Remote component quality control system based on digital twinning and wireless communication

Abstract

The application relates to the technical field of component quality control, and discloses a remote component quality control system based on digital twinning and wireless communication. A component data acquisition module of the system obtains multi-dimensional data of a component to be monitored through a wireless sensor network, generates a quality basic data set in association with a component identification code; a regional association analysis module establishes an adjacent component association mapping according to a spatial topological relationship, and forms a quality association partition map; an environmental response evaluation module analyzes the influence degree of environmental factors on the performance of the component, and generates an environmental response analysis report; a defect positioning module identifies problem points and constructs a defect distribution map; and a quality early warning module integrates the problem points and associated parameters, labels the location of potential safety hazards, and outputs a quality evaluation and risk early warning report. The system realizes remote and comprehensive control of the quality of the component, and helps to timely discover and respond to quality problems.

Description

Technical Field

[0001] This invention relates to the field of component quality control technology, specifically to a remote component quality control system based on digital twins and wireless communication. Background Technology

[0002] In modern engineering construction and industrial production, the quality of components directly affects the operational stability and safety of the entire project or equipment. Whether it is concrete components in building construction, core parts in machinery manufacturing, or structural components in the transportation sector, quality control has always been a key focus of the industry. With the continuous expansion of project scale and the increasing complexity of components, traditional quality control methods have gradually revealed many shortcomings.

[0003] In the traditional model, staff often need to inspect components on-site, obtaining quality data by manually measuring geometric dimensions and using simple instruments to test material parameters. This method not only consumes a lot of manpower and time but is also inefficient, making it difficult to meet the needs of large-scale batch inspection of components. Furthermore, manual inspection is susceptible to subjective factors, making it difficult to guarantee the accuracy and consistency of the results. This is especially true when dealing with parameters that are difficult to observe directly, such as subtle surface features and internal stress distribution; traditional inspection methods fall short in comprehensively capturing the quality information of components.

[0004] Traditional quality control methods lack systematic analysis and correlation processing of component quality data. In practical applications, components do not exist in isolation; adjacent components have close spatial topological relationships and their performance is mutually influential. However, traditional methods often only assess the quality of individual components, ignoring the interrelationships between components and failing to promptly detect overall quality problems that may be caused by mismatches in the parameters of adjacent components. Furthermore, components are continuously affected by environmental factors such as temperature and load during use. These factors gradually affect the components over time, causing changes in their performance. Traditional quality control lacks a dynamic assessment mechanism for the impact of environmental factors, making it difficult to predict potential component quality defects caused by environmental factors. Remedial action is often only taken after defects have appeared and caused some impact, missing the optimal opportunity for timely intervention.

[0005] When quality problems occur in components, traditional methods struggle to quickly and accurately locate the defects. Due to the lack of comprehensive integration and visualization of component quality data, staff must spend considerable time investigating to pinpoint the defects. This not only prolongs the problem-solving cycle but may also lead to further escalation of the problem due to delayed defect location, increasing engineering maintenance costs and safety risks. With the development of remote engineering and intelligent manufacturing, the need for remote, real-time, and precise control of component quality is increasingly urgent. Traditional quality control methods can no longer meet the new requirements of industry development, necessitating a new technological solution to overcome existing bottlenecks. Summary of the Invention

[0006] The purpose of this invention is to provide a remote component quality control system based on digital twins and wireless communication to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides a remote component quality control system based on digital twins and wireless communication, the system comprising:

[0008] The component data acquisition module acquires the geometric dimensions and material parameters of the component to be monitored through a wireless sensor network, collects the surface morphology features and internal stress distribution data of the component, associates the component identification code and calculates the corresponding quality evaluation index, and generates a basic dataset of component quality.

[0009] The regional correlation analysis module extracts the quality evaluation indicators and corresponding spatial location information from the component quality basic dataset, establishes the correlation mapping of adjacent component units according to spatial topology, marks parameter matching and abnormal areas in adjacent component units, and forms a quality correlation partition map.

[0010] The environmental response assessment module obtains component units in the parameter matching area of ​​the quality correlation zoning map, extracts temperature change curves and load fluctuation data, analyzes the change law in combination with material durability parameters, calculates the degree of influence of environmental factors on component performance, and generates an environmental response analysis report.

[0011] The defect location module identifies problem points in component units whose impact exceeds a preset threshold and are located in abnormal areas in the environmental response analysis report, and constructs a component quality defect distribution map.

[0012] The quality early warning module integrates all problem points and their associated parameters in the component quality defect distribution map, marks the location of components with potential safety hazards, and outputs a component quality assessment and risk early warning report.

[0013] Preferably, the component quality basic dataset includes quality evaluation indicators, component spatial coordinates, and standardized material parameters. The quality correlation zoning map specifically includes parameter matching area marking, abnormal area marking, and parameter difference degree between adjacent component units. The environmental response analysis report includes the influence coefficient of temperature change on materials, the influence coefficient of load fluctuation on structure, and a comparison of performance changes under different environmental conditions. The component quality defect distribution map includes the spatial coordinates of defect points, the temperature-load composite characteristics of defect points, and the material parameters and design deviation values ​​of defect points. The component quality assessment and risk warning report includes a list of quality problem locations and a comprehensive judgment result of multiple parameters for problem points.

[0014] Preferably, the component data acquisition module includes:

[0015] The geometric measurement submodule acquires the three-dimensional coordinate data and acquisition time information of the component to be monitored through distributed sensors, collects the surface roughness and internal defect detection data at the corresponding locations, and records the acquisition results as two types of indicators: surface quality parameters and internal quality parameters, forming the original parameter set of the component.

[0016] The parameter standardization submodule performs standardization processing on the surface quality parameters and internal quality parameters in the original parameter set of the component, establishes a correspondence between the processed results and the spatial position of the component, calculates the comprehensive evaluation value of the standardized surface parameters and standardized internal parameters as the quality evaluation index, and generates a basic dataset of component quality.

