A growth monitoring system for aquaculture organisms

Abstract

The application relates to the technical field of aquaculture, and discloses a growth monitoring system for aquatic breeding organisms, which comprises a data acquisition module, a data processing module, a sensitivity analysis module, a water quality growth modeling module and an anomaly detection module; the data acquisition module is used for collecting water quality parameters, biological growth information and external environment data in real time; the data processing module is used for cleaning, denoising, fusing and normalizing the collected data; the sensitivity analysis module is used for analyzing the influence of the water quality parameters on the biological growth and identifying the parameters; the water quality growth modeling module is used for establishing a nonlinear relationship model of the water quality parameters and the biological growth; and the anomaly detection module is used for detecting abnormal patterns of the water quality or the biological growth and triggering an alarm. In the application, the water quality parameters and the biological growth information are collected in real time, and data processing and analysis technologies are combined, so that the breeding environment and feeding strategies can be dynamically optimized, and therefore the biological growth speed, the feed utilization rate and the overall breeding efficiency can be significantly improved.

Description

Technical Field

[0001] This invention relates to the field of aquaculture technology, and in particular to a growth monitoring system for aquaculture organisms. Background Technology

[0002] In modern aquaculture, with the continuous expansion of the industry scale and the gradual improvement of technology, how to achieve precise monitoring of biological growth and intelligent management of the aquaculture environment has become a key research direction for the industry. The growth rate and health status of aquaculture organisms are closely related to the water quality environment (such as dissolved oxygen, ammonia nitrogen, temperature, salinity, and pH value), and fluctuations in water quality parameters can significantly affect aquaculture efficiency and even lead to serious economic losses. In addition, the external environment (such as air temperature, humidity, and air pressure) also affects the aquaculture environment through water heat exchange and dissolution characteristics.

[0003] In current technologies, most aquaculture still relies on traditional experience-based management and single-parameter monitoring equipment, such as manually sampling water quality or using a single sensor to record dissolved oxygen changes. While these methods can provide some environmental information, they generally suffer from insufficient real-time performance, limited monitoring range, and strong reliance on manual intervention. This leads to delayed responses to water quality deterioration or abnormal growth, increasing the risk of disease and mortality in farmed organisms.

[0004] Especially in high-density aquaculture models such as recirculating aquaculture systems, organisms are more sensitive to changes in the aquatic environment. Dynamic fluctuations in water quality and improper control of feeding often lead to uneven growth or health problems. The lack of monitoring methods for the growth status of individual organisms makes it impossible to detect potential problems in a timely manner, resulting in low management efficiency and high risks in aquaculture farms. Summary of the Invention

[0005] To overcome the above deficiencies, this invention provides a growth monitoring system for aquaculture organisms, aiming to improve the lack of effective means for real-time monitoring of biological growth and water quality in existing aquaculture, as well as the inability to achieve intelligent environmental control.

[0006] In a first aspect, the present invention provides the following technical solution: a growth monitoring system for aquaculture organisms, comprising:

[0007] The data acquisition module is used to collect water quality parameters, biological growth information, and external environmental data in real time.

[0008] The data processing module is used to clean, denoise, fuse, and normalize the collected data;

[0009] The sensitivity analysis module is used to analyze the impact of water quality parameters on biological growth and identify key water quality parameters.

[0010] The water quality growth modeling module is used to establish a nonlinear relationship model between water quality parameters and biological growth.

[0011] Anomaly detection module is used to detect abnormal patterns in water quality or biological growth and trigger alarms.

[0012] An automated control module is used to automatically adjust oxygenation equipment, water circulation control, and feed feeding strategies based on abnormal detection results.

[0013] The visualization and decision support module is used to monitor data changes in real time and generate analysis reports.

[0014] Preferably, the data acquisition module includes:

[0015] The water quality sensor submodule is used to collect dissolved oxygen, ammonia nitrogen, temperature, salinity and pH value in real time;

[0016] The biological growth image acquisition submodule is used to acquire biological images through a high-definition camera and extract body length and estimate weight based on a deep learning algorithm. The weight estimation formula is W = k·L. 3 , where k is the calibration coefficient for a specific variety;

[0017] The external environment data acquisition submodule is used to collect air temperature, relative humidity, and air pressure.

[0018] Preferably, the data processing module includes:

[0019] The missing value imputation submodule is used to impute short-term missing data using linear interpolation and to impute long-term missing data based on the ARIMA model.

[0020] The data denoising submodule is used to process high-frequency noise in sensor-acquired data through wavelet transform and Kalman filtering.

[0021] The data normalization submodule is used to normalize water quality, biological, and external environmental data to the interval [0,1]. Its normalization formula is as follows:

[0022] The multi-source data fusion submodule is used to integrate water quality, biological growth, and external environmental data into a unified format dataset D. t =[X t ,L t W t ,T air ,RH,P].

