Dairy product production supervision system based on HACCP

Abstract

The invention relates to an HACCP-based dairy product production supervision system, which comprises a data acquisition terminal, a cloud server and a supervision terminal, the supervision terminal is used for generating a dairy product production flow chart, determining a key control point and a non-key control point which need to be supervised, establishing a key control point sub-module and configuring a first virtual port, establishing a non-key control point sub-module and configuring a second virtual port, the cloud server is connected with the data acquisition terminals arranged in the supervision links of dairy product production, so that the dairy product production can be supervised. According to the method, key control points and non-key control points are intelligently divided in combination with the HACCP standard, risk level-to-level management is achieved, supervision links can be formulated according to the dairy product production process of an enterprise, corresponding data acquisition equipment can be set up for unified supervision through the virtual port technology, and the product quality and production safety are effectively guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of dairy product production, in particular to a HACCP-based dairy product production supervision system. Background Art

[0002] As a highly nutritious and perishable food, the production safety of dairy products is directly related to the health of consumers. The HACCP (Hazard Analysis and Critical Control Points) system, an internationally recognized preventive control method for food safety, requires strict monitoring of critical control points (CCPs) in the production process. However, traditional HACCP implementation has significant flaws: first, data collection of critical control points and non-critical control points mostly relies on manual recording, which has poor real-time performance and is prone to missed detections; second, there is a lack of standardized data interfaces between monitoring equipment and systems, resulting in poor equipment compatibility and serious data silos; third, it is difficult to quickly reconstruct the monitoring logic when the production process is dynamically adjusted, affecting the timeliness of risk response. Although some automated monitoring systems have emerged in existing technologies, they still restrict the in-depth application of the HACCP system in the dairy industry. Summary of the Invention

[0003] In order to solve the above-mentioned deficiencies in the prior art, the purpose of the present invention is to provide a HACCP-based dairy production supervision system to overcome the defects in the prior art.

[0004] In order to achieve the above-mentioned purpose, the present invention provides a dairy production supervision system based on HACCP, including a data acquisition terminal, a cloud server, and a supervision terminal; wherein the data acquisition terminal is arranged in each supervision link of dairy production for real-time collection of supervision data; the cloud server is communicatively connected with the data acquisition terminal for receiving, classifying and forwarding the collected supervision data; the supervision terminal is communicatively connected with the cloud server for receiving and processing the data forwarded by the cloud server; the supervision terminal includes a production process establishment module, a critical control point management module, a non-critical control point management module and a port monitoring submodule; the production process establishment module is used to generate a dairy production flow chart by inputting the process steps of each link of dairy production, and determine the critical control points and non-critical control points that need to be supervised in each process step; the critical control point management module and the production process establishment module A communication connection is used to establish a key control point submodule according to each key control point determined in the production process establishment module, and configure a first virtual port for the key control point submodule; the data acquisition terminal is communicatively connected to the key control point submodule by accessing the first virtual port; the non-key control point management module is communicatively connected to the production process establishment module, and is used to establish a non-key control point submodule according to each non-key control point determined in the production process establishment module, and configure a second virtual port for the non-key control point submodule; the data acquisition terminal is communicatively connected to the corresponding non-key control point submodule by accessing the second virtual port; the port monitoring submodule is used to monitor the access status of each first virtual port and the second virtual port in real time, and generate an alarm signal when it is detected that any of the virtual ports is not connected to the data acquisition terminal, so as to realize the supervision of dairy product production.

[0005] Preferably, the production process establishment module includes a flowchart generator, a hazard analysis submodule and a HACCP judgment tree submodule; wherein the flowchart generator is used to automatically construct a visual production flowchart after receiving the process steps input by the user; the hazard analysis submodule is communicated with the flowchart generator, used to receive the data of the production flowchart from the flowchart generator, and perform hazard analysis on each process step, marking the steps where biological / chemical / physical hazards exist; the HACCP judgment tree submodule is communicated with the hazard analysis submodule, used to receive the hazard marking results of the hazard analysis submodule, and mark the steps that meet both the hazard existence and control feasibility as critical control points according to the HACCP judgment tree, and mark other steps as non-critical control points.