[0017] Preferably, the regional correlation analysis module includes:

[0018] The indicator extraction submodule obtains the quality evaluation indicators and corresponding spatial location data from the component quality basic dataset, determines the relative positional relationship of all component units based on spatial relationships, calls the component location information set, calculates and sorts the proximity relationships of component units in space based on the adjacent distance standard, and forms a component unit proximity relationship sequence.

[0019] The difference calculation submodule calculates the difference in quality evaluation indicators between each pair of adjacent component units based on the proximity relationship sequence of the component units, and integrates them to generate a quality difference sequence.

[0020] The spatial consistency determination submodule extracts the surface parameter change trend and internal parameter change trend between adjacent component units based on the quality difference sequence, classifies and marks each pair of component units according to the consistency of the two parameter change directions, records and groups the parameter matching and parameter abnormality segments respectively, and forms a quality correlation partition map.

[0021] Preferably, the environmental response assessment module includes:

[0022] The environmental data extraction submodule selects the component units marked as parameter matching areas according to the quality correlation partition map, monitors the temperature gradient data and load change data of each component unit within the observation period, arranges them in time sequence to form temperature change curves and load fluctuation datasets, and constructs a set of environmental influencing factors.

[0023] The performance impact calculation submodule calculates the temperature change rate and load change amplitude between consecutive time points in the monitoring data of each component unit based on the set of environmental impact factors. It performs correlation analysis on the temperature change rate and load change amplitude under the same material conditions, and quantitatively evaluates the effect intensity of the component performance by comprehensively analyzing the two types of environmental indicators. It integrates the impact coefficient sequence of each component unit and generates an environmental response analysis report.

[0024] Preferably, the defect location module includes:

[0025] The monitoring data extraction submodule, based on the environmental response analysis report, filters out component units whose impact exceeds a preset threshold and component units located in abnormal areas, extracts continuous monitoring records of component units in chronological order, and collects extreme temperature data and load peak data at each time point to form a component status monitoring dataset.

[0026] The parameter deviation calculation submodule calls the component condition monitoring dataset to extract the temperature change and load change of the component unit at two consecutive monitoring nodes, integrates them into temperature change sequence and load change sequence, and establishes a component condition change set.

[0027] The problem point identification submodule extracts environmental humidity data and vibration frequency data for the corresponding time period based on the component state change set, determines whether the temperature change and load change both exceed the set abnormal state standard, determines whether the humidity fluctuation and vibration intensity simultaneously meet the defect judgment condition, marks the monitoring points that meet the conditions as problem points, and constructs a component quality defect distribution map.

[0028] Preferably, the quality early warning module includes:

[0029] The multi-parameter comprehensive judgment submodule obtains all problem points and their associated parameter information in the component quality defect distribution map, calculates the comprehensive risk assessment value of each problem point, and forms a comprehensive risk assessment sequence.

[0030] The risk classification output submodule, based on the comprehensive risk assessment sequence, filters out points whose impact exceeds material safety standards, whose quality evaluation indicators are lower than design requirements, and whose spatial consistency is marked as abnormal areas. It extracts the corresponding component number, location information, and region, marks them as quality abnormalities and potential safety hazards, and classifies and outputs the points that meet the comprehensive conditions according to the hierarchical structure of the digital twin model, generating a component quality assessment and risk warning report.

[0031] Preferably, the system further includes:

[0032] The digital twin construction module receives the component quality basic dataset and component quality assessment and risk warning report in real time, and establishes a three-dimensional digital twin model that includes material properties, geometric features and defect distribution;

[0033] The model update module updates the material parameters of the digital twin model in real time based on the dynamic change information in the environmental response analysis report. When a new record is detected in the component quality defect distribution map, the defect spatial distribution of the digital twin model is updated synchronously.

[0034] Preferably, the system further includes:

[0035] The wireless transmission module uses multi-protocol converged communication technology to synchronously transmit the component quality basic dataset, quality correlation zoning map and environmental response analysis report to the cloud digital twin platform;

[0036] The communication optimization module dynamically adjusts the wireless transmission communication protocol and data compression strategy based on the data characteristics in the set of component state changes.

[0037] Preferably, the system further includes:

[0038] The edge computing unit is deployed on the component monitoring terminal, and the computing logic of the component data acquisition module and the regional correlation analysis module is pre-set to perform localization processing and preliminary quality analysis on the raw monitoring data;

[0039] The cloud collaboration module receives the preprocessed component quality basic dataset from the edge computing unit, performs deep calculations by the environmental response assessment module and the defect location module, and feeds back the component quality assessment and risk warning report to the edge computing unit for local storage.

[0040] Compared with the prior art, the beneficial effects of the present invention are:

[0041] This remote component quality control system, based on digital twins and wireless communication, acquires geometric dimensions, material parameters, surface morphology features, and internal stress distribution data of the components under monitoring through a component data acquisition module using a wireless sensor network. It then associates this data with component identification codes to calculate quality evaluation indicators and generate a basic dataset. This system revolutionizes the traditional manual on-site inspection model, eliminating the need for frequent on-site visits by personnel to remotely collect component quality data. The wireless sensor network comprehensively captures multiple dimensions of component quality parameters, accurately acquiring both external geometric features and internal stress states. Furthermore, the data acquisition process is free from subjective human interference, ensuring the objectivity and completeness of the collected quality data and providing a comprehensive and reliable data foundation for subsequent quality analysis and evaluation.

[0042] The regional correlation analysis module extracts quality evaluation indicators and spatial location information, establishes correlation mappings between adjacent component units according to spatial topological relationships, and marks parameter matching and abnormal areas to form a quality correlation zoning map, effectively solving the problem of traditional quality control neglecting component correlation. By constructing correlation mappings between adjacent components, the parameter matching between different component units can be clearly presented, allowing staff to intuitively understand whether there are abnormal areas with parameter mismatches between adjacent components. This enables timely detection of potential overall quality hazards caused by component correlation issues, avoiding the limitations of a local perspective caused by evaluating individual components, achieving systematic control over component quality, and improving the comprehensiveness of quality control from an overall perspective.