[0023] Preferably, the sensitivity analysis module includes:

[0024] The parameter sensitivity analysis submodule is used to calculate the nonlinear contribution of water quality parameters to the growth rate based on a generalized linear model. The model formula is as follows: Where G represents the growth rate, f i (X i ) represents the nonlinear effect of water quality parameters;

[0025] The response curve analysis submodule is used to plot the response curves of key water quality parameters and growth rate. The fitting formula is as follows: Where a is the growth peak, b is the optimal water quality value, and c is the parameter distribution width.

[0026] Preferably, the water quality growth modeling module includes:

[0027] The deep learning modeling submodule is used to construct the nonlinear relationship between biological growth rate and water quality parameters. It uses a deep neural network for prediction, with the network input being the water quality parameter X. t The output is the growth rate G, and the network's loss function is...

[0028] The water quality threshold calculation submodule is used to determine the optimal range of water quality parameters based on the growth response curve [X]. min ,X max ], defined as satisfying G(X i )≥0.8·G max The range of water quality parameters.

[0029] Preferably, the anomaly detection module includes:

[0030] The water quality anomaly detection submodule is used to detect water quality anomalies using an LSTM time series prediction model. The prediction formula is as follows: When predicting residuals If the set threshold is exceeded, an alarm will be triggered;

[0031] The abnormal growth identification submodule classifies biological growth data based on the K-Means clustering algorithm to identify abnormal individuals.

[0032] Preferably, the automated control module includes:

[0033] The oxygenation regulation submodule is used to automatically adjust the operating speed of the oxygenation pump based on the deviation of dissolved oxygen. The adjustment formula is ΔDO = DO target -DO current ;

[0034] The water circulation control submodule dynamically adjusts the water circulation frequency based on a reinforcement learning algorithm, with the reward function being: Where b i These are the optimal values ​​for water quality parameters;

[0035] The feeding optimization submodule adjusts the feed amount based on the growth prediction model.

[0036] Secondly, the present invention provides the following technical solution: a method for monitoring the growth of aquatic organisms, comprising the following steps:

[0037] S1. Acquire water quality environmental data, including dissolved oxygen, ammonia nitrogen, temperature, salinity and pH value, through multi-parameter water quality sensors; collect multi-angle image data of aquaculture organisms through underwater high-definition cameras; extract organism length and weight information based on computer vision technology; collect external environmental data such as air temperature, humidity and air pressure using external environmental monitoring equipment; integrate and store all collected data according to timestamps.

[0038] S2. Process the collected data, including using interpolation methods to fill in short-term missing data and time series prediction methods to fill in long-term missing data, applying signal filtering technology to noise-process water quality and biological data, and normalizing multidimensional data to form a standardized data input set.

[0039] S3. By analyzing the correlation between water quality parameters and biological growth data, the contribution of each parameter to the growth rate is calculated, key water quality parameters that have a significant impact on growth are identified, and based on the selected key parameters, a response curve of the growth rate as a function of water quality parameters is generated to describe the specific pattern of the relationship between water quality and growth.

[0040] S4. Based on the selected key water quality parameters, construct a nonlinear relationship model between biological growth rate and water quality parameters, and use this model to analyze the optimal range of water quality parameters. Through the parameter response curve fitting results, extract the water quality threshold range that can support optimal growth.

[0041] S5. Perform time series prediction on the real-time collected water quality data, compare it with the actual monitoring values ​​to identify potential water quality anomalies, group biological growth data into individuals using cluster analysis, mark abnormal individuals that are different from the normal growth pattern, and record the detection results as abnormal events.

[0042] S6. Adjust the aquaculture environment parameters based on the abnormal detection results, including controlling the dissolved oxygen concentration by adjusting the oxygenation equipment, adjusting the operating frequency of the water circulation system based on real-time water quality data, adjusting the feed amount based on the prediction results of the biological growth model, and recording the parameter changes during the control process.

[0043] S7. The collected water quality parameters, biological growth information and external environmental data are displayed in real time through charts and graphs, and the time change trend is presented in a dynamic manner. When water quality or growth abnormalities are detected, alarm prompts and suggestions for adjusting environmental parameters are generated. Based on historical data, growth trend analysis reports and parameter management suggestion documents are generated for the breeding cycle.

[0044] Thirdly, the invention provides the following technical solution: a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described image recognition method.

[0045] Fourthly, the present invention provides the following technical solution: a readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the above-described image recognition method.

[0046] The present invention has the following beneficial effects:

[0047] 1. In this invention, by collecting water quality parameters and biological growth information in real time and combining them with data processing and analysis technology, the aquaculture environment and feeding strategy can be dynamically optimized, thereby significantly improving the biological growth rate, feed utilization rate and overall aquaculture efficiency.

