Aquaculture disease multi-dimensional traceability internet of things monitoring system and prevention and control closed loop method

Abstract

The application discloses an aquaculture disease multi-dimensional traceability Internet of Things monitoring system and a prevention and control closed-loop method, relates to the technical field of aquaculture disease prevention and control, and comprises the following steps: deploying an Internet of Things sensor array at a pond mouth, continuously collecting water quality parameters, aquaculture operation and seedling feed batch information, constructing a multi-source correlation database after aligning unified timestamps, and when a disease is triggered, three-dimensional traceability is realized by tracing back the environment and operation track from time, searching for the correlation pond mouth disease time sequence from space, and comparing the health status of seedling feed from batch. Evidence is divided into two levels, i.e., a main cause and an auxiliary cause, the main cause is independently concluded, the auxiliary cause is focused on the cause, a structured report is output, linkage is performed with insurance execution and archiving, and a monitoring-traceability-prevention and control-feedback closed loop is formed. Through the above method, the application realizes multi-dimensional accurate traceability of diseases, solves the problems of traditional data isolation and difficult cause finding, realizes efficient traceability, credible cause finding and traceable responsibility, supports scientific prevention and control, reduces losses, and promotes the digital upgrading of aquaculture disease management.

Description

Technical Field

[0001] This invention relates to the field of aquaculture disease prevention and control technology, and in particular to an Internet of Things (IoT) monitoring system and closed-loop prevention and control method for multidimensional tracing of aquaculture diseases. Background Technology

[0002] Aquaculture is an important sector of agricultural production in my country, and disease control is one of the core issues affecting aquaculture success. Current disease management models in aquaculture generally suffer from the following two fundamental flaws: First, the existing disease management system can only respond after the fact and cannot trace the cause of the disease. After fish farmers discover visible abnormalities in their fish populations, they contact veterinarians or distributors for diagnosis and medication advice. This entire response chain is time-consuming, while the spread of aquatic diseases can often cause large-scale losses within 24 hours. A deeper problem is that even after a treatment is completed, neither the fish farmers nor the veterinarians can determine the true cause of the disease—whether it is due to water quality deterioration, improper handling, external introduction, or batch issues with seedlings or feed. Because the cause is unknown, it is extremely common for the same disease to recur in the same fish farm, leaving fish farmers in a state of constant reactive response.

[0003] Second, existing IoT monitoring systems suffer from isolated data, making it difficult to effectively correlate data from multiple sources. While some IoT monitoring products for aquaculture are available on the market that can collect water quality parameters in real time and issue early warnings for exceeding limits, these systems have significant technical limitations: First, the monitoring data only covers water quality environmental parameters; aquaculture operation records and seedling / feed batch information are outside the system and cannot be included in unified management. Second, there is a lack of a unified time alignment mechanism between data from different sources and with different collection frequencies, making it impossible to correlate and query environmental parameters, operational behaviors, and batch status on the same time dimension. Third, the data analysis capabilities of existing systems are limited to threshold judgments for single ponds and single parameters; they cannot retrieve spatial correlation information across ponds or compare historical health status across batches, and the accumulation of data cannot be transformed into effective traceability capabilities.

[0004] The two aforementioned deficiencies have led to a situation where the treatment of aquatic diseases remains at the level of "knowing that the disease is present," without addressing questions such as "why the disease occurred, where it came from, and what batch it was from." This not only results in the overuse of animal health products and waste of resources but also leaves a long-standing lack of objective evidence for attributing responsibility among aquaculture farmers, fish veterinarians, and animal health companies.

[0005] Currently, there is no systematic method for tracing the source of diseases in aquaculture, nor is there an IoT monitoring system that can unify and correlate multi-source data across time, space, and batch dimensions for disease attribution. Therefore, it is necessary to provide a system and method that can reconstruct the complete pathogenesis chain and form a closed-loop prevention and control system when a disease occurs.

[0006] To address these issues, there is an urgent need for an IoT monitoring system and a closed-loop prevention and control method for multidimensional tracing of aquatic diseases. Summary of the Invention

[0007] To address the aforementioned issues, this application proposes an IoT monitoring system and a closed-loop prevention and control method for multidimensional tracing of aquatic diseases. This addresses the technical problems in existing technologies where aquatic disease treatment can only be a reactive response and cannot trace the cause of the disease, as well as the isolation of multi-source heterogeneous data that cannot be effectively correlated.

[0008] On the one hand, this application proposes an Internet of Things (IoT) monitoring system for multi-dimensional tracing of aquatic diseases, including: The data acquisition module is deployed at the aquaculture pond. It continuously collects water quality environmental parameters, aquaculture operation records and seedling and feed batch information through an Internet of Things (IoT) sensor array to obtain multi-source heterogeneous data. The multi-source heterogeneous data is then uploaded to the data association storage module in real time through the IoT transmission protocol. The data association storage module aligns the above-mentioned multi-source heterogeneous data through a unified timestamp alignment protocol, and forms a multi-source association database with the pond number and time unit as the primary key. The data association storage module also stores a pond relationship table for spatial tracing. The multidimensional tracing module includes time-dimensional tracing, spatial-dimensional tracing, and batch-dimensional tracing. When a disease event is triggered, the tracing is carried out along the three dimensions of time, space, and batch. The evidence in each dimension is divided into two levels: primary factors and secondary factors. The primary factors are determined independently, and the secondary factors further focus on the source of the disease after the attribution direction is established by the primary factors. Finally, a structured evidence report sorted by confidence level is output. The prevention and control closed-loop module pushes the source tracing report to the fish doctor's terminal. After the fish doctor reviews it, an electronic prescription is issued. The animal health department is linked through the Internet of Things interface to carry out the work, and the treatment process and effect data are fully archived to form a closed operation link of monitoring, source tracing, prevention and control and feedback.

