Management method and system of tuna intelligent pond culture platform

Abstract

The invention relates to the field of tuna culture management, and discloses a management method and system for a tuna intelligent pond culture platform, and the method comprises the following steps: collecting the culture associated data of tunas in combination with underwater monitoring equipment, overwater monitoring equipment and water quality monitoring equipment, and constructing a tuna culture evaluation model; carrying out risk assessment on the tuna culture state of the pond based on the tuna culture assessment model; and performing water quality parameter adjustment on the target pond area with the culture risk, and performing visual data display processing through the tuna intelligent pond culture platform. According to the invention, a data monitoring platform can be provided for intelligent culture of tunas, help can be provided for tuna culture more efficiently, the purpose of system automation is achieved, and meanwhile, the harm in the tuna culture process is reduced.

Description

Technical Field

[0001] The present invention relates to the field of tuna farming management, and in particular to a management method and system for an intelligent tuna pond farming platform. Background Art

[0002] Tuna is one of the most sought-after marine fish in the global seafood market. Its high market value has triggered overfishing, leading to a sharp decline in populations across major oceans. Sustained market demand has driven up tuna prices, creating new opportunities for tuna farming and product processing. Tuna farming has become one of the most profitable aquaculture industries globally, attracting research from numerous countries.

[0003] Although various positive measures have been taken to develop the tuna industry, such as enhancing the industry's status, increasing scientific research efforts, and coordinating talent training, and some achievements have been made, my country's tuna industry still faces many practical problems:

[0004] 1. Tuna resources in the South China Sea are already fully exploited. The exploitation rates for yellowfin and bigeye tuna in the South China Sea are 63.07% and 62.47% respectively, indicating that the resources are fully exploited.

[0005] Second, the amount of resources in the South China Sea fluctuates greatly from year to year and seasonally, making it impossible for professional fishing boats to operate year-round;

[0006] Third, the catch rate of tuna longline fishing in the South China Sea is relatively low, making longline operations difficult to achieve economic returns. Fourth, tuna fishing vessels lack low-energy preservation technology. Sashimi-grade tuna requires stringent quality standards, and offshore net voyages can take up to two months. Existing fishing vessels lack ultra-low-temperature quick-freezing equipment, so upon returning to port, the tuna generally does not meet raw consumption standards, resulting in purchase prices of only 4-9 yuan per jin (approximately 100 catties). Subjectively, research into tuna aquaculture technology in my country started relatively late, resulting in a severe lack of biological data. Overall scientific research capabilities are low, and the majority of tuna in the market still comes from wild catches. Relevant professional technical teams have yet to be established, particularly those with strong and comprehensive tuna industry expertise.

[0007] A method for artificially breeding tuna is proposed, and combined with an intelligent pond breeding platform for tuna, it is used for tuna breeding management, including adjusting parameters such as environmental water flow rate, environmental noise and light intensity, bait type and feeding frequency, water temperature and temperature difference control based on tuna behavior, feeding activity and survival rate. Therefore, a management method and system for the intelligent pond breeding platform for tuna are proposed. Summary of the Invention

[0008] The present invention overcomes the deficiencies of the prior art and provides a management method and system for an intelligent tuna pond aquaculture platform.

[0009] In order to achieve the above object, the technical solution adopted by the present invention is:

[0010] A first aspect of the present invention provides a management method for an intelligent tuna pond aquaculture platform, comprising the following steps:

[0011] Tuna farming-related data are collected through underwater monitoring equipment, surface monitoring equipment, and water quality monitoring equipment, and the tuna farming-related data are preprocessed on the tuna intelligent pond farming platform to obtain preprocessed tuna farming-related data;

[0012] Based on the pre-processed tuna farming related data, a tuna farming assessment model is established, and the real-time risk status of the target pond is evaluated based on the tuna farming assessment model;

[0013] Adjust the water quality parameters of the target ponds according to the risk status of tuna farming in different areas of the target ponds;

[0014] Through the tuna intelligent pond farming platform, the tuna farming related data of the target pond is displayed in real time in a visual view.