[0048] 2. In this invention, through real-time monitoring and anomaly detection of water quality parameters, environmental problems can be quickly identified and the automated control system can be triggered to adjust water parameters, thereby effectively maintaining the stability of the aquaculture environment and reducing the impact of environmental fluctuations on organisms.

[0049] 3. In this invention, by intelligently detecting abnormal water quality and growth, and combining this with an early warning mechanism to remind aquaculture managers to intervene in a timely manner, the risks of aquaculture caused by water quality deterioration, disease transmission, or poor individual growth can be significantly reduced. Attached Figure Description

[0050] Figure 1 This is a system architecture diagram of a growth monitoring system for aquaculture organisms proposed in this invention.

[0051] Figure 2 This is a data acquisition module architecture diagram of a growth monitoring system for aquaculture organisms proposed in this invention.

[0052] Figure 3 This is a data processing module architecture diagram of a growth monitoring system for aquaculture organisms proposed in this invention.

[0053] Figure 4 This is a diagram illustrating the architecture of a sensitivity analysis module in a growth monitoring system for aquaculture organisms proposed in this invention.

[0054] Figure 5 This is an architecture diagram of a water quality growth modeling module for a growth monitoring system for aquaculture organisms proposed in this invention.

[0055] Figure 6 This is an architecture diagram of an anomaly detection module for a growth monitoring system of aquatic organisms proposed in this invention.

[0056] Figure 7 This is an architecture diagram of an automated control module for a growth monitoring system of aquatic organisms proposed in this invention.

[0057] Figure 8 This is a flowchart illustrating the steps of a method for monitoring the growth of aquatic organisms proposed in this invention. Detailed Implementation

[0058] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Example 1

[0060] Reference Figure 1 In the first embodiment of the present invention, the present invention provides a growth monitoring system for aquaculture organisms, comprising:

[0061] Reference Figure 2 The data acquisition module is used to collect water quality parameters, biological growth information, and external environmental data in real time.

[0062] include:

[0063] The water quality sensor submodule is used to collect dissolved oxygen, ammonia nitrogen, temperature, salinity and pH value in real time;

[0064] The biological growth image acquisition submodule is used to acquire biological images through a high-definition camera and extract body length and estimate weight based on a deep learning algorithm. The weight estimation formula is W = k·L. 3 , where k is the calibration coefficient for a specific variety;

[0065] The external environment data acquisition submodule is used to collect air temperature, relative humidity, and air pressure.

[0066] Specifically, the water quality sensor submodule is used to collect key water quality parameters of aquaculture water in real time, including dissolved oxygen (DO), nitrogen (NH3), temperature (T), salinity (S), and pH value.

[0067] The sensor is installed in the central area of ​​the aquaculture pond to ensure the representativeness of the data collected. The specific implementation of this submodule includes the following:

[0068] Sensor equipment: A multi-parameter water quality sensor (such as YSIProDSS) is used to simultaneously monitor multiple water quality parameters. The sensor features high accuracy and a waterproof design, enabling it to withstand long-term underwater operation.

[0069] Data acquisition frequency: The sensor records water quality data every minute and uploads the data to the system's storage unit through an integrated data interface.

[0070] Data formatting: Water quality data is preprocessed by the sensor's internal processor and formatted as a timestamped multidimensional array, for example:

[0071] X t =[DO t ,NH 3- N t ,T t ,S t pH t ]

[0072] Where t represents the time step of data collection.

[0073] Interface communication: The collected data is sent to the central processing unit via wireless communication protocols (such as LoRa, Wi-Fi) or wired transmission for subsequent processing and analysis.

[0074] The biological growth image acquisition submodule is used to acquire image data of farmed organisms in real time using a high-definition underwater camera, and then uses deep learning algorithms to extract the organisms' body length and estimate their weight. The specific implementation of this submodule includes the following:

[0075] Camera equipment: High-resolution underwater cameras with waterproof features are selected and installed symmetrically on the walls of the aquaculture pond to ensure coverage of the entire aquaculture area.

[0076] Image acquisition: The camera captures dynamic images of the farmed organisms at fixed time intervals (e.g., every 10 minutes) in high-definition JPEG or RAW file format.

[0077] Image preprocessing: Denoising and enhancement of the acquired raw images, including brightness adjustment, contrast optimization and gamma correction.

[0078] Use deep learning-based object detection models (such as YOLO or Mask R-CNN) to segment the contour regions of farmed organisms and label individuals.

[0079] Body length extraction: Edge detection is performed on the segmented biological contour, and the body length of the organism is obtained by calculating the maximum diameter of the contour.

[0080] Body length extraction: Edge detection is performed on the segmented biological contour, and the body length of the organism is obtained by calculating the maximum diameter of the contour.

[0081] The steps for extracting body length include:

[0082] 1. Convert the pixel length to the actual length by calculating based on the camera's pixel resolution and calibration ratio.

[0083] 2. Formula for calculating body length:

[0084]

[0085] Where, N pixcls is the outline pixel length, and scale is the scaling factor from pixel to actual length.

