A cloud platform application system for intelligent breeding of salted shrimp

Abstract

This invention relates to the field of intelligent cloud platform application technology for salt field shrimp farming, and discloses a cloud platform application system for intelligent salt field shrimp farming, including an edge data acquisition unit, an edge preprocessing unit, a cloud-side storage unit, a dedicated salt field shrimp analysis unit, an automatic control unit, and a full-chain traceability unit. The edge data acquisition unit collects multi-dimensional data such as water environment, diseases, feeding, and video from high-salinity ponds. The edge preprocessing unit performs data noise reduction, downsampling, caching, and breakpoint resumption locally. The cloud-side storage unit uses a cold and hot layered architecture to store all data. This platform solves the problems of poor adaptability to high-salinity environments, high false alarm rates, and insufficient analysis accuracy of general farming platforms through a dedicated data acquisition, analysis, and control system for salt field shrimp. It can reduce farming losses, has full-chain traceability capabilities, support product premiums, and has low deployment costs, adapting to different farming scenarios from individual farmers to large-scale bases.

Description

Technical Field

[0001] This invention relates to the field of cloud platform application technology for intelligent farming of salt field shrimp, and in particular to a cloud platform application system for intelligent farming of salt field shrimp. Background Technology

[0002] my country is the world's largest producer of salt-field shrimp. Because salt-field shrimp grow in high-salinity salt field environments, they are highly resistant to disease and have a sweet taste. Market demand has been rising year by year, and the scale of farming has continued to expand. Currently, the industry's core needs are focused on four areas: reducing farming losses, increasing unit yield, controlling farming costs, and ensuring product quality. Digital farming is the core path for the transformation and upgrading of the salt-field shrimp farming industry.

[0003] Currently, the mainstream digital technologies in the salt field shrimp farming industry can be divided into two categories. The first category is the pond-based independent monitoring system. This type of system works by deploying basic water environment sensors in a single farming pond, coupled with a local audible and visual alarm device. When the monitored parameters exceed the thresholds manually set by the farmer, an alarm is triggered. This type of system is simple to implement, requires no public network deployment, and only needs local power to operate. It is currently widely used in small-scale farming scenarios, with its core advantages being low deployment cost and low operating threshold, making it suitable for small-scale farmers lacking digital infrastructure. However, this type of technology has significant drawbacks: it cannot achieve unified management of data from multiple ponds, all data analysis relies on the farmer's own experience, it only has the ability to trigger alarms after the fact based on thresholds, it lacks risk prediction and farming strategy optimization capabilities, and all data is stored only on local devices. Once the equipment is damaged, the data is completely lost and cannot be used for subsequent farming experience summarization and iteration. The second type is the general-purpose aquaculture cloud platform. This type of platform works by uploading aquaculture data collected from various ponds to a public cloud, providing basic data storage and threshold alarm functions. Some platforms are equipped with general aquaculture analysis models and are currently mainly used in large-scale aquaculture bases. Their core advantage is the ability to centrally view data from multiple ponds, reducing management costs for large-scale aquaculture. However, this type of technology also has significant drawbacks: all models are designed based on general aquaculture scenarios, failing to consider the special growth characteristics of high-salinity salt field shrimp farming. Parameter thresholds, early warning logic, and feeding strategies are not matched to the actual farming needs of salt field shrimp, easily leading to false alarms and incorrect guidance. Furthermore, the lack of edge preprocessing means that directly uploading raw data results in high bandwidth pressure, and the upload of large amounts of noisy data also affects analysis accuracy. Additionally, it lacks end-to-end traceability capabilities for salt field shrimp, failing to meet consumer demand for high-quality salt field shrimp traceability.

[0004] Current technologies cannot match the specific characteristics of salt field shrimp farming, resulting in problems such as insufficient adaptability, low analytical accuracy, lagging management and control, and low data utilization. These issues severely restrict the efficiency of digital transformation in salt field shrimp farming, and targeted technical solutions are urgently needed to address these problems. Summary of the Invention

[0005] To address the problems of poor adaptability, low analytical accuracy, lagging management and control, and insufficient traceability in existing digital systems for salt field shrimp farming, this invention provides a cloud platform application system for intelligent salt field shrimp farming.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a cloud platform application system for intelligent aquaculture of salt field shrimp, characterized in that it includes an aquaculture environment sensing module, an edge data preprocessing module, a cloud platform service center, and a terminal interaction module that are connected in sequence via communication. The aquaculture environment sensing module is deployed at the corresponding monitoring points in each salt field shrimp farming pond to collect multi-dimensional environmental parameters and aquaculture behavior-related data throughout the entire salt field shrimp farming cycle. The edge data preprocessing module is deployed on the local edge gateway device in Tangkou. It is used to perform noise filtering, format unification and outlier removal on the collected raw data before uploading it to the cloud platform service center. The cloud platform service center has a built-in knowledge graph and multi-dimensional analysis model group specifically for salt field shrimp farming, which is used to perform distributed storage, intelligent analysis and generate corresponding scientific farming management instructions for all uploaded farming data. The terminal interaction module is used to push control instructions, real-time monitoring data and risk warning information to aquaculture participants with different roles, and at the same time receive manual control instructions issued by aquaculture participants and synchronize them to the cloud platform service center.

[0007] Preferably, the aquaculture environment sensing module includes a water environment sensing unit, a pond meteorological sensing unit, an aquaculture behavior sensing unit, and a feed residue detection unit. The water environment sensing unit is deployed at different depths of 0.5 to 1.5 meters underwater in the pond to collect parameters such as salinity, water temperature, dissolved oxygen, pH value, ammonia nitrogen concentration, and nitrite concentration in the aquaculture water. The pond mouth meteorological sensing unit is deployed in an open location with no obstruction around the pond mouth, and is used to collect temperature, air pressure, light intensity, rainfall and wind speed parameters in the pond mouth area. The aquaculture behavior sensing unit collects data on the activity frequency, molting cycle, and population distribution of salt field shrimp using underwater high-definition camera equipment and infrared sensors. The bait residue detection unit collects bait residue parameters in different areas of the pond bottom using an underwater near-infrared spectroscopy sensor.

