Real estate system virtual-real mapping inspection method, device and equipment based on digital twinning and medium

Abstract

The invention provides a property system virtual-real mapping inspection method, device and equipment based on digital twinning and a medium, and belongs to the technical field of property inspection. Real-time mapping of an equipment entity and a virtual model is achieved through layered design of a physical layer, a data layer, a twinning layer and an application layer; collecting static / dynamic data and reducing noise, and establishing a global coordinate system; constructing a 1: 1 parameterized model and binding equipment attributes; the state analysis model is used for evaluating the equipment health degree, and virtual-real identification and work order visualization are triggered when abnormity occurs; the inspection path is dynamically optimized, and the processing efficiency is improved in combination with AR assistance and dual-stage verification; and finally, updating the model weight and adjusting the inspection strategy through data-driven acceptance feedback. Real-time synchronization and intelligent decision making of the equipment state are realized, the inspection efficiency is improved, the fault omission ratio is reduced, the resource allocation is optimized, and the operation and maintenance transparency and reliability are enhanced.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of property inspection, and particularly relates to a virtual-real mapping inspection method, device, equipment and medium for a property system based on digital twinning. BACKGROUND

[0002] The property service industry is an important part of urban infrastructure operation, covering residential communities, commercial complexes, office buildings, industrial parks and other formats. Intelligent property has become the mainstream direction of industry development, aiming to improve service quality, improve energy efficiency and enhance user experience through digital means.

[0003] In the intelligent property architecture of the related art, physical devices, data collection, virtual display, and functional applications are mixed, and the hierarchical division is not clear. In the intelligent property architecture of the related art, the virtual model of the property is mostly to display the appearance of the device, and lacks attribute association between devices, and cannot query the linkage relationship with other devices; for the device analysis model, the physical properties are not configured, and the physical risks cannot be predicted through the virtual model, resulting in that the virtual model does not match the actual operation and maintenance, and cannot meet the use requirements.

[0004] The device exception records, repair work orders, and historical data of the architecture are stored in different Excel tables, the work orders are in the OA system, and the historical data are in the database. The inspection personnel need to query and integrate information one by one, which is time-consuming and easy to miss key content, cannot provide technical support for inspection, and the inspection personnel only carry paper manuals and cannot obtain device historical data in real time, need to rely on experience for judgment, and cannot meet the inspection requirements. SUMMARY

[0005] The application provides a virtual-real mapping inspection method for a property system based on digital twinning, which realizes real-time synchronization of device state and intelligent decision-making, improves inspection efficiency, reduces fault omission rate, optimizes resource allocation, and enhances operation and maintenance transparency and reliability.

[0006] The method comprises: S101: constructing a layered framework and a lightweight virtual model of a digital twin of a property system; S102: collecting static data and dynamic data of the property system, and performing noise reduction processing on the collected dynamic data; establishing a global coordinate system with a building reference point as the origin, and converting the position information of each device into coordinate data and storing; S103: based on the static data and the dynamic data, constructing a parameterized model according to a 1:1 ratio for key devices using a building BIM model, exporting the parameterized model into a preset format; assigning identification information to each device, binding the device with the corresponding node in the digital twin model, and associating the bound node with static attributes, dynamic attributes and logical attributes; configuring rigid physical properties and state visualization rules for the model; S104: Obtain running data of the equipment; input the running data into a preset equipment state analysis model; S105: The equipment state score output by the equipment state analysis model; if the equipment state score exceeds a preset threshold or a device abnormality work order submitted by a user is received, an abnormal state is identified for a corresponding equipment node in the parameterized model; and repair work order information related to the equipment abnormality is superimposed on a label of the corresponding equipment node in the digital twin model to realize visual display of the abnormal state; S106: Based on the abnormal equipment state and the coordinate position information, an optimal inspection path is generated and pushed to a mobile terminal; after an inspection personnel arrives at the scene, virtual-real synchronization is triggered by scanning a code to obtain historical data of the equipment and a perspective view of hidden engineering, and state confirmation and processing are performed; S107: After processing, the stability of the equipment parameters is verified, the weight of the neural network model is updated, and subsequent inspection strategies and monitoring frequencies are adjusted.

[0007] The application also provides a virtual-real mapping inspection device for a property system based on a digital twin, which comprises: a twin framework modeling module, configured to build a hierarchical framework of a digital twin of the property system and a lightweight virtual model; a coordinate system binding module, configured to collect static data and dynamic data of the property system, perform noise reduction processing on the collected dynamic data, establish a global coordinate system with a building reference point as an origin, and convert position information of each equipment into coordinate data and store the coordinate data; a scene-equipment binding module, configured to build a parameterized model according to a 1:1 ratio for key equipment based on the static data and the dynamic data, export the parameterized model into a preset format, assign identification information to each equipment, bind the equipment with a corresponding node in the digital twin model, and associate the bound node with static attributes, dynamic attributes and logical attributes; and configure rigid physical attributes and state visualization rules for the model; a running data input module, configured to obtain running data of the equipment; and input the running data into a preset equipment state analysis model; an abnormality visualization module, configured to output an equipment state score by the equipment state analysis model; if the equipment state score exceeds a preset threshold or a device abnormality work order submitted by a user is received, an abnormal state is identified for a corresponding equipment node in the parameterized model; and repair work order information related to the equipment abnormality is superimposed on a label of the corresponding equipment node in the digital twin model to realize visual display of the abnormal state; a virtual-real mapping module, configured to generate an optimal inspection path based on an abnormal equipment state and coordinate position information, and push the optimal inspection path to a mobile terminal; after an inspection personnel arrives at the scene, virtual-real synchronization is triggered by scanning a code to obtain historical data of the equipment and a perspective view of hidden engineering, and state confirmation and processing are performed; and The verification updating module is configured to verify the stability of the device parameters after the processing is completed, update the weight of the neural network model, and adjust the subsequent inspection strategy and monitoring frequency.

[0008] According to another embodiment of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the property system virtual-real mapping inspection method based on digital twinning when executing the program.

[0009] According to still another embodiment of the present application, a storage medium is also provided, having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the property system virtual-real mapping inspection method based on digital twinning.

[0010] As can be seen from the above technical solutions, the present application has the following advantages: The property system virtual-real mapping inspection method based on digital twinning provided by the present application realizes the interaction of device operation data, virtual mapping, and corresponding monitoring by decomposing the property system into a physical layer, a data layer, a twinning layer, and an application layer. The responsibilities among the layers are clear, the virtual-real interaction provides a clear path, and the overall scalability of the system is ensured. The BIM model is converted into a three-dimensional scene model, and the key device 1:1 parameterized model ensures the form consistency of the virtual model and the physical device. The static properties, dynamic properties, and logical properties of the node are associated, and the rigid body physical properties and state visualization rules are configured, so that the virtual model is a visualization carrier, and the full life cycle information of the device can be quickly queried directly through the virtual model; the mapping of the physical device and the virtual node is realized by binding the device identification and the node, ensuring that the subsequent state monitoring and abnormal identification can be accurately located to the specific device. The abnormal identification and work order information based on the state score are superimposed, so that the inspection personnel do not need to simultaneously view the fault report, the maintenance work order, and the device drawing, but can directly obtain the device abnormal level, the fault description, the required spare parts for maintenance, and the historical fault record through the twinning model, and the binding of the abnormal identification and the work order realizes the abnormal discovery and maintenance follow-up. The optimal path is generated based on the abnormal state and the coordinate position, the devices with severe abnormalities can be inspected first, and then the devices with slight abnormalities are processed, the inspection efficiency is improved, the inspection personnel can directly obtain the historical data and the hidden engineering perspective view at the physical device, and the difficulty of on-site judgment and processing is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0011] In order to more clearly illustrate the technical solutions of the present application, the drawings required in the description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating any creative labor.

