Flat industrial control integration architecture and method of ot / it / ct fusion

Abstract

The application provides a flat OT / IT / CT fusion industrial control integrated architecture and method, relates to the technical field of industrial control, and comprises an industrial control end, a core processing unit and an industrial automation management and control platform.The industrial control end is in communication connection with the core processing unit through an industrial time-sensitive network.The core processing unit comprises a memory relational database and an industrial intelligent agent TSN switch.The memory relational database receives parameters collected by the industrial control end, and the industrial intelligent agent performs real-time analysis on the parameters and generates control decision instructions for feedback.The industrial automation management and control platform manages the industrial control end and issues control instructions.The application reduces the level through a flat architecture, shortens the time delay, simplifies system integration, reduces the cost and improves the intelligent level.

Description

Technical Field

[0001] This invention relates to the field of industrial control technology, specifically to a flattened integrated industrial control architecture and method that converges OT / IT / CT. Background Technology

[0002] Currently, in the field of industrial automation control, traditional control systems typically follow a multi-level pyramid architecture, generally divided into management, control, and field levels. The management level provides the human-machine interface and is responsible for exchanging information with external systems; the control level acts as the core hub to implement system functions; and the field level consists of instruments and control devices, responsible for monitoring equipment status and executing control commands. However, this multi-level architecture has many limitations in practical applications. First, the vertical communication structure between levels is prone to communication delays and bottlenecks when the system is large-scale or the network load is too high, seriously affecting the real-time performance of control. Second, seamless integration between levels, especially between heterogeneous systems, requires complex engineering configurations and a large amount of integration work, leading to a significant increase in the initial construction cost and subsequent maintenance complexity of the system, easily forming "information silos." In addition, when it is necessary to add new functional modules or connect new field devices, it is often necessary to coordinate and upgrade the hardware and software of multiple levels, resulting in a long construction cycle and limited system flexibility and scalability. Finally, in the context of Industry 4.0 and smart manufacturing, traditional architectures are inadequate in achieving horizontal collaboration between devices, cloud-edge-device collaboration, big data analysis, and artificial intelligence applications, and are difficult to adapt to new production models that require high flexibility, rapid response, and intelligent decision-making. Summary of the Invention

[0003] To address the issues of high communication latency, complex system integration, and poor flexibility in existing multi-level architectures, this application proposes a flattened industrial control integrated architecture and method that integrates OT / IT / CT. By utilizing an in-memory relational database and industrial time-sensitive networks, it achieves flattened management, thereby reducing latency, simplifying system integration, and improving system scalability.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: A flat, integrated industrial control architecture that converges OT / IT / CT, including an industrial control terminal, a core processing unit, and an industrial automation management and control platform; The industrial control terminal is used to control industrial equipment, and it communicates with the core processing unit through an industrial time-sensitive network. The industrial control terminal is used to control industrial equipment, and the industrial control terminal communicates with the core processing unit through a TSN (Time-Sensitive Network) switch. The core processing unit includes a memory relational database, an industrial intelligent agent, and a TSN switch. The memory relational database is used to store real-time parameters collected by the industrial control terminal and share them with the digital middle platform. The industrial intelligent agent is communicatively connected to the memory relational database and is used to perform real-time analysis on the parameters of industrial equipment transmitted through the memory relational database, and generate corresponding control decision commands based on the analysis results and send them back to the industrial automation management and control platform. The industrial automation management and control platform configures the operating parameters of the industrial control terminal, synchronously sends them to the industrial control terminal, and is also responsible for remotely upgrading and maintaining the industrial control terminal program, monitoring the actual operation of the industrial control terminal, and providing open interface services to the digital middleware. The aforementioned digital platform is used to receive industrial equipment parameters, industrial intelligent agent analysis results, and industrial control terminal operation status data uploaded from the memory relational database, and to coordinate and analyze the data based on the industrial equipment parameters, analysis results, and operation status data.

[0005] In this solution, OT represents operational control technology for industrial production sites, focusing on real-time control of industrial equipment, on-site data acquisition, and actuator motion scheduling, and is the core underlying technology of industrial control. IT represents the technology system surrounding data storage, processing, analysis, and management, encompassing databases, cloud computing, big data analytics, and platform-based management, focusing on information flow and value extraction. CT represents the technology system enabling data transmission and interconnection between different devices and levels, encompassing network protocols, switches, network interface cards (NICs), and communication link design, serving as a bridge for data interaction.

[0006] Preferably, the core processing unit adopts a redundant architecture, which consists of at least two independent physical hosts, on which an in-memory relational database, an industrial intelligent agent, and an industrial automation management and control platform are deployed.

