A digital twin-based intelligent inspection method and system for business hall equipment

Abstract

This invention relates to the field of equipment operation and maintenance management technology in business halls, and discloses a digital twin-based intelligent inspection method and system for business hall equipment. The method includes: synchronously collecting multi-source equipment operation data and spatial location sensor data in the business hall; outputting a standardized runtime sequence dataset and a digital twin baseline model of the business hall after preprocessing; identifying abnormal equipment states through degradation feature extraction and trend prediction algorithms, and outputting single-equipment abnormality indicators, degradation index, and equipment health level parameters; extracting equipment topology relationships and risk propagation characteristics to determine regional risk levels and inspection priorities; generating dynamic inspection instructions based on the determination results and equipment importance; achieving virtual-real linkage through digital twins during inspection execution, and iteratively updating the twin model and evaluation baseline after task completion. This invention achieves accurate identification, intelligent scheduling, and closed-loop management of business hall equipment inspections by integrating equipment degradation states and topology-related risks for hierarchical inspection scheduling.

Description

Technical Field

[0001] This invention relates to the field of operation and maintenance management technology for business hall equipment, and in particular to an intelligent inspection method and system for business hall equipment based on digital twins. Background Technology

[0002] As the core carrier of offline services, the business hall is equipped with various facilities such as power equipment, self-service terminals, and security and fire protection equipment. The operating status of these facilities directly affects service continuity and site safety. Traditional inspection methods mainly rely on manual inspections along fixed cycles and routes, resulting in low inspection efficiency, uneven workload for personnel, and high rates of missed or incorrect inspections. Furthermore, inspection results depend on personnel experience, making it difficult to quantify and assess equipment deterioration trends.

[0003] Existing intelligent inspection solutions monitor parameters by adding online sensors, but most only set threshold alarms for individual devices, lacking topological correlation analysis between devices and global risk assessment, and failing to identify cascading failures caused by risk propagation. Furthermore, they lack digital twin technology, resulting in a lack of a visualized inspection platform that links virtual and physical elements, leading to a disconnect between inspection task scheduling and on-site execution, and delayed anomaly response. In addition, traditional solutions often use fixed inspection cycles and paths, failing to dynamically adjust inspection priorities based on the actual health status of equipment. This easily leads to resource misallocation, such as insufficient inspection of level-one risk devices and excessive inspection of level-three risk devices, making it difficult to balance operational efficiency and cost. Therefore, an intelligent inspection method that integrates digital twins, risk propagation analysis, and dynamic task scheduling is needed. Summary of the Invention

[0004] To address the aforementioned technical shortcomings, the purpose of this invention is to propose an intelligent inspection method and system for business hall equipment based on digital twins. This aims to solve the technical problems in existing technologies that rely solely on single-device threshold alarms, lack topological correlation and global risk assessment, resulting in delayed risk identification, rigid inspection scheduling, disconnect between virtual and real systems, and insufficient closed-loop operation and maintenance capabilities.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides an intelligent inspection method for business hall equipment based on digital twins.

[0006] The aforementioned intelligent inspection method for business hall equipment based on digital twins includes: Step S10: Synchronously collect multi-source equipment operation data and spatial point sensor data in the business hall. After data cleaning, spatiotemporal alignment and twin modeling preprocessing, output a standardized runtime sequence dataset, equipment spatial topology relationship and business hall digital twin benchmark model. Step S20: Based on the standardized runtime sequence dataset obtained in step S10, an anomaly feature extraction and trend prediction algorithm is used to perform equipment degradation status detection, and output single equipment anomaly flags, degradation index and equipment health level parameters. Step S30: Based on the equipment spatial topology relationship in step S10 and the single equipment anomaly flag, degradation index and equipment health level parameters output in step S20, regional risk assessment and inspection priority determination are performed using correlation propagation analysis and global risk fusion mechanism, and regional risk level, inspection priority ranking and key equipment list are output. Step S40: Based on the regional risk level, inspection priority ranking and key equipment list output in step S30, generate intelligent inspection instructions using a dynamic task allocation and path planning mechanism, and output inspection task packages, optimal inspection paths and equipment inspection key point instructions. Step S50: Based on the inspection instructions output in step S40, the digital twin baseline model of the business hall in step S10, and the equipment health level parameters output in step S20, the inspection task is executed using a digital twin-driven virtual-physical linkage inspection execution and closed-loop feedback mechanism, the digital twin model is updated synchronously, and the inspection results and model iteration status are output.

[0007] Preferably, step S10, which involves simultaneously collecting multi-source equipment operation data and spatial location sensor data in the business hall, and performing data cleaning, spatiotemporal alignment, and twin modeling preprocessing to obtain a standardized runtime sequence dataset, equipment spatial topology relationships, and a digital twin baseline model of the business hall, specifically includes: Step S101: Simultaneously acquire the operating parameters of power equipment, self-service terminals and security equipment in the business hall using a multi-source heterogeneous acquisition method, including voltage, current, shell temperature and operating status code, and at the same time collect ambient temperature and humidity, smoke detector, access control point sensor data, as well as equipment spatial coordinates and electrical connection relationship data. Step S102: Perform missing value interpolation and completion, outlier removal and dimensional standardization on the collected multi-source data, and perform millisecond-level spatiotemporal alignment according to a unified timestamp to generate a standardized runtime sequence dataset with equal intervals. Step S103: Based on the spatial coordinates of the equipment, the topological connection relationship and the 3D scene base map of the business hall, construct the equipment-level digital twin, map the equipment attribute parameters, real-time operating status and spatial position relationship, and generate the business hall digital twin benchmark model.