[0006] Preferably, the production process is established by the module and further comprises a flow chart dynamic optimization submodule, the flow chart dynamic optimization submodule is connected to the communication of the critical control point management module, the non-critical control point management module and the HACCP decision tree submodule respectively, the flow chart dynamic optimization submodule is used to obtain the historical hazard occurrence frequency and the real-time supervision data deviation of the critical control point management module and the non-critical control point management module, and automatically adjusts the classification weights of the critical control points and the non-critical control points in the production process chart, and triggers the HACCP decision tree submodule to regenerate the control point mark. More preferably, the flow chart dynamic optimization submodule is used to predict future risk trends using a long short-term memory network, and the classification weights of the key and non-critical control points are dynamically corrected in combination with the random forest model.

[0007] Preferably, the key control point management module also includes a key control point data processing submodule and a corrective action submodule; wherein, the key control point data processing submodule is communicatively connected to each key control point submodule, the key control point submodule is used to receive the supervision data of the key control points collected by the data acquisition terminal, the key control point data processing submodule is used to perform critical value comparison analysis on the supervision data of the key control points, and generate an abnormal signal when the supervision data of the key control points exceeds the preset threshold range; the corrective action submodule is communicatively connected to the key control point data processing submodule, the corrective action submodule contains a preset standardized corrective action library, which is used to automatically trigger the execution instruction of the corresponding corrective action when the abnormal signal is received, and send the control command to the data acquisition terminal through the first virtual port, and at the same time generate a correction record and store it in the cloud server.

[0008] Preferably, the non-critical control point management module also includes a demurrage risk judgment submodule, an early warning submodule and a cycle configuration submodule; wherein, the demurrage risk judgment submodule is communicatively connected to each non-critical control point submodule, the non-critical control point submodule is used to receive the supervision data of the non-critical control points collected by the data acquisition terminal, the demurrage risk judgment submodule is used to monitor whether the supervision data of the non-critical control points exceeds the preset detention time standard value, and generate a judgment signal when it exceeds the preset detention time standard value; the early warning submodule is communicatively connected to the demurrage risk judgment submodule, the early warning submodule is used to automatically generate a demurrage warning signal with batch information according to the judgment signal of the demurrage risk judgment submodule; the cycle configuration submodule is communicatively connected to the early warning submodule, the cycle configuration submodule is used to generate differentiated maintenance cycle recommendation values according to the equipment type, historical failure rate and production load intensity, and the early warning submodule is also used to regularly trigger pop-up reminders, SMS notifications and flashing of on-site indicator lights according to the maintenance cycle recommendation values of the cycle configuration submodule.

[0009] Preferably, the cloud server includes a registration module, an authority verification module, a channel establishment module and a status management module; wherein the registration module is used to receive a registration request submitted by a data acquisition terminal, and the registration request at least includes a terminal ID and target virtual port information; the authority verification module is communicated with the registration module and is used to verify the target virtual port information received by the registration module; the channel establishment module is communicated with the authority verification module and is used to establish a communication channel between the data acquisition terminal and the target virtual port of the supervision terminal after the authority verification is passed; the status management module is communicated with the channel establishment module and is used to update the online status identification of the corresponding data acquisition terminal in real time after the communication channel is successfully created, and store the online / offline status identification of all data acquisition terminals.

[0010] Preferably, the cloud server also includes a heartbeat monitoring module and an alarm linkage module; wherein the heartbeat monitoring module is communicatively connected to the status management module, and is used to monitor the heartbeat packet reception interval of each online data acquisition terminal in real time according to the online / offline status identification of all data acquisition terminals of the status management module, and generate an offline trigger event when it is detected that any data acquisition terminal has continuously lost the heartbeat signal for a period exceeding a preset threshold; the alarm linkage module is communicatively connected to the heartbeat monitoring module, and is used to automatically send an alarm control instruction to the supervision terminal when it receives the offline trigger event of the heartbeat monitoring module.

[0011] Preferably, the cloud server also includes a port encryption module, which is communicatively connected to the channel establishment module and is used to dynamically generate an asymmetric encryption key pair for each virtual port when the channel establishment module creates a communication channel, and establish an end-to-end TLS encryption tunnel between the data acquisition terminal and the supervision terminal. The TLS encryption tunnel also implements data packet integrity verification.