[0007] Preferably, the physical host is equipped with two independent network cards, which are connected to the industrial production network and the enterprise management network respectively. The communication between the industrial control terminal and the core processing unit is completed through the industrial production network, and the communication between the digital middle platform and the core processing unit is completed through the industrial production network or the enterprise management network according to the data type.

[0008] Preferably, the TSN switch uses a TSN chip to connect to the industrial control terminal and a TSN network card to connect to the memory relational database and industrial intelligent agent.

[0009] Preferably, the industrial control terminal consists of multiple integrated PLC controllers, each of which is connected to several industrial devices via an industrial bus.

[0010] In addition, the present invention also provides a flat industrial control integration method for OT / IT / CT convergence, including the following steps: the industrial control terminal collects the operating parameters of the connected industrial equipment, transmits the operating parameters to the memory relation database of the core processing unit through an industrial time-sensitive network, and at the same time the industrial control terminal receives the control instructions issued by the memory relation database and controls the operation of the industrial equipment according to the control instructions; The memory-based relational database stores real-time parameters collected by the industrial control terminal and shares them with the digital platform; the industrial intelligent agent is connected to the memory-based relational database to perform real-time analysis on the parameters of industrial equipment transmitted through the memory-based relational database, and generates corresponding control decision commands based on the analysis results and sends them back to the industrial automation management and control platform. The industrial automation management and control platform configures the operating parameters of the industrial control terminal, synchronously sends them to the industrial control terminal, is responsible for remote upgrades and maintenance of the industrial control terminal program, monitors the actual operation of the industrial control terminal, and provides open interface services to the digital middleware. The digital middle platform receives industrial equipment parameters, industrial intelligent agent analysis results, and industrial control terminal operation status data uploaded from the memory relational database, and performs overall planning and data analysis based on the industrial equipment parameters, analysis results, and operation status data.

[0011] As an alternative, a redundant backup step is also set up, specifically: when the memory relational database performs standardized storage and forwarding of top-level control instructions and operating parameters, it synchronously transmits the data to the memory relational database of other physical hosts to perform synchronous backup of the data.

[0012] Preferably, the industrial intelligent agent obtains the operating parameters of the industrial equipment from the memory relational database. When performing real-time analysis on the operating parameters, it determines whether the operating status of the industrial equipment is abnormal based on the preset control, analysis and decision linkage relationship. If an abnormality is determined, a corresponding fault control instruction is generated as a control decision instruction.

[0013] Beneficial effects: This invention establishes a flattened industrial control architecture by setting up a core processing unit and utilizing an in-memory relational database as a data interaction hub. This architecture connects downwards to the industrial control terminal, upwards to the digital platform, and horizontally to industrial intelligent agents. This simplifies the traditional three-layer architecture to a two-layer one, reducing intermediate layers, effectively shortening data transmission latency, and improving the system's real-time response capability. Simultaneously, based on the microsecond-level read / write speed and SQL openness of the in-memory relational database, it achieves open and real-time control, enabling rapid adaptation to equipment from different manufacturers, simplifying system integration solutions, and reducing overall project construction costs. Furthermore, the introduction of industrial time-sensitive networks and industrial intelligent agents establishes a control, analysis, and decision-making linkage mechanism, ensuring not only low latency and high reliability of network transmission but also enabling real-time intelligent analysis and decision-making of industrial equipment, thus improving the system's intelligence level. The core processing unit adopts a redundant architecture design and uses dual network cards to isolate the production network from the management network, ensuring high system reliability while achieving maintenance-free operation, further reducing maintenance costs. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the overall structure of the flat industrial control integrated architecture that integrates OT / IT / CT according to an embodiment of the present invention.

[0015] The components include: 1. Digital middleware, 2. Core processing unit, 3. Industrial control terminal, 4. TSN switch, 21. In-memory relational database, 22. Industrial intelligent agent, and 23. Industrial automation management and control platform. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0018] Example 1: like Figure 1As shown, this embodiment provides a flattened integrated industrial control architecture that converges OT / IT / CT. The term "flattened" in this embodiment refers to a structure that contrasts with the traditional three-tiered pyramid architecture of "management level - control level - field level" in industrial control systems. This embodiment, through the integrated design of the core processing unit, moves or merges the functions of the intermediate control layer in the traditional architecture, thereby logically forming a two-tiered architecture of "field device layer - core processing layer - digital management platform layer," effectively reducing intermediate data transmission links.

