Aquaculture pathogenic microorganism early warning method and system and storage medium

Abstract

The invention relates to the field of aquaculture monitoring, and discloses an aquaculture pathogenic microorganism early warning method and system and a storage medium. The method comprises the following steps: acquiring environmental data through a sensor to form a preliminary data set; classifying environment modes based on the data set and judging a risk interval, and determining a pathogenic microorganism breeding potential acceleration area; predicting the breeding trend of pathogenic microorganisms and evaluating the risk level; triggering environment simulation according to the risk level, and generating a water nutrition balance scheme; tracking breeding change of pathogenic microorganisms, generating an alarm, and constructing an early warning model; updating environment data, activating a prevention and control protocol, and determining a control threshold value; the breeding density is adjusted to achieve sustainable risk reduction. According to the method and the system, the support vector machine, the random forest and other algorithms are combined with real-time monitoring and simulation analysis, so that the problem of difficulty in timely early warning of outbreak of pathogenic microorganisms in aquaculture is solved, and the accuracy of disease prevention and control and the sustainability of a culture environment are improved.

Description

Technical Field

[0001] This application relates to the field of aquaculture monitoring, and in particular to a method, system and storage medium for early warning of pathogenic microorganisms in aquaculture. Background Technology

[0002] With the widespread adoption of high-density aquaculture, the complexity and dynamism of the aquaculture environment have intensified. The rapid proliferation of pathogenic microorganisms in the water has become a major risk factor for large-scale disease outbreaks, directly impacting survival rates and economic returns. Therefore, achieving accurate early warning and proactive intervention regarding their reproductive trends has become a critical issue that urgently needs to be addressed in this field.

[0003] Currently, some existing technologies attempt to control microbial risks through regular monitoring of water quality parameters or human experience, but these methods still have several shortcomings. Firstly, at the data perception level, most methods rely on single or a few environmental indicators, lacking simultaneous collection and in-depth fusion analysis of multi-dimensional parameters such as stocking density, waste accumulation levels, water nutrient indicators, and initial values ​​of pathogenic microorganisms. For example, in typical high-density shrimp farming ponds, dissolved oxygen data alone is insufficient to predict harmful bacterial outbreaks caused by feed residues, because the dynamic relationship between organic matter accumulation and microbial reproduction is not analyzed, resulting in an inability to fully capture the complete causal chain of risk evolution. Secondly, at the risk identification and early warning level, traditional methods are mostly based on static thresholds or linear rules, failing to effectively handle the nonlinear interactions and spatiotemporal heterogeneity of parameters in high-density farming environments. Thirdly, at the model construction and decision support level, existing early warning models often disconnect historical patterns from real-time dynamics, resulting in limited prediction accuracy and a lack of closed-loop verification and adaptive optimization of the effectiveness of control measures. Fourth, although some solutions use automated equipment, their functions are mostly limited to data collection and display, failing to achieve closed-loop control throughout the entire process from risk identification, environmental simulation, nutritional regulation to stocking density adjustment. This results in slow system response, low prevention and control efficiency, and makes it difficult to achieve the goal of sustainable risk management of the aquaculture environment.

[0004] To address the above shortcomings, this application constructs a full-process early warning and intervention mechanism that combines multi-source sensor data fusion, intelligent algorithm analysis such as support vector machines and random forests, and dynamic simulation and feedback control. This mechanism covers environmental monitoring, risk identification, early warning modeling, and density regulation, solving the problems of low accuracy in predicting microbial reproduction trends and delayed response in traditional methods. It also improves the timeliness, accuracy, and system sustainability of disease prevention and control in aquaculture. Summary of the Invention

[0005] This application provides a method, system, and storage medium for early warning of pathogenic microorganisms in aquaculture, which solves the problems of low accuracy in predicting microbial reproduction trends and delayed response in traditional methods, and improves the timeliness, accuracy, and system sustainability of disease prevention and control in aquaculture.

[0006] In a first aspect, this application provides a method for early warning of pathogenic microorganisms in aquaculture, the method comprising: Step S101: Collect aquaculture environmental data through sensors to obtain a preliminary dataset of environmental changes; Step S102: Based on the preliminary dataset of environmental changes, classify environmental patterns and determine risk intervals to identify potential areas of accelerated pathogenic microbial reproduction; Step S103: Predict the trend of pathogenic microorganism reproduction rate and assess the risk level through the potential accelerated reproduction area of ​​the pathogenic microorganism; Step S104: Based on the risk level, trigger environmental change simulation to obtain a water nutrient balance scheme; Step S105: Using the water nutrient balance scheme, track changes in the reproduction rate of pathogenic microorganisms and generate alarm signals to construct a disease outbreak early warning model; Step S106: Based on the disease outbreak early warning model, update the aquaculture environment data and activate the prevention and control protocol to determine the pathogen control threshold; Step S107: By adjusting the breeding density through the pathogen control threshold, a sustainable risk reduction state of the aquaculture environment is obtained.

[0007] Secondly, this application provides an early warning system for pathogenic microorganisms in aquaculture, used to implement the aforementioned early warning method for pathogenic microorganisms in aquaculture, the system comprising: The data acquisition module is used to collect aquaculture environmental data through sensors, integrate aquaculture density data, water nutrient indicators, waste accumulation levels and initial values ​​of pathogenic microorganism reproduction, and obtain a preliminary dataset of environmental changes after preprocessing and updating. The region determination module is used to classify environmental patterns based on the preliminary dataset of environmental changes using a support vector machine algorithm, determine high-risk areas, identify areas of nutritional abnormalities, and comprehensively determine areas with potential accelerated reproduction of pathogenic microorganisms. The risk assessment module is used to compare historical and current reproduction data through the potential accelerated reproduction areas of the pathogenic microorganisms, predict the reproduction rate trend using the random forest algorithm, calculate the degree of proximity of the critical concentration, and obtain the risk level after calibration. The scheme generation module is used to determine whether the assessment value exceeds a preset threshold based on the risk level, trigger environmental change simulation, analyze the impact of nutrient distribution, adjust parameters to generate a water nutrient balance scheme, and monitor and verify the effect. The model building module is used to track changes in the reproduction rate of pathogenic microorganisms according to the water nutrient balance scheme, generate alarm signals, build a preliminary early warning model by combining historical data, and adjust the parameter weights to obtain a disease outbreak early warning model. The threshold determination module is used to integrate risk warning and economic benefit indicators based on the disease outbreak early warning model, update aquaculture environment data, activate prevention and control protocols, determine pathogen control thresholds, and optimize the triggering conditions of prevention and control protocols. The density adjustment module is used to generate and distribute aquaculture density adjustment report based on the pathogen control threshold. It adjusts the density through an automated module to match dynamic environmental requirements and monitors and analyzes data to obtain a sustainable risk reduction status.

[0008] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when run by a processor, causes the processor to execute the aforementioned method for early warning of pathogenic microorganisms in aquaculture.

[0009] This application proposes a method, system, and storage medium for early warning of pathogenic microorganisms in aquaculture, solving the problems of low accuracy in predicting microbial reproduction trends and delayed response in traditional methods, thereby improving the timeliness, accuracy, and sustainability of disease prevention and control in aquaculture. Compared with the prior art, the beneficial effects of the technical solution of this application are at least as follows: First, by integrating multi-source sensor data and adopting data fusion methods, a comprehensive environmental information system was constructed, including aquaculture density, water nutrient indicators, waste accumulation levels, and initial values ​​of pathogenic microorganisms. This system enables comprehensive perception and dynamic analysis of the aquaculture environment from multiple dimensions and parameters, overcoming the limitations of traditional methods that rely on a single or few environmental indicators.

[0010] Secondly, the support vector machine algorithm is used to classify environmental patterns and identify risk intervals, and the random forest algorithm is combined to predict the reproduction trend of pathogenic microorganisms. This can effectively handle the nonlinear interaction and dynamic fluctuation of environmental parameters in high-density aquaculture environments, and improve the accuracy of risk identification and the timeliness of early warning.