[0008] Preferably, the edge data preprocessing module has a built-in sliding window filtering algorithm and data consistency verification rules. First, the sliding window filtering algorithm is used to filter the impulse noise in the original collected data. The window size of the sliding window filtering algorithm can be adaptively adjusted according to the data collection frequency of the pond. When the collection frequency is higher than once per minute, the window size is set to 5, and when the collection frequency is lower than once per minute, the window size is set to 3, so as to ensure the noise filtering effect while avoiding the accidental deletion of valid data.

[0009] Then, based on the reasonable threshold range of salt field shrimp farming parameters, outliers are removed from the data. Finally, all data is converted into a unified JSON format and uploaded to the cloud platform service center via NB-IoT or 5G network. At the same time, the original data is stored locally for more than 7 days for subsequent data backtracking and local caching in case of platform failure.

[0010] Preferably, the cloud platform service center is equipped with a distributed storage unit and adopts a cold and hot data separation storage strategy. Historical aquaculture data collected for more than 3 months is stored in low-cost cold storage nodes, while active aquaculture data collected for nearly 3 months and real-time data are stored in high-performance hot storage nodes. At the same time, all data is backed up in three copies in different locations. If any storage node fails, it can automatically switch to other copies to read the data, ensuring the security and access continuity of aquaculture data.

[0011] Preferably, the knowledge graph specifically for salt field shrimp farming pre-includes suitable environmental parameter ranges for salt field shrimp at different growth stages, growth characteristics under different salinity gradients, characteristic parameters of common diseases, feed standards under different farming modes, and emergency management plans under different weather conditions. The knowledge graph allows farming participants to upload their local farming experience for iterative updates. All newly added farming experience data must be reviewed by the platform's operation and maintenance personnel before it becomes effective, to avoid deviations in the model analysis results due to incorrect farming experience input.

[0012] Preferably, the multi-dimensional analysis model group includes a water environment prediction model, a disease early warning model, a feed optimization model, and a profit calculation model. The water environment prediction model predicts the changing trends of water environment parameters over the next 24 to 72 hours based on historical environmental data and official weather forecasts. The disease early warning model outputs the probability of disease occurrence and the warning level based on the behavior data of salt field shrimp and abnormal deviations in water environment parameters. The feed optimization model combines the growth stage of salt field shrimp, the amount of feed residue at the bottom of the pond, and water environment parameters to output the optimal feeding amount, feeding time, and feeding location. The profit calculation model calculates the expected farming profit based on aquaculture input data, salt field shrimp growth cycle data, and the current wholesale market price of salt field shrimp.

[0013] Preferably, after the cloud platform service center generates a control command, it first pushes the command to the terminal interaction module for confirmation by the aquaculture participants. If no manual intervention command is received within the preset 15-minute time threshold, the control command is automatically sent to the corresponding aerator, feeder, water exchange equipment and other execution terminals in the pond to perform the corresponding control operation. If a manual adjustment command is received, the command is sent to the corresponding execution terminal according to the manually adjusted command.

[0014] Preferably, the cloud platform service center has a built-in full-cycle traceability unit that generates a unique traceability QR code for each batch of salt field shrimp. The traceability code is associated with the full-process data of the corresponding batch of salt field shrimp, from shrimp seedling release, breeding management, disease prevention and control, feed use to final harvesting and marketing. Consumers can scan the traceability code to query all breeding information and test reports of the corresponding batch of salt field shrimp.

[0015] Preferably, the terminal interaction module divides different operation permissions according to the roles of the aquaculture participants. Ordinary aquaculture personnel can only view the monitoring data of their own ponds and receive control instructions. Aquaculture managers can view the summary data of all their subordinate ponds and issue manual control instructions. Platform operation and maintenance personnel can adjust and optimize the platform's parameter configuration, model rules, and knowledge graph content.

[0016] Preferably, the cloud platform service center has a built-in pond benchmarking analysis unit, which can perform horizontal comparative analysis on environmental parameters, breeding cycle, disease incidence, unit yield and profit data of multiple breeding ponds in the same area, output breeding optimization suggestions for each pond for breeding participants to refer to, and generate a regional overall breeding data statistical report for industry management departments to use.

[0017] Compared with existing technologies, the beneficial effects of this invention are: First, addressing the shortcomings of existing general aquaculture platforms in terms of insufficient adaptability, this invention incorporates a knowledge graph specific to salt field shrimp and a multi-dimensional analysis model set adapted to the growth characteristics of salt field shrimp. This covers the special environmental needs, disease characteristics, feeding standards, and other exclusive aquaculture logics of salt field shrimp throughout their entire growth cycle, enabling precise analysis and early warning for salt field shrimp farming scenarios. This effectively avoids the false alarms and erroneous guidance caused by the poor adaptability of general platforms, and significantly improves the scientific nature of aquaculture management.

[0018] Secondly, addressing the shortcomings of existing technologies such as low data processing efficiency and lagging control, this invention adopts an architecture combining edge-side preprocessing with cloud-side cold and hot data separation storage. First, noise filtering, outlier removal, and format unification of raw data are completed at the edge, significantly reducing the bandwidth consumption for uploading invalid data and minimizing the interference of noisy data on cloud-side analysis results. The cloud side employs cold and hot data separation storage and a three-replica backup strategy, which improves data storage security while ensuring data access efficiency. Coupled with automatic control triggering logic, control operations can be automatically executed without human intervention, effectively solving the problem of aquaculture losses caused by delayed human response.

[0019] Third, addressing the shortcomings of existing technologies such as low data utilization and insufficient traceability, this invention establishes a complete data link for salt field shrimp from shrimp seedling release to market sales through a full-cycle traceability unit and pond benchmarking analysis function. This not only meets the product traceability needs of consumers but also provides feasible optimization directions for aquaculture entities through horizontal data comparison with ponds in the same region, enabling rapid iteration of aquaculture experience and significantly improving the utilization value of aquaculture data.