[0019] Industrial control terminal 3 is used to control industrial equipment and communicates with core processing unit 2 via TSN switch 4. Specifically, the industrial control terminal is located at the field device layer and is the core carrier of OT (Operational Technology). The industrial control terminal can be multiple integrated PLC controllers, embedded controllers, or intelligent I / O modules. Industrial equipment includes, but is not limited to, robotic arms, conveyor belts, CNC machine tools, sensors, actuators, and other field physical devices. The industrial control terminal connects to these industrial devices via industrial buses (such as ProfiNet, EtherCat, Modbus, etc.), and is responsible for collecting the operating parameters of the equipment (such as temperature, pressure, position, speed, etc.) and executing specific control actions. The industrial time-sensitive network, as the supporting network of CT (Computer-Sensitive Network), provides a microsecond-level low-latency transmission channel for data interaction between the industrial control terminal and the core processing unit, ensuring the real-time nature of control commands and feedback data.

[0020] The core processing unit includes a memory relational database 21, an industrial intelligent agent 22, and a TSN switch. The memory relational database is used to store real-time parameters collected by the industrial control terminal and share them with the digital middle platform. The industrial intelligent agent is communicatively connected to the memory relational database and is used to perform real-time analysis on the parameters of industrial equipment transmitted through the memory relational database, and generate corresponding control decision instructions based on the analysis results and send them back to the industrial automation management and control platform. The industrial automation management and control platform 23 configures the operating parameters of the industrial control terminal, synchronously sends them to the industrial control terminal, and is responsible for remotely upgrading and maintaining the industrial control terminal program, monitoring the actual operation of the industrial control terminal, and providing open interface services to the digital middleware.

[0021] In this embodiment, the core processing unit serves as the central nervous system of the entire architecture, achieving a deep integration of control, analysis, and management. Specifically, real-time data collected by the industrial control terminal is directly written to an in-memory relational database. The industrial intelligent agent reads data from the database for intelligent analysis (such as fault prediction and parameter optimization), and finally, the data is distributed from the industrial automation management platform to the industrial control terminal. This star-shaped interaction model centered on the database decouples the system components, ensuring that upgrades or changes to any component do not affect the operation of other components, greatly improving the system's scalability and maintenance efficiency.

[0022] Digital Platform 1 receives industrial equipment parameters, industrial intelligence analysis results, and operational status data from the industrial control terminal uploaded to the in-memory relational database. It then coordinates and analyzes these data based on the industrial equipment parameters, analysis results, and operational status data. Located at the top layer of the architecture, the digital platform falls under the IT category. It does not directly participate in the underlying real-time control loop but is responsible for production scheduling, plant-wide data visualization, energy efficiency analysis, and integration with enterprise ERP and MES systems. The digital platform obtains cleaned and standardized real-time production data through the in-memory relational database of its core processing unit, thereby achieving transparent management of the production process. Since all data is uniformly provided by the in-memory relational database, the digital platform does not need to concern itself with the heterogeneity of the underlying equipment (such as different brands of PLCs), focusing only on the data itself, thus achieving the flat management goal of "data as a service."

[0023] Through the above scheme, this embodiment constructs a flat architecture that deeply integrates OT / IT / CT. Compared with the traditional multi-layer architecture, this embodiment reduces the data forwarding links in the intermediate layers, significantly reducing communication latency; through the unified data hub of the in-memory relational database, it solves the data integration problem between heterogeneous systems, realizing data standardization and real-time sharing; at the same time, the introduction of industrial intelligent agents endows the system edge with intelligent decision-making capabilities, enabling the control system to respond quickly to production changes, improving production efficiency and intelligence level.

[0024] Example 2: Based on Example 1, this embodiment provides a detailed description of the communication network architecture between the industrial control terminal and the core processing unit, as well as the specific hardware form of the industrial control terminal.

[0025] Regarding the implementation of the communication network, the industrial time-sensitive network is equipped with a TSN switch. The TSN switch uses a TSN chip to connect to the industrial control unit and a TSN network interface card (NIC) to connect to the memory-based relational database and industrial intelligent agents. Specifically, the TSN switch acts as the central hub node of the entire industrial time-sensitive network, undertaking the task of precise scheduling and forwarding of data frames across the network. The industrial control unit (such as a PLC controller) integrates or externally connects to a TSN NIC or TSN chip, enabling it to have the physical interface capability to access the TSN network. The server NIC of the core processing unit is also connected to the TSN switch via fiber optic cable or twisted-pair cable, thus constructing a star topology high real-time network architecture.