[0015] Furthermore, in a preferred embodiment of the present invention, the tuna farming related data is collected by underwater monitoring equipment, surface monitoring equipment, and water quality monitoring equipment, and the tuna farming related data is preprocessed on the tuna intelligent pond farming platform to obtain preprocessed tuna farming related data, specifically:

[0016] Obtain a tuna breeding pond, mark it as a target pond, and obtain a tuna intelligent pond breeding platform, wherein the tuna intelligent pond breeding platform is a control platform that can monitor, regulate and analyze data of the target pond;

[0017] Install underwater monitoring equipment and surface monitoring equipment in the target pond, wherein the underwater monitoring equipment and the surface monitoring equipment are cameras for collecting real-time image data from underwater and on the surface of the target pond;

[0018] Installing a water quality monitoring device in the target pond, wherein the water quality monitoring device detects the water quality parameters of the target pond in real time, wherein the water quality parameters of the target pond include water temperature parameters, salinity parameters, pH value, dissolved oxygen content, and turbidity parameters of the pond water;

[0019] The real-time image data of the underwater and water surface of the target pond and the water quality parameters of the target pond are collectively referred to as tuna farming related data;

[0020] A data exchange module and a routing module are installed in the underwater monitoring equipment, the surface monitoring equipment, and the water quality monitoring equipment. The data exchange module can convert the data format of the collected tuna farming related data to obtain a tuna farming related data signal, and transmit the tuna farming related data signal to the tuna intelligent pond farming platform for storage through the routing module;

[0021] In the tuna intelligent pond farming platform, data recovery and data synchronization processing are performed on tuna farming related data signals, wherein the data synchronization processing is to control the collection time and collection area of all tuna farming related data signals to be equal, thereby obtaining recovered tuna farming related data, and data preprocessing is performed on the recovered tuna farming related data to obtain preprocessed tuna farming related data;

[0022] The data preprocessing is to perform data filtering and data noise reduction on the restored tuna farming related data.

[0023] Furthermore, in a preferred embodiment of the present invention, a tuna farming assessment model is established based on the pre-processed tuna farming related data, and the real-time risk status of the target pond is assessed based on the tuna farming assessment model, specifically:

[0024] In the tuna intelligent pond farming platform, based on the water quality parameters in the tuna farming related data, the key feature thresholds of the tuna farming assessment model are constructed and calibrated as target key feature thresholds. The target key feature thresholds include water quality mutation degree, tuna population stress index, feeding competition entropy, and thermal stratification coefficient.

[0025] Obtain an LSTM blank model, input tuna farming-related data into the input layer of the LSTM blank model, and import the target key feature threshold into the fully connected network of the LSTM blank model;

[0026] Performing a graph convolution operation within the input layer of the LSTM blank model, and importing the convolution feature values outputted from the graph convolution operation into a fully connected network for multimodal fusion, thereby obtaining a tuna farming assessment model. Combining the tuna farming assessment model, different key features of tuna farming-related data are output;

[0027] After different key features are output, a historical data network is introduced to retrieve the tuna farming risks corresponding to different key features of tuna farming related data in the historical data network, and a farming risk comparison map is constructed;

[0028] Combined with the aquaculture risk comparison map, tuna zoning risk prediction is carried out in the target ponds.

[0029] Furthermore, in a preferred embodiment of the present invention, the risk assessment and prediction of tuna by region is performed in the target pond in combination with the aquaculture risk control map, specifically:

[0030] Gridding the target ponds and analyzing tuna farming-related data using the intelligent tuna pond farming platform to determine the coordinates of the tuna in the target ponds. The coordinates of the tuna in the target ponds are then mapped into different grid areas.

[0031] A collection of aquaculture risk comparison maps was used to determine the tuna aquaculture risks in different grid areas, and the fuzzy Delphi method was used to determine the weights of different aquaculture risks;

[0032] Based on different aquaculture risk weights, the real-time risk coefficients of different grid areas are calculated, wherein the real-time risk coefficients are calculated based on the proportion of tuna aquaculture risks corresponding to different key characteristics in the aquaculture risk comparison map;

[0033] A real-time risk coefficient of danger is preset, and grid areas where the real-time risk coefficient is greater than the danger risk coefficient are marked as high-risk breeding areas.

[0034] Furthermore, in a preferred embodiment of the present invention, the water quality parameters of the target pond are adjusted according to the risk status of tuna farming in different areas of the target pond, specifically:

[0035] For high-risk aquaculture areas, the tuna aquaculture-related data is divided to obtain the corresponding tuna aquaculture-related data in the high-risk aquaculture areas, and marked as high-risk area aquaculture-related data;

[0036] Water is extracted from the bottom of a high-risk aquaculture area and marked as a target purified water body. A screen filter and a mechanical filter are obtained at the same time. The target purified water body is introduced into the screen filter and the mechanical filter to filter and remove impurities, thereby obtaining a preliminary filtered water body and preliminary filtered impurities.