[0008] Preferably, step S20, based on the standardized runtime sequence dataset obtained in step S10, involves using an anomaly feature extraction and trend prediction algorithm to perform equipment degradation status detection and outputting single-equipment anomaly flags, degradation indexes, and equipment health level parameters. This step specifically includes: Step S201: Extract time-domain statistical features and frequency-domain energy features from the standardized runtime sequence dataset, and construct a single-device health feature vector. The time-domain statistical features include mean, variance, and peak margin, and the frequency-domain energy features include the energy proportion of a specified frequency band. Step S202: Use a long short-term memory network to predict the trend of equipment operating parameters, obtain the parameter prediction baseline, calculate the relative deviation between the current measured value and the prediction baseline, and combine the health feature vector to calculate the equipment degradation index. The calculation formula is as follows:

[0009] in The comprehensive score is the normalized health feature vector. The relative deviation of the parameters, The proportion of high-frequency anomalous energy. , , These are the pre-calibrated feature weight coefficients. + =1; Step S203: When the degradation index exceeds the preset abnormal threshold and the trend prediction deviation is greater than the set limit, set the single device abnormal flag and record the degradation value; otherwise, clear the abnormal flag. At the same time, divide the device health level according to the degradation range, and output the single device abnormal flag, degradation index and device health level parameters. The device health level includes four levels: normal, attention, abnormal and fault, which are used for the inspection priority weighting in step S30 and the detection depth matching in step S50.

[0010] Preferably, the feature weight coefficient in step S202 is taken as: =0.4、 =0.35、 =0.25, the anomaly threshold is set to 0.6, when the degradation index A value ≥0.6 indicates abnormal deterioration of the equipment.

[0011] Preferably, in step S30, based on the equipment spatial topology relationship of step S10 and the single equipment anomaly flag, degradation index, and equipment health level parameters output in step S20, the step of performing regional risk assessment and inspection priority determination using correlation propagation analysis and global risk fusion mechanism, and outputting regional risk level, inspection priority ranking, and key equipment list specifically includes: Step S301: Construct a device association network based on the spatial topology of the devices, define the electrical connection association weight and the spatial adjacency association weight respectively, and calculate the risk propagation coefficient by combining the single device degradation index to obtain the indirect degradation impact value of the associated devices; Step S302: Divide the statistical units according to the functional areas of the business hall, integrate the direct deterioration degree and indirect impact value of all equipment in the area, calculate the comprehensive risk value of the area, and divide the area into risk levels. A comprehensive risk value ≥ 0.8 is a level 1 risk area, 0.5 ≤ comprehensive risk value < 0.8 is a level 2 risk area, and a comprehensive risk value < 0.5 is a level 3 risk area. Step S303: Introduce equipment importance weights and match them with the level correction coefficients corresponding to the equipment health level parameters output in step S20. Combine the single equipment deterioration degree and the risk level of the area to calculate the inspection priority score. Generate the inspection priority ranking in descending order of the scores and include the top 30% of the equipment in the key equipment list. Output the area risk level, inspection priority ranking and key equipment list.

[0012] Preferably, step S40, based on the regional risk level, inspection priority ranking, and key equipment list output in step S30, generates intelligent inspection instructions using a dynamic task allocation and path planning mechanism, and outputs inspection task packages, optimal inspection paths, and equipment inspection point instructions, specifically includes: Step S401: Based on the inspection priority ranking and the list of key equipment, match the skill tags, current location and assigned task load of the inspection personnel, perform dynamic task allocation, and generate the initial inspection task package for each inspection personnel. Step S402: Based on the spatial layout of the business hall and the coordinates of the equipment locations, an improved genetic algorithm is used to plan the optimal inspection path, satisfying the constraints of prioritizing the inspection of high-priority equipment and minimizing the total path mileage, while avoiding areas obstructed by passageways. Step S403: Output hierarchical intelligent inspection instructions. For equipment in the first-level risk area, the instructions include mandatory inspection labels, key inspection points, and shortened inspection cycles. For equipment in the second-level risk area, the instructions include key inspection items and regular inspection cycles. The instructions include at least the inspection personnel label, task equipment list, optimal inspection path, inspection items for each device, reference thresholds, and abnormal handling procedures.