[0026] To achieve time-sensitive characteristics, the TSN network in this embodiment uses the IEEE 802.1AS universal precise time protocol to synchronize the clock across the entire network, ensuring that the clock deviation between the TSN switch and each terminal device (industrial control terminal, core processing unit) is controlled within nanosecond precision. Simultaneously, the IEEE 802.1Qbv time-aware shaper is used for traffic scheduling, reserving dedicated time slots for control command data streams. This means that even with a large amount of non-real-time background traffic in the network (such as log uploads and video surveillance data), control commands can be transmitted without conflict and with low latency within the predetermined time slots. Compared to the "best-effort" transmission mechanism of traditional Ethernet, the TSN architecture in this embodiment can stably control end-to-end transmission latency within microseconds, and jitter within microseconds or even lower, thereby meeting the stringent requirements of industrial control for real-time performance and determinism, and avoiding the data packet loss or uncontrollable latency problems caused by network congestion in traditional architectures.

[0027] Regarding the specific form of the industrial control terminal, it refers to multiple integrated PLC controllers, each communicating with several industrial devices via an industrial bus. In practical applications, there are often numerous heterogeneous devices on-site, such as robotic arms, conveyor belts, and sensors from different brands. This embodiment adopts an "edge integration" strategy, utilizing multiple PLC controllers as edge control nodes. These PLC controllers can be general-purpose controllers developed based on the Codesys platform or dedicated controllers from specific brands. Each PLC controller connects to several industrial devices via an industrial bus (such as ProfiNet, EtherCat, ModbusTCP, etc.), responsible for underlying logic control, data acquisition, and action execution. This design retains the efficiency and compatibility of traditional industrial buses in device-level control without requiring large-scale modifications to the underlying devices.

[0028] Subsequently, these distributed PLC controllers no longer upload data through traditional multi-layered gateways. Instead, they are uniformly configured with TSN interfaces and directly connected to TSN switches, achieving "distributed control, centralized access." This architecture retains the flexibility of field control while establishing a high-speed data channel from the field layer to the core processing unit. This allows the core processing unit to directly obtain raw data from the underlying layers and issue high-priority control commands, truly realizing a flattened network architecture.

[0029] Example 3: Based on Example 1, this embodiment provides a detailed description of the high availability architecture of the core processing unit, the network security isolation mechanism, and the centralized management function of the industrial automation control platform.

[0030] Regarding the architecture of the core processing unit, a redundant architecture is adopted, consisting of at least two independent physical hosts. Each physical host deploys an in-memory relational database, an industrial intelligent agent, and an industrial automation control platform. Specifically, the redundant architecture preferably employs a dual-machine hot standby mode. The two physical hosts are, respectively, the active host and the standby host in standby mode. The two hosts are connected via a dedicated heartbeat cable or high-speed intranet to monitor each other's operating status in real time. When the host experiences hardware failure or software anomaly, the standby host can take over the host's IP address and services within milliseconds, ensuring uninterrupted service. In particular, due to the deployment of an in-memory relational database within the core processing unit, the data synchronization mechanism between the primary and standby hosts is greatly simplified and highly efficient. The in-memory relational database on the primary host synchronizes the latest industrial equipment parameters and control commands to the standby host's in-memory database through real-time log replication or shared storage mechanisms. This high-speed, memory-based synchronization mechanism ensures data consistency during switching, avoiding the data loss risk caused by the large synchronization latency of traditional disk databases, thereby achieving "zero-interruption" operation of the industrial control system.

[0031] To further ensure the network security of the industrial control system, the physical host is equipped with two independent network interface cards (NICs), one connected to the industrial production network and the other to the enterprise management network. Communication between the industrial control terminal and the core processing unit is completed through the industrial production network, while communication between the digital middleware and the core processing unit is completed through either the industrial production network or the enterprise management network, depending on the data type. This is a physical isolation strategy. The first NIC is bound to the industrial production network IP address and connects to the underlying industrial control terminal via a TSN switch, dedicated to transmitting high-real-time control commands and equipment status data, ensuring the closed and secure nature of the OT domain. The second NIC is bound to the enterprise management network IP address and connects to the enterprise's office network or cloud server, dedicated to non-real-time data interaction with the digital middleware, such as uploading production reports and integrating ERP data. Because the two NICs are physically independent, viruses or malicious attacks in the enterprise management network cannot directly penetrate the industrial production network through the network layer, thus eliminating network security risks at the physical level and building a robust "security defense depth."