[0011] Third, by introducing an environmental change simulation mechanism, a water nutrient balance plan can be dynamically generated according to the risk level, and a prevention and control protocol and microbial control threshold can be triggered based on the early warning model, realizing closed-loop control from risk perception, simulation intervention to threshold regulation, and enhancing the system's ability to adapt to and continuously optimize the dynamic aquaculture environment.

[0012] Fourth, by linking the pathogen control threshold with the adjustment of stocking density and using an automated execution module to optimize density, the risk of pathogen reproduction can be effectively reduced, promoting the aquaculture environment to reach and maintain a sustainable state of reduced risk, and enhancing the initiative and long-term effectiveness of aquaculture management. Attached Figure Description

[0013] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a flowchart illustrating a method for early warning of pathogenic microorganisms in aquaculture according to this application; Figure 2 This is a map showing the environmental parameters and high-risk zones in this application. Figure 3 This is a prediction chart of the reproduction trend of pathogenic microorganisms in this application; Figure 4 This is a graph showing the pathogen concentration tracking and alarm generation in this application; Figure 5 This is a schematic diagram of the structure of an early warning system for pathogenic microorganisms in aquaculture according to this application. Detailed Implementation

[0015] This application provides a method, system, and storage medium for early warning of pathogenic microorganisms in aquaculture. The terms "first," "second," "third," "fourth," etc. (if present)," in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of a method for early warning of pathogenic microorganisms in aquaculture, as described in this application, includes: Step S101: Collect aquaculture environmental data through sensors to obtain a preliminary dataset of environmental changes.

[0017] In one specific embodiment, step S101 may specifically include the following steps: The sensors collect data in real time on aquaculture density, water nutrient indicators, waste accumulation levels, and initial values ​​of pathogenic microorganism reproduction in the aquaculture environment. The data fusion method is used to integrate the aquaculture density data, the water nutrient index, the waste accumulation level, and the initial reproduction value of pathogenic microorganisms to form comprehensive environmental information. Based on the comprehensive environmental information, a preliminary dataset reflecting the dynamic changes in the aquaculture environment is constructed. The preliminary dataset is preprocessed to remove outliers and standardize the data to obtain a preliminary dataset of environmental changes suitable for subsequent analysis.

[0018] Specifically, in aquaculture scenarios, multiple types of sensors are deployed to achieve real-time data acquisition. Stocking density data is monitored and acquired by density sensors, specifically the number of cultured organisms per cubic meter of water, such as 300 fish per cubic meter in fish farming and 500 shrimp per cubic meter in shrimp farming. Water nutrient indicators are collected by water quality sensors, primarily covering nitrogen and phosphorus content, for example, specific data such as 2 mg / L nitrogen and 3 mg / L phosphorus. Waste accumulation levels are tracked by waste sensors, measured by organic matter concentration, such as 0.5 g / L organic matter concentration. Initial pathogenic microorganism reproduction values ​​are captured by microbial sensors, characterized by bacterial counts, such as per milliliter. The number of bacteria.

[0019] After data collection, a data fusion method was used to integrate the four types of data to form comprehensive environmental information. This process employed a weighted average fusion method combined with a Kalman filter algorithm to process data noise. During the weighted average fusion, specific weights were assigned: aquaculture density data had a weight of 0.4, water nutrient indicators had a weight of 0.3, waste accumulation levels had a weight of 0.2, and initial pathogen reproduction values ​​had a weight of 0.1. This was achieved using the formula... The comprehensive environmental information vector is calculated, where The integrated environmental information vector consists of A representing aquaculture density data, B representing water nutrient indicators, C representing waste accumulation levels, and D representing the initial value of pathogenic microorganism reproduction. The Kalman filter algorithm calculates state estimates through a prediction step, then corrects these estimates using new measurements, recursively eliminating data noise. This ensures that the integrated environmental information accurately reflects the interactions between different data points, solving the problem of inaccurate environmental assessments caused by biases in single data sources.

[0020] A preliminary dataset reflecting the dynamic changes in the aquaculture environment was constructed based on comprehensive environmental information. The comprehensive environmental information was organized into a structured table by timestamp, with each row corresponding to complete data for a specific time point. Each column contained weighted values ​​for four indicators: aquaculture density, water nutrition, waste accumulation, and microbial reproduction, clearly presenting the dynamic relationship of each indicator over time. The preliminary dataset underwent preprocessing to remove outliers and perform standardization. Outlier removal employed a box plot method, setting 1.5 times the interquartile range as the criterion; data points exceeding this range were identified as outliers and removed to avoid interference from extreme data in subsequent analysis. Standardization was performed using the min-max standardization method, using the formula... Scale all data to the range of 0 to 1, where E is the standardized data value and F is the original data value before preprocessing. This represents the minimum value of the corresponding indicator in the initial dataset. This represents the maximum value of the corresponding indicator in the initial dataset. The processed data forms the initial dataset of environmental changes, suitable as input for subsequent algorithms such as support vector machines.

[0021] Step S102: Based on the preliminary dataset of environmental changes, classify environmental patterns and determine risk intervals to identify potential areas of accelerated pathogen reproduction.

[0022] In one specific embodiment, step S102 may specifically include the following steps: For the preliminary dataset of environmental changes, record time-series features; The support vector machine algorithm is used to process the preliminary environmental change dataset and the time series features to obtain the water nutrient distribution pattern and the rate of waste accumulation. Determine whether the rate of waste accumulation in a certain area exceeds a preset accumulation threshold; if so, mark the area as a high-risk zone. For the high-risk area, relevant environmental features are extracted from the preliminary environmental change dataset to identify key factors affecting waste accumulation; By analyzing the nutrient distribution patterns in the water bodies, areas of nutrient excess or deficiency in aquaculture water bodies can be identified. Based on the high-risk zones and the areas of nutrient excess or deficiency, the potential areas for accelerated pathogenic microbial reproduction are determined.

[0023] Specifically, time-series features are recorded for the preliminary environmental change dataset, and feature values ​​such as mean, variance, and autocorrelation coefficient are extracted. For example, if the waste accumulation level increases from 0.3 g / L to 0.7 g / L within 24 hours, the calculated mean is 0.5 g / L, variance is 0.04, and autocorrelation coefficient is 0.8. These feature values ​​provide data support for subsequent environmental pattern classification. Furthermore, the aquaculture density data, water nutrient indicators, waste accumulation level, initial values ​​of pathogenic microorganism reproduction, and time-series features (such as mean, variance, and autocorrelation coefficient) from the preliminary environmental change dataset are integrated into an input feature matrix. Each row of the feature matrix corresponds to an environmental data vector at a given time point, and each column corresponds to a specific environmental parameter. For example, the vector at a certain time point includes a stocking density of 300 fish / m³, nitrogen content of 2 mg / L, waste accumulation of 0.5 g / L, and initial values ​​of microorganisms. Data including CFU / mL, mean 0.5 g / L, variance 0.04, and autocorrelation coefficient 0.8.

[0024] When processing the feature matrix using the Support Vector Machine (SVM) algorithm, the algorithm's normalization parameter is set to 1.0 and the gamma value to 0.1. A radial basis function is selected as the kernel function. Classification is achieved by mapping low-dimensional feature vectors to a high-dimensional feature space and constructing a maximum margin hyperplane. Specifically, for the water nutrient distribution pattern, the time-series characteristics of nutrient indicators such as nitrogen and phosphorus content are used as the core classification criteria, outputting pattern categories such as uniform distribution, local clustering, and stratified distribution. For the rate of waste accumulation, the linear regression slope of the waste accumulation level is used as the classification index, outputting low-rate (…). The study established three velocity levels: high velocity (0.01~0.02 g / (Lh)), medium velocity (0.01~0.02 g / (Lh)), and high velocity (>0.02 g / (Lh)). This allowed the study to establish a preliminary dataset of environmental changes and the correspondence between water nutrient distribution patterns and waste accumulation rates, thus solving the technical problem that traditional methods could not quantify environmental patterns and waste accumulation patterns.