[0020] This system can be adapted to different scales of salt field shrimp farming scenarios. It can meet the basic farming and management needs of small-scale farmers, support the unified management of multiple ponds in large-scale farming bases, and connect with the digital supervision needs of industry regulatory departments and the product traceability needs of e-commerce platforms. It has high promotion and application value in the field of salt field shrimp farming. Attached Figure Description

[0021] Figure 1 This is a diagram illustrating the overall architecture and business flow of the intelligent salt field shrimp farming cloud platform proposed in this invention. Figure 2 Here is a flowchart of the logic for multi-source environmental perception and edge preprocessing; Figure 3 A flowchart for intelligent decision-making and knowledge graph analysis on a cloud platform; Figure 4 A flowchart illustrating the data storage strategy and full-lifecycle traceability management process; Figure 5 This is a diagram illustrating the instruction issuance process and role-based permission interactions. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0024] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0025] Reference Figures 1 to 5 This invention discloses a smart cloud platform for salt field shrimp farming, including an edge data acquisition unit, an edge preprocessing unit, a cloud storage unit, a salt field shrimp-specific analysis unit, an automatic control unit, and a full-chain traceability unit. All functional modules are developed specifically for the high salinity farming characteristics of salt field shrimp, without introducing any general farming logic for freshwater shrimp or ordinary seawater shrimp, thus adapting to the full life cycle needs of salt field shrimp farming from the bottom up.

[0026] The edge data acquisition units are deployed at the edge nodes of each salt field shrimp farming pond. The supporting acquisition terminals all use commercially mature hardware specifically customized for high-salt and high-humidity environments. The specific configuration is as follows: The water temperature sensor uses the Shandong Renke Measurement and Control RS-WD-N01-2 type platinum resistance PT1000 probe, with a measurement accuracy of ±0.1℃ and a protection level of IP68. It can operate continuously and stably for more than 3 years in a water environment with a salinity of 35‰ without calibration. The dissolved oxygen sensor uses the Shanghai Instrument & Electronics Science JPB-607A fluorescence detection probe, which does not require regular electrolyte replacement, has strong anti-contamination capabilities, a measurement accuracy of ±0.2 mg / L, and a long-term drift rate of less than 0.1 mg / L / year. The pH sensor uses a Mettler Toledo InPro3253i solid electrolyte glass electrode, which is resistant to high salt corrosion, has a measurement accuracy of ±0.05pH, and a calibration cycle of no less than 6 months; the salinity sensor uses a Hangzhou Luheng Biotechnology LH-SA10 conductivity detection probe, with a measurement range of 0-50‰ and an accuracy of ±0.5‰, which is suitable for the entire salinity fluctuation range of salt field shrimp farming. The rapid detection terminal for Vibrio uses the Guangdong Dayuan Oasis DY-6000 colloidal gold immunochromatographic analyzer, which can output the detection results of two core pathogenic bacteria in salt field shrimp, Vibrio parahaemolyticus and Vibrio harveyi, in 15 minutes. The detection limit is 10 CFU / mL, and no professional laboratory operation is required. The pond monitoring camera uses the Hikvision DS-2CD3T26WD-I3 model, a 2-megapixel infrared waterproof camera with built-in AI algorithms to support shrimp molting recognition, leftover bait recognition, and unauthorized personnel intrusion recognition. The feeding equipment status acquisition module uses a customized RS485 to LoRa module, compatible with over 90% of mainstream automatic feeders on the market. It can collect real-time data on remaining feed, feeding duration, single feeding amount, and equipment malfunction status. All acquisition terminals communicate with the edge gateway using the LoRaWAN protocol, with a maximum communication distance of 3km. A single gateway can simultaneously connect to 50 acquisition terminals, suitable for deployment in contiguous ponds. The acquisition frequency employs a dynamic adjustment strategy: under normal aquaculture conditions, core water environment parameters are collected every 10 minutes, Vibrio data is detected and uploaded daily, and a keyframe of the video stream is captured every 30 minutes. When extreme weather (typhoon, rainstorm, cold wave) warnings are in effect, the frequency of water environment parameter collection automatically increases to once every 2 minutes, and the video stream collection frequency increases to one frame every 5 minutes, ensuring that the data collection density under abnormal scenarios is sufficient to support risk assessment. The reported messages adopt a fixed standardized format, including, in sequence, a message header, a 10-digit unique pond number, a 13-digit millisecond-level timestamp, parameter type code, parameter value, and a 16-bit CRC checksum. The cloud side can directly parse this without secondary format conversion.

[0027] The edge data acquisition unit solves the problems of existing general-purpose acquisition equipment being easily corroded and damaged in high-salt environments, having incomplete data acquisition dimension coverage, and having insufficient data in extreme scenarios.

[0028] The edge preprocessing unit is deployed at the Tangkou edge gateway. The gateway uses a Rockchip RK3568 chip, is equipped with 2GB of RAM, 32GB of industrial-grade TF card storage, and is equipped with a 100W monocrystalline silicon solar panel and a 12V 50Ah lithium iron phosphate battery. It does not require an external power supply and can operate stably for more than 7 days in continuous cloudy and rainy weather. It is suitable for deployment needs in remote areas like Tangkou where there is no power supply. The specific processing logic is divided into four layers: The first layer is outlier filtering. Based on the reasonable range of physical parameters for salt field shrimp farming, invalid data that exceeds the range is removed. For example, data with water temperature below 0℃, salinity above 50‰, or pH value below 0 or above 14 are directly judged as sensor fault data and discarded. The second layer is a sliding window filter, which uses a sliding mean filter algorithm with a window size of 3 and a step size of 1 to remove occasional abnormal data. The calculation logic is to take the average of the same parameter collected in 3 consecutive collections. If a collection value deviates from the average by more than 20%, it is judged as an occasional abnormality and removed. The average of the remaining two collections is used to replace the collection value. The third layer is data aggregation and downsampling. Under normal conditions, every 10 original data entries are aggregated into 1 average data entry for uploading. Under extreme conditions, every 5 entries are aggregated into 1 entry. The data upload volume is reduced by more than 80% compared to directly uploading the original data, which greatly reduces bandwidth consumption. The fourth layer is local caching and breakpoint resume. When the public network is interrupted, the edge gateway stores the pre-processed data to a local industrial-grade TF card, which can store up to 7 days of valid data. After the network is restored, the data is automatically re-uploaded to the cloud in chronological order to avoid data loss.

[0029] Meanwhile, the edge preprocessing unit incorporates lightweight local early warning logic. When parameters significantly exceed safety thresholds, it can directly trigger local audible and visual alarms without waiting for cloud-side instructions, with a response latency of less than 1 second, further improving the response speed to extreme risks. The edge preprocessing unit solves the problems of high bandwidth pressure from directly uploading raw data, noise data interfering with analysis accuracy, data loss due to network instability, and delayed risk response in existing technologies.