[0032] In terms of centralized management of underlying devices, the industrial automation control platform, acting as the "management wing" of the core processing unit, provides full lifecycle management capabilities for industrial control terminals. In traditional control architectures, PLC program updates often require engineers to travel to the field with a programming computer and directly connect to the PLC via serial or Ethernet ports for programming, which is inefficient and carries the risk of misoperation. In this embodiment, the industrial automation control platform implements OTA (Over-the-Air) remote upgrade functionality. When a program update for a PLC controller is needed, maintenance personnel upload the new control program file to the control platform. The control platform breaks down the program file into data packets suitable for network transmission or converts it into standard SQL instruction format, writing it into a specific instruction table in the in-memory relational database. The client program running on the PLC controller side monitors or polls the instruction table in the in-memory relational database in real time. Once an upgrade instruction for this device is detected, the corresponding data packet is downloaded via the TSN network. After downloading, the PLC controller verifies and reassembles the data packet and writes it into its own program storage area, completing the remote upgrade. This process is carried out entirely through a unified in-memory relational database, enabling "penetrating" management of lower-level machines, greatly reducing operation and maintenance costs, and improving system maintainability.

[0033] Example 4: This embodiment provides a flattened, integrated industrial control method that converges OT / IT / CT. This method is implemented based on the architecture described in any one of embodiments 1 to 3, focusing on describing the flow and interaction logic of data between components, embodying a dynamic closed-loop process of "control-analysis-decision".

[0034] In step S100, the industrial control terminal collects the operating parameters of the connected industrial equipment and transmits the operating parameters to the memory relation database of the core processing unit through an industrial time-sensitive network. At the same time, the industrial control terminal receives the control instructions issued by the memory relation database and controls the operation of the industrial equipment according to the control instructions.

[0035] Specifically, the industrial control unit (such as a PLC controller) reads operating parameters of connected industrial equipment, such as sensors and actuators, in real time via fieldbuses (such as ProfiNet and EtherCat). These parameters include, but are not limited to, analog or digital signals such as temperature, pressure, flow rate, motor speed, and valve opening. The industrial control unit packages and encapsulates the collected raw data and sends it to the core processing unit via an industrial time-sensitive network composed of TSN switches through a configured TSN network card or TSN chip. Due to the time-sensitive nature of the TSN network, this transmission process can ensure microsecond-level low latency and extremely low jitter, ensuring the real-time performance of the control data. At the same time, the industrial control unit receives instructions or instruction notifications through asynchronous messages, reads the instructions, parses them into specific control actions (such as adjusting motor frequency or opening valves), and executes them, thereby achieving a high-speed closed loop control circuit.

[0036] Step S200: The memory relational database stores the real-time parameters collected by the industrial control terminal and shares them with the digital middle platform.

[0037] Specifically, the in-memory relational database, as the data hub of the entire architecture, undertakes the critical tasks of data cleaning and standardization. Since the underlying industrial equipment may come from different manufacturers and have varying data formats, the in-memory relational database, upon receiving raw data packets from the industrial control terminal, parses and stores them into standardized relational table structures according to preset mapping rules. For example, it stores temperature sensor data from different brands in a standardized format of "device ID-timestamp-temperature value." Similarly, for top-level control commands issued by the digital platform (such as production plan changes and process parameter adjustments), the in-memory relational database also converts them into standard data records for storage. This standardized storage mechanism eliminates the need for upper-layer applications to concern themselves with the heterogeneity of the underlying equipment, achieving "one-time data collection, network-wide sharing." The forwarding process refers to the in-memory relational database pushing the stored data in real time to industrial intelligent agents or industrial automation management platforms that have subscribed to relevant data topics, or waiting for these modules to actively read the data.

[0038] In step S300, the industrial intelligent agent communicates with the memory relational database to perform real-time analysis on the parameters of industrial equipment transmitted through the memory relational database, and generates corresponding control decision commands based on the analysis results and sends them back to the industrial automation management and control platform.

[0039] Specifically, the industrial agent, acting as an intelligent decision-making unit at the edge, operates on a "read-calculate-write-back" logic. The industrial agent reads the latest industrial equipment operating parameters from an in-memory relational database via a high-speed SQL query interface. Because the read / write speed of the in-memory database reaches microsecond levels, the industrial agent can obtain the field status almost without delay. The industrial agent has pre-built analysis algorithms or rule engines, such as PID parameter self-tuning algorithms, equipment fault diagnosis models, or energy consumption optimization strategies. The industrial agent performs real-time calculations on the read parameters to determine whether the equipment is in its optimal operating state or whether there are any abnormal risks. If the analysis indicates that adjustment is needed (e.g., temperature exceeds a preset threshold), the industrial agent immediately generates the corresponding control decision command (such as reducing heating power) and writes the command back to the command table in the in-memory relational database. This process eliminates the need for multi-layered forwarding in traditional SCADA or MES systems, achieving rapid self-decision-making at the edge.

[0040] Step S400 Industrial Automation Management and Control Platform configures the operating parameters of the industrial control terminal, synchronously sends them to the industrial control terminal, is responsible for remote upgrades and maintenance of the industrial control terminal program, monitors the actual operation of the industrial control terminal, and provides open interface services to the digital middleware platform.