[0037] The initially filtered impurities include tuna feces and tuna bait residues, and the initially filtered impurities are washed to obtain tail water of the initially filtered impurities;

[0038] obtaining a protein skimmer, and performing protein separation on the tail water from which impurities have been initially filtered using the protein skimmer to obtain a secondary filtered aquaculture water body;

[0039] Obtain secondary filtered aquaculture water samples, and simultaneously introduce ozone sterilization equipment to continuously sterilize the secondary filtered aquaculture water. During the continuous sterilization process, samples of the secondary filtered aquaculture water are taken in real time and calibrated as real-time sterilized water samples;

[0040] Monitor the microbial content in real-time sterilized water samples and secondary filtered aquaculture water samples, calculate the difference in different microbial content, and preset the standard deviation;

[0041] If, during the continuous sterilization process, the differences in all microbial contents are greater than the standard deviation, the current secondary filtered aquaculture water body is judged to be a qualified aquaculture water body;

[0042] The qualified aquaculture water body is treated by gravity flow so that the qualified aquaculture water body flows into the high-risk aquaculture area, and the target water volume threshold in the high-risk aquaculture area is determined. If the volume of the water body in the high-risk aquaculture area is not maintained at the target water volume threshold after the qualified aquaculture water body is treated by gravity flow, the high-risk aquaculture area is subjected to adaptive water volume control treatment by a water pump until the volume of the water body in the high-risk aquaculture area is maintained within the target water volume threshold, thereby obtaining a qualified aquaculture area.

[0043] Furthermore, in a preferred embodiment of the present invention, the tuna farming related data of the target pond is displayed in real time in a visual view on the tuna intelligent pond farming platform, specifically:

[0044] On the tuna intelligent pond farming platform, the tuna farming related data of the target pond is digitally displayed and converted into graphics and text;

[0045] The tuna farming related data is converted into a visual view, including a visual view of water quality parameters of the target pond and a video view captured by underwater monitoring equipment and surface monitoring equipment;

[0046] After the intelligent tuna pond farming platform converts tuna farming-related data into text and graphics, it uses the key features of the tuna farming-related data output by the tuna farming assessment model to conduct real-time risk classification of high-risk farming areas.

[0047] If there is a high-risk breeding area in the target pond, an early warning alarm will be generated in the tuna intelligent pond breeding platform, and the water quality parameters of the high-risk breeding area will be adjusted in combination with the tuna intelligent pond breeding platform.

[0048] A second aspect of the present invention further provides a management system for an intelligent tuna pond farming platform, the management system comprising a memory and a processor, wherein a management method is stored in the memory, and when the management method is executed by the processor, the following steps are implemented:

[0049] Tuna farming-related data are collected through underwater monitoring equipment, surface monitoring equipment, and water quality monitoring equipment, and the tuna farming-related data are preprocessed on the tuna intelligent pond farming platform to obtain preprocessed tuna farming-related data;

[0050] Based on the pre-processed tuna farming related data, a tuna farming assessment model is established, and the real-time risk status of the target pond is evaluated based on the tuna farming assessment model;

[0051] Adjust the water quality parameters of the target ponds according to the risk status of tuna farming in different areas of the target ponds;

[0052] Through the tuna intelligent pond farming platform, the tuna farming related data of the target pond is displayed in real time in a visual view.

[0053] This invention addresses the technical deficiencies in the prior art and has the following beneficial effects: It combines underwater monitoring equipment, surface monitoring equipment, and water quality monitoring equipment to collect tuna farming-related data, which is used to construct a tuna farming assessment model. Based on this tuna farming assessment model, a risk assessment of the tuna farming status of the pond is performed. Water quality parameters are adjusted for target pond areas with farming risks, and the data is visualized and processed through an intelligent tuna pond farming platform. This invention provides a data monitoring platform for intelligent tuna farming, more efficiently facilitating tuna farming and achieving system automation while reducing hazards associated with the tuna farming process. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, without paying any creative work, they can also obtain drawings of other embodiments based on these drawings.

[0055] Figure 1 A flow chart showing a management method for an intelligent tuna pond farming platform;

[0056] Figure 2 A flow chart of a method for establishing a tuna aquaculture assessment model is shown;

[0057] Figure 3 The program view of a management system of an intelligent tuna pond farming platform is shown. DETAILED DESCRIPTION

[0058] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present application and the features therein can be combined with each other.