[0009] Preferably, the water quality environmental parameters are continuously collected by a sensor array deployed at the pond at a frequency of minutes, and uploaded in real time to the data association and storage module via an Internet of Things transmission protocol. Specifically, these parameters include: instantaneous values, rate of decrease, and daily fluctuation range of dissolved oxygen; instantaneous values ​​and rate of change of water temperature per unit time; instantaneous values ​​of ammonia nitrogen and nitrite and their cumulative trends at continuous collection points; instantaneous values ​​of pH and their shift direction and rate; and instantaneous values ​​of turbidity and abrupt changes. Each water quality environmental parameter is collected along with a collection timestamp and pond number, and stored as a structured numerical record. The aquaculture operation records are manually entered by aquaculture personnel or automatically triggered by equipment, with the time of event occurrence as the node. Specifically, they include: the time of each feeding, the amount of feed fed, and the batch number of feed used; the time of each water change or water addition operation and the amount of water; the start and stop time and operating status of the aeration equipment; and the time of each medication administration, the type of medication, and the dosage. The above records are stored in an event-based structured record, and each record is accompanied by an event timestamp and pond number. The batch information of seedlings and feed is entered once when the seedlings are put into the pond or the feed is put into the warehouse. Specifically, it includes: the seedling supplier, place of origin, time of entering the pond, batch number, and stress conditions such as transportation time, temperature difference and density when the seedlings are introduced; and the brand, batch number, time of entering the warehouse, time of opening and storage temperature and humidity conditions of the feed. The above information is stored statically in batches and is linked to the feeding records and pond numbers in the aquaculture operation records through the batch number.

[0010] Preferably, water quality parameters, including dissolved oxygen, water temperature, ammonia nitrogen, nitrite, pH, and turbidity, are continuously collected by a sensor array at a frequency of minutes and uploaded in real time via an IoT transmission protocol. Each collection is accompanied by a collection timestamp and pond number, and stored as a structured numerical record. Aquaculture operation records, including feeding time, feeding amount, water change operations, aeration equipment operation status, and medication records, are entered with the event occurrence time as the node, and each record is accompanied by an event timestamp and pond number. Seedling and feed batch information, including seedling source, pond entry time, batch number, and stress conditions, as well as feed brand, batch number, storage time, opening time, and storage conditions, are statically entered by batch and linked to feeding records and pond numbers via batch numbers. The beneficial effect is that these three types of data comprehensively cover the environmental, human, and material factors contributing to disease, providing a sufficient data foundation for multi-dimensional traceability.

[0011] Preferably, the data association and storage module uses minutes as the smallest time unit to perform unified timestamp alignment on three types of data: water quality environmental parameters, aquaculture operation records, and seedling and feed batch information. The specific alignment method is as follows: For water quality environmental parameters, each collection record is already accompanied by a minute-level timestamp, which is directly mapped to the corresponding time unit without conversion; For aquaculture operation records, each event-type record is mapped to the corresponding time unit interval by its event timestamp. Multiple operation records that occur in the same time unit are stored in parallel under that time unit. For seedling and feed batch information, the batch number is used as an index to map the effective start time to end time of the batch into a continuous time unit interval. Each time unit in the interval carries the status mark of the corresponding batch, so that the batch information can be directly associated with and queried with the environmental parameters and operation records in the same time unit. After the above alignment process, the three types of data are integrated into a multi-source associated database using the pond number and time unit as the joint primary key. Any record in the database corresponds to a complete snapshot of the status of a specific pond within a specific time unit, including the instantaneous values ​​and statistical characteristics of all environmental parameters, all operation event entries, and the currently valid seedling and feed batch markers within that time unit. Environmental changes, operational behaviors, and batch status at any point in time can be completely associated and queried in the same record using the pond number and time unit, providing a unified data retrieval foundation for the multi-dimensional traceability module. The unified timestamp alignment mechanism solves the technical problem that multi-source heterogeneous data cannot be associated due to different collection frequencies, enabling the accurate restoration of the temporal relationship of cross-type data.

[0012] Preferably, when a disease event is triggered, the multi-dimensional tracing module initiates a parallel tracing search in three dimensions to the multi-source association database, using the disease confirmation time point and the number of the affected pond as indexes. The three dimensions are time dimension, space dimension and batch dimension. The tracing factors in the three dimensions are divided into primary factors and auxiliary factors. The primary factors have independent conclusion-making ability, while the auxiliary factors further focus on the source of the disease after the primary factors have established the attribution direction. After summarizing all evidence items across the three dimensions, the confidence level is calculated based on the quantity and strength of supporting evidence for each etiological pathway. The results are then sorted and a structured source tracing report is output. The two-level classification of primary and secondary factors ensures that the source tracing conclusions possess both conclusive and refined attribution capabilities. The confidence level ranking ensures the interpretability of the conclusions, avoiding the unaccountability issues associated with black-box, single-conclusion approaches.

[0013] Preferably, the time dimension is based on the time point of disease confirmation, and the historical data of the diseased ponds within the preset time window are traced back to check the temporal relationship between the trajectory of environmental parameter changes and the aquaculture operation behavior. The spatial dimension starts with the affected ponds and searches for related ponds that have water system connectivity or geographical proximity to them, comparing their disease outbreak timeline with the usage records of shared facilities. The batch dimension uses the currently valid seedling and feed batch numbers of the diseased ponds as an index to search all ponds using the same batch within the system and compare their corresponding health status.

[0014] Preferably, the time-dimensional traceability factors include primary time-based traceability factors and secondary time-based traceability factors, as detailed below: Key factors in time-based tracing include: dissolved oxygen decline trajectory, ammonia nitrogen and nitrite accumulation trends, sudden changes in water temperature, abnormal deviations in feeding amounts, and the temporal relationship of medication records. Dissolved oxygen decline trajectory: Examine the rate of decline, duration, and intraday fluctuation range of dissolved oxygen within the retrospective window to see if there is an abnormal narrowing. A sustained low or rapid decline independently constitutes a hypoxia-related pathogenic factor. Ammonia nitrogen and nitrite accumulation trends: Examine the monotonic increase and slope changes at continuous sampling points within the retrospective window. Sustained accumulation exceeding the pathogenic threshold independently constitutes a water quality deterioration factor. Water temperature fluctuation amplitude: Examine whether the daily temperature difference within the retrospective window exceeds the set threshold or whether there are any unnatural jumps within a short period. Exceeding the limit independently constitutes a stress-related pathogenic factor. Abnormal feeding deviation: Examine the deviation of each feeding amount within the retrospective window from the historical average of the pond. Significantly higher deviations independently constitute an overfeeding-related pathogenic factor. Medication record time sequence: Examine the interval between the type and dosage of the previous medication and the current onset time. An interval matching the action cycle of a specific drug independently constitutes an improper medication-related secondary infection factor. The auxiliary factors for time-based tracing include sudden changes in turbidity, the direction and rate of pH shift, abnormal feeding times, and the operating status of aeration equipment.

