A full life cycle digital twin management system and modeling method

Abstract

The application relates to the technical field of digital twin management, and discloses a full-life-cycle digital twin management system and a modeling method, which comprise a data fusion unit, a model driving unit, a decision optimization unit and a trusted storage unit. The modeling method comprises the following steps: step S1, analyzing a product design bill of materials; step S2, establishing a topological mapping relationship between a sensor node and a virtual model area; step S3, running a Monte Carlo simulation to predict a defect probability; and step S4, binding a new virtual model version with an entity product serial number and chaining. According to the problem of data dispersion in the prior art, the system automatically imports three types of heterogeneous data, namely design geometric dimensions, manufacturing conditions and maintenance monitoring, into a shared cache pool through a main data table of a unified metadata structure, realizes cross-stage semantic alignment and time sequence synchronization, eliminates data format differences, and significantly improves data consistency.

Description

Technical Field

[0001] This invention relates to the field of digital twin management technology, specifically to a full lifecycle digital twin management system and modeling method. Background Technology

[0002] With the advancement of intelligent manufacturing, digital twin technology, as a core tool for interaction between physical entities and virtual spaces, has been initially applied in product design, manufacturing, and equipment operation and maintenance. Its core is to construct a dynamic virtual mapping of physical objects through IoT sensors, 3D modeling, and real-time data analysis to achieve status monitoring and decision optimization.

[0003] A search revealed Chinese patent CN119624082A, which discloses a product management system and method for the entire product lifecycle, relating to the field of product management technology. The system includes: dividing the product lifecycle and clarifying the product data management content for each stage; using AI technology to perform preliminary analysis of cached product data; and recursively retrieving product data for each child node based on the product's root node number and version number combined with the preliminary analysis report. This invention, by dividing the product lifecycle, clarifying the data management content for each stage, and introducing AI and digital twin technologies, achieves automated data collection, anomaly detection, and simulation modeling of product data.

[0004] Although the aforementioned patents divide the lifecycle into design, prototyping, procurement, maintenance, and mass production stages, the data for each stage is still scattered across independent systems, such as PDM, ERP, and MES systems. This dispersion leads to a lack of data standards, heterogeneous data formats across different systems, and a lack of unified metadata specifications, hindering cross-stage data integration. Furthermore, design changes cannot be synchronized to the manufacturing or maintenance stages in real time and require manual transmission, resulting in version mismatches. For example, there is the problem of using outdated design drawings during the prototyping stage. Based on this, the present invention designs a full lifecycle digital twin management system and modeling method to solve the above problems. Summary of the Invention

[0005] The purpose of this invention is to provide a full lifecycle digital twin management system and modeling method, which solves the problem of severe system dispersion in the background technology.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] A full lifecycle digital twin management system and modeling method, comprising:

[0008] The data fusion unit is configured as follows: based on a unified metadata structure, the master data table integrates geometric dimensions and material properties data from the product design stage, equipment operating condition data from the manufacturing stage, and sensor monitoring data from the maintenance stage, and establishes a shared cache pool for multi-source heterogeneous data.

[0009] The model-driven unit, which is communicatively connected to the data fusion unit, is configured to: perform dynamic incremental updates on the parameterized virtual model based on the cross-stage data stream in the shared cache pool, with the update trigger condition being that the real-time physical quantity deviation exceeds a set threshold m.

[0010] The decision optimization unit, which is communicatively connected to the model driving unit, is configured to: generate optimization instructions containing a process parameter adjustment range of ±5% and work order instructions containing a maintenance item list based on the updated virtual model, and send them to the execution terminal.

[0011] The trusted evidence storage unit, which communicates with each unit, is configured to bind and store the unique serial number of the physical product with the version number of the virtual model in three stages: when the design is completed, when the manufacturing process is changed, and when the maintenance event ends, forming an immutable twin entity association chain.

[0012] Preferably, the data fusion unit includes:

[0013] Design a data interface for extracting design master data from the product data management system;

[0014] The manufacturing data interface collects real-time current, pressure, and displacement data from the manufacturing execution system via industrial communication protocols.