[0025] When determining whether the waste accumulation rate in a certain area exceeds a preset accumulation threshold, the preset threshold is determined based on the aquaculture species and scenario. For example, in fish farming, it is set to 0.015 g / (Lh). If the waste accumulation rate in a certain area output by the support vector machine is 0.018 g / (Lh), exceeding this threshold, the area is marked as a high-risk zone, and its spatial coordinates (such as the X1-Y1 to X2-Y2 range of the aquaculture pond) and time series characteristics are recorded to provide location data for subsequent analysis. For high-risk zones, environmental characteristic data such as water temperature, pH value, dissolved oxygen content, and aquaculture density are extracted from the preliminary environmental change dataset. Key factors are screened by calculating the Pearson correlation coefficient between each feature and the waste accumulation rate. For example, if the correlation coefficient between water temperature and waste accumulation rate is 0.85 and the correlation coefficient between dissolved oxygen content and waste accumulation rate is -0.7, then high water temperature and low dissolved oxygen are identified as key factors affecting waste accumulation. This process clarifies the correlation logic between environmental parameters and waste accumulation within the high-risk zone, solving the problem that traditional methods struggle to locate the key driving factors of waste accumulation.

[0026] When identifying nutrient excess or deficiency areas by analyzing the nutrient distribution patterns in water bodies, if the pattern is a localized cluster with nitrogen content exceeding 3 mg / L and phosphorus content exceeding 1.5 mg / L, the clustered area is marked as a nutrient excess area; if the nitrogen content is below 1 mg / L and the phosphorus content is below 0.5 mg / L, it is marked as a nutrient deficiency area. At the same time, the spatial range of these areas and the specific values ​​of nutrient indicators are recorded. For example, a nutrient excess area has a nitrogen content of 3.2 mg / L and a phosphorus content of 1.6 mg / L, and the spatial range covers the X3-Y3 to X4-Y4 area of ​​the aquaculture pond. Using a spatial overlay analysis method, the spatial coordinates of high-risk intervals are matched with the spatial coordinates of nutrient-overabundant / nutrient-deficient areas. If an area belongs to both a high-risk interval and a nutrient-overabundant area, for example, the overlap between the high-risk interval X1-Y1 to X2-Y2 and the nutrient-overabundant area X3-Y3 to X4-Y4 is X5-Y5 to X6-Y6, then this overlap area is identified as a potential accelerated area for pathogenic microorganism reproduction. If a high-risk interval overlaps with a nutrient-deficient area, and the initial microbial values ​​indicate the presence of suitable conditions for the reproduction of specific pathogenic microorganisms (such as Vibrio) (e.g., Vibrio easily reproduces in an environment with insufficient nutrients but abundant organic matter), it is also included in the potential accelerated area. Simultaneously, the acceleration potential score of the region is calculated using the formula G = 0.6H + 0.4I, where G is the acceleration potential score, H is the waste accumulation rate level score in the high-risk area (high speed is 3 points, medium speed is 2 points, low speed is 1 point), and I is the nutritional abnormality score (excess / deficiency exceeding the standard value by 20% is 3 points, 10%~20% is 2 points, and <10% is 1 point). If a region has H of 3 points and I of 3 points, then G is 3 points. This further quantifies the potential acceleration risk, solves the technical problem that traditional methods cannot accurately locate high-risk breeding areas of pathogenic microorganisms, and provides a clear target area for subsequent early warning.

[0027] Step S103: Predict the trend of pathogenic microorganism reproduction rate and assess the risk level through the potential accelerated reproduction area of ​​the pathogenic microorganism.

[0028] In one specific embodiment, step S103 may specifically include the following steps: Obtain the historical reproduction rate records of pathogenic microorganisms corresponding to the potential accelerated reproduction areas of the pathogenic microorganisms; The historical reproduction rate records of the pathogenic microorganisms are compared with the current reproduction data of the pathogenic microorganisms collected by the sensors to extract the trend characteristics of the reproduction rate of the pathogenic microorganisms. The random forest algorithm is used to analyze the changing trend characteristics and predict the future trend of pathogen reproduction rate in the potentially accelerated region of pathogen reproduction; Based on the future trend of the pathogenic microorganism's reproduction rate, calculate the degree of closeness between the pathogenic microorganism concentration and the preset critical concentration; Based on the degree of proximity, a preliminary risk level assessment value is determined. This preliminary risk level assessment value is then calibrated by combining historical risk assessment data and actual disease occurrence data to assess the risk level.

[0029] Specifically, when obtaining historical reproduction rate records of pathogens corresponding to areas with potential accelerated pathogen reproduction, it is necessary to retrieve historical data for that area over the past 30 days from the database. Data dimensions include pathogen concentrations at different times of the day, corresponding environmental parameters (such as water temperature, dissolved oxygen, and nitrogen content), and stocking density. For example, the historical record of a potential accelerated area shows that the pathogen concentration from day 1 to day 30 was... The concentration fluctuated within a range, with peak concentrations occurring on days 5, 12, and 20, respectively. CFU / mL CFU / mL The CFU / mL values ​​were all measured, with peak water temperatures above 25°C and dissolved oxygen below 5 mg / L. These historical data need to be precisely correlated with the spatial coordinates and time-series characteristics of potential acceleration areas to ensure a complete match between the data source and the target region. This provides a reliable benchmark for subsequent comparative analysis and addresses the technical problem of prediction bias caused by the lack of regionalized historical data support in traditional methods.

[0030] When comparing historical reproduction rate records of pathogenic microorganisms with current pathogenic microorganism reproduction data collected by sensors, the current data must cover the same parameter dimensions as the historical data. For example, the pathogenic microorganism concentration collected at a certain point in time is... CFU / mL, water temperature 26℃, dissolved oxygen 4.8 mg / L, stocking density 320 fish / m³. By calculating the concentration differences under the same environmental parameters, a differential sequence was constructed. For example, within the range of water temperature 25–26℃ and dissolved oxygen 4.5–5 mg / L, the historical average concentration was [missing value]. CFU / mL, the difference between the current concentration and this average value is CFU / mL; simultaneously extract trend characteristics, including the slope of the linear regression of concentration differences (e.g., the slope of the difference sequence over the past 7 days). CFU / mL / day), fluctuation range (e.g., the standard deviation of the difference series is...). CFU / mL) and frequency of peak occurrence (e.g., concentration exceeding [a certain value] in the past 10 days). The CFU / mL was measured three times. These characteristics directly reflect the direction and intensity of the current reproductive rate change relative to historical levels, establishing a correspondence between historical data, current data, and trend characteristics.

[0031] When using the random forest algorithm to analyze trend characteristics, the algorithm parameters are set as follows: 200 trees, maximum depth of each tree 15, and feature selection ratio 0.8. The extracted linear regression slope, fluctuation amplitude, peak frequency, and corresponding environmental parameters (water temperature, dissolved oxygen, nitrogen content) are used as input features to construct a prediction model. The algorithm generates multiple decision trees by randomly sampling samples and features. Each decision tree is split based on a feature threshold (e.g., by slope). (CFU / mL / day is used as a dividing node to determine the strength of the reproductive rate growth trend), and the prediction result is finally determined by majority vote. For example, after inputting the current trend characteristics, the model outputs a prediction curve of pathogen concentration for the next 7 days, showing the concentration from day 1 to day 7 from... CFU / mL gradually increased to The CFU / mL value predicts a continuous increase in the future. This process integrates multi-dimensional trend characteristics with environmental parameters to solve the technical problem that traditional single-parameter prediction is difficult to cope with dynamic environmental changes, thereby improving the robustness of the prediction.