[0086] Body weight estimation: Estimating biological body weight using an empirical formula relating body length and body weight.

[0087] W = k·L 3

[0088] Where W represents body weight, L represents body length, and k is a calibration coefficient for a specific breed, which needs to be calibrated based on experimental data.

[0089] The system automatically records and stores body length and weight information, generating a biological growth dataset.

[0090] Real-time performance and scalability: Supports real-time processing of multi-channel camera data and can be scaled to more varieties and larger breeding areas.

[0091] The external environment data acquisition submodule is used to monitor the environmental conditions outside the farm, including air temperature (T). air The relative humidity (RH) and air pressure (P) provide a basis for analyzing the impact of the external environment on aquaculture water bodies and biological growth.

[0092] The specific implementation of this submodule includes the following: Sensor equipment: It adopts an integrated meteorological and body temperature sensor, and has a temperature and humidity sensor and a barometric pressure sensor to support all-weather operation.

[0093] Data acquisition methods: Temperature data is measured in real time by a thermistor with an accuracy range of ±0.1℃; humidity data is acquired by a capacitive humidity sensor with a measurement range of 0-100%; and air pressure data is measured by a piezoresistive air pressure sensor with the recording unit being hPa.

[0094] Data Acquisition Frequency and Communication: Data is acquired every 30 minutes. Higher acquisition frequencies may be triggered in cases of significant weather changes (such as sharp temperature drops), or data may be uploaded to the central data processing system via wired or wireless transmission interfaces.

[0095] Data fusion: External environmental data, water quality parameters, and biological growth information are synchronized over time to form a unified data input format for subsequent analysis.

[0096] The three sub-modules work together to collect water quality, biological, and external environmental data in real time, and integrate the data according to timestamps to form a complete multidimensional dataset.

[0097] The dataset is structured as follows:

[0098] D t =[X t ,L t W t ,T air ,RH,P]

[0099] Where: X t =[DO t ,NH3-N t ,T t ,S t pH t [ ] represents water quality parameters; ##·L t W t For the length and weight of an organism; T air RH and P are external environmental parameters.

[0100] Through this data acquisition module, the system can obtain comprehensive information about the aquaculture environment and organisms in real time, providing basic support for subsequent data processing, modeling, and regulation.

[0101] Reference Figure 3 The data processing module is used to clean, denoise, fuse, and normalize the collected data.

[0102] Includes: a missing value imputation submodule, used to impute short-term missing data using linear interpolation, and to impute long-term missing data based on the ARIMA model;

[0103] The data denoising submodule is used to process high-frequency noise in sensor-acquired data through wavelet transform and Kalman filtering.

[0104] The data normalization submodule is used to normalize water quality, biological, and external environmental data to the interval [0,1]. Its normalization formula is as follows:

[0105] The multi-source data fusion submodule is used to integrate water quality, biological growth, and external environmental data into a unified format dataset D. t =[X t ,L t W t ,T air ,RH,P].

[0106] Specifically, the missing value imputation submodule is used to handle short-term and long-term missing data in the collected data. For short-term missing data, linear interpolation is used for imputation; for long-term missing data, prediction imputation is performed based on a time series ARIMA model, thereby ensuring the integrity and continuity of the data.

[0107] The data denoising submodule processes high-frequency noise in the acquired data using wavelet transform and Kalman filtering. Wavelet transform decomposes the data signal and removes high-frequency noise components, while Kalman filtering further smooths the data and reduces random errors during sensor acquisition.

[0108] The data normalization submodule maps water quality parameters, biological growth data, and external environmental data to a unified numerical range [0,1] to eliminate the influence between different data units.

[0109] The multi-source data fusion submodule integrates water quality parameters, biological growth information, and external environmental data into a unified time-series format dataset. The fused dataset is labeled according to timestamps, forming a multi-dimensional data structure that includes water quality, biological, and external environmental variables, providing complete input for subsequent modeling and analysis.

[0110] Reference Figure 4 The sensitivity analysis module is used to analyze the impact of water quality parameters on biological growth and to identify the parameters.

[0111] Includes: a parameter sensitivity analysis submodule, used to calculate the nonlinear contribution of water quality parameters to the growth rate based on a generalized linear model, the model formula being... Where G represents the growth rate, f i (X i ) represents the nonlinear effect of water quality parameters;

[0112] The response curve analysis submodule is used to plot the response curves of key water quality parameters and growth rate. The fitting formula is as follows: Where a is the growth peak, b is the optimal water quality value, and c is the parameter distribution width.

[0113] Specifically, the parameter sensitivity analysis submodule calculates the nonlinear contribution of water quality parameters to biological growth rate based on a generalized additive model. The generalized additive model establishes the nonlinear relationship between growth rate and multiple water quality parameters by decomposing the independent effects of water quality parameters.

