Dynamic interaction system for executing mechanism and measuring point of thermal power plant

Abstract

The invention discloses a thermal power plant execution mechanism and measuring point dynamic interaction system, and particularly relates to the technical field of execution interaction, the thermal power plant execution mechanism and measuring point dynamic interaction system comprises a perception data fusion module, a digital twin deduction module and a diagnosis prediction maintenance module, the output end of the perception data fusion module is in communication connection with the digital twin deduction module; and the input end of the diagnosis prediction maintenance module is in communication connection with the output end of the digital twinborn deduction module. The sensing data fusion module is used for integrating heterogeneous data sources through the multi-source data acquisition unit, the semantic understanding unit is used for constructing a measuring point knowledge graph, rapid analysis and processing of a large amount of data are achieved, faults are accurately positioned and the service life of a part is predicted through anomaly detection, reason analysis, service life prediction and work order generation, and the service life of the part is predicted. The security cooperative processing is more accurate, and the security is higher, so that the problem that the security cooperative processing is poor in accuracy and security is solved.

Description

Technical Field

[0001] This invention relates to the field of execution interaction technology, and more specifically, to a dynamic interaction system between actuators and measuring points in thermal power plants. Background Technology

[0002] The dynamic interaction system between actuators and measuring points in thermal power plants plays a key role in improving operational efficiency, ensuring equipment safety, optimizing environmental performance, and achieving intelligent management through real-time data interaction and intelligent control.

[0003] Among the existing publicly available documents, patent publication number CN120540236A discloses an integrated electrical control system for thermal power plants. This technology improves the system's collaborative working capability and control efficiency, enables timely and accurate diagnosis and handling of equipment faults, enhances the system's reliability and stability, and simultaneously achieves efficient energy utilization and conservation. However, this technology still has the following drawbacks.

[0004] During the dynamic interaction of measurement points in thermal power plants, the large amount of data generated during the process makes it difficult to quickly analyze and safely process the data, resulting in poor accuracy and security in the safety collaborative processing. Therefore, a dynamic interaction system between thermal power plant actuators and measurement points is provided. Summary of the Invention

[0005] To overcome the above-mentioned defects of the prior art, the present invention provides the following technical solution: a dynamic interaction system between the actuator and the measuring point of a thermal power plant, including a sensing data fusion module, a digital twin inference module, and a diagnosis, prediction and maintenance module, wherein the output end of the sensing data fusion module is communicatively connected to the digital twin inference module, and the input end of the diagnosis, prediction and maintenance module is communicatively connected to the output end of the digital twin inference module. It also includes an autonomous collaborative optimization module, a human-computer interaction decision-making module, and a security boundary protection module; The output of the autonomous collaborative optimization module is connected to the human-computer interaction decision module, and the output of the diagnostic prediction and maintenance module is connected to the autonomous collaborative optimization module. The human-computer interaction decision module is connected to the input of the security boundary protection module.

[0006] In a preferred embodiment, the perceptual data fusion module includes a multi-source data acquisition unit, a large model processing unit, a semantic understanding unit, a data alignment unit, and a data lake interface; Both the multi-source data acquisition unit and the semantic understanding unit are communicatively connected to the large model processing unit, and both the semantic understanding unit and the data lake interface are communicatively connected to the data alignment unit.

[0007] In a preferred embodiment, the processing order of the multi-source data acquisition unit, the large model processing unit, and the semantic understanding unit is arranged sequentially from front to back, and the processing order of the data alignment unit and the data lake interface is arranged sequentially from front to back.

[0008] In a preferred embodiment, the digital twin simulation module includes a large model simulation unit, a real-time synchronization unit, and a performance evaluation unit; Both the large model simulation unit and the performance evaluation unit are communicatively connected to the real-time synchronization unit.

[0009] In a preferred embodiment, the diagnostic prediction and maintenance module includes an anomaly detection unit, a cause analysis unit, a lifespan prediction unit, and a work order generation unit. After the anomaly detection unit detects abnormal data, the cause analysis unit analyzes it and the lifespan prediction unit makes a prediction. After the prediction, the work order generation unit generates a work order.

[0010] In a preferred embodiment, the anomaly detection unit processes data in a faster order than the cause analysis unit, and the lifetime prediction unit processes data in a faster order than the work order generation unit.