[0015] Sudden turbidity changes: After a decrease in dissolved oxygen or accumulation of ammonia nitrogen has been established as the primary attribution, the synchronous change in turbidity is used to distinguish the specific type of water quality deterioration. Increased turbidity points to organic matter decomposition or overfeeding, while a sudden drop in turbidity points to algal community collapse. pH shift direction and rate: After an abnormal water quality has been established as the primary attribution, the pH shift direction is used to further determine the current state of the water body. Increased pH combined with increased dissolved oxygen points to excessive algal growth, while decreased pH combined with increased turbidity points to large-scale decomposition of organic matter. Abnormal feeding time: After an abnormal deviation in feeding amount has been established as the primary attribution, whether the feeding time falls during a period of low dissolved oxygen is used to distinguish between simple overfeeding and overfeeding combined with inappropriate timing. Aeration equipment operation status: After a decrease in dissolved oxygen has been established as the primary attribution, the start-up and shutdown records and fault records of the aeration equipment are used to distinguish whether the decrease in dissolved oxygen is due to natural environmental deterioration or equipment failure, directly affecting the attribution of responsibility.

[0016] Preferably, the spatial dimensional tracing factors include primary spatial tracing factors and secondary spatial tracing factors, as detailed below: The main factors for spatial source tracing include the shared relationship of the intake channel and the disease records of upstream ponds, the chronological order of disease occurrence in adjacent ponds, and the record of the order of use of shared tools; The shared inlet channel relationship and upstream pond disease records: Check whether upstream ponds sharing the same inlet channel with the affected pond have disease records in the retrospective window. If so, and the interval between the inlet operation time of the affected pond and the current onset time matches the incubation pattern of the pathogen, it independently constitutes a disease attribution due to external transmission via the inlet channel. The chronological order of onset time in adjacent ponds: Check the onset time records of ponds geographically adjacent to the affected pond in the retrospective window. If the onset times of multiple adjacent ponds show a clear chronological order and the direction is consistent with the water flow path or operation path, it independently constitutes a disease attribution due to spatial diffusion. Records of shared tool usage sequence: Check the usage sequence records of nets, aeration equipment, feeding tools, etc., among multiple ponds. If the usage sequence matches the onset time sequence of the affected pond, it independently constitutes a disease attribution due to mechanical transmission via shared tools. Spatial source tracing auxiliary factors include the relationship between the drainage direction and downstream ponds, and records of the introduction of foreign fish fry or broodstock.

[0017] The relationship between drainage direction and downstream ponds: After the disease has been confirmed in upstream ponds and the shared inlet channel has been established as the main attribution direction, the drainage direction of the diseased ponds and the distribution of its downstream associated ponds are used to further predict potential disease spread paths and issue preventive warnings to downstream associated ponds. Records of the introduction of foreign fish fry or broodstock: When the time of onset is found to be highly consistent with the time of introduction of foreign fish groups, the introduction records serve as an auxiliary factor to corroborate the batch attribution evidence at the batch level. This is used to further focus the attribution of exogenous introduction to the introduction batch itself and distinguish between two exogenous introduction paths: transmission through the water system and transmission carried by the introduced batch.

[0018] Preferably, the traceability factors at the batch level include primary batch traceability factors and secondary batch traceability factors, as detailed below: The main factors for batch traceability include the clustered disease records of the same batch of seedlings in multiple unrelated farms, and the clustering of diseases within the scope of use of the same batch of feed; For clustered disease records of the same batch of seedlings in multiple unrelated farms, the system is indexed by the currently valid seedling batch number of the affected pond. All ponds using the same batch of seedlings are searched within the system. If other farms with no water system connection, geographical proximity, or shared facilities with the affected pond show similar diseases within a similar time period, after ruling out the possibility of spatial transmission, this independently constitutes a disease attribution for the seedling batch quality problem. For disease clustering within the usage range of the same batch of feed, the system is indexed by the currently valid feed batch number of the affected pond. All ponds using the same batch of feed are searched within the system. If other ponds with no spatial connection to the affected pond show similar diseases within a similar time period, this independently constitutes a disease attribution for the feed batch quality problem. Batch traceability auxiliary factors include abnormal turbidity response after feeding, feed storage condition records, and stress assessment records at the time of seedling introduction.

[0019] Abnormal turbidity response after feeding: After establishing the feed batch as the primary attribution direction, examine the response relationship between feeding operations and turbidity changes in affected ponds during the period of using this batch of feed. If the turbidity shows an abnormal increase beyond the historical normal range after feeding, it is used to further distinguish whether the feed batch problem is due to abnormal water solubility of the feed itself or improper feeding operations, focusing on the source of the attribution. Feed storage condition records: After establishing the feed batch as the primary attribution direction, examine the batch's storage time, opening time, storage temperature and humidity, and usage cycle. If the storage conditions are abnormal for a long period, it is used to distinguish whether the problem originates from the quality of the factory batch or improper storage, directly affecting the direction of responsibility attribution. Stress assessment records during seedling introduction: After establishing the seedling batch as the primary attribution direction, examine the stress condition records such as transportation time, temperature difference, and density during the introduction of this batch of seedlings. This is used to distinguish between two situations: the batch itself carrying pathogens and the batch itself being healthy but becoming infected due to decreased immunity caused by stress, providing a basis for determining the responsibility of the seedling supplier.

[0020] The three-dimensional factor design covers the main pathogenic pathways of aquaculture diseases, and the differentiated positioning of the main and auxiliary factors enables the system to output effective source tracing conclusions under different evidence conditions.

[0021] Preferably, the prevention and control closed-loop module includes: Fisherman Doctor's End Review Unit: Fisherman Doctor reviews the complete chain of evidence to make a professional diagnosis of the cause of the disease; Electronic prescription issuance unit: A comprehensive electronic prescription is generated based on the cause of the disease, pests, products, dosage, time window, and prevention recommendations, and the fisherman's signature is recorded. IoT-linked execution unit: The IoT-linked execution unit is used for automatically sending or manually executing operation confirmation records; Effect monitoring and data archiving unit: Records data evaluation during the recovery period. If the expected results are not met, a secondary warning is triggered. If the expected results are met, the entire process data is archived and preserved.

