Intelligent disease early warning and precise prevention and control system and method for aquaculture

Abstract

The invention discloses an intelligent disease early warning and precise prevention and control system and method for aquaculture, and relates to the technical field of aquaculture. Comprising an underwater robot module, the underwater robot module is based on a sensor carried by an underwater robot, underwater information collection is achieved, and fodder feeding is achieved through an integrated fodder feeding device; the route planning module is used for planning an operation route of the underwater robot; the aquatic product monitoring and analyzing module is used for analyzing the aquaculture condition and the water quality condition of aquatic products according to the information collected by the underwater robot; and the disease early warning module is used for carrying out disease early warning based on the analysis result of the aquatic product monitoring and analysis module. According to the invention, through precise feeding route planning of the underwater robot based on coordinates, feed can be accurately fed to an area where aquatic animals are located; the traditional feeding mode easily causes feed dispersion and waste, but the system can greatly improve the feed utilization rate; the accurate feeding of the feed helps the aquatic animals to better ingest nutrition and promote the growth of the aquatic animals.

Description

Technical Field

[0001] The present invention relates to the technical field of aquaculture, and particularly to an intelligent disease early warning and precise prevention and control system and method for aquaculture. Background Art

[0002] With the continuous growth of the global population and the rapid development of the economy, the demand for aquatic products by people has been increasing continuously. Aquaculture, as an important way to provide aquatic products, is playing an increasingly crucial role in ensuring the food supply for humans. However, the traditional aquaculture model faces many severe challenges.

[0003] In the feeding link, the commonly used broadcast or manual fixed-point feeding methods have significant drawbacks. Broadcast feeding is difficult to ensure uniform distribution of feed, and a large amount of feed is scattered in non-aquatic animal activity areas, not only causing great waste of feed, but also easily polluting the water environment. Manual fixed-point feeding depends on the experience of workers and is difficult to accurately grasp the feeding position and quantity, resulting in insufficient feeding for some aquatic animals and excessive feed in other areas.

[0004] In terms of aquaculture environment monitoring, most farms still rely on manual sampling at regular intervals and empirical judgment to evaluate the water quality. This method has obvious lag and inaccuracy, and cannot reflect the dynamic changes of water quality in real time. For example, when the water quality is found to deteriorate, it often has caused certain stress or damage to aquatic animals. Moreover, manual monitoring is difficult to cover a large area of aquaculture waters comprehensively, and some local environmental changes cannot be detected in time, bringing potential risks to the healthy growth of aquatic animals.

[0005] In the field of disease prevention and control, traditional disease diagnosis mainly relies on visual observation of the external symptoms of aquatic animals and simple laboratory tests. This method is difficult to accurately diagnose some early or latent diseases, and often misses the best treatment opportunity. Once a disease breaks out, due to the lack of precise prevention and control means and effective drug guidance, farmers usually adopt a large-dose, broad-spectrum drug treatment method. This not only easily leads to excessive drug residues, affecting the quality and safety of aquatic products, but also may cause drug resistance problems due to drug abuse, making subsequent disease prevention and control more difficult. In addition, traditional prevention and control decisions are mainly based on personal experience and limited information, lacking scientificity and systematicness, and it is difficult to formulate the most optimized prevention and control plan. Summary of the Invention

[0006] The purpose of the present invention is to solve the deficiencies existing in the prior art, and to propose an intelligent disease early warning and precise prevention and control system and method for aquaculture.

[0007] To achieve the above purpose, the present invention adopts the following technical solutions: An intelligent disease early warning and precise prevention and control system for aquaculture, comprising: Underwater robot module. Based on the sensors carried by the underwater robot, the underwater robot module realizes the collection of underwater information and the feeding of feed through the integrated feeding device. Route planning module, used to plan the operation route of the underwater robot. Aquaculture monitoring and analysis module, which analyzes the aquaculture situation and water quality situation according to the information collected by the underwater robot. Disease warning module, which conducts disease warnings based on the analysis results of the aquaculture monitoring and analysis module. Prevention and control decision-making module, which generates a prevention and control decision-making plan according to the disease warning information of the disease warning module.

