as high-density farming, pathogen infection, presence of predators), etc. Through systematic screening and evaluation, establish a comprehensive list of candidate stress factors.

[0097] Classify and prioritize these candidate stress factors. The classification process considers the nature, controllability, and potential impact degree on the seedlings of the factors. The prioritization is based on the universality, severity, and research value of the factors. This helps to ensure that the experiments cover the most important and representative stress factors. Design multi-level intensity gradients for each stress factor. For example, for the factor of temperature, four levels will be set: normal temperature, slightly elevated, moderately elevated, and severely elevated. This gradient design allows observing the responses of the seedlings to different degrees of stress, and helps to determine the critical thresholds and dose-response relationships. Based on these preparations, start constructing a multi-source stress factor exposure experiment matrix. This matrix includes not only the experimental designs of single factors, but also takes into account the combined effects of multiple factors. For example, design an experimental group where the temperature rises and the dissolved oxygen decreases simultaneously to simulate the complex situations that may occur in the real environment. The design of the experimental matrix needs to balance comprehensiveness and feasibility to ensure that the experiments can be completed within a reasonable time and resource range.

[0098] During the actual stress experiments, strictly control the experimental conditions to ensure that except for the preset stress factors, other environmental parameters remain stable. During the experiment, continuously record data such as the behavioral changes, physiological indicators, and survival rate of the seedlings. Multidimensional monitoring techniques mentioned above, including high-definition video analysis, acoustic monitoring, and automated physiological parameter detection, etc., will be used during these data collection processes. After the experimental data collection is completed, conduct in-depth statistical analysis. This includes methods such as analysis of variance, regression analysis, and multivariate statistics, etc. The purpose is to quantify the impact degree and thresholds of each stress factor on the seedlings. Through these analyses, establish a quantitative relationship model between the stress factors and the growth parameters of the seedlings. These models can predict the growth conditions and behavioral characteristics that the seedlings may exhibit under given stress conditions.

[0099] To verify and optimize these models, a series of validation experiments will be conducted. These experiments may include testing the predictive ability of the models under different stress conditions, or verifying the applicability of the models under conditions closer to the actual aquaculture environment. Based on the validation results, necessary adjustments and optimizations will be made to the models to improve their accuracy and practicality. Finally, all experimental data, analysis results, and optimized models will be integrated into a comprehensive stress response database. This database not only contains the original experimental data, but also various analysis reports, model parameters, and prediction tools. It is designed as a dynamic and continuously updated system that can be improved and expanded with new research findings and data accumulation.

[0100] For example, assume that in the previous behavioral analysis, it was found that some seedlings showed abnormal aggregation behavior and reduced feeding under specific conditions. Based on this observation, they focused on water temperature increase and dissolved oxygen decrease as stress factors for research. In the experimental design, they set four water temperature levels (25°C, 28°C, 31°C, 34°C) and three dissolved oxygen levels (7mg / L, 5mg / L, 3mg / L) to form a 4x3 experimental matrix. Under each condition, behavioral indicators such as the swimming speed, aggregation degree, and feeding frequency of the seedlings will be monitored, and physiological indicators such as the growth rate, metabolic rate, and stress hormone level will be recorded. By analyzing these data, it will be found that when the water temperature rises to 31°C and the dissolved oxygen drops to 5mg / L, the seedlings begin to show obvious stress responses, such as a 50% increase in aggregation behavior and a 30% decrease in feeding frequency. This finding will be recorded in the stress response database and may be used to set warning thresholds in aquaculture management. At the same time, the effect of adding certain stress relievers (such as vitamin C) under such conditions will be further explored to provide possible solutions for dealing with similar situations.

[0101] Step S107: Generate a stress management plan according to the stress response database, where the stress management plan includes: stress source classification data, warning indicators, and a set of mitigation measures.

[0102] Specifically, systematically classify and cluster analyze the data in the stress response database. Using machine learning algorithms such as hierarchical clustering or K-means clustering, classify various stress factors according to their nature, impact degree, and action mechanism. The purpose is to establish a structured stressor classification system, which may include major categories such as physical stress (e.g., temperature, light, water flow), chemical stress (e.g., pH value, dissolved oxygen, pollutants), and biological stress (e.g., density, pathogens, predation pressure). Each major category may contain multiple specific stressors. Next, for each type of stressor, extract key warning indicators from the database. These indicators generally include two aspects: environmental parameter indicators and biological response indicators. Environmental parameter indicators may include the critical thresholds of specific stressors, such as the rate and amplitude of temperature change, the lowest acceptable level of dissolved oxygen, etc. Biological response indicators may include specific change patterns in seedling behavior, such as abnormal aggregation behavior, a significant decrease in feeding frequency, a change in swimming pattern, etc. Through statistical analysis and machine learning models, the warning thresholds of these indicators can be determined, and a multi-level warning system can be established.

[0103] After determining the stressor classification and warning indicators, proceed to design a set of targeted mitigation measures. This process requires comprehensive consideration of the experimental results in the stress response database, effective methods reported in the literature, and actual aquaculture experience. For each stressor, develop multiple levels of mitigation measures, including preventive measures, early intervention measures, and emergency response measures. These measures may involve the adjustment of environmental parameters (such as adjusting water temperature, increasing aeration), the change of nutritional strategies (such as adding specific stress relievers, adjusting feed formulations), drug intervention (such as using immunopotentiators), and other aspects. To make the stress management plan more intelligent and dynamic, receive real-time monitoring data of the aquaculture environment, automatically identify potential stress risks according to the preset warning indicators, and recommend the most suitable coping strategies from the set of mitigation measures.

[0104] Integrate the stressor classification data, warning indicators, and set of mitigation measures into a complete stress management plan. This plan not only includes static guiding documents but may also include interactive digital platforms or mobile applications so that aquaculture personnel can conveniently query and use them.

[0105] For example: Suppose in the stress response database, it is found that high temperature is a common and significantly influential stressor. Through data analysis, they determine that when the water temperature rises rapidly (such as by more than 3°C within 24 hours) or continuously exceeds 32°C, it will trigger the heat stress response of eel larvae. Therefore, in the stress management plan, high temperature stress will be classified as a type of physical stress, and its warning indicators are set as "the water temperature rise rate > 3°C / day within 24 hours" or "continuous water temperature > 32°C". For this kind of high temperature stress, the set of mitigation measures may include the following levels of strategies: 1) Preventive measures: Reduce the stocking density in advance and increase water body shading when high temperature weather is predicted; 2) Early intervention: When the water temperature starts to rise, increase the aeration intensity, reduce the feeding amount, and add antioxidants such as vitamin C; 3) Emergency response: When the water temperature reaches a dangerous level, immediately start the cooling equipment, consider partial water change, and use stress relievers such as melatonin if necessary.

[0106] In practical applications, if the aquaculture detects that the water temperature has risen by 2°C within 12 hours, the decision support automatically issues a yellow warning and recommends taking early intervention measures. If the temperature continues to rise and exceeds 32°C, it is upgraded to a red warning and emergency response measures are recommended.

[0107] In the embodiments of the present application, through multi-factor collaborative optimization processing of the pre-collected incubation environment parameters, a dynamic regulation parameter set is obtained, realizing precise control and dynamic adjustment of the incubation environment, effectively avoiding the problem of mismatched environmental parameters caused by traditional static control methods, and providing a more stable and suitable growth environment for yellow eel seedlings; based on the dynamic regulation parameter set, a phased nutritional requirement analysis is carried out on yellow eel seedlings to obtain a multi-dimensional nutritional supplementation plan, making the nutritional supply more precise and personalized, effectively improving the bait utilization rate, and promoting the healthy growth of seedlings; through the analysis of the microbial community structure of the incubation water body, a microbial regulation plan is obtained, realizing precise regulation of the water body micro-ecosystem, improving the water quality stability and self-purification ability, and creating a better micro-environment for seedling growth; based on the dynamic regulation parameter set, the multi-dimensional nutritional supplementation plan and the microbial regulation plan, multi-parameter real-time monitoring of the incubation water body is carried out to obtain a water quality dynamic regulation strategy, including a hierarchical early warning threshold and a regulation measure selection matrix, making the water quality management more proactive and precise, being able to prevent and respond to various water quality problems in a timely manner, and ensuring the stability of the seedling growth environment; through multi-dimensional data collection and analysis of yellow eel seedlings in the optimized incubation water body, an analysis result of behavioral characteristics is obtained, realizing comprehensive and real-time monitoring of the growth status and behavioral characteristics of seedlings, and providing reliable data support for subsequent management decisions; an analysis of a multi-source stress factor exposure experimental plan is carried out on the analysis result of behavioral characteristics, and a stress response database is obtained through experiments, systematically revealing the influence mechanism of various potential stress factors on yellow eel seedlings, and providing a scientific basis for stress prevention and control; finally, a stress management plan formulated according to the stress response database, including stress source classification data, early warning indicators and a set of mitigation measures, realizes precise identification, timely early warning and effective mitigation of various stress factors, greatly improving the stress resistance and survival rate of yellow eel seedlings. This systematic, dynamic and precise incubation management method not only significantly improves the hatching survival rate and growth quality of yellow eel seedlings, but also reduces the labor intensity and management cost, and improves the hatching efficiency. The method provided by the present invention realizes the all-round optimization and precise control of the yellow eel seedling hatching process by combining modern information technology, biotechnology and refined management, and improves the efficiency and accuracy of dynamic monitoring and optimization of yellow eel seedling hatching in the indoor environment.

