Intelligent power plant dynamic security inspection method and system based on digital twinning

Abstract

This invention relates to the field of power plant safety management technology, specifically to a dynamic safety inspection method and system for smart power plants based on digital twins. The method includes: establishing a digital twin model of the power plant and deploying sensors to acquire real-time data, performing real-time mapping, and dynamically sensing the power plant's operating status; dividing the power plant area into multiple sub-regions, determining the risk level of each sub-region, and generating a dynamic adaptive inspection path; conducting sub-region inspections according to the inspection path, collecting current inspection data, dynamically adjusting the risk level of each sub-region, and issuing anomaly warnings based on the risk level. The system includes: a real-time operating status dynamic sensing module, a risk-adaptive inspection path planning module, and an inspection anomaly warning module. Through the above methods, a deep integration of dynamic path planning and real-time anomaly warning is achieved, along with a closed-loop mechanism for updating and optimizing inspection strategies, which can meet the high-precision and high-efficiency requirements of smart power plants for safety inspections.

Description

Technical Field

[0001] This invention relates to the field of power plant safety management technology, and in particular to a method and system for dynamic safety inspection of smart power plants based on digital twins. Background Technology

[0002] As the power industry transforms towards intelligence and digitalization, smart power plants are becoming a development trend. Safety inspections are a crucial link in ensuring the stable operation of power plants, but traditional inspection methods, primarily relying on manual inspections and fixed sensor monitoring, have many limitations. Manual inspections are affected by personnel experience, workload, and environmental factors, resulting in low efficiency, high missed inspection rates, and difficulty in real-time dynamic monitoring of equipment operating status. Fixed sensor monitoring has limited coverage, failing to achieve full-scene, all-round status perception; furthermore, sensor data is isolated and lacks effective fusion analysis, leading to delayed risk prediction and difficulty in responding to sudden failures.

[0003] Digital twin technology, by constructing a virtual mapping model of a physical entity, enables real-time monitoring, simulation analysis, and optimized control of physical objects, providing a new solution for safety inspections in smart power plants. Currently, some power plants are attempting to apply digital twin technology to the inspection field, but existing methods mostly remain at the level of static modeling and data display, failing to achieve deep integration of dynamic path planning and real-time anomaly early warning. Furthermore, the lack of a closed-loop mechanism for model updates and inspection strategy optimization fails to meet the high-precision and high-efficiency requirements of smart power plants for safety inspections.

[0004] Therefore, there is an urgent need for a dynamic safety inspection method for smart power plants based on digital twins to solve the problems of traditional inspections and improve the level of power plant safety management. Summary of the Invention

[0005] The purpose of this invention is to provide a dynamic safety inspection method and system for smart power plants based on digital twins. It aims to solve the technical problems in existing technologies that mostly stay at the level of static modeling and data display, fail to achieve deep integration of dynamic path planning and real-time anomaly early warning, and lack a closed-loop mechanism for model updates and inspection strategy optimization, thus failing to meet the high-precision and high-efficiency requirements of smart power plants for safety inspection.

[0006] To achieve the above objectives, this invention employs a dynamic safety inspection method for smart power plants based on digital twins, comprising the following steps: Establish a digital twin model of the power plant and deploy sensors to acquire real-time data from the power plant, perform real-time mapping, and dynamically perceive the operating status of the power plant. The power plant area is divided into multiple sub-regions, the risk level of each sub-region is determined, and a dynamic adaptive inspection path is generated. The inspection is carried out in sub-areas according to the inspection route, the current inspection data is collected, the risk level of the sub-area is dynamically adjusted, and anomaly warnings are issued according to the risk level.

[0007] Among the steps involved in establishing a digital twin model of the power plant, deploying sensors to acquire real-time data, performing real-time mapping, and dynamically sensing the power plant's operating status: Geometric, attribute, and process data were collected separately, a hierarchical model was built, and collaborative associations were performed to obtain a three-dimensional model of the power plant. Establish core points, deploy sensors at the core points, and acquire real-time data on equipment, environment, and personnel respectively; Set preset values, compare the changes in real-time data with the preset values, and trigger model parameter update requests based on the comparison results; Construct an interface that integrates equipment, environment, and personnel to display equipment operating parameters, a 3D model of the power plant, environmental parameters, and personnel trajectory data.

[0008] Among the steps involved in establishing a core point, deploying sensors at the core point, and acquiring real-time data on equipment, environment, and personnel: A dual-link transmission network was established to transmit real-time data, and invalid data from offline sensors was removed.

[0009] Among them, the steps of setting preset values, comparing the changes in real-time data with the preset values, and triggering model parameter update requests based on the comparison results are as follows: When the change is greater than or equal to the preset value, a model parameter update request is triggered and executed.

[0010] In the steps of dividing the power plant area into multiple sub-regions, determining the risk level of each sub-region, and generating a dynamic adaptive inspection path: Obtain the power plant's area, divide the area into multiple sub-regions, assign a unique code to each sub-region, and bind information to each sub-region; Select core indicators from multiple dimensions, determine the weights of the indicators, and construct a risk assessment indicator system; Dynamically set path planning constraints and generate dynamic adaptive inspection paths based on the risk level area division results.