[0008] Preferably: The underwater robot is equipped with multiple thrusters, which are distributed in different directions of the robot, and the flexible steering and precise positioning of the robot are realized through the vector control algorithm. Let the coordinates of the robot in the three-dimensional space be (x, y, z), where the x-axis direction is the length direction of the aquaculture farm, the y-axis direction is the width direction, and the z-axis direction is the water depth direction. The motion speed vector of the robot is expressed as: where , , are the speed components of the robot in the x, y, and z directions respectively; by precisely controlling the rotation speed of each thruster, the precise control of the robot's motion speed vector is realized, and then the robot can move to any position in the aquaculture farm.

[0009] Preferably: The sensor system of the underwater robot module includes: Vision sensor: Installed at the front end of the robot, used to collect real-time image information in the aquaculture farm; by processing and analyzing consecutive image frames, the position, size, and posture information of aquatic animals are obtained. Water quality sensor: Distributed on the surface of the robot, used to monitor water quality parameters such as water temperature, dissolved oxygen content, and pH value in real time. Acoustic sensor: Used to detect the sound signals emitted by aquatic animals. Aquatic animals of different species and health conditions will emit sound signals within a specific frequency range.

[0010] Preferably: Before conducting route planning, the route planning module establishes a three-dimensional space model of the aquaculture farm; divides the aquaculture farm into multiple grid areas, and the size of each grid is Δx × Δy × Δz, and its vertex coordinates are , where ; ; Measure and record environmental factors such as water depth, water flow velocity, and sediment type in each grid area through sensors carried by the underwater robot, and construct an environmental factor matrix E: Among them, represents the comprehensive environmental factor value of the th grid area, and its calculation method is: In the formula, is the water depth factor is the water flow velocity factor, is the sediment factor, , , are the weight coefficients of the corresponding factors.

[0011] Preferably: The route planning module comprehensively considers the environmental factors of the aquaculture farm, the distribution and behavior trends of aquatic animals, and aims to minimize the feeding time, maximize the feed utilization rate and the feeding uniformity of aquatic animals, and establishes an optimized feeding route model; Let the sequence of grid areas passed by the feeding route be , where K is the total number of grid areas on the feeding route; then the feeding route optimization problem is expressed as: Among them, is the total feeding time, is the feeding uniformity index of aquatic animals, represents the distance between adjacent grid areas and , is the average moving speed of the robot, is the maximum distance limit for a single movement of the robot, is the total amount of feed for each feeding; By solving this optimization problem, the optimal feeding route is obtained.

[0012] Preferably: The underwater robot moves in the aquaculture farm according to the route determined by the feeding route planning module, and real-time collects images, sounds and water quality data of aquatic animals; For image data, first perform grayscale processing to convert the color image into a grayscale image; Then use an image enhancement algorithm to enhance the contrast of the image and highlight the characteristics of aquatic animals; Use a filtering algorithm to remove noise interference in the image; For sound data, perform filtering and noise reduction processing to extract effective sound signal segments; Normalize the collected water quality data and convert it to a unified numerical range.

[0013] Preferably, the aquatic behavior monitoring and analysis module uses the image information collected by the underwater robot's vision sensor to count the distribution density of aquatic animals in each grid area within different time periods. Meanwhile, combined with the behavioral characteristics of aquatic animals, it classifies their behavior patterns; through long-term monitoring and analysis of the distribution and behavior of aquatic animals, it establishes their behavior logic model to predict their distribution and behavior trends at specific times and in specific environments.

[0014] Preferably, the aquatic behavior monitoring and analysis module extracts the shape features, texture features, and color features of aquatic animals from the preprocessed images; combined with the frequency features and intensity features of the sound signals, it constructs a comprehensive feature vector of aquatic animals, and classifies and identifies the comprehensive feature vector through a trained machine learning model to determine the current behavior state of the aquatic animals.

[0015] Preferably, the disease warning module uses machine learning algorithms to perform fusion analysis on the behavior data, water quality data, and historical disease data of aquatic animals; by learning the variable relationships in the training data, it obtains a conditional probability table, and thus calculates the probability of disease occurrence based on the current observed data. When the probability of disease occurrence exceeds the set threshold, it issues a disease warning message.

