Livestock full-link digital collaborative supervision method and system based on cloud computing

Abstract

This application relates to the field of big data processing technology and provides a cloud computing-based digital collaborative supervision method and system for the entire livestock industry chain. The method includes: acquiring a business dataset; generating an initial report based on the business dataset under abnormal conditions; acquiring trigger fields from the initial report; identifying the business domain corresponding to the trigger fields; acquiring a set of departments based on the business domain when the trigger fields meet cross-departmental characteristics; generating a department response sequence based on the priority weight of departments within the business domain; sending the department response sequence to subsystems of multiple departments and acquiring feedback information from the subsystems; comparing the feedback timestamp with a preset time sequence; issuing a suspension command to the subsystem when the feedback timestamp is earlier than the feedback timestamp of the preceding department; acquiring a priority adjustment value and updating the department response sequence based on the priority adjustment value; and issuing commands to the corresponding subsystems based on the updated department response sequence, which can prevent time sequence conflicts in the execution of commands from different departments.

Description

Technical Field

[0001] This application belongs to the field of big data processing technology, specifically a cloud computing-based digital collaborative supervision method and system for the entire livestock industry chain. Background Technology

[0002] Existing livestock supervision often requires cross-departmental collaboration. However, due to the fragmented business processes between different departments and the reliance on manual communication for information transmission, there is a lack of means to control the execution sequence of each department when facing emergencies. This often leads to time misalignments and conflicts when different departments execute instructions, thereby affecting the accuracy and response speed of supervision. Summary of the Invention

[0003] To address the above problems, this invention provides a cloud computing-based digital collaborative supervision method and system for the entire livestock industry chain.

[0004] To achieve the above objectives, in a first aspect, the present invention employs a cloud computing-based digital collaborative supervision method for the entire livestock supply chain. The method includes: acquiring a business dataset; generating an initial report based on the business dataset when the business dataset meets a preset anomaly pattern; acquiring a trigger field from the initial report; identifying the business domain corresponding to the trigger field; acquiring a department set based on the business domain when the trigger field meets preset cross-departmental characteristics, the department set including multiple departments; sorting the departments based on preset priority weights within the business domain to generate a department response sequence; sending the department response sequence to the subsystems of multiple departments and acquiring feedback information output by the subsystems, wherein the feedback information includes a feedback timestamp; comparing the feedback timestamp with a preset time sequence in the department response sequence; issuing a logic lock instruction to the corresponding subsystem to suspend the instruction execution process when the feedback timestamp is earlier than the feedback timestamp of the preceding department; and acquiring a preset priority adjustment value and updating the department response sequence based on the priority adjustment value, and issuing instructions to the corresponding subsystem based on the updated department response sequence.

[0005] In some implementations, when the business dataset meets a preset anomaly pattern, an initial report is generated based on the business dataset, including: the business dataset includes epidemic prevention records and circulation records; comparing the data in the epidemic prevention records and circulation records with preset anomaly thresholds and time thresholds, wherein the time feature value is the duration for which the data remains within the preset anomaly threshold range; when the anomaly thresholds and time thresholds meet a preset risk range, obtaining an anomaly pattern, and matching the anomaly pattern with preset risk features to obtain a risk identification record; and obtaining a risk level based on the risk identification record, and obtaining the initial report based on the matching result of the risk level data and historical data records.

[0006] In some implementations, generating the department response sequence includes: matching the trigger field with a preset domain mapping table to obtain the business domain corresponding to the trigger field, and obtaining a preset department association matrix based on the business domain; obtaining a collaboration feature score based on the department association matrix, the trigger field, and business values ​​in the business dataset; obtaining the department set if the collaboration feature score meets a preset cross-department feature threshold; and obtaining the priority weight score of the department based on the department association matrix and the department set, and sorting them in descending order based on the priority weight score to generate the department response sequence.

[0007] In some implementations, generating the department association matrix includes: obtaining a business association score for each department based on the response latency and processing success rate of the departments for different business areas within a preset historical period; normalizing the department set based on the business association score to obtain weight coefficients; and generating the department association matrix based on the mapping relationship between the business area identifier, the department identifier, and the weight coefficients; wherein the business area identifier and the department identifier are the codes for the business area and the department, respectively.

[0008] In some implementations, obtaining the priority adjustment value includes: obtaining the misalignment duration of the feedback timestamp relative to the preset time sequence; obtaining the business importance weight corresponding to the subsystem, and establishing a product mapping relationship between the misalignment duration and the business importance weight; and determining the corresponding weight correction score based on the product mapping relationship, and determining the weight correction score as the priority adjustment value.

[0009] In some implementations, updating the department response sequence based on the priority adjustment value includes: reordering each department node in the department response sequence based on the priority adjustment value to obtain an initial execution sequence; performing conflict detection on the initial execution sequence to remove redundant instructions with temporal overlap and obtain a conflict-free process; and extracting key node data from the conflict-free process and optimizing the sorting of the key node data based on a decision tree algorithm to generate the updated department response sequence.

[0010] In some implementations, the method further includes: obtaining an execution confirmation signal returned by the subsystem based on the instruction; obtaining instruction coverage data based on the confirmation signal and the department set; identifying abnormal data of the subsystem based on the instruction coverage data; classifying the abnormal data based on a support vector machine algorithm; and obtaining the risk diffusion status of the subsystem. Based on the risk diffusion status, obtain the instruction interaction permissions of the subsystem, and update the regulatory record based on the permission restriction results.