[0022] On the other hand, this application proposes a closed-loop prevention and control method for an IoT monitoring system for multi-dimensional tracing of aquatic diseases, including: During the source tracing report receiving phase, the prevention and control closed-loop module receives the structured evidence report output by the multi-dimensional source tracing module. The report lists multiple etiological pathways sorted by confidence level. Each pathway is accompanied by specific evidence items supporting the pathway and their source dimensions, fully preserving the accessibility of the evidence chain. During the prescription issuance stage, the structured evidence report is pushed to the fish doctor's terminal through a standardized interface. After reviewing the complete evidence chain, the fish doctor with professional qualifications makes a diagnosis based on professional judgment and issues an electronic prescription that includes the confirmed diagnosis, the type and dosage of animal health products, the medication time window, and preventive recommendations for related risk ponds. The system does not automatically generate prescriptions, and the right to issue prescriptions belongs to the fish doctor. During the execution phase, the electronic prescription is sent to the pond execution terminal via the Internet of Things interface; if there is automatic delivery equipment on site, quantitative delivery is triggered directly; if manual operation is required, an execution instruction containing specific operation parameters is pushed to the aquaculture personnel, and a confirmation receipt is required after the operation is completed. During the effect monitoring phase, after the animal protection measures are implemented, the system continuously collects water quality environmental parameters and biological behavior data during the recovery period, compares them with the historical baseline before the onset of the disease, and assesses the recovery trend. If the recovery trend does not meet expectations, a secondary warning is triggered, and the fish veterinarian is notified to reassess the prescription. During the archiving phase, the source tracing report, electronic prescription, execution record, and recovery period data of this disease event are archived as a complete disease event record into a multi-source associated database, forming a traceable digital certificate, and serving as historical data accumulation for subsequent source tracing and comparison in the same breeding scenario.

[0023] The closed-loop process clearly defines the boundaries between system-assisted decision-making and the professional judgment of fish doctors, ensuring both the professionalism and compliance of prevention and control measures, and achieving complete retention of the responsibility chain through data archiving.

[0024] In summary, the IoT monitoring system and closed-loop prevention and control method for multi-dimensional tracing of aquatic diseases of the present invention have the following advantages compared with traditional technologies: 1. This application uses minute-level time units as the benchmark, unifies three types of data timestamps, and constructs a pond-time joint primary key database, which allows for real-time correlation and querying of environmental changes, operational behaviors, and batch status, completely solving the problem of data fragmentation, providing a unified, complete, and accurate data foundation for multi-dimensional traceability, and significantly improving the accuracy and efficiency of traceability; 2. After the disease is triggered, this application performs a three-dimensional synchronous search: time-based backtracking of the environment and operation sequence, spatial search of water systems and adjacent ponds, and batch comparison of seedling and feed distribution; the main factors are independently determined, and the auxiliary factors focus on details, and reports are output according to the confidence level, covering the entire pathogenesis of the disease, including the environment, human factors, and materials, accurately locating the cause of the disease, avoiding blind treatment, and reducing disease recurrence; 3. The sensor array in this application collects water quality data at the minute level, and the equipment / manual staff simultaneously record operation and batch information; when the network is unstable, 24 hours of data are cached locally and automatically retransmitted after recovery to ensure data continuity. The complete data covers the entire disease cycle, providing all-time and complete data support for source tracing and avoiding source tracing deviations due to data loss. 4. This application's traceability report directly links to Yu Doctor, which professionally issues electronic prescriptions; IoT-linked automatic / manual execution; real-time monitoring of recovery effects, with abnormalities triggering secondary warnings; full data archiving to form digital vouchers, clarifying the boundaries between system assistance and doctor decision-making, improving the professionalism and efficiency of treatment, completely preserving the chain of responsibility, and solving the problem of lack of evidence for attributing responsibility; 5. The source tracing report, prescription, execution record, and effect data for each disease event in this application are fully archived and included in the multi-source association database. When the disease occurs in the same farm or pond of the same batch, the historical data can be used for source tracing comparison in spatial and batch dimensions, continuously enriching the source tracing evidence library, so that the source tracing accuracy and reliability of the system can be continuously improved, forming a virtuous cycle of becoming more and more accurate with use.

[0025] The technical method of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the overall architecture of the multi-dimensional traceability IoT monitoring system for aquaculture diseases of the present invention; Figure 2 This is a schematic diagram of the data categories in the data acquisition module of the present invention; Figure 3 This is a schematic diagram of the multi-source data timestamp alignment and associated storage structure of the present invention; Figure 4 This is a schematic diagram of the three-dimensional source tracing reasoning framework of the present invention; Figure 5 This is a schematic diagram illustrating the hierarchical expansion of factors in the three-dimensional tracing of the present invention. Figure 6 This is a schematic diagram of the closed-loop operation link for epidemic prevention and control in this invention. Detailed Implementation

[0027] The technical method of the present invention will be further described below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of this application.

[0028] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the scope of this application and its application or use.

[0029] Techniques, systems, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, they should be considered part of the instruction manual.

[0030] In all the examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0031] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0032] This invention provides an IoT monitoring system and closed-loop prevention and control method for multi-dimensional tracing of aquatic diseases, with specific embodiments as follows: Example 1 like Figure 1 As shown, the IoT monitoring system for multi-dimensional tracing of aquaculture diseases of the present invention consists of four modules connected in sequence: a data acquisition module, a data association and storage module, a multi-dimensional tracing module, and a prevention and control closed-loop module. Data is transmitted between the modules through a standardized data interface.

[0033] The data acquisition module is deployed on one side of the aquaculture pond, directly facing the aquaculture environment. It is responsible for continuously collecting three types of raw data: water quality parameters, aquaculture operation records, and batch information of seedlings and feed. The collected data is then uploaded to the data association and storage module in real time via an Internet of Things (IoT) transmission protocol. In aquaculture environments with unstable network signals, the data acquisition module can cache data locally and automatically re-upload it once the network is restored, ensuring data continuity unaffected by network fluctuations.

[0034] The data association and storage module receives three types of raw data from the data acquisition module, performs unified timestamp alignment processing, and integrates them into a multi-source associated database using the pond number and time unit as the joint primary key. It then provides a unified data retrieval interface to the multi-dimensional traceability module. The data association and storage module also maintains a pond basic information table, recording the water system connectivity and geographical adjacency relationships between ponds, which the multi-dimensional traceability module can access during spatial traceability.

[0035] Upon receiving a disease event trigger signal, the multi-dimensional traceability module initiates a parallel traceability search across three dimensions, indexed by the disease confirmation time and the affected pond number. The search results are then categorized into primary and secondary factors, and the confidence level is calculated to generate a structured traceability report, which is then transmitted to the prevention and control closed-loop module. Disease event trigger signals can be manually reported by aquaculture personnel or automatically triggered by the data acquisition module when persistently abnormal water quality parameters are detected.

