Power plant personnel position dynamic monitoring and early warning method and system based on BIM

Abstract

The invention discloses a BIM-based power plant personnel position dynamic monitoring and early warning method and system, and relates to the related field of power safety monitoring and alarming, and the method comprises the steps: carrying out the BIM twinborn modeling based on regional distribution data and equipment attribute data, carrying out the risk region labeling of a power plant BIM twinborn body, and obtaining a power plant three-dimensional risk map; matching and binding the power plant personnel area permission list and the power plant three-dimensional risk map, and establishing a power plant personnel-permission area set; monitoring personnel position change data by adopting a multi-mode fusion positioning engine, and performing behavior trajectory prediction to determine a movement prediction trajectory; and carrying out risk behavior verification and identification on the mobile prediction trajectory by combining a zero-trust personnel monitoring mechanism and a power plant personnel-permission region set, and carrying out dynamic grading early warning based on risk behavior characteristics. The technical problem that early warning accuracy and timeliness are insufficient in existing power plant personnel position dynamic monitoring is solved, and the technical effect of improving monitoring accuracy and risk early warning timeliness is achieved.

Description

Technical Field

[0001] This application relates to the field of power safety monitoring and alarm, and in particular to a BIM-based method and system for dynamic monitoring and early warning of personnel location in power plants. Background Technology

[0002] Power plant environments are complex, with numerous and widely distributed pieces of equipment. Different areas present different safety risks, and there are also clear requirements for personnel's work permissions and activity ranges. If personnel violate regulations by entering risk areas or engage in abnormal behavior, it could very likely lead to a safety accident and affect the stable operation of the power plant.

[0003] Currently, the main approach for monitoring and providing risk warnings for personnel locations in power plants is based on a combination of single-location technology and simple area access control settings. This involves deploying positioning devices throughout the power plant to acquire personnel location information, and then, based on pre-defined area access rules, determining whether personnel have entered unauthorized areas and issuing warnings. However, this single-location technology-based method is prone to signal interference in complex environments, leading to inaccurate positioning. Furthermore, relying solely on simple area access control settings for warnings lacks in-depth analysis and prediction of personnel behavior trajectories, making it difficult to accurately identify potential and concealed risk behaviors, such as personnel lingering on the edge of risk areas or moving along specific abnormal paths. This results in insufficient accuracy and timeliness of warnings, easily leading to false alarms and missed alarms. Therefore, current technologies for dynamic monitoring of personnel locations in power plants suffer from insufficient accuracy and timeliness in providing warnings. Summary of the Invention

[0004] This application provides a BIM-based method and system for dynamic monitoring and early warning of personnel location in power plants. It employs a BIM twin constructed from power plant area and equipment data, annotates risks to obtain a 3D risk map, acquires a personnel area permission list, matches it with the 3D risk map to establish a personnel-permission area set, constructs a multi-modal fusion positioning engine to monitor personnel location, predicts behavioral trajectories, and verifies trajectories using a zero-trust mechanism and the personnel-permission area set. Through these technical means, it identifies risk characteristics and provides tiered early warnings, thus solving the technical problems of insufficient accuracy and timeliness in existing dynamic monitoring of personnel location in power plants. This achieves the technical effect of improving monitoring accuracy and the timeliness of risk early warnings.

[0005] This application provides a BIM-based method for dynamic monitoring and early warning of personnel location in power plants, comprising: performing BIM twin modeling based on the regional distribution data and equipment attribute data of the target power plant to generate a power plant BIM twin; marking risk areas on the power plant BIM twin to obtain a three-dimensional risk map of the power plant; obtaining a power plant personnel area permission list; matching and binding the power plant personnel area permission list with the power plant three-dimensional risk map to establish a power plant personnel-permission area set; constructing a multi-mode fusion positioning engine; using the multi-mode fusion positioning engine to monitor the target power plant personnel location change data; predicting behavioral trajectories based on the target power plant personnel location change data to determine the target movement prediction trajectory; combining a zero-trust personnel monitoring mechanism and the power plant personnel-permission area set to verify and identify risk behaviors on the target movement prediction trajectory to obtain power plant personnel risk behavior characteristics; and performing dynamic hierarchical early warning based on the power plant personnel risk behavior characteristics.

[0006] In a possible implementation, generating a power plant BIM twin involves the following processes: cleaning and standardizing the regional distribution data and equipment attribute data of the target power plant to obtain standard regional distribution data and standard equipment attribute data; performing 3D modeling based on the standard regional distribution data and standard equipment attribute data to obtain a set of 3D models of the power plant area and power plant equipment models; mapping the set of power plant equipment models to the 3D model of the power plant area for spatial topological association to construct a power plant spatial topology model; collecting historical operation datasets of power plant equipment, and performing BIM twin modeling on the power plant spatial topology model based on the historical operation datasets of power plant equipment to generate a power plant BIM twin.

[0007] In a possible implementation, the step of performing BIM twin modeling on the power plant spatial topology model based on the historical operation dataset of the power plant equipment to generate a power plant BIM twin involves the following processes: identifying the process flow of the power plant spatial topology model, constructing the power plant operation process logic, and determining the power plant twin operation task set based on the power plant operation process logic; sequentially associating each operation task in the power plant twin operation task set with the historical operation dataset of the power plant equipment to obtain a power plant operation task association dataset; performing power plant operation simulations based on the power plant operation task association dataset to obtain a power plant operation task virtual model set; and coupling the power plant operation task virtual model set to the power plant spatial topology model for BIM twin modeling to generate the power plant BIM twin.

[0008] In a possible implementation, obtaining the power plant's 3D risk map involves the following processes: collecting real-time operating data of power plant equipment; synchronously updating the power plant's BIM twin based on the real-time operating data to generate an updated power plant BIM twin; establishing a power plant risk assessment system, which includes equipment-level risk assessment rules and regional-level risk assessment rules; using the power plant risk assessment system to classify the updated power plant BIM twin into risk assessment categories to obtain a set of power plant regional risk levels; and performing risk semantic association and regional boundary labeling on the updated power plant BIM twin according to the set of power plant regional risk levels to obtain the power plant's 3D risk map.

[0009] In a possible implementation, the establishment of the power plant personnel-permission area set involves the following processes: designing an area coding system, the coding content of which includes the spatial location and risk level of the area; coding and marking the three-dimensional risk map of the power plant according to the area coding system to obtain a power plant risk coded area set; sequentially matching and filtering the permissions of each person in the power plant personnel area permission list with the power plant risk coded area set to obtain an accessible coded area set; and binding the power plant personnel area permission list and the accessible coded area set to establish the power plant personnel-permission area set.

[0010] In a possible implementation, the following processing is performed: the multi-mode fusion positioning engine includes: outdoor sub-meter level positioning using BeiDou RTK differential technology, indoor centimeter level positioning integrating UWB tags, and Bluetooth transition area positioning.

[0011] In a possible implementation, the multi-mode fusion positioning engine is used to monitor the location change data of personnel in the target power plant, and the following processing is performed: the multi-mode fusion positioning engine is used to monitor the RTK positioning change data, UWB positioning change data, and Bluetooth positioning change data of personnel in the target power plant; the RTK positioning change data, UWB positioning change data, and Bluetooth positioning change data are time-synchronized and coordinate-normalized to obtain usable RTK positioning change data, usable UWB positioning change data, and usable Bluetooth positioning change data; multi-mode fusion positioning is performed on the usable RTK positioning change data, usable UWB positioning change data, and usable Bluetooth positioning change data based on a Kalman filter, and the target power plant personnel location change data is output.

[0012] In a possible implementation, determining the target movement prediction trajectory involves the following steps: accessing the work order of the target power plant personnel; determining the basic movement prediction trajectory based on the work order; constructing an LSTM prediction network structure based on the basic movement prediction trajectory; arranging the target power plant personnel's position change data according to temporal information to obtain the power plant personnel's position temporal change data; and using the LSTM prediction network structure to perform sample partitioning training and trajectory prediction inference on the power plant personnel's position temporal change data to determine the target movement prediction trajectory.