[0032] When calculating the approximation of a predetermined critical concentration based on the future trend of pathogenic microorganism reproduction rate, the predetermined critical concentration is determined according to the disease susceptibility threshold of the cultured species, for example, set at [value missing] in fish farming. CFU / mL. The difference between the daily concentration values ​​in the prediction curve and the critical concentration is calculated to obtain the proximity sequence, using the formula J=KL, where J is the proximity (unit: CFU / mL), K is the preset critical concentration, and L is the predicted concentration. For example, the predicted concentration on day 7 is... CFU / mL, then J= = CFU / mL, and simultaneously calculate the rate of change of proximity (e.g., the rate of decrease of J over the past 3 days is...). (CFU / mL / day), quantifying the speed at which risk approaches.

[0033] When determining the initial risk level assessment value based on the degree of proximity, the degree of proximity is divided into multiple intervals and corresponding to different assessment values, for example, J> CFU / mL corresponds to an assessment value of 1 (low risk). CFU / mL <J≤ CFU / mL corresponds to an assessment value of 3 (medium risk), J≤ CFU / mL corresponds to an assessment value of 5 (high risk), and the assessment value is adjusted based on the rate of change in proximity (e.g., the rate of decrease exceeds...). (If the CFU / mL / day is used to evaluate the value, add 1). For example, if the current J= CFU / mL and the rate of decrease is The initial assessment value is 4 (CFU / mL / day). When calibrating by combining historical risk assessment data and actual disease occurrence data, historical assessment values ​​and actual disease occurrences within the same proximity interval over the past 12 months are retrieved to calculate the calibration coefficient, using the formula M=N / P, where M is the calibration coefficient, N is the number of actual disease occurrences within that interval, and P is the number of historical assessments within that interval. The following calibration rules are adopted: when the calibration coefficient M>0.8, the calibrated assessment value = min(5, initial assessment value + 2); when 0.5<calibration coefficient M≤0.8, the calibrated assessment value = min(5, initial assessment value + 1); when the calibration coefficient M≤0.5, the calibrated assessment value = the initial assessment value. For example, if N=8, P=10, and the initial assessment value is 3, then M=0.8, and the calibrated assessment value = min(5, 3+1)=4. Calibration using historical data reduces prediction errors, solves the technical problem of low correlation between traditional risk assessment and actual disease occurrence, and ensures the accuracy of risk level assessment.

[0034] Step S104: Based on the risk level, trigger environmental change simulation to obtain a water nutrient balance scheme.

[0035] In one specific embodiment, step S104 may specifically include the following steps: Obtain the risk level assessment value corresponding to the risk level, determine whether the risk level assessment value exceeds the preset median threshold, and if so, trigger the environmental change simulation of the aquaculture environment. The environmental change simulation was used to analyze the impact of nutrient distribution in aquaculture water on the reproduction of pathogenic microorganisms. Based on the impact analysis results, the water nutrient input parameters are adjusted to generate an optimized water nutrient balance scheme.

[0036] Specifically, risk level assessment values ​​corresponding to risk levels are obtained, with a range of 1 to 5, where 1 corresponds to low risk, 2-3 to medium risk, and 4-5 to high risk. The preset median threshold is determined based on the risk tolerance of the aquaculture scenario; for example, it is set to 3 in fish farming. If the risk level assessment value is 4, exceeding the preset median threshold of 3, an environmental change simulation of the aquaculture environment is triggered. The trigger signal simultaneously carries the current environmental parameters of the potential acceleration area, including water nutrient indicators (nitrogen content 3 mg / L, phosphorus content 1.2 mg / L), waste accumulation level (0.6 g / L), water temperature (26℃), and stocking density (320 fish / cubic meter), ensuring that the simulation process is based on the actual environmental state of the target area. The environmental change simulation uses a time series analysis method, selecting the Autoregressive Integrated Moving Average (ARIMA) model, setting the model parameters p=2, d=1, and q=2. The water nutrient distribution data (such as daily nitrogen and phosphorus concentrations at different depths) and pathogenic microorganism reproduction rate data of the area over the past 30 days are input to construct a correlation model between nutrient distribution and reproduction rate. During the simulation, multiple nutrient distribution scenarios were generated by changing the nitrogen and phosphorus inputs. For example, the nitrogen input was adjusted from the current 0.8 kg / day to 0.6 kg / day, 0.7 kg / day, and 0.9 kg / day, and the phosphorus input was adjusted from 0.3 kg / day to 0.2 kg / day, 0.25 kg / day, and 0.35 kg / day. Each scenario corresponded to a predicted value for the pathogenic microorganism's reproduction rate. By calculating the difference between the predicted reproduction rate under different scenarios and the current reproduction rate, the degree of influence of the nutrient distribution on pathogenic microorganism reproduction was obtained. The formula is Q = RS, where Q is the degree of influence, R is the adjusted predicted reproduction rate, and S is the current reproduction rate (e.g., the current reproduction rate is 5 × 10⁻⁶). 4 CFU / mL . In a certain scenario, the predicted value is 3.5 × 10⁻⁶. 4 CFU / mL . If the time is 10, then Q = -1.5 × 10 4 CFU / mL . The figure for 1 day indicates that the nutritional adjustment reduced the reproductive rate by 1.5 × 10⁻⁶. 4 CFU / mL . (Days) Establish the correspondence between nutrient input parameters, nutrient distribution scenarios and the degree of influence, and clarify the regulatory effect of different nutrient adjustments on reproductive rate.

[0037] When adjusting water nutrient input parameters based on the impact analysis results, priority should be given to the parameter combination that maximizes the reduction in pathogenic microorganism reproduction rate while meeting the growth requirements of aquaculture organisms. For example, when the nitrogen input is 0.7 kg / day and the phosphorus input is 0.25 kg / day, the impact level Q = -2 × 10⁻⁶. 4 CFU / mL .If, for a given period of time, the nitrogen content in the water is maintained at 2.2 mg / L and the phosphorus content at 0.8 mg / L, which meets the nutritional range required for fish growth (nitrogen 1.5-2.5 mg / L, phosphorus 0.5-1.0 mg / L), then this parameter combination will be used as the core adjustment indicator. Simultaneously, the water exchange rate will be adjusted (e.g., increasing from 0.5 times / day to 0.8 times / day) to help maintain nutrient balance, resulting in an optimized nutrient balance plan. The plan must clearly record the comparison of nutrient parameters before and after the adjustment (e.g., nitrogen input decreasing from 0.8 kg / day to 0.7 kg / day, phosphorus input decreasing from 0.3 kg / day to 0.25 kg / day), the expected nutrient distribution (uniform distribution, nitrogen and phosphorus concentration deviation ≤10% at all depths), and the expected effect on controlling the reproductive rate (a decrease of 2 × 10⁻⁶). 4 CFU / mL . This process, by quantifying the correlation between nutrient adjustment and reproduction rate, solves the technical problem that traditional methods struggle to accurately regulate water nutrients to inhibit the reproduction of pathogenic microorganisms, ensuring that the solution is both risk-controlled and adaptable to aquaculture production.

[0038] Step S105: Using the water nutrient balance scheme, track changes in the reproduction rate of pathogenic microorganisms and generate alarm signals to construct a disease outbreak early warning model.

[0039] In one specific embodiment, step S105 may specifically include the following steps: Based on the aforementioned water nutrient balance scheme, the impact of reducing waste accumulation in aquaculture water bodies on the reproduction of pathogenic microorganisms was simulated. Time series analysis was used to track the changing trends of pathogenic microorganism reproduction rates; Determine whether the trend of the pathogenic microorganism reproduction rate at a certain moment is close to the preset critical concentration of pathogenic microorganisms; if so, generate an alarm signal. Using the alarm signal as the core input, and combining it with historical disease outbreak data in aquaculture scenarios, a preliminary early warning model is constructed. For the aforementioned preliminary early warning model, key time points in which the reproduction rate of pathogenic microorganisms changed significantly were recorded; The correlation between the key time points and aquaculture environmental parameters is analyzed, and the parameter weights of the preliminary early warning model are adjusted according to the correlation to finally obtain a disease outbreak early warning model.

