Intelligent fish culture automatic control system and method based on Internet of Things

Abstract

The invention discloses an intelligent fish culture automatic control system and method based on the Internet of Things, and relates to the technical field of aquaculture control, the system comprises a data sensing unit used for collecting water quality parameters, environment parameters and fish behavior information of a culture environment in real time by using a multi-parameter water quality sensor and an environment sensor; the intelligent control unit is used for executing a multi-parameter association control algorithm and an early warning control strategy based on the breeding environment data and executing a control instruction to adjust breeding environment parameters; the network communication unit is used for establishing data communication connection between the local and the cloud; the remote application unit is used for remote monitoring and manual intervention; the intelligent control unit comprises a main control module which is used for establishing an ammonia nitrogen driven dissolved oxygen early warning control strategy according to the water quality parameters and generating a stage treatment instruction in combination with the environmental parameters. Multiple sensors and an automatic control algorithm are integrated, water quality parameters can be monitored in real time, and the fish culture environment can be automatically adjusted.

Description

Technical Field

[0001] This invention relates to the field of aquaculture control technology, and more specifically, to an intelligent fish farming automatic control system and method based on the Internet of Things. Background Technology

[0002] With the rapid development of modern aquaculture, traditional farming methods relying on manual experience are no longer sufficient to meet the demands of aquaculture and production. Traditional fish farming suffers from problems such as lagging water quality monitoring, extensive feeding control, and low environmental regulation efficiency. Real-time monitoring and control of water quality parameters such as dissolved oxygen, pH, and ammonia nitrogen concentration directly affect the healthy growth of fish and the profitability of aquaculture. Manual monitoring methods often suffer from insufficient monitoring frequency and slow response, easily leading to problems such as oxygen deficiency and ammonia poisoning in fish. Simultaneously, difficulties in controlling feed intake result in high feed waste rates, and uneaten feed further deteriorates the water quality, affecting fish growth. These problems severely restrict the improvement of economic efficiency in aquaculture.

[0003] To address the problems of traditional aquaculture models, the Internet of Things (IoT) has begun to be applied in the field of smart aquaculture in recent years. By deploying wireless sensor networks and remote control systems, preliminary intelligent management of the aquaculture environment has been achieved. Existing IoT-based aquaculture systems mainly use sensors for basic water quality monitoring and transmit data to a monitoring center via wireless communication, in conjunction with basic automated equipment to achieve remote monitoring functions. These technologies have improved the level of automation in aquaculture to a certain extent, reduced the burden of manual management, and provided a preliminary technological foundation for the digitalization of aquaculture.

[0004] However, existing technologies still have significant shortcomings. On the one hand, while existing systems can monitor single water quality parameters, they lack the ability to establish a dynamic correlation model between dissolved oxygen, ammonia nitrogen concentration, and pH value. They cannot automatically activate oxygenation pumps and circulating filtration systems when ammonia nitrogen concentration rises, and cannot achieve multi-parameter coordinated optimization control. Furthermore, traditional automated equipment lacks intelligent decision-making capabilities and cannot dynamically adjust control strategies according to the fish's growth stage. On the other hand, existing systems can only perform single-point monitoring and passive alarms, lacking the ability to integrate multi-dimensional sensor data such as dissolved oxygen, pH, ammonia nitrogen, temperature, and turbidity to establish a predictive model for water quality deterioration trends. This prevents early warning and intervention before water quality anomalies occur. Simultaneously, they lack rapid response mechanisms for emergencies such as equipment failure and extreme weather, resulting in insufficient preventative protection capabilities and often only allowing for reactive handling after problems occur. In addition, existing technologies still rely on fixed time intervals and feeding amounts for feeding control, failing to dynamically adjust based on the actual feeding behavior and growth needs of the fish, thus causing feed waste and water pollution.

[0005] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0006] To address the problems in related technologies, this invention proposes an intelligent fish farming automatic control system and method based on the Internet of Things, which has the advantages of multi-parameter correlation control, intelligent feeding optimization, and remote monitoring and anomaly handling, thereby solving the problems of single-parameter control limitations, serious feed waste, and high manual supervision costs in existing technologies.

[0007] Therefore, the specific technical solution adopted by the present invention is as follows:

[0008] According to one aspect of the present invention, an Internet of Things (IoT)-based intelligent fish farming automatic control system is provided, the IoT-based intelligent fish farming automatic control system comprising:

[0009] The data sensing unit is used to collect aquaculture environment data in real time using multi-parameter water quality sensors and environmental sensors. The aquaculture environment data includes water quality parameters, environmental parameters and fish behavior information.

[0010] The intelligent control unit is used to execute multi-parameter correlation control algorithms and early warning control strategies based on aquaculture environment data, and to execute control commands to adjust aquaculture environment parameters;

[0011] The network communication unit is used to establish a data communication connection between the local area and the cloud, and to transmit control commands and monitoring data.

[0012] The remote application unit is used to receive remote monitoring requests and respond to manual intervention instructions;

[0013] The intelligent control unit includes a main control module for establishing an ammonia nitrogen-driven dissolved oxygen early warning control strategy based on water quality parameters and generating graded treatment instructions in combination with environmental parameters; an actuator for adjusting aquaculture environmental parameters according to control instructions; and a storage module for storing the control strategy.

[0014] According to another aspect of the present invention, an intelligent fish farming automatic control method based on the Internet of Things (IoT) is also provided, the method comprising:

[0015] Using multi-parameter water quality sensors and environmental sensors, aquaculture environment data is collected in real time, including water quality parameters, environmental parameters, and fish behavior information;

[0016] Based on aquaculture environment data, a multi-parameter correlation control algorithm and early warning control strategy are executed, and control commands are executed to adjust aquaculture environment parameters;

[0017] Establish a data communication connection between the local machine and the cloud to transmit control commands and monitoring data;

[0018] It receives remote monitoring requests and responds to manual intervention instructions.

[0019] The beneficial effects of this invention are as follows:

[0020] (1) The present invention establishes an ammonia nitrogen-driven dissolved oxygen early warning control strategy through a water quality early warning module, realizing a multi-sensor data fusion and collaborative control mechanism, effectively solving the limitations of single parameter control; the system is based on the dissolved oxygen concentration, ammonia nitrogen concentration, water temperature and pH value data collected in real time by multi-parameter water quality sensors, uses time series alignment to eliminate the sampling delay between sensors, establishes a water quality parameter matrix, and establishes the basic coupling coefficient under different temperature ranges through a piecewise linear regression algorithm, and then corrects it through fish species-specific correction factors and density correction factors to form a complete coupling prediction model; when the ammonia nitrogen concentration rise rate exceeds the critical slope calculated by the system, the system can automatically predict the expected decrease of dissolved oxygen, start the oxygenation pump in advance and adjust its operating power, so as to increase the fish survival rate by 20% to 30%, effectively preventing fish mortality caused by water quality deterioration.