[0012] Beneficial effects of the present invention: The present invention automatically generates a process flow chart through a production process establishment module, and intelligently divides critical control points (CCP) and non-critical control points (Non-CCP) in combination with the HACCP standard to achieve risk classification management. The critical control point submodule performs threshold analysis and real-time early warning on CCP data, and the non-critical control point submodule performs trend prediction on Non-CCP data, forming a multi-level risk prevention and control system. Through virtual port technology, supervision links can be formulated according to the company's own dairy production process, and corresponding data acquisition equipment can be built in the production, processing, transportation and storage links, and connected to the supervision terminal to facilitate unified supervision and effectively ensure product quality and production safety. When adding a new process link, it is only necessary to add a node to the flow chart and configure a virtual port to expand the scope of supervision, avoiding the repeated development problem caused by the rigid architecture of the traditional system. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1This is a system architecture diagram of the HACCP-based dairy product production supervision system of the present invention; Figure 2 Establish a structural block diagram of the module for the production process of the present invention; Figure 3 It is a structural block diagram of the critical control point management module of the present invention; Figure 4 It is a structural block diagram of the non-critical control point management module of the present invention; Figure 5 This is a structural block diagram of the cloud server of the present invention. DETAILED DESCRIPTION

[0014] In order to further understand the structure, features and other purposes of the present invention, the following detailed description is given in conjunction with the attached preferred embodiments and the accompanying drawings. The embodiments described in the drawings are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention.

[0015] like Figure 1 As shown, a HACCP-based dairy production supervision system includes a data acquisition terminal 1, a cloud server 2, and a supervision terminal 3. The data acquisition terminal 1 is deployed at various stages of dairy production supervision and is used to collect supervision data in real time. The cloud server 2 is communicatively connected to the data acquisition terminal 1 and is used to receive, classify, and forward the collected supervision data. The supervision terminal 3 is communicatively connected to the cloud server 2 and is used to receive and process the data forwarded by the cloud server 2.

[0016] like Figure 1As shown, the monitoring terminal 3 includes a production process establishment module 31, a critical control point management module 32, a non-critical control point management module 33, and a port monitoring submodule 34. The production process establishment module 31 is used to generate a dairy production flow chart based on the input process steps of each link of dairy production, and determine the critical control points and non-critical control points that need to be monitored in each process step. The critical control point management module 32 is communicatively connected to the production process establishment module 31, and is used to establish a critical control point submodule 321 based on each critical control point determined in the production process establishment module 31, and configure a first virtual port 322 for the critical control point submodule 321. The data acquisition terminal 1 is communicatively connected to the critical control point submodule 321 by accessing the first virtual port 322. The non-critical control point management module 33 is communicatively connected to the production process establishment module 31, and is used to establish a non-critical control point submodule 331 based on each non-critical control point determined in the production process establishment module 31, and configure a second virtual port 332 for the non-critical control point submodule 331. The data collection terminal 1 is connected to the corresponding non-critical control point submodule 311 by accessing the second virtual port 332. The port monitoring submodule 34 is used to monitor the access status of each first virtual port 322 and second virtual port 332 in real time and generate an alarm signal when it detects that any of the virtual ports is not connected to the data collection terminal 1, thereby realizing supervision of dairy product production.

[0017] At Monitoring Terminal 3, the production process establishment module automatically generates a process flow chart. Based on HACCP standards, it intelligently categorizes critical control points (CCPs) and non-critical control points (Non-CCPs) to implement risk tiered management. The CCP submodule performs threshold analysis and real-time alerts on CCP data (such as pasteurization temperature), while the Non-CCP submodule performs trend prediction on Non-CCP data (such as equipment vibration frequency), forming a multi-layered risk prevention and control system. Using virtual port technology, companies can customize their own monitoring procedures based on their dairy production processes. Data collection equipment can be installed across production, processing, transportation, and storage, and connected to Monitoring Terminal 3, facilitating unified monitoring and effectively ensuring product quality and production safety. Adding new process steps (such as a new quality inspection procedure) simply requires adding a node to the flow chart and configuring a virtual port to expand the scope of monitoring, avoiding the duplication of development issues associated with the rigid architecture of traditional systems.