[0043] The environmental response assessment module targets component units in parameter-matching areas of the quality correlation zoning map. It extracts temperature change curves and load fluctuation data, analyzes the variation patterns using material durability parameters, and calculates the impact of environmental factors on component performance. This generates an environmental response analysis report, filling the gap in traditional quality control that lacks dynamic assessment of environmental factors. This module can track changes in environmental factors in real time, deeply analyze the specific impacts of these changes on component performance, and use scientific analysis methods to understand the changing patterns of component performance under the influence of environmental factors. This allows staff to anticipate the potential impact of environmental factors on component quality, enabling them to take corresponding preventative measures to avoid long-term environmental factors causing component quality defects, extend component service life, and reduce the incidence of quality problems caused by environmental factors.

[0044] The defect location module identifies component units with problem points in environmental response analysis reports that exceed preset thresholds and are located in abnormal areas. It then constructs a component quality defect distribution map, significantly improving the efficiency and accuracy of defect location. By integrating environmental impact analysis results with abnormal area information, it can quickly pinpoint component units with quality issues and specific problem points, presenting them intuitively in the form of a distribution map. Staff no longer need to conduct extensive on-site inspections; they can clearly understand the location and distribution of defects simply by looking at the defect distribution map. This provides clear guidance for timely defect repair, shortens the problem-solving cycle, and reduces additional losses caused by untimely defect location.

[0045] The quality early warning module integrates all problem points and their associated parameters from the component quality defect distribution map, marks the locations of components with potential safety hazards, and outputs quality assessment and risk warning reports. This module systematically integrates scattered defect information, clearly identifying the locations of components with potential safety hazards. This allows staff to accurately grasp the component quality risk status, prepare for risk responses in advance, and prevent safety accidents. Simultaneously, the output quality assessment report provides comprehensive evidence for component maintenance and replacement decisions, helping to optimize component management processes, reduce management costs, and improve the overall operational reliability of the project or equipment. Attached Figure Description

[0046] Figure 1 This is a timing diagram of the remote component quality control system based on digital twin and wireless communication described in this invention.

[0047] Figure 2 A schematic diagram illustrating the working principle of a remote component quality control system.

[0048] Figure 3 This is a schematic diagram illustrating the working principle of the component data acquisition module. Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] Please see Figure 1 The present invention provides a remote component quality control system based on digital twin and wireless communication. The system includes: a component data acquisition module, a regional correlation analysis module, an environmental response assessment module, a defect location module, and a quality early warning module.

[0051] The component data acquisition module collects geometric dimensions, material parameters, surface morphology features, and internal stress distribution data of the components under monitoring through a wireless sensor network deployed on them. This data is associated with the component identification code, and quality evaluation indicators are calculated and integrated into a basic component quality dataset. The regional correlation analysis module extracts quality evaluation indicators and corresponding spatial location information from the basic component quality dataset. Based on spatial topology, it establishes a correlation mapping between adjacent component units, marking parameter-matching and abnormal regions to form a quality correlation zoning map. The environmental response assessment module acquires temperature change curves and load fluctuation data for component units in parameter-matching regions of the quality correlation zoning map. It analyzes the changing patterns of environmental factors in conjunction with material durability parameters, calculates the impact of environmental factors on component performance, and generates an environmental response analysis report. The defect location module identifies problem points in component units located in abnormal regions whose impact exceeds a preset threshold in the environmental response analysis report, constructing a component quality defect distribution map. The quality early warning module integrates all problem points and their associated parameters from the component quality defect distribution map, marks the locations of components with potential safety hazards, and outputs a component quality assessment and risk warning report.

[0052] Example 1: See Figure 2 The construction of the component quality baseline dataset begins with the comprehensive acquisition of on-site monitoring data by the component data acquisition module. A wireless sensor network, with a distributed architecture, covers key parts of the monitored components. Each sensor node is configured with a unique identifier and bound to the component's identity information. Geometric dimensional data is acquired using a laser rangefinder and optical encoder, forming a three-dimensional point cloud data stream. Material parameters are measured in real time using an embedded acoustic wave detector and resistance strain gauges. Surface morphology features are obtained using high-resolution scanning technology to acquire micro-texture information. Internal stress distribution data is sampled at a frequency of thousands of times per second using a fiber optic grating sensor array. All sensor data undergoes analog-to-digital conversion and baseline calibration before transmission. After being correlated with spatial coordinates via timestamps, the embedded processor calculates preliminary quality evaluation indicators. These indicators integrate twelve dimensions, including dimensional tolerances and material uniformity. The final component quality baseline dataset adopts a hierarchical storage structure, with raw monitoring data, calibrated data, and evaluation indicators stored in different database partitions.

[0053] In terms of workflow, the regional correlation analysis module, when generating the quality correlation zoning map, first establishes a spatial index model to convert the component unit coordinates into a topological mesh, uses the Delaunay triangulation algorithm to determine adjacent component units, marks parameter matching regions based on the sliding window data consistency check, identifies abnormal regions using the isolated forest algorithm, calculates the parameter differences between adjacent component units and introduces weighting coefficients, and finally stores the zoning map as a vector layer, supporting multi-angle rendering. The environmental response analysis report data originates from a temperature sensor chain and a dynamic weighing system. It uses a cumulative damage model to calculate the influence coefficient of temperature changes on materials, calculates the influence coefficient of load fluctuations on the structure based on fatigue damage theory, and compares performance changes under different environments by constructing a multi-dimensional data cube.