[0013] In a possible implementation, the process of obtaining the risk behavior characteristics of power plant personnel involves the following steps: combining the zero-trust personnel monitoring mechanism and the power plant personnel-permission area set to determine multi-factor behavior authentication rules; performing behavior verification on the target movement prediction trajectory based on the multi-factor behavior authentication rules to obtain the multi-factor verification results of power plant personnel; and identifying risk behaviors based on the multi-factor verification results of power plant personnel to obtain the risk behavior characteristics of power plant personnel.

[0014] In a possible implementation, determining the multi-factor behavioral authentication rules involves the following steps: determining a multi-factor authentication method based on the zero-trust personnel monitoring mechanism, wherein the multi-factor authentication method includes employee badge recognition, facial recognition, and location matching; collecting and analyzing normal behavior data and features of the power plant personnel-permission area set according to the multi-factor authentication method to construct a power plant personnel behavior feature baseline library; and performing authentication logic parsing based on the power plant personnel behavior feature baseline library to determine the multi-factor behavioral authentication rules.

[0015] This application also provides a BIM-based dynamic monitoring and early warning system for power plant personnel locations, including: a BIM twin modeling module, used to perform BIM twin modeling based on the regional distribution data and equipment attribute data of the target power plant, generate a power plant BIM twin, and mark risk areas on the power plant BIM twin to obtain a three-dimensional risk map of the power plant; a permission risk matching and binding module, used to obtain a power plant personnel area permission list, match and bind the power plant personnel area permission list with the power plant three-dimensional risk map, and establish a power plant personnel-permission area set; a behavior trajectory prediction module, used to construct a multi-mode fusion positioning engine, use the multi-mode fusion positioning engine to monitor the target power plant personnel location change data, perform behavior trajectory prediction based on the target power plant personnel location change data, and determine the target movement prediction trajectory; and a risk behavior verification and identification module, used to combine a zero-trust personnel monitoring mechanism and the power plant personnel-permission area set to verify and identify risk behaviors on the target movement prediction trajectory, obtain power plant personnel risk behavior characteristics, and perform dynamic hierarchical early warning based on the power plant personnel risk behavior characteristics.

[0016] The proposed BIM-based dynamic monitoring and early warning method and system for power plant personnel locations involves several steps. First, BIM twin modeling is performed based on the target power plant's regional distribution data and equipment attribute data to generate a power plant BIM twin. Risk areas are then labeled on this BIM twin to obtain a 3D risk map of the power plant. Next, a list of personnel area permissions is obtained, and this list is matched and bound to the 3D risk map to establish a personnel-permission area set. Then, a multi-mode fusion positioning engine is constructed to monitor changes in the target power plant personnel's location data. Based on this data, behavioral trajectory prediction is performed to determine the predicted movement trajectory. Finally, a zero-trust personnel monitoring mechanism and the personnel-permission area set are combined to verify and identify risk behaviors on the predicted movement trajectory, resulting in risk behavior characteristics of the power plant personnel. Dynamic, tiered early warnings are then issued based on these characteristics. This approach effectively improves the accuracy of monitoring and the timeliness of risk warnings. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0018] Figure 1 A flowchart illustrating the BIM-based dynamic monitoring and early warning method for personnel location in power plants, provided in an embodiment of this application.

[0019] Figure 2 This is a schematic diagram of the structure of a BIM-based dynamic monitoring and early warning system for personnel location in a power plant, provided in an embodiment of this application. Reference numerals: BIM twin modeling module 10, permission and risk matching and binding module 20, behavior trajectory prediction module 30, risk behavior verification and identification module 40. Detailed Implementation

[0020] The above description is merely an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below.

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application will be provided in conjunction with the accompanying drawings. The described embodiments should not be considered as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] In the following description, references to "some embodiments" describe a subset of all possible embodiments; however, it is understood that "some embodiments" may be the same or different subsets of all possible embodiments and may be combined with each other without conflict. The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices. 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 application belongs. The terminology used herein is for the purpose of describing embodiments of this application only.

[0023] This application provides a BIM-based method for dynamic monitoring and early warning of personnel location in power plants, such as... Figure 1 As shown, the method includes: Step S100: Based on the regional distribution data and equipment attribute data of the target power plant, perform BIM twin modeling to generate a power plant BIM twin, and mark the risk areas of the power plant BIM twin to obtain a three-dimensional risk map of the power plant.

[0024] Specifically, three-dimensional spatial data of the power plant is acquired using LiDAR scanning or oblique photogrammetry, and structural information such as equipment outlines and pipeline routes is extracted by combining CAD drawings. Equipment operating parameters are collected through IoT sensors, such as temperature and pressure sensors, as equipment attribute data. Point cloud filtering algorithms are used to eliminate scanning noise, and ICP algorithms are used for multi-view point cloud registration. A three-dimensional model of the target power plant is constructed using BIM software, and equipment attribute data is embedded into model components in IFC format to generate a power plant BIM twin. This power plant BIM twin is a digital mirror image of the physical power plant, containing geometry, equipment attributes, and real-time operating data. Risk area types are defined according to power safety work regulations, such as high-voltage hazard areas and flammable and explosive areas, and the boundary coordinates of these areas are calculated using GIS spatial analysis tools. Semantic annotation technology is used to store risk levels and protection requirements in JSON format in the model metadata, and risk area annotations are overlaid on the BIM model to form an interactive safety visualization platform, namely a three-dimensional risk map of the power plant.

[0025] In one possible implementation, the step S100 of generating the power plant BIM twin further includes step S110, which involves data cleaning and standardization of the regional distribution data and equipment attribute data of the target power plant to obtain standard regional distribution data and standard equipment attribute data. Specifically, the regional distribution data comes from CAD drawings, LiDAR point clouds, GIS maps, and site plans, and may have issues such as inconsistent coordinate systems, geometric errors, semantic gaps, and heterogeneous multi-source data. An example of data cleaning and standardization is as follows: For cases where CAD drawings use a local coordinate system and LiDAR uses the WGS84 geographic coordinate system, a seven-parameter transformation method is used to convert the local coordinate system to WGS84. The seven parameters include translation (X / Y / Z), rotation (Rx / Ry / Rz), and scaling (S). For noise in the LiDAR point cloud, statistical outlier filtering is used for noise removal, for example, using a method based on a neighborhood point threshold to delete isolated points with less than 5 points. For registration errors, the ICP algorithm is used to fine-tune the point cloud positions for registration optimization. To address inconsistencies in naming the same area across different drawings, a region naming mapping table was constructed to standardize the naming format. NLP was used to extract keywords and match them with standard terminology. After data cleaning and standardization, standard region distribution data was obtained.

[0026] Equipment attribute data originates from equipment ledgers, SCADA systems, inspection records, purchase contracts, etc., and suffers from issues such as missing attributes, inconsistent units, inconsistent coding, and time-series data breakpoints. Examples of data cleaning and standardization are as follows: For the issue of missing key attributes in equipment ledgers, a rule engine is used to define required fields, automatically marking missing records. Web scraping is then used to complete the missing attributes from documents such as purchase contracts and manuals. For the issue of inconsistent units for data such as temperature and pressure, a unit conversion table is constructed, conversion rules are defined, and a Python script is used to automatically traverse data columns, convert according to the rules, and label the original units. For the issue of non-standard equipment number coding, coding rules are defined according to the IEC 61360 standard, and regular expressions are used to unify the coding format. For breakpoints in SCADA data, short-term breakpoints are filled with the average of preceding and following data; long-term breakpoints are filled with a weighted average of K nearest neighbor samples selected from historical data of similar operating conditions. After data cleaning and standardization, standard equipment attribute data is obtained.

[0027] Step S120: Based on the standard regional distribution data and standard equipment attribute data, perform 3D modeling to obtain a 3D model of the power plant area and a set of power plant equipment models. Specifically, firstly, perform 3D modeling of the power plant area, using tools such as "walls," "floors," and "roofs" in BIM software to construct the main building structure, and creating irregular structures such as turbine halls and boiler rooms through the "mass" function. Add IFC attributes to model components to support subsequent spatial queries. Use LOD technology to simplify the model, displaying low-precision geometry from a distance, for example, using cuboids to replace complex valves. Then, perform 3D modeling of the power plant equipment, creating parametric templates for equipment such as turbines and generators in CAD software, and generating different model types by modifying dimensional parameters. Store the equipment models in the product data management system according to voltage level to support quick retrieval.