[0041] Specifically, the industrial automation management platform is primarily responsible for the full lifecycle operation and maintenance management of industrial control terminals. The platform periodically reads operational status information from an in-memory relational database, such as CPU load, memory usage, network connection status, and program version number. Maintenance personnel can intuitively monitor the status of all controllers through the platform's human-machine interface. When centralized management operations are required, such as over-the-air (OTA) updates or parameter configuration distribution, maintenance personnel execute the operations on the platform, and the platform writes the generated control instructions (such as program packages and configuration files) into the in-memory relational database. The industrial control terminals receive these instructions or instruction notifications via asynchronous messages. This asynchronous message subscription method avoids the complexity of establishing point-to-point connections between the maintenance platform and each controller, achieving centralized, penetrating "one-to-many" management.

[0042] Through the above steps, this embodiment constructs a closed-loop data flow throughout the entire process. It should be understood that all data interactions in the above steps, whether it's the issuance of control commands or the uploading of status data, utilize an in-memory relational database as the sole intermediary for read and write operations, rather than point-to-point direct connections between modules. This database-based star-shaped interaction logic not only decouples the functional modules, making system expansion more flexible, but also ensures the real-time performance and determinism of the entire control method through the high-speed read and write characteristics of the in-memory database, truly achieving deep integration and flattened control of OT / IT / CT.

[0043] Example 5: This embodiment, based on the integrated control method described in Embodiment 4, provides a detailed explanation of the data high availability guarantee mechanism within the core processing unit and the specific decision-making logic of the industrial intelligent agent. These two aspects are crucial for ensuring the stable operation and intelligent control of the system in complex industrial environments.

[0044] Regarding high availability assurance for data, this embodiment also includes a redundant backup step. Specifically, when the in-memory relational database standardizes the storage and forwarding of top-level control commands and operating parameters, it synchronously transmits the data to the in-memory relational databases of other physical hosts for synchronous backup. Specifically, a high-speed data synchronization channel is established between the primary physical host and the backup physical host. The in-memory relational database uses a WAL (Write-Ahead Logging) mechanism for data management. When the host receives operating parameters uploaded from the industrial control terminal or top-level control commands issued by the digital platform, the database engine first writes these operation records to the log buffer, and then modifies the data pages in memory. This embodiment utilizes this mechanism to transmit the host's WAL logs to the backup physical host in real time via the network. After receiving the log stream, the backup host replays it in its local in-memory database in the same order, thereby achieving real-time mirror replication of the data. This log stream-based synchronization mechanism has the advantages of extremely low data transmission volume and minimal impact on host performance compared to traditional full data copying. Because in-memory databases offer extremely fast read and write speeds, this synchronization process can be completed in milliseconds or even microseconds, ensuring near-zero latency data consistency between the primary and backup machines. In the event of a primary physical host failure, the backup physical host can immediately take over the service with a minimal data loss window, thus guaranteeing the business continuity of the industrial control system.

[0045] Regarding the intelligent decision-making logic of the industrial intelligent agent, in this embodiment, the industrial intelligent agent obtains the operating parameters of industrial equipment from an in-memory relational database. When performing real-time analysis of these operating parameters, it determines whether the operating status of the industrial equipment is abnormal based on a preset control, analysis, and decision-making linkage relationship. If an abnormality is determined, a corresponding fault control instruction is generated as a control decision instruction. Specifically, the industrial intelligent agent has pre-set analysis models or rule bases for different industrial equipment. The so-called "pre-set control, analysis, and decision-making linkage relationship" can be a set of threshold rules based on expert experience, or a prediction model trained based on machine learning.

[0046] Taking a specific temperature control scenario as an example: An industrial intelligent agent reads temperature sensor values ​​from a reactor from a memory-based relational database at a set interval (e.g., every 100 milliseconds). Assume the preset linkage is "when the temperature exceeds 85℃ and the heating rate is greater than 2℃ / second, it is determined to be an abnormal temperature rise." The industrial intelligent agent compares the read real-time temperature value with the heating rate calculated from historical data, and then compares it with a preset threshold. If the comparison result meets the abnormal condition, the industrial intelligent agent immediately triggers the decision logic, generating a fault control instruction to "reduce heating power by 20%." This instruction is not sent directly to the underlying controller but is written back to the instruction table in the memory-based relational database. At this point, the memory-based relational database, acting as a data hub, notifies the industrial automation management platform to record the abnormal event and simultaneously pushes the instruction to the corresponding industrial control terminal (e.g., PLC controller) in real time via the TSN network. The industrial control terminal executes the instruction, thereby quickly suppressing the abnormal temperature rise.