[0011] In a preferred embodiment, the autonomous collaborative optimization module includes an execution tuning unit, a dynamic optimization unit, an actuator health unit, and a safety processing unit; Both the execution tuning unit and the actuator health unit are electrically connected to the dynamic optimization unit, and the actuator health unit is connected to the safety processing unit via a wireless network.

[0012] In a preferred embodiment, the execution tuning unit, dynamic optimization unit, actuator health unit, and safety processing unit are arranged sequentially from front to back.

[0013] In a preferred embodiment, the human-computer interaction decision module includes an AR visualization unit, a scenario simulation unit, and an alarm management unit; The AR visualization unit, alarm management unit, and scenario simulation unit are all connected to the output of the human-computer interaction decision module, and the processing order of the AR visualization unit, scenario simulation unit, and alarm management unit is parallel.

[0014] In a preferred embodiment, the security boundary protection module includes a large-scale security engine unit, an interlocking protection unit, and an emergency self-healing unit; both the large-scale security engine unit and the emergency self-healing unit are communicatively connected to the interlocking protection unit.

[0015] The technical effects and advantages of this invention are as follows: 1. This invention utilizes a perception data fusion module to integrate heterogeneous data sources through a multi-source data acquisition unit, and a large model processing unit to repair and fill in the data. A semantic understanding unit constructs a knowledge graph of measurement points, and a data alignment unit ensures spatiotemporal alignment, enabling rapid analysis and processing of large amounts of data. A large model simulation unit is used to establish a traditional model, and a real-time synchronization unit combines deep learning and physical constraints to achieve millisecond-level real-time synchronization between the digital twin and the physical entity, improving the accuracy of collaboration. Through anomaly detection, cause analysis, life prediction, and work order generation, faults are accurately located and component life is predicted, generating maintenance plans. Safe collaborative processing is more accurate and safer.

[0016] 2. This invention employs an execution tuning unit to decompose the operating target and tune PID parameters in real time to improve control; an actuator health unit to assess the status of the actuator and avoid operating with defects; a safety processing unit to verify the safety of instructions and prevent dangerous operations, achieving safe collaborative processing with high accuracy and safety; an AR visualization unit to enable immersive inspection and maintenance and automatically generate reports; and a scenario simulation unit to simulate set scenarios, quickly analyze data and provide comprehensive results to assist in safe execution decisions.

[0017] 3. This invention uses a large-scale model security engine unit to monitor the execution mechanism's instructions and feedback in real time, preventing malicious attacks or logical errors, forming a dynamic security envelope, and converting complex safety procedures into executable logical rules for real-time verification. After system optimization, the large-scale model automatically verifies the completeness and effectiveness of the interlocking protection logic. The emergency self-healing unit automatically triggers a safety mode under extreme conditions, buying time for manual intervention. Overall, it achieves rapid analysis of large amounts of data, making safety collaborative processing more accurate and significantly improving security. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the operating structure of the dynamic interaction system between the actuator and the measuring point in a thermal power plant according to the present invention.

[0019] Figure 2 This is a schematic diagram of the operation of the digital twin simulation module of the present invention.

[0020] Figure 3 This is a schematic diagram of the operation of the diagnostic prediction and maintenance module of the present invention.

[0021] Figure 4 This is a schematic diagram of the autonomous collaborative optimization module of the present invention.

[0022] Figure 5 This is a schematic diagram of the operation of the human-computer interaction decision module of the present invention.

[0023] Figure 6 This is a schematic diagram of the operation of the human-computer interaction decision module of the present invention.

[0024] Figure 7 This is a schematic diagram of the operation of the security boundary protection module of the present invention.