[0013] Preferably, in step S50, the inspection instructions output in step S40, the digital twin baseline model of the business hall in step S10, and the equipment health level parameters output in step S20, employ a digital twin-driven virtual-physical linkage inspection execution and closed-loop feedback mechanism to execute the inspection task, synchronously update the digital twin model, and output the inspection results and model iteration status. Specifically, this includes the following steps: Step S501: Based on the digital twin benchmark model of the business hall in step S10, the inspection personnel receive inspection instructions through mobile terminals, perform inspections sequentially according to the planned path, and transmit on-site inspection data in real time, including on-site photos, measured parameter readings and abnormal phenomenon records. The digital twin benchmark model of the business hall synchronously maps the on-site inspection status, highlights the equipment to be inspected, inspected and abnormal equipment, and adjusts the depth of the detection items according to the equipment health level parameters. Special verification items are added to abnormal and fault-level equipment. Step S502: For abnormal equipment confirmed by the inspection, trigger the graded early warning sign in the digital twin benchmark model of the business hall, generate corresponding handling suggestions, and feed back the abnormal data verified on site to the degradation detection module to update the feature baseline and prediction model parameters. Step S503: After the inspection task is completed, based on the full inspection data and the equipment operation data of the same period, iteratively update the equipment attribute parameters and health assessment baseline of the digital twin benchmark model of the business hall, generate a standardized inspection report, output the inspection results and model iteration status, and form an operation and maintenance closed loop.

[0014] This invention also provides a digital twin-based intelligent inspection system for business hall equipment, comprising: Multi-source data acquisition and twin modeling module: used to synchronously collect multi-source equipment operation data and spatial point sensor data in the business hall. After data cleaning, spatiotemporal alignment and twin modeling preprocessing, it outputs a standardized runtime sequence dataset, equipment spatial topology relationship and business hall digital twin benchmark model. Equipment deterioration status detection module: Based on the standardized runtime sequence dataset obtained by the multi-source data acquisition and twin modeling module, it uses anomaly feature extraction and trend prediction algorithms to perform equipment deterioration status detection and outputs single-equipment anomaly flags, deterioration index and equipment health level parameters. Regional Risk and Inspection Priority Assessment Module: Based on the spatial topology of equipment based on the multi-source data acquisition and twin modeling module and the output of the equipment deterioration status detection module, the module outputs single equipment anomaly markers, deterioration index and equipment health level parameters. It uses correlation propagation analysis and global risk fusion mechanism to perform regional risk assessment and inspection priority determination, and outputs regional risk level, inspection priority ranking and key equipment list. The dynamic planning module for inspection tasks is used to generate intelligent inspection instructions based on the regional risk level, inspection priority ranking and key equipment list output by the regional risk and inspection priority assessment module. It adopts a dynamic task allocation and path planning mechanism to generate intelligent inspection instructions and output inspection task packages, optimal inspection paths and equipment inspection key point instructions. Twin-driven inspection execution and closed-loop module: Based on the inspection instructions output by the inspection task dynamic planning module, the digital twin benchmark model of the business hall from the multi-source data acquisition and twin modeling module, and the equipment health level parameters output by the equipment deterioration status detection module, the module executes the inspection task using a digital twin-driven virtual-real linkage inspection execution and closed-loop feedback mechanism, updates the digital twin model synchronously, and outputs the inspection results and model iteration status.

[0015] The present invention also provides a digital twin-based intelligent inspection device for business hall equipment, the digital twin-based intelligent inspection device for business hall equipment includes: a memory, a processor, and a digital twin-based intelligent inspection program for business hall equipment stored in the memory and executable on the processor. When the digital twin-based intelligent inspection program for business hall equipment is executed by the processor, it implements the above-mentioned method.

[0016] The present invention also provides a computer program product, the computer program product including a digital twin-based intelligent inspection program for business hall equipment, which implements the above-described method when executed by a processor.

[0017] The beneficial effects of this invention are as follows: By synchronously collecting multi-source equipment operation data and spatial topology data, a digital twin benchmark model of the business hall is constructed, realizing the virtual-real mapping of equipment status, spatial location and operation and maintenance process. This provides a visual carrier and unified data foundation for the entire inspection process, solving the problems of scattered traditional inspection data and disconnect between the field and the back-end.

[0018] A device degradation detection mechanism that combines time-domain and frequency-domain feature extraction with trend prediction replaces fixed threshold alarms, enabling the identification of early, slow degradation trends and reducing the missed detection rate. Combined with risk propagation analysis based on topological associations, it achieves a progressive assessment from single device health to regional global risk, giving inspection decisions a global perspective and enabling early identification of potential cascading faults.

[0019] By driving the dynamic allocation and path optimization of inspection tasks through risk classification and priority ranking, the fixed-cycle and fixed-route mode is replaced. Inspection resources are tilted towards first-level risk and high-importance equipment, improving the utilization rate of operation and maintenance resources. At the same time, the layered inspection instructions take into account the detection depth of different risk levels, avoiding over-inspection and under-inspection.

[0020] Based on the virtual-physical linkage execution and closed-loop feedback mechanism of digital twins, the inspection process is visualized, the anomaly handling is coordinated, and the model iteration is normalized. The inspection data is used to optimize the degradation assessment model and the twin model in reverse, forming a continuously iterative operation and maintenance closed loop, and continuously improving the accuracy of inspections and equipment control capabilities. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the first embodiment of the intelligent inspection method for business hall equipment based on digital twins according to the present invention. Detailed Implementation

[0022] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0023] Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Example 1: As Figure 1 The diagram shown is a flowchart of the first embodiment of the intelligent inspection method for business hall equipment based on digital twins according to the present invention. The first embodiment of the intelligent inspection method for business hall equipment based on digital twins according to the present invention is presented.