[0028] Step S130: Map the power plant equipment model set to the 3D model of the power plant area for spatial topological association, and construct a power plant spatial topology model. Specifically, the ICP algorithm is used to unify the coordinate system of the equipment model with the coordinate system of the area model. The Delaunay triangulation algorithm is used to calculate the spatial adjacency between equipment, such as whether the horizontal distance between the steam turbine and the feedwater pump is less than 5 meters. Boolean operations are used to determine whether the equipment is located in a specific area, such as whether the control cabinet is within the polygon of the main control room. A pipeline network topology diagram is constructed based on graph theory algorithms, and node types such as valves and elbows are marked. The equipment models are imported into the area model in IFC format, and dynamic association is achieved through the "Link Model" function of the BIM software.

[0029] Step S140: Collect historical operation datasets of power plant equipment. Based on these datasets, perform BIM twin modeling on the power plant's spatial topology model to generate a power plant BIM twin. Specifically, obtain equipment operation data, such as turbine vibration values ​​and transformer oil temperatures, from the SCADA system, import equipment defect records from the inspection system, and associate them with corresponding model components. Use Python scripts to write historical data into model metadata, for example, binding turbine vibration values ​​to the real-time vibration attributes of rotor components. Use MongoDB to store time-series data to support queries by time range. Subscribe to SCADA data via the MQTT protocol to update model attributes periodically. Use Git to manage model change history, recording the time, operator, and modified content of each update. Simulate equipment failures to verify the consistency of behavior between the twin and the physical entity.

[0030] In one possible implementation, the step of performing BIM twin modeling on the power plant spatial topology model based on the historical operation dataset of the power plant equipment to generate a power plant BIM twin, further includes step S141, which involves identifying the process flow in the power plant spatial topology model, constructing the power plant operation process logic, and determining the power plant twin operation task set based on the power plant operation process logic. Specifically, the operation of a power plant includes multiple process flows, such as fuel delivery, combustion, steam generation, and power generation in the power generation process. Based on the power plant spatial topology model, these process flows are identified, and the roles and relationships of each piece of equipment in the process flow are determined. For example, a flowchart is used to identify the connection sequence and operation logic of equipment such as fuel delivery pipelines, burners, boilers, steam turbines, and generators in the BIM model. By analyzing the power plant's design drawings, operation manuals, and historical operation records, a complete power plant operation process logic is constructed. Based on this process logic, a power plant twin operation task set is determined, such as fuel delivery tasks, combustion control tasks, steam regulation tasks, and power generation monitoring tasks, with each task corresponding to the operation and running status of a series of equipment.

[0031] Step S142 involves sequentially associating each operational task in the power plant twin operation task set with the historical operation dataset of the power plant equipment to obtain a power plant operation task association dataset. Specifically, for each power plant twin operation task, data related to that task is extracted from the historical operation dataset of the power plant equipment. For example, for the fuel delivery task, historical data such as fuel flow rate, pressure, and temperature are extracted; for the combustion control task, data such as burner ignition time, combustion intensity, and flue gas composition are extracted. Through a data association algorithm, this historical data is bound to the corresponding operational task to form a power plant operation task association dataset. This dataset not only contains the operating parameters of the equipment but also reflects the changing patterns and mutual influences of these parameters in different operational tasks, providing a data foundation for operation simulation.

[0032] Step S143: Power plant operation simulations are performed based on the associated dataset of power plant operation tasks to obtain a virtual model set for power plant operation tasks. Specifically, computer simulation technology is used to simulate each operation task based on data in the associated dataset. For example, computational fluid dynamics software is used to simulate fuel flow in pipelines, and simulation parameters are set based on historical data such as fuel flow rate and pressure to predict the effectiveness of fuel delivery and potential problems under different operating conditions. For combustion control tasks, a thermodynamic model is used in conjunction with historical burner operating data to simulate heat release and flue gas generation during combustion, and combustion efficiency and environmental performance are evaluated. In this way, a virtual model is generated for each operation task, and these virtual models together constitute the virtual model set for power plant operation tasks.

[0033] Step S144: Couple the power plant operation task virtual model set to the power plant spatial topology model for BIM twin modeling to generate the power plant BIM twin. Specifically, each virtual model in the power plant operation task virtual model set is coupled with the power plant spatial topology model. For example, the fuel transportation virtual model is combined with the spatial location and topological relationship of fuel pipelines, and the combustion control virtual model is matched with the spatial layout of equipment such as burners and boilers. Data interaction and dynamic updates between the virtual models and the spatial topology model are performed through data interfaces and algorithms. The power plant BIM twin generated in this way not only includes the physical spatial structure and equipment attributes of the power plant, but also integrates the virtual simulation results of operation tasks, which can reflect the operating status and changing trends of the power plant in real time, providing comprehensive and accurate information support for the intelligent management and decision-making of the power plant.

[0034] In one possible implementation, obtaining the three-dimensional risk map of the power plant, step S100 further includes step S150: collecting real-time operating data of the power plant equipment, and synchronously updating the power plant BIM twin based on the real-time operating data of the power plant equipment to generate an updated power plant BIM twin. Specifically, operating parameters of the equipment are acquired in real time through sensors deployed on various key equipment in the power plant, such as temperature sensors, pressure sensors, and vibration sensors. These sensors transmit data to the data acquisition system via Internet of Things (IoT) technology. For example, real-time vibration data of the steam turbine and oil temperature data of the transformer are synchronized to the power plant BIM twin in real time via the MQTT protocol or other data transmission protocols. The BIM software receives this real-time data through a preset data interface and updates the attribute values ​​of the corresponding equipment in the model. For example, the real-time vibration value of the steam turbine is bound to the vibration attribute of the steam turbine rotor component in the model to dynamically update the power plant BIM twin and generate an updated power plant BIM twin.

[0035] Step S160: Establish a power plant risk assessment system, which includes equipment-level risk assessment rules and regional-level risk assessment rules. Specifically, the power plant risk assessment system includes assessment rules at two levels: equipment-level risk assessment rules and regional-level risk assessment rules. Equipment-level risk assessment rules are used to assess the safety risks of the equipment itself, such as high-voltage equipment, flammable and explosive equipment, etc. These equipment risk assessment rules are based on the equipment's operating parameters and safety standards. For example, if the equipment's operating parameters exceed the safe range or the equipment is under maintenance, these devices are marked as high-risk. Regional-level risk assessment rules are used to assess the overall risk of equipment within a region. For example, if a region contains multiple high-risk devices or the equipment within a region is operating unstably, that region is marked as a high-risk region. These assessment rules are developed through an expert system and stored in a risk assessment database.

[0036] Step S170: The power plant risk assessment system is used to classify the power plant BIM update twin for risk assessment, resulting in a set of risk levels for different areas within the power plant. Specifically, the established power plant risk assessment system is used to assess the risk of each device and area within the power plant BIM update twin. For device-level risk assessment, the risk value of each device is calculated based on its real-time operating data and preset assessment rules. For example, if the transformer oil temperature exceeds a safety threshold, the device is marked as high-risk. For area-level risk assessment, the risk value of all devices within the area is considered, along with environmental factors, to calculate the risk level of each area. Finally, a set of risk levels for different areas within the power plant is obtained, containing risk level information for each area, such as "High-voltage area: High risk" and "Main control room: Low risk."

[0037] Step S180: Risk semantic association and area boundary labeling are performed on the power plant BIM update twin according to the power plant area risk level set to obtain the power plant's 3D risk map. Specifically, the risk level of each area is first stored as a semantic tag in the BIM model's metadata. For example, the "high risk" tag is associated with the high-voltage area's model metadata, and the "low risk" tag is associated with the main control room's model metadata. Then, using GIS spatial analysis tools, area boundaries are labeled in the BIM model based on the boundary coordinates of the risk areas. For example, the boundary of the high-voltage area is drawn in the BIM model using a polygon drawing tool, making it visible on the visualization platform. In this way, risk level and area boundary information are displayed on the power plant's BIM update twin, forming a power plant 3D risk map. This map not only shows the power plant's 3D structure and equipment distribution but also displays the risk level of each area through colors, labels, etc., providing a visualization tool for dynamic monitoring and early warning.