[0016] Preferably, the prevention and control method of the prevention and control system includes the following steps: S1: The route planning module plans the operation route of the underwater robot. S2: Based on the planned operation route, the underwater robot uses the sensors carried by the underwater robot to collect underwater information, and through the integrated feed delivery device, it realizes the delivery of feed. S3: The aquatic monitoring and analysis module analyzes the aquaculture situation and water quality situation of aquatic products according to the information collected by the underwater robot. S4: The disease warning module issues a disease warning based on the analysis results of the aquatic monitoring and analysis module. S5: The prevention and control decision-making module generates a prevention and control decision-making plan according to the disease warning information of the disease warning module.

[0017] The beneficial effects of the present invention are as follows: 1. Through the accurate feeding route planning based on coordinates by the underwater robot of the present invention, the feed can be accurately delivered to the area where the aquatic animals are located; the traditional feeding method is prone to feed dispersion and waste, while this system can greatly improve the feed utilization rate; accurate feed delivery helps aquatic animals better absorb nutrients, promote growth, and shorten the breeding cycle.

[0018] 2. The present invention monitors and regulates the water quality environment in real time to ensure that aquatic animals always live in a suitable water environment; it can promptly detect and handle environmental problems, avoid stress responses and diseases of aquatic animals caused by water quality deterioration, etc., and reduce the mortality rate.

[0019] 3. The automated data collection, analysis and decision-making functions of the system of the present invention greatly reduce the workload and labor intensity of manual management. Brief Description of the Drawings

[0020] Figure 1 is a framework diagram of an intelligent disease early warning and precise prevention and control system for aquaculture proposed by the present invention; Figure 2 is a flow chart of the prevention and control method of an intelligent disease early warning and precise prevention and control system for aquaculture proposed by the present invention. Detailed Embodiments

[0021] The technical solutions of the present invention will be further described in detail below in conjunction with the specific embodiments.

[0022] Embodiment: An intelligent disease early warning and precise prevention and control system for aquaculture includes: An underwater robot module, which realizes the collection of underwater information based on the sensors carried by the underwater robot, and realizes the feeding of feed through the integrated feed feeding device; among them, the feed feeding device can be a feed storage tank with an electrically controlled output nozzle, etc., and reference can also be made to the prior art, which will not be elaborated here; A route planning module for planning the operation route of the underwater robot; An aquaculture monitoring and analysis module, which analyzes the aquaculture situation and water quality situation of aquatic products according to the information collected by the underwater robot; A disease early warning module, which conducts disease early warning based on the analysis results of the aquaculture monitoring and analysis module; A prevention and control decision-making module, which generates a prevention and control decision-making plan according to the disease early warning information of the disease early warning module.

[0023] Among them, the underwater robot is equipped with multiple thrusters, which are distributed in different directions of the robot, and realizes the flexible turning and precise positioning of the robot through the vector control algorithm; assuming the coordinates of the robot in the three-dimensional space are (x, y, z), where the x-axis direction is the length direction of the aquaculture farm, the y-axis direction is the width direction, and the z-axis direction is the water depth direction, the motion speed vector of the robot can be expressed as: Among them, , , are the velocity components of the robot in the x, y, and z directions respectively; by precisely controlling the rotation speed of each thruster, the motion speed vector of the robot can be realized Precise control is achieved, enabling the robot to move to any position within the farm.

[0024] Among them, the sensor system of the underwater robot module includes: Vision sensor: Installed at the front end of the robot, it is used to collect real-time image information within the farm; by processing and analyzing consecutive image frames, information such as the position, size, and posture of aquatic animals can be obtained; for example, for an image captured at a certain moment, target detection algorithms are used to identify individual aquatic animals in the image, and geometric measurement algorithms are used to calculate the position of the centroid coordinates of each individual in the image. Water quality sensor: Distributed on the surface of the robot, it monitors water quality parameters such as water temperature, dissolved oxygen content, and pH in real time. Acoustic sensor: Used to detect sound signals emitted by aquatic animals. Aquatic animals of different species and health conditions emit sound signals within a specific frequency range; by analyzing parameters such as the frequency characteristics and intensity characteristics of the sound signals, the behavior status and health conditions of aquatic animals can be judged.