[0025] In the first embodiment, the intelligent inspection method for business hall equipment based on digital twins includes: Step S10: Synchronously collect multi-source equipment operation data and spatial location sensor data in the business hall. After data cleaning, spatiotemporal alignment and twin modeling preprocessing, output a standardized runtime sequence dataset, equipment spatial topology relationship and business hall digital twin benchmark model.

[0026] It should be noted that in this step, the multi-source equipment operation data covers the operation parameters and status codes of all types of business hall equipment, including power distribution equipment, self-service terminals, security monitoring, and fire protection facilities. Spatial point sensor data includes location data from environmental sensors, access control systems, smoke detectors, etc., as well as the spatial coordinates and electrical connection relationships of the equipment. Data cleaning uses interpolation algorithms to fill in missing values ​​caused by communication interruptions and uses the 3σ criterion to remove outliers caused by acquisition glitches. Spatiotemporal alignment unifies data sources with different sampling frequencies to the same timestamp, ensuring the temporal consistency of subsequent correlation analysis. The digital twin baseline model uses the three-dimensional scene of the business hall as a base map, attaching twins of all equipment and carrying three core types of information: attributes, status, and location. The output standardized runtime sequence dataset is directly used for degradation detection in step S20, while the equipment spatial topology relationship and digital twin model provide support for risk propagation analysis in step S30 and virtual-physical linkage in step S50.

[0027] Understandably, through this preprocessing step, the scattered heterogeneous sensor data is integrated into a standardized dataset that is unified in time and space, and a digital twin carrier corresponding one-to-one with the physical business hall is built. This not only ensures the data quality of subsequent algorithm inputs, but also provides a foundation for full-process visual management and control, enabling inspections to be upgraded from offline discrete operations to a digital model that links online and offline operations.

[0028] It should be understood that data collection in traditional inspection schemes is often carried out independently by separate systems, lacking a unified spatiotemporal benchmark and visualization platform, which makes it difficult to link and analyze data and present the overall on-site status. This step breaks down data barriers through multi-source synchronous collection, standardized alignment, and twin modeling, achieving an initial mapping between physical and digital spaces, laying the foundation for subsequent intelligent analysis and scheduling.

[0029] For example, a business hall may contain dozens of types of equipment, such as power distribution cabinets, self-service payment machines, cameras, and smoke detectors, from different manufacturers' systems, with sampling frequencies ranging from 1Hz to 10Hz. After unified collection in this step, all data is aligned to a time sequence with a 1-second interval. Simultaneously, virtual twins of each device are placed in the actual location within a 3D twin scene, displaying parameters such as temperature and voltage in real time. When subsequent steps detect abnormal temperatures in the power distribution cabinet, the device can be directly located in the twin model, and its upstream and downstream electrical connections can be viewed.

[0030] Step S20: Based on the standardized runtime sequence dataset obtained in step S10, an anomaly feature extraction and trend prediction algorithm is used to perform equipment degradation status detection, and output single-device anomaly flags, degradation index and equipment health level parameters.

[0031] It should be noted that this step first extracts multi-dimensional health features from time-series data to comprehensively characterize the equipment's operating status; then, it uses a Long Short-Term Memory (LSTM) network to learn the trend patterns of normal equipment operation, generating a predictive baseline, and using the deviation between the measured value and the baseline as an important basis for assessment of degradation; finally, it integrates multiple features to calculate a degradation index between 0 and 1, with a larger value indicating more severe degradation. Single-device anomaly indicators are triggered by both the degradation threshold and trend deviation, avoiding misjudgments caused by fluctuations in a single indicator. Equipment health levels are divided according to degradation intervals, providing a grading basis for subsequent inspection strategy formulation. The output degradation index and health level are input into step S30 for risk assessment, and simultaneously input into step S50 to guide the focus of on-site inspections.

[0032] Understandably, the purpose of this step is to transform raw operational data into quantifiable equipment health status indicators. Compared to fixed threshold alarms that can only identify out-of-limit faults, this method can capture early signs of degradation through trend prediction and multi-feature fusion, enabling proactive fault warnings. The quantified degradation index also provides continuous input for subsequent priority ranking, making inspection scheduling more precise.

[0033] It should be understood that traditional equipment monitoring often uses single upper and lower thresholds, which cannot distinguish between normal fluctuations and early degradation, nor can it assess the degree of degradation, often resulting in a passive situation of "not repairing unless it breaks down, and rushing to repair when it does break down." The degradation assessment mechanism in this step realizes the quantitative classification of equipment health, moves the maintenance checkpoint forward, and shifts from post-fault repair to condition-based preventive inspection.

[0034] For example, the operating temperature of a power distribution cabinet in a business hall was slowly rising for a long time, but had not yet exceeded the fixed alarm threshold of 60°C. However, according to the trend prediction model, its temperature rise rate had deviated from the normal baseline. Combined with the voltage fluctuation characteristics, the degradation index was calculated to be 0.68, which exceeded the abnormal threshold of 0.6. The abnormal flag was set, the health level was judged as "abnormal", and it was promptly pushed to the subsequent risk assessment stage, thus avoiding equipment failure caused by the continuous rise in temperature.