[0011] In some implementations, after updating the regulatory records, the method further includes: extracting historical response duration data based on the regulatory records; segmenting the historical response duration data to obtain duration distribution results; comparing the duration distribution results with a preset threshold to determine key points; and obtaining abnormal fluctuation points in the historical response duration data corresponding to the key points; obtaining upstream and downstream interaction data of the key points; analyzing the abnormal patterns of the upstream and downstream interaction data and the abnormal fluctuation points based on a random forest algorithm to generate an optimization suggestion sequence; and adjusting the resource allocation strategy of each subsystem based on the optimization suggestion sequence to obtain an optimized configuration scheme.

[0012] In some implementations, the method further includes: optimizing the performance indicators of the subsystem based on the configuration scheme using a genetic algorithm, wherein the performance indicators include at least one of data processing speed, response time, and throughput; determining triggering rules based on the optimized performance indicators; and outputting an alarm signal based on the triggering rules in the event of abnormal performance indicators.

[0013] Secondly, this application also provides a cloud-based digital collaborative supervision cloud system for the entire livestock supply chain. When executed, the cloud system implements the steps of any of the cloud-based digital collaborative supervision methods for the entire livestock supply chain described above.

[0014] This application discloses a cloud-based digital collaborative supervision method and system for the entire livestock supply chain. It utilizes trigger fields, business domains, and priority weights to construct departmental response sequences, enabling automatic task allocation across departments. When feedback timestamps do not conform to the preset sequence, a suspension command is issued to the subsystem process, preventing timing conflicts between different departments. By obtaining priority adjustment values ​​and updating the departmental response sequence, timing conflicts are corrected, avoiding time misalignment and timing conflicts, and ensuring the accuracy of the supervision method when the feedback timestamps of the subsystems are misaligned.

[0015] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0016] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 A schematic diagram of a cloud-based digital collaborative supervision cloud system for the entire livestock industry chain, provided as an embodiment of this application; Figure 2 A flowchart illustrating a cloud-based digital collaborative supervision method for the entire livestock supply chain, provided as an embodiment of this application; Figure 3 A flowchart illustrating a cloud-based digital collaborative supervision method for the entire livestock supply chain, provided as an embodiment of this application; Figure 4 A flowchart illustrating a cloud-based digital collaborative supervision method for the entire livestock supply chain, provided as an embodiment of this application; Figure 5 A flowchart illustrating a cloud-based digital collaborative supervision method for the entire livestock supply chain, provided as an embodiment of this application; Figure 6 A flowchart illustrating a cloud-based digital collaborative supervision method for the entire livestock supply chain, provided as an embodiment of this application; Figure 7 A flowchart illustrating a cloud-based digital collaborative supervision method for the entire livestock supply chain, provided as an embodiment of this application; Figure 8 A flowchart illustrating a cloud-based digital collaborative supervision method for the entire livestock supply chain, provided as an embodiment of this application; Figure 9 A flowchart illustrating a cloud-based digital collaborative supervision method for the entire livestock supply chain, provided as an embodiment of this application; Figure 10This is a flowchart illustrating a cloud-based digital collaborative supervision method for the entire livestock industry chain, provided as an embodiment of this application.

[0017] Figure label: Cloud system 1000; Subsystem 101. Detailed Implementation

[0018] To enable those skilled in the art to better understand the technical solution, the present invention will be described in detail below with reference to embodiments. The description in this part is only exemplary and explanatory, and should not be used to limit the scope of protection of the present invention in any way.

[0019] In modern agricultural development, animal husbandry serves as a crucial pillar for ensuring food safety and public health, making its end-to-end regulation particularly critical. Especially in areas such as animal disease prevention, quarantine, and market circulation, efficient collaborative management directly impacts the effectiveness of disease control and the stability of market order. However, traditional regulatory methods often face complex challenges related to cross-departmental collaboration, necessitating the use of emerging technologies such as cloud computing to achieve digital transformation and improve regulatory efficiency and response speed.

[0020] Currently, while there have been some digitalization attempts in livestock supervision, most remain at the level of internal departmental information management, lacking cross-departmental collaborative capabilities. Business processes between different departments are fragmented, and information transmission relies on manual communication, resulting in slow response times and significant coordination difficulties when facing emergencies. This problem not only increases regulatory costs but may also amplify risks due to delays, especially in incidents involving public safety, where the impact is particularly severe.

[0021] From a technical perspective, the challenge of full-chain livestock supervision lies in integrating performance indicators scattered across multiple departments into a unified automated process. Of particular concern is the identification of triggering conditions in different business domains, which becomes a primary obstacle. This is because event triggering conditions vary greatly across different scenarios, making it difficult to pre-set fixed rules. This problem further leads to confusion in the order of business execution, and inter-departmental collaboration often results in time misalignments or conflicting instructions. For example, when a slaughterhouse discovers a suspected disease and uploads a report, the agricultural department may need to issue an isolation order immediately, but the market supervision department may fail to suspend the sale of related products due to untimely information synchronization, ultimately leading to the spread of potential risks. Therefore, to solve the above problems, this application provides a cloud-based digital collaborative supervision method for the entire livestock supply chain (…). Figure 2-10 (as shown) and system ( Figure 1 (As shown).

[0022] Please see Figure 1 and Figure 2 This application adopts a cloud computing-based digital collaborative supervision method for the entire livestock industry chain, the method including: 01: Obtain the business dataset, and generate an initial report based on the business dataset if the business dataset meets the preset exception pattern; 02: Obtain the trigger field from the initial report, identify the business domain corresponding to the trigger field, and if the trigger field meets the preset cross-departmental characteristics, obtain the department set based on the business domain. The department set includes multiple departments, and sort them based on the preset priority weight of the departments under the business domain to generate the department response sequence. 03: Send the department response sequence to subsystem 101 of multiple departments and obtain the feedback information output by subsystem 101, wherein the feedback information includes the feedback timestamp; 04: Compare the feedback timestamp with the preset timing in the department response sequence. If the feedback timestamp is earlier than the feedback timestamp of the preceding department, issue a suspension command to the corresponding subsystem 101; and 05: Obtain the preset priority adjustment value, update the department response sequence based on the priority adjustment value, and issue instructions to the corresponding subsystem 101 based on the updated department response sequence.