[0012] Figure 1 A property system virtual-real mapping inspection method flowchart based on digital twinning is provided. Figure 2 A property system virtual-real mapping inspection device schematic diagram based on digital twinning is provided. Figure 3 An electronic device schematic diagram is provided. DETAILED DESCRIPTION

[0013] The property system virtual-real mapping inspection device based on digital twinning disclosed in the present application is related to the cooperative work between multiple component structures and multiple operation steps, and specifically relates to a digital twinning five-dimensional structure, which divides the property system digital twinning framework into a physical layer, a data layer, a twinning layer, and an application layer.

[0014] The physical layer is composed of physical entities, control modules, and sensing modules, and is an actually existing physical entity structure, including various device entities, fixed terminals, security devices, monitoring modules, sensors, software resources, etc. The sensing module is used to obtain real-time data information of various devices.

[0015] The data layer is a bridge for realizing real-time mapping and interaction between physical entities and virtual twins, and is mainly used to realize data collection, processing, and transmission. Through data processing methods such as cleaning and filtering, data is screened and optimized to ensure the reliability and accuracy of the data. In the palletizing robot workstation, the data layer collects real-time state data of the physical layer, such as monitoring information and sensor data, and transmits it to the twinning layer for processing and analysis.

[0016] The twinning layer is the core of the digital twinning system, and includes geometric models, physical models, and behavior models. It combines the characteristics of the model layer and the physical layer, and through the construction of accurate models of physical objects in a virtual environment, the digital twinning platform server controls, optimizes, and monitors these models, performs dynamic simulation and real-time feedback, and realizes data interaction and co-evolution between real space and virtual space.

[0017] The application layer mainly includes virtual simulation, virtual-real coordination, data visualization monitoring, and other functions. A virtual simulation environment is built in the digital twinning platform system, and the workstation is simulated and tested through the HMI human-machine interface and optimized. The software-in-the-loop and hardware-in-the-loop methods are used for virtual-real coordination to realize real-time mapping and data monitoring management between the virtual space and the real space of the workstation, and to simulate the state of each device in the system in real time, thereby reducing the debugging cost and cycle.

[0018] The device state analysis model in this embodiment can analyze the device data of the inspection process and further analyze the device running state.

[0019] Optionally, the device state analysis model can be implemented based on a BP neural network, the BP neural network model is a multilayer feedforward network trained by an error back propagation algorithm, is one of the most widely used neural network models at present, is particularly suitable for processing data without a clear mathematical relationship between known conditions and results, and can establish a certain relationship between conditions and results, and effectively avoid a large number of experiments. The input layer is used to manage the received and processed raw data, and the output layer is used to manage the results of the fitted model, and the connection weight between neurons can reflect the degree of connection between neurons.

[0020] The input value can be mapped from zero to infinity or infinitesimal to 0 or 1, without affecting the trend of the original input value. The network is adjusted according to the direction of the negative gradient to obtain the optimal solution.

[0021] The tansig function has an s-shaped tangent characteristic as a transfer function, can reflect the data region from (-∞, +∞) to (0, 1), and can help the model to achieve the expected function. When the output values are all positive, the tansig parameter can be selected. There are many training methods for the BP network, and when selecting the method, the characteristics of the research problem and the sample capacity should be considered comprehensively.

[0022] The property system virtual-real mapping inspection device can connect devices in the physical world through the Internet, and a large network system for information exchange and communication. Intelligent devices with data processing and communication functions begin to be connected and interacted through the Internet in batches. The Internet of Things realizes comprehensive perception of the environment and object state through various sensors and information collection devices in the property system. These sensors can monitor temperature, humidity, light, sound, position and other physical quantities, so as to realize real-time monitoring and data collection of the devices in the property system.

[0023] The application also sets up a central server and multiple participating clients. The central server is used to coordinate the data processing process of the property system inspection process, can also initialize the global model, collect model updates from each participating client, aggregate these updates to form a new global model, and distribute back to each participating client. Each participating client independently trains the model on local data, calculates the model update, and sends the update back to the central server for updating the global model. This process is repeated throughout the learning cycle until the model reaches the predetermined performance indicators or convergence conditions.

[0024] The application establishes a virtual-real mapping (Cyber-Physical Mapping) of property system inspection. The virtual-real mapping realizes intelligent decision-making through virtual control of real and virtual-real cooperation by constructing a two-way data loop of physical entities and virtual models through sensor networks, communication protocols and digital modeling. A dynamic correspondence between physical space and digital space is established: the state of the physical entity (such as temperature, position, vibration) is mapped to the virtual model in real time, and the simulation results of the virtual model guide the control and optimization of the physical entity.

[0025] The property system of the application classifies each device into the following devices for communication, and each device in the current property system is shown in Table 1.

[0026] Table 1: Devices in the property system

[0027] To build a data network for the overall linkage of the project, the point data of key equipment and facilities is completed from the dimension of the property system. For example, the air conditioning system in the project is upgraded and reconstructed, aiming to improve the sensing ability of the system and ensure the completeness of the data. On the one hand, data acquisition devices are installed at the air conditioning terminals (such as fan coil, fresh air unit, etc.), covering key operating parameters such as temperature, humidity, wind speed, and valve opening, to realize comprehensive sensing of the terminal operating state.

[0028] The virtual-real mapping inspection method of the property system based on digital twinning related to the application will be described in detail below. In order to illustrate but not to limit, specific details such as specific system structure, technology are proposed to thoroughly understand the embodiments of the application. However, it should be clear to those skilled in the art that the application can also be implemented in other embodiments without these specific details.

[0029] It should be understood that when used in the specification of the application, the term "comprising" indicates the presence of the described features, whole, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, whole, steps, operations, elements, components and / or sets thereof. The terms "comprise", "include", "have" and their variants mean "including but not limited to", unless otherwise specifically emphasized.

[0030] The phrase "one embodiment" or "some embodiments" or similar phrases as used in this application means that a specific feature, structure or characteristic described in the embodiment is included in one or more embodiments of the application. Therefore, the phrases "in one embodiment", "in some embodiments", "in other some embodiments", "in further some embodiments" appearing in different places in this application do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized.

[0031] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0032] Please refer to Figure 1 Fig. 1 shows a flowchart of a virtual-real mapping inspection method of a property system based on digital twinning in an embodiment, and the method comprises the following steps. S101: A hierarchical framework of digital twinning of the property system is constructed, including a physical layer with device entities, a data layer for realizing mapping and interaction between the device entities and virtual twins, a twinning layer for virtual modeling and simulation, virtual-real coordination, and data monitoring, and an application layer.

[0033] In some embodiments, the physical layer not only includes device entities in the property system, such as air conditioning units, water pumps, elevators, power distribution equipment, etc., but also covers temperature sensors, vibration sensors, current sensors, valves, relays, and building floor layouts, pipeline routes, machine room locations, etc. associated with the devices. All entities have physical properties and operating states that can be perceived.

[0034] The data layer is used to establish bidirectional mapping and real-time interaction between physical entities and virtual twins, specifically including formulating data transmission protocols, constructing data synchronization methods, and building data storage modules. The twinning layer constructs virtual models based on the geometric shape, physical properties, and operating rules of physical entities, not only realizing 1:1 geometric modeling, but also incorporating the virtual mirror of the heat conduction coefficient of device materials, the friction coefficient of mechanical parts, and the elevator start-stop logic, air conditioning refrigeration capacity adjustment rules. The application layer provides specific functions based on the twinning model, including simulation functions such as simulating the operating state of a water pump impeller after wear and tear and simulating energy consumption changes under different operating modes; virtual-real coordination functions support remote control of physical devices through virtual models; and data monitoring functions real-time display device operating current, temperature, pressure, energy consumption indicators, fault warning information, etc.