[0036] After receiving the structured traceability report, the prevention and control closed-loop module pushes it to the fish doctor terminal through a standardized interface. After the fish doctor reviews and confirms the report and issues an electronic prescription, the prescription is sent to the pond execution terminal through the Internet of Things interface, linking animal health execution and continuously collecting recovery period data. All process data is archived to the data association storage module, completing a complete closed-loop operation.

[0037] The data flow between the modules of this system is a combination of unidirectional sequential transmission and end-point return: the forward flow is the acquisition module, storage module, traceability module, and closed-loop module; the return flow is the closed-loop module writing archived data back to the storage module, so that the handling results of each disease event become historical data accumulation for the next traceability.

[0038] Example 2 like Figure 2 As shown, the data collected by the data acquisition module is divided into three categories. The acquisition methods, recording formats and storage methods of each category are different, which together constitute the original data foundation for multidimensional traceability.

[0039] The first category consists of water quality environmental parameters. A sensor array is deployed at fixed underwater locations in the pond, continuously collecting six parameters—dissolved oxygen, water temperature, ammonia nitrogen, nitrite, pH, and turbidity—at a frequency of minutes. The dissolved oxygen sensor records both instantaneous values ​​and daily fluctuations, used for subsequent source tracing to determine the rate of decline and the narrowing of fluctuations; the water temperature sensor records instantaneous values ​​and the magnitude of change per unit time, used to identify sudden events; the ammonia nitrogen and nitrite sensors record instantaneous values, and the data association and storage module calculates the cumulative trend at continuous collection points after receiving the data; the pH sensor records instantaneous values ​​and the direction of change; and the turbidity sensor records instantaneous values ​​and sudden events. Each collection of six parameters is accompanied by a collection timestamp and pond number, uploaded and stored as a structured numerical record. The sensor array maintains a real-time connection with the data association and storage module via an IoT transmission protocol, synchronously caching the most recent 24 hours of collected data locally to cope with network interruptions.

[0040] The second category is aquaculture operation records. These records are entered based on the time of the event, manually by aquaculture personnel via mobile devices or automatically triggered by relevant equipment. Feeding operation records include the time, amount, and batch number of each feeding, with the amount precisely recorded in grams or kilograms. These records are compared in real-time with the pond's historical average feeding values, and are automatically flagged when deviations exceed a set threshold. Water change or addition operation records include the time and volume of water used. Aeration equipment operation status is automatically reported by the equipment controller, with start and stop times reported automatically; fault status is automatically triggered by the fault detection module. Medication operation records include the time, type of medication, and dosage. Each operation record includes an event timestamp and pond number, and is stored as an event-based structured record.

[0041] The third category is batch information for seedlings and feed. Seedling batch information is entered once upon arrival at the pond, including the supplier's name, origin, arrival time, batch number, and stress assessment data such as transportation time, temperature difference during transportation, and loading density. Feed batch information is entered upon feed storage, including feed brand, batch number, storage time, opening time, and temperature and humidity conditions of the storage environment. This batch information is statically stored using the batch number as the primary key. The batch number also serves as a foreign key, linking to feeding records and pond numbers in the aquaculture operation log, ensuring that each feeding operation can be traced back to the corresponding feed batch information, and each pond can be traced back to the corresponding seedling batch information.

[0042] Example 3 like Figure 3 As shown, after receiving the three types of raw data, the data association storage module performs unified timestamp alignment processing, mapping the data with different collection frequencies and recording granularities to a time axis in minutes, forming a multi-source association database that can be queried across different types.

[0043] The specific alignment process is as follows: For water quality environmental parameters, each collection record is accompanied by a timestamp with minute-level precision, directly mapped to the corresponding time unit without conversion. Each minute corresponds to one complete record of six parameters. For aquaculture operation records, each event-type record is mapped to the corresponding time unit with the minute portion of its event occurrence timestamp. Multiple operation events occurring within the same time unit are stored side-by-side under that time unit, distinguished by the operation type field. If multiple similar operations occur within the same minute, they are distinguished by sub-sequence numbers to ensure no data loss. For seedling and feed batch information, the batch number is used as an index to map the effective start time (seedling entry time or feed opening time) to the end time (seedling cleaning time or feed exhaustion time) of the batch to a continuous time unit interval. Each time unit within the interval carries a status marker corresponding to the batch number, ensuring continuous queryability of batch information on the timeline.

[0044] After the alignment process described above, the three types of data are integrated into a multi-source relational database using the pond number and time unit as the joint primary key. Any record in the database corresponds to a complete snapshot of the status of a specific pond within a specific minute time unit. This snapshot includes the values ​​of six parameters for the current time unit: dissolved oxygen, water temperature, ammonia nitrogen, nitrite, pH, and turbidity; all operational events that occurred within the current time unit; and valid seedling batch numbers and feed batch numbers within the current time unit.

[0045] The data association and storage module also maintains a pond relationship table, recording the water system connectivity and geographical adjacency relationships between all ponds within the system. Water system connectivity relationships include shared inlet channels and drainage flow direction relationships, recorded using fields such as connectivity type, upstream / downstream direction, and connecting medium. Geographical adjacency relationships are recorded using fields such as geographical distance and boundary contact between ponds. The pond relationship table is configured and entered by administrators when ponds are connected to the system, and is used by the multi-dimensional traceability module for spatial tracing.

[0046] Example 4 like Figure 4 As shown, after a disease event is triggered, the multi-dimensional tracing module initiates parallel tracing searches in three dimensions—time, space, and batch—to the multi-source associated database, using the disease confirmation time point and the number of the affected pond as the search entry point. The three dimensions are executed synchronously and independently. After the search is completed, the evidence is summarized and the confidence level is calculated.

[0047] There are two ways to trigger disease events. The first is through proactive reporting by aquaculture personnel. When aquaculture personnel observe visible abnormalities in the fish population, they can report the disease event via mobile device, recording the discovery time, the pond number where the disease occurred, and a description of the visible symptoms. The system will then use the reporting time as the disease confirmation time point to initiate source tracing. The second is through automatic system triggering. When the data acquisition module detects that the water quality parameters of a certain pond continuously deviate from the normal range at multiple consecutive collection points, the system will automatically generate a disease warning event. The system will then use the warning trigger time as the disease confirmation time point to initiate source tracing and simultaneously push a warning notification to the aquaculture personnel.