[0038] Step S200: Obtain the power plant personnel area permission list, match and bind the power plant personnel area permission list with the power plant three-dimensional risk map, and establish a power plant personnel-permission area set.

[0039] Specifically, power plant operation involves various personnel with different responsibilities, such as operators, maintenance personnel, and managers, each with specific area access permissions. These permissions are pre-set based on the personnel's job responsibilities and safety requirements. For example, operators are allowed access to the main control room and routine operation areas, but not to the high-voltage equipment area; maintenance personnel can enter the equipment maintenance area, but can only enter high-risk areas under specific times or conditions. The power plant personnel area access permission list is stored in the power plant's personnel management system, containing personnel identity information and area access permission information. Personnel identity information includes name, employee number, and position; area access permission information includes the areas that personnel can enter, as well as the conditions for entry, such as whether special permission or accompaniment is required.

[0040] The system extracts the power plant personnel area permission list from the personnel management system and integrates it with the area information in the power plant's 3D risk map, mapping each person's permission information to the corresponding area on the 3D risk map. For example, if an operator's permission list includes "control room," the control room area on the power plant's 3D risk map is associated with that operator's permission. When personnel permissions change, such as job changes or temporary task adjustments, the permission list is updated promptly and synchronized to the power plant's 3D risk map to ensure the accuracy and real-time nature of permission information. Through this matching and binding process, a power plant personnel-permission area set is established. This set is a database or data structure that stores each person's permission information and their corresponding area information. By monitoring personnel locations in real time and comparing them with this set, it is possible to determine whether personnel have entered unauthorized areas, thereby issuing timely warning signals to ensure personnel safety.

[0041] In one possible implementation, the establishment of the power plant personnel-authority area set, step S200 further includes step S210, designing an area coding system. The coding content of the area coding system includes the spatial location and risk level of the area. Specifically, the area coding system includes not only the spatial location information of the area but also the risk level information of the area. The area coding system can adopt a composite coding method, where the first few digits or letters represent the spatial location of the area, such as floor, room number, equipment area, etc., and the last few digits represent the risk level, such as 1 for low risk, 2 for medium risk, and 3 for high risk. For example, the code "1A-3" can represent area A on floor 1, and this area is a high-risk area.

[0042] Step S220: Encode and identify the three-dimensional risk map of the power plant according to the regional coding system to obtain a set of power plant risk coding regions. Specifically, according to the designed regional coding system, each region in the three-dimensional risk map of the power plant is coded and identified, and the spatial location and risk level information of each region are converted into corresponding codes, which are then stored in the metadata of the three-dimensional risk map of the power plant. In this way, each region within the power plant has a unique code identifier, forming a set of power plant risk coding regions.

[0043] Step S230 involves sequentially matching and filtering the permissions of each person in the power plant personnel area permission list with the power plant risk-coded area set to obtain an accessible coded area set. Specifically, for each person in the power plant personnel area permission list, their permission information is matched against the power plant risk-coded area set to check whether the area identifier in each person's permission list matches the code in the power plant risk-coded area set. For example, if an operator's permission list contains "Control Room (Low Risk)," a matching code is searched in the power plant risk-coded area set. In this way, the area codes that each person can actually access are filtered out, forming an accessible coded area set. This set identifies the areas that each person is allowed to enter, providing a specific area range for personnel location monitoring and risk warning.

[0044] Step S240 involves binding the power plant personnel area permission list and the set of accessible coded areas to establish the power plant personnel-permission area set. Specifically, each person in the power plant personnel area permission list is bound to their set of accessible coded areas to form a complete power plant personnel-permission area set. When personnel enter an unauthorized area, the system can quickly identify and issue an early warning, effectively ensuring personnel safety and the safe operation of the power plant.

[0045] Step S300: Construct a multi-mode fusion positioning engine, use the multi-mode fusion positioning engine to monitor the location change data of personnel in the target power plant, perform behavioral trajectory prediction based on the location change data of personnel in the target power plant, and determine the target movement prediction trajectory. The multi-mode fusion positioning engine includes: outdoor sub-meter level positioning using Beidou RTK differential technology, indoor centimeter level positioning integrating UWB tags, and Bluetooth transition area positioning.

[0046] Specifically, to achieve precise monitoring of personnel locations within the power plant, a multi-mode fusion positioning engine was constructed. This engine combines multiple positioning technologies to adapt to different environments and scenarios, ensuring accuracy and continuity of positioning. Specifically, outdoor positioning utilizes BeiDou RTK differential technology, a high-precision satellite positioning technology that achieves sub-meter (less than 1 meter) positioning accuracy through differential calculations between a base station and a rover station. Multiple BeiDou RTK base stations are deployed in the outdoor area of ​​the power plant. These base stations transmit differential data to rover stations (personnel-carried positioning devices) via wireless communication. After receiving the differential data, the rover station combines it with BeiDou satellite signals to calculate its own position coordinates, achieving sub-meter positioning accuracy.

[0047] Indoor positioning utilizes UWB (Ultra-Wideband) technology, a high-precision indoor positioning technology capable of achieving centimeter-level accuracy. UWB determines a person's location by transmitting and receiving ultra-wideband radio frequency signals and calculating the signal's time of flight. Multiple UWB base stations are installed in indoor areas of the power plant, such as the main control room and equipment rooms. UWB tags worn by personnel communicate with these base stations, and by calculating the signal's time of flight, the personnel's coordinates are determined, achieving centimeter-level positioning accuracy.

[0048] Bluetooth technology, a short-range wireless communication technology, is used for positioning in transitional areas, such as the boundary between indoor and outdoor spaces. Bluetooth beacons are deployed in these transitional areas, such as doorways and corridors. Bluetooth devices carried by personnel receive signals from these beacons, and the approximate location is calculated using a signal strength algorithm, thus enabling positioning within the transitional area.

[0049] The multi-mode fusion positioning engine collects real-time location data of personnel, including sub-meter accuracy outdoors, centimeter accuracy indoors, and Bluetooth positioning data in transitional areas. Positioning devices transmit this location data to a backend server via wireless communication. The backend server receives data from different positioning technologies and performs data fusion. For example, when a person moves from indoors to outdoors, the system automatically switches from UWB positioning to BeiDou RTK differential positioning to ensure continuity and accuracy. Based on the collected data on changes in personnel location, machine learning algorithms or data analysis models are used to predict the person's behavioral trajectory. For example, by analyzing a person's historical walking paths and speeds, their future direction and location of movement can be predicted.

[0050] In one possible implementation, the step S300, which involves using the multi-mode fusion positioning engine to monitor the location change data of personnel at the target power plant, further includes step S310, which involves using the multi-mode fusion positioning engine to monitor the RTK location change data, UWB location change data, and Bluetooth location change data of the personnel at the target power plant. Specifically, the multi-mode fusion positioning engine monitors the location change data of power plant personnel in real time, including location coordinates and timestamps, through the integration of multiple positioning technologies.

[0051] Step S320 involves performing time synchronization and coordinate normalization on the RTK positioning change data, UWB positioning change data, and Bluetooth positioning change data to obtain usable RTK positioning change data, usable UWB positioning change data, and usable Bluetooth positioning change data. Specifically, by analyzing the timestamp of each positioning data, data from different sources are aligned to a unified time base. For example, if the timestamp of the RTK data is UTC time, while the timestamps of the UWB and Bluetooth data are local time, the timestamps of the UWB and Bluetooth data need to be converted to UTC time to ensure that the time base of all data is consistent. Since the RTK, UWB, and Bluetooth positioning systems may use different coordinate systems, these coordinate systems need to be unified into a standard coordinate system. For example, the geographic coordinates (latitude and longitude) of the RTK are converted to a unified coordinate system within the power plant, such as the power plant's planar coordinate system, and the local coordinates of UWB and Bluetooth, i.e., coordinates relative to the base station or beacon, are converted to absolute coordinates in the same coordinate system. After time synchronization and coordinate normalization, usable RTK positioning change data, usable UWB positioning change data, and usable Bluetooth positioning change data are obtained.