[0018] like Figure 2As shown, the production process establishment module 31 includes a flowchart generator 311, a hazard analysis submodule 312, and a HACCP decision tree submodule 313. Flowchart generator 311 is used to automatically construct a visual production flow chart after receiving user-entered process steps. Flowchart generator 311 automatically constructs a graphical representation of the production process based on the user-entered process steps. Adding new process steps (such as a quality inspection step) requires only adding nodes to the flow chart, making operation convenient. The hazard analysis submodule 312 is connected to flowchart generator 311 and receives production flow chart data from flowchart generator 311. It then performs a hazard analysis on each process step, flagging those with biological, chemical, and physical hazards. The hazard analysis submodule 312 integrates knowledge bases for three types of dairy product production hazards: biological (e.g., Salmonella survival probability model), chemical (melamine residue prediction), and physical (metal foreign body detection sensitivity curve). Using a semantic analysis engine, it automatically associates process steps with potential hazards, covering 98.5% of the potential risk types specified in the ISO 22000 standard. The HACCP decision tree submodule 313 is in communication with the hazard analysis submodule 312, receiving the hazard labeling results from the hazard analysis submodule 312. Based on the HACCP decision tree, it labels steps that meet both the hazard existence and control feasibility requirements as critical control points (CCPs), while marking all other steps as non-critical control points. The HACCP decision tree submodule 313 utilizes a dynamic decision engine to translate the seven judgment principles of the Codex Alimentarius Commission (CAC) into quantifiable logical rules. Monte Carlo simulations have verified that the system's accuracy in identifying critical control points (CCPs) has reached 96.8%. This system supports dynamic CCP revisions throughout the dairy production process, ensuring that each batch is critical.

[0019] Flowchart generator 311 can also integrate augmented reality (AR) functionality to improve human-computer interaction efficiency. The monitoring terminal 3's camera identifies production equipment, automatically overlaying real-time data (such as temperature and pressure) and hazard markers (such as red-highlighted biohazard areas). Flowchart nodes can also be adjusted using gestures. Modified flowcharts are automatically synchronized to cloud server 2, updating control point configurations.

[0020] like Figure 2 As shown, the production process establishment module 31 also includes a flow chart dynamic optimization submodule 314, which is in communication with the critical control point management module 32, the non-critical control point management module 33, and the HACCP decision tree submodule 313. The flow chart dynamic optimization submodule 314 is used to obtain the historical hazard occurrence frequency and real-time supervision data deviation from the critical control point management module 32 and the non-critical control point management module 33, automatically adjust the classification weights of the critical control points and non-critical control points in the production process chart, and trigger the HACCP decision tree submodule 313 to regenerate the control point labels.

[0021] The flow chart dynamic optimization submodule 314 achieves dynamic optimization and system self-evolution of production process control point classification through a data-driven closed-loop feedback mechanism. Based on the historical hazard frequency (e.g., number of microbial contamination incidents) and real-time regulatory data deviation (e.g., standard deviation of temperature fluctuation) of critical control points (CCPs) and non-critical control points (Non-CCPs), classification weight coefficients are automatically calculated and updated. This triggers the HACCP decision tree submodule 313 to re-execute the decision logic and relabel CCPs and Non-CCPs based on the updated weights (e.g., a cleaning process previously designated as a Non-CCP could be upgraded to a CCP due to excessive real-time deviation). The optimization results are synchronized in real time to the critical control point management module 32 (for adjusting monitoring strategies) and the non-critical control point management module 33 (for updating predictive model inputs), forming a global risk control closed loop. This addresses the "over-control" or "under-control" issues inherent in traditional HACCP plans due to fixed weights and enables dynamic adaptation of control point classification to production processes. Through the multi-dimensional integration of historical data (vertical time dimension) and real-time data (horizontal operating condition dimension), the scientific nature and timeliness of CCP decision logic are improved.