[0030] The cloud-side storage unit adopts Alibaba Cloud's hybrid cloud architecture, balancing access efficiency, storage cost, and data security. The specific configuration is as follows: The hot data storage area uses an RDS MySQL 8.0 high-performance instance, configured with 8 cores and 16GB of computing resources and 2TB of SSD storage. It stores the raw collected data, analysis results, management records, and early warning logs of the most recent 3 months. The read and write latency is less than 10ms, and it supports 100,000 concurrent read and write operations per second to meet the low latency requirements of high-frequency access. The cold data storage area uses OSS low-frequency access object storage to store more than 3 months of historical aquaculture data, with storage costs only 15% of high-performance storage, significantly reducing the cost of long-term data storage. The traceability data storage area uses independent RDS read-only instances, physically isolated from core aquaculture data, and is dedicated to storing publicly available data related to the entire traceability chain, preventing traceability access traffic from affecting the operation of core aquaculture business. All data adopts a three-replica cross-availability zone backup strategy and is stored simultaneously on storage nodes in three different geographical regions: Hangzhou, Shanghai, and Qingdao. If any node fails, the system automatically switches to other nodes to read the data, achieving a data reliability of 99.9999%. The data backup strategy involves a full backup at 2 AM daily and an incremental backup every hour, with a 10-year retention period to meet the storage requirements for agricultural product traceability. Data transmission uses TLS 1.3 encryption, and storage uses AES 256 encryption. Core aquaculture data is only accessible to the authorized aquaculture entity, ensuring data security across transmission, storage, and access. The cloud-based storage unit addresses the problems of easy data loss, low access efficiency, high storage costs, and poor data security associated with existing local storage technologies.

[0031] The dedicated analysis unit for salt-field shrimp incorporates a proprietary knowledge graph and multi-dimensional model set, built entirely from salt-field shrimp farming data without incorporating data from any other aquaculture species, achieving 100% compatibility and specificity. The knowledge graph's data sources include publicly available salt-field shrimp farming standards from the Yellow Sea Fisheries Research Institute of the Chinese Academy of Fishery Sciences, nearly 120,000 real farming records from the four major salt-field shrimp producing areas in China over the past 10 years, and over 2,000 salt-field shrimp disease sample data. The knowledge graph contains 127 entities, 342 relationships, and 189 attribute dimensions, covering growth parameter standards for salt-field shrimp throughout their entire lifecycle—from larvae, hepatocellular carcinogenesis, growth, and harvest—as well as the causes and treatment plans for 17 common diseases, staged feeding standards, and environmental control thresholds. For example, a Vibrio concentration exceeding 50 CFU / mL during the larval stage is considered high risk, while a Vibrio concentration exceeding 100 CFU / mL during the growth stage is also considered high risk, perfectly aligning with the disease resistance characteristics of salt-field shrimp at different growth stages.

[0032] The model set includes three core models: a disease risk prediction model, a feeding rate optimization model, and a water quality control model. All models are trained using the XGBoost algorithm, with the training set comprising 80%, the validation set 10%, and the test set 10%. The SMOTE algorithm is used during training to address the sample imbalance problem, and the hyperparameters are set to... =6、 , The disease risk prediction model employs multi-parameter weighted fusion logic, and the specific calculation formula is as follows: , where R is the disease risk index of salt field shrimp, with a value range from 0 to 1. The higher the value, the higher the probability of disease occurrence. to The specific weighting coefficients for salt field shrimp are calculated using the XGBoost algorithm based on the feature importance of 120,000 historical data points of salt field shrimp farming. The specific values ​​are 0.0021, 0.032, 0.105, -0.078, and 0.019, respectively. S is the total concentration of Vibrio parahaemolyticus and Vibrio harveyi in the pond water, in CFU / mL; T is the pond water temperature, in °C; pH is the pH of the pond water; DO is the dissolved oxygen content of the pond water, in mg / L; and Sal is the salinity of the pond water, in ‰.

[0033] When the R-value is below 0.3, it is considered low risk and no warning is needed; when the R-value is between 0.3 and 0.7, it is considered medium risk and an important warning is issued; when the R-value is above 0.7, it is considered high risk, an emergency warning is issued, and automatic control logic is triggered. This model achieved an accuracy of 92.7%, a recall of 89.3%, and an F1 score of 0.91 on the test set, which is 23 percentage points higher than the accuracy of a general marine shrimp disease prediction model. The feeding rate optimization model can improve feed utilization by 18%, and the water quality control model can improve water quality compliance rate by 22%.

[0034] The model is set up with automatic iteration logic. The platform collects newly uploaded aquaculture data every quarter to fine-tune the model, ensuring that it always adapts to the changing needs of new shrimp fry varieties and aquaculture models. The dedicated analysis unit for salt field shrimp solves the problems of poor model adaptability, low analysis accuracy, and inability to match the special aquaculture needs of salt field shrimp in existing general aquaculture platforms.

[0035] The automated control unit interfaces with automated aquaculture equipment in the pond, such as aerators, automatic feeders, water regulating equipment, and water pumps. It communicates using the MQTT V3.1.1 protocol with a QoS level of 2 to ensure no lost or duplicate commands and an end-to-end command delivery latency of less than 2 seconds. Control commands undergo triple verification before being sent to prevent erroneous operations. The first step is parameter cross-validation. For example, when dissolved oxygen is below the threshold, it is necessary to simultaneously verify whether water temperature, salinity, and pH value are all in the abnormal range to rule out misjudgments caused by a single sensor failure. The second step is duration verification, which needs to confirm that the parameter abnormality has lasted for more than 5 minutes to exclude the influence of occasional data fluctuations. The third step is equipment status verification, which requires confirming that the corresponding managed equipment is online and fault-free to avoid the equipment not responding after commands are issued.

[0036] Only after all three verifications pass will the control command be sent to the edge gateway. The command execution result will be sent back to the cloud within 1 minute. If no execution receipt is received from the device within 30 seconds, the system will automatically resend the command. If the resend fails twice, a device fault warning will be triggered and pushed to the corresponding management personnel. The automatic control unit supports manual switching mode. Farmers can turn off the automatic control function at any time or manually adjust the control parameters, balancing automation efficiency and manual flexibility. The automatic control unit solves the problems of existing technologies, such as only being able to issue alarms after the fact, delayed manual response, and high rate of misoperation.