[0023] Specifically, in the method of 01, the business dataset represents the collection formed after big data collection and processing of industrial big data. This application uses livestock big data as an example. For instance, the business dataset is the original regulatory data collected from the entire chain of livestock breeding, circulation, and slaughter. For instance, this application can utilize a cloud-based distributed data collection cloud system (hereinafter referred to as Cloud System 1000) to collect data from 500 livestock regulatory stations nationwide every hour. Preset anomaly patterns are rules used to determine whether business data belongs to a risk state. The preset anomaly patterns are objective indicators obtained by organizing historical data. Historical big data represents various business indicators and their corresponding risk event records stored during the entire chain of supervision within a preset historical period. Cloud System 1000 extracts the statistical inflection point of sudden change in business security by fitting and analyzing the statistical distribution pattern of risk occurrence in historical big data, and determines it as a quantitative critical value. By analyzing the negative correlation between vaccination rates and disease outbreak frequency over the past five years, Cloud System 1000 discovered that when the vaccination rate falls below 70%, the probability distribution of disease transmission risk exhibits a step increase. Based on this, Cloud System 1000 automatically configures the quantitative threshold for business domains to 70%. If the actual measured business data shows a centralized vaccination rate of 60%, Cloud System 1000 determines that it meets a preset anomaly pattern, and then obtains the time, location, and involved objects of the data, and encapsulates it to generate an initial report.

[0024] In the method described in 02, the trigger field represents regulatory-oriented keywords extracted from the initial report. The business domain represents the functional scope corresponding to the trigger field, such as animal disease prevention and market access. Cross-departmental features are used to determine whether a risk involves collaboration among multiple functional departments. The department set represents all relevant collaborating departments identified after identification. Priority weights represent the degree of functional coreness of each department in a specific business domain. For example, after the cloud system 1000 identifies the business domain corresponding to the trigger field (such as "disease"), if it determines that it meets the cross-departmental feature, it obtains the set of departments involved and sorts them in descending order according to the priority weights of each department, thereby generating a departmental response sequence and realizing the automated division of regulatory tasks.

[0025] In the method described in 03, subsystem 101 represents the business terminals corresponding to each department in the department set. Each subsystem 101 is connected to and communicates with the distributed data acquisition cloud system 1000 of the cloud computing platform. The distributed data acquisition cloud system 1000 can send instructions to subsystem 101, and subsystem 101 can also send feedback information back to the distributed data acquisition cloud system 1000. The feedback information represents the execution status data sent back to the cloud system 1000 by each subsystem 101 after receiving instructions or executing tasks. The feedback timestamp represents the time node when the subsystem 101 outputs the feedback information. For example, the cloud system 1000 synchronizes the generated department response sequence to the mobile office APP or PC management backend of each department (i.e., subsystem 101) through the API interface, and listens for the feedback information returned by each subsystem 101, which includes processing status, latitude and longitude, and millisecond-level feedback timestamps, to realize the real-time issuance and return of instructions, providing a time reference for subsequent timing verification.

[0026] In the method of 04, the preset timing sequence represents the ideal processing time order that each department should follow in the department response sequence, and the preceding department represents the department that is arranged before the current processing department in the department response sequence. The suspension command is used to forcibly pause the program that is processing instructions inside subsystem 101. Specifically, cloud system 1000 retrieves the preset timing sequence and compares it with the feedback timestamps of each subsystem 101. If it finds that the feedback timestamp of the second department in the sequence is earlier than the feedback timestamp of its preceding department, cloud system 1000 determines that a logical preemption has occurred and sends a suspension command to the department's subsystem 101. This prevents instruction conflicts and data inconsistencies caused by the different execution paces of subsystem 101 in the environment of distributed cloud system 1000.

[0027] In the method described in 05, the priority adjustment value represents a pre-set numerical item used to correct the original weights in response to process execution deviations. The updated department response sequence represents a new sorting sequence that meets the current actual execution requirements after logical correction. The instruction represents the specific business operation instruction combination finally issued to each subsystem 101. Specifically, based on the identified timing conflicts, the cloud system 1000 obtains a priority adjustment value such as "-1" or "+1" to compensate for the original weights, thereby generating an updated department response sequence. Based on this new sequence, it reissues instructions to the relevant subsystems 101 to avoid timing conflicts.

[0028] In the regulatory method of this application, a departmental response sequence is constructed using trigger fields, business domains, and priority weights to achieve automatic allocation of cross-departmental tasks. When the feedback timestamp does not conform to the preset sequence, a suspension command is issued to the subsystem 101 process to prevent timing conflicts in the execution of commands between different departments. By obtaining the priority adjustment value and updating the departmental response sequence to issue commands, the timing conflict is corrected, ensuring the accuracy of the regulatory method when the feedback timestamp of subsystem 101 is misaligned.

[0029] Please see Figure 1 and Figure 3 In some implementations, if the business dataset meets a preset anomaly pattern, an initial report is generated based on the business dataset, including: 011: The business dataset includes epidemic prevention records and circulation records. The data in the epidemic prevention records and circulation records are compared with preset anomaly thresholds and time thresholds. The time threshold is the duration for which the data remains within the anomaly threshold range. 012: When the anomaly threshold and time threshold meet the preset risk range, obtain the anomaly pattern and obtain risk identification records based on the anomaly pattern; and 013: Obtain the risk level based on the risk identification record, and obtain the initial report based on the risk level data.