[0011] In the step of calculating the risk level based on real-time data from the power plant: Based on real-time data from the power plant, extract data for each indicator and calculate the risk score for the sub-region based on the indicator data. The sub-regions are divided into multiple risk level zones based on the risk score.

[0012] After the steps of dynamically setting path planning constraints and generating a dynamic adaptive inspection path based on the risk level area division results: Obtain risk level adjustment data for the region and update the inspection route based on the adjustment data.

[0013] Among the steps, the process of conducting sub-area inspections according to the inspection route, collecting current inspection data, dynamically adjusting the risk level of the sub-area, and issuing anomaly warnings based on the risk level includes: Perform sub-area inspection operations, collect current inspection data, and generate sub-area inspection reports; Compare the current inspection data with the real-time power plant data acquired by the sensors to verify the current inspection data; Based on inspection data, the core indicators in the risk assessment indicator system are updated, and the risk levels of sub-regions are dynamically adjusted.

[0014] This includes the steps of updating the core indicators in the risk assessment indicator system based on inspection data and dynamically adjusting the risk levels of sub-regions: Based on the adjusted risk levels of the sub-regions, a multi-level early warning mechanism is set up and a response is initiated.

[0015] This invention also provides a smart power plant dynamic safety inspection system based on digital twins, including a real-time operating status dynamic perception module, a risk-adaptive inspection path planning module, and an inspection anomaly early warning module; wherein: The real-time operating status dynamic sensing module is used to establish a digital twin model of the power plant, and deploy acquisition sensors to obtain real-time data of the power plant, perform real-time mapping, and dynamically sense the operating status of the power plant. The risk adaptive inspection path planning module is used to divide the power plant area into multiple sub-regions, determine the risk level of the sub-regions, and generate dynamic adaptive inspection paths. The inspection anomaly early warning module is used to conduct sub-area inspections according to the inspection path, collect current inspection data, dynamically adjust the risk level of the sub-area, and issue anomaly early warnings based on the risk level.

[0016] This invention discloses a dynamic safety inspection method and system for smart power plants based on digital twins. The method comprises a real-time operational status dynamic perception module, a risk-adaptive inspection path planning module, and an inspection anomaly early warning module, which perform the following steps: establishing a digital twin model of the power plant and deploying sensors to acquire real-time data, performing real-time mapping, and dynamically perceiving the power plant's operational status; dividing the power plant area into multiple sub-regions, determining the risk level of each sub-region, and generating a dynamic adaptive inspection path; conducting sub-region inspections according to the inspection path, collecting current inspection data, dynamically adjusting the risk level of each sub-region, and issuing anomaly early warnings based on the risk level; through the above methods, a deep integration of dynamic path planning and real-time anomaly early warning is achieved, along with a closed-loop mechanism for updating and optimizing inspection strategies, which can meet the high-precision and high-efficiency requirements of smart power plants for safety inspections. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of the steps of the intelligent power plant dynamic safety inspection method based on digital twins according to the present invention.

[0019] Figure 2 This is a flowchart of steps S100 of the present invention.

[0020] Figure 3 This is a flowchart of steps S200 of the present invention.

[0021] Figure 4 This is a flowchart of steps S300 of the present invention.

[0022] Figure 5 This is a schematic diagram of the structural principle of the intelligent power plant dynamic safety inspection system based on digital twins according to the present invention.

[0023] Figure 6 This is a schematic diagram of the electronic device of the present invention.

[0024] 401 - Real-time operational status dynamic perception module, 402 - Risk adaptive inspection path planning module, 403 - Inspection anomaly early warning module. Detailed Implementation

[0025] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0026] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0027] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0028] Please see Figures 1-4 This invention provides a dynamic safety inspection method for smart power plants based on digital twins, comprising the following steps: S100: Establish a digital twin model of the power plant and deploy sensors to acquire real-time data from the power plant, perform real-time mapping, and dynamically perceive the operating status of the power plant.

[0029] In this embodiment, a digital twin model of the power plant is established, and sensors are deployed to acquire real-time data from the power plant, which is then mapped in real time to dynamically perceive the power plant's operating status. The specific process is as follows: S101: Collect geometric, attribute, and process data respectively, build a hierarchical model and perform collaborative association to obtain a three-dimensional model of the power plant; S102: Establish a core point, deploy sensors at the core point to acquire real-time data from equipment, environment, and personnel, build a dual-link transmission network, use the dual-link transmission network to transmit real-time data, and remove invalid data from sensors in offline states. S103: Set a preset value, compare the change in real-time data with the preset value, and if the change is greater than or equal to the preset value, trigger a model parameter update request and execute it; S104: Construct an interface for perceiving equipment, environment, and personnel, which displays equipment operating parameters, a 3D model of the power plant, environmental parameters, and personnel trajectory data, respectively.

[0030] In the above process, LiDAR scanning technology is used to perform a panoramic scan of the power plant building, equipment appearance, and pipeline routing to generate 3D point cloud data. For core equipment such as boilers, turbines, and transformers, detailed parameters such as equipment installation coordinates, interface dimensions, and moving part strokes are collected to ensure that the geometric data integrity is ≥99%. The point cloud data is imported into BIM software (such as Revit) for model reconstruction. Coordinate calibration tools are used to ensure that the geometric error between the virtual model and the physical equipment is ≤0.5%, meeting the visualization accuracy requirements during inspection.

