An artificial intelligence-based aquaculture monitoring system

Abstract

The application relates to the technical field of automation monitoring, in particular to an aquatic breeding monitoring system based on artificial intelligence, which comprises a water quality dynamic analysis module, an environmental fluctuation tracking module, a biological behavior analysis module, an optimization control decision module and a risk prediction management module.In the application, through real-time monitoring of water quality, external environment and biological behavior feedback, dynamic analysis is realized, the influence of water quality abnormal fluctuation and environmental change on biological behavior is accurately predicted, the fluctuation range of environmental factors is tracked and the change of external interference factors is identified, breeding control is optimized and targeted early warning and risk management are provided, multi-dimensional data integration and intelligent analysis improve the stability of aquatic breeding, reduce the influence of environmental fluctuation on breeding state, guarantee biological health and production safety, promote efficient resource management and controllability of breeding risk, avoid human errors and response delay, and improve the efficiency and safety of the breeding process.

Description

Technical Field

[0001] This invention relates to the field of automated monitoring technology, and in particular to an artificial intelligence-based aquaculture monitoring system. Background Technology

[0002] The field of automated monitoring technology mainly involves the monitoring and data acquisition of various environments and equipment through automated devices and systems. This ensures the real-time acquisition of key parameters and effective analysis and response in various working scenarios. Core aspects of this field include sensor technology, data acquisition and transmission technology, real-time monitoring systems, fault detection and alarms, and automated control technology. Automated monitoring is increasingly widely used across various industries, especially in aquaculture, environmental monitoring, and intelligent manufacturing, where it plays a role in improving efficiency, ensuring production safety, and optimizing resource management. Traditional aquaculture monitoring systems rely on manual labor or simple equipment to monitor aquaculture conditions such as water quality, temperature, and dissolved oxygen. These systems depend on periodic checks and manual recording, resulting in low efficiency and the presence of human error. To address these issues, aquaculture monitoring systems employ multiple sensors for real-time data acquisition and transmit the data to the monitoring center via wireless networks. Traditional monitoring systems typically use a single sensor, leading to long data acquisition cycles and poor real-time performance.

[0003] Existing technologies rely on manual labor or simple equipment to monitor the aquaculture environment, which suffers from problems such as low inspection frequency, long data collection cycles, and poor real-time performance. This makes it impossible to detect water quality anomalies or environmental fluctuations in a timely manner, thus affecting aquaculture efficiency. Since data collection mainly relies on timed inspections and manual recording, there are significant errors and delays, failing to effectively capture the immediate impact of environmental changes on biological behavior. This results in a weak ability to respond to emergencies during the aquaculture process. Traditional systems use a single sensor, which limits the ability to monitor complex environments and makes it difficult to achieve comprehensive and accurate analysis of aquaculture status. This increases the uncertainty of risk management and fails to fully consider the potential impact of external interference factors on aquaculture stability. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides an artificial intelligence-based aquaculture monitoring system. The technical solution is as follows: On the one hand, an artificial intelligence-based aquaculture monitoring system is provided, which includes: The water quality dynamic analysis module extracts water quality change trends, environmental fluctuation ranges, and biological behavior patterns based on water quality parameters, external environmental indicators, and biological behavior feedback in aquaculture waters. It stores and analyzes the expansion trend of abnormal water quality fluctuations in time series, judges the cumulative impact of environmental fluctuations, identifies the stability distribution of biological behavior, and obtains an aquaculture state stability dataset. Based on the aquaculture status stability dataset, the environmental fluctuation tracking module analyzes the changing trend of environmental factor fluctuation range, judges the cumulative changes of abnormal biological behavior, counts the fluctuation frequency of short-term environmental compensation capacity, identifies the range of change of external disturbance factors, and obtains a list of aquaculture risk status. Based on the list of aquaculture risk conditions, the biological behavior analysis module analyzes the changing trends of biological adaptability, identifies the rate of behavioral recovery, judges adaptability potential, and generates a list of biological health status. Based on the biological health status list, the optimized control decision module extracts initial environmental records, external interference factors, and control stability indicators, identifies the range of state stability changes, analyzes the degree of external interference, and obtains an intelligent control scheme for aquaculture monitoring.

[0005] As a further aspect of the present invention, the aquaculture state stability dataset includes water quality fluctuation expansion rate, environmental factor distribution, biological behavior stability, and consistency of external disturbances; the aquaculture risk status list includes environmental fluctuation amplitude, cumulative impact of abnormal behavior, frequency of changes in compensatory capacity, and range of external disturbance fluctuations; the biological health status list includes biological adaptability, behavioral recovery rate, and adaptability potential; and the intelligent control scheme for aquaculture monitoring includes state stability range, degree of external disturbance, and risk of control implementation.

[0006] As a further aspect of the present invention, the water quality dynamic analysis module includes: The water quality anomaly analysis submodule extracts the distribution and expansion trend of water quality anomaly areas based on water quality parameters, external environmental indicators and biological behavior feedback in aquaculture waters, analyzes the changes in anomaly areas and collects them to the corresponding time points, arranges the anomaly expansion values ​​in chronological order, judges the fluctuation of anomaly expansion, and obtains the water quality fluctuation expansion rate. The environmental factor distribution submodule calls the water quality fluctuation expansion rate to filter time points, extracts environmental factor fluctuation data, counts fluctuation categories and proportions, analyzes changes in the proportion of fluctuation categories, and generates values ​​for changes in the proportion of environmental factor fluctuation categories. The biological behavior monitoring submodule calls the percentage change value of the environmental factor fluctuation category, compares it with the biological behavior standard, and counts the percentage of time points when the behavior is abnormal and does not meet the standard, thus obtaining the breeding status stability dataset.

[0007] As a further aspect of the present invention, the environmental fluctuation tracking module includes: The environmental fluctuation analysis submodule extracts the fluctuation range of environmental factors based on the aquaculture state stability dataset, analyzes the fluctuation amplitude of each factor within a continuous time window, filters factors whose fluctuations exceed the environmental stability threshold, judges the trend of environmental fluctuation changes, and obtains the environmental fluctuation coefficient. Based on the environmental fluctuation coefficient, the behavioral anomaly monitoring submodule extracts records of abnormal biological behavior, identifies the number and proportion of abnormal records, judges the trend of abnormal changes, and obtains the cumulative rate of change of behavioral anomalies. The external disturbance assessment submodule extracts external disturbance factor data based on the cumulative change rate of the abnormal behavior, statistically analyzes the numerical range and variation range of the differential factors, and obtains a list of aquaculture risk status.