[0030] Specifically, in the 011 method, the disease prevention record represents data on the immunization, vaccination, and health monitoring status of livestock herds. For example, the disease prevention record includes, but is not limited to, vaccination rates and immunization coverage rates. The circulation record represents the flow data of livestock products at various transaction and transportation stages. For example, the circulation record includes, but is not limited to, livestock transaction volume, quarantine pass rate, and daily increase in livestock transactions. Preset anomaly thresholds represent critical values ​​used to determine when the disease prevention record and circulation record deviate from the normal range. Each type of data corresponds to different anomaly thresholds and time thresholds. For example, the anomaly threshold for the vaccination rate can be 70%, and the time threshold can be 15 hours, etc.

[0031] In the 012 method, the risk interval represents the risk judgment domain jointly constituted by the anomaly threshold and the time threshold. For example, if the vaccination rate is below 70% or the quarantine pass rate is below 90%, a sliding window is used to calculate the duration for which the vaccination rate in a certain area is below 70%. If it exceeds 12 hours, it is considered an abnormal pattern. Simultaneously, if the daily increase in livestock transactions exceeds 30% and the quarantine pass rate is below 90%, the abnormal pattern could be a low immunization coverage accompanied by high transaction volume. A risk identification record is then generated; for example, the risk identification record could be a suspected outbreak risk of disease. That is, it can be understood that the abnormal pattern is the result of a combined judgment of multiple data points. By combining the judgment of the risk identification record, the complexity of the judgment can be increased, and the accuracy of the judgment can be improved. The risk identification record can be generated from past livestock risk cases and related feature databases stored in the cloud system 1000. Corresponding and comparing the epidemic prevention records with the circulation records can improve the accuracy of abnormal pattern judgment.

[0032] In the 013 method, the risk level is characterized by an assessment level based on the severity and scope of impact of the risk label record. For example, if the risk label record is "suspected major animal epidemic risk," the corresponding risk level can be assessed as "high"; if the risk label record is "general quarantine procedure violation," the risk level can be assessed as "medium" or "low." By matching risk intervals and risk characteristics, data can be transformed into risk label records with semantic information, thus realizing the identification of livestock risks.

[0033] Please see Figure 1 and Figure 4 In some implementations, generating a departmental response sequence includes: 021: Match the trigger field with the preset domain mapping table to obtain the business domain corresponding to the trigger field, and obtain the preset department association matrix based on the business domain; 022: Obtain collaboration feature scores based on the department association matrix, trigger fields, and business values ​​in the business dataset; if the collaboration feature scores meet the preset cross-department feature threshold, obtain the department set; and 023: Obtain the priority weight score of the departments based on the department association matrix and department set, and sort them in descending order based on the priority weight score to generate the department response sequence.

[0034] Specifically, in the 021 method, trigger fields are extracted from the initial report using natural language processing. For example, trigger fields include keywords such as the suspected disease upload time and quarantine instruction requirements. The domain mapping table represents a database of pre-defined correspondences between keywords and business domains. The Cloud System 1000 uses a keyword matching algorithm to compare the extracted trigger fields (such as "disease" and "quarantine") with the domain mapping table to determine the business domain to which the report belongs, such as animal disease prevention or market circulation. The department association matrix represents a pre-defined association data structure in the Cloud System 1000 used to define the collaborative relationships between different business domains and various functional departments. Logically, it is similar to a multi-table join query in a database, using business domain identifiers as indexes to quickly retrieve departments with overlapping functions or necessary collaboration with that domain. It should be noted that the department association matrix not only stores the departmental relationships but also pre-defined functional allocation weights for each department under different business domains to reflect the focus of each department's regulatory functions in different risk scenarios.

[0035] In the 022 method, the business value represents the quantitative data carried in the trigger field. The business value is the data inherent in the business dataset. For example, if the trigger field is "number of abnormal livestock deaths," the corresponding business value could be 50 heads. The collaboration feature score is used to determine whether cross-departmental collaboration is needed. Cloud System 1000 calculates the weights based on the association weights provided by the department association matrix, combined with the category of the trigger field and the business value. For example, setting the weight of "epidemic" to 0.8 and "isolation" to 0.7, if the combined score exceeds 1.2, or if the abnormal number in the business value accounts for 100% and the keyword matching degree exceeds 0.6, then the collaboration feature score is determined to meet the preset cross-departmental feature threshold, and the multiple departments involved are obtained to form a department set. If the collaboration feature score does not reach the cross-departmental feature threshold, then the business domain is determined to involve only a single department, and no further cross-departmental collaboration logic needs to be initiated. Introducing the department association matrix can identify collaboration needs, determine the necessity of cross-departmental collaboration, and avoid wasting resources.

[0036] In the 023 method, the priority weight score represents the order of intervention of each department in a set of departments within a specific business area. Since different types of risks impose different functional requirements on each department, Cloud System 1000 calculates the score for each department based on a departmental association matrix for the current business area. For example, in the field of animal disease prevention, the functional weight of the agricultural department is set to 0.6, and the functional weight of the market supervision department is set to 0.4. If the proportion of anomalies is 120%, then the priority weight score of the agricultural department is 0.72, and the score of the market supervision department combined with the circulation impact factor is 0.32. Cloud System 1000 sorts the departments in descending order based on these scores, with the highest-scoring department intervening first, thus generating a departmental response sequence. For example, the generated sequence might be that the agricultural department prioritizes disease prevention and control, followed by the market supervision department implementing circulation supervision, thereby ensuring the orderly distribution of regulatory tasks.