[0031] Collect equipment factory parameters (such as rated power, rated temperature, vibration threshold), operating data (fault records, maintenance logs, repair reports), and production process flow diagrams (steam circulation process, electrical wiring diagrams, media transport paths); organize the data according to "equipment ID-parameter type-data format-update frequency" to form a structured data dictionary, and endow the model with physical attributes and functional logic.

[0032] Building a layered model: Equipment twin models are 3D visualization models of single devices built using the Unity engine. They integrate finite element analysis modules (such as ANSYS plugins) to simulate the operating states of equipment under different conditions (e.g., vibration distribution changes when the turbine rotor speed increases from 3000 r / min to 3200 r / min, temperature field distribution when the transformer load rate increases from 80% to 100%). Each equipment model is bound to a unique identifier ID, corresponding one-to-one with the sensor IDs of the physical equipment, supporting precise data association and status synchronization.

[0033] An environmental twin model is constructed within a virtual power plant space, including modules simulating parameters such as temperature, humidity, air pressure, dust concentration, and levels of harmful gases (CO, SO2). Based on the actual functional zones of the power plant (boiler workshop, power distribution room, control room, desulfurization zone), environmental monitoring sub-regions are defined, with at least one set of environmental sensors deployed in each sub-region to ensure comprehensive environmental data collection. Environmental parameters are integrated into the model in real-time via a data interface, enabling synchronized changes between the virtual and physical environments.

[0034] The process twin sub-model, based on Petri net technology, constructs a power plant production process model, reproducing the entire process logic of power generation, transmission, distribution, and environmental treatment (e.g., boiler ignition → steam generation → turbine drive → generator power generation → grid connection). The process model is linked with equipment and environmental models via the OPCUA industrial communication protocol. When a step in the process is triggered (e.g., boiler ignition), the equipment model (induced draft fan start-up, coal feeder coal delivery) and the environmental model (boiler workshop temperature increase, flue gas concentration change) are automatically activated, achieving multi-dimensional data collaborative interaction.

[0035] Equipment twin models, environmental twin models, and process twin models are combined to form a three-dimensional model of the power plant.

[0036] Sensor network deployment and real-time data acquisition: The core equipment sensors are deployed as follows: piezoelectric vibration sensors (measurement range 0~5000Hz, accuracy ±0.1mm / s) are installed at the turbine bearings to monitor the equipment vibration status in real time; pressure sensors (measurement range 0~1.6MPa, accuracy ±0.5%FS) are installed on the top of the transformer tank to prevent the risk of tank leakage; wireless temperature sensors (transmission distance ≤100m, refresh rate 1 time / minute) are installed at cable joints to avoid overheating and failure; and ultrasonic leak sensors (detection sensitivity ≤0.1bar pressure difference) are deployed for easily leaking pipelines and valves to achieve early leak warning.

[0037] Environmental and personnel sensor deployment, at a rate of 500m. 2 The density of Zone 1 is assessed by installing temperature and humidity sensors (measurement range 0~60℃, 20%~90%RH) and dust concentration sensors (detection range 0~100mg / mm³) inside the factory. 3 UWB personnel positioning base stations (positioning accuracy ≤30cm, supporting simultaneous positioning of more than 50 people) are installed at key nodes of the inspection channel (entrance, corner, and equipment-dense areas) to track the location of inspection personnel in real time; in high-risk areas such as high-voltage power distribution areas and fuel storage areas, additional 4K resolution video surveillance cameras (supporting motion detection and infrared night vision) are deployed to achieve visual monitoring.

[0038] Dual-link transmission network setup: A "5G + Industrial Ethernet" dual-link transmission architecture is adopted. Sensor data is preferentially transmitted to the factory edge computing gateway via 5G edge nodes (latency ≤20ms). Critical data (voltage, current, temperature) from core equipment (such as generators and main transformers) are simultaneously transmitted via industrial Ethernet backup to ensure no data loss. The edge gateway deploys data filtering rules to automatically remove invalid data when sensors are offline and outliers exceeding their measurement range (e.g., a temperature sensor displaying -50℃ is identified as faulty data and marked).

[0039] Lightweight data preprocessing: Kalman filtering algorithms are deployed on the edge gateway to eliminate noise in time-series data such as vibration and temperature (e.g., filtering high-frequency fluctuations caused by electromagnetic interference); frame extraction and feature recognition are performed on video data (e.g., using AI algorithms to identify whether personnel are wearing safety helmets or whether equipment is obstructed by foreign objects). The preprocessed data is pushed to the digital twin platform via the MQTT protocol, with the push frequency graded according to data importance (core equipment data once / second, environmental data once / 30 seconds, and video data once / 5 seconds).

[0040] Set a data synchronization trigger threshold. When the change in sensor data exceeds a preset value (e.g., temperature change ≥ 2℃, vibration amplitude change ≥ 0.2mm / s), the virtual model parameters will be automatically updated. For example, if the temperature of the physical transformer rises from 45℃ to 48℃ (exceeding the 2℃ threshold), the temperature display on the digital twin platform will be updated synchronously, and the temperature color will gradually change from "normal blue" to "warning yellow," intuitively reflecting the change in equipment status. If the vibration amplitude rises from 0.3mm / s to 0.6mm / s (exceeding the 0.2mm / s threshold), a red warning icon will flash at the corresponding location on the model (e.g., at the bearing).