[0108] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0109] (1) Perform time series decomposition on the water temperature, dissolved oxygen, pH value, light intensity, salinity time series change data and hardness historical data of the pre-collected incubation environment parameters to obtain the trend term, seasonal term and random term of each parameter;

[0110] (2) Perform polynomial fitting on the trend term of each parameter to obtain the long-term change trend function of each parameter;

[0111] (3) Perform Fourier analysis on the seasonal terms of each parameter to obtain the periodic change characteristics of each parameter;

[0112] (4) Perform autocorrelation analysis on the random terms of each parameter to obtain the short-term fluctuation characteristics of each parameter;

[0113] (5) Based on the long-term change trend function of each parameter, the periodic change characteristics of each parameter, and the short-term fluctuation characteristics of each parameter, construct a time series prediction model for each parameter to obtain the predicted time series data of each parameter;

[0114] (6) Perform correlation analysis on the predicted time series data of each parameter to obtain the correlation coefficient matrix between parameters;

[0115] (7) According to the correlation coefficient matrix between parameters, perform principal component analysis on each parameter to obtain the main influencing factors and their weights, and based on the main influencing factors and their weights, perform weighted combination on the predicted time series data of each parameter to obtain a comprehensive environmental index;

[0116] (8) According to the comprehensive environmental index, perform collaborative optimization on each parameter through a dynamic programming algorithm to obtain a set of dynamic regulation parameters.

[0117] Specifically, perform time series decomposition on the historical data of water temperature, dissolved oxygen, pH value, light intensity, salinity, and hardness. Use the Seasonal and Trend decomposition using Loess (STL) method to decompose the time series data of each parameter into a trend term, a seasonal term, and a random term. The trend term reflects the long-term change trend of the parameter, the seasonal term represents the periodic change pattern, and the random term contains irregular fluctuations and noise. Perform polynomial fitting on the trend terms of each parameter. Use the least squares method to select a polynomial function of an appropriate order to fit the trend term data. By comparing the fitting effects of polynomials of different orders, select the function that can best describe the long-term change trend. For example, for the water temperature trend, it may be found that a cubic polynomial can well capture the slow change trend of the temperature.

[0118] For the seasonal term, use the Fourier analysis method to identify and quantify the periodic change characteristics. Fourier analysis transforms the time series data into the frequency domain, thereby revealing the main periodic components in the data. By analyzing the results of the Fourier transform, the main periodic lengths and intensities of each parameter can be determined. For example, it may be found that there are obvious daily and annual change cycles in the dissolved oxygen content.

[0119] The analysis of the random terms is carried out through autocorrelation analysis. Autocorrelation analysis calculates the correlation of a time series with itself at different time lags, which helps to reveal the short-term dependence relationships and fluctuation patterns in the data. By analyzing the autocorrelation function and the partial autocorrelation function, potential patterns in the random terms, such as short-term persistence or periodic fluctuations, can be identified.

[0120] Based on the above analysis results, a time series prediction model is constructed for each parameter. The long-term trend function, periodic change characteristics, and short-term fluctuation characteristics are comprehensively utilized. Commonly used prediction models include the autoregressive integrated moving average (ARIMA) model, the exponential smoothing model, or more complex machine learning models such as the long short-term memory (LSTM) neural network. Through these models, the predicted time series data of each parameter can be generated. Correlation analysis is performed on the predicted time series data of each parameter, and the Pearson correlation coefficient between the parameters is calculated to form a correlation coefficient matrix. This reveals the mutual relationships between different environmental parameters and helps to understand the mutual influences between the parameters.

[0121] Based on the correlation coefficient matrix, principal component analysis (PCA) is carried out. PCA can identify the main influencing factors that contribute the most to the overall variation and calculate the weights of each factor. It effectively reduces the dimension of the data while retaining the most important information.

[0122] Finally, using the main influencing factors and their weights, the predicted time series data of each parameter are weighted and combined to obtain a comprehensive environmental index. This index can comprehensively reflect the overall situation of the hatching environment. Based on this comprehensive index, the parameters are synergistically optimized through the dynamic programming algorithm, and finally a set of dynamic regulation parameters is obtained. The dynamic programming algorithm can efficiently find the optimal parameter combination by decomposing a complex problem into a series of sub-problems and avoiding repeated calculations during the solution process.

[0123] For example: Suppose when analyzing water temperature data, the time series decomposition shows obvious annual periodic changes and a slowly rising long-term trend. The following trend function may be obtained by polynomial fitting: T(t) = 25.2 + 0.01t - 0.0002t 2 , where T represents temperature and t represents time (unit: days). Fourier analysis may reveal a strong 365-day period and a weak 30-day period, corresponding to annual and monthly changes respectively. Autocorrelation analysis may show a short-term correlation of 3 - 5 days, indicating short-term persistence in temperature changes.

[0124] When building a prediction model, the ARIMA(2,1,1) model may be selected to capture these features. Correlation analysis may find that the water temperature is strongly negatively correlated with dissolved oxygen (correlation coefficient is -0.85) and weakly positively correlated with pH value (correlation coefficient is 0.3). Principal component analysis may identify that water temperature and dissolved oxygen are the two main influencing factors, explaining 60% and 25% of the total variation respectively. Based on these analysis results, the comprehensive environmental index may be defined as: I = 0.6T ′ + 0.25O ′ - 0.1P ′ - 0.05S ′ , where T ′ , O ′ , P ′ and S ′ represent the standardized water temperature, dissolved oxygen, pH value and salinity respectively. The dynamic programming algorithm will, under the guidance of this comprehensive index, search for a parameter combination that can keep the index within the ideal range to form a set of dynamic regulation parameters.

[0125] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0126] (1) Segment the data of water temperature, dissolved oxygen and pH value in the hatching environment parameters to obtain the growth stage division criteria for the eel larvae.

[0127] (2) According to the growth stage division criteria, segment the growth cycle of the eel larvae to obtain the time range of each growth stage, and conduct morphological measurements on the eel larvae in each growth stage to obtain data on body length, body weight and the development degree of feeding organs.

[0128] (3) According to the data of body length, body weight and the development degree of feeding organs, conduct palatability analysis on the eel larvae in each growth stage to obtain the target bait types for each stage.

[0129] (4) Conduct nutritional component analysis on the target bait types for each stage to obtain the trace element content data of each bait type, and calculate the proportion of nutritional components of the bait formula according to the trace element content data of each bait type and the nutritional requirements of the eel larvae in each growth stage to obtain the proportion of nutritional components in each growth stage.

[0130] (5) Collect the feeding behaviors of the eel larvae in each growth stage to obtain the target feeding frequency and single feeding amount, and calculate the addition amount of trace elements according to the proportion of nutritional components in each growth stage and the single feeding amount to obtain the addition amount of trace elements in each growth stage.

[0131] (6) Measure the growth indicators of the rice field eel seedlings at each growth stage to obtain data on body length growth rate, body weight growth rate, and survival rate. Then, conduct a correlation analysis between the data of body length growth rate, body weight growth rate, and survival rate and the nutritional components through multiple regression analysis to obtain the influence weights of each nutritional component on the seedling growth indicators. Finally, generate a multi-dimensional nutritional supplement plan based on the influence weights of each nutritional component on the seedling growth indicators and the addition amounts of trace elements at each growth stage.