[0015] Maintain the data interface to receive data streams from vibration and temperature sensors deployed on the device itself.

[0016] Preferably, the model driving unit comprises:

[0017] A multiphysics coupling engine is used to simultaneously perform structural stress field and temperature field simulation calculations;

[0018] The deviation response module, when the real-time sensor measurement value y and the virtual model output value y0 satisfy:

[0019] At that time, the model state correction is activated.

[0020] Preferably, the state correction employs a Kalman filter algorithm to update the model parameters according to the state equation:

[0021] ;

[0022] x is the updated virtual model state vector, which includes key variables such as position, stress, and temperature.

[0023] x - The prior state estimate before the update;

[0024] K is the Kalman gain matrix, used to dynamically adjust the observation weights;

[0025] y represents the physical quantity measurement value acquired by the sensor in real time;

[0026] H is the observation matrix, used to map the state vector to the sensor measurement space;

[0027] Each revision generates a new version identifier with a timestamp and marks the incremental change point.

[0028] Preferably, the decision optimization unit includes:

[0029] The lifespan prediction module trains a time series model based on historical equipment operating data and outputs the remaining lifespan (RUL) and a 90%-95% confidence interval.

[0030] The instruction decision module generates a load rate adjustment instruction of ±5% when the lower limit of the RUL confidence interval is lower than the preset safety threshold n.

[0031] Preferably, the optimization instructions include:

[0032] Machining parameter commands, spindle speed adjustment range;

[0033] Maintenance work order instructions, preventive maintenance item list and execution time window.

[0034] Preferably, the trusted evidence storage unit performs an initial binding operation at the end of the design phase, associating and storing the following data:

[0035] Unique product serial number;

[0036] Initial virtual model version number;

[0037] Design and review digital signatures;

[0038] Bind operation timestamp;

[0039] Phase end marker code.

[0040] Preferably, the trusted evidence storage unit adopts a dual-blockchain architecture:

[0041] Model version chain: Stores full model snapshots and differences between adjacent versions, with a storage interval of ≤24 hours;

[0042] Product event chain: Stores records of changes to manufacturing process parameters and hash values ​​of maintenance work order execution results.

[0043] According to a second aspect of the present invention, a modeling method for full lifecycle digital twin management is also proposed, the modeling method comprising the following steps:

[0044] Step S1: Analyze the product design bill of materials and extract ≥8 parameters including geometric dimensions, tolerances, and material properties to construct the master data table;

[0045] Step S2: Establish the topological mapping relationship between sensor nodes and virtual model regions;

[0046] Step S3: Run Monte Carlo simulation to predict the defect probability P. When P exceeds the preset defect safety threshold λ, trigger the parameter optimization algorithm to generate adjustment instructions.

[0047] Step S4: Within 24 hours of the manufacturing process change event, bind the new virtual model version to the physical product serial number on the blockchain.

[0048] Preferably, the execution process of the parameter optimization algorithm in step S3 includes:

[0049] a) Construct a response surface model of process parameters and defect probabilities:

[0050] ;

[0051] Where S is the spindle speed, F is the feed rate, and C is the depth of cut;

[0052] b) Solving the objective function based on gradient descent:

[0053] Where S is the spindle speed, F is the feed rate, C is the depth of cut, λ is the preset defect safety threshold, and α is the weighting coefficient;

[0054] Constraints: S min Minimum spindle speed, S max The maximum spindle speed is given by C, ΔF is the feed rate change, and C is the maximum spindle speed. 安全 To ensure a safe cutting depth.

[0055] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0056] 1. In view of the problem of data dispersion in the prior art, this system automatically merges three types of heterogeneous data, namely design geometry, manufacturing conditions and maintenance monitoring, into a shared cache pool through a master data table with a unified metadata structure. This achieves cross-stage semantic alignment and time synchronization, eliminates data format differences and significantly improves data consistency.

[0057] 2. In response to the problem that "design changes cannot be synchronized in real time and require manual transmission, which leads to version misalignment", the model-driven unit uses real-time sensor deviation as the trigger condition to perform millisecond-level incremental updates on the virtual model using Kalman filtering and automatically generates a version identifier with a timestamp. The design, manufacturing and maintenance parties can work together on the same model version, eliminating the risk of using outdated drawings during the trial production stage.