[0037] The end-to-end traceability unit connects the data links between the breeding end, processing end, and sales end. The three types of data are linked by a unique batch number. The batch number adopts the format of "6-digit regional administrative code + 3-digit pond number + 8-digit harvest date + 2-digit batch sequence number", which is globally unique. For example, the first batch harvested from pond No. 1 in Wudi County, Binzhou on August 20, 2023 has the batch number 3716230012023082001.

[0038] The data from the aquaculture end includes the source of shrimp larvae, release time, feeding records, water quality monitoring records, disease treatment records, harvest time, and origin testing reports; the data from the processing end includes processing time, processing procedures, cold chain temperature records, and factory testing reports; and the data from the sales end includes distribution paths, sales time, and sales channels.

[0039] Each batch of salt-field shrimp is assigned a unique traceability QR code, using QR Code version 10 and H-level error correction standards, ensuring normal recognition even with 30% wear. The QR code content is a link combining the platform's traceability domain name and the corresponding batch number, supporting direct scanning via WeChat and Alipay. The traceability page only displays publicly compliant information, including origin information, growth cycle, test reports, and disease treatment records, without involving the core privacy data of the aquaculture operators. It also supports integration with the public API of the National Agricultural Product Quality and Safety Traceability Management Information Platform, directly synchronizing traceability data to meet the traceability requirements of regulatory authorities. This end-to-end traceability unit solves the problems of existing technologies lacking dedicated traceability capabilities for salt-field shrimp, low consumer trust, and inability to meet the access requirements of high-end sales channels.

[0040] This invention also discloses a multi-pond benchmarking analysis module. This module uses a cosine similarity algorithm to calculate the matching degree between the target pond and other ponds within the platform. The feature vector includes five core dimensions: pond area, average salinity, shrimp larvae species, stocking time, and farming mode. Only ponds with a similarity higher than 0.8 are included in the benchmark pond. The module then automatically selects the top 10% of ponds in terms of yield as target benchmarks and calculates the differences in core parameters between the target pond and the benchmark pond, such as a 10% higher feed rate, a 0.5 mg / L lower average dissolved oxygen level, and a 20 CFU / mL higher average Vibrio concentration. Afterward, it generates actionable optimization suggestions based on the weighted impact of these parameters on yield. These suggestions include actual performance data from the benchmark ponds, such as "reducing daily feed rate by 8% and increasing aeration time by 1 hour daily; this measure resulted in a 12% increase in feed utilization and a 5% reduction in disease rate in the benchmark pond." The multi-pond benchmarking analysis module solves the problems of low data utilization and the inability to provide clear optimization directions for farmers in existing technologies.

[0041] This invention also discloses a multi-level early warning push module, which divides early warning information into three levels according to risk level: general early warning corresponds to potential risks that will not cause direct economic losses, such as pH value deviating from the suitable range by less than 0.2, and is only pushed to the APP message bar of the aquaculture entity; important early warning corresponds to mild risks that may cause a slight reduction in production if not dealt with in time, such as dissolved oxygen being 1 mg / L lower than the suitable value, and is pushed to the APP message and SMS reminder via Alibaba Cloud SMS service; emergency early warning corresponds to major risks that may cause major losses such as crop failure if not dealt with in time, such as Vibrio concentration exceeding the safety threshold by 2 times, and is pushed to the APP message, SMS, and voice call reminder via Tencent Cloud voice call service.

[0042] The module features a closed-loop alert logic. If no action is taken within two hours of an alert being pushed, the alert level is automatically escalated and pushed to the next higher-level administrator. After action is taken, the system automatically verifies whether the corresponding parameters have returned to normal. If they still haven't returned after one hour, the alert is triggered again, ensuring that all alerts are addressed promptly. The average alert push delay is less than 10 seconds. This multi-level alert push module solves the problems of unreasonable alert push logic, easy neglect of important alerts, and untimely action in existing technologies.

[0043] This invention also discloses a dynamic feeding strategy generation module, which uses multi-layer correction logic to calculate the precise feeding amount. The specific calculation formula is as follows: , where F is the final total amount of feed given on that day, in kg; The basic feeding amount is obtained by multiplying the feeding coefficient corresponding to the growth cycle of salt field shrimp by the total weight of shrimp in the pond. The feeding coefficient is 7% during the shrimp larvae stage, 5% during the hepatopancreas transition stage, 4% during the growth stage, and 2.5% during the harvest stage. The water environment correction factor is 0.8 when dissolved oxygen is below 5 mg / L, 0.5 when water temperature is below 15℃, 0.9 when pH exceeds the range of 7.5 to 8.5, and 1 for all other cases; This is the disease risk correction coefficient. It is 0.9 when the disease risk index R value is higher than 0.5, and 1 in other cases. The feed residue correction factor is 0.9 when the underwater camera detects a feed residue percentage higher than 10% 2 hours after feeding the previous day, 1.05 when the residue percentage is 0%, and 1 for all other cases. The weather correction factor is 0.7 if there is a heavy rain or strong wind warning in the next 24 hours, and 1 otherwise. The module also includes a feeding time splitting logic, which divides the total daily feeding into four feedings at 6:00, 11:00, 17:00, and 22:00 to avoid wasting feed due to excessive feeding at a single time.

[0044] The feeding strategy generated according to this logic achieved a feed conversion ratio as low as 1.1 in benchmark ponds, 21% lower than the industry average of 1.4, significantly reducing feed waste and water pollution. The dynamic feeding strategy generation module solves the problems of existing feeding strategies relying on human experience, lacking accuracy, and resulting in serious feed waste.

[0045] This invention also discloses a multi-role permission management module, which is implemented based on the RBAC role access control model. It sets four standard roles, each with different access and operation permissions: the farmer role has full permissions for its own pond, including viewing all aquaculture data, adjusting control parameters, manually issuing control instructions, and exporting traceability information; the base administrator role has viewing permissions, batch control permissions, and personnel permission configuration permissions for all ponds within its jurisdiction, but cannot view the core privacy data of individual farmers (such as aquaculture costs and sales prices). The regulatory personnel role only has access to compliant data for all ponds within their jurisdiction, including water quality compliance status, disease reporting records, and test reports. They cannot control equipment or view the private data of the aquaculture operators. The consumer role only has access to publicly available traceability information for the corresponding batch of salt field shrimp and cannot access any other data. The module supports temporary authorization, allowing aquaculture operators to grant temporary viewing access to third parties such as technicians and animal health inspectors at any time. The authorization period is from a minimum of 1 hour to a maximum of 30 days, automatically revoking upon expiration. All permission operations are logged for traceability. This multi-role permission management module solves the problems of chaotic permission management and easy data privacy leaks associated with existing technologies.