[0025] The attached diagram is labeled as follows: 1. Perception Data Fusion Module; 2. Digital Twin Inference Module; 3. Diagnosis, Prediction, and Maintenance Module; 4. Autonomous Collaborative Optimization Module; 5. Human-Computer Interaction Decision Module; 6. Security Boundary Protection Module; 7. Multi-Source Data Acquisition Unit; 8. Large Model Processing Unit; 9. Semantic Understanding Unit; 10. Data Alignment Unit; 11. Data Lake Interface; 12. Large Model Simulation Unit; 13. Real-Time Synchronization Unit; 14. Performance Evaluation Unit; 15. Anomaly Detection Unit; 16. Root Cause Analysis Unit; 17. Lifetime Prediction Unit; 18. Work Order Generation Unit; 19. Execution Tuning Unit; 20. Dynamic Optimization Unit; 21. Actuator Health Unit; 22. Safety Processing Unit; 23. AR Visualization Unit; 24. Scenario Inference Unit; 25. Alarm Management Unit; 26. Large Model Safety Engine Unit; 27. Interlock Protection Unit; 28. Emergency Self-Healing Unit. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] like Figure 1 - Figure 7 The system illustrates a dynamic interaction system between actuators and measuring points in a thermal power plant. It includes a sensing data fusion module 1, a digital twin simulation module 2, and a diagnosis, prediction, and maintenance module 3. The output of the sensing data fusion module 1 is communicatively connected to the digital twin simulation module 2, and the input of the diagnosis, prediction, and maintenance module 3 is communicatively connected to the output of the digital twin simulation module 2. The system also includes an autonomous collaborative optimization module 4, a human-machine interaction decision-making module 5, and a safety boundary protection module 6. The output of the autonomous collaborative optimization module 4 is communicatively connected to the human-machine interaction decision-making module 5, and the output of the diagnosis, prediction, and maintenance module 3 is communicatively connected to the autonomous collaborative optimization module 4. The human-machine interaction decision-making module 5 is communicatively connected to the input of the safety boundary protection module 6.

[0028] In this embodiment, as Figure 2As shown, the perception data fusion module 1 includes a multi-source data acquisition unit 7, a large model processing unit 8, a semantic understanding unit 9, a data alignment unit 10, and a data lake interface 11. The multi-source data acquisition unit 7 and the semantic understanding unit 9 are both communicatively connected to the large model processing unit 8, and the semantic understanding unit 9 and the data lake interface 11 are both communicatively connected to the data alignment unit 10. The multi-source data acquisition unit 7 integrates heterogeneous data sources such as DCS, SIS, PLC, vibration sensors, and video images. After acquisition, the large model processing unit 8 uses the large model Time-LLM to identify data anomalies, noise, and missing data. The semantic understanding unit 9 utilizes the natural language processing capabilities of the large model to automatically parse the measurement point names and descriptions. The data alignment unit 10 solves the time and space alignment technology for data with different sampling frequencies and transmission delays, ensuring consistent data alignment processing. Finally, the processed high-quality data is written to the real-time data lake through the data lake interface 11.

[0029] In this embodiment, as Figure 2 As shown, the processing order of the multi-source data acquisition unit 7, the large model processing unit 8, and the semantic understanding unit 9 is arranged sequentially from front to back, as is the processing order of the data alignment unit 10 and the data lake interface 11. Processing is faster through the multi-source data acquisition unit 7, the large model processing unit 8, and the semantic understanding unit 9, and further accelerated through the data alignment unit 10 and the data lake interface 11.

[0030] In this embodiment, as Figure 3 As shown, the digital twin simulation module 2 includes a large-scale model simulation unit 12, a real-time synchronization unit 13, and a performance evaluation unit 14; both the large-scale model simulation unit 12 and the performance evaluation unit 14 are communicatively connected to the real-time synchronization unit 13. Simulation processing is performed through the digital twin simulation module 2. The large-scale model simulation unit 12 uses fluid mechanics and thermodynamics to establish a traditional white-box model of the equipment, and performs model processing operations. The real-time synchronization unit 13 utilizes LSTM and Transformer deep learning networks, and incorporates physical laws as constraints into the data-driven model using the large-scale model Physics-Informed.