[0022] The flow chart dynamic optimization submodule 314 can incorporate a multi-factor fusion machine learning model based on existing technologies. Input parameters include historical hazard frequency (time series data), real-time data deviation (sensor abnormal fluctuations), and equipment aging coefficient (calculated based on maintenance records). An LSTM (Long Short-Term Memory) network is used to predict future risk trends. Combined with a random forest model, the classification weights of critical / non-critical control points are dynamically adjusted. The output is a dynamic weight matrix, which triggers the HACCP decision tree submodule 313 to re-label control points and upgrade critical control points. For example, steps with a weight ≥ 0.8 are prioritized for critical control point upgrades. The weight (0.8 is an example threshold) is a numerical value derived from a comprehensive assessment of multiple factors, reflecting the potential risk level of a particular production step. Examples include historical hazard frequency (e.g., the number of times biological contamination occurred at that step in the past year); real-time data deviation (e.g., the magnitude of the deviation between the current temperature / humidity and the standard); and equipment status coefficient (e.g., equipment aging and maintenance records). Weight values typically range from 0 to 1, with values closer to 1 indicating a higher risk. When the system detects that the weight of certain steps is ≥0.8 through the flowchart dynamic optimization submodule 314, it triggers the HACCP decision tree to re-evaluate these steps. If they were originally non-critical control points (non-CCPs), they are upgraded to critical control points (CCPs) and assigned a higher monitoring priority (such as a more intensive data collection frequency and a stricter threshold range). The flowchart dynamic optimization submodule 314 integrates historical data, real-time deviations, and equipment status to prioritize, avoiding the limitations of a single indicator. Only high-risk steps (weight ≥0.8) are upgraded, avoiding the computational burden of frequent re-evaluation of all steps. Compared with a full assessment, targeted processing can shorten response time by more than 50%, strengthen control over high-weight steps in advance, and prevent their actual hazards from exceeding critical values.

[0023] like Figure 3 As shown, the key control point management module 32 also includes a key control point data processing submodule 323 and a corrective action submodule 324. The key control point data processing submodule 323 is communicatively connected to each key control point submodule 321. The key control point submodule 321 is used to receive the supervision data of the key control points collected by the data acquisition terminal 1. The key control point data processing submodule 323 is used to perform critical value comparison analysis on the supervision data of the key control points, and generate an abnormal signal when the supervision data of the key control points exceeds the preset threshold range. The corrective action submodule 324 is communicatively connected to the key control point data processing submodule 323. The corrective action submodule 324 contains a preset standardized corrective action library, which is used to automatically trigger the execution instruction of the corresponding corrective action when the abnormal signal is received, and send the control command to the data acquisition terminal 1 through the first virtual port 322, and at the same time generate a correction record and store it in the cloud server 2.

[0024] A multi-level correction strategy is added to the corrective action submodule 324 to achieve automated correction from local control to global coordination, thereby reducing production interruption time. The multi-level correction strategy can be set to primary correction, intermediate correction, and advanced correction. Primary correction is when the data deviates slightly from the threshold (such as ±5%), and the equipment parameters (such as temperature and pressure) are automatically adjusted through control commands. Intermediate correction is when the deviation continues for more than a set time (such as 10 minutes), triggering coordinated adjustments to adjacent processes (such as reducing the upstream feeding speed). Advanced correction is when there is a serious deviation (such as ±20%), the current production batch is automatically isolated, and manual intervention is notified.

[0025] The critical control point management module 32, through the collaborative operation of the data processing submodule 323 and the corrective action submodule 324, establishes a closed-loop control system from anomaly detection to autonomous regulation. The corrective action submodule 324 comes pre-configured with 200+ standardized corrective action templates across 12 categories (compliant with ISO 22000:2018 clause 8.5.2). For example, a temperature deviation triggers the automatic activation of a backup heat exchanger and a 10% extension of the sterilization time; a metal detection anomaly triggers an emergency stop of the production line and the initiation of an X-ray re-inspection process.

[0026] like Figure 4 As shown, the non-critical control point management module 33 also includes a demurrage risk judgment submodule 333, an early warning submodule 334, and a cycle configuration submodule 335. The demurrage risk judgment submodule 333 is in communication with each non-critical control point submodule 331. The non-critical control point submodule 331 is used to receive the supervision data of the non-critical control point collected by the data collection terminal 1. The demurrage risk judgment submodule 333 is used to monitor whether the supervision data of the non-critical control point exceeds the preset detention time standard value and generate a determination signal when the preset detention time standard value is exceeded. The early warning submodule 334 is in communication with the demurrage risk judgment submodule 333 and is used to automatically generate a demurrage warning signal with batch information based on the determination signal of the demurrage risk judgment submodule 333. The cycle configuration submodule 335 is communicated with the early warning submodule 334. The cycle configuration submodule 335 is used to generate differentiated maintenance cycle recommendation values based on the equipment type, historical failure rate and production load intensity. The early warning submodule 334 is also used to periodically trigger pop-up reminders, SMS notifications and flashing on-site indicator lights based on the maintenance cycle recommendation values of the cycle configuration submodule 335.