[0021] (2) The feeding control module of the present invention collects fish activity image data through an underwater camera, counts the proportion of fish in the feeding area, and calculates the feeding amount adjustment coefficient by combining the historical feeding completion time and the nutritional needs of the current breeding stage. This realizes fish behavior analysis and dynamic optimization of feeding amount based on machine vision. The system performs batch feeding control according to the preset feeding interval, avoiding feed waste and water pollution caused by feeding a large amount at once, and reducing feed costs by more than 15%. At the same time, the environmental adjustment module generates graded control instructions based on changes in water temperature, water quality parameters and fish behavior status, and intelligently adjusts the operating status of actuators such as heating or cooling devices, ultraviolet sterilization lamps and circulating water valves. This realizes on-demand start-up and shutdown of equipment and power adjustment, saving 25% of energy consumption compared with the traditional continuous operation mode, and significantly reducing the breeding operation cost.

[0022] (3) This invention establishes a data communication connection between the local area and the cloud through a network communication unit, and the remote application unit receives remote monitoring requests and responds to manual intervention instructions, thereby realizing remote monitoring and management of the fish farming environment. The system has a complete anomaly handling mechanism, which can monitor the operating parameters of the actuators at regular intervals, and monitor the changes in air pressure, water temperature and light intensity. When the extreme weather threshold is exceeded, the system classifies the anomaly level according to the magnitude of the change, executes graded response control and sends an early warning signal. In addition, the system monitors abnormal fish behavior through an underwater camera. When the fish aggregation ratio exceeds the anomaly threshold or abnormal features are detected on the body surface, the system automatically starts the ultraviolet sterilization lamp and adjusts the circulating water valve in combination with water quality risk parameter assessment, and implements differentiated treatment. This intelligent remote monitoring and anomaly handling mechanism greatly reduces the need for manual intervention, making the system particularly suitable for large-scale intensive fish farming scenarios. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram of an intelligent fish farming automatic control system based on the Internet of Things according to an embodiment of the present invention;

[0025] Figure 2 This is a detailed implementation diagram of an intelligent fish farming automatic control system based on the Internet of Things according to an embodiment of the present invention;

[0026] Figure 3 This is a flowchart illustrating an intelligent fish farming automatic control method based on the Internet of Things according to an embodiment of the present invention. Detailed Implementation

[0027] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments.

[0028] According to an embodiment of the present invention, an intelligent fish farming automatic control system and method based on the Internet of Things is provided.

[0029] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, according to one embodiment of the present invention, an intelligent fish farming automatic control system based on the Internet of Things (IoT) is provided, the intelligent fish farming automatic control system based on the IoT includes:

[0030] The data sensing unit 1 is used to collect aquaculture environment data in real time using multi-parameter water quality sensors and environmental sensors. The aquaculture environment data includes water quality parameters, environmental parameters and fish behavior information.

[0031] The intelligent control unit 2 is used to execute multi-parameter correlation control algorithms and early warning control strategies based on aquaculture environment data, and to execute control commands to adjust aquaculture environment parameters;

[0032] Network communication unit 3 is used to establish a data communication connection between the local area and the cloud, and to transmit control commands and monitoring data;

[0033] Remote application unit 4 is used to receive remote monitoring requests and respond to manual intervention instructions;

[0034] The intelligent control unit 2 includes a main control module for establishing an ammonia nitrogen-driven dissolved oxygen early warning control strategy based on water quality parameters and generating graded treatment instructions in combination with environmental parameters, an actuator for adjusting aquaculture environmental parameters according to control instructions, and a storage module for storing control strategies.

[0035] Specifically, the automatic control system provided by this invention is based on a classic four-layer Internet of Things (IoT) architecture. The perception layer corresponds to the data perception unit 1, which uses a distributed sensor network comprised of multi-parameter water quality sensors, underwater cameras, and environmental sensors to collect real-time water quality parameters, environmental parameters, and fish behavior information from the aquaculture environment. The network layer corresponds to the network communication unit 3, which establishes a data communication connection between the local area and the cloud via a wireless communication module using low-power wide-area network technologies such as LoRa or NB-IoT. This uploads the sensor data collected on-site to the cloud and simultaneously receives control commands from the cloud. The control layer corresponds to the intelligent control unit 2, whose main control module is responsible for executing multi-parameter correlation control algorithms and early warning control strategies based on the collected data, and for adjusting aquaculture environment parameters by executing control commands through actuators. The application layer corresponds to the remote application unit 4, which receives remote monitoring requests through a cloud control platform and a user control terminal, and responds to manual intervention commands. The entire system, through the IoT architecture of "sensing-networking-controlling-application", achieves full-link coverage from data sensing, network transmission, intelligent processing to application services, and builds a complete smart aquaculture IoT ecosystem. It effectively solves the problems of low frequency of manual monitoring, poor control accuracy, and difficulty in remote management in traditional aquaculture models, and realizes the digitalization, networking and intelligence of the fish farming process.

[0036] In one embodiment, such as Figure 2 As shown, the data sensing unit 1 includes a multi-parameter water quality sensor for monitoring water quality parameters, an underwater camera for monitoring fish density and feeding behavior, and an environmental sensor for monitoring environmental parameters.

[0037] The network communication unit 3 includes a wireless communication module for local control data transmission; the remote application unit 4 includes a cloud control platform for remote control and a user control terminal for manual control intervention.

[0038] The multi-parameter water quality sensor includes a dissolved oxygen sensor for detecting dissolved oxygen, a pH sensor for detecting pH value, an ammonia nitrogen sensor for detecting ammonia nitrogen concentration, a water temperature sensor for detecting temperature, and a turbidity sensor for detecting turbidity; the actuator includes an oxygenation pump for regulating dissolved oxygen, a heating and cooling device for regulating water temperature, an automatic feeder for controlling the amount of fish bait fed, a circulating water valve for water purification control, and an ultraviolet germicidal lamp for water sterilization control.