[0058] 3. In view of the problem that the lack of unified metadata standards hinders cross-stage integration, this invention adopts a trusted evidence storage unit with a dual blockchain architecture. The model version chain stores full model snapshots and differences, while the product event chain stores process changes and maintenance result hashes. The two are bound by the entity serial number and the virtual version number to form an immutable twin entity association chain, realizing full lifecycle data traceability and auditability. Attached Figure Description

[0059] Figure 1 This is a diagram illustrating the contents of the full lifecycle digital twin management system of the present invention.

[0060] Figure 2 This is a diagram of the data fusion unit of the present invention;

[0061] Figure 3 This is a flowchart of the modeling method of the present invention;

[0062] Figure 4 This is a flowchart of the trusted evidence storage chain mechanism of the present invention. Detailed Implementation

[0063] 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.

[0064] Example 1;

[0065] Please see Figures 1-4In this embodiment of the invention, a full lifecycle digital twin management system includes: a data fusion unit, configured to: integrate geometric dimensions and material properties data from the product design stage, equipment operating condition data from the manufacturing stage, and sensor monitoring data from the maintenance stage based on a master data table with a unified metadata structure, and establish a shared cache pool for multi-source heterogeneous data; a model-driven unit, communicatively connected to the data fusion unit, configured to: perform dynamic incremental updates on the parameterized virtual model based on the cross-stage data flow in the shared cache pool, with the update trigger condition being that the real-time physical quantity deviation exceeds a set threshold m; a decision optimization unit, communicatively connected to the model-driven unit, configured to: generate optimization instructions containing a process parameter adjustment range of ±5% and work order instructions containing a maintenance item list based on the updated virtual model, and issue them to the execution terminal; and a trusted evidence storage unit, communicatively connected to each unit, configured to: bind and store the unique serial number of the physical product with the version number of the virtual model in three stages: when the design is completed, when the manufacturing process is changed, and when the maintenance event ends, forming an immutable twin entity association chain.

[0066] The data fusion unit includes: a design data interface for extracting master design data from the product data management system; a manufacturing data interface for acquiring real-time current, pressure, and displacement data from the manufacturing execution system via industrial communication protocols; and a maintenance data interface for receiving data streams from vibration and temperature sensors deployed on the equipment body.

[0067] The model-driven unit includes: a multiphysics coupling engine for synchronously performing structural stress and temperature field simulation calculations; and a deviation response module, which responds when the real-time sensor measurement value y and the virtual model output value y0 satisfy the following conditions:

[0068] At that time, the model state correction is activated.

[0069] State correction employs the Kalman filter algorithm to update model parameters according to the state equations:

[0070] ;

[0071] x is the updated virtual model state vector, which includes key variables such as position, stress, and temperature.

[0072] x - The prior state estimate before the update;

[0073] K is the Kalman gain matrix, used to dynamically adjust the observation weights;

[0074] y represents the physical quantity measurement value acquired by the sensor in real time;

[0075] H is the observation matrix, used to map the state vector to the sensor measurement space;

[0076] Each revision generates a new version identifier with a timestamp and marks the incremental change point.

[0077] The decision optimization unit includes: a lifespan prediction module, which trains a time series model based on the equipment's historical operating data and outputs the remaining lifespan (RUL) and a 90%-95% confidence interval; and an instruction decision module, which generates a load rate adjustment instruction of ±5% when the lower limit of the RUL confidence interval is lower than the preset safety threshold n.

[0078] The optimization instructions include: machining parameter instructions, spindle speed adjustment range; maintenance work order instructions, preventive maintenance item list and execution time window.

[0079] The trusted evidence storage unit performs its first binding operation at the end of the design phase, linking the following data for evidence storage: product unique serial number; initial virtual model version number; design review digital signature; binding operation timestamp; phase end marker code.

[0080] The trusted evidence storage unit adopts a dual blockchain architecture: Model version chain: stores full model snapshots and adjacent version difference data, with an evidence storage interval of ≤24 hours; Product event chain: stores manufacturing process parameter change records and maintenance work order execution result hash values.

