Stamping equipment construction method based on digital twinning

Abstract

The invention discloses a stamping equipment construction method based on digital twinning, and the method comprises the steps: firstly constructing a digital twinning system of stamping equipment, which comprises a physical layer, a digital layer, a virtual layer and an application layer; then, three-dimensional modeling and multi-physical-quantity simulation of the stamping equipment are carried out on a physical layer, and the real form of the equipment is reflected; real-time data acquisition and processing of the stamping equipment are carried out on the digital layer, a digital twin engine is constructed, virtual-real combination and real-time linkage processing are carried out on the virtual layer, and data analysis and optimization of the stamping equipment are completed; finally, man-machine interaction and visualization processing are carried out on an application layer, user interactivity and the intelligent level of equipment are improved, and intelligent decision support is carried out through data integration and dynamic adjustment. According to the stamping equipment construction method, high efficiency and intelligence of stamping equipment construction are achieved, the design period is remarkably shortened, the production efficiency and the equipment performance are improved, and meanwhile the cost and the operation and maintenance risk are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of digital twins and stamping equipment, and in particular to a method for constructing stamping equipment based on digital twins. Background Art

[0002] Traditional methods for constructing stamping equipment rely primarily on repeated testing and debugging of physical prototypes. During this process, designers must repeatedly manufacture and test physical prototypes to verify equipment performance, optimize process parameters, and resolve potential issues. This approach not only consumes significant material and labor costs, but also results in long prototype manufacturing and testing cycles, leading to low overall development efficiency and difficulty meeting the modern manufacturing industry's demand for rapid response to market demands. Furthermore, the debugging process of physical prototypes is often limited by the actual production environment, making it difficult to fully simulate complex operating conditions. This can lead to unforeseen failures or performance deviations during actual operation.

[0003] With the rapid development of information technology, digital twin technology has gradually become one of the key technologies for the intelligent transformation of the manufacturing industry. By integrating multidisciplinary modeling, real-time data interaction, and simulation analysis, digital twins can build high-fidelity virtual dynamic models for physical objects, achieving a deep integration of physical space and cyberspace. In the field of stamping equipment, digital twin technology can accurately simulate the operating status of the equipment, the production process, and its interaction with the environment, providing strong support for the optimized design and efficient operation and maintenance of the equipment. However, there are still problems such as insufficient model accuracy, limited real-time performance, and weak coordination. Some simulation models fail to fully consider actual factors such as material nonlinearity and mold wear, resulting in deviations between virtual and physical entities. At the same time, the synchronous fusion and efficient processing capabilities of multi-source heterogeneous data (such as mechanical signals and temperature fields) need to be improved. Twin data in the design, production, and operation and maintenance links have not yet formed a closed-loop feedback loop, which restricts the realization of full life cycle optimization.

[0004] Therefore, there is an urgent need for a stamping equipment construction method based on digital twins. Through high-precision modeling, real-time interaction and intelligent decision-making, the limitations of traditional methods can be solved, and the stamping equipment can be further developed towards intelligence and efficiency. Summary of the Invention

[0005] The problem to be solved by the present invention is to provide a method for constructing stamping equipment based on digital twins, use digital twin technology to create a virtual model of the stamping equipment, achieve accurate simulation and verification in the design stage, and then optimize the design scheme.

[0006] The present invention adopts the following technical solution: a method for constructing a stamping equipment based on digital twin, comprising the following steps:

[0007] Step 1: Build a digital twin system for stamping equipment, including the physical layer, digital layer, virtual layer, and application layer.

[0008] Step 2: Perform three-dimensional modeling and multi-physical quantity simulation of stamping equipment at the physical layer;

[0009] Step 3: Collect and process real-time data of stamping equipment at the digital layer and build a digital twin engine;

[0010] Step 4: Perform virtual-real integration and real-time linkage processing in the virtual layer to complete stamping equipment data analysis and optimization;

[0011] Step 5: Perform human-computer interaction and visualization processing at the application layer, and provide intelligent decision support through data integration and dynamic adjustment.

[0012] Preferably, in step 2, the three-dimensional modeling of the stamping equipment is performed by parametric modeling using three-dimensional modeling software, combined with reverse engineering, to build a 1:1 true restoration three-dimensional model based on the stamping equipment data, which is consistent with the physical stamping equipment in terms of geometric dimensions and material properties;

[0013] 3D modeling software, including: SolidWorks, CATIA, AutoCAD Inventor, NX;

[0014] Stamping equipment data, including: geometric data, material properties, and process parameters.