[0025] Among them, before route planning, the route planning module establishes a three-dimensional space model of the farm; the farm is divided into multiple grid areas, and the size of each grid is Δx×Δy×Δz, and its vertex coordinates are Among them, ; ; ; Through the sensors carried by the underwater robot, environmental factors such as water depth, water flow velocity, and bottom substrate type in each grid area are measured and recorded to construct an environmental factor matrix E: Among them, represents the comprehensive environmental factor value of the th grid area, and its calculation method is: In the formula, is the water depth factor, is the water flow velocity factor, is the bottom substrate factor, , , are the weight coefficients of the corresponding factors.

[0026] Among them, the route planning module comprehensively considers the farm environmental factors, the distribution and behavior trends of aquatic animals, and establishes a feeding route optimization model with the goal of minimizing the feeding time, maximizing the feed utilization rate, and the feeding uniformity of aquatic animals; let the sequence of grid areas passed by the feeding route be , where K is the total number of grid areas on the feeding route; then the feeding route optimization problem is expressed as: Among them, is the total feeding time, is the feeding uniformity index of aquatic animals (which can be calculated by the matching degree between the feed delivery amount and the distribution density of aquatic animals), represents the adjacent grid areas and the distance between them, is the average moving speed of the robot, is the maximum distance limit for a single movement of the robot, is the total feed amount for each feeding; by solving this optimization problem, the optimal feeding route is obtained.

[0027] Among them, the underwater robot moves in the aquaculture farm according to the route determined by the feeding route planning module, and real-time collects images, sounds and water quality data of aquatic animals; for the image data, first perform grayscale processing to convert the color image into a grayscale image; then use an image enhancement algorithm (such as histogram equalization) to enhance the contrast of the image and highlight the characteristics of aquatic animals; use a filtering algorithm (such as Gaussian filtering) to remove noise interference in the image; for the sound data, perform filtering and noise reduction processing to extract effective sound signal segments; perform normalization processing on the collected water quality data to convert it into a unified numerical range for subsequent comprehensive analysis.

[0028] Among them, the aquatic behavior monitoring and analysis module uses the image information collected by the underwater robot vision sensor to count the distribution density of aquatic animals in each grid area within different time periods ; at the same time, combined with the behavioral characteristics of aquatic animals (such as swimming speed, aggregation degree, etc.), classify their behavior patterns; for example, define active behavior patterns, foraging behavior patterns, rest behavior patterns, etc., and calculate the probability distribution of aquatic animals in different grid areas under each behavior pattern , among which, represents the behavior pattern category, represents the grid area number; through long-term monitoring and analysis of the distribution and behavior of aquatic animals, establish their behavior logic model and predict their distribution and behavior trends at specific times and in specific environments.

[0029] Among them, the aquatic behavior monitoring and analysis module extracts the shape features (such as perimeter, area, circularity, etc.), texture features (such as gray-level co-occurrence matrix features), and color features (such as RGB color mean, standard deviation, etc.) of aquatic animals from the preprocessed images; combines the frequency features and intensity features of the sound signals to construct a comprehensive feature vector of aquatic animals, and classifies and identifies the comprehensive feature vector through a trained machine learning model to determine the current behavior state of the aquatic animals (such as normal swimming, foraging, disease infection, etc.).

[0030] Among them, the disease warning module formulates a series of disease warning rules according to the behavior characteristics of aquatic animals and water quality monitoring data. For example: If aquatic animals frequently show abnormal swimming behaviors (such as significantly reduced swimming speed, out-of-control direction, etc.) in a certain area and the duration exceeds a certain threshold (such as more than 5 minutes continuously), a suspected disease warning is issued.

[0031] When the water quality parameters exceed the appropriate range (such as the water temperature is higher or lower than the set upper and lower limits, the dissolved oxygen content is lower than the set lower limit, etc.) and last for a certain period of time (such as more than 10 minutes continuously), a water quality abnormality warning is issued, indicating that there may be a disease risk.