[0052] Step S330: Multimodal fusion positioning is performed on the available RTK positioning change data, available UWB positioning change data, and available Bluetooth positioning change data based on a Kalman filter, outputting the target power plant personnel location change data. Specifically, a Kalman filter is a recursive data filter that can estimate the system's state variables based on the system's dynamic model and observation data. In the positioning scenario, the Kalman filter can combine observation data from different positioning technologies to estimate the precise location of personnel. Specifically, the state variables and covariance matrix of the Kalman filter are initialized based on the initial values ​​of the available RTK, UWB, and Bluetooth positioning data. The personnel's position at the next moment is predicted based on the personnel's motion model, such as uniform motion or acceleration motion. The predicted position is compared with the new available RTK, UWB, and Bluetooth positioning data, and the predicted position is corrected using the Kalman filter's update formula to obtain the accurate personnel position. Finally, the fused personnel location change data is output, including precise location coordinates and a timestamp. Multimodal fusion positioning using a Kalman filter can effectively reduce the error of a single positioning technology and improve positioning accuracy and reliability.

[0053] In one possible implementation, step S300, determining the target movement prediction trajectory, further includes step S340: accessing the work orders of the target power plant personnel, and determining the basic movement prediction trajectory based on the work orders. Specifically, accessing the work orders of the target power plant personnel records information such as the personnel's work tasks, work locations, and work processes within a specific time period. Based on these work orders, the personnel's basic movement prediction trajectory is determined, i.e., the initial path and time schedule determined based on the work schedule and work process information.

[0054] Step S350: Based on the basic motion prediction trajectory, construct an LSTM prediction network structure. Specifically, LSTM is a deep learning model suitable for processing sequential data and capable of capturing the temporal dependencies and dynamic changes in the data. The location and time data from the basic motion prediction trajectory are used as input, and the data is preprocessed, including normalization and serialization operations, to make it suitable for input into the LSTM model. The LSTM prediction network structure is constructed, including an input layer, multiple LSTM layers, and an output layer. The LSTM layers can capture the time-series features in the location data, and the output layer is used to generate the predicted trajectory points. The LSTM model is trained using historical personnel location data, which includes the actual movement trajectories of personnel and their corresponding timestamps. Through training, the model learns the patterns and rules of personnel movement.

[0055] Step S360: Arrange the target power plant personnel location change data according to time sequence information to obtain power plant personnel location time sequence change data. Specifically, sort the data points in the target power plant personnel location change data according to time sequence to form a complete power plant personnel location time sequence change data sequence. For example, the timestamp of data point 1 is 8:00:00, the timestamp of data point 2 is 8:00:05, the timestamp of data point 3 is 8:00:10, and so on. Format the time sequence change data into a format suitable for input into the LSTM model, for example, combine the location coordinates and timestamps into a serialized data structure.

[0056] Step S370: An LSTM prediction network structure is used to perform sample partitioning, training, and trajectory prediction inference on the time-series data of the power plant personnel's location changes, thereby determining the target's predicted movement trajectory. Specifically, the constructed LSTM prediction network structure is used to process the time-series data of the power plant personnel's location changes. First, the time-series data is divided into a training set and a test set. The training set is used to train the LSTM model, and the test set is used to evaluate the model's predictive performance. Through training, the LSTM model learns the patterns and rules of personnel movement. Then, the trained LSTM model is used to perform prediction inference on the test set, predicting the future movement trajectory points of the personnel based on their historical location change data. Finally, based on the prediction results of the LSTM model, the predicted movement trajectory of the target power plant personnel is determined.

[0057] Step S400: Combine the zero-trust personnel monitoring mechanism and the power plant personnel-authority area set to verify and identify the risk behavior of the target movement prediction trajectory, obtain the risk behavior characteristics of the power plant personnel, and perform dynamic hierarchical early warning based on the risk behavior characteristics of the power plant personnel.

[0058] Specifically, the zero-trust personnel monitoring mechanism refers to the fact that in the scenario of monitoring the location of personnel in a power plant, the security of every movement and behavior of personnel is not assumed, but needs to be continuously verified to ensure that only authorized behaviors that comply with security policies are allowed.

[0059] The predicted movement trajectories of personnel are compared with the power plant's personnel-authorized area set to check whether the planned entry area is within their authorized scope. For example, if the predicted trajectory shows that an operator is about to enter a high-voltage equipment area, but according to the authorized area set, the operator does not have access to this area, this is identified as a potential risky behavior. The behavior of personnel is further evaluated by combining the area risk levels in the power plant's 3D risk map. If personnel enter a high-risk area, even if they have the relevant permissions, more stringent verification is required based on factors such as the purpose of their actions and the duration of their stay. For example, if maintenance personnel enter a high-risk flammable and explosive area to repair equipment, it needs to be verified whether they are carrying the necessary safety protection equipment and whether the maintenance operation complies with safety regulations. Analyzing the predicted movement trajectories of personnel identifies abnormal behavioral patterns, such as personnel frequently entering and exiting different areas in a short period of time, or staying in certain areas for much longer than the normal working time; these may be signs of risky behavior. Through machine learning algorithms and other means, a model of normal behavioral patterns is established by analyzing and learning from a large amount of historical personnel behavior data, thereby promptly detecting abnormal behaviors that deviate from normal patterns.

[0060] Through the above verification and identification process, the characteristics of risky behaviors of power plant personnel were summarized, including unauthorized entry into specific areas, abnormal duration of stay in high-risk areas, and movement trajectories inconsistent with equipment operation. Based on the real-time behavior trajectory of personnel and changes in the environment, the warning status is adjusted in real time. That is, the warning system does not statically issue alarms, but reacts accordingly to the development and changes in personnel behavior. For example, when personnel first show signs of entering an unauthorized area, the warning system initially issues a low-level warning; if the personnel continue to move in that direction and approach the boundary of the area, the warning level gradually increases; if the personnel promptly adjust their route and leave the danger zone, the warning level decreases or is even lifted.

[0061] Based on the severity of the risky behavior, early warning information is divided into different levels. For example, early warnings are divided into three levels: low-risk warning, medium-risk warning, and high-risk warning. A low-risk warning simply reminds personnel to pay attention to safety or informs them that they are about to exceed their authorized scope; a medium-risk warning requires personnel to immediately stop their current behavior or report to management personnel; a high-risk warning will immediately activate the emergency plan and take mandatory measures, such as issuing an emergency evacuation signal through the positioning device carried by the personnel, while notifying on-site staff and safety management personnel to intervene.

[0062] In one possible implementation, obtaining the risk behavior characteristics of power plant personnel, step S400 further includes step S410, which combines the zero-trust personnel monitoring mechanism and the power plant personnel-authority area set to determine multi-factor behavior authentication rules. Specifically, the multi-factor behavior authentication rules are formulated based on the power plant personnel-authority area set and various data collected by the zero-trust monitoring mechanism. These rules include multiple aspects. Specifically, the first is authorization verification, which checks whether personnel have the authority to enter the target area. Authorization verification is based on the personnel's identity information and their corresponding authorization scope in the power plant personnel-authority area set. For example, only authorized maintenance personnel can enter the equipment maintenance area. The second is area risk level verification, which combines the area risk level in the power plant's three-dimensional risk map and determines whether additional verification measures are needed based on the area risk level entered by the personnel. For example, entering high-risk areas requires stricter verification, such as additional safety training certificates, special permits, etc. There is also time verification, which verifies whether the time when personnel enter the area complies with work arrangements and safety regulations. For example, some areas are prohibited from entering during specific time periods, or the personnel's behavior trajectory is inconsistent with the work schedule, which also requires verification. For example, if equipment maintenance is being carried out in a certain area at night, entry by unauthorized personnel during that time would be considered abnormal behavior. There's also behavioral trajectory verification, which analyzes personnel's historical behavioral trajectories and current predicted movement trajectories to determine if their behavior conforms to normal work processes and patterns. If personnel's behavioral trajectories are abnormal, such as frequently entering multiple unrelated areas or staying in a certain area for an excessively long time, this may indicate risky behavior. Finally, there's equipment operation verification; if personnel enter equipment operation areas, it needs to be verified whether they have the qualifications and authorization to operate the corresponding equipment.