[0048] After the three-dimensional source tracing search is launched in parallel, each dimension executes independently according to the following logic: The time dimension uses the disease confirmation time as the endpoint and retrieves historical records within a preset time window (default 72 hours, adjustable according to the aquaculture species and disease type). It examines the environmental parameter changes and operational behavior sequence of the pond item by item to identify the occurrence of primary and secondary factors. The spatial dimension uses the affected pond number as the search condition. It retrieves all associated pond numbers with water system connectivity or geographical proximity from the pond relationship table, and then searches the multi-source association database for disease event records and shared facility usage records of each associated pond within the retrospective time window to determine the external source entry path and diffusion direction. The batch dimension uses the currently valid seedling batch number and feed batch number of the affected pond as the search condition. It searches the multi-source association database for all other ponds using the same batch within the system to obtain their recent disease event records and health status data, identifying batch-specific clustering disease characteristics.

[0049] The search results from the three dimensions generate separate lists of evidence entries. Each evidence entry includes the evidence type, the dimension it belongs to, the supported etiological pathway, the evidence strength rating, and the corresponding original data record reference. After all evidence entries are aggregated, they are grouped by etiological pathway, and the overall confidence level for each pathway is calculated. The confidence level comprehensively considers the number of evidence entries supporting the pathway, the strength rating of each piece of evidence, and the cross-verification of evidence from different dimensions. Finally, the results are sorted from highest to lowest confidence level, and a structured source tracing report is output.

[0050] Example 5 like Figure 5 As shown, the factors in the three tracing dimensions are divided into two levels: primary factors and secondary factors. These two levels of factors play different logical roles in the tracing reasoning. The following uses a typical aquaculture scenario to illustrate the specific operation of each dimension's factors.

[0051] Taking a grass carp farm in Guangdong as an example, during the spring warming period, some ponds experienced abnormal situations such as fish surfacing and reduced feeding. The farmers reported the disease event through a mobile device, and the system initiated three-dimensional traceability.

[0052] In the time-based tracing, the system retrieved historical data from the pond over the past 72 hours and found the following: Dissolved oxygen began a monotonous downward trend 48 hours before the onset of the disease, and the daily fluctuation range narrowed significantly in the 24 hours before the onset, with the lowest value dropping to 3.2 mg / L. The downward trajectory of dissolved oxygen was determined to be the main factor, independently constituting the cause of hypoxia-induced disease. Ammonia nitrogen began to accumulate continuously 36 hours before the onset of the disease, reaching 0.8 mg / L by the time the disease occurred, exceeding the upper limit of suitable water quality for grass carp. The accumulation trend of ammonia nitrogen was determined to be the main factor, independently constituting the cause of water quality deterioration. The water temperature showed a continuous widening of the daily temperature difference in the 72 hours before the onset of the disease, with the maximum daily temperature difference reaching 9℃, exceeding the stress threshold for grass carp. The sudden change in water temperature was determined to be the main factor, independently constituting the cause of stress-induced disease. After identifying the primary factors, the system further examined auxiliary factors: turbidity increased significantly 48 hours before the onset of the disease, coinciding closely with the time of dissolved oxygen decline. This indicated that the sudden increase in turbidity pointed to the large-scale decomposition of organic matter, further focusing the cause of water quality deterioration on the increased oxygen consumption due to the decomposition of bottom organic matter. Feeding records from 36 hours before the onset of the disease showed that feeding occurred during the early morning when dissolved oxygen was at its lowest. This abnormal feeding time was identified as an auxiliary factor, further revising the attribution of hypoxia to feeding during the low dissolved oxygen period, which exacerbated the hypoxia. Aeration equipment operation records showed that the aerator operated normally with no fault records in the 24 hours before the onset of the disease, ruling out equipment failure and confirming that the decline in dissolved oxygen was due to natural environmental deterioration rather than equipment issues.

[0053] In the spatial tracing, the system searched upstream ponds connected to the current pond by waterways and found that the upstream pond sharing the same inlet channel had disease records 5 days before the onset of the disease. Furthermore, this upstream pond had performed a water change operation 3 days before the onset of the disease. The timing of the water change coincided with the incubation period of the pathogen, leading to the conclusion that the shared inlet channel and the disease records of the upstream pond were the primary factors, constituting the exogenous transmission path. Simultaneously, the system searched the onset time of geographically adjacent ponds and found no clear chronological order, ruling out geographically adjacent transmission paths. No records of shared tool usage across ponds were found, ruling out mechanical transmission paths. Regarding auxiliary factors, the system examined the drainage direction of the pond and found that it flowed to two downstream ponds. After establishing the exogenous transmission path, the system automatically issued preventative warnings to the two downstream ponds.

[0054] In batch-level tracing, the system searched for other ponds using the same seedling batch number within the system, finding three ponds. One of these was located in the same aquaculture farm (geographically adjacent to the affected pond and not included in batch attribution), while the other two were located in different farms and had no spatial connection to the affected pond. These two ponds had no recent disease records, ruling out seedling batch issues. Searching by feed batch number, five ponds using the same feed batch were found. Four had no disease records, and one had a disease record but shared an inlet channel with the affected pond, making spatial transmission a possibility that could not be ruled out. The feed batch attribution had low confidence and was retained as a low-confidence path. Regarding auxiliary factors, the system checked the pond's turbidity response records after feeding and found no abnormal increase in turbidity after the most recent feeding, further reducing the likelihood of a feed batch issue being the cause.

[0055] After summarizing the three dimensions, the source tracing report for this disease event lists three causal pathways in order of confidence: The first is exogenous input combined with environmental deterioration, supported by evidence including disease records upstream of the intake canal, the trajectory of dissolved oxygen decline, the trend of ammonia nitrogen accumulation, and sudden changes in water temperature, with the highest confidence level; the second is natural environmental deterioration, supported by evidence including the trajectory of dissolved oxygen decline, the trend of ammonia nitrogen accumulation, and sudden changes in water temperature, with the next highest confidence level; the third is feed batch issues, with weak supporting evidence and a low confidence level. The report also lists the focused conclusions of each auxiliary factor regarding the attribution direction of the main factor, for the fisheries veterinarian's review.

[0056] Example 6 like Figure 6 As shown, after receiving the structured traceability report output by the multidimensional traceability module, the prevention and control closed-loop module executes the five stages in sequence: traceability report reception, prescription issuance, implementation, effect monitoring, and archiving, forming a complete closed-loop operation chain.