[0039] Specifically, in terms of hardware deployment, the dissolved oxygen sensor is installed at half the water depth of the fishpond, the pH sensor and ammonia nitrogen sensor are deployed in the middle and lower layers of the water, the water temperature sensor and turbidity sensor are distributed at different water depths, the underwater camera is installed facing the feeding area, the light intensity sensor is placed above the water surface, and the air pressure sensor is installed in the control box; the oxygenation pump and circulating water valve are controlled by the PID algorithm of the main control module to maintain the dissolved oxygen concentration above 5mg / L, the automatic feeder is installed above the feeding area, the heating and cooling devices are installed on the pond wall and the bottom of the water, respectively, and the ultraviolet germicidal lamp is arranged in the circulating water pipe; the wireless communication module in network communication unit 3 is installed in the waterproof control box to ensure stable signal transmission.

[0040] In one embodiment, the main control module includes:

[0041] The water quality early warning module is used to establish a coupled prediction model based on water quality parameters of the aquaculture environment, combined with fish species information and density monitoring data, and to calculate environmental load parameters based on fish density and tolerance data, determine the critical slope of ammonia nitrogen concentration increase, so as to establish an ammonia nitrogen-driven dissolved oxygen early warning and control strategy.

[0042] The feeding control module is used to calculate the proportion of fish in the feeding area based on fish activity image data, combine the historical feeding completion time with the nutritional needs of the current breeding stage, calculate the feeding amount adjustment coefficient, and execute batch feeding control according to the preset feeding interval.

[0043] The environmental control module is used to generate graded control commands based on environmental parameters monitored by environmental sensors. When water quality parameters continuously exceed the standard or fish behavior is abnormal, the module adjusts the operating status of the actuators in real time.

[0044] It is important to note that in freshwater fish farming environments, ammonia nitrogen concentration is a core control factor affecting fish survival and water quality stability. This is mainly due to the production mechanism and biotoxicity characteristics of ammonia nitrogen. Fish excrement, undigested feed residue, and the decomposition of organic matter continuously produce ammonia nitrogen. An increase in ammonia nitrogen concentration not only directly poisons fish but, more importantly, consumes a large amount of dissolved oxygen in the water through biological oxidation, creating a vicious cycle of "increased ammonia nitrogen → decreased dissolved oxygen → fish stress → decreased disease resistance → increased mortality." Compared to relatively stable environmental factors such as temperature and pH, ammonia nitrogen concentration changes are the most sensitive and respond the fastest, often showing a significant increase within 2-4 hours after feeding. If oxygen is not replenished in time, dissolved oxygen will drop below the safe threshold within 6-8 hours. Therefore, this invention uses ammonia nitrogen concentration as the dominant variable for predictive water quality control, enabling an active control strategy of early detection, early prediction, and early intervention. This is more scientific and effective than the traditional passive monitoring of dissolved oxygen changes and subsequent response, and it also meets the practical needs of modern intensive aquaculture for refined water quality management.

[0045] In one embodiment, when the water quality early warning module establishes a coupled prediction model based on aquaculture environment water quality parameters, combined with fish species information and density monitoring data, it includes:

[0046] Based on real-time data of dissolved oxygen concentration, ammonia nitrogen concentration, water temperature and pH value collected by multi-parameter water quality sensors, time series alignment is used to eliminate sampling delay between sensors, a water quality parameter matrix is ​​established, and historical datasets are generated by combining information on farmed fish species and density monitoring data.

[0047] Based on the correlation data between the rate of change of ammonia nitrogen concentration and the rate of dissolved oxygen consumption in historical datasets, a piecewise linear regression algorithm is used to establish the basic coupling coefficients for different temperature ranges.

[0048] A coupling prediction model is established by correcting the basic coupling coefficient using fish species-specific correction factors and density correction factors.

[0049] In one embodiment, when the water quality early warning module calculates environmental load parameters based on fish density and tolerance data to determine the critical slope for the rise in ammonia nitrogen concentration, in order to establish an ammonia nitrogen-driven dissolved oxygen early warning control strategy, it includes:

[0050] Based on the current ammonia nitrogen concentration and the coupled prediction model, the expected decrease in dissolved oxygen within the future time window is calculated to obtain the predicted value of dissolved oxygen concentration, and the prediction confidence interval is set in combination with the fish species' tolerance.

[0051] Based on the current fish density and daily feeding amount, combined with the fish's basal metabolic parameters and tolerance data, the environmental load parameters and dual-parameter baseline values ​​are calculated to determine the critical slope for the increase in ammonia nitrogen concentration.

[0052] When the rate of increase of ammonia nitrogen concentration exceeds the critical slope, the expected decrease in dissolved oxygen is calculated using a coupled prediction model, and the start-up timing and operating power of the oxygenation pump are set according to the expected decrease.

[0053] Specifically, in this embodiment, a multi-parameter water quality sensor continuously collects dissolved oxygen concentration, ammonia nitrogen concentration, water temperature, and pH data of the aquaculture water at 1-minute intervals. The system identifies sampling delays caused by differences in sensor response times by comparing the timestamps of the data from each sensor. A linear interpolation algorithm is used to align the data from all sensors to a standard time point, ensuring accurate matching of multi-parameter data at the same moment. The collected raw data undergoes a 5-point median filter to remove high-frequency noise, and outliers exceeding three times the standard deviation of the normal range are removed using the 3σ criterion. The processed dissolved oxygen concentration, ammonia nitrogen concentration, water temperature, and pH value are then compared using backward differential... The rate of change calculated by the method is combined into an 8-dimensional water quality parameter matrix; underwater cameras collect images of fish activity every 5 minutes, and the improved YOLO algorithm is used to identify the fish outlines. The number of fish is calculated by the calibration relationship between pixel area and actual size. Combined with the length, width and height of the pond, the current fish density is obtained by dividing the number of fish by the water volume; the system sets corresponding biological parameters such as average weight of individual fish, basal oxygen consumption rate and ammonia nitrogen production rate according to the current fish species being farmed; the above water quality monitoring data, fish density values ​​and fish biological parameters are combined with sampling intervals of 10 minutes to construct a three-dimensional historical dataset containing time series, environmental conditions and biological characteristics.