[0008] As a further aspect of the present invention, the biological behavior analysis module includes: The biological adaptability assessment submodule extracts biological adaptability indicators and short-term adaptability based on the aquaculture risk status list, identifies the proportion of short-term adaptability to total adaptability, filters biological numbers whose short-term adaptability exceeds the benchmark value, collects the changes in biological adaptability, and obtains the biological adaptability volatility. The behavior recovery rate analysis submodule extracts behavior recovery data based on the bio-adaptive volatility, identifies the behavior recovery cycle and average recovery days, analyzes changes in the behavior recovery rate, and generates the behavior recovery variability rate. The adaptation potential calculation submodule evaluates the adaptation data based on the behavior recovery change rate, identifies the adaptation interval and adaptation ratio, analyzes the adaptation trend, and obtains a list of biological health status.

[0009] As a further aspect of the present invention, the collection of biological adaptability changes refers to the collection of biological adaptability fluctuations by statistically analyzing the number of fluctuations and the amplitude of fluctuations in the adaptability of biological numbers within a target time window. The behavioral recovery data refers to data obtained from physiological indicator data and behavioral performance data of biological IDs.

[0010] As a further aspect of the present invention, the optimization and control decision module includes: The biological health status list assessment submodule extracts biological adaptation data, behavioral recovery data and adaptation data based on the biological health status list, calculates the adaptation ability index, recovery rate and adaptation potential, and obtains the biological health status list index. The state stability analysis submodule extracts the initial environmental records based on the biological health status inventory index, determines whether the state deviation exceeds the state deviation threshold, analyzes the state change trend, and obtains the state stability change range. The risk assessment submodule for regulation implementation captures external disturbance fluctuation data and regulation impact data based on the state stability change range, calculates the external disturbance fluctuation rate, analyzes the stability of regulation implementation, and screens aquaculture units whose regulation risk exceeds the benchmark threshold, thereby obtaining an intelligent regulation scheme for aquaculture monitoring.

[0011] As a further aspect of the present invention, the state deviation threshold is the upper limit of the root mean square deviation value of the degree of exponential deviation in a preset list for judging the health status of organisms based on the baseline established by the initial environmental records. The state change trend is predicted by using a moving average algorithm to forecast the future trend of the biological health status inventory index based on the changes in the biological health status inventory index. The benchmark threshold refers to the original control effect data and loss data of the breeding unit.

[0012] As a further aspect of the present invention, the system also includes a risk prediction and management module: Based on the aforementioned intelligent control scheme for aquaculture monitoring, the risk prediction and management module extracts the current dynamic state of aquaculture, monitors the expansion of water quality fluctuations, fluctuations of environmental factors, and changes in biological behavior, analyzes the range of dynamic fluctuations, identifies the magnitude of changes in the stability of aquaculture status, judges abnormal changes in control, and obtains the intelligent early warning level for aquaculture risks. The intelligent early warning levels for aquaculture risks include the dynamic fluctuation range of the status, the magnitude of changes in status stability, and abnormal changes in regulation.

[0013] As a further aspect of the present invention, the risk prediction and management module includes: The dynamic state monitoring submodule, based on the aforementioned intelligent control scheme for aquaculture monitoring, monitors the expansion of water quality fluctuations, fluctuations of environmental factors, and changes in biological behavior, identifies fluctuation delays, environmental anomalies, and behavioral changes, and obtains the dynamic state fluctuation range. Based on the dynamic fluctuation range of the state, the state fluctuation analysis submodule analyzes the trend of changes in the breeding state, screens breeding units with fluctuation expansion, environmental abnormalities and behavioral decline, judges the degree of change in state stability, identifies the state persistence and abnormal concentration intervals, and obtains the amplitude of changes in the stability of the breeding state. The regulation anomaly judgment submodule detects the frequency of regulation changes, implementation deviations, and aquaculture records based on the stability variation range of the aquaculture status, and screens aquaculture units with unstable regulation implementation to obtain the intelligent early warning level of aquaculture risk.

[0014] The beneficial effects of the technical solutions provided by the embodiments of the present invention include at least the following: By monitoring water quality parameters, external environmental factors, and biological behavior feedback in real time, dynamic analysis of the aquaculture environment is achieved. It can accurately predict the impact of abnormal water quality fluctuations and environmental changes on biological behavior, track the fluctuation range of environmental factors, and identify changes in external disturbance factors, thereby optimizing aquaculture regulation and providing targeted early warning and risk management. Through the integration and intelligent analysis of multi-dimensional data, the stability of aquaculture is effectively improved, the impact of environmental fluctuations on aquaculture status is reduced, and biological health and production safety are ensured. It provides an intelligent regulation mechanism based on real-time data and trend analysis, which can automatically identify and respond to various environmental and biological anomalies, promote efficient resource management and controllable aquaculture risks, avoid errors and delayed responses caused by human operation, and improve the efficiency and safety of the aquaculture process. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of an artificial intelligence-based aquaculture monitoring system provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the system framework of the present invention; Figure 3 This is a flowchart of the water quality dynamic analysis module in this invention; Figure 4 This is a flowchart of the environmental fluctuation tracking module in this invention; Figure 5 This is a flowchart of the biological behavior analysis module in this invention; Figure 6 This is a flowchart of the optimized control decision-making module in this invention; Figure 7 This is a flowchart of the risk prediction and management module in this invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] This invention provides an artificial intelligence-based aquaculture monitoring system, such as... Figure 1-2 The diagram shown illustrates an AI-based aquaculture monitoring system, which includes: The water quality dynamic analysis module extracts water quality change trends, environmental fluctuation ranges, and biological behavior patterns based on water quality parameters, external environmental indicators, and biological behavior feedback in aquaculture waters. It stores and analyzes the expansion trend of abnormal water quality fluctuations in time series, judges the cumulative impact of environmental fluctuations, identifies the stability distribution of biological behavior, and obtains an aquaculture state stability dataset. The environmental fluctuation tracking module analyzes the changing trends of environmental factor fluctuation range based on the aquaculture status stability dataset, judges the cumulative changes of abnormal biological behavior, counts the fluctuation frequency of short-term environmental compensation capacity, identifies the range of change of external disturbance factors, and obtains a list of aquaculture risk status. The biological behavior analysis module analyzes the changing trends of biological adaptability based on the list of aquaculture risk status, identifies the rate of behavioral recovery, judges adaptability potential, and generates a list of biological health status. The optimized control decision module is based on the biological health status list, extracts initial environmental records, external disturbance factors, and control stability indicators, identifies the range of state stability changes, analyzes the degree of external disturbance, and obtains an intelligent control scheme for aquaculture monitoring. The risk prediction and management module is based on the intelligent control scheme for aquaculture monitoring. It extracts the current dynamic status of aquaculture, monitors the expansion of water quality fluctuations, environmental factor fluctuations, and changes in biological behavior, analyzes the range of dynamic fluctuations, identifies the magnitude of changes in the stability of aquaculture status, judges abnormal changes in control, and obtains the intelligent early warning level of aquaculture risk.