[0040] Specifically, when simulating the impact of waste accumulation reduction on pathogenic microorganism reproduction in aquaculture waters based on a nutrient balance scheme, a correlation model between waste accumulation and reproduction rate needs to be constructed based on the nutrient input parameters and water exchange rate determined in the scheme. By inputting the current waste accumulation level and the expected reduction in waste accumulation after the scheme is implemented, the waste accumulation values ​​at different time points are simulated. Combined with the linear relationship between waste accumulation and pathogenic microorganism reproduction rate in historical data, the predicted reproduction rate values ​​at corresponding time points are calculated, establishing a correspondence between the reduction in waste accumulation and changes in reproduction rate. This addresses the technical problem that traditional methods struggle to quantify the impact of waste accumulation control on reproduction rate.

[0041] When using time series analysis to track the changing trends of pathogenic microorganism reproduction rates, an autoregressive integral moving average (ARIMA) model was selected, with parameters p=3, d=1, and q=3. Actual reproduction rate data collected daily after the implementation of the scheme were input. The model was decomposed into trend, seasonal, and residual terms to extract the long-term trend and short-term fluctuations of the reproduction rate. Simultaneously, the simulated predicted reproduction rate was compared with the actual tracking data, the error was calculated, and the model parameters were adjusted through error feedback to ensure the accuracy of trend tracking. This process solved the technical problem of unstable reproduction rate tracking under dynamically changing environmental parameters, providing reliable data support for subsequent alarm triggering.

[0042] When determining whether the trend of pathogenic microorganism reproduction rate at a certain moment is approaching a preset critical concentration, the preset critical concentration is determined based on the disease outbreak threshold of the cultured species (e.g., 1.2 × 10⁻⁶ in fish farming). 6 Based on the observed reproductive rate trends, the remaining time to the critical concentration from the current rate is calculated using the formula T = (UV) / W, where T is the remaining time, U is the preset critical concentration, and V is the concentration corresponding to the current reproductive rate (e.g., 2.6 × 10⁻⁶ CFU / mL). 4 CFU / mL . Days x 7 + initial concentration 7 x 10 5 CFU / mL = 8.82 × 10 5 CFU / mL), W is the average daily reproduction rate (e.g., the average for the past 7 days is 3.4 × 10⁻⁶). 4 CFU / mL . If T ≤ 10 days, the trend is determined to be approaching the critical concentration, and an alarm signal is generated. The signal includes the degree of approach (e.g., 8 days remaining), current environmental parameters (water temperature 25℃, dissolved oxygen 5.2mg / L), and coordinates of the potential risk area, which solves the technical problem that traditional methods are difficult to accurately determine when the risk is approaching.

[0043] When constructing a preliminary early warning model using alarm signals as the core input, historical disease outbreak data from aquaculture scenarios over the past 24 months were collected. This included alarm signal characteristics before an outbreak (such as approach time and rate trend), corresponding environmental parameters (water temperature, nutrient indicators, and waste accumulation), and disease occurrence (such as whether an outbreak occurred and its intensity). Twelve features from the alarm signals (such as remaining time, rate slope, and environmental parameter deviation) were used as the input vector, and disease occurrence was used as the output label. A logistic regression algorithm was employed to construct the model, with a regularization parameter λ=0.01. The loss function was minimized using gradient descent to obtain the initial weights for each input feature (e.g., remaining time weight -0.3, rate slope weight 0.5). During model training, 5-fold cross-validation was used to ensure the generalization ability of the preliminary model under different aquaculture scenarios, thus solving the technical problem of low reliability of the early warning model due to a lack of historical data support.

[0044] When recording key time points where the reproduction rate of pathogenic microorganisms changes significantly, the moment when the probability output by the preliminary early warning model exceeds 0.6 (i.e., is judged as high risk) is marked as a key time point. Simultaneously, environmental parameters (e.g., water temperature 26℃, nitrogen content 2.3 mg / L) and reproduction rate data (e.g., a sudden increase in rate to 4 × 10⁻⁶) are recorded at that moment. 4 CFU / mL . (Days) To form a dataset corresponding to key time points and environmental parameters. When analyzing the correlation between key time points and environmental parameters, the Pearson correlation coefficient between each environmental parameter and the key time point is calculated. For example, the correlation coefficient between water temperature and key time points is 0.75, and the correlation coefficient between nitrogen content and key time points is 0.68. Based on the magnitude of the correlation coefficient, the parameter weights of the preliminary early warning model are adjusted. The weight of water temperature is increased from 0.2 to 0.4, and the weight of nitrogen content is increased from 0.15 to 0.3. At the same time, the weight of parameters with low correlation (such as dissolved oxygen, correlation coefficient 0.2) is reduced, and finally, a disease outbreak early warning model is obtained. The adjusted model improves the accuracy of key time point identification by 15%, solves the technical problem of insufficient correlation between model parameters and actual environment leading to prediction bias, and ensures the practicality of the early warning model.

[0045] Step S106: Based on the disease outbreak early warning model, update the aquaculture environment data and activate the prevention and control protocol to determine the pathogen control threshold.

[0046] In one specific embodiment, step S106 may specifically include the following steps: By integrating the pathogenic microorganism risk early warning mechanism with the economic benefit indicators of aquaculture through the disease outbreak early warning model, a comprehensive evaluation basis is formed. A feedback loop processing method is used to input the preliminary dataset of environmental changes collected in real time into the disease outbreak early warning model, and to update the aquaculture environment data in real time. If the alarm signal triggered by the disease outbreak early warning model persists, the preset aquaculture pathogen control protocol will be activated. Based on the aforementioned prevention and control protocol, and in conjunction with updated aquaculture environment data, the pathogen control threshold was determined. Record the correlation between the pathogen control threshold and the corresponding aquaculture environmental parameters; Analyze the aforementioned relationships, optimize the triggering conditions of subsequent prevention and control protocols based on the analysis results, and ensure that the prevention and control protocols are adapted to the dynamic changes in the aquaculture environment.

[0047] Specifically, when integrating pathogen risk warning mechanisms with economic benefit indicators of aquaculture through a disease outbreak early warning model, a quantitative correlation needs to be established between the risk level (values ​​1-5) output by the risk warning mechanism and the economic benefit indicators (yield per unit water volume, feed conversion rate, and disease loss rate). Economic benefit indicators are obtained through historical data statistics from aquaculture scenarios, and the integration formula is set as X=0.6Y+0.4Z, where X is the comprehensive evaluation value, Y is the standardized risk level value (risk level 1 corresponds to 0.2, 2 to 0.4, 3 to 0.6, 4 to 0.8, and 5 to 1.0), and Z is the standardized economic benefit value (actual yield per unit water volume / benchmark value × 0.3 + (2 - actual feed conversion rate / benchmark value) × 0.3 + (1 - actual disease loss rate / benchmark value) × 0.4). This formula calculates the comprehensive evaluation basis, solving the technical problem of traditional risk warning neglecting economic benefits, leading to excessively high prevention and control costs.

[0048] When updating aquaculture environmental data using a feedback loop processing method, the feedback loop cycle is set to 1 hour. A preliminary dataset of real-time environmental changes (culture density, nitrogen content, waste accumulation level, and pathogen concentration) is input into the disease outbreak early warning model. The model adjusts its internal parameters, such as correcting the weights of features like nitrogen content and waste accumulation, using a gradient descent algorithm (learning rate 0.01, 50 iterations), to match the risk level output by the model with the real-time environmental state. After each loop, the updated environmental data is associated with and stored along with the comprehensive assessment value X, forming a dynamically updated environment-assessment database. This addresses the technical problem of traditional static data failing to reflect real-time environmental changes and ensures consistency between environmental data and the early warning model output.