[0081] The working principle of this invention is as follows: the operating mechanism is based on the closed loop of "data-model-decision-evidence storage". When the system is started, the data fusion unit first performs semantic alignment and time synchronization on the multi-source heterogeneous data such as three-dimensional geometric dimensions, tolerance chains, material stress-strain curves, and heat treatment parameters in the product design stage, CNC spindle current, hydraulic system pressure, and servo motor displacement in the manufacturing stage, and accelerometers, infrared thermocouples, and acoustic emission sensors in the maintenance stage, based on the master data table with a unified metadata structure, and builds a shared cache pool with a version control mechanism.

[0082] After receiving the cross-stage data stream pushed by the data fusion unit, the model-driven unit activates the multiphysics coupling engine to simultaneously solve the structural stress field and transient temperature field based on the finite volume method (FVM), forming a high-fidelity parametric virtual model. This model incorporates a Kalman filter-based state estimator. Its observation equations use the Jacobian matrix H to map the state vectors (nodal displacements, von Mises stress, and temperature gradient) to the sensor measurement space. When the Euclidean distance between the real-time sensor measurement value y and the model output value y0 exceeds a set threshold m (default 0.05 times the measurement range), an incremental update mechanism is triggered. The update process dynamically calculates the Kalman gain matrix K by solving the Riccati equation, achieving a Bayesian optimal estimate of the model parameters.

[0083] The decision optimization unit, based on the updated virtual model, calls the bidirectional long short-term memory (Bi-LSTM) time series model trained in the life prediction module. Inputs include 12-dimensional features such as vibration spectral entropy, temperature cycle amplitude, and load power spectral density. Outputs the remaining service life (RUL) and its 90%-95% confidence interval. When the lower limit of the RUL confidence interval is lower than a preset safety threshold n (e.g., 100 hours), the instruction decision module uses a reinforcement learning-based policy gradient algorithm to search for the optimal action within the feasible region of process parameters. This generates optimized instructions with adjustment ranges of ±5% for spindle speed, ±3% for feed rate, and ±2% for depth of cut, as well as preventative maintenance work orders including bearing lubrication, tool replacement, and thermal balance calibration. These work orders are then sent to the CNC controller and CMMS system for execution via the MQTT protocol.

[0084] The trusted evidence storage unit employs a dual-blockchain architecture to achieve immutable evidence storage of the twin entity association chain. The model version chain is built on Hyperledger Fabric, storing full model snapshots and adjacent version difference data, and uses the IPFS distributed file system to store large-capacity geometric models, with an evidence storage interval of ≤24 hours. The product event chain is built on an Ethereum private chain, storing manufacturing process parameter change records, Keccak-256 hash values ​​and digital signatures of maintenance work order execution results. The initial binding operation is executed at the end of the design phase, and the unique product serial number (GTIN+ serial number based on the GS1 standard), the initial virtual model version number, the design review digital signature, the binding operation timestamp (ISO8601 format, accurate to milliseconds), and the phase end marker code (UUIDv4) are associated and stored through a smart contract. Subsequently, at the end of manufacturing process changes and maintenance events, on-chain updates are automatically triggered through an event-driven mechanism.

[0085] Example 2;

[0086] Please see Figures 1-4 In this embodiment of the invention, the modeling method includes the following steps: Step S1, parse the product design bill of materials, extract ≥8 parameters including geometric dimensions, tolerances, and material properties to construct a master data table; Step S2, establish a topological mapping relationship between sensor nodes and virtual model regions; Step S3, run Monte Carlo simulation to predict the defect probability P, and when P exceeds the preset defect safety threshold λ, trigger the parameter optimization algorithm to generate adjustment instructions; Step S4, within 24 hours after a manufacturing process change event occurs, bind the new virtual model version to the physical product serial number on the blockchain.