[0041] A three-in-one perception interface integrating "equipment-environment-personnel" is constructed, which displays a list of equipment operating parameters (including real-time values, threshold ranges, and status indicators), a virtual power plant 3D model (which can be scaled and rotated, and details can be viewed by clicking on the equipment), and the real-time location and movement trajectory of inspection personnel.

[0042] S200: Divide the power plant area into multiple sub-regions, determine the risk level of each sub-region, and generate a dynamic adaptive inspection path.

[0043] In this implementation, the power plant area is divided into multiple sub-regions, the risk level of each sub-region is determined, and a dynamic adaptive inspection path is generated. The specific process is as follows: S201: Obtain the power plant area, divide the power plant area into multiple sub-regions, assign a unique code to each sub-region, and bind information to each sub-region; S202: Select multi-dimensional core indicators, determine indicator weights, and construct a risk assessment indicator system; S203: Extract data for each indicator based on real-time data from the power plant, and calculate the risk score for the sub-region based on the indicator data; S204: Divide the sub-region into multiple risk level areas based on the risk score; S205: Dynamically set path planning constraints and generate dynamic adaptive inspection paths based on the risk level area division results; S206: Obtain risk level adjustment data for the area and update the inspection path based on the adjustment data.

[0044] In the above process, the core principle for division is to follow the principle of "strong equipment correlation, close physical distance, and similar risk types" to avoid efficiency losses caused by cross-regional inspections. For example, the "boiler body - feedwater pump - economizer - air preheater" area is divided into the "boiler sub-area" (all equipment are boiler system-related components), and the "main transformer - high-voltage switchgear - surge arrester - voltage transformer" area is divided into the "distribution electronics area" (all are electrical system equipment); the area of ​​each sub-area is controlled between 500 and 1000 m². 2 Ensure that each inspection by patrol personnel takes no more than 30 minutes to avoid fatigue.

[0045] Multi-dimensional auxiliary adjustment: combining equipment importance (prioritizing the division of sub-areas for core equipment, such as generator set areas) and environmental risks (dust concentration ≥50mg / m³). 3 Areas with temperatures ≥40℃ are divided into separate sub-areas, and personnel accessibility is fine-tuned (high-altitude equipment platforms, underground cable trenches, and enclosed spaces are divided into separate sub-areas). For large single pieces of equipment such as desulfurization towers and chimneys, sub-areas are divided in layers from top to bottom (e.g., desulfurization towers are divided into "spray layer sub-areas," "demister layer sub-areas," and "slurry pool sub-areas") to ensure that the equipment in each sub-area can be fully inspected.

[0046] Each sub-region is assigned a unique code (e.g., "GL-01" represents boiler sub-region 01, "PD-03" represents power distribution sub-region 03), and is marked with different colored blocks on the power plant map of the digital twin platform, displaying sub-region details: including equipment list (e.g., GL-01 includes "#1 boiler body, #1 feed water pump, #1 economizer"), number of sensors (8 vibration sensors, 3 temperature sensors), number of historical faults (3 faults in the past year), and inspection cycle (1 time / day).

[0047] Basic information entry: The inspection requirements of sub-areas (such as the maximum inspection time of 25 minutes, the list of tools to be carried: infrared thermometer, vibration detector) and equipment maintenance cycle (such as the key inspection of the seals of the water pump every three months) are entered into the platform, and the scope of the impact of equipment failure in the sub-area is associated (such as the failure of the boiler sub-area may lead to the shutdown of the entire plant, while the local failure of the electronic distribution area only affects part of the power supply), providing basic data for risk level judgment.

[0048] Construction of a risk assessment indicator system: The core indicators were selected from three dimensions: "equipment status, environmental parameters, and historical risks," comprising eight key indicators as follows: The degree to which real-time equipment parameters exceed the standard (such as the percentage of temperature exceeding the threshold, the multiple by which vibration amplitude exceeds the standard). Equipment uptime (number of days since the last maintenance; exceeding the maintenance period increases the risk). Deviation of environmental parameters (such as the number of times the dust concentration exceeds the standard value, or the degree by which the temperature deviates from the normal range). Number of failures in the sub-region over the past three months (the more frequent the failures, the higher the risk); The number of high-risk equipment in a sub-region (e.g., the proportion of special equipment; the higher the proportion, the higher the risk). Frequency of personnel operations (average number of people entering per day; higher personnel density leads to increased risks). Emergency access channels are unobstructed (blockage increases risk, unobstructed channels reduce risk). External environmental factors (such as extreme weather such as heavy rain, high temperature, and lightning) can increase the probability of equipment failure.

[0049] Indicator weights and scoring criteria: The Analytic Hierarchy Process (AHP) was used to determine the indicator weights. "Degree of exceeding equipment real-time parameters" had the highest weight (0.25), followed by "Number of historical failures" (0.2). Other indicators were weighted according to their importance (0.15~0.05). A 100-point quantitative standard was established: 100 points for no equipment parameters exceeding the standard, 60 points for exceeding the standard by 5%~10%, and 30 points for exceeding the standard by more than 10%; 100 points for normal environmental parameters, 80 points for slight deviations (≤10%), and 40 points for serious deviations (>20%); 100 points for zero historical failures, 80 points for 1~2 failures, and 50 points for 3 or more failures.