[0035] The embodiment also constructs a lightweight virtual model. Specifically, the construction is performed through element screening, geometric simplification, data compression, and interaction adaptation. Under the premise of combining the location, operating state, and connection relationship of property operation equipment, the geometric complexity, data volume, and computing power demand of the model are reduced. Here, the objects strongly related to equipment management, inspection, and security monitoring are first locked through element screening to avoid non-key elements occupying resources. Then, the model can dynamically adapt the precision according to different use scenarios through LOD geometric simplification and topology optimization, balancing the visual effect and resource consumption. Through data dimension reduction and compression, the static attribute redundancy and dynamic data transmission volume are reduced, ensuring that the model can be smoothly loaded on devices with limited network bandwidth and general hardware computing power; finally, through interaction logic adaptation, smooth basic support is provided for subsequent device state visualization, inspection path generation, and other functions.

[0036] S102: Collect static data and dynamic data of the property system, and perform noise reduction processing on the collected dynamic data; establish a global coordinate system with a building reference point as the origin, and convert the location information of each device into coordinate data and store it.

[0037] In some embodiments, the static data is the inherent attribute of the device, such as device model, factory number, rated parameters such as rated voltage / current / power, installation date, manufacturer, maintenance period, total number of floors, building area per floor, machine room location coordinates, pipeline diameter and material, wall thickness, etc. It can be entered from the original archives of the property.

[0038] The dynamic data of the embodiment is the real-time operating state data of the device, including current, voltage, power factor, speed, vibration frequency, bearing temperature, water flow rate, pipeline pressure, medium temperature, ambient temperature, humidity, dust concentration, etc., which are collected in real time by sensors deployed at key positions of the device. The collection frequency is set according to the importance of the parameters.

[0039] The dynamic data noise reduction processing of the embodiment is aimed at the noise mixed in the sensor collection process. The sliding average method is used to filter short-term fluctuations for current data, and wavelet transform is used to separate noise and effective signals for vibration data, retaining the true trend of the data. The global coordinate system is established with the vertex of the southeast corner of the building bottom layer as the origin, the X-axis along the east-west direction of the building, the Y-axis along the north-south direction, and the Z-axis along the vertical ground direction, forming a three-dimensional Cartesian coordinate system; the device location conversion is performed by field measurement, the measurement results are converted into three-dimensional coordinate values, and the values are stored in the coordinate database after being associated with the unique identification of the device.

[0040] S103: Based on static data and dynamic data, a parameterized model of the key equipment is constructed in a 1:1 scale using the building BIM model, the parameterized model is exported in a preset format, identification information is assigned to each equipment, the equipment is bound to a corresponding node in the digital twin model, and static attributes, dynamic attributes and logical attributes are associated with the bound node; and the model is configured with rigid body physical attributes and state visualization rules.

[0041] In some embodiments, the modeling tool selects a professional software that supports BIM import and three-dimensional modeling; when the building BIM model is converted into a three-dimensional scene model, the building structure information in the BIM model is first extracted, the redundant construction information in the BIM is removed, and lightweight processing is performed, and then imported into a rendering tool to construct a three-dimensional scene model containing the overall space of the building.

[0042] The 1:1 parameterized modeling of the key equipment in this embodiment is based on static data and dynamic data, and parameterized models of motor diameter, pump length, material density, thermal conductivity, valve opening angle range, and motor speed interval are constructed. The model can simulate different operating states by adjusting parameters.

[0043] This embodiment can preset the export format selection to support formats that support cross-platform loading and attribute association, such as GLB format that supports geometry, texture, and attribute information packaging, and FBX format that supports model animation and physical property export, which facilitates subsequent loading and calling in the digital twin platform.

[0044] The device identification information set in this embodiment uses the coding rules of device type, device location coordinates and number. The binding of the device and the virtual node is achieved through database association, and the device ID and the ID of the device node in the virtual model are stored in the association table, and the ID is matched when calling.

[0045] The static attributes associated with the bound node include fixed information such as device model, rated parameters, and installation date, the dynamic attributes include real-time current, temperature, and running state, and the logical attributes include the association relationship between the device and other devices, and the connection logic between the device and the pipeline. The rigid body physical property configuration is based on a physical engine, which sets the mass, collision volume and friction coefficient of the device model to ensure that the virtual model meets the real physical law when colliding with other models.

[0046] In some specific embodiments, step S103 specifically includes the following modes: S1031: According to the application scenario of the digital twin model, the model application level is divided, the building structure framework, device connection interface and key component size are retained, and the decorative components are converted and the key size deviation between the model and the physical entity is verified through size comparison.

[0047] Specifically, after exporting the BIM model in IFC format using Revit, processing is performed in Blender, and the model is divided into "global browsing layer", "local maintenance layer", and "component disassembly layer". The global browsing layer retains the load-bearing walls, elevator shafts, and main pipelines of the building structure. The local maintenance layer retains the device shell, interface size, and wiring terminal for maintenance. The component disassembly layer retains the assembly relationship of the device internal compressor cylinder and elevator speed limiter. After conversion, the key dimensions of the physical equipment are measured using a laser range finder.

[0048] S1032: When performing 1:1 parameterized modeling of key equipment, implant dynamic working condition adaptation components. For the running characteristics of elevators and air conditioning equipment, preset the motion constraint range of the model components, and reserve the associated interface with the equipment running parameters. Optionally, the motion constraint range is the lifting height range of the elevator car and the opening angle range of the air conditioning valve.

[0049] In this embodiment, SolidWorks is used to establish parameterized models of key equipment such as elevators and power distribution cabinets. In the air conditioning valve model, the valve opening angle corresponding to the opening degree of 0-100% is preset, and the interface for connecting with the data layer valve control signal is reserved. When the model is exported, the parameterized driving file is saved to ensure that the digital twin platform can drive the model component motion by inputting the running parameters.

[0050] S1033: Assign a unique identification information to each device, including basic identification, subsystem code, parent device code, installation location coordinates, commissioning time, and maintenance period. Establish a bidirectional mapping relationship between the identification information and the digital twin model node.

[0051] In this embodiment, the device basic identification is used as an index to associate the identification to realize the belonging association of the device, subsystem, and parent-child devices. The space-time identification realizes the spatial positioning and full life cycle tracking of the device. The bidirectional mapping realizes one-to-one storage of the identification information and the model node ID through the index table in the database. When querying, the index table is matched quickly to ensure accurate positioning.

[0052] S1034: When associating device attributes, build a cross-layer collaborative attribute pool. Bind static attributes with device specifications, connect dynamic attributes with real-time data streams of data layer edge nodes, and associate logical attributes with application layer scene function modules to realize cross-layer real-time synchronization of attribute data.

[0053] The embodiment builds a device attribute library in the digital twin platform. A static attribute module stores device material density, rated power, and is associated with a device factory specification book. A dynamic attribute module interfaces with a data layer edge node through an MQTT protocol and receives device operation parameters in real time. A logical attribute module is associated with an application layer scene rule. An attribute pool sets a data synchronization trigger. When a physical layer device parameter is updated or an application layer rule is adjusted, the attribute library data is automatically updated to ensure that a virtual model can accurately map a physical device.