[0132] Specifically, segment the data of water temperature, dissolved oxygen, and pH value in the hatching environment parameters. Using a clustering analysis method, such as the K-means clustering algorithm, divide the time series data of these environmental parameters into several relatively stable stages. The environmental characteristics of each stage correspond to the physiological development stage of the rice field eel seedlings, thereby obtaining the growth stage division criteria. According to this division criterion, segment the entire growth cycle of the rice field eel seedlings to determine the specific time range of each growth stage. In each divided growth stage, conduct detailed morphological measurements on the rice field eel seedlings. This includes measuring the body length using a precision digital microscope, determining the body weight using a microelectronic balance, and evaluating the development degree of the feeding organs through microscopic observation. These data provide a basis for subsequent analysis of bait palatability.

[0133] Based on the morphological data, especially the data on body length, body weight, and development degree of the feeding organs, conduct an analysis of bait palatability. Using a multi-factor decision analysis method, consider factors such as the oral cavity size, digestive ability, and nutritional requirements of the seedlings to determine the most suitable bait type for each growth stage. For example, rotifers or small plankton may be selected in the early stage, while compound feed may be turned to in the later stage. Conduct a detailed nutritional component analysis on the determined target bait types for each stage. Using analytical techniques such as high performance liquid chromatography (HPLC) and atomic absorption spectrometry (AAS), accurately measure the contents of proteins, fats, carbohydrates, and various trace elements in each bait. Combining the specific nutritional requirements of the rice field eel seedlings at each growth stage, optimize the bait formula through linear programming method to calculate the optimal nutritional component ratio for each growth stage.

[0134] To determine the optimal feeding strategy, the feeding behaviors of rice field eel seedlings at various growth stages are observed and recorded. This process uses high-speed imaging technology to capture data such as the feeding frequency, single feeding amount, and feeding duration of the seedlings. By analyzing this data, the target feeding frequency and single feeding amount for each growth stage are determined. Combining the previously obtained nutrient component ratios, the precise calculation of the trace element addition amount is carried out to ensure that each feeding can meet the nutritional requirements of the seedlings. Finally, the growth indicators of rice field eel seedlings at various growth stages are continuously measured, including regularly measuring the body length and weight, calculating the growth rate, and recording the survival rate. Through multiple regression analysis, a relationship model between these growth indicators and various nutrient components is established. Using the least squares method, the influence weights of each nutrient component on the growth of the seedlings are calculated. Based on these weights and the previously determined trace element addition amounts, a comprehensive multi-dimensional nutritional supplementation plan is formulated.

[0135] For example: Suppose when segmenting the data of water temperature, dissolved oxygen, and pH value, the K-means clustering algorithm identifies three main growth stages: the newly hatched stage (0 - 7 days), the metamorphosis stage (8 - 21 days), and the seedling stage (22 - 45 days). In the newly hatched stage, morphological measurements show that the average body length increases from 3 mm to 5 mm, and the body weight increases from 0.1 mg to 0.3 mg, and the feeding organs are not fully developed. Based on these data, the analysis of bait palatability determines that the target bait type for the newly hatched stage is a mixture of rotifers and microalgae.

[0136] Nutrient component analysis shows that this mixed bait contains 40% protein, 10% fat, 30% carbohydrate, and various trace elements. Through linear programming optimization, the optimal nutrient ratio for the newly hatched stage is determined to be 45% protein, 12% fat, and 28% carbohydrate. Feeding behavior analysis finds that the daily feeding frequency of newly hatched seedlings is 8 - 10 times, and the single feeding amount is about 5% of the body weight. Based on these data, the daily trace element addition amount for the newly hatched stage is calculated. For example, the vitamin C addition amount is 0.5 mg / L of water body.

[0137] The results of multiple regression analysis show that in the newly hatched stage, the influence weight of protein on the body length growth rate is 0.6, on the body weight growth rate is 0.5, and on the survival rate is 0.4. Based on these data, a nutritional supplementation plan for the newly hatched stage is formulated, including 8 feedings per day, with each feeding amount being 5% of the seedling body weight, the protein content in the bait remaining at 45%, and adding a specific dose of trace elements.

[0138] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0139] (1) Sample the hatching water body, extract and amplify microbial DNA from the water sample through high-throughput sequencing technology to obtain the original sequence data of the microbial community, and perform quality control and chimera removal on the original sequence data to obtain the effective sequence data;

[0140] (2) Perform operational taxonomic unit clustering analysis on the effective sequence data to obtain the species composition and abundance information of the microbial community;

[0141] (3) Calculate the α-diversity index and β-diversity index based on the species composition and abundance information to obtain the diversity assessment results of the microbial community, and perform redundancy analysis (RDA) on the microbial community structure and environmental factors according to the diversity assessment results and the dynamic regulation parameter set to obtain the influence degree of key environmental factors on the microbial community structure;

[0142] (4) Screen out the microbial species beneficial to the growth of rice field eel seedlings according to the influence degree and the multi-dimensional nutrient supplementation plan to obtain the target microbial flora combination;

[0143] (5) Conduct a culture experiment on the target microbial flora combination, measure its growth curve and metabolites to obtain the target culture conditions and yield data, and design the formula and production process of the microbial agent according to the target culture conditions and yield data to obtain the microbial agent sample;

[0144] (6) Conduct stability and activity tests on the microbial agent sample to obtain the storage conditions and shelf life of the agent, design the delivery strategy of the microbial agent according to the storage conditions and shelf life, the delivery strategy includes the delivery time series and dosage, and generate a microbial regulation plan through the delivery strategy.

[0145] Specifically, scientifically sample the hatching water body, use a sterile water sampler to collect water samples at different water layers and positions to ensure that the samples can represent the microbial distribution of the entire water body. The collected water samples are immediately stored at low temperature to maintain the original state of the microbial community. Use high-throughput sequencing technology to extract and amplify microbial DNA from the water sample. Use a commercial DNA extraction kit, such as the PowerWater DNA Isolation Kit, to extract the total DNA from the water sample. Then, use universal primers to perform PCR amplification on the V3-V4 region of the bacterial 16S rRNA gene. After purification of the amplification product, construct a sequencing library and perform high-throughput sequencing on the Illumina MiSeq platform to obtain the original sequence data.

[0146] Quality control and chimera removal are performed on the obtained raw sequence data. The quality control process includes removing low-quality reads, trimming adapter sequences, filtering short sequences, etc., and is completed using QIIME2 software. Chimera removal uses the UCHIME algorithm to identify and remove artificial sequences generated during the PCR process. These processing steps ensure the accuracy of subsequent analyses and obtain high-quality valid sequence data. Operational taxonomic unit (OTU) clustering analysis is performed on the valid sequence data. Using the UPARSE algorithm, sequences with a similarity of 97% are clustered into one class, and each class represents a potential species or genus. By aligning the representative sequences of each OTU with a reference database (such as Greengenes or SILVA), species annotation is performed to obtain the species composition and relative abundance information of the microbial community.

[0147] Based on the species composition and abundance information, α-diversity indices and β-diversity indices are calculated. α-diversity indices include the Shannon index, Simpson index, Chao1 index, etc., which reflect the species diversity within a single sample. β-diversity indices such as the Bray-Curtis distance and UniFrac distance are used to evaluate the community differences between different samples. These indices together constitute the diversity assessment results of the microbial community. Combining the diversity assessment results and the dynamic regulation parameter set, redundancy analysis (RDA) is performed. RDA is a multivariate statistical method used to reveal the relationship between environmental factors and community structure. Through RDA analysis, the influence degrees of key environmental factors such as water temperature, pH, dissolved oxygen, etc. on the microbial community structure are obtained, and the amount of community variation explained by each factor is expressed as a percentage.

[0148] Based on the RDA analysis results and the multi-dimensional nutritional supplement plan, the microbial species beneficial to the growth of Monopterus albus juveniles are screened. Combining literature data and expert knowledge, microorganisms with functions such as water quality purification, nutrient transformation, or pathogen inhibition are selected to form a target microbial flora combination. The screened target microbial flora is cultured in the laboratory, and its growth curve and metabolites are measured. By adjusting conditions such as the composition of the culture medium, temperature, pH, etc., the culture method of each microorganism is optimized to obtain the optimal culture conditions and yield data. Based on these data, the formulation and production process of the microbial preparation are designed, and a microbial preparation sample is prepared.

[0149] A series of stability and activity tests are performed on the microbial preparation sample, including temperature tolerance, pH tolerance, shelf life, etc. Through these tests, the optimal storage conditions and expiration date of the preparation are determined. According to the test results, the delivery strategy of the microbial preparation is designed, including the delivery time sequence and dosage, to form a complete microbial regulation plan.

[0150] For example: In an eel fry hatching system, 1 million raw sequences were obtained through high-throughput sequencing. After quality control and chimera removal, 800,000 high-quality sequences were retained. OTU clustering analysis identified 1,000 OTUs, and the dominant bacterial communities included Pseudomonas, Lactobacillus, and Bacillus, etc. The α-diversity analysis showed that the Shannon index was 4.5, indicating a relatively high community diversity. The results of RDA analysis showed that water temperature and pH value explained 30% and 25% of the variation in the microbial community, respectively.