[0050] Risk score calculation: The digital twin platform collects data on each indicator in real time and calculates the risk score for each sub-region according to the formula "Risk Score = Σ (Indicator Score × Indicator Weight)". For example, in a certain power distribution area, the "Degree of Exceedance of Real-Time Equipment Parameters" scores 60 points (weight 0.25), the "Number of Historical Failures" scores 80 points (weight 0.2), the "Degree of Deviation of Environmental Parameters" scores 100 points (weight 0.15), and all other indicators score 90 points. The final risk score is 60 × 0.25 + 80 × 0.2 + 100 × 0.15 + 90 × 0.4 = 82 points.

[0051] Risk Level Classification and Updates: Sub-areas are divided into four levels based on scores: Low Risk (85-100 points, green), Medium Risk (60-84 points, yellow), High Risk (30-59 points, orange), and Very High Risk (0-29 points, red). An update schedule is established, such as recalculating risk scores every 15 minutes. If a sub-area's risk level changes (e.g., from medium risk (82 points) to low risk (86 points), or from high risk (58 points) to high risk), the sub-area color on the digital twin map is automatically updated, and a level change notification is pushed to the inspection management terminal (e.g., "The risk level of the PD-03 electronics distribution area has increased from yellow to orange; please prioritize inspection").

[0052] The path planning constraints are set as follows: Time constraints: The total time for a single inspection task is ≤2 hours (including round-trip travel and data recording time). Safety constraints dictate that the route must avoid areas of equipment failure, temporary work areas, and emergency access routes (to avoid hindering emergency rescue efforts). Priority constraints apply: high-risk sub-areas must be inspected first, and their inspection time must account for no less than 40% of the total inspection time. Distance constraints: the inspection path distance between adjacent sub-regions is ≤500 meters to avoid backtracking and increasing invalid routes.

[0053] Dynamic constraint updates: The platform acquires real-time dynamic information about the power plant (such as sudden equipment failures requiring closure of a sub-area, temporary adjustments to inspection personnel, and the addition of new emergency inspection tasks) and automatically updates the constraints. For example, if "PD-02 power distribution area is closed due to equipment maintenance," this area will be automatically excluded during route planning; if a new emergency task "#2 generator abnormal noise" is added, the sub-area where the equipment is located (such as "FD-01 generator area") will be immediately included in the priority inspection scope.

[0054] Algorithm parameter optimization: Based on the traditional A* algorithm, a "risk level weight coefficient" (1.5 for high risk, 1.2 for medium risk, and 1.0 for low risk) and a "path smoothing coefficient" are introduced (to avoid sharp turns of more than 90 degrees in the path and reduce personnel fatigue). The center point of the sub-region is set as the path node, and the cost between nodes = physical distance × risk level weight coefficient. The algorithm generates the optimal path by minimizing the total cost (the lower the total cost, the better the path).

[0055] Path generation and visualization: The platform assigns tasks based on the number of inspectors (e.g., 3 inspectors) and the risk level of each sub-area: 1 inspector is responsible for high-risk / very high-risk areas (e.g., "GL-02 boiler sub-area" and "PD-03 electronics distribution area"), and 2 inspectors are responsible for medium- and low-risk areas. The generated path is displayed on the digital twin map as a "colored line + arrow," marking the estimated inspection time for each sub-area (e.g., 15 minutes for GL-01 sub-area) and key inspection equipment (#1 boiler body furnace wall temperature, feedwater pump sealing status). Inspectors can view the path navigation via mobile terminals, which displays their current location, distance to the next sub-area (e.g., "200 meters to PD-03 sub-area"), and remaining time (e.g., "45 minutes remaining for inspection").

[0056] The platform employs a dynamic path adjustment mechanism, adjusting trigger scenarios: when events occur such as "the risk level of a sub-area suddenly rises to extremely high risk," "inspection personnel discover unprepared anomalies (such as pipeline leaks)," or "the power plant temporarily adds an emergency task," the platform automatically triggers path adjustments. For example, if inspection personnel discover a pipeline leak in the "GL-02 sub-area" (risk level rises from orange to red), the platform immediately calculates and adjusts the path: prioritizing the personnel to complete a detailed inspection of the leak area (adding inspection items: leak location, leak volume), and postponing inspection tasks in low-risk areas (such as the "HJ-01 environmental protection sub-area").

[0057] Synchronization and Traceability After Adjustment: After the path is adjusted, the updated path will be pushed to the mobile terminal and simultaneously synchronized to the inspection monitoring panel. Managers can view the reason for the adjustment (e.g., "A pipeline leak was found in the GL-02 sub-area, priority inspection") and compare the paths before and after the adjustment (e.g., the original path "GL-01→HJ-01→PD-03" is adjusted to "GL-01→GL-02→PD-03"), ensuring that the inspection task is controllable and traceable.

[0058] S300: Conducts sub-area inspections based on the inspection path, collects current inspection data, dynamically adjusts the risk level of sub-areas, and issues anomaly warnings based on the risk level.