[0054] S1035: When configuring the model state visualization rule, a scene-based dynamic rule library is built. A visualization unit is configured according to the property operation and maintenance scene. A color gradient is used to identify the state. An animation and a sound-light linkage prompt are superimposed in the emergency rescue scene. Energy consumption data and a heat map are associated in the energy consumption audit scene.

[0055] It should be noted that in the visualization configuration module of the digital twin platform, a rule library is built according to the daily inspection, emergency rescue, and energy consumption audit scenes. In the daily inspection scene, a device is displayed in yellow when normal and in red when faulty. In the emergency rescue scene, a faulty device is superimposed with a red flashing animation, and a sound-light prompt of the platform is linked. The fault type is labeled beside the model. In the energy consumption audit scene, device energy consumption data are associated with a heat map, and the energy consumption value per unit time is labeled. When the scene is switched, the corresponding visualization rule is loaded. This helps the inspection personnel quickly capture key information and improves the operation and maintenance response efficiency.

[0056] S104: Obtain the operation data of the device; and input the operation data into a preset device state analysis model.

[0057] In some embodiments, the operation data is obtained based on an Internet of Things sensing method, and data preprocessing is based on data quality improvement principles. Cleaning and filling processing eliminates data noise and missing, avoiding input of incorrect data to the model. Normalization processing eliminates the influence of parameter magnitude, ensuring that the weight evaluation of the model to each parameter meets the requirements. The device state analysis model is based on machine learning principles. The mapping relationship between operation data and device state is learned through historical data, such as learning that a current exceeding 4A and a temperature exceeding 60°C correspond to a motor overload state. After inputting the real-time preprocessed data, the model outputs the current state evaluation result of the device through internal calculation, such as forward propagation of a neural network and node judgment of a decision tree.

[0058] S105: The device state score output by the device state analysis model; if the device state score exceeds a preset threshold value or a device abnormality work order submitted by a user is received, an abnormal state is identified for a corresponding device node in the parameterized model; and device abnormality related maintenance work order information is superimposed on the label of the corresponding device node in the digital twin model, realizing visual display of the abnormal state.

[0059] In some embodiments, the device state score is calculated by the model based on the pre-processed operation data, the score range is set to 0-100, 0 represents the best device state, and 100 represents device failure, the score is calculated by weighted summation method: according to the importance of each operation parameter, the weights are set as follows: motor current weight 0.3, temperature weight 0.25, vibration frequency weight 0.2, rotating speed weight 0.15, and environmental temperature weight 0.1. First, the score of each parameter is calculated, for example, 0 for current within the rated range, 20 for exceeding 10% of the rated value, 50 for exceeding 30%, and 80 for exceeding 50%.

[0060] In this embodiment, the total score is obtained by weighted summation. The preset threshold is set according to the device type and industry standard, and the threshold can be dynamically adjusted according to historical failure data. The work order information needs to include device location description, abnormal phenomenon, submission time, contact person, etc. The system matches the device unique ID through the device description in the work order. The maintenance work order information is superimposed on the device node label, the label is hidden by default, and the label is displayed when the mouse hovers or clicks on the device node. The information contained in the label includes work order number, abnormal phenomenon description, emergency level, submission time, assigned maintenance personnel, and expected completion time.

[0061] The abnormal state identification of this embodiment quickly locates the abnormal device in the virtual model, avoids the maintenance personnel to find one by one in the huge building model, improves the abnormal discovery efficiency; the work order information superposition realizes the direct association of abnormality and processing task, the maintenance personnel can understand the work order details synchronously when checking the abnormality, without switching the system to query, reduces the information disconnection, and improves the maintenance response speed.

[0062] S106: Based on the abnormal device state and coordinate position information, an optimal inspection path is generated and pushed to a mobile terminal; after the inspection personnel arrive at the scene, the virtual-real synchronization is triggered by scanning the code to obtain the device historical data and hidden engineering perspective view, and the state confirmation and processing are performed.

[0063] In some embodiments, the input factors for generating the optimal inspection path include: coordinate position of the abnormal device, device abnormal priority, such as score exceeding threshold value 90 points or more for high priority, 70-90 points for medium priority, and less than 70 points for low priority; and the passing constraint between devices and the current position of the inspection personnel.

[0064] In this embodiment, the abnormal devices to be inspected are first screened, and then the route is planned based on the shortest path priority principle.

[0065] The path output includes the inspection sequence, the expected arrival time of each device, and the walking / car navigation route. The mobile terminal is a smart phone or an industrial tablet provided for the inspection personnel.

[0066] The path pushing is triggered by an APP message notification, the planned path is displayed on an APP map interface, and real-time navigation is supported. The device scanning is implemented through a two-dimensional code posted on the device. After the inspection personnel scan the two-dimensional code using the APP scanning function, the virtual-real synchronization is triggered through the device ID. The real-time state of the device node in the digital twin model is displayed on the APP end. The device historical data and the hidden engineering perspective view are called from the database.

[0067] The state confirmation and processing are performed by the inspection personnel: comparing the virtual model data with the actual state of the on-site device to confirm the abnormal reason; and processing according to the abnormal reason, during which the operation steps can be recorded on the APP end, and the on-site photos can be uploaded to the system.

[0068] In some specific embodiments, step S106 specifically includes the following manner: Step S1061: Mark the abnormal device as a visual identifier in the digital twin model through the binding relationship between the device and the digital twin model node, and superimpose the maintenance work order information; obtain the physical coordinates of the abnormal device, bind the coordinate information with the model node, and ensure the accurate correspondence between the virtual mark and the physical position.

[0069] Step S1062: Take the abnormal device as a high-priority node, construct a graph theory model G=(V, E) containing device location, physical obstacles, and passing priority, wherein the node V is a device / critical location, the edge E is a passable path and is assigned a weight, and the weight = fault severity x physical distance x obstacle coefficient; an improved Dijkstra algorithm is used to calculate the obstacle avoidance shortest path, and a real-time inspection route containing the optimal sequence and avoiding congestion areas is generated and pushed to the mobile terminal of the inspection personnel.

[0070] In this embodiment, when constructing the graph theory model, the node V includes an abnormal device node, a stairway, a fire passage, and an obstacle node; and the edge E represents the passable path between the nodes.

[0071] The weight calculation of this embodiment is: weight = fault severity x physical distance x obstacle coefficient, no obstacle = 1, and obstacle = 2. When the improved Dijkstra algorithm is used, the path with the smallest weight is preferentially expanded, and the path is updated in real time. After the path is generated, it is pushed to the mobile terminal.

[0072] Step S1063: Associate the historical operation data of the abnormal device, the hidden engineering perspective view, and the device technical manual to generate a work order information containing a recommended inspection step, and embed an AR perspective portal in the work order information to support the inspection personnel to view the internal structure of the device through the mobile terminal.

[0073] The system of the embodiment retrieves the running data of the abnormal device from the historical database through SQL query, extracts the concealed engineering data from the BIM model, and generates an intelligent work order containing charts and three-dimensional views. When the inspection personnel trigger by scanning the device two-dimensional code, the spatial feature points of the current environment are obtained by calling SLAM positioning, and the concealed engineering view in the BIM model is superimposed on the real scene picture.

[0074] Step S1064: The inspection personnel scans the device two-dimensional code using the mobile terminal, calls SLAM positioning to calculate the global coordinates of the device, real-time highlights the nodes of the twin model, and activates the AR perspective function, aligns the real scene photo with the model view angle; After the inspection personnel completes the maintenance, the two-dimensional code is scanned again to call the device state analysis model to verify the current parameters in real time. If the parameters meet the standard, the node state of the twin model is updated to green; If the parameters do not meet the standard, the red alarm is maintained and the manual acceptance process is triggered.