[0114] Model formula:

[0115]

[0116] Where: G: biological growth rate (unit: mm / day or g / day); β0: model bias term; fi (X i ): The nonlinear effect function of the i-th water quality parameter; X i : The i-th water quality parameter (such as dissolved oxygen, ammonia nitrogen, etc.); n: The total number of water quality parameters; ∈: Random error.

[0117] Parameter contribution calculation: X for each water quality parameter i The contribution of f is through its nonlinear effect. i (X i Variance calculation:

[0118]

[0119] Based on the contribution results, key water quality parameters that contribute significantly to biological growth rate were selected.

[0120] Implementation process: 1. Input the processed multidimensional water quality parameters and growth rate data;

[0121] 2. Use a generalized additive model to fit the growth rate G and water quality parameter X. i Relationship;

[0122] 3. Calculate the nonlinear contribution of each parameter to the growth rate;

[0123] 4. Output the contribution ranking results of the parameters and mark the key water quality parameters.

[0124] The response curve analysis submodule is used to plot the response curves between key water quality parameters and growth rate, and further quantifies the response relationship through fitting formulas. The fitting results provide the optimal range of key water quality parameters and the variation pattern of the growth rate.

[0125] Fitting formula: Key water quality parameter X i With growth rate G(X) i The response relationship was fitted using a Gaussian distribution curve:

[0126]

[0127] Where: G(X) i Growth rate versus parameter X i The response value;

[0128] a: Peak growth rate, representing the maximum growth rate under optimal water quality conditions;

[0129] b: Optimal water quality value, i.e., the parameter value at which the growth rate reaches its peak;

[0130] c: Parameter distribution width, reflecting the sensitivity of the parameter to the growth rate.

[0131] Parameter meaning:

[0132] The size of 'a' determines the water quality parameter X. i The intensity of the potential impact on biological growth;

[0133] The value of b is used to guide actual water quality control strategies;

[0134] The range of 'c' reflects the flexibility of water quality parameter regulation.

[0135] Implementation process:

[0136] 1. Input the selected key water quality parameters and their corresponding growth rate data.

[0137] 2. Fitting key water quality parameter X using nonlinear least squares method i The response curve.

[0138] 3. Calculate the growth peak value a, the optimal water quality value b, and the distribution width c.

[0139] 4. Output the fitting results and growth rate response curve, and store them for subsequent modeling and optimization.

[0140] Reference Figure 5 The water quality growth modeling module is used to establish a nonlinear relationship model between water quality parameters and biological growth.

[0141] Includes: a deep learning modeling submodule, used to construct the nonlinear relationship between biological growth rate and water quality parameters, using a deep neural network for prediction, with the network input being the water quality parameter X. t The output is the growth rate G, and the network's loss function is...

[0142] The water quality threshold calculation submodule is used to determine the optimal range of water quality parameters based on the growth response curve [X]. min ,X max ], defined as satisfying G(X i )≥0.8·G max The range of water quality parameters.

[0143] Specifically, the deep learning modeling submodule uses deep neural networks to construct a nonlinear relationship model between water quality parameters and biological growth rate, thereby achieving accurate prediction of biological growth rate.

[0144] Input layer: Water quality parameters [DO, NH3-N, T, S, pH], with an input dimension of n, where n represents the number of water quality parameters.

[0145] Hidden layer: Includes multiple fully connected layers. The network structure is designed with three hidden layers, with the number of nodes being [64, 128, 64].

[0146] Output layer: Biological growth rate G, which is a scalar output.

[0147] Training Process: Input Data: Preprocessed water quality parameters are used as the model input, and biological growth rate is used as the model output target. Activation Function: The hidden layer uses the ReLU activation function, and the output layer uses the linear activation function. Optimizer: The Adam optimization algorithm is used, with a learning rate set to 0.001. Loss Function: The mean squared error is used to measure the difference between the model's predicted values ​​and the actual values. The loss function is defined as follows:

[0148]

[0149] in, This represents the growth rate predicted by the model. This represents the actual growth rate, and n is the number of training samples.

[0150] Model training steps:

[0151] The input data is divided into training and testing sets, usually in an 80:20 ratio.

[0152] Optimize model parameters on the training set by continuously adjusting weights and biases by minimizing the loss function.

[0153] Validate the model performance on the test set and evaluate the accuracy of the predictions.

[0154] The water quality threshold calculation submodule, based on the output of the deep learning modeling submodule and combined with the growth response curve, determines the optimal range of water quality parameters to guide water quality regulation.

[0155] Growth response curve fitting:

[0156] Using key water quality parameters and predicted growth rate data, a growth response curve was fitted.

[0157] The formula for the response curve is expressed as:

[0158]

[0159] in:

[0160] G(X i Growth rate affects key parameter X i The response value;

[0161] a: Peak growth;

[0162] b: The optimal value of the key parameter.