[0054] The component quality defect distribution map relies on multi-source data fusion technology to jointly calculate the spatial coordinates of defect points, extract composite features using a time series alignment algorithm, and quantify deviation values ​​through a digital twin model. The component quality assessment and risk warning report employs a multi-level review mechanism, using a fuzzy inference system to calculate the comprehensive judgment result. The report output module supports automatic generation of inspection report templates. During data transmission, monitoring data is encapsulated using a lightweight message queue telemetry transmission protocol, and integrity is ensured through verification. The cloud processing platform establishes a distributed computing cluster to process data, and a real-time stream processing module identifies abnormal patterns. The system's visualization interface constructs a three-dimensional interactive model, with zonal maps overlaid with heatmaps, environmental response analysis reports displayed using dynamic curves, and defect distribution maps presented as 3D marker points.

[0055] Example 2: See Figure 3 The geometric measurement submodule of the component data acquisition module collects data through a distributed sensor network deployed on the component surface. Each sensor node is equipped with a multi-axis accelerometer and a laser displacement sensor, capturing the three-dimensional coordinate changes of the component at a sampling frequency of 200 times per second. A mesh network is established between sensor nodes using a self-organizing network communication protocol, and a time synchronization algorithm ensures that the timestamp deviation of all node data is less than 1 millisecond. Surface roughness is measured using a contact probe profilometer, where the probe moves along the component surface at a constant speed, recording the depth data of microscopic undulations. Internal defect detection utilizes ultrasonic phased array technology, employing a multi-channel ultrasonic probe to emit and receive acoustic signals, constructing a cross-sectional image of the component's internal structure. The acquired raw data is recorded in two categories: surface quality parameters, including characteristic values ​​such as the arithmetic mean deviation of the profile and the average width of the profile unit; and internal quality parameters, covering indicators such as the defect area ratio and porosity distribution density.

[0056] The parameter standardization submodule performs multi-stage processing on the original parameter set. Surface quality parameters first undergo outlier filtering, using a sliding window detection method to identify and remove abrupt changes caused by sensor interference. Internal quality parameters undergo dimensional normalization, converting parameters with different physical meanings into dimensionless ratios, and mapping each parameter to the [0,1] interval using a minimum-maximum standardization method. Material type correction coefficients are introduced during standardization, employing different transformation functions for components made of different materials such as concrete and steel. A mapping relationship is established between the processed parameters and the spatial location of the component, achieving rapid association between parameters and location through spatial indexing technology in a spatial database. The comprehensive evaluation value of standardized surface parameters and standardized internal parameters is calculated using principal component analysis, extracting the main characteristic components from multiple parameters and determining the weight coefficient of each component based on the variance contribution rate. The final generated quality evaluation index is a continuous value between 0 and 100, which is positively correlated with the component quality. The complete component quality dataset is encapsulated in JSON format and consists of three parts: a file header, a data body, and a checksum. The file header records the dataset version number and generation timestamp, while the data body stores the standardized parameters and evaluation metrics of all components in array form.

[0057] The regional association analysis module's index extraction submodule reads quality evaluation indicators and spatial coordinate information from the quality foundation dataset. Spatial relationship calculation employs a quadtree index algorithm, dividing the component's location into different levels of grid cells. Adjacent component cells are determined based on a spatial distance threshold, dynamically adjusted according to 1.5 times the average component size. A kd-tree nearest neighbor search algorithm quickly finds adjacent cells within a specified range around each component cell. The component location information set is stored in a spatial database, supporting range queries and nearest neighbor queries based on R-tree indexes. Topological relationship verification is introduced during the proximity calculation process; in addition to distance conditions, topological constraints such as face adjacency or edge adjacency must be met. The final generated component cell proximity relationship sequence is arranged in ascending order of spatial distance, with each record containing fields such as the main component ID, adjacent component IDs, and actual spacing. The difference calculation submodule, after reading the proximity relationship sequence, uses a dynamic programming algorithm to batch calculate the difference in quality evaluation indicators between adjacent cells. The difference calculation considers not only the current indicator difference but also the similarity of trends over time, evaluating the consistency of long-term behavior by calculating the dynamic time-warped distance between the indicator sequences of two component cells.

[0058] The spatial consistency determination submodule performs multi-scale analysis on the quality difference sequence. Surface parameter variation trends are extracted using wavelet transform to capture variation features across different frequency bands, while internal parameter variation trends are obtained through empirical mode decomposition to obtain intrinsic mode functions. The consistency judgment of parameter variations between adjacent component units is based on trend correlation analysis, calculating the Pearson correlation coefficient and lag cross-correlation value of the parameter variation curves of the two units. The determination of parameter matching regions requires simultaneous satisfaction of both trend correlation thresholds and difference thresholds, while parameter anomaly regions are identified using an outlier detection algorithm. Detected anomaly segments are clustered and grouped according to anomaly type and spatial distribution. The final generated quality correlation zoning map is stored in vector graphics format, containing multiple layers: the basic component layer displays component locations and numbers; the parameter matching region layer is labeled with green semi-transparent polygons; the anomaly region layer is marked with red warning icons; and the difference layer uses gradient color levels to represent the degree of difference.

[0059] At the data communication level, the distributed sensor network employs a time-division multiple access protocol to avoid signal collisions, with each sensor node transmitting data within a designated time slot. Gateway devices perform preliminary verification and filtering of the received data, then transmit the compressed data packets to the cloud processing platform via a 4G / 5G wireless network. The cloud server adopts a distributed architecture, using the Hadoop framework for parallel processing of large-scale data, while real-time data streams are processed asynchronously via an Apache Kafka message queue. The database system uses a hybrid storage scheme, with recent data stored in a Redis in-memory database for real-time querying, and historical data archived in an HBase distributed database for long-term storage. The system's visualization interface adopts a B / S architecture, allowing users to access 3D models and data analysis results through a web browser. The 3D engine is developed based on WebGL technology, supporting rotation, scaling, and translation operations on component models, with quality correlation partition maps displayed on the model surface as overlays.