[0151] Based on these results, three beneficial bacteria were screened out: Bacillus subtilis, Lactobacillus plantarum, and Pseudomonas fluorescens. Laboratory culture showed that these three bacteria grew best under the conditions of 30 °C and pH 7.0, and the colony count could reach 10 9 CFU / mL within 24 hours. The prepared microbial preparation could maintain its activity for 6 months when stored at 4 °C. The final formulated dosing strategy was to dose twice a week, with a dose of 10 6 CFU / mL of water body.

[0152] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0153] (1) Continuously sample and detect the dissolved oxygen, pH value, ammonia nitrogen, nitrite, and nitrate contents in the hatching water body to obtain time series data of water quality parameters, and perform trend analysis on the time series data of water quality parameters to obtain the change trends and fluctuation ranges of each water quality parameter;

[0154] (2) According to the change trends and fluctuation ranges of each parameter, by dynamically regulating the water quality index data in the parameter set, set multi-level warning thresholds for each water quality parameter to obtain a classification warning threshold table, and perform correlation analysis on the thresholds at each level in the classification warning threshold table and the multi-dimensional nutrition supplement plan to obtain the interaction relationship between each water quality parameter and bait feeding;

[0155] (3) According to the interaction relationship between each water quality parameter and bait feeding and the microbial regulation plan, classify and sort out the adjustment measures under different water quality conditions to obtain a preliminary adjustment measure library, and perform effect evaluation experiments on each measure in the preliminary adjustment measure library to obtain the response time and influence degree data of each adjustment measure;

[0156] (4) Based on the response time and impact degree data of each adjustment measure, prioritize and optimize the combination of adjustment measures to obtain an adjustment measure selection matrix for different water quality conditions, and fuse the hierarchical early warning threshold table and the adjustment measure selection matrix to obtain a water quality dynamic adjustment strategy.

[0157] Specifically, continuously sample and detect the key water quality parameters in the hatching water body. Use a high-precision sensor network, including dissolved oxygen sensors, pH electrodes, ammonia nitrogen selective electrodes, nitrite and nitrate ion selective electrodes, etc. These sensors collect data at a frequency of once every 5 minutes to form time series data of water quality parameters. Conduct trend analysis on the collected time series data. Using the time series decomposition method, decompose the data of each parameter into a trend term, a seasonal term, and a random term. The trend term reflects the long-term change trend, the seasonal term represents the periodic change, and the random term contains short-term fluctuations. By analyzing these components, obtain the change trend and fluctuation range of each water quality parameter. For example, it may be found that the dissolved oxygen shows a daily periodic change, while the ammonia nitrogen shows a slow upward trend.

[0158] According to the analysis results and the water quality index requirements in the dynamic regulation parameter set, set multi-level early warning thresholds for each water quality parameter. Using the fuzzy logic method, divide the safe range of each parameter into four levels: normal, attention, warning, and danger. For example, the pH value may be divided into 7.0 - 7.5 as normal, 6.8 - 7.0 or 7.5 - 7.7 as attention, 6.5 - 6.8 or 7.7 - 8.0 as warning, and below 6.5 or above 8.0 as danger. These thresholds form a hierarchical early warning threshold table. Conduct correlation analysis between the hierarchical early warning threshold table and the multi-dimensional nutrient supplementation plan. Use the cross-correlation analysis method to calculate the correlation coefficient and time lag between each water quality parameter and the bait feeding amount and frequency. Through this analysis, the impact degree and time relationship of bait feeding on water quality parameters can be revealed. For example, it may be found that the ammonia nitrogen concentration starts to rise 2 hours after bait feeding.

[0159] Based on the interactive relationship between water quality parameters and bait delivery and the microbial regulation scheme, the regulation measures under different water quality conditions are classified and sorted. The decision tree algorithm is used to formulate corresponding regulation measures according to different combinations of water quality conditions. For example, when dissolved oxygen decreases and ammonia nitrogen increases, possible regulation measures include increasing aeration, reducing feeding, and adding specific microbial preparations. These measures constitute the preliminary regulation measure library. The effect evaluation experiment is carried out on each measure in the preliminary regulation measure library. Different water quality conditions are simulated in a controlled environment, various regulation measures are applied, and the change process of water quality parameters is recorded. Through these experiments, the response time (the time from implementation to the effect) and impact degree (the magnitude of water quality improvement) data of each regulation measure are obtained. According to the response time and impact degree data, the regulation measures are prioritized and combined and optimized. The multi-objective optimization algorithm is used to consider factors such as the effect, implementation difficulty and cost of the measures, and the optimal combination of regulation measures is selected for different water quality conditions. For example, for mild ammonia nitrogen exceeding the standard, increasing the delivery of microbial preparations may be preferred, while for severe exceeding the standard, a combination of partial water change and reduced feeding may be selected. These optimized combinations of measures form a matrix for regulating measure selection.

[0160] Finally, the hierarchical warning threshold table and the adjustment measure selection matrix are fused to obtain a complete water quality dynamic adjustment strategy. This strategy is a complex decision support system that can automatically identify potential risks based on real-time water quality data and recommend the most suitable adjustment measures.

[0161] For example: In a rice field eel hatching system, continuous monitoring data showed that dissolved oxygen showed periodic changes within 24 hours, ranging from 5.5-7.5 mg / L. The pH value was generally stable between 7.2-7.4, but it would drop briefly by 0.1-0.2 units within 1-2 hours after feeding each day. The ammonia nitrogen concentration showed a slow upward trend, increasing from 0.2 mg / L to 0.5 mg / L. Based on these data, the warning thresholds for dissolved oxygen were set as: normal (>6.5 mg / L), caution (6.0-6.5 mg / L), warning (5.5-6.0 mg / L), and danger (<5.5 mg / L).

[0162] Correlation analysis found that ammonia nitrogen concentration began to rise 2 hours after each feeding, and the rate of increase was positively correlated with the feeding amount. The effect evaluation experiment showed that increasing the amount of microbial preparations could reduce ammonia nitrogen concentration by 20% within 3 hours, while partial water changes (30% water volume) could reduce ammonia nitrogen concentration by 50% within 1 hour. Based on these data, the optimized regulation strategy is: when the ammonia nitrogen concentration reaches 0.4 mg / L, first increase the amount of microbial preparations; if the ammonia nitrogen concentration is still higher than 0.5 mg / L after 6 hours, perform a 30% partial water change.

[0163] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0164] (1) Adjust the water quality parameters of the hatching water body to obtain a target water body that meets the requirements of the classification warning threshold. In the target water body, record the high-definition video of the eel seedlings to obtain the original data of the seedlings' swimming trajectories.

[0165] (2) Perform image processing on the original data of the seedlings' swimming trajectories to obtain the motion parameters of the individual seedlings, including swimming speed, turning frequency, and activity range. Then, perform statistical analysis on the motion parameters to obtain the group behavior characteristics, including aggregation degree, distribution uniformity, and overall activity.

[0166] (3) Collect and analyze the feeding behavior of the eel seedlings in the target water body to obtain the feeding behavior data. Then, perform time series analysis on the feeding behavior data to obtain the feeding rhythm and the change curve of satiety.

[0167] (4) Collect the sound signals of the eel seedlings through an underwater microphone, perform spectral analysis on the sound signals to obtain the acoustic feature spectrum, and perform multi-dimensional feature analysis on the motion parameters, group behavior characteristics, feeding behavior data, and acoustic feature spectrum to obtain a multi-dimensional behavior feature dataset.

[0168] (5) Perform clustering analysis on the multi-dimensional behavior feature dataset to classify the normal, sub-healthy, and diseased states, obtain the classification result of the behavior state, and perform correlation analysis on the classification result of the behavior state and the physiological indicators to obtain the analysis result of the behavior characteristics.