[0015] Preferably, in step 2, multi-physical quantity simulation is performed, integrating multiple physical quantities on the basis of the three-dimensional model, including friction, gravity, and resistance, to perform simulation, restore the operating state of the stamping equipment under different working conditions, analyze the root cause of the abnormal state, determine that the abnormality is an imbalance in the coupling of multiple physical quantities or external interference exceeds the design tolerance, predict the abnormality risk in advance and perform parameter optimization;

[0016] Specific working conditions include: process parameter conditions, material property conditions, mold and equipment conditions, environmental conditions, special scenario conditions, and multi-physics field coupling conditions.

[0017] Preferably, in step 3, data collection and processing is carried out by using the industrial Internet and radio frequency identification methods to collect physical layer data in real time, and perform cleaning, correlation and mining processes to form the data foundation of the digital twin system;

[0018] Physical layer data, including: equipment operating status data, mold and workpiece interaction data, environment and auxiliary system data, engineering control and abnormal signal data;

[0019] The digital twin engine is used to connect the physical layer and the virtual layer to transmit real-time data, dynamically update the three-dimensional model, and drive the virtual-reality interaction in the virtual layer.

[0020] Preferably, in step 4, the virtual layer, the virtual-reality interaction includes virtual-reality mapping and real-time linkage, and the virtual-reality mapping of the physical space and the information space of the stamping equipment is performed through the digital twin method to achieve the coordination of the physical space and the information space, including the following sub-steps:

[0021] Step 4.1, coordinate system construction and dynamic conversion:

[0022] Laser trackers are deployed at key motion nodes of the physical equipment to establish a world coordinate system. In the virtual 3D model, a local coordinate system is defined with the mold center as the origin, and coordinate mapping is performed using a homogeneous transformation matrix. Kalman filtering is used to fuse multi-sensor data to correct coordinate system drift caused by mechanical vibration or thermal deformation in real time and perform dynamic compensation.

[0023] Step 4.2: Data synchronization and real-time linkage;

[0024] Step 4.3: Multi-physics data analysis and optimization;

[0025] Step 4.4: Abnormal state prediction and self-healing:

[0026] The punching pressure curve is obtained through feature modeling. Under normal working conditions, the punching pressure curve has a bimodal characteristic. Under abnormal working conditions, the punching pressure curve has a single peak or abnormal amplitude. The dynamic time warping algorithm is used to calculate the similarity between the current pressure curve and the standard template. When the similarity is less than the preset threshold, an alarm is triggered. The convolutional neural network is used to determine the abnormality type and adjust the parameters.

[0027] Step 4.5: Virtual-reality interaction closed-loop verification:

[0028] Conduct confidence assessment of digital twins, define consistency indicators between virtual models and physical equipment, quantify uncertainty propagation through Monte Carlo simulation, and continuously obtain stamping equipment operation data to update operating parameters.

[0029] Preferably, in step 4.2, data synchronization uses the IEEE 1588 precision time protocol to synchronize the physical sensor with the virtual model clock, and the edge computing node caches and sorts the data stream to align the timestamps;

[0030] Real-time linkage, including physical to virtual linkage and virtual to physical linkage;

[0031] Physical to virtual linkage: stamping equipment sensor data drives the finite element simulation of the virtual 3D model in real time to update the stress field distribution;

[0032] Virtual-to-physical linkage generates optimization instructions based on simulation results, calculates real-time stamping speed, and feeds back to the PLC controller via the OPC UA protocol.

[0033] Preferably, in step 4.3, the multi-physical quantity data analysis and optimization includes the following sub-steps:

[0034] Step 4.3.1. Construct an objective function to minimize energy consumption and mold wear while satisfying the forming quality constraints; use the gradient descent method to search for the local optimal solution for continuous variables and the genetic algorithm to handle the global optimization of discrete variables;

[0035] Step 4.3.2: De-noise the input data using wavelet transform, extract key data features, and obtain the mean value of the impact force and the peak value of the spectrum energy;

[0036] Step 4.3.3: Build a proxy model based on historical data and establish a mapping relationship between process parameters and objective functions;

[0037] Step 4.3.4: Online optimization of the proxy model. After each stamping cycle, the proxy model is updated and the optimal parameter combination is solved.