[0054] It should be noted that in this embodiment, the system employs multi-level image processing technology to achieve accurate fish identification and counting, specifically including the following steps:

[0055] The first step is image preprocessing. The system preprocesses and optimizes the raw images captured by the underwater camera: First, it uses the Limiting Contrast Adaptive Histogram Equalization (CLAHE) algorithm to enhance the local contrast of the image. In this embodiment, the limitation factor of the CLAHE algorithm is set to 3.0, and the grid size is set to 8×8 pixels. Then, a bilateral filter is used to remove noise while preserving edge information. The filter window size is set to 9×9, the color space standard deviation is set to 75, and the coordinate space standard deviation is set to 75. Finally, gamma correction is used to adjust the image brightness. The gamma value is dynamically set between 0.7 and 1.3 according to the water depth and lighting conditions.

[0056] The second step is fish detection and localization. A YOLO-v5 model trained on an underwater fish dataset is used for target detection. The model takes a pre-processed image of 416×416 pixels as input and outputs the bounding box coordinates (x, y, w, h), confidence score, and category information for each fish. The system sets the confidence threshold to 0.5 and the non-maximum suppression (NMS) threshold to 0.4. Post-processing is performed on the detection results to remove candidate boxes with an area less than 100 pixels or abnormal aspect ratios (less than 0.3 or greater than 3.0).

[0057] The third step is fish contour extraction. Since YOLO-v5 outputs rectangular bounding boxes that cannot recognize fish contours, this invention extracts the fish contour within each valid bounding box. First, the bounding box region is converted to a grayscale image, and then the Canny edge detection algorithm is used to extract edge information. In this embodiment, the low threshold of the Canny algorithm is set to 50, the high threshold is set to 150, and the Sobel operator size is 3×3. Morphological closure operations are used to connect broken edges, with a 5×5 elliptical kernel as the structuring element. The cv2.findContours() function is used to extract the contour, the RETR_EXTERNAL mode is used to extract the outer contour, and the CHAIN_APPROX_SIMPLE method is used to compress the contour points. Contours with an area between 50 and 5000 pixels are retained as valid fish contours through contour area filtering.

[0058] The fourth step is the separation of overlapping fish bodies. To address the problem of inaccurate counting caused by overlapping fish distribution, this invention employs a separation algorithm based on geometric feature analysis to handle overlapping fish distribution: calculating the pixel area S, pixel perimeter P, convex hull area (pixel area) CH, and boundary rectangle area BR for each contour; defining the contour complexity index C = P 2 / (4πS), when C is greater than 1.5, it is determined to be a possible overlapping area; calculate the convexity value H=S / CH, when H is less than 0.75, it is confirmed to be overlapping; for the confirmed overlapping area, use the cv2.fitEllipse() function to perform ellipse fitting, calculate the fitting error E=|S-πab| / S (where a and b are the major and minor semi-axes of the ellipse), when E is less than 0.3, it is determined to be a single fish, when E is greater than or equal to 0.3, it is determined to be multiple fish overlapping; for overlapping cases, estimate the number of fish based on the ratio of the outline area to the standard single fish area, the calculation formula is N=round(S / S st ), where S st This refers to the standard area for a single fish in the current aquaculture stage.

[0059] The fifth step is multi-frame fusion statistics. To improve counting accuracy, the system performs correlation analysis on the detection results of five consecutive frames: using a centroid-based tracking algorithm, fish are identified as the same entity when the centroid distance between two frames is less than 30 pixels; a sliding window voting mechanism is adopted, using the result that appears most frequently within five frames as the final count value; and outlier detection is set up, excluding the result of a frame if the count result of a certain frame deviates from the average of the previous four frames by more than 20%.

[0060] It should be noted that the above-mentioned multi-level image processing techniques are all implemented based on standard functions of the OpenCV computer vision library, including mature algorithms such as histogram equalization, bilateral filtering, Canny edge detection, morphological operations, contour extraction, ellipse fitting, and camera calibration. System parameters can be adjusted and optimized according to the actual aquaculture environment. Based on the correlation data between the rate of change of ammonia nitrogen concentration and the rate of dissolved oxygen consumption in historical datasets, a piecewise linear regression algorithm is used to establish the basic coupling coefficients under different temperature ranges. The basic coupling coefficients are then corrected using fish-specific correction factors and density correction factors to establish a coupling prediction model.

[0061] Specifically, in this embodiment, the system extracts data pairs of ammonia nitrogen concentration change rate (ΔNH3-N / Δt) and dissolved oxygen consumption rate (Δdissolved oxygen / Δt) from historical datasets. Based on water temperature range, the data is divided into four temperature intervals: 15-20℃, 20-25℃, 25-30℃, and 30-35℃. For each temperature interval, piecewise linear regression is performed using the least squares method to establish a linear relationship of Δdissolved oxygen / Δt = α × ΔNH3-N / Δt + β, where α is the basic coupling coefficient for that temperature interval, and β is the intercept term. The fish species-specific correction factor is obtained by calculating the ratio of the actual oxygen-nitrogen ratio of the current fish species to the oxygen-nitrogen ratio of the standard fish species. The actual oxygen-to-nitrogen ratio is equal to the oxygen consumption of the fish per unit time divided by the ammonia nitrogen production. The standard oxygen-to-nitrogen ratio for fish species is set at 4.57 (based on the physiological and metabolic data of carp). The density correction factor is calculated using an exponential function, with the formula e^(0.1×(ρ-ρ0) / ρ0), where ρ is the current fish density and ρ0 is the recommended stocking density. When the density exceeds the recommended value, the correction factor is greater than 1, indicating that the impact of ammonia nitrogen on dissolved oxygen is aggravated under high-density conditions. The final coupling coefficient is equal to the basic coupling coefficient multiplied by the fish species-specific correction factor and then multiplied by the density correction factor. This coupling coefficient quantifies the intensity of the impact of changes in ammonia nitrogen concentration on dissolved oxygen consumption under specific stocking conditions.

[0062] Specifically, the expression for the coupled prediction model is:

[0063] R DO =K(T,pH)×[α×dC NH4 / dt+β×ρ f +γ×F s];

[0064]

[0065] In the formula, R DO dC represents the dissolved oxygen consumption rate; K(T,pH) is the temperature-pH combined correction coefficient; α, β, and γ are coupling coefficients obtained through regression from historical data; dC NH4 / dt represents the rate of change in ammonia nitrogen concentration; ρ f F is the fish density correction factor. s Fish species-specific correction factor; C DO (t) represents the dissolved oxygen concentration at time t; R photo Δt represents the photosynthetic oxygenation rate; Δt represents the prediction time window.