[0031] In this embodiment, as Figure 4As shown, the diagnostic prediction and maintenance module 3 includes an anomaly detection unit 15, a cause analysis unit 16, a lifespan prediction unit 17, and a work order generation unit 18. After the anomaly detection unit 15 detects abnormal data, the cause analysis unit 16 analyzes it, and the lifespan prediction unit 17 makes predictions. The work order generation unit 18 then generates work orders. The processing order of the anomaly detection unit 15 is faster than that of the cause analysis unit 16, and the processing order of the lifespan prediction unit 17 is faster than that of the work order generation unit 18. The anomaly detection unit 15 combines multiple signals such as vibration, temperature, pressure, and current to detect whether early and complex anomaly data are the same as the set data, constructing a fault knowledge base of historical cases, maintenance records, and technical standards. Using large-scale model retrieval enhancement generation technology, when an anomaly is detected, it automatically retrieves similar cases and solutions, and then uses a large-scale model graph neural network (GNN) to analyze the causal graph between measurement points, locate the fault propagation path, and accurately locate the root cause. The lifespan prediction unit 17, based on digital twin and performance degradation data, predicts the RUL (Lifespan Limit of Components) of key components, providing a basis for spare parts procurement and maintenance planning, and generating maintenance work orders.

[0032] In this embodiment, as Figure 5 As shown, the autonomous collaborative optimization module 4 includes an execution tuning unit 19, a dynamic optimization unit 20, an actuator health unit 21, and a safety processing unit 22. Both the execution tuning unit 19 and the actuator health unit 21 are electrically connected to the dynamic optimization unit 20, and the actuator health unit 21 is connected to the safety processing unit 22 via a wireless network. The execution tuning unit 19, dynamic optimization unit 20, actuator health unit 21, and safety processing unit 22 are arranged sequentially from front to back. The execution tuning unit 19 receives the minimum coal consumption for power supply from the operating target and decomposes it into a series of specific control command sequences. The dynamic optimization unit 20 coordinates the boiler, turbine, and environmental protection facilities to avoid operation with defects before sending commands to the actuators through the actuator health unit 21, while meeting environmental and safety constraints. Finally, the safety processing unit 22 utilizes the logical reasoning capabilities of the large model.

[0033] In this embodiment, as Figure 6As shown, the human-computer interaction decision module 5 includes an AR visualization unit 23, a scenario simulation unit 24, and an alarm management unit 25. The AR visualization unit 23, alarm management unit 25, and scenario simulation unit 24 are all communicatively connected to the output of the human-computer interaction decision module 5. The processing order of the AR visualization unit 23, scenario simulation unit 24, and alarm management unit 25 is parallel. The safety boundary protection module 6 includes a large-scale model safety engine unit 26, an interlocking protection unit 27, and an emergency self-healing unit 28. The large-scale model safety engine unit 26 and emergency self-healing unit 28 are both communicatively connected to the interlocking protection unit 27. Through the AR visualization unit 23, operators can directly ask questions using voice or text. After understanding the question, the large model calls the backend module and generates the answer. Through the scenario simulation unit 24, operators can set the main steam pressure to increase by 0.5 MPa, and the system uses digital twins and the large model for simulation.

[0034] In this embodiment, as Figure 7 As shown, the safety boundary protection module 6 includes a large-scale safety engine unit 26, an interlocking protection unit 27, and an emergency self-healing unit 28; both the large-scale safety engine unit 26 and the emergency self-healing unit 28 are communicatively connected to the interlocking protection unit 27. By monitoring the action commands and feedback of all actuators, using a sequence model to detect abnormal operating modes, and calculating the upper and lower safety limits of each key parameter under the current operating conditions in real time, a dynamic safety envelope is formed. After the complex safety procedures and twenty-five countermeasures text knowledge are modified and optimized by the interlocking protection unit 27 using the system, the emergency self-healing unit 28 isolates faulty equipment under extreme operating conditions, resulting in better interactive processing.

[0035] The working principle of the dynamic interaction system between the actuator and the measuring point in a thermal power plant according to the present invention is as follows: Step 1: During the perception data fusion process, the multi-source data acquisition unit 7 in the perception data fusion module 1 integrates heterogeneous data sources such as DCS, SIS, PLC, vibration sensors, and video images. After acquisition, the large model processing unit 8 uses the large model Time-LLM to identify data anomalies, noise, and missing data, and performs intelligent repair and filling. Simultaneously, the semantic understanding unit 9 utilizes the natural language processing capabilities of the large model to automatically parse the measurement point names and descriptions, and automatically construct a measurement point knowledge graph. Then, the data alignment unit 10 solves the temporal and spatial alignment technology for data with different sampling frequencies and transmission delays, ensuring that the data is processed in the same alignment manner. Finally, the processed high-quality data is written to the real-time data lake through the data lake interface 11, providing unified data services for upper-layer applications and supplying service time.