[0023] The aquaculture stability dataset includes water quality fluctuation spread rate, environmental factor distribution, biological behavior stability, and consistency of external disturbances. The aquaculture risk status list includes environmental fluctuation amplitude, cumulative impact of abnormal behavior, frequency of changes in compensatory capacity, and range of external disturbance fluctuations. The biological health status list includes biological adaptability, behavioral recovery rate, and adaptability potential. The intelligent control scheme for aquaculture monitoring includes state stability range, degree of external disturbance, and risk of control implementation. The intelligent early warning level for aquaculture risks includes dynamic fluctuation range of state, amplitude of state stability change, and abnormal changes in control.

[0024] Specifically, such as Figure 2 , 3 As shown, the water quality dynamic analysis module includes: The water quality anomaly analysis submodule extracts the distribution and expansion trend of water quality anomaly areas based on water quality parameters, external environmental indicators and biological behavior feedback in aquaculture waters, analyzes the changes in anomaly areas and collects them to the corresponding time points, arranges the anomaly expansion values ​​in chronological order, judges the fluctuation of anomaly expansion, and obtains the water quality fluctuation expansion rate. It receives water quality parameters, external environmental indicators, and biological behavior feedback, such as dissolved oxygen 0.5 mg / L, pH 8.9, ammonia nitrogen 0.6 mg / L, nitrite 0.15 mg / L, and water temperature 31.2℃, and compares them with preset normal ranges (dissolved oxygen 5.0-8.0 mg / L, pH 7.5-8.5, ammonia nitrogen <0.2 mg / L, nitrite <0.05 mg / L, and water temperature 25-30℃). Parameters outside the range are marked as abnormal points, and their geographical locations are identified to form the distribution of water quality abnormal areas. For example, if sensors A, B, and C detect dissolved oxygen below 4.0 mg / L at 10:00, their area is identified as abnormal. The area is updated every 15 minutes, calculating the increase or decrease in the area or the number of affected sensors. For example, if the area expands from 3 to 4, the expansion amount is 1 sensor. The expansion amount is arranged in chronological order as an abnormal expansion value sequence, such as [0 (10:00), 1 (10:15), 0 (10:30), 2 (10:45)]. By comparing the abnormal expansion values ​​at adjacent time points, it is determined whether the expansion trend is increasing, decreasing, or stable. For example, a change from 0 to 2 indicates accelerated expansion. The water quality fluctuation expansion rate is obtained, which is quantified as the average percentage change in the area of ​​the abnormal region or the number of affected sensors per unit time. For example, if the area increases by 20% in 1 hour, the water quality fluctuation expansion rate is 20% / hour.

[0025] The environmental factor distribution submodule calls the water quality fluctuation expansion rate to filter time points, extracts environmental factor fluctuation data, statistically analyzes the fluctuation categories and proportions, analyzes the changes in the proportions of fluctuation categories, and generates the change values ​​of the proportions of environmental factor fluctuation categories. The system receives water quality fluctuation expansion rates, for example, 15% at 2:00 PM, 20% at 3:00 PM, and 8% at 4:00 PM on a given day. A time point filtering threshold of 10% is set, filtering time points where the water quality fluctuation expansion rate exceeds 10%, such as 2:00 PM and 3:00 PM. From the filtered time points, corresponding environmental factor fluctuation data are extracted, including water temperature changes of 2℃ / hour, light intensity changes of 3000 lux / hour, and rainfall changes of 5 mm / hour. Environmental factor fluctuation data are categorized: water temperature fluctuations exceeding 1℃ / hour are classified as "temperature fluctuations," light intensity fluctuations exceeding 2000 lux / hour are classified as "light intensity fluctuations," and rainfall fluctuations exceeding 3 mm / hour are classified as "rainfall fluctuations." The frequency and percentage of each type of fluctuation are also statistically analyzed. For example, if there are 10 environmental fluctuation events within one hour, temperature fluctuations occur 5 times (50%), light intensity fluctuations occur 3 times (30%), and rainfall fluctuations occur 2 times (20%). The current percentage of fluctuation categories is compared with the average percentage of the previous period (e.g., the previous day), and the change in the percentage of each category is calculated. For example, the previous day, the percentage of temperature fluctuation was 40%, the percentage of sunshine fluctuation was 40%, and the percentage of rainfall fluctuation was 20%. Currently, the percentage of temperature fluctuation has increased by 10%, the percentage of sunshine fluctuation has decreased by 10%, and the percentage of rainfall fluctuation remains unchanged. The changes are integrated into the change value of the percentage of environmental factor fluctuation categories.

[0026] The biological behavior monitoring submodule calls the change value of the proportion of environmental factor fluctuation categories, compares it with the biological behavior standard, and counts the proportion of time points when the behavior is abnormal and does not meet the standard, thus obtaining the breeding status stability dataset. The system receives changes in the percentage of environmental factor fluctuations, such as a 10% increase in the percentage of temperature fluctuations and a 10% decrease in the percentage of light fluctuations. Preset biological behavior standards are established, such as a feeding frequency of 2-3 times per day for 15-20 minutes, a cluster density change rate of less than 5%, and a swimming speed of 0.5-1.0 m / s. When environmental factors fluctuate, biological behavior data, including food intake, activity level, and cluster distribution, are monitored and compared with the preset biological behavior standards. For example, if the percentage of temperature fluctuations increases, the feeding frequency decreases to once per day for 5 minutes, the cluster density change rate reaches 15%, and the swimming speed decreases to 0.3 m / s, the behavior is marked as abnormal and not meeting the standards. The system also counts the number of monitoring points or the duration of abnormal and non-compliant biological behavior within a specific time period, such as a day, and calculates its percentage in the total monitoring time points. For example, if 24 time points are monitored throughout the day, and biological behavior is non-compliant at 8 time points, the percentage is 33.3%. This percentage reflects the degree of discomfort of organisms under the current environmental fluctuations. The dataset of breeding status stability is obtained, which includes the percentage of time points where behavior is abnormal and the corresponding environmental fluctuation information. For example, when the percentage of temperature fluctuations increases by 10%, the percentage of time points where behavior is abnormal and the standards are not met is 33.3%.