[0075] Specifically, when scanning the code, the mobile terminal calls SLAM positioning to scan the environment around the device, extracts feature points and matches them with the pre-stored device spatial feature library, and calculates the global coordinates of the device. The twin model finds the corresponding node through coordinate matching, calls the AR rendering engine to superimpose the model view angle on the real scene picture.

[0076] When the code is scanned again, the device state analysis model is called to calculate the current parameters in real time. If the current value decreases from 12A to 5A, the color of the model node is updated to green. If the current value is still 11A, the manual acceptance process is triggered and the detection report is required to be uploaded.

[0077] Step S1065: Continuously monitor the parameter stability of the repaired abnormal device. If the parameter fluctuation is ≤±5%, it is determined that the repair is successful, the work order is archived, and a repair report is generated; If the parameters continue to exceed the standard, the high-response manual troubleshooting process is triggered.

[0078] The effective maintenance data is encrypted and uploaded to the federal learning center node. The center node fuses the gradient parameters of each edge device through weighted average, generates a global optimization model, and distributes it to the terminal to improve the prediction accuracy of similar faults. The faulty device is added to the monitoring queue, and the recurrence probability prediction model is trained through historical data to generate the recurrence probability.

[0079] The repair result verification of the embodiment is accurate, and the secondary fault is reduced. The federal learning improves the model accuracy without sharing the original data, and protects the device data privacy.

[0080] S107: Verify the device parameter stability after processing is completed, update the neural network model weight, and adjust the subsequent inspection strategy and monitoring frequency.

[0081] In some embodiments, the device parameter stability verification is performed on the processed parameters such as motor current, temperature, pressure, and rotating speed, and the verification duration is set according to the device type, for example, 20 minutes for motor type devices and 10 minutes for water pump type devices. During the verification process, the parameter data is collected in real time, and the judgment standard is that the parameter fluctuation range is within ±5% of the rated value of the device. If the parameter continuously stabilizes within the fluctuation range within 20 minutes, it is determined to be stable. If the parameter exceeds the fluctuation range, it is determined to be unstable and needs to be reprocessed.

[0082] The input data for updating the weight of the neural network model of the embodiment is the effective data of the current maintenance, including: abnormal parameters before processing, processing measures, stable parameters after processing, and processing results.

[0083] The weight updating of the neural network model of the embodiment is based on the iterative optimization of machine learning. The effective data of the current maintenance is taken as a new training sample, the model prediction error is calculated through the back propagation algorithm after inputting the model, and the connection weight of the related neurons is adjusted, so that the recognition accuracy of the model for similar faults is improved. The patrol strategy adjustment is based on the current state of the device and the historical fault data. The system reads the fault times and processing results of the device from the database, generates a new patrol strategy through the rule engine, and updates the patrol task list. The monitoring frequency adjustment is realized by modifying the sensor acquisition parameters. The system sends acquisition frequency instructions to the corresponding sensors according to the device classification, and the sensors adjust the acquisition interval after receiving the instructions to realize the dynamic change of the monitoring frequency, ensure the maintenance quality, reduce the risk of fault recurrence, update the neural network weight, let the model continuously learn new fault data, and continuously optimize the fault recognition ability, improve the patrol pertinence, and improve the monitoring frequency. The equipment that has just been processed or has high risk is monitored, and the recurring faults are found in time.

[0084] In an embodiment of the present application, based on step S105, a possible embodiment will be given below to illustrate the specific implementation scheme. Step S105 specifically includes the following modes: S1051: Implementing identification on the abnormal state of the device, and dividing the device state score into slight abnormality, moderate abnormality, and severe abnormality levels according to the device state score. Different levels adopt different visual identification.

[0085] Optionally, the slight abnormality is a yellow frame flicker, the moderate abnormality is an orange filling and a frame flicker, and the severe abnormality is a red filling, a high-frequency flicker, and a magnification effect.

[0086] In the embodiment, the abnormality level is determined by comparing the state score with the preset level threshold, a rendering instruction of the corresponding visual identification is triggered, the appearance change of the device node in the digital twin model directly reflects the emergency degree of the abnormality, and the priority can be directly judged intuitively without checking the specific score.

[0087] S1052: Verify the abnormal state, combine the score output by the device state analysis model, call the last 3 historical failure records of the device, the same type of device running data and related environmental parameters at the same period, if the data are all abnormal, confirm the abnormality is valid, otherwise mark it as a suspected abnormality and trigger secondary detection; The embodiment uses historical failures, same type of device benchmarks, and environmental influences to construct an abnormality verification matrix, and judges the authenticity of the abnormality through the consistency of the data. Based on the verification, the false positive rate is reduced, and the secondary detection of suspected abnormalities can capture potential failures in time, improving the accuracy of abnormality judgment.

[0088] S1053: Superimpose maintenance work order information in the digital twin model node label, including work order number, fault description, urgency, expected processing time, required spare parts list, historical maintenance records, and device maintenance manual excerpts, and realize information on-demand display through the hierarchical folding function in the label. In the embodiment, the device node label displays the work order number, fault description, and urgency. It can also display the expected processing time and required spare parts.

[0089] Optionally, it can also display the maintenance records in a period of time and the processing steps for the corresponding fault in the maintenance manual.

[0090] S1054: When the device state changes from abnormal to normal, enters the maintenance state, or completes the maintenance but is not stable, update the identification style and synchronize the state conversion timestamp.

[0091] The embodiment presets the device state conversion rule, and the normal state corresponds to a green static model; the maintenance state corresponds to blue; the unstable state corresponds to red; when the state is converted, the conversion time is recorded and stored in the state log. By monitoring the device state score change and maintenance process node in real time, the switching instruction of the identification style is triggered, so that the identification state of the virtual model is synchronized with the actual processing progress of the physical device, realizing full-process visual tracking.

[0092] S1055: Realize cross-scene linkage display of abnormal state, in the emergency rescue scene, the identification of the serious abnormal device is highlighted, and the state of its associated devices is also displayed; in the energy consumption audit scene, the identification of the abnormal device is associated with the display of the percentage of its energy consumption deviation from the benchmark value and the degree of influence on the overall energy consumption.

[0093] Optionally, according to the currently activated application scene, the preset device association table is called, the association line is drawn in the virtual model through the rendering engine, and the related data is displayed, realizing the collaborative visualization of abnormal devices and associated systems, and improving the collaborative processing capability.

[0094] In an embodiment of the present application, based on step S1054, a possible embodiment will be given below to specifically illustrate the non-limiting embodiment. Step S1054 specifically includes the following manner: Step S10541: According to the device type and the abnormal level, the triggering rule is defined.

[0095] Exemplarily, the elevator door machine abnormality needs to meet three conditions of state score ≤0.4, continuous 3 times sampling score drop ≥0.1, and environment humidity > 85%; the air conditioner compressor abnormality needs to score ≤0.3, does not recover for 10 minutes, and outlet temperature > 45°C. The triggering conditions of different device types are dynamically configured through the device archive library.

[0096] Step S10542: When the abnormal device state changes, the state check is triggered, and the associated state information is superimposed in the identification.

[0097] Exemplarily, when the elevator is abnormal, the linked access control system and fire alarm device are identified through the BIM association relationship table, the real-time state interface of the access control system is called, whether the access control is locked, and the access record is abnormal. If the access control system detects the abnormal access times, a yellow warning label of the access control system is superimposed beside the elevator abnormality identification. If the fire alarm device triggers the smoke concentration exceeding the standard, a red emergency label is superimposed, and the content is: smoke exceeds the standard. The linkage rule is dynamically maintained through the device association relationship graph to ensure the consistency of the cross-system state display.