[0037] Please see Figure 1 and Figure 5 In some implementations, generating the departmental association matrix includes: 0221: Based on the response delay and processing success rate of departments for different business areas within a preset historical period, obtain the business relevance score of each department; 0222: Normalize the department set based on business relevance scores to obtain weight coefficients; and 0223: Generate a department association matrix based on the mapping relationship between business domain identifier, department identifier and weight coefficient; where the business domain identifier and department identifier are the codes of the business domain and department, respectively.

[0038] Specifically, in the 0221 method, the preset historical period characterizes the time span for extracting historical regulatory data from the system; for example, it can be set to the past 30 days, a quarter, or a year. Response latency characterizes the time difference between receiving an instruction and outputting feedback information for each department, used to assess the department's response speed. Processing success rate characterizes the ratio of the number of tasks completed as required by each department to the total number of tasks in the historical period. Business relevance score characterizes a comprehensive evaluation value based on historical performance, used to quantify the degree of matching between the department and a specific business area. For example, if the agricultural department's average response latency for animal disease prevention in the past 30 days is 1.5 hours and the processing success rate is 98%, then a higher business relevance score is obtained through a preset evaluation formula (such as a weighted ratio of success rate to response latency).

[0039] In the 0222 method, normalization is used to transform business relevance scores of different dimensions or magnitudes into a mathematical process with unified dimensions. This eliminates calculation biases caused by large differences in the original scores, ensuring that departments in the department set are compared under the same benchmark. The weight coefficient represents a standardized value obtained after normalization, used to characterize the importance of a department in a specific business area, and its value typically ranges from 0 to 1.

[0040] In the 0223 method, business domain identifiers represent the digital encoding of different business domains, and department identifiers represent the digital encoding of participating functional departments. By establishing a mapping relationship between business domain identifiers, department identifiers, and weight coefficients, the system ultimately generates a department response matrix. For example, the department response matrix can be constructed using the topological relationships between departments built from a knowledge graph. Cloud System 1000 abstracts business domain identifiers and department identifiers as core nodes in the graph, and defines the normalized weight coefficients as attribute edges connecting the nodes. Through this graph-based organization, for example, the business domain "Animal Epidemic Prevention" serves as the central node, pointing to the department identifiers "Agricultural Department" and "Market Supervision Department" through attribute edges. The attribute edges record corresponding weight coefficients, such as 0.6 and 0.4. When a new business requirement is triggered, Cloud System 1000 does not need to perform a full table scan, but instead quickly locates the associated path through the node index of the graph, thereby generating a department association matrix for the current business scenario.

[0041] Please see Figure 1 and Figure 6 In some implementations, obtaining the priority adjustment value includes: 051: Obtain the duration of the misalignment between the feedback timestamp and the preset time sequence; 052: Obtain the business importance weight corresponding to subsystem 101, and establish a product mapping relationship between the misalignment duration and the business importance weight; and 053: Determine the corresponding weight correction score based on the product mapping relationship, and set the weight correction score as the priority adjustment value.

[0042] Specifically, in method 051, the preset timing sequence represents the pre-arranged task execution plan that each subsystem 101 in the department response sequence should follow under ideal collaborative conditions. The feedback timestamp represents the physical time when subsystem 101 actually executes the task and returns feedback information in the cloud computing environment. The misalignment duration represents the absolute deviation of the feedback timestamp from the preset timing sequence. For example, cloud system 1000 parses the timestamp in the feedback timestamp. Assuming the sequence contains five business instructions with preset timing sequences of T1=10:00:00, T2=10:00:02, and T3=10:00:01, the timestamps are sorted in ascending order using a sorting algorithm (such as quicksort), resulting in the correct execution plan being T1, T3, and T2. If the actual feedback timestamp of T3 is 10:00:02.5, while T2 has already completed feedback at 10:00:02, it is determined to be a time misalignment, and the misalignment duration is calculated as 0.5 seconds by subtracting the two.

[0043] In method 052, cloud system 1000 obtains the business importance weights of each subsystem 101. For example, the weight of subsystem 101 responsible for epidemic isolation instructions (such as T3) is set to 0.8, while the weight of subsystem 101 responsible for routine data aggregation (such as T2) is set to 0.6. In method 053, the weight adjustment score is an adjustment index used to characterize the severity of timing conflicts, derived from the product mapping result. The priority adjustment value is the final quantitative result of the adjustment score. The calculated priority adjustment value for T3 is 0.4, and the priority adjustment value for T2 is 0.3. By comparing the magnitude of the priority adjustment values, cloud system 1000 can identify which subsystem 101 has a greater impact from timing misalignment, thus providing a quantitative basis for updating the department response sequence in step 05. Therefore, by calculating the misalignment duration, the execution deviation of subsystem 101 is perceived. Secondly, by introducing business importance weights and establishing a product mapping relationship, the limitations of comparing time in a single dimension are overcome. This ensures that priority adjustment considers not only "how long the misalignment lasted" but also "who is in the misalignment," guaranteeing that instructions receive higher priority scheduling weights when timing deviations occur. For example, the product mapping relationship can be "adjustment value = misalignment duration × weight." By integrating the misalignment duration with the business importance weight, a single time-dimensional deviation is mapped into a comprehensive feature score reflecting the degree of business impairment. For example, even if subsystem 101, which executes daily reports, experiences a significant misalignment duration due to network fluctuations, after product mapping with a weight coefficient of 0.3, its final quantification result will still be lower than that of the core business subsystem 101, which has a smaller misalignment duration but a weight coefficient of 0.8.