[0027] Specifically, such as Figure 2 , 4As shown, the environmental fluctuation tracking module includes: The environmental fluctuation analysis submodule extracts the fluctuation range of environmental factors based on the aquaculture status stability dataset, analyzes the fluctuation amplitude of each factor within a continuous time window, filters factors whose fluctuations exceed the environmental stability threshold, judges the trend of environmental fluctuation changes, and obtains the environmental fluctuation coefficient. The system receives a dataset on the stability of aquaculture conditions, which includes the percentage of time points where aquatic behavior was abnormal and did not meet standards (e.g., 33.3% on a certain day). It also records the fluctuation data of environmental factors within the corresponding time period. The system extracts the fluctuation range of environmental factors corresponding to the abnormal behavior. For example, during periods of abnormal behavior, water temperature fluctuates from 26℃ to 30℃, pH from 7.8 to 8.3, and dissolved oxygen from 5.5 mg / L to 3.5 mg / L. A continuous time window of 6 hours is set, and the fluctuation amplitude of each environmental factor within this window is analyzed. The fluctuation amplitude is the difference between the maximum and minimum values. For example, water temperature fluctuation amplitude is 4℃, pH fluctuation amplitude is 0.5, and dissolved oxygen fluctuation amplitude is 2.0 mg / L. Environmental stability thresholds are set: water temperature fluctuation threshold is 2℃, pH fluctuation threshold is 0.3, and dissolved oxygen fluctuation threshold is 1.5 mg / L. Factors are filtered when the calculated fluctuation amplitude exceeds the corresponding threshold. For example, a water temperature fluctuation of 4℃ exceeds the 2℃ threshold, and a dissolved oxygen fluctuation of 2.0 mg / L exceeds the 1.5 mg / L threshold. Trend analysis is performed on the selected environmental factors, such as by using linear regression, to determine whether the fluctuations are continuously increasing, decreasing, or stable. For example, if the water temperature fluctuation amplitude shows an upward trend, it is judged as an increasing fluctuation trend. The fluctuation coefficient is calculated as the weighted average of the fluctuation amplitudes of all environmental factors that exceed the threshold. The weights are set based on the degree of influence of the factors on the abnormal biological behavior.

[0028] The behavioral anomaly monitoring submodule extracts records of abnormal biological behavior based on the environmental fluctuation coefficient, identifies the number and proportion of abnormal records, judges the trend of abnormal changes, and obtains the cumulative rate of change of behavioral anomalies. The system receives the environmental fluctuation coefficient, for example, calculated to be 0.75. It then accesses the historical biological behavior database of the breeding unit to extract records of biological behaviors marked as abnormal. These records include the behavior type, occurrence time, duration, and number of individuals involved. For example, if there are 12 abnormal records in the past 24 hours, including 5 instances of decreased feed intake, 4 instances of reduced activity levels, and 3 instances of cluster dispersal, the system calculates the total number of abnormal records and compares it with the total number of behavior monitoring records to determine the percentage of abnormal records. For example, if there are 1000 records in total and 12 are abnormal, the percentage is 1.2%. The system compares the number and percentage of abnormal records across different time periods to determine the trend of abnormal changes. For example, if the number of abnormal records in the previous 24 hours was 8 (0.8%), and both the number and percentage are currently increasing, it indicates an increasing trend of abnormalities. The system calculates the average growth rate based on the magnitude of change in the number or percentage of abnormal records within a continuous time window. For example, if the percentages of abnormalities in the past three consecutive 24-hour periods were 0.8%, 1.2%, and 1.5%, the cumulative rate of change in behavioral abnormalities is calculated to be 0.35% / 24 hours.

[0029] The external disturbance assessment submodule extracts external disturbance factor data based on the cumulative change rate of abnormal behavior, statistically analyzes the numerical range and variation range of differential factors, and obtains a list of aquaculture risk status. The system receives the cumulative rate of change of abnormal behavior, for example, an increase of 0.35% per day. Based on this rate, it extracts external interference factor data that highly overlaps with the time period of abnormal behavior. This includes weather forecasts of heavy rainfall, typhoons, extreme high or low temperatures, pesticide spraying records in nearby farmland, construction noise data outside the aquaculture farm, and the density of passing ships. The extracted external interference factors are compared with historical data to statistically analyze the difference range between the factor values ​​and those under normal conditions, as well as the range of variation during the abnormal period. For example, during a period with a high cumulative rate of change of abnormal behavior, the aquaculture unit experienced 3 hours of continuous heavy rainfall, with a rainfall amount of 50 mm. The rainfall difference range is [mm], and the range of variation is [mm]. Simultaneously, nearby construction noise was detected to continuously exceed 70 decibels during the abnormal period. The noise difference range is [decibels], and the range of variation is [decibels]. The difference range and range of variation identify external interference factors that affect the aquaculture organisms. The analysis results are integrated to list all identified external interference factors, their difference ranges, ranges of variation, and contributions to the cumulative rate of change of abnormal behavior, thus obtaining a list of aquaculture risk status.