[0032] Among them, the disease warning module uses machine learning algorithms (such as Bayesian networks) to perform fusion analysis on the behavior data, water quality data, and historical disease data of aquatic animals. The structure of the Bayesian network model includes nodes (representing various variables, such as aquatic animal behavior, water quality parameters, disease types, etc.) and directed edges (representing the causal relationships between variables). By learning the variable relationships in the training data, a conditional probability table is obtained, so that the probability of a certain disease occurring can be calculated according to the current observed data. When the probability of a certain disease occurring exceeds the set threshold, a disease warning message is issued.

[0033] Among them, the prevention and control decision-making module collects and collates various common prevention and control methods and drug information for aquatic diseases, and establishes a prevention and control plan library. The prevention and control plan library contains the symptom descriptions, cause analyses, recommended prevention and control drugs and their usage methods, dosages, precautions, etc. for each disease. For example, for bacterial enteritis, the symptoms are that aquatic animals have loss of appetite, abdominal distension, abnormal feces, etc., and the main causes are bacterial infection and water quality deterioration. The recommended prevention and control drugs include oxytetracycline, florfenicol, etc., and the usage method is to mix with bait for feeding. The dosage is calculated according to the weight of the aquatic animals (such as using 10-20 milligrams of oxytetracycline per kilogram of body weight), and the precautions include avoiding mixing with alkaline drugs, etc.

[0034] Among them, the prevention and control decision-making module intelligently matches the most suitable prevention and control plan from the prevention and control plan library according to the disease type and severity determined by the disease early warning module. At the same time, considering the actual situation of the farm (such as water quality conditions, breeding scale, drug residue limits, etc.) and cost factors, the prevention and control plan is optimized and adjusted. For example: If the water quality of the farm is good and the breeding scale is small, a biological control method with less side effects and relatively low cost can be selected (such as putting beneficial microbial agents); If the disease is relatively serious and the condition needs to be controlled urgently, chemical drug treatment is preferred, but attention should be paid to controlling the drug dosage and usage frequency to avoid excessive drug residues.

[0035] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution of the present invention and its inventive concept, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. An intelligent disease warning and precise prevention and control system for aquaculture, characterized in that, Including: An underwater robot module, which realizes the collection of underwater information based on the sensors carried by the underwater robot and realizes the feeding of feed through the integrated feed delivery device; A route planning module for planning the operation route of the underwater robot; An aquaculture monitoring and analysis module, which analyzes the aquaculture situation and water quality situation of aquatic products according to the information collected by the underwater robot; A disease warning module, which conducts disease warning based on the analysis results of the aquaculture monitoring and analysis module; A prevention and control decision-making module, which generates a prevention and control decision-making plan according to the disease warning information of the disease warning module.

2. The intelligent disease early warning and precise prevention and control system for aquaculture according to claim 1, wherein The underwater robot is equipped with multiple thrusters, which are distributed in different orientations of the robot. The flexible steering and precise positioning of the robot are achieved through a vector control algorithm. Let the coordinates of the robot in the three-dimensional space be (x, y, z), where the x-axis direction is the length direction of the farm, the y-axis direction is the width direction, and the z-axis direction is the water depth direction. The motion velocity vector of the robot is expressed as: where are the velocity components of the robot in the x, y, and z directions respectively. By precisely controlling the rotation speeds of the respective thrusters, the precise control of the motion velocity vector of the robot is achieved, and thus the robot can move to any position within the farm.

3. The intelligent disease early warning and precise prevention and control system for aquaculture according to claim 1, characterized in that, The sensor system of the underwater robot module includes: A vision sensor: installed at the front end of the robot for real-time collection of image information in the farm; by processing and analyzing consecutive image frames, obtaining the position, size, and posture information of aquatic animals; A water quality sensor: distributed on the surface of the robot for real-time monitoring of water quality parameters such as water temperature, dissolved oxygen content, and pH value; An acoustic sensor: used to detect the sound signals emitted by aquatic animals, and different species and health conditions of aquatic animals will emit sound signals within a specific frequency range.