[0098] Step S10543: A state change time axis is established for each device, and the time stamp, triggering condition, associated parameter and historical identification style of each state conversion are recorded.

[0099] The operation and maintenance personnel can view the complete time axis by calling the identification, support filtering according to time range and key parameter comparison. Through the historical state backtracking, the operation and maintenance personnel can analyze the development process of the abnormality, identify the potential root cause, and avoid repeated faults.

[0100] Step S10544: Based on the device historical state data and the current state, the ARIMA time series model is used to predict the abnormal development trend in the future preset time period, and the prediction result is superimposed in the identification.

[0101] This embodiment trains the ARIMA time series model based on the device historical state data, such as the score, temperature and current in the past 30 days, inputs the current score, temperature and time to predict the score change in the next 2 hours. The prediction result is displayed in the form of probability interval, and is superimposed on the current identification through the semi-transparent gray label. The prediction model is automatically updated every hour to help the operation and maintenance personnel prepare maintenance spare parts and tools in advance, and reduce the fault response time.

[0102] In an embodiment of the present application, based on step S107, a possible embodiment will be given below to specifically and non-limitingly illustrate the specific implementation thereof. Step S107 specifically comprises the following manner: Step S1071: continuously collect the operating parameters of the repaired equipment, calculate the mean and standard deviation of the parameters, set a stability determination rule, and comprehensively determine whether the equipment is stable in combination with the historical normal operating range of the equipment.

[0103] Optionally, the stability determination rule can be that the continuous 10 sampling points all fall within the interval [μ-1.5σ, μ+1.5σ].

[0104] In this embodiment, after the equipment is repaired, the temperature, current, vibration frequency and other parameters are collected at a preset time interval as the sampling interval, the mean μ and the standard deviation σ of each parameter are calculated. If the continuous 10 sampling points all fall within the interval, and the temperature does not exceed the best range marked by the manufacturer, it is determined that the equipment is stable, and it is ensured that the equipment state meets the long-term operation requirements.

[0105] Step S1072: if the parameter exceeds the threshold value twice, according to the threshold value amplitude and the equipment importance level, match the troubleshooting strategies of different response levels.

[0106] Optionally, the troubleshooting strategy can be based on the equipment severe threshold value trigger to arrive within a preset time period and generate information containing troubleshooting key points.

[0107] In this embodiment, the equipment importance level is predefined, such as important equipment can be elevator, fire fighting, air conditioning, power distribution. General equipment is lighting, access control. The threshold value level is mild: +5% of the threshold value; moderate: +15% of the threshold value; severe: +25% of the threshold value. When the equipment exceeds the threshold value twice, the response strategy is matched through the equipment level x the threshold value level. The task list is automatically pushed to the mobile terminal of the corresponding personnel through the work order system, and is associated with the historical fault cases of the equipment.

[0108] Step S1073: incremental update of the model based on federated learning, classify the effective maintenance data according to the equipment type and the fault mode, upload to the federated learning center node through the Paillier algorithm; the federated learning center node performs weighted average on the encrypted gradients of the same type of equipment, generates a global optimization model and distributes it to the edge equipment.

[0109] In this embodiment, the equipment locally trains a BP neural network sub-model to extract the weight update amount related to fault prediction. The center node collects the encrypted gradients of all similar equipment, performs weighted average according to the equipment data volume, generates a globally optimized weight, and distributes it to all edge equipment. The edge equipment merges the globally optimized weight with the local weight to complete the model update. The model update efficiency is improved, and the prediction accuracy is improved.

[0110] Step S1074: Based on the current device state and historical inspection data, the priority in the global inspection route is calculated, and a path optimization algorithm is called to generate an inspection route containing the maintenance device, balancing the global inspection efficiency and device risk.

[0111] The embodiment is based on the current state of the device and historical data, and calculates the device risk index. Illustratively, the device is divided into key inspection, regular inspection and low-frequency inspection according to the risk index. A genetic algorithm is called to generate an inspection route for each inspection importance, and the key inspection device is preferentially covered. After dynamic adjustment of the inspection route, the key device inspection coverage rate is improved.

[0112] Step S1075: Based on the recurrence probability prediction and visualization of historical data, the associated features of the current fault of the device are extracted and input into a pre-trained recurrence probability prediction model, and the recurrence probability value in the future preset time period is output. A dynamic trend label is added to the device node in the twin model, and a preventive maintenance suggestion is generated.

[0113] The embodiment predicts the recurrence risk in advance, and the operation and maintenance personnel can strengthen the monitoring, the visualized label intuitively displays the risk level, and the analysis time is reduced.

[0114] In an embodiment of the present application, the method further comprises the following steps: S2011: In the virtual-real mapping inspection of the property system based on digital twinning, on the basis of original operation data acquisition and state analysis, the virtual-real data is fused, and the fused data is compared with the standard operation state data set of the device in the digital twinning model in each dimension, and the specific parts of the device with cooperative anomalies between the data are marked, providing abnormal feature association prompts for inspection personnel.

[0115] The embodiment utilizes the complementarity of virtual-real data, combines environmental data to exclude external interference, and narrows down the abnormal positioning range to specific device parts through data association. The abnormal position positioning accuracy is improved, and a clear troubleshooting direction is provided for the inspection personnel.

[0116] S2012: Before the execution of the inspection path optimization link, based on the physical connection relationship and functional dependence relationship of the device in the digital twinning model, an associated influence graph is constructed for the device marked with an anomaly, and a list of other devices affected by the device anomaly and an influence level are evaluated. The list and the influence level are superimposed on the inspection path planning layer of the digital twinning model, and the abnormal devices and the associated devices with an influence level greater than or equal to 3 are preferentially covered when the path is optimized.

[0117] The embodiment first constructs an equipment correlation graph in the digital twin model, sets a weight for each correlation; after marking the abnormal equipment, the system traverses the first-level correlated equipment in the graph, such as the direct correlation equipment of the water pump, i.e., the water supply valve and the pressure gauge, and the second-level correlated equipment, such as the floor water pipe correlated with the water supply valve, and calculates the affected degree. Based on the physical connection and functional dependence logic between the equipment, the abnormal equipment may spread the fault through the link, and the influence range and degree are quantified through the correlation graph to identify potential hidden dangers in advance.

[0118] S2013: When the inspection personnel arrive at the target equipment to perform the troubleshooting task, the digital twin model is started to realize real-time visualized cooperation with the on-site inspection. The inspection personnel takes an equipment image through the mobile terminal, matches the corresponding equipment node in the digital twin model, and synchronizes the visual angle.

[0119] The mobile terminal of the inspection personnel in the embodiment supports image recognition and AR function. When the equipment is photographed, the equipment identifier is recognized, the visual angle is synchronized through the feature point alignment after the digital twin model node is matched. After the cooperation request is initiated, a video stream channel is established, the AR picture of the inspection personnel is transmitted to the background in real time, and the technical personnel mark the troubleshooting key points on the model, which are superimposed on the on-site AR view in real time. Based on the visual angle synchronization of the digital twin model and the AR augmented reality technology, the communication error between the on-site and the background is eliminated.

[0120] S2014: After each inspection work order is archived, the actual operation parameters of the equipment in this inspection, the actual measured value of the abnormal position are extracted, and the preset parameters and the predicted abnormal position data of the corresponding equipment node in the digital twin model are compared to calculate the deviation rate. The deviation rate = (|actual value-model value| / model value) x 100%.

[0121] If the deviation rate of a certain parameter is continuously more than ±5% for three times, the equipment service life and environmental change data are called to adjust the parameter reference value and physical property of the equipment node in the model to realize the self-calibration of the model.