[0059] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0060] Figure 1 A flow chart showing a method for managing a tuna intelligent pond aquaculture platform is provided, comprising the following steps:

[0061] S102: Collecting tuna farming-related data through underwater monitoring equipment, surface monitoring equipment, and water quality monitoring equipment, and preprocessing the tuna farming-related data on the tuna intelligent pond farming platform to obtain preprocessed tuna farming-related data;

[0062] S104: establishing a tuna farming assessment model based on the pre-processed tuna farming related data, and assessing the real-time risk status of the target pond based on the tuna farming assessment model;

[0063] S106: Adjusting water quality parameters of the target pond based on the risk status of tuna farming in different areas of the target pond;

[0064] S108: The tuna farming related data of the target pond is displayed in real time in a visual view on the tuna intelligent pond farming platform.

[0065] Furthermore, in a preferred embodiment of the present invention, the tuna farming related data is collected by underwater monitoring equipment, surface monitoring equipment, and water quality monitoring equipment, and the tuna farming related data is preprocessed on the tuna intelligent pond farming platform to obtain preprocessed tuna farming related data, specifically:

[0066] Obtain a tuna breeding pond, mark it as a target pond, and obtain a tuna intelligent pond breeding platform, wherein the tuna intelligent pond breeding platform is a control platform that can monitor, regulate and analyze data of the target pond;

[0067] Install underwater monitoring equipment and surface monitoring equipment in the target pond, wherein the underwater monitoring equipment and the surface monitoring equipment are cameras for collecting real-time image data from underwater and on the surface of the target pond;

[0068] Installing a water quality monitoring device in the target pond, wherein the water quality monitoring device detects the water quality parameters of the target pond in real time, wherein the water quality parameters of the target pond include water temperature parameters, salinity parameters, pH value, dissolved oxygen content, and turbidity parameters of the pond water;

[0069] The real-time image data of the underwater and water surface of the target pond and the water quality parameters of the target pond are collectively referred to as tuna farming related data;

[0070] A data exchange module and a routing module are installed in the underwater monitoring equipment, the surface monitoring equipment, and the water quality monitoring equipment. The data exchange module can convert the data format of the collected tuna farming related data to obtain a tuna farming related data signal, and transmit the tuna farming related data signal to the tuna intelligent pond farming platform for storage through the routing module;

[0071] In the tuna intelligent pond farming platform, data recovery and data synchronization processing are performed on tuna farming related data signals, wherein the data synchronization processing is to control the collection time and collection area of all tuna farming related data signals to be equal, thereby obtaining recovered tuna farming related data, and data preprocessing is performed on the recovered tuna farming related data to obtain preprocessed tuna farming related data;

[0072] The data preprocessing is to perform data filtering and data noise reduction on the restored tuna farming related data.

[0073] It should be noted that equipment for monitoring water quality parameters and capturing surface and underwater video footage must be installed in the target tuna ponds, capturing relevant data. The water quality monitoring equipment consists of integrated multi-parameter water quality sensors for real-time monitoring. These sensors include sensors for temperature, salinity, pH, dissolved oxygen, and turbidity, accurately measuring key water quality indicators. The sensors transmit the collected data to a local platform or shore-based management platform for real-time analysis and monitoring of the water quality of the aquatic product growing environment in the aquaculture area. The water quality parameters are then transmitted to the intelligent tuna pond aquaculture platform. The surface monitoring equipment is a Starlight-class network explosion-proof infrared camera, a high-performance, marine-specific surface video surveillance device with outstanding technical features and comprehensive monitoring capabilities. The camera features intelligent detection capabilities, including boundary crossing and area intrusion detection, effectively enhancing monitoring effectiveness. Image enhancement features include backlight compensation (BLC), glare suppression, 3D digital noise reduction, and an excellent 120dB wide dynamic range, ensuring clear and detailed images under varying lighting conditions. The underwater monitoring equipment is a four-in-one underwater digital HD camera. The collected data is fused to generate relevant tuna farming data, which is then uploaded to the platform for storage and analysis. Any noise present during the upload process requires filtering and noise reduction. After uploading, the data undergoes spatiotemporal synchronization to ensure consistency and integrity.

[0074] Furthermore, in a preferred embodiment of the present invention, the water quality parameters of the target pond are adjusted according to the risk status of tuna farming in different areas of the target pond, specifically:

[0075] For high-risk aquaculture areas, the tuna aquaculture-related data is divided to obtain the corresponding tuna aquaculture-related data in the high-risk aquaculture areas, and marked as high-risk area aquaculture-related data;

[0076] Water is extracted from the bottom of a high-risk aquaculture area and marked as a target purified water body. A screen filter and a mechanical filter are obtained at the same time. The target purified water body is introduced into the screen filter and the mechanical filter to filter and remove impurities, thereby obtaining a preliminary filtered water body and preliminary filtered impurities.