[0047] This process demonstrates a closed-loop logic of "control-analysis-decision": the industrial control unit is responsible for executing control, the industrial intelligent agent is responsible for data analysis and strategy generation, and the in-memory relational database is responsible for the data flow and decoupling throughout the entire process. Through this mechanism, the system can automatically respond to some common equipment anomalies without human intervention, greatly improving the safety and automation level of the production process. It should be understood that the temperature control described above is only an example; in other embodiments, the industrial intelligent agent can also perform comprehensive analysis based on multi-dimensional parameters such as pressure, flow rate, and vibration to achieve more complex fault prediction and health management (PHM) functions.

[0048] Example 6: This embodiment uses the intelligent transformation of a welding workshop as an example to illustrate in detail the application effect of the flat industrial control integrated architecture that converges OT / IT / CT in a real production scenario. The welding workshop originally deployed multiple welding robots of different brands. Due to the inconsistent communication protocols and control interfaces used by the robots, severe "information silos" formed between the devices, making collaborative operation impossible. At the same time, the traditional control architecture had multiple layers, resulting in a control command issuance delay of up to 50ms, which was insufficient to meet the stringent real-time requirements of high-precision welding processes.

[0049] To address the aforementioned issues, the workshop is modified using the architecture described in this embodiment. First, edge integration is performed at the equipment layer. The industrial control terminal consists of multiple integrated PLC controllers, each communicating with several industrial devices via an industrial bus. Specifically, for welding robots of different brands, a suitable PLC controller is configured as its control core. These PLC controllers connect to the robot's servo drivers, welding power supplies, and other devices via ProFiNet or EtherCat buses, responsible for underlying logic control and data acquisition.

[0050] Subsequently, a high real-time communication network was constructed. The industrial time-sensitive network is equipped with a TSN switch, and the industrial control unit is configured with a TSN network card or TSN chip. The industrial control unit communicates with the TSN switch via the TSN network card or TSN chip, and the core processing unit communicates with the TSN switch. During the upgrade, a TSN network card supporting the IEEE 802.1AS protocol was added to each PLC controller, and connected to the TSN switch in the central cabinet of the workshop via fiber optic cable. The server of the core processing unit is also connected to this TSN switch, thus constructing a deterministic network with a star topology. This upgrade method does not require replacing the existing welding robot equipment; only edge controllers and network interfaces need to be added, greatly reducing the upgrade cost.

[0051] During production, the system demonstrates excellent real-time control and intelligent decision-making capabilities. The industrial agent retrieves operating parameters of industrial equipment from an in-memory relational database. When analyzing these parameters in real time, it determines whether the equipment's operating status is abnormal based on a preset control, analysis, and decision-making linkage. If an abnormality is detected, a corresponding fault control instruction is generated as a control decision instruction. For example, when a welding robot is performing spot welding, the PLC controller collects parameters such as welding current, voltage, and electrode pressure in real time and writes them to the core processing unit's in-memory relational database via the TSN network at microsecond speeds. The industrial agent reads this data in real time and analyzes it using a preset welding quality prediction model. When the agent detects abnormal fluctuations in the welding current (such as a sudden drop of more than 10%), it determines a risk of poor welding. At this time, the agent immediately generates a fault control instruction to "increase welding pressure" or "extend energizing time" and writes it back to the in-memory relational database. The PLC controller receives the instruction notification via asynchronous messages, reads the instruction from the database, and executes the adjustment. Actual testing showed that the end-to-end latency from the occurrence of an anomaly to the completion of instruction execution was reduced from 50ms in the traditional architecture to less than 5ms, effectively avoiding the occurrence of soldering defects.

[0052] Furthermore, operational efficiency has been significantly improved. The industrial automation management platform centrally manages PLC controllers and simultaneously distributes PLC controller operating programs and remote upgrade commands to the PLC controllers via an in-memory relational database. When the process department optimizes the welding control algorithm, maintenance personnel no longer need to go to the workshop site; they only need to upload the new PLC program file to the digital platform's management system. The management platform breaks down the program file into data packets and writes them into the in-memory relational database. After detecting the upgrade command in the database, the on-site PLC controller automatically downloads and completes the OTA remote upgrade. This process achieves unified management and penetrating maintenance of heterogeneous equipment, completely solving the tedious problem of engineers having to reprogram each device individually in the traditional model, improving operational efficiency by over 80%.