[0030] Specifically, such as Figure 2 , 5 As shown, the biological behavior analysis module includes: The biological adaptability assessment submodule extracts biological adaptability indicators and short-term adaptability based on the list of aquaculture risk status, identifies the proportion of short-term adaptability to total adaptability, filters biological numbers whose short-term adaptability exceeds the benchmark value, collects the changes in biological adaptability, and obtains the biological adaptability volatility. Collecting changes in biological adaptability refers to collecting data by statistically analyzing the number and magnitude of fluctuations in the adaptability of biological numbers within a target time window. The system receives a list of aquaculture risk conditions, detailing external disturbances and their impact on abnormal biological behavior. Examples include sudden salinity drops due to heavy rainfall and stress caused by construction noise. Based on the risk, corresponding biological adaptation indicators are extracted, including physiological indicators (such as blood glucose levels, cortisol concentration, and immunoglobulin levels) and behavioral data (such as time required to resume feeding and exploratory behavior in new environments). Short-term adaptation capacity is assessed, which is the ability of organisms to rapidly adjust their physiology and behavior to adapt to specific external disturbances. For example, after heavy rainfall, monitoring shows that fish cortisol concentration increases from 20 ng / mL to 50 ng / mL but recovers to 30 ng / mL within 6 hours. The proportion of short-term adaptation capacity to total adaptation is calculated by comparing short-term adaptation capacity with the species' maximum adaptation potential under ideal conditions. For example, if a certain fish species can tolerate a maximum water temperature fluctuation of 8°C, but its current short-term adaptation capacity only tolerates a 3°C fluctuation, then the short-term adaptation capacity accounts for 37.5% of the total adaptation. A baseline value of 60% was set, and biological individuals or groups with a short-term fit rate below 60% were selected, such as fish groups A001 and A003, whose short-term fit rates were 35% and 45%, respectively. The fit rate changes of the selected biological numbers were collected. During the collection process, the number of fluctuations and the amplitude of the fit rate fluctuations of the biological numbers within the target time window (e.g., the past 7 days) were statistically analyzed. For example, the cortisol concentration of fish group A001 exceeded the normal range 3 times within 7 days, with an average fluctuation amplitude of 25 ng / mL. Based on the number of fluctuations and the amplitude, the biological fit rate was calculated. More fluctuations and larger amplitudes indicate a higher biological fit rate, and the biological fit rate was obtained.

[0031] The behavior recovery rate analysis submodule extracts behavior recovery data based on bioadaptive volatility, identifies the behavior recovery cycle and average recovery days, analyzes changes in behavior recovery rate, and generates behavior recovery variability rate. Behavioral recovery data refers to data obtained from physiological indicator data and behavioral performance data of biometrics; The system receives biological fitness volatility, for example, the biological fitness volatility of a certain fish population is 0.2. Based on the volatility, behavioral recovery data is extracted from the physiological indicator data (such as heart rate, respiratory rate, blood glucose, blood oxygen saturation) and behavioral performance data (such as feeding activity, schooling behavior, swimming patterns) of the organisms. The data records the process by which various physiological and behavioral indicators of the organisms recover to normal levels after experiencing external disturbances or stress. For example, after a water quality abnormality event, it took 12 hours for the fish's heart rate to drop from 90 beats / minute to the normal level of 60 beats / minute, and 18 hours for feeding activity to recover from cessation to normal levels. The system analyzes the recovery process, identifies the behavioral recovery cycle, that is, the time required from the abnormal state to complete recovery. For example, the heart rate recovery cycle is 12 hours, and the feeding recovery cycle is 18 hours. The system also calculates the average recovery days. For example, in the past month, the fish population experienced 5 similar stress events, and its average recovery days were 1.5 days. Compare the average recovery days for different time periods or different stress events to analyze the changing trend of behavioral recovery rate. For example, if the average recovery days increase from 1.2 days last month to 1.5 days, it indicates a decrease in behavioral recovery rate. Quantify the degree of change in behavioral recovery rate. For example, if the average recovery days increase by 25%, the behavioral recovery change rate is +25%, and obtain the behavioral recovery change rate.

[0032] The adaptation potential calculation submodule evaluates adaptation data based on the behavioral recovery change rate, identifies adaptation intervals and adaptation ratios, analyzes adaptation trends, and obtains a list of biological health status. A change in the rate of change of receiving behavior, for example, +25%, indicates a decrease in the biological recovery rate. Based on this change rate, combined with biological adaptation indicators, behavioral recovery data, and environmental factor fluctuation data, the biological adaptation data is evaluated. For example, considering the average cortisol concentration of the fish population, feeding recovery time, group stability, and the range of environmental temperature fluctuations experienced, cluster analysis is used to identify the range of environmental factors that the organism can effectively adapt to, i.e., the adaptation range. For example, for temperature, the adaptation range for the fish population is 25℃-30℃, and for pH, the adaptation range is 7.5-8.5. At the same time, the adaptation ratio of the organism is calculated, which is the ratio of the actual adaptation capacity of the organism in the current environment to its theoretical maximum adaptation capacity. For example, theoretically, the maximum temperature fluctuation range that the fish population can tolerate is 10℃ (23℃-33℃), and the current actual adaptation range is 5℃ (25℃-30℃), then the adaptation ratio is 50%. By comparing the adaptation range and adaptation ratio at different time periods, and analyzing the adaptation trend, such as the shrinking adaptation range and the decline in the adaptation ratio, it is found that the biological adaptation potential is weakening. Based on the above assessment results, the adaptation range, adaptation ratio and adaptation trend of each breeding unit or individual organism are listed in detail, and a health score is given to obtain a list of biological health status.

[0033] Specifically, such as Figure 2 , 6As shown, the optimized control decision-making module includes: The biological health status assessment submodule extracts biological adaptation data, behavioral recovery data, and adaptation data based on the biological health status list, calculates the adaptation ability index, recovery rate, and adaptation potential, and obtains the biological health status list index. The system receives a list of biological health statuses, including the fit range, fit ratio, and fit trend for each breeding unit or individual organism. For example, breeding unit A has a fit range of 25℃-30℃, a fit ratio of 50%, and a declining fit trend. The system extracts biological fit data (such as stress hormone levels and immune indicators), behavioral recovery data (such as recovery time and degree), and fit data (such as historical environmental tolerance range) from the list. A preset algorithm is used to calculate the fit index. The calculation method is as follows: normalized average cortisol concentration, normalized immunoglobulin levels, and normalized feed recovery rate are multiplied by weights of 0.4, 0.3, and 0.3, respectively, and then summed. If the normalized cortisol concentration value is 0.7, the normalized immunoglobulin level value is 0.8, and the normalized feed recovery rate value is 0.9, then the fit index = 0.4 × 0.7 + 0.3 × 0.8 + 0.3 × 0.9 = 0.28 + 0.24 + 0.27 = 0.79. Simultaneously, the behavioral recovery rate is calculated, which is the reciprocal of the time required for an organism to recover from a stress state to a normal state. For example, if the average recovery time is 24 hours, the recovery rate is 1 / 24 hours, or approximately 0.0417 times / hour. Adaptability is calculated by assessing an organism's survival and reproductive capacity in extreme environments. For example, the calculation method is: normalize the maximum tolerance temperature and multiply by 0.6, plus normalize the pH tolerance range and multiply by 0.4. If the normalized value of the maximum tolerance temperature is 0.75 and the normalized value of the pH tolerance range is 0.8, then the adaptability = 0.6 × 0.75 + 0.4 × 0.8 = 0.77. The Biohealth Status Index is a weighted average of the above-mentioned Adaptability Index, Recovery Rate, and Adaptability Potential. The weights are allocated according to their importance to the overall health status. To obtain the Biohealth Status Index, for example, Biohealth Status Index = 0.05 × Adaptability Index + 0.03 × Recovery Rate + 0.02 × Adaptability Potential. Substituting these values ​​into the calculation, the result is: 0.05 × 0.79 + 0.03 × 0.0417 + 0.02 × 0.77 = 0.056151.