[0077] The initially filtered impurities include tuna feces and tuna bait residues, and the initially filtered impurities are washed to obtain tail water of the initially filtered impurities;

[0078] obtaining a protein skimmer, and performing protein separation on the tail water from which impurities have been initially filtered using the protein skimmer to obtain a secondary filtered aquaculture water body;

[0079] Obtain secondary filtered aquaculture water samples, and simultaneously introduce ozone sterilization equipment to continuously sterilize the secondary filtered aquaculture water. During the continuous sterilization process, samples of the secondary filtered aquaculture water are taken in real time and calibrated as real-time sterilized water samples;

[0080] Monitor the microbial content in real-time sterilized water samples and secondary filtered aquaculture water samples, calculate the difference in different microbial content, and preset the standard deviation;

[0081] If, during the continuous sterilization process, the differences in all microbial contents are greater than the standard deviation, the current secondary filtered aquaculture water body is judged to be a qualified aquaculture water body;

[0082] The qualified aquaculture water body is treated by gravity flow so that the qualified aquaculture water body flows into the high-risk aquaculture area, and the target water volume threshold in the high-risk aquaculture area is determined. If the volume of the water body in the high-risk aquaculture area is not maintained at the target water volume threshold after the qualified aquaculture water body is treated by gravity flow, the high-risk aquaculture area is subjected to adaptive water volume control treatment by a water pump until the volume of the water body in the high-risk aquaculture area is maintained within the target water volume threshold, thereby obtaining a qualified aquaculture area.

[0083] It's important to note that after identifying high-risk aquaculture areas, the water within these areas must be filtered and cleaned to ensure that tuna farming in these areas poses no risk to the fish. Poor water quality can lead to unhealthy conditions during the aquaculture process. Water from the bottom of the aquaculture pond is pumped through screen filters and mechanical filters to remove particulate matter such as fish feces and leftover bait. This water is then flushed through an automatic backwash system to produce a pre-filtered tailwater. This pre-filtered tailwater requires further treatment, as it could pose a risk to tuna farming. This pre-filtered tailwater then passes through a protein skimmer to further remove fine suspended particles. Finally, it passes through an ozone sterilization system to eliminate pathogenic microorganisms and algae in the secondary filtered aquaculture water before being released into the aquaculture ponds. Water sampling is required before ozone sterilization and continues during the sterilization process. Real-time microbial content analysis of these samples is performed to assess the sterilization effectiveness. If the difference in all microbial content values is greater than the standard deviation, ozone sterilization treatment is no longer necessary to improve sterilization efficiency and the resulting tail water can be redirected to the aquaculture area. Because the volume of tail water may change after sterilization, adaptive water volume control treatment is required in conjunction with water pumps in high-risk aquaculture areas to prevent the water volume in the area from decreasing or increasing and to maintain it at a certain value.

[0084] Furthermore, in a preferred embodiment of the present invention, the tuna farming related data of the target pond is displayed in real time in a visual view on the tuna intelligent pond farming platform, specifically:

[0085] On the tuna intelligent pond farming platform, the tuna farming related data of the target pond is digitally displayed and converted into graphics and text;

[0086] The tuna farming related data is converted into a visual view, including a visual view of water quality parameters of the target pond and a video view captured by underwater monitoring equipment and surface monitoring equipment;

[0087] After the intelligent tuna pond farming platform converts tuna farming-related data into text and graphics, it uses the key features of the tuna farming-related data output by the tuna farming assessment model to conduct real-time risk classification of high-risk farming areas.

[0088] If there is a high-risk breeding area in the target pond, an early warning alarm will be generated in the tuna intelligent pond breeding platform, and the water quality parameters of the high-risk breeding area will be adjusted in combination with the tuna intelligent pond breeding platform.

[0089] It should be noted that the intelligent tuna pond farming platform is a data visualization platform. By visualizing tuna farming-related data, tuna farmers can directly understand the specific water quality data in the target pond. At the same time, it can calibrate and grade the risks of different areas in real time, so as to know when the water quality parameters of the target pond need to be adjusted.

[0090] Figure 2 A flow chart of a method for establishing a tuna farming assessment model is shown, comprising the following steps:

[0091] S202: establishing a tuna farming assessment model based on the pre-processed tuna farming related data, and assessing the real-time risk status of the target pond based on the tuna farming assessment model;

[0092] S204: Based on the aquaculture risk comparison map, conduct zoning risk assessment and prediction for tuna in the target pond.