[0066] It should be noted that α is the dominant coupling coefficient of ammonia nitrogen change rate, reflecting the degree of direct impact of ammonia nitrogen concentration change on dissolved oxygen consumption. It is obtained by analyzing the synchronous change relationship between ammonia nitrogen concentration increase and dissolved oxygen decrease in historical data and using least squares regression. β is the density-load coupling coefficient, reflecting the influence of fish density on oxygen consumption intensity. The higher the density, the more severe the dissolved oxygen consumption caused by a unit change in ammonia nitrogen. It is determined by regression analysis of comparative data under different density conditions. γ is the fish species-specific coupling coefficient. Considering the metabolic differences and environmental sensitivity of different fish species, grass carp, common carp, crucian carp and other different fish species have different γ values. It is determined separately by regression analysis of specific aquaculture data for each fish species.

[0067] Specifically, in this embodiment, the system uses a first-order autoregressive model to predict the evolution curve of ammonia nitrogen concentration over the next two hours based on the current ammonia nitrogen concentration and the ammonia nitrogen change trend over the previous 30 minutes. The predicted ammonia nitrogen concentration change rate is then substituted into the coupled prediction model, and the expected dissolved oxygen consumption is calculated using the formula: Expected Dissolved Oxygen Consumption = Coupling Coefficient × Ammonia Nitrogen Concentration Change. Simultaneously, considering the oxygenation contribution of algal photosynthesis, the photosynthetic oxygenation rate is estimated based on the current light intensity and chlorophyll content in the water. The predicted dissolved oxygen value is then calculated using the formula: Predicted Dissolved Oxygen Value = Current Dissolved Oxygen Concentration - Expected Dissolved Oxygen Consumption + Photosynthetic Oxygenation. The lethal dissolved oxygen concentration of fish is used as the absolute lower limit, and a safety margin of 1.5 mg / L is added to this as the warning threshold. The prediction confidence interval is determined by calculating the standard deviation of historical prediction errors, with the upper and lower boundaries being the predicted value ± 1.96 × the prediction standard deviation, respectively. When the lower confidence interval of the predicted dissolved oxygen concentration is lower than the warning threshold, the system generates a graded warning signal based on the expected decrease. A decrease of 0.5-1.0 mg / L triggers a level one warning and initiates intermittent aeration, while a decrease of more than 1.0 mg / L triggers a level two warning and initiates continuous enhanced aeration, thus realizing a dissolved oxygen warning control strategy based on an ammonia nitrogen-driven mechanism.

[0068] In one embodiment, based on the current fish density and daily feed intake, combined with the fish's basal metabolic parameters and tolerance data, environmental load parameters and dual-parameter baseline values ​​are calculated to determine the critical slope for the increase in ammonia nitrogen concentration, including:

[0069] Based on the current fish density and daily feeding data, combined with the basal metabolic parameters of different fish species, the basal load values ​​of ammonia nitrogen production and dissolved oxygen consumption in the water body are calculated, and then corrected by the current water temperature and pH conditions to obtain the environmental load parameters.

[0070] Specifically, in this embodiment, the system uses image data collected by an underwater camera to count the current fish density using a fish counting algorithm. Combined with the daily feeding data recorded by the automatic feeder, the system reads the basic metabolic parameters of the current farmed fish species from the storage module, including the daily ammonia nitrogen production coefficient and dissolved oxygen consumption coefficient per individual fish. The system calculates the basic ammonia nitrogen load value in the water body according to the formula: ammonia nitrogen production equals fish density multiplied by average fish weight multiplied by ammonia nitrogen production coefficient. The system also calculates the basic dissolved oxygen load value according to the formula: dissolved oxygen consumption equals fish density multiplied by average fish weight multiplied by dissolved oxygen consumption coefficient. At the same time, temperature correction coefficient and pH correction coefficient are introduced to correct the basic load value for environmental conditions. The temperature correction coefficient is calculated using the Q10 temperature coefficient method as 2^[(current water temperature - 20) / 10]. The pH correction coefficient is calculated using the Henderson-Hasselbalch equation based on the dissociation equilibrium of ammonia nitrogen under different pH conditions. Finally, the environmental load parameter is equal to the basic load value multiplied by the temperature correction coefficient and then by the pH correction coefficient.

[0071] Based on the tolerance data of fish at different growth stages, the current stage is determined by the current number of days of culture and the size of the fish. The lower limit of dissolved oxygen requirement and the upper limit of ammonia nitrogen tolerance for this stage are obtained. The values ​​are then adjusted in combination with environmental load parameters to generate a dual-parameter baseline value under the current culture conditions.

[0072] Specifically, in this embodiment, the system calculates the number of days of rearing based on the start time and current time, and combines this with regularly measured fish size data (body length, weight) to determine the current growth stage through fish growth stage classification standards. The growth stages include fry stage (0-30 days), juvenile stage (31-90 days), adult stage (91-180 days), and marketable stage (over 181 days). The system reads tolerance parameters for different growth stages from the storage module, including the minimum dissolved oxygen requirement and maximum ammonia nitrogen tolerance concentration for that stage. Based on the tolerance parameters of the current growth stage, the system sets initial values ​​for the lower limit of dissolved oxygen requirement and the upper limit of ammonia nitrogen tolerance. Then, it dynamically adjusts these values ​​based on the environmental load parameters calculated above. The adjustment formula is: the adjusted value of the lower limit of dissolved oxygen requirement equals the initial lower limit multiplied by (1 + environmental load parameter / standard load parameter), and the adjusted value of the upper limit of ammonia nitrogen tolerance equals the initial upper limit divided by (1 + environmental load parameter / standard load parameter). The standard load parameter is the load benchmark value under recommended rearing density conditions. Finally, a dual-parameter benchmark value adapted to the current rearing environment and fish growth status is generated.

[0073] Based on the dual-parameter benchmark values, a safe threshold for dissolved oxygen and ammonia nitrogen warning threshold are set, and the critical slope for the increase in ammonia nitrogen concentration is calculated.