[0034] The state stability analysis submodule extracts initial environmental records based on the biological health status inventory index, determines whether the state deviation exceeds the state deviation threshold, analyzes the state change trend, and obtains the state stability change range. The state deviation threshold is the upper limit of the root mean square deviation value of the index deviation in the preset list of biological health status assessment based on the baseline established by the initial environmental records. The trend of status change is predicted by using a moving average algorithm to forecast the future trend of the biological health status inventory index based on the changes in the biological health status inventory index. The system receives the biohealth status inventory index, for example, an index of 0.78 for a certain aquaculture unit. It also extracts the initial environmental records for that unit, including baseline parameters such as initial water temperature of 28℃, pH of 7.8, and dissolved oxygen of 6.5 mg / L. The deviation of the current biohealth status inventory index from its historical average or the set baseline value is calculated. For example, if the historical average index is 0.85, the current deviation is 0.78 - 0.85 = -0.07. The state deviation threshold is set based on the baseline established from the initial environmental records. Statistical analysis of the biohealth status inventory index over the past year yields a mean of 0.85 and a standard deviation of 0.03. The calculated root mean square deviation upper limit threshold is 0.06. If the absolute value of the current deviation of 0.07 exceeds 0.06, the state deviation is considered to have exceeded the state deviation threshold. The trend of status change is predicted using a moving average algorithm (e.g., 5-day moving average) to forecast the future trend of the biological health status inventory index. For example, if the index for the past 5 days was 0.82, 0.80, 0.79, 0.78, and 0.77, and the moving average has been declining, the predicted future trend is downward. The range of status stability change is determined based on whether the status deviation exceeds the threshold and the status change trend, to determine whether the breeding status is stable, fluctuating, declining, or rising. For example, if the deviation exceeds the threshold and the trend is downward, the range of status stability change is marked as "severe decline".

[0035] The risk assessment submodule for regulation implementation captures external disturbance fluctuation data and regulation impact data based on the range of state stability changes, calculates the external disturbance fluctuation rate, analyzes the stability of regulation implementation, screens aquaculture units whose regulation risk exceeds the benchmark threshold, and obtains an intelligent regulation scheme for aquaculture monitoring. External disturbance volatility is expressed by the formula: ; in, Represents volatility caused by external disturbances. Representing the Interference at any time affects the data. This represents the average value of the data affected by interference. The standard deviation of the data represents the impact of interference. Representing the Real-time control data, This represents the average value of the control data. Represents the total number of time intervals; The baseline threshold refers to the original control effect data and loss data of the breeding unit.

[0036] The system receives data on the stability variation range of the receiving state. For example, if the current stability variation range is "severely decreased," it captures real-time data on external disturbance fluctuations in the aquaculture unit, such as the fluctuation range of water temperature, rainfall, and light intensity over the past 24 hours. It also records the impact data of regulatory measures, i.e., the effects of the last regulatory measures (e.g., turning on the aerator or adjusting feed) on water quality and biological behavior. For example, how much did dissolved oxygen increase after the aerator was turned on, how much did feed intake change after the feed adjustment, and the external disturbance fluctuation rate. The calculations aim to quantify the comprehensive impact of external environmental disturbances on aquaculture, among which, Representing the Interference at any given time can affect the data; for example, the average value of the change in dissolved oxygen (DO) per unit time. and standard deviation This reflects the overall level and dispersion of the data affected by interference, while Representing the Real-time control data, such as the average value of the aerator's power adjustment per unit time. Total number of time Used to assess the overall strength and sustainability of regulatory measures; The advantage of this formula lies in its comprehensive assessment of the combined volatility of external disturbances under regulatory conditions by multiplying the standardized absolute deviation of the disturbance impact data by the root mean square value of the volatility of the control data. For example, if the disturbance impact data fluctuates drastically and the control data also shows significant volatility, this formula can accurately capture this dual instability, providing a precise quantitative basis for subsequent risk assessment. The external disturbance volatility rate is calculated as follows: Assuming the following data is collected within the monitoring period N=5 time points: Disturbance impact data [1.2, 1.5, 0.8, 1.0, 1.3] (Unit: mg / L, representing negative changes in DO) Regulation data [0.6, 0.7, 0.5, 0.8, 0.6] (Unit: kW, representing aerator power); First, calculate the average value of the data affected by interference. and standard deviation : ; ; Next, the average value of the control data is calculated. : ; Then, substitute the value into the formula: ; Calculate the summation term: ; ; Substitute calculation : ; Calculated external disturbance volatility The value is approximately 0.08776. This result indicates that during the current monitoring period, the impact of external environmental disturbances on the aquaculture units is relatively small, and the fluctuation of control measures is also at a low level. The baseline threshold is set based on historical control effect data and loss data of the aquaculture units. For example, through the analysis of cases where control failures in the past two years resulted in losses exceeding 10%, the baseline threshold for external disturbance volatility was set at 0.12. The value is compared with the benchmark threshold of 0.12. For example, if 0.08776 is less than 0.12, it indicates that the stability of the control implementation is good. The stability of the control implementation is analyzed, for example, if... If the value remains below the threshold, the control measures are considered stable; if it exceeds the threshold, they are considered unstable. Aquaculture units whose control risk exceeds the benchmark threshold are then identified. For example, if another aquaculture unit's... If the value is 0.15, the unit is selected, and the risk assessment results of each aquaculture unit are integrated. Specific control strategies and early warning measures are proposed for high-risk units to obtain an intelligent control scheme for aquaculture monitoring.