[0063] Step S420: Verify the target movement prediction trajectory based on the multi-factor behavior authentication rules to obtain the power plant personnel multi-factor verification result. Specifically, according to the multi-factor behavior authentication rules determined in step S410, verify the target movement prediction trajectory to ensure that the personnel's behavior meets safety requirements. The specific verification process is as follows: Compare the personnel's movement prediction trajectory with the power plant personnel-authorized area set to check if the area the personnel plan to enter is within their authorized scope. If the personnel's prediction trajectory shows that they will enter an unauthorized area, the verification result is a failure. Based on the risk level of the area the personnel are entering, and in conjunction with the multi-factor behavior authentication rules, verify whether the personnel meet the additional conditions for entering the area. For example, for high-risk areas, check whether the personnel are carrying the necessary safety protection equipment and have completed relevant safety training. Compare the personnel's movement prediction trajectory with the work schedule and safety regulations to verify whether their entry time into the area meets the requirements. If the personnel's behavior trajectory shows that they enter an area outside of working hours or during prohibited periods, the verification result is a failure. Analyze the personnel's historical behavior trajectory and current movement prediction trajectory to determine whether their behavior conforms to normal work processes and patterns. If personnel exhibit abnormal behavior, such as frequently entering multiple unrelated areas or spending excessive time in a particular area, this may indicate risky behavior, and the verification result will be "failed." If personnel enter the equipment operation area, their qualifications and authorization to operate the corresponding equipment will be verified. This is done by checking personnel training records, operating certificates, and other information to ensure they can safely operate the equipment. If personnel lack the appropriate operating qualifications, the verification result will be "failed." After the above multi-factor behavioral verification, a multi-factor verification result for power plant personnel is generated. The result can be either "pass" or "fail." "Pass" indicates that the personnel's behavior complies with all certification rules, while "fail" indicates that the personnel's behavior does not comply with the certification rules.

[0064] Step S430 involves identifying risk behaviors based on the multi-factor verification results of the power plant personnel to obtain their risk behavior characteristics. Specifically, risk-characteristic behavioral patterns are extracted from the verification results. For verification failures, the specific reasons for the failures are analyzed. Examples include personnel entering unauthorized areas, failing to carry necessary safety equipment in high-risk areas, or entering specific areas outside of working hours. The personnel's historical behavioral data and current verification results are combined to analyze whether their behavioral patterns are abnormal. For example, repeated verification failures or frequent abnormal behavioral trajectories indicate risky behavior. Based on the reasons for verification failures and the degree of abnormality in the behavioral patterns, the personnel's behavior is assessed for risk levels, which can be categorized into low, medium, and high risk. For example, personnel entering unauthorized areas without serious consequences can be assessed as medium risk; personnel performing dangerous operations in high-risk areas without taking any safety measures can be assessed as high risk. Through the risk behavior identification process, the risk behavior characteristics of power plant personnel are finally obtained, including personnel entering unauthorized areas, failing to take safety measures in high-risk areas, abnormal behavior patterns, and multiple verification failures.

[0065] In one possible implementation, the step S410 of determining the multi-factor behavioral authentication rules further includes step S411, which determines a multi-factor authentication method based on the zero-trust personnel monitoring mechanism. The multi-factor authentication method includes employee badge recognition, facial recognition, and location matching. Specifically, multiple authentication methods are used to comprehensively verify the identity and behavior of personnel to improve the accuracy and reliability of authentication and reduce the vulnerabilities and risks that may exist with a single authentication method. Specifically, the employee badge recognition system reads information such as electronic tags or barcodes on the employee badge and compares it with employee badge information in the personnel management system to confirm the legitimacy of the personnel's identity. Facial recognition technology uses a camera to capture facial images of personnel and verifies their identity by analyzing and comparing the images. Compared with employee badge recognition, facial recognition can prevent personnel from using other people's employee badges or forging identities. Facial recognition is also used for real-time monitoring of personnel behavior, such as detecting whether personnel are wearing safety helmets. Location matching obtains the real-time location information of personnel through positioning technology and compares it with the personnel's authorized area. If a person's location does not match the authorized area or their behavior does not conform to expectations, then it can be determined that they have engaged in risky behavior.

[0066] Step S412: Normal behavior data and feature analysis are performed on the power plant personnel-authorized area set according to the multi-factor authentication method to construct a baseline database of power plant personnel behavior characteristics. Specifically, behavioral data of power plant personnel under normal working conditions is collected. The collection process is as follows: When personnel enter the power plant area, their identity information is collected through work badge recognition and facial recognition systems, and compared with the information in the personnel management system to ensure the legality and validity of the personnel's identity. Positioning technology is used to collect personnel movement trajectories in real time, including walking routes, stopping locations, and stopping times within the power plant. When personnel operate equipment, operation records are collected through the equipment management system, including operation time, equipment type, and operation steps. Simultaneously, relevant data about the personnel's environment, such as area risk level and equipment operating status, are collected.

[0067] Feature analysis is performed on the collected normal behavior data to extract key characteristics of human behavior, including personal identification, behavioral trajectory, dwell time, and equipment operation. For example, by analyzing a person's behavioral trajectory, features such as the distribution of dwell time in different areas and walking speed can be extracted; by analyzing a person's equipment operation records, features such as operation frequency and proficiency can be extracted.

[0068] The data after feature analysis is stored in the power plant personnel behavior characteristic baseline library. This baseline library contains various behavioral characteristics of power plant personnel under normal working conditions, which are used as a reference for subsequent behavior verification and risk identification.

[0069] Step S413: Based on the power plant personnel behavior characteristic baseline database, perform authentication logic parsing to determine the multi-factor behavior authentication rules. Specifically, based on the data in the power plant personnel behavior characteristic baseline database, perform logic parsing on the multi-factor authentication method to determine the multi-factor behavior authentication rules. The specific parsing process is as follows: Determine the rules for work badge recognition, such as the validity verification of the work badge and the matching rules between the work badge and the personnel identity. For example, the work badge must be valid, and the information on the work badge must be consistent with the information in the personnel management system for the work badge recognition to be considered successful. Determine the rules for face recognition, such as the matching degree of facial features and the accuracy of recognition. For example, the matching degree of facial features must reach a certain threshold, and the recognition process must be completed within a specified time for the face recognition to be considered successful. Determine the rules for location matching, such as whether the personnel's location is within the authorized area and whether the behavioral trajectory meets expectations. For example, the personnel's location must be within the authorized area, and the similarity between the behavioral trajectory and the normal behavioral trajectory in the baseline database must reach a certain threshold for the location matching to be considered successful.

[0070] Based on the logic of the various authentication methods described above, multi-factor behavior authentication rules are established. For example, all three authentication methods—employee badge recognition, facial recognition, and location matching—must pass to be considered valid and legal for an employee's behavior. Alternatively, certain weights can be assigned, such as 30% for employee badge recognition, 40% for facial recognition, and 30% for location matching. Only when the overall score reaches a certain threshold is the employee's behavior considered valid and legal. This approach ensures that employee behavior complies with the power plant's safety requirements while improving the accuracy and reliability of authentication.

[0071] This application's embodiments employ a BIM twin constructed based on power plant area and equipment data, annotating risks to obtain a 3D risk map, acquiring a personnel area permission list, matching it with the 3D risk map to establish a personnel-permission area set, constructing a multi-mode fusion positioning engine to monitor personnel locations, predicting behavioral trajectories, and combining a zero-trust mechanism and the personnel-permission area set to verify trajectories, identify risk characteristics, and provide tiered early warnings. These technical means solve the technical problems of insufficient early warning accuracy and timeliness in existing dynamic monitoring of power plant personnel locations, achieving the technical effect of improving monitoring accuracy and the timeliness of risk early warnings.