[0074] Specifically, in this embodiment, the system sets the control target for water quality management based on dual-parameter benchmark values. A safety margin of 1.0 mg / L is added to the lower limit adjustment value of dissolved oxygen demand as the dissolved oxygen safety threshold, and a safety margin of 20% is reduced to the upper limit adjustment value of ammonia nitrogen tolerance as the ammonia nitrogen warning threshold. By analyzing historical ammonia nitrogen concentration change data, the system calculates the time window required for the ammonia nitrogen concentration to rise from the current value to the warning threshold. Combining the correlation between dissolved oxygen and ammonia nitrogen in the aforementioned coupled prediction model, the system determines the corresponding ammonia nitrogen change rate at which the dissolved oxygen concentration drops to the safety threshold within this time window. The critical slope for the rise in ammonia nitrogen concentration is calculated as (ammonia nitrogen warning threshold - current ammonia nitrogen concentration) divided by the prediction time window. The prediction time window is dynamically determined based on the current environmental load parameters and historical change trends, typically set to 30–120 minutes. When the real-time monitored rate of change in ammonia nitrogen concentration exceeds the critical slope, the system determines that there is a risk of insufficient dissolved oxygen and immediately initiates a warning control strategy, including early activation of oxygenation equipment and adjustment of the feeding plan.

[0075] In one embodiment, the feeding control module calculates the feeding amount adjustment coefficient based on the percentage of fish in the feeding area according to the fish activity image data, combined with the historical feeding completion time and the nutritional needs of the current breeding stage, and performs batch feeding control according to the preset feeding interval.

[0076] Specifically, the system uses an underwater camera to capture high-definition images of the feeding area every 5 minutes, employs the YOLO-v5 algorithm to identify individual fish in the images and count their numbers, calculating the proportion of fish in the feeding area to the total fish population; it reads historical feeding completion time data from the storage module for the past 30 days, uses a moving average algorithm to calculate the average feeding completion time under different feeding amounts, and establishes a correlation curve between feeding amount and feeding time; it determines the growth stage of the fish based on the current number of days of rearing, and obtains the daily nutrient requirements and ideal feeding frequency for that stage from a pre-set nutrient requirement database; and then... The formula for calculating the feeding adjustment coefficient is: the adjustment coefficient equals (the proportion of fish in the feeding area × the historical feeding efficiency coefficient × the stage nutritional requirement coefficient), where the feeding efficiency coefficient is determined based on the ratio of the historical feeding completion time to the standard time, and the stage nutritional requirement coefficient is set based on the metabolic level of the current growth stage; the system performs batch feeding according to the preset feeding interval, and the amount of each feeding is equal to the total daily feeding amount divided by the number of feedings and then multiplied by the adjustment coefficient. The feeding situation is monitored in real time during the feeding process. When the feeding completion time of two consecutive feedings exceeds 1.5 times the expected time, the amount of the next feeding is automatically reduced by 10%.

[0077] In one embodiment, the environmental control module generates graded control commands based on environmental parameters monitored by environmental sensors. When water quality parameters continuously exceed standards or fish behavior is abnormal, the module adjusts the operating status of the actuators in real time, including:

[0078] When the water temperature deviates from the target range by more than the preset temperature threshold, the heating or cooling device will be activated proportionally based on the difference between the ambient temperature and the target temperature.

[0079] Specifically, in this embodiment, the system collects water temperature data every minute using a water temperature sensor and compares the current water temperature with the target temperature range set according to the fish species and growth stage. When the water temperature deviates from the target range by more than a preset temperature threshold (set to ±1.5℃), the system calculates the temperature deviation value as the current water temperature minus the median of the target temperature range. The heating or cooling requirement is determined based on the magnitude and direction of the temperature deviation value. When the deviation is positive, the cooling device is activated; when the deviation is negative, the heating device is activated. The device's starting power is calculated using a proportional control algorithm, where the power percentage equals the absolute value of the temperature deviation divided by the maximum allowable deviation value and then multiplied by 100%. The maximum power is limited to 80% of the device's rated power. The system also monitors the ambient temperature change trend. When the ambient temperature change rate exceeds 0.5%, the temperature control device is activated in advance for pre-adjustment to prevent large fluctuations in water temperature from causing stress to the fish.

[0080] When the fluctuation range of water quality parameters exceeds the preset fluctuation threshold and abnormal fish behavior is detected, the ultraviolet sterilization lamp is activated to run for the preset disinfection time, and the circulating water valve is adjusted to the preset opening degree.

[0081] When abnormal operation of the actuator or extreme weather is detected, graded response control is implemented according to the fault type and abnormality level, and an early warning signal is sent.

[0082] In one embodiment, when the change in water quality parameters exceeds a preset fluctuation threshold and abnormal fish behavior is determined, activating the ultraviolet sterilization lamp for a preset disinfection time and adjusting the circulating water valve to a preset opening degree includes:

[0083] Data on dissolved oxygen, pH, ammonia nitrogen concentration, and turbidity are collected periodically. When the change of any parameter within a preset time window exceeds a preset multiple of the daily fluctuation benchmark value, it is marked as a risk parameter and recorded.

[0084] The underwater camera captures images of fish activity at regular intervals, and the number and distribution of individual fish in the images are counted. When the proportion of fish gathering exceeds the abnormal gathering threshold and the duration exceeds the preset duration, or when the area of ​​abnormal features on the body surface of a single fish exceeds the preset proportion, it is determined to be abnormal fish behavior.

[0085] When the number of risk parameters reaches the preset number and is determined to be abnormal fish behavior, the ultraviolet sterilization lamp is activated to run at the preset power for the preset disinfection time, and the opening of the circulating water valve is adjusted to the preset purification flow level.

[0086] A differentiated treatment mechanism should be established based on the number of risk parameters and the severity of abnormal fish behavior.

[0087] Specifically, in this embodiment, the system collects dissolved oxygen, pH, ammonia nitrogen concentration, and turbidity data every 5 minutes using a multi-parameter water quality sensor. Based on historical data from the past 72 hours, it calculates the daily fluctuation baseline value for each parameter. When the change of any parameter within a 30-minute time window exceeds twice its daily fluctuation baseline value, the parameter is marked as a risk parameter and a timestamp is recorded. Simultaneously, an underwater camera captures images of fish activity every 10 minutes. A deep learning algorithm is used to statistically analyze the number and spatial distribution of individual fish in the images, calculating the aggregation ratio of fish in a specific area of ​​the water. When the aggregation ratio exceeds 60% and lasts for more than 30 minutes, or when image analysis detects abnormal features such as white spots or ulcers on the body surface of a single fish covering more than 5% of its total body surface area, the system determines it as abnormal fish behavior. When the number of risk parameters reaches two or more and abnormal fish behavior is simultaneously determined, the system considers the fish behavior as abnormal. When an abnormal behavior occurs, the system initiates a differentiated treatment mechanism based on the number of risk parameters and the severity of the abnormal behavior. This includes a Level 1 response (2 risk parameters, minor abnormal behavior): the UV sterilizer is activated at 40% power for 45 minutes, and the circulating water valve opening is adjusted to 1.3 times the normal flow rate. A Level 2 response (3-4 risk parameters, moderate abnormal behavior): the UV sterilizer is activated at 70% power for 90 minutes, and the circulating water valve opening is adjusted to 1.6 times the normal flow rate. A Level 3 response (5 or more risk parameters, severe abnormal behavior): the UV sterilizer is activated at 100% power for 120 minutes, and the circulating water valve opening is adjusted to the maximum purification flow rate. During the treatment process, the system continuously monitors changes in water quality parameters and fish behavior. When three consecutive tests show a decrease in the number of risk parameters and a return to normal fish behavior, the treatment intensity is gradually reduced until the system returns to normal operation.