[0049] If the alarm signal triggered by the disease outbreak early warning model persists (with a set duration threshold of 2 hours), the preset prevention and control protocol is activated. The protocol includes tiered response measures, such as different levels of measures (increasing water exchange rate, administering microbial inhibitors, isolating diseased fish areas, etc.) corresponding to different alarm signal durations. During activation, the system automatically retrieves the execution parameters of the corresponding measures and records the activation time, measure type, and initial environmental parameters in the prevention and control log, resolving the technical problem of risk spread caused by the delayed triggering of traditional prevention and control protocols. When determining the pathogen control threshold based on the prevention and control protocol and updated aquaculture environmental data, the control threshold needs to be dynamically adjusted according to different prevention and control measure levels. Different control threshold calculation formulas are set for different measure levels, incorporating real-time environmental parameters (water temperature, dissolved oxygen) as variables. For Level 1 prevention and control measures, the control threshold... Where Ba is the base threshold, which is 1.0 × 10⁻⁶. 6 CFU / mL, Tb is water temperature, Tb0 is the reference water temperature (25℃), Db is dissolved oxygen, Db0 is the reference dissolved oxygen (5 mg / L), K1 and K2 are adjustment coefficients (0.02 and 0.03, respectively). A correspondence between control measure levels, environmental parameters, and control thresholds is established to address the technical problem that traditional fixed control thresholds cannot adapt to dynamic environments.

[0050] When recording the correlation between pathogen control thresholds and corresponding aquaculture environmental parameters, the control thresholds and real-time environmental parameters (water temperature, nitrogen content, dissolved oxygen, and stocking density) should be stored as key-value pairs to form a correlation table, providing structured data for subsequent analysis. When analyzing the correlations to optimize the triggering conditions of the control protocol, the Pearson correlation coefficient is used to calculate the correlation strength between the control thresholds and each environmental parameter. The triggering conditions are adjusted based on the correlation strength, adding environmental parameter constraints to the original triggering conditions. Simultaneously, a dynamic adjustment coefficient for the triggering conditions is set: Dynamic adjustment coefficient = 1 + 0.05 × (real-time water temperature - reference water temperature) - 0.03 × (real-time dissolved oxygen - reference dissolved oxygen). By adjusting the coefficient, the trigger duration threshold is corrected to ensure that the control protocol adapts to the dynamic changes in the aquaculture environment, solving the technical problem of poor adaptability caused by traditional fixed triggering conditions.

[0051] Step S107: By adjusting the breeding density through the pathogen control threshold, a sustainable risk reduction state of the aquaculture environment is obtained.

[0052] In one specific embodiment, step S107 may specifically include the following steps: Based on the pathogen control threshold, a targeted output report is generated, which includes suggestions for adjusting breeding density and corresponding environmental adaptation requirements. The targeted output reports are distributed to the central control system of the aquaculture farm; An automated execution module is used to adjust the stocking density of aquaculture according to the stocking density adjustment recommendations; The adjusted stocking density is matched with the dynamic adaptation requirements of the aquaculture environment; Based on the matching results, the parameters of the aquaculture environment are continuously collected by sensors to form an environmental parameter monitoring dataset. By analyzing the environmental parameter monitoring dataset, it is determined whether the risk of pathogenic microorganism reproduction in the aquaculture environment has been reduced to a preset safe range and can be maintained stably. If so, the sustainable risk reduction status of the aquaculture environment is obtained.

[0053] Specifically, when generating targeted output reports based on pathogen control thresholds, the logical correlation between control thresholds and stocking density must be quantified, and the density adjustment formula should be set as: Adjusted stocking density = Current stocking density × (Current pathogen concentration / Pathogen control threshold). The report must clearly indicate the adjustment recommendations (such as the specific numerical value of the adjusted stocking density from the current stocking density) and environmental adaptation requirements (such as the range to be maintained for dissolved oxygen in the water after adjustment, and the range to be controlled for nitrogen and phosphorus nutrient indicators), and also include the calculation process for the control thresholds and density adjustments to ensure the traceability of the report data.

[0054] When targeted output reports are distributed to the central control system of aquaculture farms, real-time report push is achieved through encrypted data transmission protocols (such as MQTT). Upon receiving the reports, the system automatically parses the density adjustment parameters and environmental adaptation requirements, stores them in the execution command database, and generates execution priority identifiers (e.g., priority is set to 1 for high-risk scenarios and 2 for medium-risk scenarios). During the distribution process, transmission time, reception status, and system feedback information must be recorded synchronously to form a distribution log, preventing command loss or delays and solving the technical problem of response lag caused by traditional manual report delivery, ensuring that adjustment commands reach the execution end in a timely manner. Further automated execution modules retrieve execution commands from the central control system and drive mechanical devices (such as zoned barriers and automatic transfer equipment) to achieve density adjustment. During the adjustment process, real-time density data is collected (every 5 minutes via a density sensor), compared with the target density, and the deviation value (e.g., the difference between the current and target densities) is calculated. A proportional-integral-derivative (PID) control algorithm (proportional coefficient set to 0.3, integral coefficient to 0.1, and derivative coefficient to 0.05) is used to dynamically correct the execution parameters. For example, when the deviation value exceeds a preset range (e.g., 5% of the target density), the operating speed of the mechanical device is adjusted until the current density stabilizes within ±3% of the target density. This closed-loop control process solves the technical problems of low accuracy and poor stability in manual density adjustment, ensuring that the density adjustment meets the reporting requirements.

[0055] When matching the adjusted stocking density with the dynamic adaptation requirements of the aquaculture environment, these requirements are generated based on historical environmental data and a density adaptation model (such as the minimum threshold for dissolved oxygen and the maximum threshold for ammonia nitrogen at a given density). Matching is achieved by calculating the degree of fit between the adjusted density and each environmental parameter. The formula for calculating the degree of fit is set as follows: ,in Score for dissolved oxygen compatibility (1 point for meeting the requirements, otherwise points are deducted according to the deviation ratio). The score is determined by the nitrogen content. For phosphorus content matching score, if A value ≥0.8 is considered a successful match. If the value is less than 0.8, an environmental parameter adjustment command will be triggered (such as starting the oxygenation equipment to increase the level). Adjust nutritional input to reduce or The matching process addresses the technical issue of risk rebound caused by incompatibility between density adjustments and environmental parameters, ensuring that density is adapted to the environment.

[0056] When continuously collecting environmental parameter data based on the matching results, the sensor sampling frequency is set according to the matching results (once every 30 minutes when the match is successful, and once every 10 minutes when the match is unsuccessful). The collected parameters include dissolved oxygen, water temperature, nitrogen content, phosphorus content, pathogen concentration, and stocking density, forming an environmental parameter monitoring dataset. The dataset needs to be stored in a time-series structure, with each record including the sampling time, parameter name, parameter value, and sensor number. Simultaneously, data cleaning algorithms (such as the 3σ criterion) are used to remove outliers (data exceeding three standard deviations from the normal range) to ensure the accuracy of the dataset. This process solves the technical problem of fragmented traditional monitoring data leading to analytical difficulties, providing a reliable data foundation for subsequent status assessment.

[0057] When analyzing environmental parameter monitoring datasets to determine the sustainable risk reduction status, a preset safety range is set (e.g., pathogen concentration below 80% of the control threshold, dissolved oxygen maintained within the required range, and nitrogen and phosphorus content meeting dynamic adaptation standards). A sliding window analysis method (window size set to 24 hours) is used to calculate the compliance rate of each parameter within the window. If the compliance rate is ≥90% for three consecutive windows, and the pathogen concentration shows no upward trend (slope ≤0 through linear regression analysis), then the sustainable risk reduction status is determined to have been achieved. During the analysis, compliance data and trend changes for each window must be recorded simultaneously to generate a status verification report. This addresses the technical problem of traditional risk status assessment lacking long-term stability evaluation, ensuring that the risk reduction effect can be stably maintained.