[0060] Example 3: Environmental Response Assessment Module: The environmental data extraction submodule filters component units with matching parameters and continuous, uniform spatial distribution from the quality correlation zoning map. The monitoring system deploys a temperature sensor array with multiple temperature measurement points on the surface of the component unit to capture temperature gradient changes, and installs dynamic load sensors to record real-time stress. Temperature gradient data is sampled every minute to form a time-series dataset, and load change data is recorded in an event-triggered manner. The observation period is generally 24 hours. The temperature change curve is connected to discrete points using spline interpolation, and the load fluctuation dataset retains the original sampling points and calculates statistical characteristic values. All environmental data is denoised before entry, and the environmental impact factor set is stored in time-aligned to ensure that the temperature and load data timestamps are synchronized. Performance Impact Calculation Submodule: Time-frequency analysis is performed on the environmental impact factor set, and the temperature change rate is calculated using the central difference method. The load change amplitude is calculated considering the load directionality. The same material conditions are determined based on the component unit material formula and age parameters. The correlation analysis uses an improved grey relational algorithm to calculate the correlation coefficient between the temperature change rate sequence and the load change amplitude sequence. An environmental impact weight factor is introduced to quantify the intensity of the component performance effect, generating an impact coefficient sequence containing instantaneous and cumulative impact values. The environmental response analysis report is a structured document that supports retrieval and comparative analysis by time range.

[0061] Defect localization module: The monitoring data extraction submodule sets multi-level filtering conditions, dynamically adjusts the impact threshold, and combines historical data and expert experience to determine abnormal areas. It extracts continuous monitoring records of component units from the time-series database, with sampling intervals set as needed. A sliding window extreme value detection algorithm is used to extract extreme temperature data, and a peak value detection algorithm is used to identify peak load data. The component condition monitoring dataset is stored in columnar format, saving data quality identifiers.

[0062] The problem identification submodule introduces a multi-parameter joint criterion. Environmental humidity data comes from a high-precision humidity sensor, with sampling frequency synchronized with temperature data. Vibration frequency data is obtained through accelerometer spectral analysis. The setting of abnormal condition standards considers component type and usage environment, employing different threshold standards for indoor and outdoor components. The defect judgment logic uses a decision tree model, first determining whether the temperature change exceeds the threshold, then checking whether the load change also exceeds the threshold, and finally making a final judgment based on a combination of humidity fluctuations and vibration intensity. Problem point labeling includes confidence assessment, calculating the confidence level based on data quality and parameter correlation. The generation of the component quality defect distribution map adopts a progressive update strategy, adding newly discovered problem points to the distribution map in real time while retaining historical problem point evolution records.

[0063] A feature-weighted algorithm was introduced during the data analysis process to calculate the comprehensive index of environmental impact:

[0064]

[0065] in: A comprehensive index representing environmental impact. The environmental weight factor represents the i-th time point. It is the rate of temperature change at the i-th time point. The load change magnitude corresponding to the i-th time point, Indicates the humidity deviation value. This is the adjustment coefficient for the influence of humidity. This calculation formula is used to quantify the combined effect of multiple environmental factors on component performance. The weighting factor is dynamically adjusted according to the sensor layout density and data reliability. The humidity deviation value is taken from the difference between the relative humidity and the ideal ambient humidity.

[0066] The system's data acquisition unit utilizes industrial-grade sensors, covering a temperature measurement range of -40℃ to 150℃ with an accuracy of ±0.5℃. The load sensor range is customized based on the actual stress conditions of the components, and its overload protection capacity is 150% of the rated load. All sensors undergo regular on-site calibration, and the calibration data is stored in the equipment management database. Data transmission employs a hybrid wired and wireless network approach, with dual-route transmission for critical data points to ensure data reliability. The data processing platform establishes a distributed computing cluster; the streaming module continuously analyzes real-time monitoring data, while the batch processing module performs in-depth data mining on historical data. The anomaly detection algorithm employs multiple models running in parallel, including a statistical process control model, a machine learning anomaly detection model, and a physical model-driven residual analysis model. The defect prediction function is implemented through a time series prediction algorithm, predicting the likelihood of defect development over a future period based on current trends.

[0067] The visualization system provides a multi-dimensional data display interface. Environmental response analysis results are displayed as heatmaps on a 3D model, with color intensity indicating the degree of impact. The defect distribution map supports time-based scrolling, dynamically displaying the generation and development process of defects. The system also provides a data comparison function, allowing monitoring data from different periods to be displayed side-by-side, facilitating the identification of trends by analysts. The report generation module automatically generates inspection reports that conform to industry standards, including basic component information, monitoring data statistics, defect distribution maps, analysis conclusions, and recommended measures. The report supports multiple output formats, including PDF, Word, and HTML5 interactive reports. All report versions are stored in the document management system, supporting version tracking and difference comparison.

[0068] Example 4: The multi-parameter comprehensive judgment submodule of the quality early warning module extracts problem point data from the component quality defect distribution map. Each problem point includes fields such as spatial coordinates, temperature load composite characteristics, and material parameter deviations. The calculation of the comprehensive risk assessment value adopts the analytic hierarchy process (AHP). First, a judgment matrix is ​​established to determine the weights of each type of parameter through pairwise comparisons. Then, the standardized parameter values ​​are weighted and summed. The weight coefficients are determined with reference to the scoring results of structural engineering experts, considering the degree of influence of parameters on structural safety. For example, the weight of material parameter deviations is higher than that of temperature load characteristics. During the calculation process, the membership function of fuzzy mathematics is introduced to handle the uncertainty of parameters, converting continuous parameter values ​​into risk membership degrees. The comprehensive risk assessment sequence is sorted from high to low according to the evaluation value, and the parameter contribution analysis of each problem point is recorded to identify the main sources of risk.