[0093] Furthermore, in a preferred embodiment of the present invention, a tuna farming assessment model is established based on the pre-processed tuna farming related data, and the real-time risk status of the target pond is assessed based on the tuna farming assessment model, specifically:

[0094] In the tuna intelligent pond farming platform, based on the water quality parameters in the tuna farming related data, the key feature thresholds of the tuna farming assessment model are constructed and calibrated as target key feature thresholds. The target key feature thresholds include water quality mutation degree, tuna population stress index, feeding competition entropy, and thermal stratification coefficient.

[0095] Obtain an LSTM blank model, input tuna farming-related data into the input layer of the LSTM blank model, and import the target key feature threshold into the fully connected network of the LSTM blank model;

[0096] Performing a graph convolution operation within the input layer of the LSTM blank model, and importing the convolution feature values outputted from the graph convolution operation into a fully connected network for multimodal fusion, thereby obtaining a tuna farming assessment model. Combining the tuna farming assessment model, different key features of tuna farming-related data are output;

[0097] After different key features are output, a historical data network is introduced to retrieve the tuna farming risks corresponding to different key features of tuna farming related data in the historical data network, and a farming risk comparison map is constructed;

[0098] Combined with the aquaculture risk comparison map, tuna zoning risk prediction is carried out in the target ponds.

[0099] It should be noted that the LSTM model is a predictive model used in this solution to predict the aquaculture risks of tuna in different areas of a pond. The LSTM model consists of an input layer, a fully connected layer, and an output layer. Tuna aquaculture data is fed into the input layer, and target key feature thresholds are introduced into the fully connected network of the blank LSTM model. The goal is to analyze the tuna aquaculture data and predict the relationship between the data and the target key feature thresholds. The target key feature thresholds include water quality mutation rate, tuna population stress index, feeding competition entropy, and thermal stratification coefficient, all of which reflect the state of tuna aquaculture in the target pond. Water quality mutation rate represents the rate of change of dissolved oxygen, tuna population stress index represents abnormal tuna behavior caused by environmental stress, feeding competition entropy indicates the uniformity of the distribution of tuna feed, and thermal stratification coefficient indicates the vertical stability of water temperature in the pond. LSTM is a predictive model, and convolution is used as a training method for the model. Therefore, graph convolution is performed on the input layer. The resulting convolutional features are then fed into a fully connected network for multimodal fusion. This allows the model to determine the state of associated data based on target key feature thresholds, resulting in a model that can assess aquaculture risks—the tuna aquaculture assessment model. By introducing a historical data network, the corresponding tuna aquaculture risks for different key features of the associated tuna aquaculture data are retrieved from the historical data network. This allows the model to construct a comparative map of aquaculture risks, determining the risk level for aquaculture tuna under different feature data.

[0100] Furthermore, in a preferred embodiment of the present invention, the risk assessment and prediction of tuna by region is performed in the target pond in combination with the aquaculture risk control map, specifically:

[0101] Gridding the target ponds and analyzing tuna farming-related data using the intelligent tuna pond farming platform to determine the coordinates of the tuna in the target ponds. The coordinates of the tuna in the target ponds are then mapped into different grid areas.

[0102] A collection of aquaculture risk comparison maps was used to determine the tuna aquaculture risks in different grid areas, and the fuzzy Delphi method was used to determine the weights of different aquaculture risks;

[0103] Based on different aquaculture risk weights, the real-time risk coefficients of different grid areas are calculated, wherein the real-time risk coefficients are calculated based on the proportion of tuna aquaculture risks corresponding to different key characteristics in the aquaculture risk comparison map;

[0104] A real-time risk coefficient of danger is preset, and grid areas where the real-time risk coefficient is greater than the danger risk coefficient are marked as high-risk breeding areas.

[0105] It should be noted that in order to classify the target pond area according to different risk levels, the target pond must first be gridded to facilitate regional classification. The number and coordinates of tuna in different areas vary, so the coordinates of the tuna in the target pond need to be mapped to different grid areas. The Fuzzy Delphi method is an algorithm that calculates the weight of certain characteristics in the risk. According to the Fuzzy Delphi method, the farming risk weights corresponding to different characteristics are obtained, and the proportion of tuna farming risks corresponding to different key characteristics is calculated to determine the high-risk farming areas in the target pond.