[0036] Step 2: During digital twin simulation, all data is processed by the digital twin simulation module 2. The large model simulation unit 12 uses fluid dynamics and thermodynamics to build a traditional white-box model of the equipment, and performs model processing operations. Simultaneously, the real-time synchronization unit 13 uses LSTM and Transformer deep learning networks to learn the dynamic characteristics of the equipment from historical data. Furthermore, it utilizes the large model Physics-Informed to incorporate physical laws as constraints into the data-driven model, achieving more accurate and physically reliable state simulation and virtual sensing prediction of parameters that are difficult to measure directly. This ensures that the states of the digital twin and the physical entity are updated synchronously in real-time every 2 milliseconds.

[0037] Step 3, during diagnostic and predictive maintenance, the diagnostic and predictive maintenance module 3 begins the process. The anomaly detection unit 15 combines multiple signals such as vibration, temperature, pressure, and current to detect whether early and complex anomalies match predefined data. The cause analysis unit 16 constructs a fault knowledge base containing historical cases, maintenance records, and technical standards. Utilizing large-scale model retrieval enhancement generation technology, when an anomaly is detected, it automatically retrieves similar cases and solutions. Then, using a large-scale model graph neural network (GNN), it analyzes the causal graph between measurement points to locate the fault propagation path and accurately pinpoint the root cause. The lifespan prediction unit 17 uses digital twin and performance degradation data to predict the relative uptime (RUL) of critical components, providing a basis for spare parts procurement and maintenance planning. Finally, the work order generation unit 18 generates a maintenance work order, including a fault description, possible causes, handling suggestions, required tools and spare parts, and safety precautions.

[0038] Step 4, during autonomous collaborative optimization, the execution tuning unit 19 in the autonomous collaborative optimization module 4 begins to receive the operating target of minimum power supply coal consumption and fastest load change, decomposes it into a series of specific control command sequences, and automatically tunes the PID parameters in real time according to changes in operating conditions to improve control quality. The dynamic optimization unit 20 coordinates the boiler, steam turbine, and environmental protection facilities to dynamically optimize economic targets while meeting environmental and safety constraints. The health unit 21 sends instructions to the actuators to assess their current health status. If the status is poor, the control strategy is adjusted or backup equipment is switched to avoid operation with defects. Then, the safety processing unit 22 uses the logical reasoning ability of the large model to verify the safety and rationality of the generated control instructions to prevent dangerous operation.

[0039] Step 5, during human-computer interaction decision-making, operators can directly ask questions via voice or text through the AR visualization unit 23 in the human-computer interaction decision-making module 5, analyzing the root cause of the induced draft fan vibration at 2 PM yesterday. After understanding the question, the large model calls the backend module and generates the answer. Through AR glasses, the data of the digital twin, early warning information, and maintenance guidance are superimposed on the physical equipment to achieve immersive inspection and maintenance. The large model is used to cluster, compress, and sort the massive alarms by root cause. The large model automatically generates shift handover reports, daily reports, and weekly reports, including key indicator analysis, abnormal event summaries, and trend predictions. At the same time, through the scenario simulation unit 24, operators can set what the impact would be if the main steam pressure were increased by 0.5 MPa. The system uses the digital twin and the large model to perform simulations and provide a comprehensive result analysis, thus performing various data prediction processing.

[0040] Step Six: During safety boundary protection, the large-scale model safety engine unit 26 in the safety boundary protection module 6 monitors the action commands and feedback of all actuators. It uses a sequence model to detect abnormal operating modes to prevent malicious attacks or logical errors. It calculates the upper and lower safety limits of each key parameter under the current operating conditions in real time, forming a dynamic safety envelope. It transforms complex safety procedures and twenty-five countermeasure text knowledge into executable logical rules through the large model for real-time verification. After the interlock protection unit 27 is modified and optimized by the system, the large model automatically verifies whether the existing interlock protection logic is still complete and effective. Furthermore, after the emergency condition self-healing unit 28, the system can automatically trigger the preset safety mode under extreme conditions to isolate faulty equipment, stabilize unit operation, and buy time for manual intervention. This interactive processing effect is better.