[0169] Specifically, the water quality parameters of the hatching water body are adjusted to reach the target state that meets the requirements of the hierarchical early warning threshold. This step uses precise water quality control equipment, such as an automatic dosing system, a microbubble aeration device, and a biological filter, etc., to precisely adjust key parameters such as dissolved oxygen, pH value, and ammonia nitrogen to ensure that the water quality is stable within the ideal range. In the adjusted target water body, high-resolution underwater cameras are used to continuously record videos of the eel seedlings. The camera system includes multiple high-definition cameras that capture the swimming trajectories of the seedlings at a rate of 60 frames per second. The video data is preprocessed, including denoising, contrast enhancement, and background subtraction, to obtain the original data of the clear swimming trajectories of the seedlings. Image processing and analysis are performed on the original data of the seedlings' swimming trajectories, and computer vision algorithms are used to extract the motion parameters of individual seedlings. This step uses target tracking algorithms, such as the Kalman filter, to track the motion trajectories of each seedling and calculate its instantaneous velocity, acceleration, and turning angle. By analyzing these data, the average swimming speed, turning frequency, and activity range of each seedling are obtained. At the same time, statistical analysis is performed on the motion data of the entire group to calculate group behavior characteristics such as the aggregation index (such as Ripley's K function), spatial distribution uniformity (such as Moran's I index), and overall activity level (such as average motion energy).

[0170] The feeding behaviors of the eel seedlings in the target water body are collected and analyzed. This process uses a specialized high-speed camera system to capture the feeding actions of the seedlings and record the frequency, duration, and intensity of each suction. Through image recognition algorithms, such as convolutional neural networks (CNNs), the feeding behaviors are automatically recognized and counted. Time series analysis is performed on the collected feeding behavior data, using autoregressive integrated moving average (ARIMA) models or wavelet analysis to reveal the periodic patterns of feeding and obtain the feeding rhythm. At the same time, by fitting the relationship between the cumulative food intake and time, the change curve of the satiety level is fitted to reflect the change of the hunger and satiety state of the seedlings. To obtain more comprehensive behavior information, a high-sensitivity underwater microphone array is used to collect the sound signals of the eel seedlings. The collected original sound signals are filtered and denoised, and then spectral analysis is performed. This step uses the fast Fourier transform (FFT) to convert the time-domain signal into the frequency domain to obtain the frequency distribution characteristics of the sound. By analyzing the energy distribution and time-varying characteristics of different frequency bands, an acoustic feature spectrum reflecting the activity state of the seedlings is obtained. The motion parameters, group behavior characteristics, feeding behavior data, and acoustic feature spectrum are subjected to multi-dimensional feature fusion to form a comprehensive multi-dimensional behavior feature dataset. This step uses feature selection algorithms, such as principal component analysis (PCA) or linear discriminant analysis (LDA), to extract the most representative and discriminative feature combinations.

[0171] Perform clustering analysis on the multi-dimensional behavioral feature dataset using the K-means or hierarchical clustering algorithm to divide the behavioral states of the seedlings into three categories: normal, sub-healthy, and diseased. The clustering results are verified and adjusted by experts to obtain the final behavioral state classification results. Subsequently, perform correlation analysis on the behavioral state classification results and physiological indicators (such as growth rate, metabolic level, immune indicators, etc.), and use multivariate statistical methods such as canonical correlation analysis (CCA) or partial least squares regression (PLS-R) to reveal the corresponding relationship between behavioral characteristics and physiological states, and finally obtain the comprehensive behavioral feature analysis results.

[0172] For example: In an eel seedling hatching system, after water quality regulation, the dissolved oxygen is stable at 7.2 mg / L, the pH value is 7.3, and the ammonia nitrogen concentration is 0.1 mg / L, all of which meet the warning threshold requirements. The high-definition video system continuously collected about 500 GB of raw video data within 24 hours. After image processing and analysis, it was found that the average swimming speed of the seedlings was 5 body lengths / s, the turning frequency was 2 times / minute, and the activity range covered 70% of the water body. Group behavior analysis showed that the aggregation index fluctuated between 0.8 and 1.2 at different times, the spatial distribution uniformity index was 0.75, and the overall activity level fluctuated significantly before and after feeding, with the peak occurring within 30 minutes after feeding. Feeding behavior analysis characterized the average feeding frequency of the seedlings as 20 times / hour, lasting 2 - 3 seconds each time. Time series analysis showed an obvious circadian rhythm, with feeding activities reaching the peak 2 hours after the light was turned on. The satiation curve presented an S shape, indicating a rapid increase within 2 hours after feeding and then flattening out.

[0173] Acoustic analysis found that the sounds produced by healthy seedlings were mainly concentrated in the 100 - 500 Hz frequency band, while there was a significant increase in energy in the 700 - 1000 Hz frequency band for sub-healthy individuals. Through multi-dimensional feature fusion and clustering analysis, the seedlings were divided into three categories: normal (accounting for 80%), sub-healthy (accounting for 15%), and suspected diseased (accounting for 5%). Correlation analysis showed that the correlation coefficient between behavioral characteristics and growth rate reached 0.85, and the daily average weight gain rate of the normal behavior group was 20% higher than that of the sub-healthy group.

[0174] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0175] (1) Perform clustering analysis on the behavioral feature analysis results to obtain the typical behavioral patterns of eel seedlings, and based on the typical behavioral patterns, screen the potential stress factors to obtain a list of candidate stress factors;

[0176] (2) Classify the factors in the list of candidate stress factors to obtain three types of stress sources: physical, chemical, and biological, and design intensity gradients for the physical, chemical, and biological stress sources to obtain a multi-level stress intensity scheme;

[0177] (3) According to the multi-level stress intensity scheme, construct an exposure experiment matrix of single-factor and multi-factor combinations to obtain a stress factor exposure experiment plan. Through the stress factor exposure experiment plan, conduct grouped experiments on eel seedlings, record the behavioral changes, physiological indicators and survival rates of the seedlings, and obtain the original stress response data;

[0178] (4) Conduct statistical analysis on the original stress response data to obtain the influence degree and threshold of each stress factor on the seedlings. According to the influence degree and threshold of each stress factor on the seedlings, establish a quantitative relationship between the stress factors and the growth parameters of the seedlings to obtain a stress response mapping table;

[0179] (5) Conduct simulation mapping analysis on the stress response mapping table to obtain a stress response database.

[0180] Specifically, the construction of the stress response database for eel seedlings is a complex and systematic process, involving multiple data analysis and experimental steps. First, conduct cluster analysis on the results of behavioral feature analysis, and use the K-means clustering algorithm to divide the behavioral patterns of the seedlings into several categories. The K-means algorithm iteratively calculates and groups similar behavioral features into one category, and finally obtains the typical behavioral patterns of eel seedlings. For example, several typical patterns such as normal activity, overexcitement, and sluggish response may be identified.

[0181] Based on these typical behavioral patterns, combined with literature research and expert knowledge, screen potential stress factors. This step uses the Delphi method, inviting aquaculture experts to evaluate and score various possible stress factors, and finally summarizing to form a list of candidate stress factors. The factors in the list are then classified into three major categories of stress sources: physical, chemical, and biological. Physical factors include temperature changes, light intensity, water flow velocity, etc.; chemical factors include pH value fluctuations, dissolved oxygen changes, heavy metal pollution, etc.; biological factors include high-density farming, pathogen infection, predation pressure, etc.

[0182] Design intensity gradients for these three types of stress sources, and set 3-5 different intensity levels for each stress factor. For example, temperature stress may be designed into four levels: normal temperature, mild increase (+2°C), moderate increase (+4°C), and severe increase (+6°C). This gradient design forms a multi-level stress intensity scheme, providing a systematic framework for subsequent experiments.

[0183] Based on the multi-level stress intensity scheme, construct an exposure experiment matrix of single-factor and multi-factor combinations. Single-factor experiments examine the influence of a single stress source, while multi-factor experiments simulate the complex situations that may occur in the actual environment. For example, a combined experiment of temperature increase and dissolved oxygen decrease may be designed. These experimental designs constitute a complete stress factor exposure experiment plan.

[0184] According to the experimental protocol, the juvenile rice field eels are grouped for experiments. Each experimental group is set with appropriate replicates and includes a control group. During the experiment, the behavioral changes of the juveniles (such as swimming speed, feeding frequency), physiological indices (such as cortisol level, metabolic rate), and survival rate are continuously recorded. These data constitute the original stress response data. Statistical analysis is performed on the original stress response data, using methods such as analysis of variance (ANOVA) and multiple regression analysis to calculate the influence degree of each stress factor on the juveniles. The influence degree can be quantified by the effect size, such as Cohen's d value. Meanwhile, through dose-response curve analysis, the threshold of each stress factor is determined, that is, the lowest intensity that causes a significant response.

[0185] Based on the influence degree and threshold data, a quantitative relationship between the stress factors and the growth parameters of the juveniles is established. This step uses multiple linear regression or non-linear regression models to describe the functional relationship between the stress intensity and the growth parameters (such as body length, weight growth rate). These relationships are summarized to form a stress response mapping table, visually showing the influence of different stress factors on the growth of the juveniles.