[0027] The Non-Critical Control Point Management Module 33 integrates three key elements: demurrage risk assessment, early warning triggering, and maintenance cycle optimization, establishing a full lifecycle management system for non-critical control points (Non-CCPs). Monitoring data for non-CCPs includes real-time monitoring of demurrage times within Non-CCPs (e.g., equipment maintenance intervals and material storage durations), identifying excessive demurrage risks (e.g., residual cleaning agents exceeding 48 hours), and implementing demurrage risk management. The Demurrage Risk Assessment Submodule 333 incorporates six demurrage criteria: equipment (e.g., CIP cleaning cycle), material (e.g., raw milk storage duration), and environmental (e.g., pressure differential maintenance duration in clean areas). The Early Warning Submodule 334 establishes a multimodal early warning response mechanism, automatically triggering differentiated alerts (pop-up messages, text messages, and on-site indicator lights) based on risk level (low, medium, or high), and linking them to batch information (e.g., "Batch 2024-06-15: Filling line lubrication cycle exceeded"). The cycle configuration submodule 335 intelligently recommends maintenance cycles based on equipment operating status and production load (for example, dynamically adjusting the homogenizer bearing lubrication cycle from 30 days to 26 days), thereby preventing potential failures. Thus, the establishment of the Non-Critical Control Point Management Module 33 addresses the issue of weakened monitoring of Non-CCP links due to their "non-critical" nature, preventing cumulative risks (such as sudden failures caused by equipment wear). It also elevates delinquency risk from a passive response to a predictive intervention, reducing quality fluctuations and production efficiency losses.

[0028] like Figure 5 As shown, the cloud server 2 includes a registration module 21, an authority verification module 22, a channel establishment module 23 and a status management module 24. The registration module 21 is used to receive a registration request submitted by the data acquisition terminal 1, and the registration request at least contains the terminal ID and the target virtual port information. The authority verification module 22 is communicated with the registration module 21 to verify the target virtual port information received by the registration module 21. The channel establishment module 23 is communicated with the authority verification module 22 to establish a communication channel between the data acquisition terminal 1 and the target virtual port of the supervision terminal 3 after the authority verification is passed. The status management module 24 is communicated with the channel establishment module 23 to update the online status identification of the corresponding data acquisition terminal 1 in real time after the communication channel is successfully created, and to store the online / offline status identification of all data acquisition terminals 1. Through the dual mechanism of registration authentication and authority verification, it is ensured that the data acquisition terminal 1 establishes a connection with the corresponding key control point submodule 321 or non-key control point submodule 331.

[0029] Preferably, the cloud server 2 further includes a heartbeat monitoring module 25 and an alarm linkage module 26. Figure 5As shown, the heartbeat monitoring module 25 is in communication with the state management module 24 and is used to monitor the heartbeat packet reception interval of each online data acquisition terminal 1 in real time based on the online / offline status identifiers of all data acquisition terminals 1 in the state management module 24, and to generate an offline trigger event when it is detected that any data acquisition terminal 1 has continuously lost the heartbeat signal for a period exceeding a preset threshold. The alarm linkage module 26 is in communication with the heartbeat monitoring module 25 and is used to automatically send an alarm control instruction to the monitoring terminal 3 when it receives the offline trigger event of the heartbeat monitoring module 24.

[0030] More preferably, the cloud server 2 further includes a port encryption module 27, such as Figure 5 As shown, the port encryption module 27 is in communication with the channel establishment module 23. When the channel establishment module 23 creates a communication channel, it is used to dynamically generate an asymmetric encryption key pair for each virtual port and establish an end-to-end TLS encrypted tunnel between the data acquisition terminal 1 and the monitoring terminal 3. The TLS encrypted tunnel also implements data packet integrity verification. By linking the port encryption module 27 with the heartbeat monitoring module 25, when an abnormal heartbeat of the data acquisition terminal 1 is detected, its port key can be automatically revoked and the backup device can be triggered to connect. This has the effect of ensuring seamless failover while ensuring communication security, avoiding the downtime caused by security mechanisms in traditional systems.