[0035] Step S30: Based on the equipment spatial topology relationship in step S10 and the single equipment anomaly flag, degradation index and equipment health level parameters output in step S20, regional risk assessment and inspection priority determination are performed using correlation propagation analysis and global risk fusion mechanism, and regional risk level, inspection priority ranking and key equipment list are output. It should be noted that the correlation propagation analysis is based on two topological relationships: electrical connection and spatial adjacency. It calculates the impact of a single device's degradation on surrounding devices, obtaining the indirect degradation impact value and simulating the risk transmission effect. Regional risk assessment is statistically analyzed according to functional zones within the business hall, such as the self-service area, power distribution room, and monitoring room. It integrates the direct degradation and indirect impact of all devices within the zone to obtain a comprehensive regional risk value and classify it into levels. The inspection priority calculation incorporates equipment importance weights and matches them with correction coefficients corresponding to the equipment's health level. Faulty devices have the highest correction coefficients, and normal devices have the lowest. Combining degradation degree and regional risk, a final priority score is obtained, and key devices are selected after ranking. The output regional risk level and priority list provides a decision-making basis for task planning in step S40.

[0036] Understandably, this step expands from single-device health assessment to regional-wide risk evaluation. It considers not only the equipment's own degradation status but also achieves differentiated weighting of degradation levels through health level coefficients, avoiding the insufficient granularity of classification caused by using only degradation degree numerical values. Furthermore, it incorporates risk transmission based on topological relationships, avoiding limitations. The priority-sorted equipment list allows inspection resources to be precisely targeted at the highest-risk objects, realizing a shift from comprehensive, average inspections to precise inspections of key equipment.

[0037] For example, the main distribution cabinet in the power distribution room has a degradation index of 0.72 and a health level of "abnormal," with a corresponding correction factor of 1.2. The downstream branch cabinet has a degradation index of 0.35 and a health level of "attention," with a correction factor of 1.0. Through electrical topology correlation analysis, the indirect impact of the main cabinet on the branch cabinet is 0.28. After integration, the overall risk value of the entire power distribution room area is 0.81, classifying it as a Level 1 risk area. Combining the equipment importance weight and the health level correction factor, the main distribution cabinet has the highest priority score and is included in the list of key equipment, followed by the downstream branch cabinet, while the normal-level equipment in other areas has a lower priority.

[0038] Step S40: Based on the regional risk level, inspection priority ranking and key equipment list output in step S30, generate intelligent inspection instructions using a dynamic task allocation and path planning mechanism, and output inspection task packages, optimal inspection paths and equipment inspection key point instructions.

[0039] It should be noted that the dynamic task allocation comprehensively considers the skills and expertise of inspection personnel, their current location, and the amount of work they have already undertaken, assigning high-priority equipment to personnel with corresponding skills to balance personnel workload. Path planning is based on the coordinates of equipment locations and uses an improved genetic algorithm to calculate the shortest inspection route in total distance while ensuring priority access to high-priority equipment, and automatically avoiding physical obstructions such as walls and counters. Layered inspection instructions are configured with different inspection depths for equipment of different risk levels: Level 1 risk equipment requires full inspection of all items and a shortened inspection cycle; Level 2 risk equipment requires focused inspection of Level 1 risk items; and Level 3 risk equipment undergoes routine inspection. The output instructions are directly sent to the inspection personnel's terminal to guide the on-site execution of step S50.

[0040] Understandably, this step transforms the static risk priority list into executable, dynamic inspection tasks and routes. Task allocation considers personnel capabilities and workload, route planning considers efficiency and priority, and hierarchical instructions consider safety and cost, making the inspection plan both scientific and operable, thus solving the problems of rigidity and inefficiency in traditional shift-based inspection scheduling.

[0041] It should be understood that traditional inspections often use fixed schedules and routes, which cannot adapt to the dynamic changes in equipment risks or adjust tasks according to the real-time status of personnel. This often results in uneven workloads among personnel and situations where level-one risk equipment cannot be inspected in a timely manner. The dynamic allocation and intelligent planning mechanism in this step enables the on-demand generation and optimal scheduling of inspection tasks, significantly improving the speed of operation and maintenance response and the efficiency of personnel.

[0042] For example, a business hall was found to have a Level 1 risk area in its power distribution room, which contained two key pieces of equipment. Based on the inspector's skill tags, the system assigned the power distribution equipment inspection task to a certified electrician. Considering the inspector's current location at the business hall entrance, the system planned the optimal route: "Entrance, main power distribution cabinet, branch power distribution cabinet, self-service area equipment," with no backtracking. Simultaneously, a mandatory inspection instruction was generated for the main power distribution cabinet, including eight key inspection points such as temperature, wiring terminals, and load current, specifying that the inspection must be completed within two hours.

[0043] Step S50: Based on the inspection instructions output in step S40, the digital twin baseline model of the business hall in step S10, and the equipment health level parameters output in step S20, the inspection task is executed using a digital twin-driven virtual-physical linkage inspection execution and closed-loop feedback mechanism, the digital twin model is updated synchronously, and the inspection results and model iteration status are output.