[0044] Please see Figure 1 and Figure 7 In some implementations, updating the departmental response sequence based on priority adjustment values ​​includes: 055: Reorder the department nodes in the department response sequence based on the priority adjustment value to obtain the initial execution sequence; 057: Perform conflict detection on the initial execution sequence to remove redundant instructions with overlapping timelines and obtain a conflict-free flow; and 059: Extract key node data from the conflict-free process, optimize and sort the key node data based on the decision tree algorithm, and generate an updated department response sequence.

[0045] Specifically, in method 055, the initial execution sequence represents the task execution queue initially reconstructed by cloud system 1000 based on feedback deviation. The reordering process represents the action of dynamically replacing the positions of each department node according to the priority adjustment value produced in step 05. For example, if cloud system 1000 calculates that the priority adjustment value of subsystem 101 responsible for epidemic isolation instructions is 0.4, while that of subsystem 101 responsible for routine data aggregation is 0.3, cloud system 1000, through comparison, finds that the former has a more severe conflict and a more urgent business, and thus moves the execution position of that node forward in the department response sequence, thereby obtaining the adjusted initial execution sequence.

[0046] In the method described in 057, conflict detection characterizes the spatiotemporal consistency verification of the initial execution sequence to identify whether there is resource contention or logical contradiction between instructions from different subsystems 101. Redundant instructions characterize instructions in the initial execution sequence that are repeatedly executed or invalidally occupied due to time overlap. For example, if the reordered sequence causes two subsystems 101 to need to access a shared database interface within the same time period, the cloud system 1000 will identify it as time overlap and remove one of the subsystems according to preset rules to release system bandwidth and computing resources, thereby obtaining a conflict-free process. This avoids the common instruction backlog and logical deadlock problems of subsystems 101 in a distributed environment.

[0047] In the 059 method, key node data represents tasks with high impact weights involving cross-departmental flow in conflict-free processes. Cloud system 1000 extracts key node data such as historical processing success rate and current task urgency of corresponding nodes in each subsystem 101 as feature inputs, and uses a decision tree algorithm for step-by-step path determination. For example, if the decision tree determines that "the historical success rate of a certain subsystem 101 is >95%" and "the current task urgency is urgent," then the optimal position of subsystem 101 in the sequence is further confirmed. Through heuristic sorting using the decision tree algorithm, an updated departmental response sequence is finally generated. The updated departmental response sequence serves as the basis for the final issuance of instructions by cloud system 1000, ensuring that the regulatory logic achieves optimal execution efficiency while avoiding physical conflicts.

[0048] Please see Figure 1 and Figure 8 In some implementations, it also includes: 071: Obtain the execution confirmation signal returned by the subsystem 101 based on the instruction, and obtain the instruction coverage data based on the confirmation signal and the department set; 072: Based on instruction coverage data, identify abnormal data in subsystem 101, classify the abnormal data using a support vector machine algorithm, and obtain the risk diffusion status of subsystem 101; and 073: Obtain instruction interaction permissions from subsystem 101 based on the risk diffusion status, and update the regulatory record based on the permission restriction results.

[0049] Specifically, in method 071, the instruction represents the task generated by cloud system 1000 based on the updated department response sequence and distributed to each subsystem 101. The execution confirmation signal represents the confirmation signal returned by subsystem 101 to cloud system 1000 after receiving or executing the instruction. The instruction coverage data represents the completeness of instruction distribution and the response ratio within each functional scope of the department set. For example, cloud system 1000 issues instructions through a distributed task scheduling system and collects the confirmation signals returned by each subsystem 101 in real time; by comparing the source of the confirmation signal with preset departments in the department set, if the response rate of subsystem 101 corresponding to each department reaches a preset threshold, complete instruction coverage data is obtained; if there is insufficient coverage, cloud system 1000 triggers a supplementary distribution process based on the instruction coverage data.

[0050] In the method of 072, anomalous data represents numerical characteristics that deviate from normal timing, logic, or success rate requirements during instruction execution, such as significant network latency or instruction execution failure rate. The cloud system 1000 identifies whether each subsystem 101 has anomalous data such as network latency or instruction execution failure rate based on instruction coverage data, and extracts the anomalous data as input vectors into the support vector machine algorithm. For example, the cloud system 1000 standardizes the indicators such as network latency, instruction execution failure rate, and resource utilization rate reported by the subsystem 101 and encapsulates them into input vectors. For example, if the network latency of a subsystem 101 is 15% and the instruction execution failure rate is 5%, a corresponding numerical vector is formed and input into the support vector machine model. The support vector machine algorithm determines the risk level by calculating the spatial distance and orientation of the vector to the classification hyperplane. For example, if the vector falls within a preset high-risk area, the cloud system 1000 determines that the risk diffusion status of the subsystem 101 is high-risk.

[0051] In the method described in 073, instruction interaction permission represents the operational authority of subsystem 101 to communicate data with other nodes in the cloud environment, call resources, or continue executing subsequent instructions. Permission restriction result represents the execution record of defensive actions such as isolation or rate limiting taken in response to the risk diffusion state. When the risk diffusion state indicates a high risk of diffusion, cloud system 1000, based on a preset prevention and control strategy, acquires and restricts the instruction interaction permission of the corresponding subsystem 101, such as closing communication ports or suspending its execution thread, to obtain a temporary state where the risk is under control. Subsequently, cloud system 1000 writes information including instruction distribution, coverage, risk diffusion state, and permission restriction result into the monitoring record. The monitoring record is stored in cloud system 1000.