[0087] The execution process of the parameter optimization algorithm in step S3 includes:

[0088] a) Construct a response surface model of process parameters and defect probabilities:

[0089] ;

[0090] Where S is the spindle speed, F is the feed rate, and C is the depth of cut;

[0091] b) Solving the objective function based on gradient descent:

[0092] Where S is the spindle speed, F is the feed rate, C is the depth of cut, λ is the preset defect safety threshold, and α is the weighting coefficient;

[0093] Constraints: S min Minimum spindle speed, S max The maximum spindle speed is given by C, ΔF is the feed rate change, and C is the maximum spindle speed. 安全 To ensure a safe cutting depth.

[0094] The working principle of this invention is as follows: the operating mechanism is based on a four-step closed loop of "data parsing - topology mapping - defect prediction - version binding". After the system starts, step S1 is executed first, which parses the product design bill of materials (BOM) through the CAD / CAE software API interface and extracts no less than 8 key parameters, including geometric dimensions (such as aperture, wall thickness, and profile), dimensional tolerances (such as ISO 2768-m grade), and material properties (such as elastic modulus, Poisson's ratio, coefficient of thermal expansion, yield strength, hardness, density, thermal conductivity, and specific heat capacity), to ensure the machine readability and scalability of the parameters.

[0095] In step S2, the system discretizes the virtual model into finite element elements using a 3D meshing algorithm (such as Delaunay tetrahedral partitioning) and establishes a topological mapping relationship between sensor nodes and model regions based on the sensor deployment drawings. The mapping process uses a KD-Tree spatial index structure to accelerate the query, associating each sensor node (such as a triaxial accelerometer or infrared thermocouple) with the nearest finite element node and recording the mapping weight matrix for spatial interpolation calculations during subsequent data assimilation.

[0096] Step S3 starts the Monte Carlo simulation engine and generates a random sample set of process parameters (spindle speed S, feed rate F, depth of cut C) based on Latin hypercube sampling (LHS). The sample size is ≥10,000 sets, and the defect probability P is statistically analyzed (defined as the volume percentage of regions with stress concentration factor >3). When P exceeds the preset defect safety threshold λ (e.g., 0.1%), the parameter optimization algorithm is triggered. This algorithm first constructs a process parameter-defect probability response surface model based on radial basis functions (RBF), whose kernel function adopts a multi-quadratic surface form. The optimal shape parameters are determined through cross-validation. Then, the constrained gradient descent method is used to solve the objective function, which is defined as a weighted combination of defect probability P and process time T (weighting coefficients α=0.7, β=0.3). The constraints include the upper limit of machine tool power, the lower limit of tool life, and the upper limit of surface roughness.

[0097] Step S4: Within 24 hours of a manufacturing process change event, the system automatically invokes a smart contract to bind the new virtual model version (including geometric updates, material parameter corrections, and boundary condition adjustments) to the physical product serial number and upload it to the blockchain. This is done through the Ethereum private blockchain's PoA consensus mechanism, ensuring an immutable association between the model version and the physical state. Simultaneously, it triggers the incremental synchronization mechanism of the model version chain, completing the closed-loop update of the digital twin model throughout its entire lifecycle.

[0098] Example 3;

[0099] Please see Figures 1-4 This provides a specific implementation example of full lifecycle digital twin management for high-pressure compressor blades of aero engines.

[0100] The data fusion unit establishes a master data table with a unified metadata structure during system initialization. Fields include: blade section chord length 82.4 mm, maximum thickness 3.2 mm, leading edge radius 0.8 mm, material TC4 titanium alloy, elastic modulus 114 GPa, Poisson's ratio 0.342, and density 4.43 g / cm³. -3 Yield strength 880 MPa. During manufacturing, real-time data from the five-axis machining center was acquired via OPC UA protocol at a sampling rate of 250 Hz: spindle current 45.7A, hydraulic pressure 7.3 MPa, X-axis displacement -12.4 μm. During maintenance, vibration signals were transmitted back by eight triaxial MEMS accelerometers arranged on the blade surface at a sampling rate of 1 kHz, with a peak-to-peak value of 12.3 ms. -2 Meanwhile, the surface temperature was measured to be 412℃ by an infrared thermocouple.