[0186] Finally, simulation mapping analysis is performed on the stress response mapping table. This step uses the Monte Carlo simulation method, by randomly generating a large number of stress scenarios, to predict the responses of the juveniles under different combined stress conditions. After verification and calibration of the simulation results, a comprehensive stress response database is formed.

[0187] For example: In a stress study of juvenile rice field eels, K-means clustering analysis identified four typical behavior patterns: normal active, over-excited, sluggish response, and abnormal aggregation. Based on these behavior patterns, experts evaluated and screened temperature, dissolved oxygen, pH value, ammonia nitrogen concentration, and density as the main stress factors. Four intensity levels were designed for temperature stress: 25°C (control), 28°C, 31°C, and 34°C. The experimental results showed that when the temperature rose to 31°C, the average swimming speed of the juveniles increased by 50%, the feeding frequency decreased by 30%, and the cortisol level increased by 2 times. Statistical analysis indicated that the influence degree (Cohen's d) of temperature on swimming speed was 1.8, and the influence degree on feeding frequency was 1.5. Through dose-response curve analysis, the threshold of temperature stress was determined to be 29.5°C. Multiple regression analysis established a relationship model between temperature and growth rate: GrowthRate = 0.15 - 0.005*(T - 25) 2 , where T is the temperature. Based on this model, the Monte Carlo simulation generated 10,000 temperature fluctuation scenarios, predicting the growth performance of the juveniles under different temperature conditions, and these data were finally integrated into the stress response database.

[0188] In a specific embodiment, the process of executing step S107 may specifically include the following steps:

[0189] (1) Perform hierarchical clustering analysis on the stress factors in the stress response database to obtain stress source classification data, and extract the characteristic indicators of various stress sources according to the stress source classification data to obtain a stress source recognition feature set;

[0190] (2) Perform principal component analysis on the stress source recognition feature set, screen out key warning indicators to obtain a warning indicator system, and design multi-level warning thresholds according to the warning indicator system to obtain warning indicators;

[0191] (3) Conduct effect evaluation in the stress response database to obtain a list of effective mitigation measures for various stress sources, and construct coping strategies for different stress sources and intensities according to the list of effective mitigation measures to obtain a mitigation measure matrix;

[0192] (4) Perform mitigation measure similarity matching on the mitigation measure matrix to obtain a mitigation measure set.

[0193] Specifically, hierarchical clustering analysis is performed on the stress factors in the stress response database. This step uses Ward's minimum variance method to organize them into a hierarchical structure based on the nature and influence characteristics of the stress factors. The clustering process starts from the individual stress factors at the bottom layer and gradually merges similar factors until a complete classification dendrogram is formed. By setting an appropriate truncation threshold, the final stress source classification data is obtained, usually including three major categories: physical, chemical, and biological, with several subcategories under each category. According to the stress source classification data, the characteristic indicators of various stress sources are extracted. This step uses feature selection algorithms such as information gain method or chi-square test to screen out the indicators from the original data that can best represent and distinguish various stress sources. These indicators may include environmental parameters (such as temperature change rate, pH value fluctuation range), physiological indicators (such as cortisol level, metabolic rate change), and behavioral indicators (such as abnormal swimming speed, decreased feeding frequency), etc. These screened indicators constitute the stress source recognition feature set.

[0194] Perform principal component analysis (PCA) on the stress source recognition feature set. Through linear transformation, PCA converts the original features into a set of mutually orthogonal new features (principal components), which are sorted according to the magnitude of the explained data variance. By setting a cumulative contribution rate threshold (usually 85% or 90%), the first few principal components are selected as key early warning indicators. These key indicators together constitute the early warning indicator system, which can reflect the stress state with the least information loss. Based on the early warning indicator system, multi-level early warning thresholds are designed. This step uses the quantile method or the expert judgment method to divide the value range of each early warning indicator into multiple levels, such as normal, mild early warning, moderate early warning, and severe early warning. Each level corresponds to a specific numerical range, and the boundary values of these ranges form the early warning thresholds. The setting of the early warning thresholds needs to consider the physiological tolerance of the eel fry and the actual needs of aquaculture production.

[0195] Evaluate the effectiveness of the mitigation measures in the stress response database. This step uses the meta-analysis method to synthesize the results of multiple experiments and calculate the effect size and 95% confidence interval of each mitigation measure. The effect size is usually expressed by the standardized mean difference (SMD) or the risk ratio (RR). By setting the effect size threshold and the statistical significance level, the effective mitigation measures for various stress sources are screened out to form a list of effective mitigation measures. According to the list of effective mitigation measures, coping strategies for different stress sources and intensities are constructed. This step uses decision tree algorithms such as C4.5 or CART to select the most suitable mitigation measure or combination of measures for each situation according to the stress source type, intensity level, and expected effect. Each path of the decision tree represents a coping strategy, and all strategies are summarized to form a mitigation measure matrix.

[0196] Finally, perform similarity matching analysis on the mitigation measure matrix. This step uses the cosine similarity or Jaccard coefficient to calculate the similarity degree between different coping strategies. Strategies with high similarity are merged or simplified to reduce redundancy and improve management efficiency. After similarity matching and optimization, the final set of mitigation measures is obtained.

[0197] For example: In a study on stress management of eel fry, hierarchical clustering analysis classifies 20 stress factors into 3 categories: physical (such as temperature, light), chemical (such as pH, ammonia nitrogen), and biological (such as density, pathogens). The feature extraction step selects 15 key indicators from 50 original indicators, including the water temperature change rate, dissolved oxygen fluctuation range, cortisol level, etc. The results of principal component analysis show that the first 5 principal components cumulatively explain 88% of the data variance, and these 5 principal components are determined as the core early warning indicators.

[0198] For the warning index of water temperature, four warning levels are set: normal (25 - 28 °C), mild warning (28 - 30 °C or 23 - 25 °C), moderate warning (30 - 32 °C or 21 - 23 °C), and severe warning (> 32 °C or < 21 °C). The effect evaluation shows that for temperature stress, the combined measures of increasing aeration and partial water replacement have the most significant effect, and the standardized mean difference (SMD) is 1.8 (95% CI: 1.5 - 2.1). The decision tree analysis generates a response strategy tree with 50 nodes, and each leaf node represents a specific response plan. The similarity analysis finds that the similarity of the response strategies for partial high temperature and low dissolved oxygen reaches 0.85, so these strategies are merged, and finally a set of mitigation measures containing 30 unique response plans is formed.

[0199] The indoor environment eel fry hatching dynamic monitoring and optimization method in the embodiments of the present application is described above. Next, the indoor environment eel fry hatching dynamic monitoring and optimization system in the embodiments of the present application will be described. Please refer to Figure 2 , an embodiment of the indoor environment eel fry hatching dynamic monitoring and optimization system in the embodiments of the present application includes:

[0200] An optimization processing module 201, configured to perform multi-factor collaborative optimization processing on the pre-collected hatching environment parameters to obtain a dynamic regulation parameter set;

[0201] A demand analysis module 202, configured to perform phased nutritional demand analysis on eel fry based on the dynamic regulation parameter set to obtain a multi-dimensional nutritional supplement plan;

[0202] A structure analysis module 203, configured to perform microbial community structure analysis on the hatching water body according to the dynamic regulation parameter set and the multi-dimensional nutritional supplement plan to obtain a microbial regulation plan;

[0203] A real-time monitoring module 204, configured to perform multi-parameter real-time monitoring on the hatching water body based on the dynamic regulation parameter set, the multi-dimensional nutritional supplement plan, and the microbial regulation plan to obtain a water quality dynamic regulation strategy, where the water quality dynamic regulation strategy includes: a hierarchical warning threshold and a regulation measure selection matrix;

[0204] An environment optimization module 205, configured to optimize the hatching water body based on the water quality dynamic regulation strategy, and perform multi-dimensional data collection and analysis on eel fry in the optimized hatching water body to obtain a behavioral characteristic analysis result;

[0205] A scenario analysis module 206 is configured to perform a multi-source stress factor exposure experiment scenario analysis on the behavioral feature analysis result, generate a multi-source stress factor exposure experiment scenario, and perform a stress experiment on the eel seedlings through the multi-source stress factor exposure experiment scenario to obtain a stress response database;

[0206] A scenario generation module 207 is configured to generate a stress management scenario according to the stress response database, wherein the stress management scenario includes: stress source classification data, early warning indicators, and a set of mitigation measures.