[0044] It should be noted that this step uses the digital twin baseline model of the business hall constructed in step S10 as the visualization carrier and data foundation. Real-time communication with the twin backend is achieved through mobile terminals. The location of inspection personnel, inspected equipment, and on-site data are synchronously mapped into the twin model, allowing managers to monitor the inspection progress globally from the backend. The depth of inspection execution is matched to the equipment health level: abnormal and fault-level equipment undergoes additional special checks such as wiring tightness and abnormal noise detection, while normal equipment undergoes routine checks. For faults verified on-site, an alert is triggered in the twin model, and a handling plan is pushed out. Simultaneously, on-site data is sent back to the degradation detection module to correct the prediction model parameters. After the task is completed, the equipment attributes of the twin baseline model are calibrated using the measured data from this inspection, and the baseline standards for health assessment are updated to ensure the model continuously adapts to equipment aging and environmental changes. Finally, a standardized inspection report and model iteration status are output, completing a full operation and maintenance closed loop.

[0045] Understandably, this step transforms the twin benchmark model built by S10 from a static modeling result into a dynamic execution carrier, realizing a closed loop of the entire lifecycle of modeling, use, and iteration; at the same time, the health level parameters output by S20 directly guide the depth of on-site detection, so that the health assessment results are not only used for decision-making and scheduling, but also implemented in the execution stage, realizing the full-link connection of assessment, decision-making, and execution.

[0046] For example, when an inspector arrives at the main distribution cabinet as instructed, and the equipment's health level is "abnormal," the system automatically adds two special tests to the inspection terminal: "infrared temperature measurement of wiring terminals" and "cabinet noise troubleshooting." The inspector measures a temperature of 58°C, discovering signs of overheating at the terminals. After taking photos and transmitting them back, the equipment immediately turns into a red warning sign in the digital twin baseline model of the business hall, and the system simultaneously displays suggestions for handling the overheating issue. After the task is completed, the system incorporates the measured data into the training set, updates the temperature prediction model baseline, and corrects the rated temperature rise attribute of the equipment in the twin baseline model, making subsequent degradation assessments more consistent with the actual condition of the equipment.

[0047] Example 2: Furthermore, the present invention provides a digital twin-based intelligent inspection system for business hall equipment, employing a digital twin-based intelligent inspection method for business hall equipment from the above embodiments. This system addresses the technical problems in existing technologies, such as reliance on single-device threshold alarms, lack of topological correlation and global risk assessment, leading to delayed risk identification, rigid inspection scheduling, disconnect between virtual and real systems, and insufficient closed-loop operation and maintenance capabilities. The beneficial effects of the digital twin-based intelligent inspection system for business hall equipment provided by the present invention are the same as those of the digital twin-based intelligent inspection method for business hall equipment provided in the above embodiments. Other technical features of the digital twin-based intelligent inspection system for business hall equipment are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0048] Example 3: This invention provides a digital twin-based intelligent inspection device for business hall equipment. The device includes: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, which, when executed, enable the at least one processor to perform the digital twin-based intelligent inspection method for business hall equipment described in Example 1. This digital twin-based intelligent inspection device for business hall equipment in this invention can include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. This digital twin-based intelligent inspection device for business hall equipment is merely an example and should not limit the functionality or scope of the invention. A digital twin-based intelligent inspection device for branch office equipment may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in a read-only memory or a program loaded from a storage device into a random access memory. The random access memory also stores various programs and data required for the operation of the digital twin-based intelligent inspection device for branch office equipment. The processing unit, read-only memory, and random access memory are interconnected via a bus. An I / O interface is also connected to the bus. Typically, the following systems can be connected to the I / O interface: input devices including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices including, for example, magnetic tapes, hard disks, etc.; and communication devices. The communication device allows the digital twin-based intelligent inspection device for branch office equipment to communicate wirelessly or wiredly with other devices to exchange data. While a digital twin-based intelligent inspection device for branch office equipment with various systems has been described, it should be understood that implementation or possession of all described systems is not required. It can be implemented alternatively or with more or fewer systems.

[0049] Example 4: This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described intelligent inspection method for business hall equipment based on digital twins. The computer program product provided by this invention can solve the technical problems in the prior art, such as relying solely on single-device threshold alarms, lacking topological correlation and global risk assessment, leading to delayed risk identification, rigid inspection scheduling, disconnect between virtual and real systems, and insufficient closed-loop operation and maintenance capabilities. Compared with the prior art, the beneficial effects of the computer program product provided by this invention are the same as those of the intelligent inspection method for business hall equipment based on digital twins provided in the above-described embodiments, and will not be repeated here.

[0050] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a read-only memory. When the computer program is executed by a processing device, it performs the functions defined in the methods of the embodiments disclosed in this invention.

[0051] It should be understood that the various parts disclosed in this invention can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0052] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the present invention and its equivalents, the present invention also intends to include these modifications and variations.