[0058] Please see Figure 2 , Figure 2An environmental parameter and high-risk zone identification map shows the changes in the standardized ratios of two environmental parameters, "waste accumulation level" and "mean trophic level," over 30 days, while also marking the time range of the "high-risk zone." Over time, the standardized ratios of waste accumulation level and mean trophic level generally show an upward trend. Within the "high-risk zone" marked in gray on the right side of the map, the fluctuation range of both parameters increases significantly, and the values ​​are in the higher range. This indicates that when the standardized ratios of environmental parameters (waste accumulation and trophic level) increase and enter a state of high fluctuation, the corresponding time interval is identified as a high-risk zone. This demonstrates the correlation between changes in environmental parameters and risk zone identification, and also verifies the effectiveness of the system in locating high-risk periods through parameter fluctuation characteristics.

[0059] Please see Figure 3 , Figure 3 This is a pathogen reproduction trend prediction map, displaying historical data, current values, and predicted trends of pathogen concentrations over 35 days, while also marking the time range of critical concentrations and acceleration zones. The map shows that pathogen concentration growth is initially slow, but the growth rate accelerates significantly after entering the acceleration zone; the current value is approaching the critical concentration, and the predicted trend indicates that the concentration will rapidly exceed the critical value. This demonstrates that pathogen reproduction has an accelerated growth phase, where the concentration quickly approaches or even exceeds the critical concentration. This highlights the role of reproduction trend prediction in identifying high-risk phases in advance and verifies that the system's early warning mechanism should focus on concentration changes within the acceleration zone.

[0060] Please see Figure 4 , Figure 4 The pathogen concentration tracking and alarm generation graph shows the trend of pathogen concentration changes over 30 to 44 days. The concentration trend shows a continuous decrease in pathogen concentration starting from day 30, consistently remaining below the warning threshold, alarm threshold, and critical concentration. This directly verifies the effectiveness of the aquatic nutrient balance scheme. For example, the scheme achieves a continuous reduction in pathogen concentration by precisely controlling nitrogen and phosphorus nutrients and combining biofloc and probiotic technologies to cut off the nutrient source for pathogens. Threshold comparisons show that the system did not trigger any warnings or alarms, indicating that the optimized nutrient balance scheme successfully controlled pathogen risk within a safe range, demonstrating the dynamic regulation and ecological synergy of the aquatic nutrient balance scheme. Figure 4 The trend fully demonstrates the significant effectiveness of the optimized water nutrient balance scheme in risk management and water quality maintenance, and provides practical evidence for the refined nutrient management of aquaculture.

[0061] Please see Figure 5 The following describes an early warning system for pathogenic microorganisms in aquaculture, as described in an embodiment of this application. The early warning system 500 for pathogenic microorganisms in aquaculture includes: The data acquisition module 501 is used to collect aquaculture environmental data through sensors, integrate aquaculture density data, water nutrient indicators, waste accumulation levels and initial values ​​of pathogenic microorganism reproduction, and obtain a preliminary dataset of environmental changes after preprocessing and updating. The region determination module 502 is used to classify environmental patterns, determine high-risk areas, identify nutrient abnormality areas, and comprehensively determine potentially accelerated areas of pathogenic microorganism reproduction based on the preliminary dataset of environmental changes and the support vector machine algorithm. Risk assessment module 503 is used to compare historical and current reproduction data through the potential accelerated reproduction area of ​​the pathogenic microorganism, predict the reproduction rate trend using the random forest algorithm, calculate the degree of proximity of the critical concentration, and obtain the risk level after calibration. The scheme generation module 504 is used to determine whether the assessment value exceeds the preset median threshold based on the risk level, trigger environmental change simulation, analyze the impact of nutrient distribution, adjust parameters to generate a water body nutrient balance scheme, and monitor and verify the effect. The model building module 505 is used to track changes in the reproduction rate of pathogenic microorganisms according to the water nutrient balance scheme, generate alarm signals, build a preliminary early warning model by combining historical data, and adjust the parameter weights to obtain a disease outbreak early warning model. The threshold determination module 506 is used to integrate risk warning and economic benefit indicators according to the disease outbreak early warning model, update aquaculture environment data, activate prevention and control protocols, determine pathogen control thresholds, and optimize the triggering conditions of prevention and control protocols. The density adjustment module 507 is used to generate and distribute a breeding density adjustment report based on the pathogen control threshold, adjust the density through an automation module to match dynamic environmental requirements, and monitor and analyze data to obtain a sustainable risk reduction status.

[0062] Through the collaborative efforts of the aforementioned components, this system constructs a comprehensive intelligent prevention and control system for aquaculture pathogens, encompassing "monitoring-identification-early warning-intervention-optimization." It achieves end-to-end automated management, from dynamic data collection of the aquaculture environment to long-term control of pathogen risks. The data acquisition module 501 collects data in real time from multiple types of sensors, including aquaculture density, water nutrient indicators, waste accumulation levels, and initial values ​​of pathogenic microorganism reproduction. Using data fusion and preprocessing technologies, it transforms the scattered raw data into a structured, highly reliable preliminary dataset of environmental changes. The region determination module 502, based on this preliminary dataset, uses a support vector machine algorithm to mine time-series features, accurately classifying water nutrient distribution patterns and waste accumulation rate levels. Combined with preset thresholds, it marks high-risk areas and identifies areas of nutrient excess / deficiency, ultimately comprehensively pinpointing areas with potential accelerated pathogenic microorganism reproduction. The risk assessment module (503) retrieves historical reproduction data of pathogenic microorganisms in potential acceleration areas, compares it with current real-time data to extract trend characteristics, uses a random forest algorithm to predict future reproduction rate trends, and then calibrates the risk level by calculating the proximity of concentration to critical values ​​and combining historical disease data, transforming the abstract "risk perception" into a quantifiable "risk level assessment value." The solution generation module (504) triggers environmental change simulations based on the risk level assessment value, analyzes the impact of water nutrient distribution on pathogenic microorganism reproduction, and generates a solution that balances "inhibition of microorganisms" with other measures. The system includes a water nutrient balance scheme for "biological reproduction" and "ensuring the growth of aquatic organisms"; the model building module 505 tracks changes in the reproduction rate of pathogenic microorganisms based on the water nutrient balance scheme, automatically generating an alarm signal when the trend approaches the critical concentration, and then constructing a preliminary early warning model by combining historical disease outbreak data. The model parameter weights are optimized by analyzing the correlation between key time points and environmental parameters, ultimately forming a dynamically adaptable disease outbreak early warning model; the threshold determination module 506 integrates the risk signals from the early warning model with aquaculture economic benefit indicators (such as yield per unit water volume and disease loss rate), and updates environmental data in real time through feedback loops. According to the protocol, if the alarm persists, the tiered prevention and control protocol is activated. The protocol dynamically determines the pathogen control threshold based on real-time environmental parameters and records the correlation between the threshold and environmental parameters to optimize the protocol triggering conditions. The density adjustment module 507 generates a report containing density adjustment suggestions and environmental adaptation requirements based on the control threshold. The automated execution module precisely adjusts the aquaculture density while continuously monitoring environmental parameters and verifying the compatibility between the density and the environment until the pathogen risk is stably reduced to a safe range. This achieves a closed loop from "short-term risk mitigation" to "long-term sustainable management," ensuring the stability of the aquaculture environment and the continuous improvement of aquaculture benefits.

[0063] This application also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the aquaculture pathogen early warning method.

[0064] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the methods and systems described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0065] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions 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 this application.