[0122] S2015: After the equipment maintenance report and the recurrence probability curve are generated, the abnormal type, the troubleshooting method, and the repair scheme are classified according to the equipment type, the service life, and the installation environment, and are stored in the property inspection knowledge base. When the same type of equipment initiates the inspection, similar cases are pushed; the recurrence probability curve is synchronized to the equipment management system to predict the high-risk period and generate a maintenance reminder. The average recurrence probability and the maintenance cost of the same type of equipment are provided to the procurement department as a procurement reference.

[0123] The embodiment processes the inspection results and stores them in the knowledge base according to the equipment type, the service life, and the environment; when the same type of equipment in other buildings is inspected, the system calculates the case similarity, pushes the case and the troubleshooting key points, and improves the rationality of the inspection process.

[0124] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0125] The following is an embodiment of a virtual-real mapping inspection device for a property system based on digital twinning provided by the embodiments of the present disclosure. The device and the virtual-real mapping inspection method for a property system based on digital twinning described above belong to the same inventive concept. Details not described in the embodiment of the virtual-real mapping inspection device for a property system based on digital twinning can be referred to the above embodiment of the virtual-real mapping inspection method for a property system based on digital twinning.

[0126] As shown in Figure 2 The device comprises: A twinning framework modeling module 201 is configured to build a hierarchical framework of a digital twin of a property system and a lightweight virtual model. A coordinate system binding module 202 is configured to collect static data and dynamic data of the property system, perform noise reduction processing on the collected dynamic data, establish a global coordinate system with a building reference point as the origin, and convert and store the position information of each device into coordinate data. A scene device binding module 203 is configured to build a parameterized model according to a 1:1 scale for key devices based on the static data and the dynamic data, using a building BIM model, export the parameterized model into a preset format, assign identification information to each device, bind the device with a corresponding node in the digital twin model, and associate the bound node with static attributes, dynamic attributes, and logical attributes. The model is configured with rigid body physical properties and state visualization rules. An operation data input module 204 is configured to obtain operation data of the device, and input the operation data into a preset device state analysis model. An abnormality visualization module 205 is configured to output a device state score from the device state analysis model. If the device state score exceeds a preset threshold value, or a device abnormality work order submitted by a user is received, an abnormal state identification is performed on the corresponding device node in the parameterized model. The device abnormality related maintenance work order information is superimposed on the label of the corresponding device node in the digital twin model, to realize the visual display of the abnormal state. A virtual-real mapping module 206 is configured to generate an optimal inspection path based on the abnormal device state and the coordinate position information, and push the optimal inspection path to a mobile terminal. After the inspection personnel arrive at the scene, the virtual-real synchronization is triggered by scanning the code, the historical data of the device and the perspective view of the concealed engineering are obtained, and the state confirmation and processing are performed. A verification and update module 207 is configured to verify the device parameter stability after the processing is completed, update the neural network model weight, and adjust the subsequent inspection strategy and monitoring frequency.

[0127] AsFigure 3 As shown, the present application also provides an electronic device, comprising a display module 103, a memory 102, a processor 101, and a computer program stored in the memory and capable of running on the processor 101, wherein the processor 101 implements the steps of the property system virtual-real mapping inspection method based on digital twinning when executing the program.

[0128] In embodiments of the present application, the electronic device includes, but is not limited to, a laptop computer, a desktop computer, a workstation, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections, and relationships, and their functions, are shown as examples only and are not meant to limit implementations of the present application described and / or claimed herein.

[0129] In embodiments of the present application, the processor 101 can be implemented by using at least one of an application-specific integrated circuit, a programmable logic device, a field programmable gate array, a processor, a controller, a microcontroller, a microprocessor, an electronic unit designed to perform the functions described herein, and in some cases, such implementation can be implemented in a controller. For software implementation, the implementation of such processes or functions can be implemented with separate software modules that allow at least one function or operation to be performed. The software code can be implemented by a software application (or program) written in any appropriate programming language, which can be stored in the memory and executed by the controller.

[0130] The display module 103 is used to display information input by the user or information provided to the user. The display module 103 can include a display panel, which can be configured in the form of a liquid crystal display, an organic light-emitting diode, etc.

[0131] The memory 102 can be used to store software programs and various data. The memory 102 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other volatile solid-state memory device.

[0132] The present application also provides a storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the steps of the property system virtual-real mapping inspection method based on digital twinning.

[0133] The storage medium can be any available medium that can be accessed by a general purpose or special purpose computer. By way of example, and not limitation, such computer-readable media can comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code means in the form of instructions or data structures and that can be accessed by a general-purpose or special-purpose computer, or a general-purpose or special-purpose processor. Also, any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or other

[0134] In this document, the terms "computer-readable medium" or "computer- readable media" is used to generally refer to media such as removable storage, volatile memory, non-volatile memory, or any other storage medium readable by a computer or a general purpose or special purpose computer. Exemplary computer-readable storage media includes storage mediums such as a magnetic, optical, or semiconductor storage medium, or any other medium which is used to provide storage "instructions" to a computer or a general purpose or special purpose computer system. Exemplary embodiments of the present application can be implemented as a method, apparatus, or article of manufacture using standard programming and / or engineering techniques. The term "article of manufacture" as used herein is intended to encompass a computer program accessible from a computer-readable medium or storage media providing program instructions to a computer or a general purpose or special purpose computer system. In addition, the article of manufacture can take many forms of media, for example, but not limited to, a

[0135] The foregoing description of the disclosed embodiments enables a person skilled in the art to implement or use the application. Numerous modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for inspecting a property management system based on digital twins, characterized in that the method... include: S101: Construct a layered framework and lightweight virtual model for digital twins of property management systems; S102: Collect static and dynamic data from the property management system, and perform noise reduction processing on the collected dynamic data; A global coordinate system is established with the building reference point as the origin, and the position information of each device is converted into coordinate data and stored. S103: Based on static and dynamic data, construct parametric models of key equipment at a 1:1 scale using the building BIM model, and export the parametric models in a preset format; assign identification information to each piece of equipment, bind the equipment to the corresponding nodes in the digital twin model, and associate static, dynamic, and logical attributes with the bound nodes; configure rigid body physical properties and state visualization rules for the model. S104: Acquire equipment operating data; input the operating data into the preset equipment status analysis model; S105: Equipment status score output by the equipment status analysis model; if the equipment status score exceeds the preset threshold, or if a user submits an equipment anomaly work order, then in the parameterized model, the corresponding equipment node is marked as having an anomaly status; the maintenance work order information related to the equipment anomaly is overlaid onto the label of the corresponding equipment node in the digital twin model to achieve a visual display of the anomaly status. S106: Based on the abnormal equipment status and coordinate location information, generate the optimal inspection path and push it to the mobile terminal; After the inspection personnel arrive at the site, they trigger virtual-real synchronization by scanning the code to obtain historical data of the equipment and perspective view of the hidden works, and then confirm and process the status. S107: After processing, verify the stability of equipment parameters, update the weights of the neural network model, and adjust the subsequent inspection strategy and monitoring frequency.