Claims

1. A method for intelligent inspection of business hall equipment based on digital twins, characterized in that, The methods include: Step S10: Synchronously collect multi-source equipment operation data and spatial point sensor data in the business hall. After data cleaning, spatiotemporal alignment and twin modeling preprocessing, output a standardized runtime sequence dataset, equipment spatial topology relationship and business hall digital twin benchmark model. Step S20: Based on the standardized runtime sequence dataset obtained in step S10, an anomaly feature extraction and trend prediction algorithm is used to perform equipment degradation status detection, and output single equipment anomaly flags, degradation index and equipment health level parameters. Step S30: Based on the equipment spatial topology relationship in step S10 and the single equipment anomaly flag, degradation index and equipment health level parameters output in step S20, regional risk assessment and inspection priority determination are performed using correlation propagation analysis and global risk fusion mechanism, and regional risk level, inspection priority ranking and key equipment list are output. Step S40: Based on the regional risk level, inspection priority ranking and key equipment list output in step S30, generate intelligent inspection instructions using a dynamic task allocation and path planning mechanism, and output inspection task packages, optimal inspection paths and equipment inspection key point instructions. Step S50: Based on the inspection instructions output in step S40, the digital twin baseline model of the business hall in step S10, and the equipment health level parameters output in step S20, the inspection task is executed using a digital twin-driven virtual-physical linkage inspection execution and closed-loop feedback mechanism, the digital twin model is updated synchronously, and the inspection results and model iteration status are output.

2. The intelligent inspection method for business hall equipment based on digital twins according to claim 1, characterized in that, In step S10, the steps of simultaneously collecting multi-source equipment operation data and spatial location sensor data in the business hall, and obtaining a standardized runtime sequence dataset, equipment spatial topology relationship, and digital twin benchmark model of the business hall through data cleaning, spatiotemporal alignment, and twin modeling preprocessing, specifically include: Step S101: Simultaneously acquire the operating parameters of power equipment, self-service terminals and security equipment in the business hall using a multi-source heterogeneous acquisition method, including voltage, current, shell temperature and operating status code, and at the same time collect ambient temperature and humidity, smoke detector, access control point sensor data, as well as equipment spatial coordinates and electrical connection relationship data. Step S102: Perform missing value interpolation and completion, outlier removal and dimensional standardization on the collected multi-source data, and perform millisecond-level spatiotemporal alignment according to a unified timestamp to generate a standardized runtime sequence dataset with equal intervals. Step S103: Based on the spatial coordinates of the equipment, the topological connection relationship and the 3D scene base map of the business hall, construct the equipment-level digital twin, map the equipment attribute parameters, real-time operating status and spatial position relationship, and generate the business hall digital twin benchmark model.

3. The intelligent inspection method for business hall equipment based on digital twins according to claim 2, characterized in that, In step S20, based on the standardized runtime sequence dataset obtained in step S10, the steps of performing equipment degradation state detection using an anomaly feature extraction and trend prediction algorithm, and outputting single-equipment anomaly flags, degradation indexes, and equipment health level parameters specifically include: Step S201: Extract time-domain statistical features and frequency-domain energy features from the standardized runtime sequence dataset, and construct a single-device health feature vector. The time-domain statistical features include mean, variance, and peak margin, and the frequency-domain energy features include the energy proportion of a specified frequency band. Step S202: Use a long short-term memory network to predict the trend of equipment operating parameters, obtain the parameter prediction baseline, calculate the relative deviation between the current measured value and the prediction baseline, and combine the health feature vector to calculate the equipment degradation index. The calculation formula is as follows: in The comprehensive score is the normalized health feature vector. The relative deviation of the parameters, The proportion of high-frequency anomalous energy. , , These are the pre-calibrated feature weight coefficients. + =1; Step S203: When the degradation index exceeds the preset abnormal threshold and the trend prediction deviation is greater than the set limit, set the single device abnormal flag and record the degradation value; otherwise, clear the abnormal flag. At the same time, classify the device health level according to the degradation range and output the single device abnormal flag, degradation index and device health level parameters.

4. The intelligent inspection method for business hall equipment based on digital twins according to claim 3, characterized in that, In step S202, the feature weight coefficient is taken as... =0.4、 =0.35、 =0.25, the anomaly threshold is set to 0.6, when the degradation index A value ≥0.6 indicates abnormal deterioration of the equipment.

5. The intelligent inspection method for business hall equipment based on digital twins according to claim 1, characterized in that, In step S30, based on the equipment spatial topology relationship of step S10 and the single equipment anomaly flag, degradation index, and equipment health level parameters output in step S20, a regional risk assessment and inspection priority determination are performed using a correlation propagation analysis and global risk fusion mechanism, outputting the regional risk level, inspection priority ranking, and key equipment list. Specifically, this includes: Step S301: Construct a device association network based on the spatial topology of the devices, define the electrical connection association weight and the spatial adjacency association weight respectively, and calculate the risk propagation coefficient by combining the single device degradation index to obtain the indirect degradation impact value of the associated devices; Step S302: Divide the statistical units according to the functional areas of the business hall, integrate the direct deterioration degree and indirect impact value of all equipment in the area, calculate the comprehensive risk value of the area, and divide the area into risk levels. A comprehensive risk value ≥ 0.8 is a level 1 risk area, 0.5 ≤ comprehensive risk value < 0.8 is a level 2 risk area, and a comprehensive risk value < 0.5 is a level 3 risk area. Step S303: Introduce equipment importance weights and match them with the level correction coefficients corresponding to the equipment health level parameters output in step S20. Combine the single equipment deterioration degree and the risk level of the area to calculate the inspection priority score, and generate an inspection priority ranking in descending order of the scores.