[0088] In one embodiment, when an actuator malfunction or extreme weather is detected, graded response control is performed based on the fault type and anomaly level, and an early warning signal is sent, including:

[0089] The operating parameters of the actuator are monitored periodically. When the operating parameters deviate from the rated value by more than the preset deviation range or the response time exceeds the preset duration, it is determined that the equipment is operating abnormally and the fault type is recorded.

[0090] When the rate of change of air pressure, the rate of change of water temperature, or the magnitude of change of light intensity exceeds the corresponding extreme weather threshold, it is judged as extreme weather and classified into multiple abnormal levels according to the magnitude of change.

[0091] Based on the type of equipment failure and the severity of extreme weather, a tiered response is implemented, reducing equipment power to a safe operating level and sending early warning signals.

[0092] Specifically, in this embodiment, the system monitors the operating parameters of actuators such as the aerator pump, heating device, cooling device, automatic feeder, circulating water valve, and ultraviolet germicidal lamp every 3 minutes, including power consumption, speed, temperature, and response time. When any operating parameter deviates from the rated value by more than ±15% or the command response time exceeds the preset duration, the system determines that the equipment is malfunctioning and records the fault type. Simultaneously, the system monitors the rate of change of air pressure (exceeding 5 hPa / hour is considered abnormal), the rate of change of water temperature (exceeding 1.5℃ / hour is considered abnormal), and the amplitude of change of light intensity (exceeding 3 times the normal daily variation is considered abnormal) through environmental sensors. When the above parameters exceed the corresponding extreme weather threshold, the system determines it as an extreme weather condition and classifies extreme weather into mild (exceeding 1-1.5 times the threshold) and moderate (exceeding 1.5-2 times the threshold) based on the magnitude of the change. The system classifies extreme weather into three levels of anomalies: severe (more than twice the threshold) and critical (more than twice the threshold). Based on the type of equipment failure and the severity of the extreme weather, the system executes a four-level response: Level 0 (minor failure or mild extreme weather) maintains normal operation and records the anomaly; Level 1 (moderate failure of a single device or moderate extreme weather) reduces the power of the relevant equipment to 70% of the safe operating level and sends a low-priority warning signal; Level 2 (moderate failure of multiple devices or severe extreme weather) reduces the power of all devices to 50% of the safe operating level and sends a medium-priority warning signal; Level 3 (severe failure of any device or multiple extreme weather indicators) shuts down non-essential equipment and reduces the power of essential equipment to the minimum safe operating level of 30%, while simultaneously sending a high-priority warning signal and activating backup power. Warning signals are pushed in real-time to the cloud control platform and user control terminals via the network communication unit.

[0093] like Figure 3 As shown, according to another embodiment of the present invention, an intelligent fish farming automatic control method based on the Internet of Things is also provided, which includes:

[0094] Using multi-parameter water quality sensors and environmental sensors, aquaculture environment data is collected in real time, including water quality parameters, environmental parameters, and fish behavior information;

[0095] Based on aquaculture environment data, a multi-parameter correlation control algorithm and early warning control strategy are executed, and control commands are executed to adjust aquaculture environment parameters;

[0096] Establish a data communication connection between the local machine and the cloud to transmit control commands and monitoring data;

[0097] It receives remote monitoring requests and responds to manual intervention instructions.

[0098] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. 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. An intelligent fish farming automatic control system based on the Internet of Things, characterized in that, include: The data sensing unit is used to collect aquaculture environment data in real time using multi-parameter water quality sensors and environmental sensors. The aquaculture environment data includes water quality parameters, environmental parameters, and fish behavior information. The intelligent control unit is used to execute multi-parameter correlation control algorithms and early warning control strategies based on aquaculture environment data, and to execute control commands to adjust aquaculture environment parameters; The network communication unit is used to establish a data communication connection between the local area and the cloud, and to transmit control commands and monitoring data. The remote application unit is used to receive remote monitoring requests and respond to manual intervention instructions; The intelligent control unit includes a main control module for establishing an ammonia nitrogen-driven dissolved oxygen early warning control strategy based on water quality parameters and generating graded treatment instructions in combination with environmental parameters; an actuator for adjusting aquaculture environmental parameters according to control instructions; and a storage module for storing the control strategy.

2. The intelligent fish farming automatic control system based on the Internet of Things according to claim 1, characterized in that, The data sensing unit includes a multi-parameter water quality sensor for monitoring water quality parameters, an underwater camera for monitoring fish density and feeding behavior, and an environmental sensor for monitoring environmental parameters. The network communication unit includes a wireless communication module for local control data transmission; the remote application unit includes a cloud control platform for remote control and a user control terminal for manual control intervention.

3. The intelligent fish farming automatic control system based on the Internet of Things according to claim 1, characterized in that, The main control module includes: The water quality early warning module is used to establish a coupled prediction model based on water quality parameters of the aquaculture environment, combined with fish species information and density monitoring data, and to calculate environmental load parameters based on fish density and tolerance data, determine the critical slope of ammonia nitrogen concentration increase, so as to establish an ammonia nitrogen-driven dissolved oxygen early warning and control strategy. The feeding control module is used to calculate the proportion of fish in the feeding area based on fish activity image data, combine the historical feeding completion time with the nutritional needs of the current breeding stage, calculate the feeding amount adjustment coefficient, and execute batch feeding control according to the preset feeding interval. The environmental control module is used to generate graded control commands based on environmental parameters monitored by environmental sensors. When water quality parameters continuously exceed the standard or fish behavior is abnormal, the module adjusts the operating status of the actuators in real time.