[0207] Through the collaborative cooperation of the above-mentioned various components, by performing multi-factor collaborative optimization processing on the pre-collected incubation environment parameters, a dynamic regulation parameter set is obtained, realizing the precise control and dynamic adjustment of the incubation environment, effectively avoiding the problem of environmental parameter mismatch caused by traditional static control methods, and providing a more stable and suitable growth environment for eel seedlings; based on the dynamic regulation parameter set, a phased nutritional requirement analysis is performed on the eel seedlings to obtain a multi-dimensional nutritional supplement plan, making the nutritional supply more accurate and personalized, effectively improving the bait utilization rate, and promoting the healthy growth of the seedlings; through the analysis of the microbial community structure of the incubation water body, a microbial regulation plan is obtained, realizing the precise regulation of the water body micro-ecosystem, improving the water quality stability and self-purification ability, and creating a better micro-environment for the growth of the seedlings; based on the dynamic regulation parameter set, the multi-dimensional nutritional supplement plan, and the microbial regulation plan, real-time multi-parameter monitoring is performed on the incubation water body to obtain a water quality dynamic regulation strategy, including a hierarchical early warning threshold and a regulation measure selection matrix, making the water quality management more proactive and accurate, being able to timely prevent and respond to various water quality problems, and ensuring the stability of the seedlings' growth environment; through the multi-dimensional data collection and analysis of the eel seedlings in the optimized incubation water body, a behavioral feature analysis result is obtained, realizing the comprehensive and real-time monitoring of the growth status and behavioral features of the seedlings, and providing reliable data support for subsequent management decisions; performing a multi-source stress factor exposure experiment scenario analysis on the behavioral feature analysis result, and obtaining a stress response database through the experiment, systematically revealing the influence mechanism of various potential stress factors on the eel seedlings, and providing a scientific basis for stress prevention and control; finally, the stress management scenario formulated according to the stress response database, including stress source classification data, early warning indicators, and a set of mitigation measures, realizes the precise identification, timely early warning, and effective mitigation of various stress factors, greatly improving the stress resistance and survival rate of the eel seedlings. This systematic, dynamic, and precise combination of incubation management methods not only significantly improves the hatching survival rate and growth quality of eel seedlings, but also reduces the labor intensity and management cost, and improves the hatching efficiency. The method provided by the present invention realizes the comprehensive optimization and precise control of the eel seedling hatching process by combining modern information technology, biotechnology, and refined management, and improves the efficiency and accuracy of the dynamic monitoring and optimization of eel seedling hatching in the indoor environment.

[0208] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions run on a computer, the computer is caused to execute the steps of the method for dynamically monitoring and optimizing the hatching of eel larvae in an indoor environment.

[0209] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described system, system and unit can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0210] The above is the case. The above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments or perform equivalent replacements for some of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for dynamically monitoring and optimizing the hatching of yellow eel fry in an indoor environment, characterized in that, The method for dynamically monitoring and optimizing the hatching of rice field eel larvae in the indoor environment includes: Performing multi-factor collaborative optimization processing on the pre-collected hatching environment parameters to obtain a dynamic regulation parameter set; Based on the dynamic regulation parameter set, analyzing the nutritional requirements of rice field eel larvae in stages to obtain a multi-dimensional nutritional supplement plan; According to the dynamic regulation parameter set and the multi-dimensional nutritional supplement plan, analyzing the microbial community structure of the hatching water body to obtain a microbial regulation plan; Based on the dynamic regulation parameter set, the multi-dimensional nutritional supplement plan and the microbial regulation plan, performing multi-parameter real-time monitoring on the hatching water body to obtain a water quality dynamic regulation strategy, where the water quality dynamic regulation strategy includes: a hierarchical early warning threshold and a regulation measure selection matrix; Based on the water quality dynamic regulation strategy, optimizing the environment of the hatching water body, and in the optimized hatching water body, collecting and analyzing multi-dimensional data of rice field eel larvae to obtain a behavioral characteristic analysis result; Analyzing the multi-source stress factor exposure experiment plan for the behavioral characteristic analysis result, generating a multi-source stress factor exposure experiment plan, and performing a stress experiment on rice field eel larvae through the multi-source stress factor exposure experiment plan to obtain a stress response database; According to the stress response database, generating a stress management plan, where the stress management plan includes: stress source classification data, early warning indicators and a set of mitigation measures.

2. The dynamic monitoring and optimization method for hatching of rice field eel fry in indoor environment according to claim 1, characterized in that, The hatching environment parameters include: water temperature, dissolved oxygen, pH value, light intensity, salinity time series change data and hardness time series change data, and the correlation coefficient matrix between the parameters. The multi-factor collaborative optimization processing of the pre-collected hatching environment parameters to obtain a dynamic regulation parameter set includes: Performing time series decomposition on the water temperature, dissolved oxygen, pH value, light intensity, salinity time series change data and hardness historical data of the pre-collected hatching environment parameters to obtain the trend term, seasonal term and random term of each parameter; Performing polynomial fitting on the trend term of each parameter to obtain the long-term change trend function of each parameter; Performing Fourier analysis on the seasonal term of each parameter to obtain the periodic change characteristics of each parameter; Performing autocorrelation analysis on the random term of each parameter to obtain the short-term fluctuation characteristics of each parameter; Based on the long-term change trend function of each parameter, the periodic change characteristics of each parameter and the short-term fluctuation characteristics of each parameter, constructing a time series prediction model for each parameter to obtain the predicted time series data of each parameter; Performing correlation analysis on the predicted time series data of each parameter to obtain the correlation coefficient matrix between the parameters; According to the correlation coefficient matrix between the parameters, performing principal component analysis on each parameter to obtain the main influencing factors and their weights, and based on the main influencing factors and their weights, performing weighted combination on the predicted time series data of each parameter to obtain a comprehensive environment index; According to the comprehensive environment index, performing collaborative optimization on each parameter through a dynamic programming algorithm to obtain a dynamic regulation parameter set.

3. The dynamic monitoring and optimization method for hatching of rice field eel seedlings in an indoor environment according to claim 2, characterized in that The multi-dimensional nutrition supplement plan includes: a correspondence table of growth stage, bait type, feeding frequency, proportion of nutritional components, and addition amount of trace elements, as well as the influence weights of each nutritional component on the growth indexes of seedlings. Based on the dynamic regulation parameter set, a phased nutritional requirement analysis is carried out on the rice field eel seedlings to obtain a multi-dimensional nutrition supplement plan, including: Segment the data of water temperature, dissolved oxygen, and pH value in the hatching environment parameters to obtain the growth stage division standard of the rice field eel seedlings; According to the growth stage division standard, segment the growth cycle of the rice field eel seedlings to obtain the time range of each growth stage, and conduct morphological measurements on the rice field eel seedlings at each growth stage to obtain data on body length, body weight, and development degree of feeding organs; According to the data of body length, body weight, and development degree of feeding organs, conduct bait palatability analysis on the rice field eel seedlings at each growth stage to obtain the target bait type at each stage; Conduct nutritional component analysis on the target bait type at each stage to obtain the trace element content data of each bait type, and calculate the proportion of nutritional components of the bait formula according to the trace element content data of each bait type and the nutritional requirements of the rice field eel seedlings at each growth stage to obtain the proportion of nutritional components at each growth stage; Collect the feeding behaviors of the rice field eel seedlings at each growth stage to obtain the target feeding frequency and single feeding amount, and calculate the addition amount of trace elements according to the proportion of nutritional components and the single feeding amount at each growth stage to obtain the addition amount of trace elements at each growth stage; Measure the growth indexes of the rice field eel seedlings at each growth stage to obtain data on body length growth rate, body weight growth rate, and survival rate, and conduct correlation analysis between the data of body length growth rate, body weight growth rate, and survival rate and nutritional components through multiple regression analysis to obtain the influence weights of each nutritional component on the growth indexes of seedlings, and generate a multi-dimensional nutrition supplement plan according to the influence weights of each nutritional component on the growth indexes of seedlings and the addition amount of trace elements at each growth stage.