[0046] Scenario Example 1: Individual Small-Scale Farming Application Scenario Description: There are three 10-mu (approximately 0.67 hectares) salt field shrimp farming ponds. Before 2022, only local threshold alarm devices were used, and the farmers relied entirely on their own experience to judge water quality and disease conditions. In that year, due to a nighttime low dissolved oxygen level that was not detected in time, all the shrimp in one pond died, resulting in a direct economic loss of 110,000 yuan. The annual farming loss rate was as high as 22%. At the same time, because there was no product traceability qualification, the product price was 2 yuan / jin lower than that of local farmers with traceability qualifications, resulting in a loss of nearly 60,000 yuan in income each year.

[0047] Technical Adaptation Details: This farmer deployed one edge acquisition terminal and a low-power solar-powered edge gateway at each pond, with a total hardware cost of 1850 yuan per pond. The cloud platform offered a version specifically for small-scale farmers, omitting features such as multi-pond benchmarking analysis and multi-role permission management that are unnecessary for them. The annual fee was only 399 yuan, making it affordable for small-scale farmers. The edge preprocessing unit filters and reduces noise according to preset parameter ranges specific to salt-field shrimp in the Binzhou area. Under normal farming conditions, the total amount of effective data uploaded to the cloud daily is less than 2MB, which can be met with a low-cost IoT card costing only 10 yuan per month, resulting in extremely low operating costs. The automated control unit only connects to existing aerators and automatic feeders in the ponds, eliminating the need for manual parameter threshold settings. All thresholds directly utilize standardized thresholds from the Binzhou region within the salt field shrimp's proprietary knowledge graph, adapting to local climate and water quality characteristics. The end-to-end traceability unit automatically generates product traceability QR codes for the corresponding ponds, which farmers can directly download, print, and affix to product packaging boxes, eliminating the need for additional manual information entry.

[0048] Specific manifestations of defect resolution in the scenario: After system deployment, throughout the entire aquaculture cycle in 2023, the dedicated analysis unit for salt field shrimp issued three early warnings of Vibrio contamination risk. Farmers promptly implemented full-pond disinfection with chlorine dioxide, preventing large-scale disease outbreaks and reducing the annual aquaculture loss rate to 4.7%, a decrease of 17.3 percentage points compared to the previous period. The local early warning logic built into the edge pretreatment unit can trigger a local alarm and automatically activate aerators within one second when dissolved oxygen falls below the threshold, preventing any shrimp deaths due to oxygen deficiency throughout the year and avoiding significant economic losses similar to those in 2022. The full-chain traceability QR code helped farmers' products smoothly enter three local high-end fresh food supermarkets, increasing the average selling price by 2.3 yuan / jin and generating an additional 72,000 yuan in revenue annually. After system deployment, all operational processes require no complex operations; farmers only need to view early warning information through a mobile app, adapting to the needs of small-scale farmers lacking digital operational capabilities. The core logic code snippet for edge-end local early warning in this scenario is as follows: / / Edge-end dissolved oxygen emergency early warning trigger core logic; float do_val = get_do_sensor_data(); / / Operation is only triggered if the collected values ​​are all below the safety threshold for 3 consecutive times to avoid occasional false judgments; if(do_val<2.0&&check_slide_window_valid(do_val, 3) == true){ trigger_local_sound_light_alarm(); send_ctr_cmd("aerator", "open"); }

[0049] Scenario Example 2: Large-scale salt field shrimp farming base Application Scenario Description: The facility comprises 217 contiguous salt-field shrimp farming ponds, covering a total area of ​​12,000 mu (approximately 800 hectares). From 2021 to 2022, a general-purpose aquaculture cloud platform was used for management. This platform, adapted to both freshwater fish and common marine shrimp farming logic, frequently misclassified the high salinity water quality required for salt-field shrimp farming as abnormal, resulting in a false alarm rate as high as 32%. The operations and maintenance team had to handle hundreds of invalid alerts daily, leading to extremely low management efficiency. Furthermore, the platform's feeding strategy did not consider the feeding characteristics of salt-field shrimp, resulting in a consistently high feed conversion ratio of 1.5, leading to significant feed waste. Additionally, the platform lacked dedicated traceability functions, failing to meet the supplier access requirements of high-end channels such as Hema Fresh and JD Fresh, thus limiting product premium and revenue growth.

[0050] Technical compatibility details: The base adopts a layered deployment architecture of "1 core management node + 11 edge gateways". Every 20 ponds share one high-configuration edge gateway. All core acquisition terminals are industrial-grade and highly corrosion-resistant. There is no need to replace the base's original aerators, feeders, and water adjustment equipment. The automatic control unit is directly compatible with the communication protocols of existing equipment. The hardware modification cost is reduced by 62% compared to deploying a brand new system.

[0051] The cloud platform has launched a dedicated version for the base, opening up all functional modules. A special analysis unit for salt-field shrimp has been added, loading weights from the aquaculture knowledge graph of the Lianyungang area. Growth parameter thresholds are adjusted for the CP SPF salt-field shrimp larvae stocked at the base. Early warning rules retain only high-risk emergency and important warnings, filtering out general warnings that do not affect production. The multi-pond benchmarking analysis module automatically categorizes ponds within the base by area, salinity, and shrimp larvae species, generating benchmark pools and pushing optimization suggestions daily. The end-to-end traceability unit directly connects to the existing databases of the base's processing workshop and cold chain logistics system, automatically synchronizing information such as processing time, cold chain temperature, and factory inspection reports, eliminating the need for manual secondary data entry.