[0037] Specifically, such as Figure 2 , 7 As shown, the risk prediction and management module includes: The dynamic state monitoring submodule is based on the intelligent control scheme for aquaculture monitoring. It monitors the expansion of water quality fluctuations, fluctuations of environmental factors and changes in biological behavior, identifies fluctuation delays, environmental anomalies and behavioral changes, and obtains the dynamic state fluctuation range. The system receives intelligent control solutions for aquaculture monitoring, including risk assessments and suggested control strategies for each aquaculture unit. For example, for a risk warning of aquaculture unit A, it suggests increasing the aeration frequency. Based on the solution, it continuously monitors the expansion of water quality fluctuations, environmental factor fluctuations, and changes in biological behavior. For example, it monitors water quality sensor data of aquaculture unit A in real time to track the expansion of areas with abnormal dissolved oxygen, while also monitoring changes in environmental factors such as ambient temperature and light intensity. It analyzes biological behavior data such as fish feeding behavior and school density captured by cameras, identifies fluctuation delay phenomena, such as a 30-minute delay in abnormal changes in water quality parameters (such as dissolved oxygen) when the water temperature drops. It also identifies abnormal environmental conditions, such as light intensity that is continuously too high or too low during abnormal periods, and identifies changes in biological behavior, such as fish changing from a normal foraging state to a static, bottom-gathering state within a specific time. Based on the monitoring and identification results, the dynamic fluctuation range of the status is quantified and obtained. This range represents the real-time variation range of the current aquaculture status in various key indicators. For example, the dissolved oxygen fluctuation range is between 3.0-5.0 mg / L, and the fish activity level fluctuation range is between 20%-50% of the normal level.

[0038] The state fluctuation analysis submodule analyzes the trend of changes in breeding state based on the dynamic fluctuation range of state, screens breeding units with fluctuation expansion, environmental abnormalities and behavioral decline, judges the degree of change in state stability, identifies the persistence of state and the concentrated interval of abnormality, and obtains the amplitude of change in breeding state stability. The dynamic fluctuation range of the receiving status is measured. For example, dissolved oxygen fluctuates between 3.0-5.0 mg / L, and fish activity levels fluctuate between 20%-50% of normal levels. Based on this dynamic fluctuation data, time series analysis methods, such as the ARIMA model, are used to analyze the changing trends of the aquaculture status and predict a continued decline in dissolved oxygen over the next 12 hours. By comparing with preset thresholds, aquaculture units exhibiting expanding fluctuations, environmental anomalies, and decreased biological behavior are identified. For example, in aquaculture unit B, the area of ​​abnormal water quality continues to expand; in aquaculture unit C, the ambient temperature consistently exceeds the normal range by more than 2°C; and in aquaculture unit D, the fish's feed intake decreases by more than 30% for two consecutive days. For the selected units, the degree of change in their state stability is judged. For example, if dissolved oxygen drops from 5.0 mg / L to 3.0 mg / L, a decrease of 40%, it is judged as "decline". At the same time, the persistence of the state is identified, that is, the duration of the abnormal state. For example, the low dissolved oxygen state has lasted for 8 hours. The abnormal concentration period is also identified, that is, the time period during which the abnormal situation occurs most often in a day. For example, the period from 2 am to 6 am is the period in which low dissolved oxygen occurs most frequently. Taking into account the degree of fluctuation expansion, the severity of environmental abnormality, the decline in biological behavior and the duration of abnormality, it is finally quantified into a value. For example, after comprehensive evaluation, the stability change of breeding unit B is -0.3 (indicating a 30% decrease in stability). The stability change of breeding state is obtained.

[0039] The abnormal control judgment submodule detects the frequency of control changes, implementation deviations and aquaculture records based on the stability changes of the aquaculture status, screens aquaculture units with unstable control implementation, and obtains the intelligent early warning level of aquaculture risk. Based on the stability variation range of the breeding status, for example, the stability variation range of breeding unit B is -0.3, the frequency of control changes in the breeding unit is detected. For example, in the past 24 hours, the aerator power was adjusted 5 times and the feeding amount was adjusted 2 times. At the same time, the control implementation deviation is detected, that is, the difference between the actual control effect and the expected control target. For example, it is expected that the dissolved oxygen should increase by 1.0 mg / L after aeration, but it only increases by 0.5 mg / L in reality, with a deviation of 50%. Simultaneously, aquaculture records are monitored, including historical yield data, disease occurrence, and feed conversion rate. Standards are set to determine the stability of control implementation. For example, if the frequency of control changes exceeds 3 times / 24 hours, or the control implementation deviation exceeds 20%, or historical aquaculture records show a yield decrease of more than 10% for two consecutive cycles, aquaculture units with unstable control implementation are screened based on the test results. For example, aquaculture unit B has a control change frequency of 7 times / 24 hours and an implementation deviation of 60%, both exceeding the preset standards, and is therefore screened as an unstable control implementation unit. A comprehensive assessment is conducted based on the instability of control implementation, the magnitude of changes in state stability, and the potential risk level. For example, the risk level is divided into three levels: "low," "medium," and "high," and corresponding early warning suggestions are given. For example, aquaculture unit B is rated as "high" risk level, and it is recommended to immediately conduct a manual inspection to obtain the intelligent early warning level of aquaculture risk.

[0040] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An artificial intelligence-based aquaculture monitoring system, characterized in that, The system includes: The water quality dynamic analysis module extracts water quality change trends, environmental fluctuation ranges, and biological behavior patterns based on water quality parameters, external environmental indicators, and biological behavior feedback in aquaculture waters. It stores and analyzes the expansion trend of abnormal water quality fluctuations in time series, judges the cumulative impact of environmental fluctuations, identifies the stability distribution of biological behavior, and obtains an aquaculture state stability dataset. Based on the aquaculture status stability dataset, the environmental fluctuation tracking module analyzes the changing trend of environmental factor fluctuation range, judges the cumulative changes of abnormal biological behavior, counts the fluctuation frequency of short-term environmental compensation capacity, identifies the range of change of external disturbance factors, and obtains a list of aquaculture risk status. Based on the list of aquaculture risk conditions, the biological behavior analysis module analyzes the changing trends of biological adaptability, identifies the rate of behavioral recovery, judges adaptability potential, and generates a list of biological health status. Based on the biological health status list, the optimized control decision module extracts initial environmental records, external interference factors, and control stability indicators, identifies the range of state stability changes, analyzes the degree of external interference, and obtains an intelligent control scheme for aquaculture monitoring.

2. The artificial intelligence-based aquaculture monitoring system according to claim 1, characterized in that: The aquaculture stability dataset includes water quality fluctuation spread rate, environmental factor distribution, biological behavior stability, and consistency of external disturbances. The aquaculture risk status list includes environmental fluctuation amplitude, cumulative impact of abnormal behavior, frequency of changes in compensatory capacity, and range of external disturbance fluctuations. The biological health status list includes biological adaptability, behavioral recovery rate, and adaptability potential. The aquaculture monitoring and intelligent control scheme includes state stability range, degree of external disturbance, and risk of control implementation.