4. The dynamic monitoring and optimization method for hatching of rice field eel fry in indoor environment according to claim 1, characterized in that, According to the dynamic regulation parameter set and the multi-dimensional nutrition supplement plan, conduct microbial community structure analysis on the hatching water body to obtain a microbial regulation plan, including: Sample the hatching water body, extract and amplify microbial DNA from the water sample through high-throughput sequencing technology to obtain the original sequence data of the microbial community, and conduct quality control and chimera removal on the original sequence data to obtain the effective sequence data; Conduct operational taxonomic unit clustering analysis on the effective sequence data to obtain the species composition and abundance information of the microbial community; Calculate the α-diversity index and β-diversity index according to the species composition and abundance information to obtain the diversity evaluation result of the microbial community, and conduct redundancy analysis (RDA) on the microbial community structure and environmental factors according to the diversity evaluation result and the dynamic regulation parameter set to obtain the influence degree of key environmental factors on the microbial community structure; According to the influence degree and the multi-dimensional nutrition supplement plan, screen out the microbial species beneficial to the growth of the rice field eel seedlings to obtain the target microbial flora combination; Cultivation experiments are carried out on the target microbial flora combination to measure its growth curve and metabolites, obtaining the target culture conditions and yield data. Based on the target culture conditions and yield data, the formulation and production process of the microbial preparation are designed to obtain a microbial preparation sample; The stability and activity of the microbial preparation sample are tested to obtain the storage conditions and shelf life of the preparation. Based on the storage conditions and shelf life, a delivery strategy for the microbial preparation is designed. The delivery strategy includes a delivery time series and dosage, and the microbial regulation plan is generated through the delivery strategy.

5. The method for dynamically monitoring and optimizing the hatching of eel fry in an indoor environment according to claim 1, wherein Based on the dynamic regulation parameter set, the multi-dimensional nutrition supplementation plan, and the microbial regulation plan, multi-parameter real-time monitoring is carried out on the hatching water body to obtain a water quality dynamic regulation strategy. Among them, the water quality dynamic regulation strategy includes: hierarchical warning thresholds and a regulation measure selection matrix, including: Continuous sampling and detection are carried out on the dissolved oxygen, pH value, ammonia nitrogen, nitrite, and nitrate contents in the hatching water body to obtain time series data of water quality parameters. Trend analysis is carried out on the time series data of water quality parameters to obtain the change trends and fluctuation ranges of each water quality parameter; According to the change trends and fluctuation ranges of the parameters, multi-level warning thresholds are set for each water quality parameter through the water quality index data in the dynamic regulation parameter set to obtain a hierarchical warning threshold table. Correlation analysis is carried out between the thresholds at each level in the hierarchical warning threshold table and the multi-dimensional nutrition supplementation plan to obtain the interaction relationship between each water quality parameter and bait feeding; According to the interaction relationship between each water quality parameter and bait feeding and the microbial regulation plan, the regulation measures under different water quality conditions are classified and sorted to obtain a preliminary regulation measure library. Effect evaluation experiments are carried out on the measures in the preliminary regulation measure library to obtain the response time and influence degree data of each regulation measure; According to the response time and influence degree data of each regulation measure, the regulation measures are sorted by priority and combined and optimized to obtain a regulation measure selection matrix for different water quality conditions. The hierarchical warning threshold table and the regulation measure selection matrix are fused with data to obtain the water quality dynamic regulation strategy.

6. The dynamic monitoring and optimization method for hatching of eel larvae in an indoor environment according to claim 1, characterized in that, Based on the water quality dynamic regulation strategy, the environment of the hatching water body is optimized. In the optimized hatching water body, multi-dimensional data of rice field eel seedlings are collected and analyzed to obtain the behavioral characteristic analysis results, including: The water quality parameters of the hatching water body are adjusted to obtain a target water body that meets the requirements of the hierarchical warning thresholds. In the target water body, high-definition video recording of rice field eel seedlings is carried out to obtain the original data of the seedlings' swimming trajectories; Image processing is carried out on the original data of the seedlings' swimming trajectories to obtain the motion parameters of individual seedlings, including swimming speed, turning frequency, and activity range. Statistical analysis is carried out on the motion parameters to obtain the group behavioral characteristics, including aggregation degree, distribution uniformity, and overall activity; Collect and analyze the feeding behaviors of Monopterus albus juveniles in the target water body to obtain feeding behavior data, and perform time series analysis on the feeding behavior data to obtain the feeding rhythm and the satiation degree change curve; Collect the sound signals of Monopterus albus juveniles through an underwater microphone, perform spectral analysis on the sound signals to obtain an acoustic feature spectrum, and perform multi-dimensional feature analysis on the motion parameters, group behavior characteristics, feeding behavior data, and acoustic feature spectrum to obtain a multi-dimensional behavior feature dataset; Perform cluster analysis on the multi-dimensional behavior feature dataset, classify normal, sub-healthy, and diseased states to obtain a behavior state classification result, and perform correlation analysis on the behavior state classification result and physiological indicators to obtain the behavior feature analysis result.

7. The dynamic monitoring and optimization method for hatching of eel fry in an indoor environment according to claim 6, characterized in that, Analyze the multi-source stress factor exposure experimental plan for the behavior feature analysis result, generate a multi-source stress factor exposure experimental plan, and conduct a stress experiment on Monopterus albus juveniles through the multi-source stress factor exposure experimental plan to obtain a stress response database, including: Perform cluster analysis on the behavior feature analysis result to obtain the typical behavior patterns of Monopterus albus juveniles, and screen potential stress factors according to the typical behavior patterns to obtain a list of candidate stress factors; Classify the factors in the list of candidate stress factors to obtain three types of stress sources: physical, chemical, and biological, and design intensity gradients for the three types of physical, chemical, and biological stress sources to obtain a multi-level stress intensity plan; According to the multi-level stress intensity plan, construct an exposure experimental matrix of single-factor and multi-factor combinations to obtain a stress factor exposure experimental plan, and conduct a grouped experiment on Monopterus albus juveniles through the stress factor exposure experimental plan, record the behavior changes, physiological indicators, and survival rates of the juveniles to obtain the original stress response data; Perform statistical analysis on the original stress response data to obtain the influence degree and threshold of each stress factor on the juveniles, and establish a quantitative relationship between the stress factors and the growth parameters of the juveniles according to the influence degree and threshold of each stress factor on the juveniles to obtain a stress response mapping table; Perform simulated mapping analysis on the stress response mapping table to obtain the stress response database.

8. The method for dynamically monitoring and optimizing the hatching of yellow eel seedlings in an indoor environment according to claim 7, characterized in that Generate a stress management plan according to the stress response database, where the stress management plan includes: stress source classification data, warning indicators, and a set of mitigation measures, including: Perform hierarchical cluster analysis on the stress factors in the stress response database to obtain stress source classification data, and extract the characteristic indicators of each type of stress source according to the stress source classification data to obtain a stress source identification feature set; Perform principal component analysis on the stress source identification feature set, screen out key warning indicators to obtain a warning indicator system, and design multi-level warning thresholds according to the warning indicator system to obtain the warning indicators; Conduct an effectiveness evaluation on the stress response database to obtain a list of effective mitigation measures for each type of stress source, and construct coping strategies for different stress sources and intensities according to the list of effective mitigation measures to obtain a mitigation measure matrix; Perform similarity matching of mitigation measures on the mitigation measure matrix to obtain the set of mitigation measures.

9. An indoor environment eel fry hatching dynamic monitoring and optimization system, characterized in that, For implementing the dynamic monitoring and optimization method of rice field eel fry hatching in an indoor environment as described in any one of claims 1-8, the dynamic monitoring and optimization system of rice field eel fry hatching in the indoor environment includes: An optimization processing module, configured to perform multi-factor collaborative optimization processing on the pre-collected hatching environment parameters to obtain a dynamic regulation parameter set; A requirement analysis module, configured to perform phased nutritional requirement analysis on the rice field eel fry based on the dynamic regulation parameter set to obtain a multi-dimensional nutritional supplement plan; A structure analysis module, configured to perform microbial community structure analysis on the hatching water body according to the dynamic regulation parameter set and the multi-dimensional nutritional supplement plan to obtain a microbial regulation plan; A real-time monitoring module, configured to perform multi-parameter real-time monitoring on the hatching water body based on the dynamic regulation parameter set, the multi-dimensional nutritional supplement plan, and the microbial regulation plan to obtain a water quality dynamic regulation strategy, wherein the water quality dynamic regulation strategy includes: a hierarchical early warning threshold and a regulation measure selection matrix; An environment optimization module, configured to optimize the environment of the hatching water body based on the water quality dynamic regulation strategy, and perform multi-dimensional data collection and analysis on the rice field eel fry in the optimized hatching water body to obtain a behavior characteristic analysis result; A plan analysis module, configured to perform a multi-source stress factor exposure experiment plan analysis on the behavior characteristic analysis result, generate a multi-source stress factor exposure experiment plan, and perform a stress experiment on the rice field eel fry through the multi-source stress factor exposure experiment plan to obtain a stress response database; A plan generation module, configured to generate a stress management plan according to the stress response database, wherein the stress management plan includes: stress source classification data, early warning indicators, and a set of mitigation measures.

Images