2. The property system virtual-real mapping inspection method based on digital twins according to claim 1, characterized in that, Step S103 specifically includes the following methods: The model application level is divided according to the application scenario of the digital twin model. The dimensions of the building structure frame, equipment connection interface and key components are retained. After the decorative components are converted, the key dimensional deviations between the model and the physical entity are verified by size comparison. When performing 1:1 parametric modeling of key equipment, a dynamic working condition adaptation component is embedded. Based on the operating characteristics of elevators and air conditioning equipment, the motion constraint range of model components is preset, and an interface for association with equipment operating parameters is reserved. When associating device attributes, a cross-layer collaborative attribute pool is built, which binds static attributes to the device specifications, connects dynamic attributes to the real-time data stream of the data layer edge nodes, and associates logical attributes with the application layer scenario function modules to achieve cross-layer real-time synchronization of attribute data. When configuring model status visualization rules, a scenario-based dynamic rule library is built. Visualization units are configured according to property operation and maintenance scenarios, and color gradients are used to identify the status. Emergency rescue scenarios are overlaid with animations and sound and light linkage prompts, and energy consumption audit scenarios are associated with energy consumption data heat maps.

3. The property system virtual-real mapping inspection method based on digital twins according to claim 1, characterized in that, Step S105 specifically includes the following methods: Abnormal equipment conditions are identified and classified into minor, moderate, and severe abnormality levels based on equipment condition scores, with different visual labels used for each level. Verify the abnormal state by combining the score output by the equipment status analysis model, retrieving the equipment's three most recent historical fault records, the concurrent operating data of similar equipment, and related environmental parameters. If all the data are abnormal, the abnormality is confirmed to be valid; otherwise, it is marked as a suspected abnormality and a secondary detection is triggered. Repair work order information is overlaid on the node labels of the digital twin model, and the information is displayed on demand through the hierarchical folding function within the labels.

4. The property system virtual-real mapping inspection method based on digital twins according to claim 1, characterized in that, Step S105 also includes the following methods: Define triggering rules based on equipment type and anomaly level; when the status of an abnormal device changes, trigger a status check and overlay associated status information into the identifier; Establish a state change timeline for each device, recording the timestamp, triggering conditions, associated parameters, and historical identifier styles for each state transition; Based on the historical status data and current status of the equipment, the abnormal development trend in the next 2 hours is predicted by the ARIMA time series model, and the prediction results are superimposed in the label.

5. The property system virtual-real mapping inspection method based on digital twins according to claim 1, characterized in that, Step S106 specifically includes the following methods: By binding the device to the nodes of the digital twin model, the abnormal device is marked as a visual identifier in the twin model and repair work order information is overlaid; the physical coordinates of the abnormal device are obtained and the coordinate information is bound to the model nodes; Using abnormal equipment as high-priority nodes, a graph theory model G=(V,E) is constructed, which includes equipment location, physical obstacles, and passage priority. Node V represents equipment / critical location, and edge E represents passable paths with assigned weights. An improved Dijkstra algorithm is used to calculate the shortest obstacle avoidance path and push it to the mobile terminal of the inspection personnel. Inspection personnel use mobile terminals to scan the device's QR code, call up SLAM positioning to calculate the device's global coordinates, the twin model highlights nodes in real time and activates the AR perspective function, aligning the real-world photo with the model's viewpoint; After completing the maintenance, the inspection personnel scan the QR code again to call up the equipment status analysis model to verify the current parameters in real time. Continuously monitor the parameter stability after the abnormal equipment is repaired. If the parameter fluctuation is ≤±5%, the repair is considered successful, the work order is archived, and a maintenance report is generated.

6. The property system virtual-real mapping inspection method based on digital twins according to claim 1, characterized in that, Step S107 specifically includes the following methods: Continuously collect the operating parameters of the equipment after repair, calculate the mean and standard deviation of the parameters, set stability judgment rules, and combine the historical normal operating range of the equipment to comprehensively judge whether the equipment is stable; If the parameter exceeds the threshold twice, different investigation strategies with different response levels will be matched according to the magnitude of the threshold exceedance and the importance level of the equipment. Based on the current equipment status and historical inspection data, the priority in the global inspection route is calculated, and the path optimization algorithm is called to generate an inspection route that includes the equipment under maintenance, balancing global inspection efficiency and equipment risk. Based on historical data, the recurrence probability prediction and visualization are used to extract the correlation features of the current equipment failure, input them into a pre-trained recurrence probability prediction model, output the recurrence probability value within a preset time period in the future, and add dynamic trend labels to the equipment nodes in the twin model to generate preventive maintenance suggestions.

7. The property system virtual-real mapping inspection method based on digital twins according to claim 1, characterized in that, The method also includes the following steps: Based on the existing operational data collection and status analysis, the virtual and real data are merged. The merged data is then compared with the standard operational status dataset of the equipment in the digital twin model dimension by dimension to mark the specific parts of the equipment where there are collaborative anomalies between the data, providing inspection personnel with abnormal feature association prompts. Before the inspection route optimization process, based on the physical connection and functional dependency of the equipment in the digital twin model, an associated impact map is constructed for the marked abnormal equipment. The list of other equipment affected by the abnormal equipment and the level of impact are evaluated. The list and the level of impact are overlaid on the inspection route planning layer of the digital twin model. During route optimization, abnormal equipment and associated equipment with an impact level ≥ 3 are covered first. When inspection personnel arrive at the target equipment to perform inspection tasks, the real-time visualization collaboration between the digital twin model and the on-site inspection is activated. The inspection personnel take pictures of the equipment through mobile terminals, match the corresponding equipment nodes in the digital twin model and synchronize the view. After each inspection work order is archived, the actual operating parameters of the equipment inspected and the measured values ​​of abnormal locations are extracted and compared with the preset parameters and predicted abnormal location data of the corresponding equipment nodes in the digital twin model. The deviation rate is calculated as follows: Deviation rate = (|actual value - model value| / model value) × 100%; After generating equipment maintenance reports and recurrence probability curves, the anomaly types, troubleshooting methods, and repair plans are categorized by equipment type, service life, and installation environment, and stored in the property inspection knowledge base.

8. A property management system virtual-real mapping inspection device based on digital twins, characterized in that, The device is used to implement the digital twin-based property system virtual-real mapping inspection method as described in any one of claims 1 to 7; The device includes: The twin framework building module is used to construct a layered framework and lightweight virtual model for building digital twins of property systems. The coordinate system binding module is used to collect static and dynamic data from the property system, and to perform noise reduction processing on the collected dynamic data; a global coordinate system is established with the building reference point as the origin, and the position information of each device is converted into coordinate data and stored. The scene device binding module, based on static and dynamic data, uses the building BIM model to construct parametric models of key equipment at a 1:1 scale, and exports the parametric models in a preset format; it assigns identification information to each device, binds the device to the corresponding node in the digital twin model, and associates static, dynamic, and logical attributes with the bound nodes; it configures rigid body physical properties and state visualization rules for the model. The operation data input module is used to acquire the equipment's operation data and input the operation data into a preset equipment status analysis model. The anomaly visualization module is used to output the equipment status score from the equipment status analysis model. If the equipment status score exceeds the preset threshold, or if a user submits an equipment anomaly work order, the corresponding equipment node is marked as having an anomaly status in the parameterized model. The maintenance work order information related to the equipment anomaly is overlaid onto the label of the corresponding equipment node in the digital twin model to achieve a visual display of the anomaly status. The virtual-real mapping module generates the optimal inspection path based on the abnormal equipment status and coordinate location information and pushes it to the mobile terminal. After the inspection personnel arrive at the site, they trigger virtual-real synchronization by scanning the code to obtain historical equipment data and perspective view of hidden works, and then confirm and process the status. The verification and update module is used to verify the stability of equipment parameters after processing, update the weights of the neural network model, and adjust the subsequent inspection strategy and monitoring frequency.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the property system virtual-real mapping inspection method based on digital twins as described in any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the digital twin-based property system virtual-real mapping inspection method as described in any one of claims 1 to 7.

Images