[0069] The risk grading output submodule sets up multi-level screening conditions. Reference values ​​for material safety standards are derived from current national standards and specifications, with corresponding allowable stress values ​​set for different material types. The comparison between quality evaluation indicators and design requirements uses relative deviation rate calculations; when measured values ​​exceed the allowable fluctuation range of design values, they are marked as unqualified. Verification of spatial consistency markings is achieved through a topological relation database, checking whether problem points are located in abnormal areas marked in the quality correlation zoning map. The comprehensive judgment of points uses a decision tree model, first screening problem points that simultaneously meet three conditions, and then further subdividing them into levels based on risk assessment values. The extraction of component numbers and location information is based on spatial database queries from the BIM model; the determination of the region is based on the component's functional positioning within the overall structure. Safety hazard marking uses a color-coding system: red indicates urgent risk, yellow indicates potential risk, and green indicates normal status. The hierarchical structure of the digital twin model is organized according to a four-level system: "structural system - substructure - component - monitoring point," supporting risk display at different granularities.

[0070] The digital twin building module employs a parametric modeling approach. Material properties are defined using physical parameters such as elastic modulus, Poisson's ratio, and coefficient of thermal expansion, imported from laboratory test data. Geometric features are established by registering laser-scanned point cloud data with the BIM model, achieving precise matching between the actual component and the theoretical model. Defect distribution integration utilizes spatial database technology, mapping the coordinates of defect points to the 3D model. Lightweight model processing uses a mesh simplification algorithm, reducing model data volume while maintaining accuracy and improving rendering efficiency. The model update module establishes a dynamic data monitoring mechanism; dynamic changes in environmental response analysis reports are transmitted via message middleware. Real-time updates of material parameters are based on a damage evolution model, automatically adjusting constitutive parameters when monitoring data indicates material performance degradation. Updates to the spatial distribution of defects employ an incremental update strategy, with new records marked through a version management mechanism, supporting retrospective analysis of change trajectories. Model consistency checks ensure the integrity and logical correctness of the model after each update.

[0071] Taking a large bridge project as an example, abnormal data was found in the bridge pier components during operational monitoring. The multi-parameter comprehensive judgment submodule collected monitoring data such as concrete carbonation depth, steel corrosion rate, and crack width, and obtained a comprehensive risk value through weighted calculation. Refer to Table 1, which shows the risk assessment details of three typical problem points.

[0072] Table 1: Risk Assessment of Bridge Components

[0073]

[0074] A digital twin model meticulously models the bridge, recording the material properties of each component based on actual mix proportions, and including manufacturing and installation deviation data in its geometric dimensions. When the monitoring system detects a decrease in the concrete strength of component P-12-45, the model update module automatically adjusts the material parameters for that area and displays a color change in the visualization interface. The defect distribution map is updated in real-time, with newly discovered crack data synchronized to the 3D model within 30 seconds. The risk warning report generation module uses a template-based design, including sections on project overview, monitoring data statistics, risk distribution map, and remediation recommendations. Report output supports multiple formats: a PDF version for archiving and an interactive HTML version for online viewing of the 3D model. The latest regulatory provisions are automatically referenced during report generation to ensure the compliance of recommended measures.

[0075] The system employs a hierarchical authorization mechanism for access control. Designers can view all model data, while inspection personnel can only view risk assessment results. Operation logs record all model modifications and report generation operations, meeting the requirements of engineering quality management. The data backup system periodically takes snapshots of the digital twin model and monitoring data, supporting status retrospection at any point in time. During bridge operation and maintenance, the system generates a weekly assessment report and performs monthly trend analysis and forecasting. When abnormal data occurs, a special assessment process is automatically triggered, organizing experts to discuss and assess high-risk points. The historical data comparison function can display the evolution of component status, providing decision support for preventative maintenance.

[0076] The system's interface design supports integration with existing engineering management software, enabling the push of risk warning information to the project management platform. The mobile application allows on-site inspectors to view component risk levels in real time and upload on-site inspection photos and notes. All data exchange uses encrypted transmission to ensure the security of engineering data. The model visualization engine supports multi-terminal access; detailed 3D models and parameter information can be viewed on computers, while simplified risk distribution maps are provided on mobile devices. Users can customize the displayed content as needed, selecting to display problem points of specific risk levels or specific types of monitoring parameters. The model measurement tool supports distance, area, and volume measurements, facilitating spatial analysis by engineers. The alarm notification system has multi-level threshold settings, immediately sending SMS and email notifications to relevant personnel when high-risk issues are detected. Alarm information includes the problem location, risk level, and preliminary analysis results, supporting one-click navigation to the detailed analysis interface. The alarm handling process implements closed-loop management, with records and traceability of the entire process from problem discovery to resolution.

[0077] Example 5: The wireless transmission module employs multi-protocol converged communication technology. For instance, in a large-scale bridge structural health monitoring project, the system deployed a hybrid communication network. The main bridge area uses 5G base stations for high-speed data transmission, remote areas utilize LoRa technology for long-distance, low-power communication, and key nodes are enhanced with a Wi-Fi Mesh network to improve coverage. The choice of communication protocol is based on real-time network status assessment. When the 5G signal strength falls below a threshold, it automatically switches to the LoRa network while maintaining data transmission continuity. The synchronous transmission mechanism uses timestamp alignment technology. The basic dataset of component quality is first grouped and packaged, with each data packet having a sequence number and checksum added. The quality correlation partition map is converted to vector data format to reduce volume, and the environmental response analysis report uses an incremental update method, transmitting only the changed parts. During transmission, AES-256 encryption is used to ensure data security. The cloud-based digital twin platform is equipped with multiple access points, and load balancing technology is used to allocate data transmission paths, ensuring stable reception under high-concurrency scenarios.