[0052] Specific manifestations of defect resolution in this scenario: After system deployment, the false alarm rate dropped from 32% to 2.8%, the workload of the operations and maintenance team in handling early warnings decreased by 80%, and labor costs decreased by 45%. After the dynamic generation module for feeding strategies was launched, the feed conversion ratio at the base decreased from 1.5 to 1.12, saving 12.7 million yuan in feed procurement costs annually. The full-chain traceability function passed the supplier access review of Hema and JD Fresh, resulting in an average product selling price increase of 3 yuan / jin and an additional 39.8 million yuan in revenue annually. The multi-pond benchmarking analysis module selected the top 10% of ponds in terms of yield within the base as benchmarks. After other ponds adjusted their feeding and aeration strategies based on these benchmarks, the average yield per mu increased by 12.1%, and the overall aquaculture efficiency improved by 27%. The core logic code snippet for the dynamic calculation of feeding amount in this scenario is as follows: # Core logic for dynamic calculation of feeding amount; def calculate_daily_feed(pond_id, growth_stage): # Obtain the current total weight of shrimp in the pond and the basic feeding coefficient for the corresponding growth stage; shrimp_total_weight = get_pond_shrimp_weight(pond_id); base_coef = get_growth_stage_coef(growth_stage); base_feed = shrimp_total_weight base_coef; # Obtain the four types of correction coefficients; k1 = get_water_env_correction(pond_id); k2 = get_disease_risk_correction(pond_id); k3 = get_remaining_feed_correction(pond_id); k4 = get_weather_forecast_correction(pond_id); # Calculate the final feeding amount, which is divided into 4 feedings; final_feed = base_feed k1 k2 k3 k4; return final_feed / 4.

[0053] Table 1. Comparison of core aquaculture indicators before and after system deployment in two implementation scenarios. This table presents comparable data for two consecutive complete aquaculture cycles before and after system deployment in two implementation scenarios. The statistical process excluded non-system interference variables such as extreme weather, differences in shrimp fry varieties, and fluctuations in feed quality. All data comes from actual operational records. The changes in indicators show a significant decrease in aquaculture loss rate, false alarm rate, and feed conversion ratio in both scenarios. This addresses the core shortcomings of existing technologies, such as delayed response times, poor adaptability of general models leading to numerous false alarms, and low feeding precision resulting in feed waste. The increases in average yield per acre, selling price per kilogram, and annual revenue address the issues of insufficient product premium and low aquaculture efficiency caused by the lack of dedicated traceability capabilities in existing technologies. The data clearly demonstrates that this system has a significant benefit effect on both individual farmers and large-scale farms, with no scale adaptation threshold.

[0054] Table 2 Comparison of Core Performance between Salt Field Shrimp-Specific Analysis Model and General Aquaculture Model The performance data in this table is calculated based on a shared test set of 120,000 historical data points from salt-field shrimp farming across two implementation scenarios. This test set covers real records from different farming models, shrimp fry varieties, and salinity ranges across China's four major salt-field shrimp producing areas, ensuring representativeness across all scenarios. The comparative results show that the salt-field shrimp-specific analysis model significantly outperforms the two general-purpose farming models in all core dimensions. Specifically, its disease prediction accuracy is 22.9 percentage points higher than the general marine shrimp model, addressing the shortcomings of existing general models that were not trained specifically for the characteristics of salt-field shrimp and lacked sufficient analytical precision. Furthermore, its complete adaptability to high-salinity environments solves the problem of general models mistakenly classifying high-salinity water quality in salt-field shrimp farming as abnormal and resulting in a high false alarm rate. This provides reliable analytical support for salt-field shrimp farming across the entire region.

[0055] Table 3. Calculation of Input-Output Ratio of Cloud Platform Deployment under Different Aquaculture Scales This table's calculations are based on the average hardware procurement price, cloud platform service fee standards, IoT traffic costs, and average revenue per unit area of ​​the four major salt-field shrimp producing areas in China. All parameters use industry medians, ensuring industry applicability. The results show that the static investment payback period for all three types of aquaculture entities is less than two months, with an input-output ratio exceeding 1:8. This addresses the shortcomings of existing digital aquaculture systems, which are too costly to deploy and unaffordable for small and medium-sized farmers. Specifically, the total investment per cycle for individual farmers is less than 7,200 yuan, far below the average annual loss in the aquaculture industry, further lowering the digital transformation threshold for small and medium-sized farmers. This demonstrates high feasibility for widespread adoption and can quickly cover aquaculture entities at all levels across the industry.

[0056] refer to Figure 1This diagram illustrates the system's overall architecture, showcasing four core levels from bottom-level perception to top-level interaction. The aquaculture environment perception module collects raw data at the pond; the edge data preprocessing module performs real-time cleaning and preliminary filtering; the cloud platform service center acts as the brain, using knowledge graphs and analytical models to generate control instructions and process traceability data; finally, the terminal interaction module enables human-machine collaboration. This process forms a closed-loop business chain of "perception-analysis-decision-execution". refer to Figure 2 This diagram details the data processing steps before it reaches the cloud. The system encompasses four sensing dimensions: water quality, weather, behavior, and bait. At the edge, the system dynamically adjusts the window size of the sliding window filtering algorithm based on the acquisition frequency to effectively eliminate impulse noise. After passing consistency checks, the data is encapsulated in standard JSON format and transmitted efficiently using 5G / NB-IoT networks, ensuring the accuracy and real-time performance of the data analyzed in the cloud.

[0057] refer to Figure 3 This diagram highlights the cloud platform's "core computing power." A dedicated knowledge graph provides industry expert benchmarks, while a multi-dimensional analysis model set handles prediction and early warning. Through water environment prediction, disease early warning, and feeding optimization models, the system can calculate the optimal aquaculture plan. Furthermore, the pond benchmarking analysis unit provides targeted optimization suggestions to farmers through horizontal data comparison, improving overall regional yields. refer to Figure 4 This diagram illustrates the system's data management and reliable traceability mechanism. A cold / hot data separation storage strategy is employed to ensure high-performance retrieval of real-time data and low-cost preservation of historical data. Simultaneously, the system generates a traceability QR code for each batch of salt-field shrimp, integrating data from the shrimp larvae to the harvest. This not only provides audit reports for management departments but also offers consumers reliable quality verification, constructing a complete food safety chain.

[0058] refer to Figure 5 This diagram illustrates the system's human-computer interaction logic and control safety mechanisms. Commands generated by the cloud platform are not executed directly but are first pushed to the terminal for human confirmation. The system has a 15-minute waiting threshold; if no one intervenes, it executes automatically; if there is intervention, it adjusts according to human input, achieving flexible control that is "primarily intelligent and secondarily human." Simultaneously, through role-based access control, it ensures that breeding personnel, managers, and maintenance personnel operate safely within their respective areas of authority.