[0059] In this implementation, sub-area inspections are conducted according to the inspection path, current inspection data is collected, the risk level of each sub-area is dynamically adjusted, and anomaly warnings are issued based on the risk level. The specific process is as follows: S301: Perform sub-area inspection operation, collect current inspection data, and generate sub-area inspection report; S302: Compare the current inspection data with the real-time power plant data acquired by the sensors to verify the current inspection data; S303: Based on inspection data, update the core indicators in the risk assessment indicator system and dynamically adjust the risk level of sub-areas; S304: Based on the adjusted risk level of the sub-region, set up a multi-level early warning mechanism and respond accordingly.

[0060] During the above process, sub-area inspection operations are performed, current inspection data is collected, and a sub-area inspection report is generated. The report includes the total number of inspected equipment, the number of normal equipment, the number of equipment awaiting verification, the average value of key parameters, and the inspection time.

[0061] During the inspection, data is transmitted using a combination of real-time upload and batch upload: 1. Quantitative parameters (such as temperature and pressure) are uploaded in real time via the terminal's 5G module (delay ≤ 3 seconds). After receiving the data, the platform automatically matches the device ID with the sub-area code. 2. Qualitative data (such as photos and videos) are uploaded in batches after each device is inspected. The platform performs preliminary verification using AI algorithms (e.g., whether the photos are clear and whether they contain key parts of the equipment; blurry photos will prompt "Please retake photos of the #1 boiler weld").

[0062] For data consistency comparison and anomaly marking, the platform compares the inspection-collected data with two types of data: 1. Real-time sensor data (e.g., if the inspected water pump outlet pressure is 1.2 MPa, compared with the sensor data of 1.18 MPa, an error ≤2% is considered normal; otherwise, it is marked "data deviation, requires verification"); 2. Historical inspection data (e.g., if the average boiler wall temperature this time is 45℃, compared with 42℃ in the same period last week, if the increase is ≥5%, it is marked "abnormal temperature rise, requires attention"). After the comparison is completed, the platform generates a "data verification report" and synchronizes it to the terminal for personnel to view. If there are 3 or more anomaly markings, the inspection of the next sub-area must be suspended, and the current data must be verified first.

[0063] Based on inspection data, the three core indicators in the S200 risk assessment system were updated, with their weights dynamically adjusted according to actual scenarios: Equipment status indicators (weight increased to 0.3, from 0.25): If the inspection finds "leakage at the sealing surface of the #1 boiler feedwater pump" (qualitative abnormality) or "furnace wall temperature 48℃ (96% of the threshold of 50℃)" (quantitatively close to the threshold), the score of this indicator will be reduced from 80 points to 60 points; if "insulation resistance value of the #2 switch cabinet in the electronic distribution area is 100MΩ (threshold ≥100MΩ)", the score of the indicator will remain at 100 points.

[0064] Environmental impact indicators (weight 0.15 unchanged): If the inspection finds that the pH value of the spray tower water is 6.0 (threshold 6.5~8.5) in the HJ-02 sub-area, the score of this indicator will drop from 90 points to 50 points; if the environmental sensor data is consistent with the inspection data (e.g., humidity 50%), the score will remain unchanged.

[0065] Historical Risk Correction Indicator (New Weight 0.1): If the anomaly found in this inspection (such as water pump leakage) is in the same location as the same type of fault in the previous month, the score of this indicator will be reduced from 80 points to 40 points, strengthening the risk control of "repeated fault areas".

[0066] Reoperator region risk: Example 1: The original risk score of sub-area GL-01 was 82 points (medium risk yellow). This inspection found that the water supply pump was leaking (equipment status index 60 points), the environment was normal (80 points), and there were no repeated faults (80 points). The adjusted score = 60×0.3+80×0.15+80×0.1+other indicators (such as historical faults 80 points ×0.2) = 18+12+8+16=54 points, and the level was reduced from "medium risk yellow" to "high risk orange".

[0067] Example 2: The original score of sub-area PD-03 was 58 points (high-risk orange). In this inspection, all equipment was normal (equipment status 100 points) and environmental parameters met the standards (100 points). The adjusted score is 100×0.3+100×0.15+90×0.1+90×0.2=30+15+9+18=72 points, and the level is upgraded to "medium-risk yellow".

[0068] The adjustment results are synchronized and visualized. After the risk level is adjusted, the color of the sub-region in the digital twin model is updated synchronously (e.g., GL-01 changes from yellow to orange, flashing as a reminder). "Risk adjustment notifications" are pushed to the terminals of inspection personnel and the back-end of management personnel (e.g., "The risk of sub-region GL-01 has been raised to orange high risk, and the water pump leakage problem needs to be closely monitored"). The notification includes "adjustment basis" (e.g., details of changes in equipment status index scores) to ensure that the adjustment is traceable.

[0069] Early warning level classification and triggering conditions: Based on the adjusted risk level of the sub-region, a four-level early warning mechanism is set up. The triggering conditions are strongly linked to the risk score to avoid over-warning or omission.

[0070] Level 1 Warning (Minor) Response: The platform displays a green alert box in the "Early Warning Center" module (e.g., "GL-01 sub-area #1 boiler temperature 48℃, approaching the threshold"). Managers can click "Mark for Attention," and the system will automatically add the device to the "Next Inspection Priority List." No on-site handling is required; the inspection personnel can continue along the original path, and the terminal only records the early warning information without interrupting the inspection process.