[0078] The communication optimization module dynamically adjusts the transmission strategy based on the data characteristics in the component state change set. In the bridge monitoring example, when the sensor detects abnormal vibration, the data sampling frequency is increased from the usual 1Hz to 100Hz. At this time, the communication optimization module automatically activates the streaming transmission mode, using the UDP protocol to reduce transmission latency. The data compression strategy is intelligently selected according to the data type. Numerical monitoring data uses lossless compression algorithms to preserve accuracy, while image data uses lossy compression to balance quality and size. The network bandwidth monitoring unit continuously evaluates the available bandwidth. When the bandwidth is below the threshold, a data hierarchical transmission mechanism is activated, prioritizing the transmission of key parameters and alarm information. The communication protocol stack supports software-defined configuration and can dynamically adjust the MTU size and retransmission strategy based on indicators such as network latency and packet loss rate to optimize transmission efficiency.

[0079] Edge computing units are deployed at monitoring terminals throughout the bridge. Each terminal is equipped with an ARM architecture processor and a dedicated FPGA chip. Pre-built data acquisition algorithms are embedded in the hardware logic, enabling low-power real-time computation. Raw sensor data is first cached locally, then digital filtering and outlier removal are performed by the edge computing units. Preliminary quality analysis uses a lightweight machine learning model to generate a simplified basic dataset of component quality locally. The terminal devices are powered by a combination of solar panels and supercapacitors, supporting offline operation and allowing continuous operation for over 72 hours when the network is interrupted. Edge computing units form a distributed computing cluster via a self-organizing network protocol, enabling collaborative processing of cross-regional correlation analysis tasks. The cloud collaboration module runs on a distributed cloud computing platform. After receiving preprocessed data from multiple edge computing units, it launches containerized computing instances to perform deep analysis. The environmental response assessment algorithm uses multi-threaded parallel computing to simultaneously process temperature-load correlation analysis for hundreds of components. The defect localization module uses a GPU-accelerated convolutional neural network to identify minute damage features. After comparing the analysis results with historical data, a risk assessment report is generated and asynchronously pushed to the edge computing units via a message queue. After receiving the report, the edge device updates its local database and adjusts the monitoring parameters according to the preset strategy, forming a closed-loop optimization system.

[0080] In a subway tunnel monitoring project, the system installs numerous sensor nodes on the tunnel lining, with edge computing units embedded within the sensors. When a train passes, generating vibration data, the edge units immediately perform spectral analysis, extract characteristic frequency components, and compress and transmit the data. The cloud integrates data from the entire line, detecting an abnormal vibration pattern in a section of the tunnel lining and automatically triggering detailed detection commands. Upon receiving the commands, the edge units adjust their sampling strategy, increasing the surface strain monitoring frequency and uploading high-density data to the cloud in segments. The communication optimization module is adaptive, activating a data caching mechanism during daytime network congestion and batch transmitting complete datasets during nighttime off-peak hours. If weak mobile network signals are detected within the tunnel, the system initiates a store-and-forward mode, exchanging data using inspection vehicles. Multi-protocol fusion supports automatic 4G / 5G switching; 5G is used at the tunnel entrance for video data transmission, while leaky cable communication is used inside the tunnel to ensure continuity. Edge computing unit resources are dynamically allocated, with routine monitoring using low resources and anomalies automatically prioritized. Local storage uses a circular buffer, where new data overwrites old data while retaining key event records. The program supports remote deployment of new algorithms to optimize the analysis model.

[0081] The cloud-based collaborative module establishes a digital twin and physical mapping, with the tunnel's 3D model including geological, structural dimensions, and material parameters. Edge units report lining deformation data, and the cloud model updates stress distribution simulations and predicts trends in real time. Early warning information is pushed to maintenance personnel to guide inspections of risk areas. The system employs a multi-layered security mechanism: end-to-end encryption for data transmission, hardware security modules on edge devices storing keys, and multi-factor authentication in the cloud. Access control is based on role-based access management; on-site operators view their assigned data, while administrators have global configuration rights. Operation logs record data exchange and processing. The maintenance management platform provides visual monitoring, displaying edge node status and communication quality. The network topology diagram shows data transmission paths, and faulty nodes automatically trigger alarms. Remote diagnostic tools support online debugging of edge units and viewing of the processing pipeline status. System resource monitoring tracks CPU, memory, and storage usage, and predictive maintenance provides early warnings before resources are exhausted.

[0082] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0083] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A remote component quality control system based on digital twin and wireless communication, characterized in that, The system includes: The component data acquisition module acquires the geometric dimensions and material parameters of the component to be monitored through a wireless sensor network, collects the surface morphology features and internal stress distribution data of the component, associates the component identification code and calculates the corresponding quality evaluation index, and generates a basic dataset of component quality. The regional correlation analysis module extracts the quality evaluation indicators and corresponding spatial location information from the component quality basic dataset, establishes the correlation mapping of adjacent component units according to spatial topology, marks parameter matching and abnormal areas in adjacent component units, and forms a quality correlation partition map. The environmental response assessment module obtains component units in the parameter matching area of ​​the quality correlation zoning map, extracts temperature change curves and load fluctuation data, analyzes the change law in combination with material durability parameters, calculates the degree of influence of environmental factors on component performance, and generates an environmental response analysis report. The defect location module identifies problem points in component units whose impact exceeds a preset threshold and are located in abnormal areas in the environmental response analysis report, and constructs a component quality defect distribution map. The quality early warning module integrates all problem points and their associated parameters in the component quality defect distribution map, marks the location of components with potential safety hazards, and outputs a component quality assessment and risk early warning report; The environmental response assessment module includes: The environmental data extraction submodule selects the component units marked as parameter matching areas according to the quality correlation partition map, monitors the temperature gradient data and load change data of each component unit within the observation period, arranges them in time sequence to form temperature change curves and load fluctuation datasets, and constructs a set of environmental influencing factors. The performance impact calculation submodule calculates the temperature change rate and load change amplitude between consecutive time points in the monitoring data of each component unit based on the set of environmental impact factors. It performs correlation analysis on the temperature change rate and load change amplitude under the same material conditions, and quantitatively evaluates the effect intensity of the component performance by comprehensively analyzing the two types of environmental indicators. It integrates the impact coefficient sequence of each component unit and generates an environmental response analysis report.