[0057] During the source tracing report receiving phase, the prevention and control closed-loop module pushes the structured source tracing report to the Fish Doctor terminal via a standardized interface. The report interface displays each etiological pathway in descending order of confidence level. Under each pathway, all evidence entries supporting that pathway are displayed. Each evidence entry is labeled with its respective dimension, evidence type, and corresponding original data record. Fish Doctors can view the evidence content item by item. The report also displays the focusing conclusions of auxiliary factors on the main factors, as well as the early warning status for associated risk ponds.

[0058] During the prescription writing stage, after reviewing the complete chain of evidence, the doctor confirms or modifies the etiology based on their professional judgment and issues an electronic prescription within the system. The electronic prescription includes the following fields: confirmed etiology, recommended animal health product name and specifications, dosage, medication time window, administration method, withdrawal period requirements, and preventative recommendations for associated risk ponds. The system does not automatically generate the prescription content; all fields are filled in and confirmed by the doctor, and the doctor digitally signs the prescription upon completion. The signature information is stored along with the prescription content.

[0059] During the execution phase, after the electronic prescription is signed by the fish veterinarian, the system sends the types, dosages, and administration time windows of the animal health products in the prescription to the pond execution terminal via an IoT interface. If the farm is equipped with automatic dispensing equipment, the system automatically triggers quantitative dispensing at the start of the administration time window, and the equipment automatically sends back an execution confirmation record after dispensing is completed. If manual operation is required on-site, the system pushes execution instructions containing specific operation parameters to the fish farmers. After the fish farmers complete the operation according to the instructions, they submit an operation confirmation receipt on their mobile devices. The receipt includes the actual operation time and the actual dosage used.

[0060] During the effectiveness monitoring phase, after the animal protection implementation confirmation record is generated, the system enters the effectiveness monitoring state, continuously collecting water quality environmental parameters and biological behavior data of the pond. The collected data is compared item by item with the historical baseline of the pond in the 30 days prior to the onset of disease, and the recovery trend score for each parameter is calculated. If each parameter reaches the historical baseline level within the preset recovery period, the system determines that the treatment effect has met expectations, automatically closes the effectiveness monitoring state, and enters the archiving stage. If the recovery trend score still does not meet the standard at the end of the preset recovery period, the system triggers a secondary warning, pushing the effectiveness monitoring data and the original prescription to the fish doctor's terminal, notifying the fish doctor to reassess the prescription plan and re-enter the prescription issuance stage until the effect meets expectations.

[0061] During the archiving phase, after the effectiveness monitoring is completed, the system writes all process data of this disease event into a multi-source associated database, forming a complete disease event record. The archived content includes: disease confirmation time and affected pond number, full text of the 3D source tracing report, the causative conclusion confirmed by the fish veterinarian, the original electronic prescription and the fish veterinarian's digital signature, execution confirmation record, effectiveness monitoring data, and final treatment conclusion. After archiving, this disease event record becomes historical data accumulation for this aquaculture scenario. When subsequent disease events occur in the same farm or related ponds from the same batch, it serves as a historical disease record in terms of spatial or batch dimensions for source tracing retrieval, continuously enhancing the system's source tracing capabilities with data accumulation.

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical methods of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical methods of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical methods to deviate from the spirit and scope of the technical methods of the present invention.

Claims

1. An IoT monitoring system for multi-dimensional tracing of aquatic diseases, characterized in that, include: The data acquisition module is deployed at the aquaculture pond. It continuously collects water quality environmental parameters, aquaculture operation records and seedling and feed batch information through an Internet of Things (IoT) sensor array to obtain multi-source heterogeneous data. The multi-source heterogeneous data is then uploaded to the data association storage module in real time through the IoT transmission protocol. The data association storage module aligns the above-mentioned multi-source heterogeneous data through a unified timestamp alignment protocol, and forms a multi-source association database with the pond number and time unit as the primary key. The data association storage module also stores a pond relationship table for spatial tracing. The multidimensional tracing module includes time-dimensional tracing, spatial-dimensional tracing, and batch-dimensional tracing. When a disease event is triggered, the tracing is carried out along the three dimensions of time, space, and batch, and finally a structured evidence report sorted by confidence level is output. The prevention and control closed-loop module pushes the source tracing report to the fish doctor's terminal. After the fish doctor reviews it, an electronic prescription is issued. The animal health department is linked through the Internet of Things interface to carry out the work, and the treatment process and effect data are fully archived to form a closed operation link of monitoring, source tracing, prevention and control and feedback.

2. The IoT monitoring system for multi-dimensional tracing of aquatic diseases according to claim 1, characterized in that, The water quality environmental parameters are continuously collected at a frequency of minutes by a sensor array deployed at the pond, and uploaded in real time to the data association and storage module via an Internet of Things (IoT) transmission protocol, specifically including: Instantaneous dissolved oxygen values, rate of decrease, and intraday fluctuations; instantaneous water temperature values ​​and changes per unit time; instantaneous ammonia nitrogen and nitrite values ​​and cumulative trends at continuous sampling points; instantaneous pH values ​​and direction and rate of shift; instantaneous turbidity values ​​and abrupt changes. Each water quality environmental parameter is collected along with a collection timestamp and pond number, and stored as a structured numerical record. The aquaculture operation records are entered manually by aquaculture personnel or automatically triggered by equipment, with the event occurrence time as the node. Specifically, they include: The time, amount, and batch number of each feeding; the time and amount of each water change or addition; the start-up and shutdown time and operating status of the aeration equipment; the time, type, and dosage of each medication; all of the above records are stored in an event-based structured record format, with each record including an event timestamp and pond number. The batch information for seedlings and feed is entered once when the seedlings are delivered to the pond or the feed is stored, specifically including: The supplier, origin, time of entry into the pond, batch number, and stress conditions such as transportation time, temperature difference and density of the seedlings; the brand, batch number, time of entry into the warehouse, time of opening, and storage temperature and humidity conditions of the feed. The above information is stored statically in batches and is linked to the feeding records and pond numbers in the aquaculture operation records through the batch number.