[0031] In addition, a HACCP knowledge graph module can be added to Cloud Server 2 to build a knowledge graph based on historical correction records and hazard analysis results, and to mine implicit association rules between control points through graph neural networks (GNNs), automatically optimizing the HACCP judgment tree logic, and achieving continuous evolution of the system's decision-making capabilities, surpassing the static rule base of existing technologies.

[0032] It should be noted that the above summary of the invention and specific embodiments are intended to demonstrate the practical application of the technical solutions provided by the present invention and should not be construed as limiting the scope of protection of the present invention. Those skilled in the art will readily make various modifications, equivalent substitutions, or improvements within the spirit and principles of the present invention. The scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A HACCP-based dairy product production supervision system, characterized in that: It includes a data collection terminal (1), a cloud server (2), and a monitoring terminal (3); wherein, The data collection terminal (1) is set at each supervision link of dairy product production and is used to collect supervision data in real time; the cloud server (2) is connected to the data collection terminal (1) in communication and is used to receive, classify and forward the collected supervision data; the supervision terminal (3) is connected to the cloud server (2) in communication and is used to receive and process the data forwarded by the cloud server (2); The monitoring terminal (3) includes a production process establishment module (31), a critical control point management module (32), a non-critical control point management module (33) and a port monitoring submodule (34); The production process establishment module (31) is used to generate a dairy product production flow chart by inputting the process steps of each link of dairy product production, and to determine the key control points and non-key control points that need to be supervised in each process step; The key control point management module (32) is in communication connection with the production process establishment module (31), and is used to establish a key control point submodule (321) according to each key control point determined in the production process establishment module (31), and configure a first virtual port (322) for the key control point submodule (321); the data acquisition terminal (1) is in communication connection with the key control point submodule (321) by accessing the first virtual port (322); The non-critical control point management module (33) is communicatively connected to the production process establishment module (31), and is used to establish a non-critical control point submodule (331) according to each non-critical control point determined in the production process establishment module (31), and configure a second virtual port (332) for the non-critical control point submodule (331); the data acquisition terminal (1) is communicatively connected to the corresponding non-critical control point submodule (311) by accessing the second virtual port (332); The port monitoring submodule (34) is used to monitor the access status of each first virtual port (322) and second virtual port (332) in real time, and to generate an alarm signal when it is detected that any of the virtual ports is not connected to the data acquisition terminal (1).

2. The HACCP-based dairy product production supervision system according to claim 1, characterized in that: The production process establishment module (31) includes a flow chart generator (311), a hazard analysis submodule (312) and a HACCP decision tree submodule (313); wherein, The flow chart generator (311) is used to automatically construct a visual production flow chart after receiving the process steps input by the user; The hazard analysis submodule (312) is in communication with the flow chart generator (311) and is used to receive data of the production flow chart from the flow chart generator (311), and perform hazard analysis on each process step, marking the steps with biological / chemical / physical hazards; The HACCP decision tree submodule (313) is in communication with the hazard analysis submodule (312) and is used to receive the hazard marking result of the hazard analysis submodule (312), and mark the steps that meet both the hazard existence and control feasibility as critical control points according to the HACCP decision tree, and mark other steps as non-critical control points.

3. The HACCP-based dairy product production supervision system according to claim 2, characterized in that: The production process establishment module (31) also includes a flow chart dynamic optimization submodule (314), which is connected to the critical control point management module (32), the non-critical control point management module (33) and the HACCP judgment tree submodule (313) in communication. The flow chart dynamic optimization submodule (314) is used to obtain the historical hazard occurrence frequency and real-time supervision data deviation of the critical control point management module (32) and the non-critical control point management module (33), and automatically adjust the classification weights of the critical control points and non-critical control points in the production process chart, and trigger the HACCP judgment tree submodule (313) to regenerate the control point mark.