[0163] c: The response width of the key parameter, reflecting its sensitivity to growth.

[0164] Optimal range of water quality parameters: The optimal range of water quality parameters calculated based on the response curve is defined as the range of parameters that meets the following conditions:

[0165] G(X i )≥0.8·G max

[0166] Among them, G max The maximum value representing the growth rate

[0167] Implementation steps:

[0168] 1. Based on the growth rate data and water quality parameters output by the model, fit the growth response curve for each key parameter;

[0169] 2. Calculate the parameter range when the growth rate reaches its peak of 80%, and extract [X]. min ,X max ];

[0170] 3. Output the optimal range of water quality parameters to optimize the aquaculture environment.

[0171] Reference Figure 6 Anomaly detection module, used to detect abnormal patterns in water quality or biological growth and trigger alarms;

[0172] include:

[0173] The water quality anomaly detection submodule is used to detect water quality anomalies using an LSTM time series prediction model. The prediction formula is as follows: When predicting residuals If the set threshold is exceeded, an alarm will be triggered;

[0174] The abnormal growth identification submodule classifies biological growth data based on the K-Means clustering algorithm to identify abnormal individuals.

[0175] Specifically, the growth anomaly identification submodule classifies biological growth data based on the K-Means clustering algorithm, and identifies abnormal individuals by analyzing their body length, weight, and growth rate.

[0176] Implementation method: The K-Means clustering algorithm is used to classify growth data into two categories: "normal growth" and "abnormal growth". Input features include body length, weight, and growth rate data. Abnormal individuals are marked based on the degree of deviation from the cluster center.

[0177] Reference Figure 7 An automated control module is used to automatically adjust the oxygenation equipment, water circulation control, and feed feeding strategy based on abnormal detection results.

[0178] Includes: an oxygenation regulation submodule, used to automatically adjust the oxygenation pump operating speed according to the dissolved oxygen deviation, with the adjustment formula being ΔDO = DO target -DO current ;

[0179] The water circulation control submodule dynamically adjusts the water circulation frequency based on a reinforcement learning algorithm, with the reward function being: Where b i These are the optimal values ​​for water quality parameters;

[0180] The feeding optimization submodule adjusts the feed amount based on the growth prediction model.

[0181] Specifically, the oxygenation equipment control submodule automatically adjusts the operating status of the oxygenation equipment based on abnormal dissolved oxygen conditions.

[0182] Control logic: When dissolved oxygen is below the set threshold, start or increase the operating speed of the oxygenation equipment;

[0183] When dissolved oxygen reaches the target value, reduce the equipment operating speed or turn off the oxygenation equipment.

[0184] The water circulation control submodule triggers automatic adjustment of the water circulation equipment based on abnormal water quality parameters (such as ammonia nitrogen and salinity).

[0185] Regulation logic:

[0186] When ammonia nitrogen or salinity exceeds the standard, increase the water circulation frequency or activate the water body drainage mode;

[0187] When the water quality returns to normal, resume the standard circulation frequency.

[0188] The feed feeding strategy optimization submodule adjusts the amount and timing of feed based on the results of abnormal growth detection and the growth rate prediction model.

[0189] Regulation logic:

[0190] For individuals with abnormal growth, increase or decrease the amount of feed given;

[0191] Adjust the daily feeding frequency according to the population growth rate.

[0192] The visualization and decision support module is used to monitor data changes in real time and generate analysis reports.

[0193] Specifically, the real-time monitoring submodule displays key parameters of the aquaculture environment and biological growth through a dynamic visual interface.

[0194] Monitoring content:

[0195] Water quality parameters: real-time curves of dissolved oxygen, ammonia nitrogen, temperature, salinity, pH, etc.

[0196] Biological data: mean body length, rate of weight gain, and individual abnormal markers;

[0197] External environmental data: trends in air temperature, humidity, and air pressure.

[0198] Presentation method:

[0199] Time series plot: Shows the historical trends of key parameters;

[0200] Real-time status graph: Presents the real-time values ​​of current parameters in the form of a dashboard or heatmap;

[0201] Abnormal alert: When a parameter exceeds the threshold range, a red marker or pop-up warning will be triggered.

[0202] Example 2:

[0203] Reference Figure 8 In a second embodiment of the present invention, the present invention provides a method for monitoring the growth of aquaculture organisms, comprising the following steps:

[0204] S1. Acquire water quality environmental data, including dissolved oxygen, ammonia nitrogen, temperature, salinity and pH value, through multi-parameter water quality sensors; collect multi-angle image data of aquaculture organisms through underwater high-definition cameras; extract organism length and weight information based on computer vision technology; collect external environmental data such as air temperature, humidity and air pressure using external environmental monitoring equipment; integrate and store all collected data according to timestamps.