[0101] The model-driven unit uses a tetrahedral mesh with 1,874,652 elements and a unit size of 0.15 mm. The coupled solver has a time step of 0.1 ms. The Kalman filter observation matrix H has 8×12 dimensions and maps the state vector to the acceleration and temperature measurement points. The deviation threshold m is set to 5% of the measurement range, meaning that if the difference between the measurement value of any channel sensor and the model output exceeds 0.615 ms... -2 Incremental updates may be triggered at 20.6℃; the updated state vector x includes a tip displacement of 0.34 mm, a maximum von Mises stress of 615 MPa, and a leading-edge temperature gradient of 38 ℃ mm. -1 .

[0102] The decision optimization unit calls a Bi-LSTM model with a training set containing 3.2 TB of running data from the past 36 months. The input feature dimension is 12, the hidden layer has 256 units, the learning rate is 0.001, and the batch size is 512. The model outputs the current blade remaining service life (RUL) as 217 hours, with a 90% confidence interval of 196 hours–238 hours. The safety threshold n is set to 200 hours. Since the lower limit of 196 hours is already below the threshold, the instruction decision module generates a process parameter adjustment instruction: the spindle speed is adjusted from 8000 rpm. -1 Reduced to 7,600 r / min -1 The feed rate was increased from 0.25 mm rev. -1 Reduced to 0.238 mm rev -1 The cutting depth was reduced from 0.8mm to 0.76mm. Simultaneously, a maintenance work order was generated: to complete three tasks before the next shift: fluorescent penetrant testing of the blade surface, ultrasonic flaw detection of the blade root tenon groove, and random sampling inspection of the coating porosity.

[0103] The trusted evidence storage unit performs its first binding at the end of the design phase. The product's unique serial number is recorded as GTIN095060001234567890, serial number S / N AE2025-0718-0001, initial virtual model version number v1.0.0, design review digital signature adopts ECDSA P-256 algorithm, signature value 3046022100… timestamp 2025-07-18T09:17:42.123Z, phase end marker code UUID 3ce3f8b0-… Manufacturing process change occurred at 2025-07-18T14:32:05Z, change record includes spindle speed adjustment -5%, feed rate adjustment -5%, and depth of cut adjustment -5%. When the maintenance event ends, the binding is re-established, and the execution results are recorded as follows: no cracks detected, coating porosity 2.1%, RUL updated to 278 h, version number upgraded to v1.1.3, binding timestamp 2025-07-18T22:45:10Z, completing the full life cycle closed loop.

[0104] Working Principle: This patent centers on a closed loop of "data fusion - model-driven - decision optimization - trusted evidence storage," spanning the entire product lifecycle. First, the data fusion unit aggregates key data from the design, manufacturing, and maintenance stages into a shared cache pool using a unified metadata structure. Then, the model-driven unit utilizes a multiphysics coupling engine and Kalman filter to dynamically update the model based on the deviation between real-time sensor readings and the virtual model, generating a new version with a timestamp. The decision optimization unit calculates the remaining lifespan based on the updated model. When the lower limit of the lifespan confidence interval falls below a set threshold, it automatically generates a ±5% adjustment instruction for process parameters and a preventative maintenance work order, which are then sent to the execution terminal. The trusted evidence storage unit binds the physical product serial number and the virtual model version number to a dual blockchain at three nodes: design completion, process change, and maintenance completion, forming an immutable twin entity association chain. The entire system achieves real-time data synchronization across stages, adaptive model evolution, closed-loop decision optimization, and full-process traceability.

[0105] 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, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A full lifecycle digital twin management system, characterized in that, include: The data fusion unit is configured as follows: based on a unified metadata structure, the master data table integrates geometric dimensions and material properties data from the product design stage, equipment operating condition data from the manufacturing stage, and sensor monitoring data from the maintenance stage, and establishes a shared cache pool for multi-source heterogeneous data. The model-driven unit, which is communicatively connected to the data fusion unit, is configured to: perform dynamic incremental updates on the parameterized virtual model based on the cross-stage data stream in the shared cache pool, with the update trigger condition being that the real-time physical quantity deviation exceeds a set threshold m. The decision optimization unit, which is communicatively connected to the model driving unit, is configured to: generate optimization instructions containing a process parameter adjustment range of ±5% and work order instructions containing a maintenance item list based on the updated virtual model, and send them to the execution terminal. The trusted evidence storage unit, which communicates with each unit, is configured to bind and store the unique serial number of the physical product with the version number of the virtual model in three stages: when the design is completed, when the manufacturing process is changed, and when the maintenance event ends, forming an immutable twin entity association chain.