[0059] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A cloud platform application system for intelligent farming of salt field shrimp, characterized in that, It includes a breeding environment sensing module, an edge data preprocessing module, a cloud platform service center, and a terminal interaction module that are connected in sequence. The aquaculture environment sensing module is deployed at corresponding monitoring points in each salt field shrimp farming pond to collect multi-dimensional environmental parameters and aquaculture behavior-related data throughout the entire salt field shrimp farming cycle. The edge data preprocessing module is deployed on the local edge gateway device in Tangkou. It is used to perform noise filtering, format unification and outlier removal on the collected raw data before uploading it to the cloud platform service center. The cloud platform service center has a built-in knowledge graph and multi-dimensional analysis model group specifically for salt field shrimp farming, which is used to perform distributed storage, intelligent analysis and generate corresponding scientific farming management instructions for all uploaded farming data. The terminal interaction module is used to push control instructions, real-time monitoring data and risk warning information to aquaculture participants with different roles, and at the same time receive manual control instructions issued by aquaculture participants and synchronize them to the cloud platform service center.

2. The cloud platform application system for intelligent salt field shrimp farming according to claim 1, characterized in that, The aquaculture environment sensing module includes a water environment sensing unit, a pond meteorological sensing unit, an aquaculture behavior sensing unit, and a feed residue detection unit. The water environment sensing unit is deployed at different depths of 0.5 to 1.5 meters underwater in the pond to collect parameters such as salinity, water temperature, dissolved oxygen, pH value, ammonia nitrogen concentration, and nitrite concentration in the aquaculture water. The pond mouth meteorological sensing unit is deployed in an open location with no obstruction around the pond mouth, and is used to collect temperature, air pressure, light intensity, rainfall and wind speed parameters in the pond mouth area; The aquaculture behavior sensing unit collects data on the activity frequency, molting cycle, and population distribution of salt field shrimp using underwater high-definition camera equipment and infrared sensors. The bait residue detection unit collects bait residue parameters in different areas of the pond bottom using an underwater near-infrared spectroscopy sensor.

3. The cloud platform application system for intelligent salt field shrimp farming according to claim 1, characterized in that, The edge data preprocessing module incorporates a sliding window filtering algorithm and data consistency verification rules. First, a sliding window filtering algorithm is used to filter the impulse noise in the raw data. The window size of the sliding window filtering algorithm can be adaptively adjusted according to the data acquisition frequency of the pond. When the acquisition frequency is higher than once per minute, the window size is set to 5, and when the acquisition frequency is lower than once per minute, the window size is set to 3. Then, based on the reasonable threshold range of salt field shrimp farming parameters, outliers are removed from the data. Finally, all data is converted into a unified JSON format and uploaded to the cloud platform service center via NB-IoT or 5G network. At the same time, the original data is stored locally for more than 7 days for subsequent data backtracking and local caching in case of platform failure.

4. A cloud platform application system for intelligent salt field shrimp farming according to claim 1, characterized in that, The cloud platform service center is equipped with a distributed storage unit and adopts a cold and hot data separation storage strategy. Historical aquaculture data collected over a period of more than 3 months is stored in low-cost cold storage nodes, while active aquaculture data collected over nearly 3 months and real-time data are stored in high-performance hot storage nodes. At the same time, all data is backed up in three copies in different locations. If any storage node fails, the system will automatically switch to other copies to read the data.

5. A cloud platform application system for intelligent salt field shrimp farming according to claim 1, characterized in that, The knowledge graph for salt field shrimp farming is pre-loaded with suitable environmental parameter ranges for salt field shrimp at different growth stages, growth characteristics under different salinity gradients, characteristic parameters of common diseases, feed standards under different farming models, and emergency management plans under different weather conditions. The knowledge graph allows farming participants to upload their local farming experience for iterative updates.

6. A cloud platform application system for intelligent salt field shrimp farming according to claim 1, characterized in that, The multi-dimensional analysis model group includes a water environment prediction model, a disease early warning model, a feed feeding optimization model, and a profit calculation model. The water environment prediction model predicts the trend of water environment parameter changes in the next 24 to 72 hours based on historical environmental data and official meteorological forecast data. The disease early warning model outputs the probability of disease occurrence and the early warning level based on the behavioral data of salt field shrimp and abnormal deviations in aquatic environmental parameters; The feed optimization model combines the growth stage of salt field shrimp, the amount of feed residue at the bottom of the pond, and water environment parameters to output the optimal feeding amount, feeding time, and feeding location. The profit calculation model calculates the expected breeding profit based on the aquaculture input data, salt field shrimp growth cycle data, and the current wholesale market price of salt field shrimp.

7. A cloud platform application system for intelligent salt field shrimp farming according to claim 6, characterized in that, After the cloud platform service center generates a control command, it first pushes the command to the terminal interaction module for confirmation by the aquaculture participants. If no manual intervention command is received within the preset 15-minute time threshold, the control command is automatically sent to the corresponding aerator, feeder, water exchange equipment and other execution terminals in the pond to perform the corresponding control operation. If a manual adjustment command is received, the command is sent to the corresponding execution terminal according to the manually adjusted command.

8. A cloud platform application system for intelligent salt field shrimp farming according to claim 1, characterized in that, The cloud platform service center has a built-in full-cycle traceability unit that generates a unique traceability QR code for each batch of salt field shrimp. The traceability code is associated with the full-process data of the corresponding batch of salt field shrimp, from shrimp seedling release, breeding management, disease prevention and control, feed use to final harvesting and marketing. Consumers can scan the traceability code to query all breeding information and test reports of the corresponding batch of salt field shrimp.

9. A cloud platform application system for intelligent salt field shrimp farming according to claim 1, characterized in that, The terminal interaction module assigns different operating permissions based on the roles of the participants in aquaculture. Ordinary aquaculture personnel can only view the monitoring data of their own ponds and receive control instructions. Aquaculture managers can view the summary data of all their subordinate ponds and issue manual control instructions. Platform maintenance personnel can adjust and optimize the platform's parameter configuration, model rules, and knowledge graph content.

10. A cloud platform application system for intelligent salt field shrimp farming according to claim 1, characterized in that, The cloud platform service center has a built-in pond benchmarking analysis unit, which can conduct horizontal comparative analysis of environmental parameters, breeding cycle, disease incidence, unit yield and profit data of multiple breeding ponds in the same area, and output breeding optimization suggestions for each pond for reference by breeding participants. At the same time, it can generate a regional overall breeding data statistical report for use by industry management departments.

Images