2. The remote component quality control system based on digital twin and wireless communication according to claim 1, characterized in that, The component quality basic dataset includes quality evaluation indicators, component spatial coordinates, and standardized material parameters. The quality correlation zoning map specifically includes parameter matching area marking, abnormal area marking, and parameter difference degree between adjacent component units. The environmental response analysis report includes the influence coefficient of temperature change on materials, the influence coefficient of load fluctuation on structure, and a comparison of performance changes under different environmental conditions. The component quality defect distribution map includes the spatial coordinates of defect points, the temperature-load composite characteristics of defect points, and the material parameters and design deviation values ​​of defect points. The component quality assessment and risk warning report includes a list of quality problem locations and a comprehensive judgment result of multiple parameters for problem points.

3. The remote component quality control system based on digital twin and wireless communication according to claim 1, characterized in that, The component data acquisition module includes: The geometric measurement submodule acquires the three-dimensional coordinate data and acquisition time information of the component to be monitored through distributed sensors, collects the surface roughness and internal defect detection data at the corresponding locations, and records the acquisition results as two types of indicators: surface quality parameters and internal quality parameters, forming the original parameter set of the component. The parameter standardization submodule performs standardization processing on the surface quality parameters and internal quality parameters in the original parameter set of the component, establishes a correspondence between the processed results and the spatial position of the component, calculates the comprehensive evaluation value of the standardized surface parameters and standardized internal parameters as the quality evaluation index, and generates a basic dataset of component quality.

4. The remote component quality control system based on digital twin and wireless communication according to claim 3, characterized in that, The regional correlation analysis module includes: The indicator extraction submodule obtains the quality evaluation indicators and corresponding spatial location data from the component quality basic dataset, determines the relative positional relationship of all component units based on spatial relationships, calls the component location information set, calculates and sorts the proximity relationships of component units in space based on the adjacent distance standard, and forms a component unit proximity relationship sequence. The difference calculation submodule calculates the difference in quality evaluation indicators between each pair of adjacent component units based on the proximity relationship sequence of the component units, and integrates them to generate a quality difference sequence. The spatial consistency determination submodule extracts the surface parameter change trend and internal parameter change trend between adjacent component units based on the quality difference sequence, classifies and marks each pair of component units according to the consistency of the two parameter change directions, records and groups the parameter matching and parameter abnormality segments respectively, and forms a quality correlation partition map.

5. The remote component quality control system based on digital twin and wireless communication according to claim 1, characterized in that, The defect location module includes: The monitoring data extraction submodule, based on the environmental response analysis report, filters out component units whose impact exceeds a preset threshold and component units located in abnormal areas, extracts continuous monitoring records of component units in chronological order, and collects extreme temperature data and load peak data at each time point to form a component status monitoring dataset. The parameter deviation calculation submodule calls the component condition monitoring dataset to extract the temperature change and load change of the component unit at two consecutive monitoring nodes, integrates them into temperature change sequence and load change sequence, and establishes a component condition change set. The problem point identification submodule extracts environmental humidity data and vibration frequency data for the corresponding time period based on the component state change set, determines whether the temperature change and load change both exceed the set abnormal state standard, determines whether the humidity fluctuation and vibration intensity simultaneously meet the defect judgment condition, marks the monitoring points that meet the conditions as problem points, and constructs a component quality defect distribution map.

6. The remote component quality control system based on digital twin and wireless communication according to claim 5, characterized in that, The quality early warning module includes: The multi-parameter comprehensive judgment submodule obtains all problem points and their associated parameter information in the component quality defect distribution map, calculates the comprehensive risk assessment value of each problem point, and forms a comprehensive risk assessment sequence. The risk classification output submodule, based on the comprehensive risk assessment sequence, filters out points whose impact exceeds material safety standards, whose quality evaluation indicators are lower than design requirements, and whose spatial consistency is marked as abnormal areas. It extracts the corresponding component number, location information, and region, marks them as quality abnormalities and potential safety hazards, and classifies and outputs the points that meet the comprehensive conditions according to the hierarchical structure of the digital twin model, generating a component quality assessment and risk warning report.

7. The remote component quality control system based on digital twin and wireless communication according to claim 6, characterized in that, The system also includes: The digital twin construction module receives the component quality basic dataset and component quality assessment and risk warning report in real time, and establishes a three-dimensional digital twin model that includes material properties, geometric features and defect distribution; The model update module updates the material parameters of the digital twin model in real time based on the dynamic change information in the environmental response analysis report. When a new record is detected in the component quality defect distribution map, the defect spatial distribution of the digital twin model is updated synchronously.

8. The remote component quality control system based on digital twin and wireless communication according to claim 7, characterized in that, The system also includes: The wireless transmission module uses multi-protocol converged communication technology to synchronously transmit the component quality basic dataset, quality correlation zoning map and environmental response analysis report to the cloud digital twin platform; The communication optimization module dynamically adjusts the wireless transmission communication protocol and data compression strategy based on the data characteristics in the set of component state changes.

9. The remote component quality control system based on digital twin and wireless communication according to claim 8, characterized in that, The system also includes: The edge computing unit is deployed on the component monitoring terminal, and the computing logic of the component data acquisition module and the regional correlation analysis module is pre-set to perform localization processing and preliminary quality analysis on the raw monitoring data; The cloud collaboration module receives the preprocessed component quality basic dataset from the edge computing unit, performs deep calculations by the environmental response assessment module and the defect location module, and feeds back the component quality assessment and risk warning report to the edge computing unit for local storage.

Images