4. The intelligent fish farming automatic control system based on the Internet of Things according to claim 3, characterized in that, Based on aquaculture environment water quality parameters, combined with fish species information and density monitoring data, a coupled prediction model was established, including: Based on real-time data of dissolved oxygen concentration, ammonia nitrogen concentration, water temperature and pH value collected by multi-parameter water quality sensors, time series alignment is used to eliminate sampling delay between sensors, a water quality parameter matrix is ​​established, and historical datasets are generated by combining information on farmed fish species and density monitoring data. Based on the correlation data between the rate of change of ammonia nitrogen concentration and the rate of dissolved oxygen consumption in historical datasets, a piecewise linear regression algorithm is used to establish the basic coupling coefficients for different temperature ranges. A coupling prediction model is established by correcting the basic coupling coefficient using fish species-specific correction factors and density correction factors.

5. The intelligent fish farming automatic control system based on the Internet of Things according to claim 3, characterized in that, Environmental load parameters were calculated based on fish density and tolerance data to determine the critical slope for the rise in ammonia nitrogen concentration, in order to establish an ammonia nitrogen-driven dissolved oxygen early warning and control strategy, including: Based on the current ammonia nitrogen concentration and the coupled prediction model, the expected decrease in dissolved oxygen within the future time window is calculated to obtain the predicted value of dissolved oxygen concentration, and the prediction confidence interval is set in combination with the fish species' tolerance. Based on the current fish density and daily feeding amount, combined with the fish's basal metabolic parameters and tolerance data, the environmental load parameters and dual-parameter baseline values ​​are calculated to determine the critical slope for the increase in ammonia nitrogen concentration. When the rate of increase of ammonia nitrogen concentration exceeds the critical slope, the expected decrease in dissolved oxygen is calculated using a coupled prediction model, and the start-up timing and operating power of the oxygenation pump are set according to the expected decrease.

6. The intelligent fish farming automatic control system based on the Internet of Things according to claim 5, characterized in that, Based on the current fish density and daily feed intake, combined with the fish's basal metabolic parameters and tolerance data, environmental load parameters and dual-parameter baseline values ​​were calculated to determine the critical slope for the increase in ammonia nitrogen concentration, including: Based on the current fish density and daily feeding data, combined with the basal metabolic parameters of different fish species, the basal load values ​​of ammonia nitrogen production and dissolved oxygen consumption in the water body are calculated, and then corrected by the current water temperature and pH conditions to obtain the environmental load parameters. Based on the tolerance data of fish at different growth stages, the current stage is determined by the current number of days of culture and the size of the fish. The lower limit of dissolved oxygen requirement and the upper limit of ammonia nitrogen tolerance for this stage are obtained. The values ​​are then adjusted in combination with environmental load parameters to generate a dual-parameter baseline value under the current culture conditions. Based on the dual-parameter benchmark values, a safe threshold for dissolved oxygen and ammonia nitrogen warning threshold are set, and the critical slope for the increase in ammonia nitrogen concentration is calculated.

7. The intelligent fish farming automatic control system based on the Internet of Things according to claim 3, characterized in that, Based on environmental parameters monitored by environmental sensors, when water quality parameters continuously exceed standards or fish behavior becomes abnormal, graded control commands are generated to adjust the operating status of the actuators in real time, including: When the water temperature deviates from the target range by more than the preset temperature threshold, the heating or cooling device will be activated proportionally based on the difference between the ambient temperature and the target temperature. When the fluctuation range of water quality parameters exceeds the preset fluctuation threshold and abnormal fish behavior is detected, the ultraviolet sterilization lamp is activated to run for the preset disinfection time, and the circulating water valve is adjusted to the preset opening degree. When abnormal operation of the actuator or extreme weather is detected, graded response control is implemented according to the fault type and abnormality level, and an early warning signal is sent.

8. The intelligent fish farming automatic control system based on the Internet of Things according to claim 7, characterized in that, When the fluctuation range of water quality parameters exceeds the preset fluctuation threshold, and abnormal fish behavior is detected, the ultraviolet sterilization lamp is activated for the preset disinfection time, and the circulating water valve is adjusted to the preset opening degree, including: Data on dissolved oxygen, pH, ammonia nitrogen concentration, and turbidity are collected periodically. When the change of any parameter within a preset time window exceeds a preset multiple of the daily fluctuation benchmark value, it is marked as a risk parameter and recorded. The underwater camera captures images of fish activity at regular intervals, and the number and distribution of individual fish in the images are counted. When the proportion of fish gathering exceeds the abnormal gathering threshold and the duration exceeds the preset duration, or when the area of ​​abnormal features on the body surface of a single fish exceeds the preset proportion, it is determined to be abnormal fish behavior. When the number of risk parameters reaches the preset number and is determined to be abnormal fish behavior, the ultraviolet sterilization lamp is activated to run at the preset power for the preset disinfection time, and the opening of the circulating water valve is adjusted to the preset purification flow level. A differentiated treatment mechanism should be established based on the number of risk parameters and the severity of abnormal fish behavior.

9. The intelligent fish farming automatic control system based on the Internet of Things according to claim 7, characterized in that, When an actuator malfunction or extreme weather is detected, a graded response control is implemented based on the fault type and anomaly level, and warning signals are sent, including: The operating parameters of the actuator are monitored periodically. When the operating parameters deviate from the rated value by more than the preset deviation range or the response time exceeds the preset duration, it is determined that the equipment is operating abnormally and the fault type is recorded. When the rate of change of air pressure, the rate of change of water temperature, or the magnitude of change of light intensity exceeds the corresponding extreme weather threshold, it is judged as extreme weather and classified into multiple abnormal levels according to the magnitude of change. Based on the type of equipment failure and the severity of extreme weather, a tiered response is implemented, reducing equipment power to a safe operating level and sending early warning signals.

10. An IoT-based intelligent fish farming automatic control method, employing any one of the IoT-based intelligent fish farming automatic control systems described in claims 1-9, characterized in that, The method includes: Using multi-parameter water quality sensors and environmental sensors, aquaculture environment data is collected in real time, including water quality parameters, environmental parameters, and fish behavior information; Based on aquaculture environment data, a multi-parameter correlation control algorithm and early warning control strategy are executed, and control commands are executed to adjust aquaculture environment parameters; Establish a data communication connection between the local machine and the cloud to transmit control commands and monitoring data; It receives remote monitoring requests and responds to manual intervention instructions.

Images