3. The IoT monitoring system for multi-dimensional tracing of aquatic diseases according to claim 1, characterized in that, The data association and storage module uses minutes as the smallest time unit to perform unified timestamp alignment on three types of data: water quality environmental parameters, aquaculture operation records, and seedling and feed batch information. The specific alignment method is as follows: For water quality environmental parameters, each collection record is already accompanied by a minute-level timestamp, which is directly mapped to the corresponding time unit without conversion; For aquaculture operation records, each event-type record is mapped to the corresponding time unit interval by its event timestamp. Multiple operation records that occur in the same time unit are stored in parallel under that time unit. For seedling and feed batch information, the batch number is used as an index to map the effective start time to end time of the batch into a continuous time unit interval. Each time unit in the interval carries the status mark of the corresponding batch, so that the batch information can be directly associated with and queried with the environmental parameters and operation records in the same time unit. After the above alignment process, the three types of data are integrated into a multi-source associated database using the pond number and time unit as the joint primary key. Any record in the database corresponds to a complete snapshot of the status of a specific pond within a specific time unit, including the instantaneous values ​​and statistical characteristics of all environmental parameters, all operation event entries, and the currently valid seedling and feed batch markers within that time unit. Environmental changes, operational behaviors, and batch status at any point in time can be completely associated and queried in the same record using the pond number and time unit, providing a unified data retrieval foundation for the multi-dimensional traceability module.

4. The IoT monitoring system for multi-dimensional tracing of aquatic diseases according to claim 1, characterized in that, When a disease event is triggered, the multi-dimensional tracing module initiates a parallel tracing search in three dimensions to the multi-source association database, using the disease confirmation time point and the number of the affected pond as indexes. The three dimensions are time dimension, space dimension and batch dimension, and the traceability factors in each of the three dimensions are divided into primary factors and auxiliary factors; After summarizing all evidence items across the three dimensions, the confidence level is calculated based on the number and strength of supporting evidence for each etiological pathway, and the results are sorted and a structured source tracing report is output.

5. The IoT monitoring system for multi-dimensional tracing of aquatic diseases according to claim 1, characterized in that, The time dimension starts from the time point of disease confirmation and traces back to the historical data of diseased ponds within a preset time window to examine the temporal relationship between changes in environmental parameters and aquaculture operations. The spatial dimension starts with the affected ponds and searches for related ponds that have water system connectivity or geographical proximity to them, comparing their disease outbreak timeline with the usage records of shared facilities. The batch dimension uses the currently valid seedling and feed batch numbers of the diseased ponds as an index to search all ponds using the same batch within the system and compare their corresponding health status.

6. The IoT monitoring system for multi-dimensional tracing of aquatic diseases according to claim 1, characterized in that, The time-dimensional tracing factors include primary time-tracing factors and secondary time-tracing factors, as detailed below: Key factors in time-based tracing include: dissolved oxygen decline trajectory, ammonia nitrogen and nitrite accumulation trends, sudden changes in water temperature, abnormal deviations in feeding amounts, and the temporal relationship of medication records. The auxiliary factors for time-based tracing include sudden changes in turbidity, the direction and rate of pH shift, abnormal feeding times, and the operating status of aeration equipment.

7. The IoT monitoring system for multi-dimensional tracing of aquatic diseases according to claim 1, characterized in that, The factors for tracing the spatial dimension include primary factors and auxiliary factors for spatial tracing, as detailed below: The main factors for spatial source tracing include the shared relationship of the intake channel and the disease records of upstream ponds, the chronological order of disease occurrence in adjacent ponds, and the record of the order of use of shared tools; Spatial source tracing auxiliary factors include the relationship between drainage direction and downstream ponds, and records of the introduction of foreign fish fry or broodstock.

8. The IoT monitoring system for multi-dimensional tracing of aquatic diseases according to claim 1, characterized in that, The batch-level traceability factors include primary batch traceability factors and secondary batch traceability factors, as detailed below: The main factors for batch traceability include the clustered disease records of the same batch of seedlings in multiple unrelated farms, and the clustering of diseases within the scope of use of the same batch of feed; Batch traceability auxiliary factors include abnormal turbidity response after feeding, feed storage condition records, and stress assessment records at the time of seedling introduction.

9. The IoT monitoring system for multi-dimensional tracing of aquatic diseases according to claim 1, characterized in that, The prevention and control closed-loop module includes: Fisherman Doctor's End Review Unit: Fisherman Doctor reviews the complete chain of evidence to make a professional diagnosis of the cause of the disease; Electronic prescription issuance unit: A comprehensive electronic prescription is generated based on the cause of the disease, pests, products, dosage, time window, and prevention recommendations, and the fisherman's signature is recorded. IoT-linked execution unit: The IoT-linked execution unit is used for automatically sending or manually executing operation confirmation records; Effect monitoring and data archiving unit: Records data evaluation during the recovery period. If the expected results are not met, a secondary warning is triggered. If the expected results are met, the entire process data is archived and preserved.

10. A closed-loop prevention and control method for an IoT monitoring system for multi-dimensional source tracing of aquatic diseases, characterized in that, include: During the source tracing report receiving phase, the prevention and control closed-loop module receives the structured evidence report output by the multi-dimensional source tracing module. The report lists multiple etiological pathways sorted by confidence level. Each pathway is accompanied by specific evidence items supporting the pathway and their source dimensions, fully preserving the accessibility of the evidence chain. During the prescription issuance stage, the structured evidence report is pushed to the fish doctor's terminal through a standardized interface. After reviewing the complete evidence chain, the fish doctor with professional qualifications makes a diagnosis based on professional judgment and issues an electronic prescription that includes the confirmed diagnosis, the type and dosage of animal health products, the medication time window, and preventive recommendations for related risk ponds. The system does not automatically generate prescriptions, and the right to issue prescriptions belongs to the fish doctor. During the execution phase, the electronic prescription is sent to the pond execution terminal via the Internet of Things interface; if there is automatic delivery equipment on site, quantitative delivery is triggered directly; if manual operation is required, an execution instruction containing specific operation parameters is pushed to the aquaculture personnel, and a confirmation receipt is required after the operation is completed. During the effect monitoring phase, after the animal protection program is completed, the system continuously collects water quality environmental parameters and biological behavior data during the recovery period, compares them with the historical baseline before the onset of the disease, and assesses the recovery trend. If the recovery trend does not meet expectations, a second warning will be triggered, and the doctor will be notified to reassess the prescription. During the archiving phase, the source tracing report, electronic prescription, execution record, and recovery period data of this disease event are archived as a complete disease event record into a multi-source associated database, forming a traceable digital certificate, and serving as historical data accumulation for subsequent source tracing and comparison in the same breeding scenario.

Images