[0053] As can be seen from the application of this embodiment, the flat architecture provided by the present invention solves the real-time communication problem of heterogeneous devices through the TSN network, breaks down information silos through the memory relational database, realizes standardized data interaction, and realizes intelligent decision-making at the edge through industrial intelligent agents, which significantly improves the production efficiency, product quality and maintenance convenience of automotive welding production lines.

[0054] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A flat, integrated industrial control architecture that converges OT / IT / CT, characterized by: This includes industrial control terminals, core processing units, and industrial automation management platforms. The industrial control terminal is used to control industrial equipment, and the industrial control terminal is connected to the core processing unit through a TSN switch. The core processing unit includes a memory relational database, an industrial intelligent agent, and a TSN switch. The memory relational database is used to store real-time parameters collected by the industrial control terminal and share them with the digital middle platform. The industrial intelligent agent is communicatively connected to the memory relational database and is used to perform real-time analysis on the parameters of industrial equipment transmitted through the memory relational database, and generate corresponding control decision commands based on the analysis results and send them back to the industrial automation management and control platform. The industrial automation management and control platform configures the operating parameters of the industrial control terminal, synchronously sends them to the industrial control terminal, and is also responsible for remotely upgrading and maintaining the industrial control terminal program, monitoring the actual operation of the industrial control terminal, and providing open interface services to the digital middleware. The aforementioned digital platform is used to receive industrial equipment parameters, industrial intelligent agent analysis results, and industrial control terminal operation status data uploaded from the memory relational database, and to coordinate and analyze the data based on the industrial equipment parameters, analysis results, and operation status data.

2. The flat, integrated industrial control architecture merging OT / IT / CT according to claim 1, characterized in that, The core processing unit adopts a redundant architecture, which consists of at least two independent physical hosts. The physical hosts are equipped with an in-memory relational database, an industrial intelligent agent, and an industrial automation management and control platform.

3. The flattened industrial control integrated architecture merging OT / IT / CT according to claim 2, characterized in that, The physical host is equipped with two independent network cards, which are connected to the industrial production network and the enterprise management network respectively. The communication between the industrial control terminal and the core processing unit is completed through the industrial production network, and the communication between the digital middle platform and the core processing unit is completed through the industrial production network or the enterprise management network according to the data type.

4. The flattened industrial control integrated architecture merging OT / IT / CT according to claim 1, characterized in that, The TSN switch uses a TSN chip to connect to the industrial control terminal and a TSN network card to connect to the memory relational database and industrial intelligent agent.

5. The flat, integrated industrial control architecture merging OT / IT / CT according to claim 1, characterized in that, The industrial control terminal consists of multiple integrated PLC controllers, each of which communicates with several industrial devices via an industrial bus.

6. A flattened industrial control integration method integrating OT / IT / CT, based on the flattened industrial control integration architecture integrating OT / IT / CT as described in any one of claims 1-5, characterized in that, Includes the following steps: The industrial control terminal collects the operating parameters of the connected industrial equipment and transmits the operating parameters to the memory relation database of the core processing unit through an industrial time-sensitive network. At the same time, the industrial control terminal receives the control instructions issued by the memory relation database and controls the operation of the industrial equipment according to the control instructions. The memory-based relational database stores real-time parameters collected by the industrial control terminal and shares them with the digital platform; the industrial intelligent agent is connected to the memory-based relational database to perform real-time analysis on the parameters of industrial equipment transmitted through the memory-based relational database, and generates corresponding control decision commands based on the analysis results and sends them back to the industrial automation management and control platform. The industrial automation management and control platform configures the operating parameters of the industrial control terminal, synchronously sends them to the industrial control terminal, is responsible for remote upgrades and maintenance of the industrial control terminal program, monitors the actual operation of the industrial control terminal, and provides open interface services to the digital middleware. The digital middle platform receives industrial equipment parameters, industrial intelligent agent analysis results, and industrial control terminal operation status data uploaded from the memory relational database, and performs overall planning and data analysis based on the industrial equipment parameters, analysis results, and operation status data.

7. The flattened industrial control integration method based on OT / IT / CT fusion according to claim 6, characterized in that, Redundant backup procedures are also included, specifically: When the memory relational database standardizes the storage and forwarding of top-level control instructions and operating parameters, it synchronously transmits the data to the memory relational databases of other physical hosts for synchronous backup.

8. The flattened industrial control integration method based on OT / IT / CT fusion according to claim 6, characterized in that, The industrial intelligent agent obtains the operating parameters of industrial equipment from the memory relational database. When performing real-time analysis on the operating parameters, it determines whether the operating status of the industrial equipment is abnormal based on the preset control, analysis and decision linkage relationship. If an abnormality is determined, a corresponding fault control instruction is generated as a control decision instruction.

Images