[0071] Level II Warning (Attention) Response: The inspection personnel's terminal will issue a "beep" sound and light alert, and the screen will display the warning details (e.g., "PD-03 sub-area #2 switchgear pressure 1.6MPa, exceeding the standard by 6.7%)." They must click "Confirm Receive" within 5 minutes. Personnel must return to the equipment for a second verification (e.g., re-measure using a spare pressure gauge). If the parameter returns to normal after verification (e.g., pressure drops to 1.4MPa), the terminal will mark "Warning Released," and the platform will update the risk level accordingly. If the parameter still exceeds the standard after verification, it will automatically escalate to a Level 3 warning.

[0072] Level 3 Emergency Response: Upon triggering, the platform immediately executes three operations: 1. Pushes an "emergency response task" to the inspection personnel's terminal (e.g., "Immediately shut off the power to switch cabinet #2 in sub-area PD-03 and take photos of the faulty area"); 2. Sends an alert to the maintenance team's WeChat group (including sub-area location, fault description, and on-site personnel contact information); 3. Marks the faulty device with a red flashing box in the digital twin model and displays "Estimated arrival time of maintenance personnel: 15 minutes." Inspection personnel must complete preliminary handling (e.g., isolating the faulty device) within 10 minutes and upload the handling progress in real time (e.g., "Power off, waiting for maintenance personnel"). The platform records the handling timeliness (if not handled within the time limit, the alert level is upgraded to Level 4).

[0073] Level IV Warning (Critical) Response: Activate the power plant's emergency response mechanism: 1. The entire plant will sound an audible and visual alarm, and the emergency broadcast will repeatedly announce, "A cable joint smoke fault has occurred in the PD-03 sub-area. Non-emergency personnel are requested to evacuate." 2. The platform will automatically trigger the "emergency shutdown procedure" (e.g., cutting off the main power supply to the area) and synchronize it to the dispatch center. 3. The emergency management team will receive a rescue instruction containing the location of the digital twin model (e.g., "The fault point is located 3 meters northwest of the PD-03 sub-area"). They can view the fault propagation simulation through the model (e.g., "The cable ignition range that may be caused by high temperature") and formulate a rescue route. 4. Inspection personnel must immediately evacuate to a safe area and report "On-site personnel have evacuated" through the terminal. After the emergency team has completed its handling, a fault review will be conducted.

[0074] Corresponding to the aforementioned embodiments of the dynamic safety inspection method for smart power plants based on digital twins, this application also provides embodiments of a dynamic safety inspection system for smart power plants based on digital twins.

[0075] Figure 5 This is a block diagram illustrating a digital twin-based smart power plant dynamic safety inspection system according to an exemplary embodiment. (Refer to...) Figure 5 The system may include: a real-time operational status dynamic perception module 401, a risk-adaptive inspection path planning module 402, and an inspection anomaly early warning module 403; wherein: The real-time operating status dynamic sensing module 401 is used to establish a digital twin model of the power plant, and deploy acquisition sensors to obtain real-time data of the power plant, perform real-time mapping, and dynamically sense the operating status of the power plant. The risk adaptive inspection path planning module 402 is used to divide the power plant area into multiple sub-regions, determine the risk level of the sub-regions, and generate dynamic adaptive inspection paths. The inspection anomaly early warning module 403 is used to perform sub-area inspections according to the inspection path, collect current inspection data, dynamically adjust the risk level of the sub-area, and issue anomaly early warnings based on the risk level.

[0076] In this embodiment, the real-time operating status dynamic perception module 401 establishes a digital twin model of the power plant and deploys sensors to acquire real-time data from the power plant, performs real-time mapping, and dynamically perceives the operating status of the power plant. The risk-adaptive inspection path planning module 402 divides the power plant area into multiple sub-regions, determines the risk level of each sub-region, and generates a dynamic adaptive inspection path. The inspection anomaly early warning module 403 performs sub-region inspections according to the inspection path, collects current inspection data, dynamically adjusts the risk level of each sub-region, and issues anomaly early warnings based on the risk level. Through the above methods, a deep integration of dynamic path planning and real-time anomaly early warning is achieved, along with a closed-loop mechanism for updating and optimizing inspection strategies, which can meet the high-precision and high-efficiency requirements of smart power plants for safety inspections.

[0077] Regarding the system in the above embodiments, the specific ways in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0078] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0079] Accordingly, this application also provides an electronic device, including: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the above-described method for dynamic safety inspection of smart power plants based on digital twins. Figure 6 The diagram shown is a hardware structure diagram of any device with data processing capabilities within a smart power plant dynamic safety inspection system based on digital twins, as provided in an embodiment of the present invention. (Except for...) Figure 6 In addition to the processor, memory, and network interface shown, any data processing device in the embodiment may also include other hardware depending on the actual function of the data processing device, which will not be described in detail here.