4. An intelligent disease warning and precise prevention and control system for aquaculture according to claim 1, characterized in that, Before route planning, the route planning module establishes a three-dimensional spatial model of the farm; divides the farm into multiple grid areas, with the size of each grid being Δx×Δy×Δz, and its vertex coordinates being , where ; ; ; The environmental factors such as water depth, water flow velocity, and substrate type of each grid area are measured and recorded by sensors carried by the underwater robot, and an environmental factor matrix E is constructed: where represents the comprehensive environmental factor value of the th grid area, and its calculation method is: In the formula, is the water depth factor, is the water flow velocity factor, is the substrate factor, , , are the weight coefficients of the corresponding factors.

5. An intelligent disease warning and precise prevention and control system for aquaculture according to claim 4, characterized in that, The route planning module comprehensively considers the environmental factors of the farm, the distribution and behavior trends of aquatic animals, and aims to minimize the feeding time, maximize the feed utilization rate, and ensure uniform feeding of aquatic animals, and establishes an optimized feeding route model; let the sequence of grid areas passed by the feeding route be , where K is the total number of grid areas on the feeding route; then the optimized feeding route problem is expressed as: Among them, is the total feeding time, is the feeding uniformity index of aquatic animals, represents the distance between adjacent grid regions and ; is the average moving speed of the robot, is the maximum distance limit for a single movement of the robot, is the total amount of feed for each feeding; By solving this optimization problem, the optimal feeding route is obtained.

6. The intelligent disease early warning and precise prevention and control system for aquaculture according to claim 1, characterized in that, The underwater robot moves in the farm according to the route determined by the feeding route planning module, and real-time collects images, sounds, and water quality data of aquatic animals; for the image data, first perform grayscale processing to convert the color image into a grayscale image; then use an image enhancement algorithm to enhance the contrast of the image and highlight the characteristics of aquatic animals; Use a filtering algorithm to remove noise interference in the image; for the sound data, perform filtering and noise reduction processing to extract effective sound signal segments; perform normalization processing on the collected water quality data to convert it to a unified numerical range.

7. An intelligent disease early warning and precise prevention and control system for aquaculture according to claim 1, characterized in that, The aquatic behavior monitoring and analysis module uses the image information collected by the underwater robot's vision sensor to count the distribution density of aquatic animals in each grid area during different time periods. Meanwhile, in combination with the behavioral characteristics of aquatic animals, it classifies their behavior patterns; through long-term monitoring and analysis of the distribution and behavior of aquatic animals, it establishes their behavioral logic model to predict their distribution and behavior trends at specific times and in specific environments.

8. An intelligent disease warning and precise prevention and control system for aquaculture according to claim 7, characterized in that, The aquaculture behavior monitoring and analysis module extracts the shape features, texture features, and color features of aquatic animals from the preprocessed images; Combines the frequency features and intensity features of the sound signal to construct a comprehensive feature vector of aquatic animals, and classifies and identifies the comprehensive feature vector through a trained machine learning model to determine the current behavior state of aquatic animals.

9. An intelligent disease early warning and precise prevention and control system for aquaculture according to claim 1, characterized in that, The disease warning module uses machine learning algorithms to perform fusion analysis on the behavior data, water quality data, and historical disease data of aquatic animals; by learning the variable relationships in the training data, obtaining a conditional probability table, and thus calculating the probability of disease occurrence according to the current observed data. When the probability of disease occurrence exceeds the set threshold, a disease warning information is issued.

10. The intelligent disease early warning and precise prevention and control system for aquaculture according to claim 1, characterized in that, The prevention and control method of the prevention and control system includes the following steps: S1: The route planning module plans the operation route of the underwater robot; S2: The underwater robot realizes the collection of underwater information based on the planned operation route and the sensors carried by the underwater robot, and realizes the feeding of feed through the integrated feed delivery device; S3: The aquaculture monitoring and analysis module analyzes the aquaculture situation and water quality situation of aquatic products according to the information collected by the underwater robot; S4: The disease warning module conducts disease warning based on the analysis results of the aquaculture monitoring and analysis module; S5: The prevention and control decision-making module generates a prevention and control decision-making plan according to the disease warning information of the disease warning module.

Images