[0072] In the above text, refer to Figure 1 This paper describes in detail a BIM-based dynamic monitoring and early warning method for personnel location in power plants according to embodiments of the present invention. Next, we will refer to... Figure 2 This invention describes a BIM-based dynamic monitoring and early warning system for the location of personnel in a power plant.

[0073] The BIM-based dynamic monitoring and early warning system for power plant personnel locations according to embodiments of the present invention addresses the technical problems of insufficient accuracy and timeliness in existing dynamic monitoring of power plant personnel locations, thereby improving the accuracy of monitoring and the timeliness of risk warnings. The BIM-based dynamic monitoring and early warning system for power plant personnel locations includes: a BIM twin modeling module 10, a permission and risk matching and binding module 20, a behavior trajectory prediction module 30, and a risk behavior verification and identification module 40.

[0074] The BIM twin modeling module 10 is used to perform BIM twin modeling based on the regional distribution data and equipment attribute data of the target power plant, generate a power plant BIM twin, and mark risk areas on the power plant BIM twin to obtain a three-dimensional risk map of the power plant; the permission risk matching and binding module 20 is used to obtain the power plant personnel area permission list, match and bind the power plant personnel area permission list with the power plant three-dimensional risk map, and establish a power plant personnel-permission area set; the behavior trajectory prediction module 30 is used to construct a multi-mode fusion positioning engine, use the multi-mode fusion positioning engine to monitor the target power plant personnel location change data, and perform behavior trajectory prediction based on the target power plant personnel location change data to determine the target movement prediction trajectory; the risk behavior verification and identification module 40 is used to combine the zero-trust personnel monitoring mechanism and the power plant personnel-permission area set to verify and identify the risk behavior of the target movement prediction trajectory, obtain the power plant personnel risk behavior characteristics, and perform dynamic hierarchical early warning based on the power plant personnel risk behavior characteristics.

[0075] The detailed description of the specific configuration of the BIM twin modeling module 10 is explained as follows: As mentioned above, to generate a power plant BIM twin, the BIM twin modeling module 10 may further include: a data preprocessing unit for cleaning and standardizing the regional distribution data and equipment attribute data of the target power plant to obtain standard regional distribution data and standard equipment attribute data; a 3D modeling unit for performing 3D modeling based on the standard regional distribution data and standard equipment attribute data to obtain a set of 3D models of the power plant area and power plant equipment models; a spatial topology association unit for mapping the set of power plant equipment models to the 3D model of the power plant area to perform spatial topology association and construct a power plant spatial topology model; and a BIM twin modeling unit for collecting historical operation datasets of power plant equipment and performing BIM twin modeling on the power plant spatial topology model based on the historical operation datasets of power plant equipment to generate a power plant BIM twin.

[0076] The step of performing BIM twin modeling on the power plant spatial topology model based on the historical operation dataset of the power plant equipment to generate a power plant BIM twin can be further divided into: a power plant operation process logic construction subunit for identifying the process flow of the power plant spatial topology model, constructing the power plant operation process logic, and determining the power plant twin operation task set according to the power plant operation process logic; an association subunit for sequentially associating each operation task in the power plant twin operation task set with the historical operation dataset of the power plant equipment to obtain a power plant operation task association dataset; a power plant operation simulation subunit for performing power plant operation simulation based on the power plant operation task association dataset to obtain a power plant operation task virtual model set; and a coupling subunit for coupling the power plant operation task virtual model set to the power plant spatial topology model for BIM twin modeling to generate the power plant BIM twin.

[0077] The BIM twin modeling module 10, which obtains the three-dimensional risk map of the power plant, may further include: a synchronization update unit for collecting real-time operating data of power plant equipment and updating the power plant BIM twin based on the real-time operating data to generate an updated power plant BIM twin; a power plant risk assessment system construction unit for building a power plant risk assessment system, which includes equipment-level risk assessment rules and regional-level risk assessment rules; a risk assessment division unit for using the power plant risk assessment system to perform risk assessment division on the updated power plant BIM twin to obtain a set of power plant regional risk levels; and a regional boundary labeling unit for performing risk semantic association and regional boundary labeling on the updated power plant BIM twin according to the set of power plant regional risk levels to obtain the three-dimensional risk map of the power plant.

[0078] The detailed description of the specific configuration of the permission risk matching and binding module 20 is explained as follows: As mentioned above, to establish a power plant personnel-permission area set, the permission risk matching and binding module 20 may further include: an area coding system design unit for designing an area coding system, wherein the coding content of the area coding system includes the area spatial location and the area risk level; a coding identification unit for coding and identifying the power plant three-dimensional risk map according to the area coding system to obtain a power plant risk coding area set; a matching and filtering unit for sequentially matching and filtering the permissions of each person in the power plant personnel area permission list with the power plant risk coding area set to obtain an accessible coding area set; and a permission binding unit for binding the power plant personnel area permission list and the accessible coding area set to establish the power plant personnel-permission area set.

[0079] The detailed description of the specific configuration of the behavior trajectory prediction module 30 is explained as follows: As mentioned above, the behavior trajectory prediction module 30 may further include: the multi-mode fusion positioning engine includes: outdoor sub-meter level positioning using Beidou RTK differential technology, indoor centimeter level positioning integrating UWB tags, and Bluetooth transition area positioning.

[0080] The behavior trajectory prediction module 30, which uses the multi-mode fusion positioning engine to monitor the location change data of personnel in the target power plant, may further include: a location change data monitoring unit for monitoring RTK location change data, UWB location change data, and Bluetooth location change data of personnel in the target power plant using the multi-mode fusion positioning engine; a time synchronization unit for synchronizing the time and normalizing the coordinates of the RTK location change data, UWB location change data, and Bluetooth location change data to obtain available RTK location change data, available UWB location change data, and available Bluetooth location change data; and a multi-mode fusion positioning unit for performing multi-mode fusion positioning on the available RTK location change data, available UWB location change data, and available Bluetooth location change data based on a Kalman filter, and outputting the location change data of personnel in the target power plant.

[0081] The behavior trajectory prediction module 30, which determines the target movement prediction trajectory, may further include: a basic movement prediction trajectory determination unit for calling the work order of the target power plant personnel and determining the basic movement prediction trajectory based on the work order; an LSTM prediction network structure construction unit for constructing an LSTM prediction network structure based on the basic movement prediction trajectory; a power plant personnel location time-series change data acquisition unit for arranging the target power plant personnel location change data according to time-series information to obtain power plant personnel location time-series change data; and a trajectory prediction inference unit for using the LSTM prediction network structure to perform sample partitioning training and trajectory prediction inference on the power plant personnel location time-series change data to determine the target movement prediction trajectory.

[0082] The detailed description of the specific configuration of the risk behavior verification and identification module 40 is explained as follows: As mentioned above, to obtain the risk behavior characteristics of power plant personnel, the risk behavior verification and identification module 40 may further include: a multi-factor behavior authentication rule determination unit used to combine the zero-trust personnel monitoring mechanism and the power plant personnel-authority area set to determine multi-factor behavior authentication rules; a behavior verification unit used to perform behavior verification on the target movement prediction trajectory based on the multi-factor behavior authentication rules to obtain the multi-factor verification results of power plant personnel; and a risk behavior identification unit used to identify risk behaviors on the multi-factor verification results of power plant personnel to obtain the risk behavior characteristics of power plant personnel.

[0083] The multi-factor behavior authentication rule determination unit may further include: a multi-factor authentication method determination subunit, used to determine multi-factor authentication methods based on the zero-trust personnel monitoring mechanism, wherein the multi-factor authentication methods include employee badge recognition, facial recognition, and location matching; a power plant personnel behavior feature baseline database construction subunit, used to collect and analyze normal behavior data and features of the power plant personnel-permission area set according to the multi-factor authentication methods, and construct a power plant personnel behavior feature baseline database; and an authentication logic parsing subunit, used to perform authentication logic parsing based on the power plant personnel behavior feature baseline database, and determine the multi-factor behavior authentication rules.