[0106] like Figure 3 As shown, the second aspect of the present invention further provides a management system for an intelligent tuna pond farming platform, the management system comprising a memory 31 and a processor 32. The memory 31 stores a management method. When the management method is executed by the processor 32, the following steps are implemented:

[0107] Tuna farming-related data are collected through underwater monitoring equipment, surface monitoring equipment, and water quality monitoring equipment, and the tuna farming-related data are preprocessed on the tuna intelligent pond farming platform to obtain preprocessed tuna farming-related data;

[0108] Based on the pre-processed tuna farming related data, a tuna farming assessment model is established, and the real-time risk status of the target pond is evaluated based on the tuna farming assessment model;

[0109] Adjust the water quality parameters of the target ponds according to the risk status of tuna farming in different areas of the target ponds;

[0110] Through the tuna intelligent pond farming platform, the tuna farming related data of the target pond is displayed in real time in a visual view.

[0111] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A management method for an intelligent tuna pond aquaculture platform, characterized in that: The following steps are involved: Tuna farming-related data are collected through underwater monitoring equipment, surface monitoring equipment, and water quality monitoring equipment, and the tuna farming-related data are preprocessed on the tuna intelligent pond farming platform to obtain preprocessed tuna farming-related data; Based on the pre-processed tuna farming related data, a tuna farming assessment model is established, and the real-time risk status of the target pond is evaluated based on the tuna farming assessment model; Adjust the water quality parameters of the target ponds according to the risk status of tuna farming in different areas of the target ponds; Through the tuna intelligent pond farming platform, the tuna farming related data of the target pond is displayed in real time in a visual view.

2. The management method of a tuna intelligent pond aquaculture platform according to claim 1, characterized in that: The tuna farming related data is collected by underwater monitoring equipment, water monitoring equipment and water quality monitoring equipment, and the tuna farming related data is preprocessed on the tuna intelligent pond farming platform to obtain preprocessed tuna farming related data, specifically: Obtain a tuna breeding pond, mark it as a target pond, and obtain a tuna intelligent pond breeding platform, wherein the tuna intelligent pond breeding platform is a control platform that can monitor, regulate and analyze data of the target pond; Install underwater monitoring equipment and surface monitoring equipment in the target pond, wherein the underwater monitoring equipment and the surface monitoring equipment are cameras for collecting real-time image data from underwater and on the surface of the target pond; Installing a water quality monitoring device in the target pond, wherein the water quality monitoring device detects the water quality parameters of the target pond in real time, wherein the water quality parameters of the target pond include water temperature parameters, salinity parameters, pH value, dissolved oxygen content, and turbidity parameters of the pond water; The real-time image data of the underwater and water surface of the target pond and the water quality parameters of the target pond are collectively referred to as tuna farming related data; A data exchange module and a routing module are installed in the underwater monitoring equipment, the surface monitoring equipment, and the water quality monitoring equipment. The data exchange module can convert the data format of the collected tuna farming related data to obtain a tuna farming related data signal, and transmit the tuna farming related data signal to the tuna intelligent pond farming platform for storage through the routing module; In the tuna intelligent pond farming platform, data recovery and data synchronization processing are performed on tuna farming related data signals, wherein the data synchronization processing is to control the collection time and collection area of all tuna farming related data signals to be equal, thereby obtaining recovered tuna farming related data, and data preprocessing is performed on the recovered tuna farming related data to obtain preprocessed tuna farming related data; The data preprocessing is to perform data filtering and data noise reduction on the restored tuna farming related data.

3. The management method of a tuna intelligent pond aquaculture platform according to claim 1, characterized in that: The method is to establish a tuna farming assessment model based on the pre-processed tuna farming related data, and to assess the real-time risk status of the target pond based on the tuna farming assessment model, specifically: In the tuna intelligent pond farming platform, based on the water quality parameters in the tuna farming related data, the key feature thresholds of the tuna farming assessment model are constructed and calibrated as target key feature thresholds. The target key feature thresholds include water quality mutation degree, tuna population stress index, feeding competition entropy, and thermal stratification coefficient. Obtain an LSTM blank model, input tuna farming-related data into the input layer of the LSTM blank model, and import the target key feature threshold into the fully connected network of the LSTM blank model; Performing a graph convolution operation within the input layer of the LSTM blank model, and importing the convolution feature values outputted from the graph convolution operation into a fully connected network for multimodal fusion, thereby obtaining a tuna farming assessment model. Combining the tuna farming assessment model, different key features of tuna farming-related data are output; After different key features are output, a historical data network is introduced to retrieve the tuna farming risks corresponding to different key features of tuna farming related data in the historical data network, and a farming risk comparison map is constructed; Combined with the aquaculture risk comparison map, tuna zoning risk prediction is carried out in the target ponds.