[0052] By comparing the execution confirmation signals with the department set, the integrity of the execution of regulatory instructions was verified, and information blind spots in the distributed environment were eliminated through a supplementary distribution mechanism. Secondly, a support vector machine algorithm was introduced to classify abnormal data, enabling the cloud system 1000 to identify abnormal data from the confirmation signals returned by each subsystem 101.

[0053] Please see Figure 1 and Figure 9 In some implementations, after updating the regulatory record, the following is also included: 081: Extract historical response duration data based on regulatory records, and segment the historical response duration data to obtain duration distribution results; 082: Compare the duration distribution results with preset thresholds to determine key points, and obtain abnormal fluctuation points in the historical response duration data corresponding to the key points; 083: Obtain upstream and downstream interaction data for key points, and analyze the abnormal patterns of upstream and downstream interaction data and abnormal fluctuation points based on the random forest algorithm to generate an optimization suggestion sequence; and 084: Adjust the resource allocation strategy of each subsystem 101 based on the optimization suggestion sequence to obtain an optimized configuration scheme.

[0054] Specifically, in the method of 081, historical response time data is a record of the execution time of instructions by each subsystem 101 within a preset period, stored in the monitoring record. Segmentation processing represents the process of dividing continuous time series data and statistically analyzing its distribution characteristics. For example, the cloud system 1000 queries historical response time data for the past 30 days from the database, and calculates the average response time, maximum value, minimum value, and standard deviation for each hour using a time series analysis algorithm and a sliding window method, thereby obtaining the duration distribution results to identify execution efficiency fluctuations within the historical period.

[0055] In the method of 082, the preset threshold represents the time limit for determining whether the response speed of subsystem 101 meets the standard. Key points represent specific subsystems 101 or task segments where the response time exceeds the preset threshold and performance bottlenecks exist. Abnormal fluctuation points represent discrete data records in historical response time data where the specific value significantly deviates from the preset threshold. For example, assuming the preset threshold is 2.0 seconds, if the time distribution results show that the average response time of a certain segment is 2.5 seconds, and 30% of the records exceed 3.0 seconds, then cloud system 1000 identifies this segment as a key point and automatically marks the corresponding abnormal data as abnormal fluctuation points, generating an abnormal distribution map.

[0056] In the method of 083, upstream and downstream interaction data represent the communication and dependency relationships between the subsystem 101 corresponding to the key point and other adjacent subsystems 101 in the performance indicator chain. Abnormal patterns represent the classification of factors causing response latency, such as data processing blockage and network transmission latency. Optimization suggestion sequences represent combinations of solutions generated for the identified classifications. For example, cloud system 1000 extracts upstream and downstream interaction data for the key point and uses a random forest algorithm to calculate the time consumption weight of each step, such as database query time accounting for 60% and network latency accounting for 25%. If the bottleneck is determined to be processing blockage, an optimization suggestion sequence is generated, the content of which includes, but is not limited to, optimizing the database index, expecting to reduce time by 40%, introducing a caching mechanism, setting a hot data cache hit rate of 90%, and using asynchronous message queues for interfaces across subsystems 101.

[0057] In the method of 084, the optimized configuration scheme represents the final set of parameters adjusted to guide the operation of each subsystem 101. For example, cloud system 1000 automatically adjusts the CPU priority, database connection pool size, or bandwidth allocation of relevant subsystems 101 according to the optimization suggestion sequence, thereby obtaining the optimized configuration scheme. Subsequently, cloud system 1000 uses the optimized configuration scheme to update regulatory records and continuously monitors changes in response time at key points to determine the final improvement effect.

[0058] Please see Figure 1 and Figure 10 In some implementations, it also includes: 091: Based on the configuration scheme, optimize the performance indicators of subsystem 101 using a genetic algorithm, wherein the performance indicators include at least one of data processing speed, response time, and throughput; 092: Determine the triggering rules based on the optimized performance metrics; and 093: In the event of abnormal performance indicators, an alarm signal will be output based on the triggering rules.

[0059] Specifically, in the method of 091, the configuration scheme is a set of parameters used to adjust the resource weights of subsystem 101. Performance metrics characterize numerical indicators used to quantitatively evaluate the business processing capabilities of subsystem 101. For example, cloud system 1000, based on the optimized configuration scheme, uses a genetic algorithm to iterate the execution parameters of subsystem 101 multiple times. During the iterations, the system improves performance metrics through selection, crossover, and mutation operations.

[0060] In the method of 092, the triggering rule represents the logical criteria preset by the cloud system 1000 for determining abnormal states and initiating alarm actions. For example, the cloud system 1000 determines the triggering rule in reverse based on the optimized performance indicators as a benchmark. For instance, if the optimized response time of subsystem 101 should be 1.8 seconds, the triggering rule can be set as "trigger an alarm if the real-time response time exceeds 20% of the benchmark value".

[0061] In the method described in 093, the alarm signal is used to alert the cloud system 1000 to regulatory risks or subsystem 101 malfunctions. For example, the cloud system 1000 monitors the operational status of each subsystem 101. In cases of abnormal performance indicators, the cloud system 1000 executes the corresponding triggering rules. For instance, when a subsystem 101 experiences a sudden congestion of computing resources, causing its processing speed to drop below a preset triggering rule threshold, the cloud system 1000 automatically outputs an alarm signal to the relevant subsystem 101.