6. The intelligent inspection method for business hall equipment based on digital twins according to claim 1, characterized in that, In step S40, based on the regional risk level, inspection priority ranking, and key equipment list output in step S30, an intelligent inspection instruction is generated using a dynamic task allocation and path planning mechanism. The steps include outputting the inspection task package, the optimal inspection path, and the equipment inspection key point instructions. Step S401: Based on the inspection priority ranking and the list of key equipment, match the skill tags, current location and assigned task load of the inspection personnel, perform dynamic task allocation, and generate the initial inspection task package for each inspection personnel. Step S402: Based on the spatial layout of the business hall and the coordinates of the equipment locations, an improved genetic algorithm is used to plan the optimal inspection path, satisfying the constraints of prioritizing the inspection of high-priority equipment and minimizing the total path mileage, while avoiding areas obstructed by passageways. Step S403: Output hierarchical intelligent inspection instructions. For equipment in the first-level risk area, the instructions include mandatory inspection labels, key inspection points, and shortened inspection cycles. For equipment in the second-level risk area, the instructions include key inspection items and regular inspection cycles. The instructions include at least the inspection personnel label, task equipment list, optimal inspection path, inspection items for each device, reference thresholds, and abnormal handling procedures.

7. The intelligent inspection method for business hall equipment based on digital twins according to claim 1, characterized in that, In step S50, the inspection instructions output in step S40, the digital twin baseline model of the business hall in step S10, and the equipment health level parameters output in step S20, employ a digital twin-driven virtual-physical linkage inspection execution and closed-loop feedback mechanism to execute the inspection task, synchronously update the digital twin model, and output the inspection results and model iteration status. Specifically, this includes: Step S501: Based on the digital twin benchmark model of the business hall in step S10, the inspection personnel receive inspection instructions through mobile terminals, perform inspections sequentially according to the planned path, and transmit on-site inspection data in real time, including on-site photos, measured parameter readings and abnormal phenomenon records. The digital twin benchmark model of the business hall synchronously maps the on-site inspection status, highlights the equipment to be inspected, inspected and abnormal equipment, and adjusts the depth of the detection items according to the equipment health level parameters. Special verification items are added to abnormal and fault-level equipment. Step S502: For abnormal equipment confirmed by the inspection, trigger the graded early warning sign in the digital twin benchmark model of the business hall, generate corresponding handling suggestions, and feed back the abnormal data verified on site to the degradation detection module to update the feature baseline and prediction model parameters. Step S503: After the inspection task is completed, based on the full inspection data and the equipment operation data of the same period, iteratively update the equipment attribute parameters and health assessment baseline of the digital twin benchmark model of the business hall, generate a standardized inspection report, output the inspection results and model iteration status, and form an operation and maintenance closed loop.

8. A digital twin-based intelligent inspection system for business hall equipment, applied to the digital twin-based intelligent inspection method for business hall equipment as described in any one of claims 1 to 7, characterized in that, The digital twin-based intelligent inspection system for business hall equipment includes: Multi-source data acquisition and twin modeling module: used to synchronously collect multi-source equipment operation data and spatial point sensor data in the business hall. After data cleaning, spatiotemporal alignment and twin modeling preprocessing, it outputs a standardized runtime sequence dataset, equipment spatial topology relationship and business hall digital twin benchmark model. Equipment deterioration status detection module: Based on the standardized runtime sequence dataset obtained by the multi-source data acquisition and twin modeling module, it uses anomaly feature extraction and trend prediction algorithms to perform equipment deterioration status detection and outputs single-equipment anomaly flags, deterioration index and equipment health level parameters. Regional Risk and Inspection Priority Assessment Module: Based on the spatial topology of equipment based on the multi-source data acquisition and twin modeling module and the output of the equipment deterioration status detection module, the module outputs single equipment anomaly markers, deterioration index and equipment health level parameters. It uses correlation propagation analysis and global risk fusion mechanism to perform regional risk assessment and inspection priority determination, and outputs regional risk level, inspection priority ranking and key equipment list. The dynamic planning module for inspection tasks is used to generate intelligent inspection instructions based on the regional risk level, inspection priority ranking and key equipment list output by the regional risk and inspection priority assessment module. It adopts a dynamic task allocation and path planning mechanism to generate intelligent inspection instructions and output inspection task packages, optimal inspection paths and equipment inspection key point instructions. Twin-driven inspection execution and closed-loop module: Based on the inspection instructions output by the inspection task dynamic planning module, the digital twin benchmark model of the business hall from the multi-source data acquisition and twin modeling module, and the equipment health level parameters output by the equipment deterioration status detection module, the module executes the inspection task using a digital twin-driven virtual-real linkage inspection execution and closed-loop feedback mechanism, updates the digital twin model synchronously, and outputs the inspection results and model iteration status.

9. A digital twin-based intelligent inspection device for business hall equipment, characterized in that, The digital twin-based intelligent inspection device for business hall equipment includes: a memory, a processor, and a digital twin-based intelligent inspection program for business hall equipment stored in the memory and executable on the processor. When the digital twin-based intelligent inspection program for business hall equipment is executed by the processor, it implements the digital twin-based intelligent inspection method for business hall equipment according to any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes a digital twin-based intelligent inspection program for business hall equipment. When the digital twin-based intelligent inspection program for business hall equipment is executed by the processor, it implements a digital twin-based intelligent inspection method for business hall equipment according to any one of claims 1 to 7.

Images