[0084] The BIM-based dynamic monitoring and early warning system for power plant personnel location provided in this embodiment of the invention can execute the BIM-based dynamic monitoring and early warning method for power plant personnel location provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0085] Although this application makes various references to certain modules in the system according to the embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy distinction between each other and are not used to limit the scope of protection of this invention. The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application. In some cases, the actions or steps described in this application can be performed in a different order than that shown in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A BIM-based dynamic monitoring and early warning method for personnel location in power plants, characterized in that, The method includes: Based on the regional distribution data and equipment attribute data of the target power plant, BIM twin modeling is performed to generate a power plant BIM twin. Risk areas are marked on the power plant BIM twin to obtain a three-dimensional risk map of the power plant. Obtain the power plant personnel area permission list, match and bind the power plant personnel area permission list with the power plant 3D risk map, and establish a power plant personnel-permission area set; A multi-mode fusion positioning engine is constructed, and the multi-mode fusion positioning engine is used to monitor the location change data of personnel in the target power plant. Based on the location change data of personnel in the target power plant, behavioral trajectory prediction is performed to determine the target's predicted movement trajectory. By combining the zero-trust personnel monitoring mechanism and the power plant personnel-authority area set, the risk behavior of the target movement prediction trajectory is verified and identified to obtain the risk behavior characteristics of power plant personnel, and dynamic hierarchical early warning is carried out based on the risk behavior characteristics of power plant personnel.

2. The BIM-based dynamic monitoring and early warning method for personnel location in power plants as described in claim 1, characterized in that, The generation of the power plant BIM twin includes: Data cleaning and standardization are performed on the regional distribution data and equipment attribute data of the target power plant to obtain standard regional distribution data and standard equipment attribute data. Based on the standard regional distribution data and standard equipment attribute data, three-dimensional modeling is performed to obtain a set of three-dimensional models of the power plant area and power plant equipment models. The power plant equipment model set is mapped to the three-dimensional model of the power plant area to perform spatial topological association, thereby constructing a power plant spatial topology model; Collect historical operation datasets of power plant equipment, and perform BIM twin modeling on the spatial topology model of the power plant based on the historical operation datasets of the power plant equipment to generate a power plant BIM twin.

3. The BIM-based dynamic monitoring and early warning method for personnel location in power plants as described in claim 2, characterized in that, The step of performing BIM twin modeling on the power plant spatial topology model based on the historical operation dataset of the power plant equipment to generate a power plant BIM twin includes: The power plant spatial topology model is labeled with process flow, the power plant operation process logic is constructed, and the power plant twin operation task set is determined based on the power plant operation process logic. Each operation task in the power plant twin operation task set is sequentially associated with the power plant equipment historical operation dataset to obtain the power plant operation task association dataset. Power plant operation simulations were performed based on the power plant operation task association dataset to obtain a virtual model set of power plant operation tasks. The virtual model set of power plant operation tasks is coupled to the power plant spatial topology model for BIM twin modeling to generate the power plant BIM twin.

4. The BIM-based dynamic monitoring and early warning method for personnel location in power plants as described in claim 1, characterized in that, The obtained three-dimensional risk map of the power plant includes: Collect real-time operating data of power plant equipment, and synchronously update the power plant BIM twin based on the real-time operating data of the power plant equipment to generate a power plant BIM update twin; Establish a power plant risk assessment system, which includes equipment-level risk assessment rules and regional-level risk assessment rules; The power plant risk assessment system is used to classify the power plant BIM update twin for risk assessment, resulting in a set of power plant area risk levels. Based on the power plant area risk level set, risk semantic association and area boundary labeling are performed on the power plant BIM update twin to obtain the power plant three-dimensional risk map.

5. The BIM-based dynamic monitoring and early warning method for personnel location in power plants as described in claim 1, characterized in that, The establishment of the power plant personnel-permission area set includes: Design a regional coding system, wherein the coding content of the regional coding system includes regional spatial location and regional risk level; The power plant's three-dimensional risk map is coded and identified according to the regional coding system to obtain a set of power plant risk coded regions. The permissions of each person in the power plant personnel area permission list are sequentially matched and filtered with the power plant risk code area set to obtain the accessible code area set. Bind the power plant personnel area permission list and the accessible coded area set to establish the power plant personnel-permission area set.

6. The BIM-based dynamic monitoring and early warning method for personnel location in power plants as described in claim 1, characterized in that, The multi-mode fusion positioning engine includes: outdoor sub-meter level positioning using BeiDou RTK differential technology, indoor centimeter level positioning integrating UWB tags, and Bluetooth transition area positioning.

7. The BIM-based dynamic monitoring and early warning method for personnel location in power plants as described in claim 1, characterized in that, The method of using the multi-mode fusion positioning engine to monitor changes in the location of personnel at the target power plant includes: The multi-mode fusion positioning engine is used to monitor the RTK positioning change data, UWB positioning change data, and Bluetooth positioning change data of personnel in the target power plant. Time synchronization and coordinate normalization are performed on the RTK positioning change data, UWB positioning change data and Bluetooth positioning change data to obtain available RTK positioning change data, available UWB positioning change data and available Bluetooth positioning change data; Based on the Kalman filter, multimodal fusion positioning is performed on the available RTK positioning change data, available UWB positioning change data, and available Bluetooth positioning change data to output the target power plant personnel location change data.

8. The BIM-based dynamic monitoring and early warning method for personnel location in power plants as described in claim 1, characterized in that, The determination of the target movement prediction trajectory includes: Call the work order of the personnel at the target power plant, and determine the basic movement prediction trajectory based on the work order; Based on the aforementioned basic mobile prediction trajectory, an LSTM prediction network structure is constructed. Arrange the target power plant personnel location change data according to time sequence information to obtain power plant personnel location time sequence change data; An LSTM prediction network structure is used to perform sample segmentation training and trajectory prediction inference on the temporal change data of the power plant personnel's location to determine the target's predicted trajectory.

9. The BIM-based dynamic monitoring and early warning method for personnel location in power plants as described in claim 1, characterized in that, The obtained risk behavior characteristics of power plant personnel include: By combining the zero-trust personnel monitoring mechanism and the aforementioned power plant personnel-authority area set, multi-factor behavioral authentication rules are determined. Based on the multi-factor behavioral authentication rules, the predicted trajectory of the target movement is verified to obtain the multi-factor verification results of the power plant personnel. Risk behavior identification is performed on the multi-factor verification results of the power plant personnel to obtain the risk behavior characteristics of the power plant personnel.

10. The BIM-based dynamic monitoring and early warning method for personnel location in power plants as described in claim 9, characterized in that, The determination of multi-factor behavioral authentication rules includes: Based on the zero-trust personnel monitoring mechanism, a multi-factor authentication method is determined, which includes employee badge recognition, facial recognition, and location matching. According to the multi-factor authentication method, normal behavior data of the power plant personnel-permission area set is collected and feature analysis is performed to construct a baseline database of power plant personnel behavior characteristics. The authentication logic is parsed based on the baseline database of power plant personnel behavior characteristics to determine the multi-factor behavior authentication rules.

11. A BIM-based dynamic monitoring and early warning system for personnel location in power plants, characterized in that, The system is used to implement the BIM-based dynamic monitoring and early warning method for personnel location in power plants according to any one of claims 1-10, and the system includes: The BIM twin modeling module is used to perform BIM twin modeling based on the regional distribution data and equipment attribute data of the target power plant, generate a power plant BIM twin, and mark the risk areas of the power plant BIM twin to obtain a three-dimensional risk map of the power plant. The permission risk matching and binding module is used to obtain the power plant personnel area permission list, match and bind the power plant personnel area permission list with the power plant 3D risk map, and establish a power plant personnel-permission area set; The behavior trajectory prediction module is used to construct a multi-mode fusion positioning engine, which monitors the location change data of personnel in the target power plant, and performs behavior trajectory prediction based on the location change data of personnel in the target power plant to determine the target's predicted movement trajectory. The risk behavior verification and identification module is used to combine the zero-trust personnel monitoring mechanism and the power plant personnel-authority area set to verify and identify the risk behavior of the target's predicted movement trajectory, obtain the risk behavior characteristics of the power plant personnel, and perform dynamic hierarchical early warning based on the risk behavior characteristics of the power plant personnel.

Images