3. The artificial intelligence-based aquaculture monitoring system according to claim 1, characterized in that: The water quality dynamic analysis module includes: The water quality anomaly analysis submodule extracts the distribution and expansion trend of water quality anomaly areas based on water quality parameters, external environmental indicators and biological behavior feedback in aquaculture waters, analyzes the changes in anomaly areas and collects them to the corresponding time points, arranges the anomaly expansion values ​​in chronological order, judges the fluctuation of anomaly expansion, and obtains the water quality fluctuation expansion rate. The environmental factor distribution submodule calls the water quality fluctuation expansion rate to filter time points, extracts environmental factor fluctuation data, counts fluctuation categories and proportions, analyzes changes in the proportion of fluctuation categories, and generates values ​​for changes in the proportion of environmental factor fluctuation categories. The biological behavior monitoring submodule calls the percentage change value of the environmental factor fluctuation category, compares it with the biological behavior standard, and counts the percentage of time points when the behavior is abnormal and does not meet the standard, thus obtaining the breeding status stability dataset.

4. The artificial intelligence-based aquaculture monitoring system according to claim 3, characterized in that: The environmental fluctuation tracking module includes: The environmental fluctuation analysis submodule extracts the fluctuation range of environmental factors based on the aquaculture state stability dataset, analyzes the fluctuation amplitude of each factor within a continuous time window, filters factors whose fluctuations exceed the environmental stability threshold, judges the trend of environmental fluctuation changes, and obtains the environmental fluctuation coefficient. Based on the environmental fluctuation coefficient, the behavioral anomaly monitoring submodule extracts records of abnormal biological behavior, identifies the number and proportion of abnormal records, judges the trend of abnormal changes, and obtains the cumulative rate of change of behavioral anomalies. The external disturbance assessment submodule extracts external disturbance factor data based on the cumulative change rate of the abnormal behavior, statistically analyzes the numerical range and variation range of the differential factors, and obtains a list of aquaculture risk status.

5. The artificial intelligence-based aquaculture monitoring system according to claim 4, characterized in that: The biological behavior analysis module includes: The biological adaptability assessment submodule extracts biological adaptability indicators and short-term adaptability based on the aquaculture risk status list, identifies the proportion of short-term adaptability to total adaptability, filters biological numbers whose short-term adaptability exceeds the benchmark value, collects the changes in biological adaptability, and obtains the biological adaptability volatility. The behavior recovery rate analysis submodule extracts behavior recovery data based on the bio-adaptive volatility, identifies the behavior recovery cycle and average recovery days, analyzes changes in the behavior recovery rate, and generates the behavior recovery variability rate. The adaptation potential calculation submodule evaluates the adaptation data based on the behavior recovery change rate, identifies the adaptation interval and adaptation ratio, analyzes the adaptation trend, and obtains a list of biological health status.

6. The artificial intelligence-based aquaculture monitoring system according to claim 5, characterized in that: The collection of biological adaptability changes refers to the collection of biological adaptability fluctuations by statistically analyzing the number and magnitude of fluctuations in the adaptability of biological IDs within a target time window. The behavioral recovery data refers to data obtained from physiological indicator data and behavioral performance data of biological IDs.

7. The artificial intelligence-based aquaculture monitoring system according to claim 5, characterized in that: The optimized control decision-making module includes: The biological health status list assessment submodule extracts biological adaptation data, behavioral recovery data and adaptation data based on the biological health status list, calculates the adaptation ability index, recovery rate and adaptation potential, and obtains the biological health status list index. The state stability analysis submodule extracts the initial environmental records based on the biological health status inventory index, determines whether the state deviation exceeds the state deviation threshold, analyzes the state change trend, and obtains the state stability change range. The risk assessment submodule for regulation implementation captures external disturbance fluctuation data and regulation impact data based on the state stability change range, calculates the external disturbance fluctuation rate, analyzes the stability of regulation implementation, and screens aquaculture units whose regulation risk exceeds the benchmark threshold, thereby obtaining an intelligent regulation scheme for aquaculture monitoring.

8. The artificial intelligence-based aquaculture monitoring system according to claim 7, characterized in that: The state deviation threshold is the upper limit of the root mean square deviation value of the degree of exponential deviation in a preset list for judging the health status of organisms based on the baseline established by the initial environmental records. The state change trend is predicted by using a moving average algorithm to forecast the future trend of the biological health status inventory index based on the changes in the biological health status inventory index. The benchmark threshold refers to the original control effect data and loss data of the breeding unit.

9. The artificial intelligence-based aquaculture monitoring system according to claim 1, characterized in that: The system also includes a risk prediction and management module: Based on the aforementioned intelligent control scheme for aquaculture monitoring, the risk prediction and management module extracts the current dynamic state of aquaculture, monitors the expansion of water quality fluctuations, fluctuations of environmental factors, and changes in biological behavior, analyzes the range of dynamic fluctuations, identifies the magnitude of changes in the stability of aquaculture status, judges abnormal changes in control, and obtains the intelligent early warning level for aquaculture risks. The intelligent early warning levels for aquaculture risks include the dynamic fluctuation range of the status, the magnitude of changes in status stability, and abnormal changes in regulation.

10. The artificial intelligence-based aquaculture monitoring system according to claim 9, characterized in that: The risk prediction and management module includes: The dynamic state monitoring submodule, based on the aforementioned intelligent control scheme for aquaculture monitoring, monitors the expansion of water quality fluctuations, fluctuations of environmental factors, and changes in biological behavior, identifies fluctuation delays, environmental anomalies, and behavioral changes, and obtains the dynamic state fluctuation range. Based on the dynamic fluctuation range of the state, the state fluctuation analysis submodule analyzes the trend of changes in the breeding state, screens breeding units with fluctuation expansion, environmental abnormalities and behavioral decline, judges the degree of change in state stability, identifies the state persistence and abnormal concentration intervals, and obtains the amplitude of changes in the stability of the breeding state. The regulation anomaly judgment submodule detects the frequency of regulation changes, implementation deviations, and aquaculture records based on the stability variation range of the aquaculture status, and screens aquaculture units with unstable regulation implementation to obtain the intelligent early warning level of aquaculture risk.

Images