2. The full lifecycle digital twin management system according to claim 1, characterized in that, The data fusion unit includes: Design a data interface for extracting design master data from the product data management system; The manufacturing data interface collects real-time current, pressure, and displacement data from the manufacturing execution system via industrial communication protocols. Maintain the data interface to receive data streams from vibration and temperature sensors deployed on the device itself.

3. The full lifecycle digital twin management system according to claim 1, characterized in that, The model-driving unit includes: A multiphysics coupling engine is used to simultaneously perform structural stress field and temperature field simulation calculations; The deviation response module, when the real-time sensor measurement value y and the virtual model output value y0 satisfy... At that time, the model state correction is activated.

4. The full lifecycle digital twin management system according to claim 3, characterized in that, The state correction updates the model parameters according to the state equation: ； x is the updated virtual model state vector, which includes key variables such as position, stress, and temperature. x - The prior state estimate before the update; K is the Kalman gain matrix, used to dynamically adjust the observation weights; y represents the physical quantity measurement value acquired by the sensor in real time; H is the observation matrix, used to map the state vector to the sensor measurement space; Each revision generates a new version identifier with a timestamp and marks the incremental change point.

5. A full lifecycle digital twin management system according to claim 1, characterized in that, The decision optimization unit includes: The lifespan prediction module trains a time series model based on historical equipment operating data and outputs the remaining lifespan (RUL) and a 90%-95% confidence interval. The instruction decision module generates a load rate adjustment instruction of ±5% when the lower limit of the RUL confidence interval is lower than the preset safety threshold n.

6. The full lifecycle digital twin management system according to claim 1, characterized in that, The optimization instructions include: Machining parameter commands, spindle speed adjustment range; Maintenance work order instructions, preventive maintenance item list and execution time window.

7. The full lifecycle digital twin management system according to claim 1, characterized in that, The trusted evidence storage unit performs its first binding operation at the end of the design phase, associating and storing the following data: Unique product serial number; Initial virtual model version number; Design and review digital signatures; Bind operation timestamp; Phase end marker code.

8. A full lifecycle digital twin management system according to claim 1, characterized in that, The trusted evidence storage unit adopts a dual-blockchain architecture: Model version chain: Stores full model snapshots and differences between adjacent versions, with a storage interval of ≤24 hours; Product event chain: Stores records of changes to manufacturing process parameters and hash values ​​of maintenance work order execution results.

9. A modeling method for full lifecycle digital twin management, implemented using the full lifecycle digital twin management system as described in any one of claims 1-8, characterized in that, The modeling method includes the following steps: Step S1: Analyze the product design bill of materials, extract geometric dimensions, tolerances, and material property parameters to construct the master data table; Step S2: Establish the topological mapping relationship between sensor nodes and virtual model regions; Step S3: Run Monte Carlo simulation to predict the defect probability P. When P exceeds the preset defect safety threshold λ, trigger the parameter optimization algorithm to generate adjustment instructions. Step S4: Within 24 hours of the manufacturing process change event, bind the new virtual model version to the physical product serial number on the blockchain.

10. The modeling method for full lifecycle digital twin management according to claim 9, characterized in that: The execution process of the parameter optimization algorithm in step S3 includes: a) Construct a response surface model of process parameters and defect probabilities: ； Where S is the spindle speed, F is the feed rate, and C is the depth of cut; b) Solving the objective function based on gradient descent: ； Constraints: Where λ is the preset defect safety threshold, α is the weighting coefficient, and S min Minimum spindle speed, S max The maximum spindle speed is given by C, ΔF is the feed rate change, and C is the maximum spindle speed. 安全 To ensure a safe cutting depth.

Images