[0205] Specifically, multi-parameter water quality sensors are used to collect real-time water quality data, including dissolved oxygen, ammonia nitrogen, temperature, salinity, and pH. Underwater high-definition cameras are used to collect multi-angle images of aquaculture organisms, and computer vision technology is used to extract organism length and weight information. External environmental data, including air temperature, humidity, and air pressure, is acquired through weather station equipment. All collected data is integrated and stored according to timestamps to form a complete dataset.

[0206] S2. Process the collected data, including using interpolation methods to fill in short-term missing data and time series prediction methods to fill in long-term missing data, applying signal filtering technology to noise-process water quality and biological data, and normalizing multidimensional data to form a standardized data input set.

[0207] Specifically, the collected data is processed, including using interpolation methods to complete short-term missing data and using time series forecasting methods to process long-term missing data. Noise filtering is performed on water quality and biological growth data to remove equipment errors and environmental interference. All data is normalized to convert data of different dimensions into standardized values, forming a high-quality data input set suitable for subsequent analysis.

[0208] S3. By analyzing the correlation between water quality parameters and biological growth data, the contribution of each parameter to the growth rate is calculated, key water quality parameters that have a significant impact on growth are identified, and based on the selected key parameters, a response curve of the growth rate as a function of water quality parameters is generated to describe the specific pattern of the relationship between water quality and growth.

[0209] Specifically, based on the selected key parameters, a response curve of growth rate as a function of water quality parameters is generated to describe the influence pattern of water quality parameters on biological growth and to provide input for subsequent modeling.

[0210] S4. Based on the selected key water quality parameters, construct a nonlinear relationship model between biological growth rate and water quality parameters, and use this model to analyze the optimal range of water quality parameters. Through the parameter response curve fitting results, extract the water quality threshold range that can support optimal growth.

[0211] Specifically, by utilizing historical data and parameter response curves, the optimal range of water quality parameters is analyzed, and the water quality threshold range supporting optimal growth is extracted, providing specific parameter references for water quality environment optimization.

[0212] S5. Perform time series prediction on the real-time collected water quality data, compare it with the actual monitoring values ​​to identify potential water quality anomalies, group biological growth data into individuals using cluster analysis, mark abnormal individuals that are different from the normal growth pattern, and record the detection results as abnormal events.

[0213] S6. Adjust the aquaculture environment parameters based on the abnormal detection results, including controlling the dissolved oxygen concentration by adjusting the oxygenation equipment, adjusting the operating frequency of the water circulation system based on real-time water quality data, adjusting the feed amount based on the prediction results of the biological growth model, and recording the parameter changes during the control process.

[0214] Specifically, dissolved oxygen concentration is controlled by adjusting aeration equipment; the operating frequency of the water circulation system is adjusted or a water exchange program is initiated based on water quality conditions; and feed feeding strategies are adjusted in conjunction with the biological growth status, including optimizing feeding amount and feeding time, and changes in all control parameters are recorded.

[0215] S7. The collected water quality parameters, biological growth information and external environmental data are displayed in real time through charts and graphs, and the time change trend is presented in a dynamic manner. When water quality or growth abnormalities are detected, alarm prompts and suggestions for adjusting environmental parameters are generated. Based on historical data, growth trend analysis reports and parameter management suggestion documents are generated for the breeding cycle.

[0216] Example 3

[0217] In a third embodiment of the present invention, based on the same inventive concept, the present invention proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of a growth monitoring method for aquaculture organisms as described in the above embodiments.

[0218] Example 4

[0219] In the fourth embodiment of the present invention, based on the same inventive concept, the present invention proposes a computer device, which includes: a processor and a memory; the processor and the memory communicate with each other; the memory is used to store instructions; the processor is used to execute the instructions in the memory to execute the growth monitoring method for aquaculture organisms of the above embodiment.