Claims

1. An aquaculture pathogenic microorganism early warning method, characterized by, The method comprises the following steps: Step S101, collecting aquaculture environment data through a sensor to obtain an initial data set of environmental changes; Step S102, classifying environmental patterns and judging risk intervals according to the initial data set of environmental changes, and determining potential acceleration areas of pathogenic microorganism reproduction; Step S103, predicting the reproduction rate trend of pathogenic microorganisms and evaluating the risk level through the potential acceleration areas of pathogenic microorganism reproduction; Step S104, triggering environmental change simulation according to the risk level to obtain a water nutrition balance scheme; Step S105, tracking the change of pathogenic microorganism reproduction rate and generating an alarm signal through the water nutrition balance scheme, and constructing a disease outbreak early warning model; Step S106, updating the aquaculture environment data according to the disease outbreak early warning model and activating the prevention and control protocol to determine the pathogenic microorganism control threshold; Step S107, adjusting the breeding density through the pathogenic microorganism control threshold to obtain a sustainable risk reduction state of the aquaculture environment.

2. The method of claim 1, wherein, The step S101 comprises: Collecting the breeding density data, water nutrition indicators, waste accumulation level and initial value of pathogenic microorganism reproduction in the aquaculture environment through a sensor in real time; Integrating the breeding density data, water nutrition indicators, waste accumulation level and initial value of pathogenic microorganism reproduction by using a data fusion method to form comprehensive environmental information; According to the comprehensive environmental information, a preliminary data set reflecting the dynamic changes of the aquaculture environment is constructed; The preliminary data set is preprocessed to eliminate outliers and standardized to obtain an environmental change preliminary data set suitable for subsequent analysis.

3. The method of claim 2, wherein, The step S102 comprises: Recording time sequence characteristics for the initial data set of environmental changes; Processing the initial data set of environmental changes and the time sequence characteristics by using a support vector machine algorithm to obtain water nutrition distribution patterns and waste accumulation speed; Judging whether the waste accumulation speed in a certain area exceeds a preset accumulation threshold, if yes, marking the area as a high-risk interval; For the high-risk interval, extracting related environmental characteristics from the initial data set of environmental changes to determine the key factors affecting waste accumulation; By analyzing the water nutrition distribution patterns, the areas of nutrient excess or deficiency in the aquaculture water body are identified; According to the high-risk interval and the areas of nutrient excess or deficiency, the potential acceleration areas of pathogenic microorganism reproduction are comprehensively determined.

4. The method of claim 1, wherein, The step S103 comprises: Obtaining the historical reproduction rate records of pathogenic microorganisms corresponding to the potential acceleration areas of pathogenic microorganism reproduction; Comparing the historical reproduction rate records of pathogenic microorganisms with the current pathogenic microorganism reproduction data collected by the sensor to extract the change trend characteristics of pathogenic microorganism reproduction rate; Analyzing the change trend characteristics by using a random forest algorithm to predict the future trend of pathogenic microorganism reproduction rate in the potential acceleration areas of pathogenic microorganism reproduction; According to the future trend of pathogenic microorganism reproduction rate, the closeness of pathogenic microorganism concentration to a preset critical concentration is calculated; Based on the proximity, a preliminary risk level assessment value is determined, the preliminary risk level assessment value is calibrated in combination with historical risk assessment data and actual disease occurrence data, and a risk level is assessed.

5. The method of claim 1, wherein, The step S104 comprises: An assessment value corresponding to the risk level is obtained, and it is determined whether the assessment value exceeds a preset median threshold value, and if so, an environmental change simulation of the aquaculture environment is triggered; Through the environmental change simulation, the influence degree of the nutrient distribution state in the aquaculture water body on the reproduction of pathogenic microorganisms is analyzed; According to the influence degree analysis result, the water body nutrient input parameter is adjusted, and an optimized water body nutrient balance scheme is generated.

6. The method of claim 1, wherein, The step S105 comprises: According to the water body nutrient balance scheme, the influence of the reduction of waste accumulation in the aquaculture water body on the reproduction of pathogenic microorganisms is simulated; A time series analysis method is used to track the change trend of the reproduction rate of pathogenic microorganisms; It is determined whether the change trend of the reproduction rate of pathogenic microorganisms at a certain moment approaches a preset critical concentration of pathogenic microorganisms, and if so, an alarm signal is generated; The alarm signal is taken as a core input, and a preliminary warning model is constructed in combination with historical disease outbreak data in the aquaculture scene; For the preliminary warning model, a key time point at which the reproduction rate of pathogenic microorganisms changes significantly is recorded; The relevance of the key time point and the aquaculture environmental parameters is analyzed, the parameter weight of the preliminary warning model is adjusted according to the relevance, and finally a disease outbreak warning model is obtained.

7. The method of claim 1, wherein, The step S106 comprises: Through the disease outbreak warning model, a pathogenic microorganism risk warning mechanism and an economic benefit index of aquaculture are integrated to form a comprehensive evaluation basis; A feedback cycle processing method is used to input the preliminary environmental change data set collected in real time into the disease outbreak warning model, and the aquaculture environmental data is updated in real time; If the alarm signal triggered by the disease outbreak warning model persists, a preset pathogenic microorganism control protocol for aquaculture is activated; According to the control protocol, in combination with the updated aquaculture environmental data, a pathogenic microorganism control threshold is determined; The correlation between the pathogenic microorganism control threshold and the corresponding aquaculture environmental parameters is recorded; The correlation is analyzed, and the triggering conditions of the subsequent control protocol are optimized according to the analysis result, so as to ensure that the control protocol is adapted to the dynamic changes of the aquaculture environment.

8. The method of claim 1, wherein, The step S107 comprises: According to the pathogenic microorganism control threshold, a targeted output report containing breeding density adjustment suggestions and corresponding environmental adaptation requirements is generated; The targeted output report is distributed to the central control system of the aquaculture farm; An automatic execution module is used to adjust the breeding density of aquaculture according to the breeding density adjustment suggestions; The adjusted breeding density is matched with the dynamic adaptation requirements of the aquaculture environment; Based on the matching result, the parameter changes of the aquaculture environment are continuously collected through sensors to form an environmental parameter monitoring data set; The environmental parameter monitoring data set is analyzed to determine whether the pathogenic microorganism reproduction risk in the aquaculture environment is reduced to a preset safe range and can be stably maintained. If yes, a sustainable risk reduction state of the aquaculture environment is obtained.

9. A system for the early warning of pathogenic microorganisms in aquaculture for implementing the method for the early warning of pathogenic microorganisms in aquaculture according to any one of claims 1 to 8, characterized in that, The aquaculture pathogenic microorganism early warning system comprises: A data acquisition module is configured to collect aquaculture environment data through a sensor, integrate breeding density data, water nutrition indicators, waste accumulation levels, and initial values of pathogenic microorganism reproduction, and obtain preliminary environmental change data sets after preprocessing and updating. A region determination module is configured to classify environmental patterns using a support vector machine algorithm based on the preliminary environmental change data sets, determine a high-risk interval, identify a nutrition abnormal region, and comprehensively determine a potential acceleration region of pathogenic microorganism reproduction. A risk assessment module is configured to compare historical and current reproduction data through the potential acceleration region of pathogenic microorganism reproduction, predict reproduction rate trends using a random forest algorithm, calculate the closeness of critical concentration, and obtain a risk level after calibration. A scheme generation module is configured to determine whether the evaluation value exceeds a preset median threshold value based on the risk level, trigger environmental change simulation, analyze the influence of nutrition distribution, adjust parameters to generate a water nutrition balance scheme, and monitor and verify the effect. A model construction module is configured to track changes in pathogenic microorganism reproduction rate based on the water nutrition balance scheme, generate an alarm signal, construct a preliminary early warning model based on historical data, and adjust parameter weights to obtain a disease outbreak early warning model. A threshold determination module is configured to integrate risk early warning and economic benefit indicators based on the disease outbreak early warning model, update breeding environment data, activate a prevention and control protocol, determine a pathogenic microorganism control threshold, and optimize the prevention and control protocol trigger condition. A density adjustment module is configured to generate a breeding density adjustment report and distribute it based on the pathogenic microorganism control threshold, adjust the density through an automatic module, match environmental dynamic requirements, and monitor and analyze data to obtain a sustainable risk reduction state.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, causes the processor to perform the aquaculture pathogenic microorganism early warning method of any one of claims 1 to 8.

Images