4. The management method of a tuna intelligent pond aquaculture platform according to claim 3, characterized in that: The above-mentioned risk assessment and prediction of tuna in target ponds is carried out by combining the aquaculture risk control map, specifically: Gridding the target ponds and analyzing tuna farming-related data using the intelligent tuna pond farming platform to determine the coordinates of the tuna in the target ponds. The coordinates of the tuna in the target ponds are then mapped into different grid areas. A collection of aquaculture risk comparison maps was used to determine the tuna aquaculture risks in different grid areas, and the fuzzy Delphi method was used to determine the weights of different aquaculture risks; Based on different aquaculture risk weights, the real-time risk coefficients of different grid areas are calculated, wherein the real-time risk coefficients are calculated based on the proportion of tuna aquaculture risks corresponding to different key characteristics in the aquaculture risk comparison map; A real-time risk coefficient of danger is preset, and grid areas where the real-time risk coefficient is greater than the danger risk coefficient are marked as high-risk breeding areas.

5. The management method of a tuna intelligent pond aquaculture platform according to claim 1, characterized in that: According to the risk status of tuna farming in different areas of the target pond, the water quality parameters of the target pond are adjusted, specifically: For high-risk aquaculture areas, the tuna aquaculture-related data is divided to obtain the corresponding tuna aquaculture-related data in the high-risk aquaculture areas, and marked as high-risk area aquaculture-related data; Water is extracted from the bottom of a high-risk aquaculture area and marked as a target purified water body. A screen filter and a mechanical filter are obtained at the same time. The target purified water body is introduced into the screen filter and the mechanical filter to filter and remove impurities, thereby obtaining a preliminary filtered water body and preliminary filtered impurities. The initially filtered impurities include tuna feces and tuna bait residues, and the initially filtered impurities are washed to obtain tail water of the initially filtered impurities; obtaining a protein skimmer, and performing protein separation on the tail water from which impurities have been initially filtered using the protein skimmer to obtain a secondary filtered aquaculture water body; Obtain secondary filtered aquaculture water samples, and simultaneously introduce ozone sterilization equipment to continuously sterilize the secondary filtered aquaculture water. During the continuous sterilization process, samples of the secondary filtered aquaculture water are taken in real time and calibrated as real-time sterilized water samples; Monitor the microbial content in real-time sterilized water samples and secondary filtered aquaculture water samples, calculate the difference in different microbial content, and preset the standard deviation; If, during the continuous sterilization process, the differences in all microbial contents are greater than the standard deviation, the current secondary filtered aquaculture water body is judged to be a qualified aquaculture water body; The qualified aquaculture water body is treated by gravity flow so that the qualified aquaculture water body flows into the high-risk aquaculture area, and the target water volume threshold in the high-risk aquaculture area is determined. If the volume of the water body in the high-risk aquaculture area is not maintained at the target water volume threshold after the qualified aquaculture water body is treated by gravity flow, the high-risk aquaculture area is subjected to adaptive water volume control treatment by a water pump until the volume of the water body in the high-risk aquaculture area is maintained within the target water volume threshold, thereby obtaining a qualified aquaculture area.

6. The management method of a tuna intelligent pond aquaculture platform according to claim 1, characterized in that: The tuna farming related data of the target pond is displayed in real time in a visual view on the tuna intelligent pond farming platform, specifically: On the tuna intelligent pond farming platform, the tuna farming related data of the target pond is digitally displayed and converted into graphics and text; The tuna farming related data is converted into a visual view, including a visual view of water quality parameters of the target pond and a video view captured by underwater monitoring equipment and surface monitoring equipment; After the intelligent tuna pond farming platform converts tuna farming-related data into text and graphics, it uses the key features of the tuna farming-related data output by the tuna farming assessment model to conduct real-time risk classification of high-risk farming areas. If there is a high-risk breeding area in the target pond, an early warning alarm will be generated in the tuna intelligent pond breeding platform, and the water quality parameters of the high-risk breeding area will be adjusted in combination with the tuna intelligent pond breeding platform.

7. A management system for an intelligent tuna pond farming platform, characterized in that: The management system includes a memory and a processor, wherein a management method program is stored in the memory. When the management method program is executed by the processor, the management method steps according to any one of claims 1 to 6 are implemented.

Images