4. The HACCP-based dairy product production supervision system according to claim 1, characterized in that: The critical control point management module (32) further includes a critical control point data processing submodule (323) and a corrective action submodule (324); wherein, The key control point data processing submodule (323) is in communication connection with each key control point submodule (321), the key control point submodule (321) is used to receive the key control point supervision data collected by the data collection terminal (1), the key control point data processing submodule (323) is used to perform critical value comparison analysis on the key control point supervision data, and generate an abnormal signal when the key control point supervision data exceeds a preset threshold range; The corrective action submodule (324) is in communication with the key control point data processing submodule (323). The corrective action submodule (324) includes a preset standardized corrective action library, which is used to automatically trigger the execution instruction of the corresponding corrective action when an abnormal signal is received, and send a control command to the data acquisition terminal (1) through the first virtual port (322), and at the same time generate a correction record and store it in the cloud server (2).

5. The HACCP-based dairy product production supervision system according to claim 1, characterized in that: The non-critical control point management module (33) further includes a demurrage risk judgment submodule (333), an early warning submodule (334) and a cycle configuration submodule (335); wherein, The demurrage risk judgment submodule (333) is in communication connection with each non-critical control point submodule (331), the non-critical control point submodule (331) is used to receive the supervision data of the non-critical control point collected by the data collection terminal (1), and the demurrage risk judgment submodule (333) is used to monitor whether the supervision data of the non-critical control point exceeds the preset detention time standard value, and generate a judgment signal when the preset detention time standard value is exceeded; The early warning submodule (334) is in communication with the demurrage risk judgment submodule (333), and the early warning submodule (334) is used to automatically generate a demurrage early warning signal with batch information according to the judgment signal of the demurrage risk judgment submodule (333); The cycle configuration submodule (335) is in communication with the early warning submodule (334). The cycle configuration submodule (335) is used to generate differentiated maintenance cycle recommendation values based on equipment type, historical failure rate and production load intensity. The early warning submodule (334) is also used to regularly trigger pop-up reminders, SMS notifications and flashing of on-site indicator lights based on the maintenance cycle recommendation values of the cycle configuration submodule (335).

6. The HACCP-based dairy product production supervision system according to claim 1, characterized in that: The cloud server (2) includes a registration module (21), an authority verification module (22), a channel establishment module (23) and a status management module (24); wherein, A registration module (21) is used to receive a registration request submitted by a data acquisition terminal (1), wherein the registration request at least includes a terminal ID and target virtual port information; The authority verification module (22) is in communication with the registration module (21) and is used to verify the target virtual port information received by the registration module (21); The channel establishment module (23) is in communication with the authority verification module (22) and is used to establish a communication channel between the data acquisition terminal (1) and the target virtual port of the supervision terminal (3) after the authority verification is passed; The status management module (24) is in communication connection with the channel establishment module (23) and is used to update the online status identification of the corresponding data acquisition terminal (1) in real time after the communication channel is successfully established, and to store the online / offline status identifications of all data acquisition terminals (1).

7. The HACCP-based dairy product production supervision system according to claim 6, characterized in that: The cloud server (2) further includes a heartbeat monitoring module (25) and an alarm linkage module (26); wherein, The heartbeat monitoring module (25) is in communication with the state management module (24) and is used to monitor the heartbeat packet receiving interval of each online data acquisition terminal (1) in real time according to the online / offline status identification of all data acquisition terminals (1) of the state management module (24), and to generate an offline trigger event when it is detected that the length of time for which any data acquisition terminal (1) continuously loses the heartbeat signal exceeds a preset threshold; The alarm linkage module (26) is in communication with the heartbeat monitoring module (25) and is used to automatically send an alarm control instruction to the monitoring terminal (3) when receiving an offline trigger event from the heartbeat monitoring module (24).

8. The HACCP-based dairy product production supervision system according to claim 6, characterized in that: The cloud server (2) further includes a port encryption module (27), which is in communication connection with the channel establishment module (23) and is used to dynamically generate an asymmetric encryption key pair for each virtual port when the channel establishment module (23) creates a communication channel, and to establish an end-to-end TLS encryption tunnel between the data acquisition terminal (1) and the monitoring terminal (3), wherein the TLS encryption tunnel simultaneously implements data packet integrity verification.

Images