[0041] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A dynamic interaction system between actuators and measuring points in a thermal power plant, comprising a sensing data fusion module (1), a digital twin simulation module (2), and a diagnostic prediction and maintenance module (3), characterized in that: The output end of the perception data fusion module (1) is communicatively connected to the digital twin inference module (2), and the input end of the diagnosis, prediction and maintenance module (3) is communicatively connected to the output end of the digital twin inference module (2). It also includes an autonomous collaborative optimization module (4), a human-computer interaction decision module (5), and a security boundary protection module (6); The output of the autonomous collaborative optimization module (4) is connected to the human-computer interaction decision module (5), and the output of the diagnostic prediction and maintenance module (3) is connected to the autonomous collaborative optimization module (4). The human-computer interaction decision module (5) is connected to the input of the security boundary protection module (6).

2. The dynamic interaction system between the actuator and measuring point in a thermal power plant according to claim 1, characterized in that: The perception data fusion module (1) includes a multi-source data acquisition unit (7), a large model processing unit (8), a semantic understanding unit (9), a data alignment unit (10), and a data lake interface (11). The multi-source data acquisition unit (7) and the semantic understanding unit (9) are both connected to the large model processing unit (8), and the semantic understanding unit (9) and the data lake interface (11) are both connected to the data alignment unit (10).

3. The dynamic interaction system between the actuator and measuring point in a thermal power plant according to claim 2, characterized in that: The processing order of the multi-source data acquisition unit (7), the large model processing unit (8), and the semantic understanding unit (9) is arranged from front to back, and the processing order of the data alignment unit (10) and the data lake interface (11) is arranged from front to back.

4. The dynamic interaction system between the actuator and measuring point in a thermal power plant according to claim 1, characterized in that: The digital twin simulation module (2) includes a large model simulation unit (12), a real-time synchronization unit (13), and a performance evaluation unit (14). The large model simulation unit (12) and the performance evaluation unit (14) are both connected to the real-time synchronization unit (13).

5. The dynamic interaction system between the actuator and measuring point in a thermal power plant according to claim 1, characterized in that: The diagnostic prediction and maintenance module (3) includes an anomaly detection unit (15), a cause analysis unit (16), a life prediction unit (17), and a work order generation unit (18). The anomaly detection unit (15) detects abnormal data, which is then analyzed by the cause analysis unit (16) and predicted by the life prediction unit (17). After prediction, the work order generation unit (18) generates the work order.

6. The dynamic interaction system between the actuator and measuring point in a thermal power plant according to claim 5, characterized in that: The processing order of the anomaly detection unit (15) is superior to that of the cause analysis unit (16), and the processing order of the life prediction unit (17) is superior to that of the work order generation unit (18).

7. The dynamic interaction system between the actuator and measuring point in a thermal power plant according to claim 1, characterized in that: The autonomous collaborative optimization module (4) includes an execution tuning unit (19), a dynamic optimization unit (20), an actuator health unit (21), and a safety processing unit (22). The execution tuning unit (19) and the actuator health unit (21) are both electrically connected to the dynamic optimization unit (20), and the actuator health unit (21) is connected to the safety processing unit (22) via a wireless network.

8. The dynamic interaction system between the actuator and measuring point in a thermal power plant according to claim 7, characterized in that: The execution tuning unit (19), dynamic optimization unit (20), actuator health unit (21), and safety processing unit (22) are arranged in sequence from front to back.

9. The dynamic interaction system between the actuator and measuring point in a thermal power plant according to claim 1, characterized in that: The human-computer interaction decision module (5) includes an AR visualization unit (23), a scenario simulation unit (24), and an alarm management unit (25). The AR visualization unit (23), alarm management unit (25), and scenario simulation unit (24) are all connected to the output of the human-computer interaction decision module (5). The processing order of the AR visualization unit (23), scenario simulation unit (24), and alarm management unit (25) is parallel.

10. The dynamic interaction system between the actuator and measuring point in a thermal power plant according to claim 1, characterized in that: The security boundary protection module (6) includes a large model security engine unit (26), an interlock protection unit (27), and an emergency self-healing unit (28). The large model safety engine unit (26) and the emergency condition self-healing unit (28) are both connected to the interlock protection unit (27).

Images