[0220] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0221] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A growth monitoring system for aquaculture organisms, characterized in that, include: The data acquisition module is used to collect water quality parameters, biological growth information, and external environmental data in real time. The data processing module is used to clean, denoise, fuse, and normalize the collected data. The sensitivity analysis module is used to analyze the impact of water quality parameters on biological growth and to identify the parameters. The water quality growth modeling module is used to establish a nonlinear relationship model between water quality parameters and biological growth. Anomaly detection module is used to detect abnormal patterns in water quality or biological growth and trigger alarms. An automated control module is used to automatically adjust the oxygenation equipment, water circulation control, and feed feeding strategy based on abnormal detection results; The visualization and decision support module is used to monitor data changes in real time and generate analysis reports; The data processing module includes: The missing value imputation submodule is used to impute short-term missing data using linear interpolation and to impute long-term missing data based on the ARIMA model. The data denoising submodule is used to process high-frequency noise in sensor-acquired data through wavelet transform and Kalman filtering. The data normalization submodule is used to normalize water quality parameters, biological growth information, and external environmental data to the interval [0,1]. Its normalization formula is as follows: ; The multi-source data fusion submodule is used to integrate water quality parameters, biological growth information, and external environmental data into a unified format dataset. ; The anomaly detection module includes: The water quality anomaly detection submodule is used to detect water quality anomalies using an LSTM time series prediction model. The prediction formula is as follows: When predicting residuals If the set threshold is exceeded, an alarm will be triggered; The abnormal growth identification submodule classifies biological growth data based on the K-Means clustering algorithm to identify abnormal individuals. The data acquisition module includes: The water quality sensor submodule is used to collect dissolved oxygen, ammonia nitrogen, temperature, salinity and pH value in real time; The biological growth image acquisition submodule is used to acquire biological images through a high-definition camera and extract body length and estimate weight based on a deep learning algorithm. The weight estimation formula is as follows: ,in For a specific variety, the calibration coefficient is used. The external environment data acquisition submodule is used to collect air temperature, relative humidity, and air pressure. The sensitivity analysis module includes: The parameter sensitivity analysis submodule is used to calculate the nonlinear contribution of water quality parameters to the growth rate based on a generalized linear model. The model formula is as follows: ,in, Indicates growth rate, This represents the nonlinear effect of water quality parameters; The response curve analysis submodule is used to plot the response curves of key water quality parameters and growth rate. The fitting formula is as follows: ,in This represents the peak growth period. For optimal water quality values, The width of the parameter distribution.

2. The growth monitoring system for aquaculture organisms according to claim 1, characterized in that, The water quality growth modeling module includes: The deep learning modeling submodule is used to construct the nonlinear relationship between biological growth rate and water quality parameters. It uses a deep neural network for prediction, with the water quality parameters as input to the network. The output is the growth rate. The network's loss function is ; The water quality threshold calculation submodule is used to determine the optimal range of water quality parameters based on the response curve. Defined as satisfying The range of water quality parameters.

3. The growth monitoring system for aquaculture organisms according to claim 2, characterized in that, The automated control module includes: The oxygenation regulation submodule is used to automatically adjust the operating speed of the oxygenation pump based on the deviation in dissolved oxygen levels. The adjustment formula is as follows: ; The water circulation control submodule dynamically adjusts the water circulation frequency based on a reinforcement learning algorithm, with the reward function being: ,in These are the optimal values ​​for water quality parameters; The feeding optimization submodule adjusts the feed amount based on the growth prediction model.

4. A method for monitoring the growth of aquaculture organisms, characterized in that, The growth monitoring system for aquaculture organisms as described in claim 3 includes the following steps: S1. Acquire water quality environmental data, including dissolved oxygen, ammonia nitrogen, temperature, salinity and pH value, through multi-parameter water quality sensors; collect multi-angle image data of aquaculture organisms through underwater high-definition cameras; extract organism length and weight information based on computer vision technology; collect external environmental data such as air temperature, humidity and air pressure using external environmental monitoring equipment; integrate and store all collected data according to timestamps. S2. Process the collected data, including using interpolation methods to fill in short-term missing data and time series prediction methods to fill in long-term missing data, applying signal filtering technology to noise-process water quality and biological data, and normalizing multidimensional data to form a standardized data input set. S3. By analyzing the correlation between water quality parameters and biological growth data, the contribution of each parameter to the growth rate is calculated, key water quality parameters that have a significant impact on growth are identified, and based on the selected key parameters, a response curve of the growth rate as a function of water quality parameters is generated to describe the specific pattern of the relationship between water quality and growth. S4. Based on the selected key water quality parameters, construct a nonlinear relationship model between biological growth rate and water quality parameters, and use this model to analyze the optimal range of water quality parameters. Through the parameter response curve fitting results, extract the water quality threshold range that can support optimal growth. S5. Perform time series prediction on the real-time collected water quality data, compare it with the actual monitoring values ​​to identify potential water quality anomalies, group biological growth data into individuals using cluster analysis, mark abnormal individuals that are different from the normal growth pattern, and record the detection results as abnormal events. S6. Adjust the aquaculture environment parameters based on the abnormal detection results, including controlling the dissolved oxygen concentration by adjusting the oxygenation equipment, adjusting the operating frequency of the water circulation system based on real-time water quality data, adjusting the feed amount based on the prediction results of the growth prediction model, and recording the parameter changes during the control process. S7. The collected water quality parameters, biological growth information and external environmental data are displayed in real time through charts and graphs, and the time change trend is presented in a dynamic manner. When water quality or growth abnormalities are detected, alarm prompts and suggestions for adjusting environmental parameters are generated. Based on historical data, growth trend analysis reports and parameter management suggestion documents are generated for the breeding cycle.

5. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the growth monitoring method for aquaculture organisms as described in claim 4.

6. A readable storage medium, characterized in that, The readable storage medium stores a computer program, which, when executed by a processor, implements the growth monitoring method for aquaculture organisms as described in claim 4.

Images