[0062] It should be noted that, in this document, the terms "comprising," "including," and any other variations are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Specific examples have been used in this document to illustrate the principles and implementation methods of the present invention. These examples are merely for the purpose of helping to understand the method and core ideas of the present invention. The above descriptions are only preferred embodiments of the present invention. It should be pointed out that, due to the limitations of written expression and the objective existence of infinite specific structures, those skilled in the art can make several improvements, modifications, or variations without departing from the principles of the present invention, and can also combine the above technical features in an appropriate manner. These improvements, modifications, variations, or combinations, or the direct application of the concept and technical solution of the present invention to other situations without modification, should all be considered within the scope of protection of the present invention.

Claims

1. A cloud-based digital collaborative supervision method for the entire livestock industry chain, characterized in that, include: Obtain the business dataset, and if the business dataset meets a preset exception pattern, generate an initial report based on the business dataset; The trigger field in the initial report is obtained, the business domain corresponding to the trigger field is identified, and if the trigger field meets the preset cross-departmental characteristics, a department set is obtained based on the business domain. The department set includes multiple departments, and the departments are sorted based on the preset priority weights under the business domain to generate a department response sequence. The department response sequence is sent to multiple subsystems of the departments, and feedback information output by the subsystems is obtained, wherein the feedback information includes a feedback timestamp; If the feedback timestamp does not conform to the preset timing sequence in the department response sequence, a suspension command is issued to the subsystem corresponding to the feedback timestamp; and Obtain a preset priority adjustment value, update the department response sequence based on the priority adjustment value, and issue instructions to the corresponding subsystem based on the updated department response sequence.

2. The regulatory method according to claim 1, characterized in that, If the business dataset meets a preset anomaly pattern, an initial report is generated based on the business dataset, including: The business dataset includes epidemic prevention records and circulation records. The data in the epidemic prevention records and circulation records are compared with preset anomaly thresholds and time thresholds. The time threshold is the duration for which the data remains within the anomaly threshold. If the anomaly threshold and the time threshold meet a preset risk range, the anomaly pattern is satisfied, and a preset risk feature is matched based on the anomaly pattern to obtain a risk identification record; and The risk level is obtained based on the risk identification record, and the initial report is obtained based on the matching result of the risk level data and historical data records.

3. The method according to claim 1, characterized in that, The generated department response sequence includes: The trigger field is matched with a preset domain mapping table to obtain the business domain corresponding to the trigger field, and a preset department association matrix is ​​obtained based on the business domain; Based on the department association matrix, the trigger field, and the business values ​​in the business dataset, a collaboration feature score is obtained; if the collaboration feature score meets a preset cross-department feature threshold, the department set is obtained; and The priority weight score of each department is obtained based on the department association matrix and the department set, and then sorted in descending order based on the priority weight score to generate the department response sequence.

4. The regulatory method according to claim 1 or 3, characterized in that, Generating the departmental association matrix includes: Based on the response delay and processing success rate of the departments for different business areas within a preset historical period, obtain the business relevance score of each department; The department set is normalized based on the business relevance score to obtain weight coefficients; and The department association matrix is ​​generated based on the mapping relationship between the business domain identifier, the department identifier, and the weight coefficient; wherein the business domain identifier and the department identifier are the codes for the business domain and the department, respectively.

5. The regulatory method according to claim 1, characterized in that, Obtaining the priority adjustment value includes: Obtain the misalignment duration of the feedback timestamp relative to the preset time sequence; Obtain the business importance weight corresponding to the subsystem, and establish a product mapping relationship between the misalignment duration and the business importance weight; and The corresponding weight correction score is determined based on the product mapping relationship, and the weight correction score is determined as the priority adjustment value.

6. The regulatory method according to claim 1, characterized in that, The step of updating the department response sequence based on the priority adjustment value includes: Based on the priority adjustment value, the nodes of each department in the department response sequence are reordered to obtain the initial execution sequence; Conflict detection is performed on the initial execution sequence to remove redundant instructions with overlapping timelines and obtain a conflict-free process; and Extract key node data from the conflict-free process, and optimize and sort the key node data based on the decision tree algorithm to generate the updated department response sequence.

7. The regulatory method according to claim 1, characterized in that, Also includes: Obtain the execution confirmation signal returned by the subsystem based on the instruction, and obtain instruction coverage data based on the confirmation signal and the department set; Based on the instruction coverage data, abnormal data of the subsystem is identified, and the abnormal data is classified based on the support vector machine algorithm to obtain the risk diffusion status of the subsystem. and Based on the risk diffusion status, obtain the instruction interaction permissions of the subsystem, and update the regulatory record based on the permission restriction results.

8. The regulatory method according to claim 7, characterized in that, The update of the regulatory records also includes: Based on the aforementioned regulatory records, historical response duration data is extracted, and the historical response duration data is segmented to obtain duration distribution results. The duration distribution results are compared with a preset threshold to determine key points, and abnormal fluctuation points in the historical response duration data corresponding to the key points are obtained. Acquire upstream and downstream interaction data of the key points, and analyze the upstream and downstream interaction data and the abnormal patterns of the abnormal fluctuation points based on the random forest algorithm to generate an optimization suggestion sequence; and Based on the optimization suggestion sequence, adjust the resource allocation strategy of each subsystem to obtain an optimized configuration scheme.

9. The regulatory method according to claim 8, characterized in that, Also includes: Based on the configuration scheme, the performance indicators of the subsystem are optimized using a genetic algorithm, wherein the performance indicators include at least one of data processing speed, response time, and throughput. Based on the optimized performance metrics, the triggering rules are determined; In the event of abnormal performance indicators, an alarm signal is output based on the triggering rules.

10. A cloud-based digital collaborative monitoring system for the entire livestock industry chain, characterized in that: When the cloud system is executed, it implements the steps of the cloud computing-based digital collaborative supervision method for the entire livestock supply chain as described in any one of claims 1 to 9.

Images