[0080] Accordingly, this application also provides a computer-readable storage medium storing computer instructions, which, when executed by a processor, implement the aforementioned method for dynamic safety inspection of a smart power plant based on digital twins. The computer-readable storage medium can be an internal storage unit of any data-processing device as described in any of the foregoing embodiments, such as a hard disk or memory. The computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, smart media card (SMC), SD card, flash card, etc., equipped on the device. Furthermore, the computer-readable storage medium can include both internal storage units of any data-processing device and external storage devices. The computer-readable storage medium is used to store the computer program and other programs and data required by the data-processing device, and can also be used to temporarily store data that has been output or will be output.

[0081] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

[0082] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A dynamic safety inspection method for smart power plants based on digital twins, characterized in that, Includes the following steps: Establish a digital twin model of the power plant and deploy sensors to acquire real-time data from the power plant, perform real-time mapping, and dynamically perceive the operating status of the power plant. The power plant area is divided into multiple sub-regions, the risk level of each sub-region is determined, and a dynamic adaptive inspection path is generated. The inspection is carried out in sub-areas according to the inspection route, the current inspection data is collected, the risk level of the sub-area is dynamically adjusted, and anomaly warnings are issued according to the risk level.

2. The method for dynamic safety inspection of smart power plants based on digital twins as described in claim 1, characterized in that, In the steps of establishing a digital twin model of the power plant, deploying sensors to acquire real-time data from the power plant, performing real-time mapping, and dynamically sensing the power plant's operating status: Geometric, attribute, and process data were collected separately, a hierarchical model was built, and collaborative associations were performed to obtain a three-dimensional model of the power plant. Establish core points, deploy sensors at the core points, and acquire real-time data on equipment, environment, and personnel respectively; Set preset values, compare the changes in real-time data with the preset values, and trigger model parameter update requests based on the comparison results; Construct an interface that integrates equipment, environment, and personnel to display equipment operating parameters, a 3D model of the power plant, environmental parameters, and personnel trajectory data.

3. The method for dynamic safety inspection of smart power plants based on digital twins as described in claim 2, characterized in that, In the steps of establishing a core point, deploying sensors at the core point, and acquiring real-time data on equipment, environment, and personnel: A dual-link transmission network was established to transmit real-time data, and invalid data from offline sensors was removed.

4. The method for dynamic safety inspection of smart power plants based on digital twins as described in claim 2, characterized in that, In the steps of setting preset values, comparing the changes in real-time data with the preset values, and triggering a model parameter update request based on the comparison results: When the change is greater than or equal to the preset value, a model parameter update request is triggered and executed.

5. The method for dynamic safety inspection of smart power plants based on digital twins as described in claim 1, characterized in that, In the steps of dividing the power plant area into multiple sub-regions, determining the risk level of each sub-region, and generating a dynamic adaptive inspection path: Obtain the power plant's area, divide the area into multiple sub-regions, assign a unique code to each sub-region, and bind information to each sub-region; Select core indicators from multiple dimensions, determine the weights of the indicators, and construct a risk assessment indicator system; Dynamically set path planning constraints and generate dynamic adaptive inspection paths based on the risk level area division results.

6. The method for dynamic safety inspection of smart power plants based on digital twins as described in claim 5, characterized in that, In the step of calculating the risk level based on real-time data from the power plant: Based on real-time data from the power plant, extract data for each indicator and calculate the risk score for the sub-region based on the indicator data. The sub-regions are divided into multiple risk level zones based on the risk score.

7. The method for dynamic safety inspection of smart power plants based on digital twins as described in claim 6, characterized in that, After dynamically setting path planning constraints and generating a dynamic adaptive inspection path based on the risk level area division results: Obtain risk level adjustment data for the region and update the inspection route based on the adjustment data.

8. The method for dynamic safety inspection of smart power plants based on digital twins as described in claim 5, characterized in that, In the steps of conducting sub-area inspections according to the inspection route, collecting current inspection data, dynamically adjusting the risk level of the sub-area, and issuing anomaly warnings based on the risk level: Perform sub-area inspection operations, collect current inspection data, and generate sub-area inspection reports; Compare the current inspection data with the real-time power plant data acquired by the sensors to verify the current inspection data; Based on inspection data, the core indicators in the risk assessment indicator system are updated, and the risk levels of sub-regions are dynamically adjusted.

9. The method for dynamic safety inspection of smart power plants based on digital twins as described in claim 8, characterized in that, After updating the core indicators in the risk assessment indicator system based on inspection data and dynamically adjusting the risk levels of sub-regions: Based on the adjusted risk levels of the sub-regions, a multi-level early warning mechanism is set up and a response is initiated.

10. A dynamic safety inspection system for smart power plants based on digital twins, applied to the dynamic safety inspection method for smart power plants based on digital twins as described in claim 1, characterized in that, This includes a real-time operational status dynamic perception module, a risk-adaptive inspection path planning module, and an inspection anomaly early warning module; among which: The real-time operating status dynamic sensing module is used to establish a digital twin model of the power plant, and deploy acquisition sensors to obtain real-time data of the power plant, perform real-time mapping, and dynamically sense the operating status of the power plant. The risk adaptive inspection path planning module is used to divide the power plant area into multiple sub-regions, determine the risk level of the sub-regions, and generate dynamic adaptive inspection paths. The inspection anomaly early warning module is used to conduct sub-area inspections according to the inspection path, collect current inspection data, dynamically adjust the risk level of the sub-area, and issue anomaly early warnings based on the risk level.

Images