[0038] Preferably, in step 5, data integration and dynamic adjustment include the following sub-steps:

[0039] Step 5.1: Utilize the Internet of Things and sensor networks to collect real-time operating data of the stamping equipment, integrate it with the digital twin 3D model, perform data analysis, and implement multi-dimensional feature extraction and prediction through wavelet transform, PCA, and LSTM methods;

[0040] Step 5.2: Identify potential problems and performance bottlenecks in stamping equipment, including: mold wear and thermal fatigue at the process level, stamping equipment stiffness and vibration resonance characteristics, process parameter mismatch, and production cycle imbalance and energy utilization bottlenecks at the system coordination level;

[0041] Step 5.3: Dynamically adjust the production plan and equipment parameters. Based on MIP, MPC and adaptive control algorithms, achieve real-time closed-loop optimization of the production plan and equipment parameters, and optimize the digital twin system structure in real time.

[0042] Preferably, in step 5.1, the data analysis method is as follows:

[0043] Step 5.1.1, Time Series Signal Analysis: Perform time-frequency analysis on the non-stationary signal using the wavelet transform method to extract the transient characteristics of vibration and pressure;

[0044] Step 5.1.2, Multivariate Statistical Modeling: Perform principal component analysis to reduce the dimensionality of high-dimensional temperature, pressure, and vibration data and extract the main eigenvectors;

[0045] Step 5.1.3, machine learning classification and prediction: Use the random forest method to determine the type of anomaly based on a multi-decision tree ensemble; use the long short-term memory network to predict the remaining life of the mold, with the input being the historical wear time series data and the output being the RUL data value;

[0046] Step 5.1.4, Optimization modeling: Dynamically optimize the multi-stage production plan with the goal of minimizing the total delay time.

[0047] Preferably, in step 5.3, the production plan and equipment parameters are dynamically adjusted as follows:

[0048] Step 5.3.1. Perform multi-source data fusion at the input layer. The multi-source data streams include: stamping equipment pressure, temperature, vibration status data, production order information, and environmental parameters.

[0049] Step 5.3.2: In the decision-making optimization engine, a dynamic scheduling model is constructed through mixed integer programming to perform predictive control and rolling optimization of stamping equipment parameters.

[0050] Step 5.3.3: Perform closed-loop control at the execution layer and adaptively adjust parameters, including: stamping speed adjustment, lubrication strategy adjustment, and production plan rescheduling;

[0051] Stamping speed adjustment: Dynamic speed reduction based on LSTM prediction of mold life; Lubrication strategy: Adjust the lubricant injection amount according to the real-time friction coefficient; Production plan rescheduling: Order insertion based on urgent order status, and reallocate workstation tasks using the Hungarian algorithm.

[0052] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0053] 1. The stamping equipment construction method of the present invention, through digital twin technology, realizes the efficiency and intelligence of stamping equipment construction, significantly shortens the design cycle, improves production efficiency and equipment performance, and reduces costs and operation and maintenance risks. It has effectively promoted the development of intelligent manufacturing and enhanced the market competitiveness of enterprises.

[0054] 2. The stamping equipment construction method of the present invention realizes the digital and intelligent management of the entire process of stamping equipment design, production, operation and maintenance through high-precision three-dimensional modeling, multi-physical quantity simulation, data acquisition and processing, digital twin engine, virtual-reality mapping and real-time linkage, data analysis and optimization, as well as human-computer interaction and visualization.

[0055] 3. The implementation of the stamping equipment construction method of the present invention will significantly improve the design accuracy, production efficiency and operation and maintenance management level of the stamping equipment, and has broad application prospects and important economic value. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a process architecture diagram of the digital twin-based stamping equipment construction method of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions of the application are further elaborated in detail below with reference to the accompanying drawings. The described embodiments are only a part of the embodiments involved in the present invention. All non-innovative embodiments of other researchers in this field on this embodiment fall within the scope of protection of the present invention. At the same time, the step numbers in the embodiments of the present invention are only set for the convenience of explanation and description, and the order between the steps is not limited in any way. The execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0058] The present invention provides a method for constructing a stamping equipment based on digital twins, comprising the following steps:

[0059] Step 1: Build a digital twin system for stamping equipment, including the physical layer, digital layer, virtual layer, and application layer.

[0060] Step 2: Perform three-dimensional modeling and multi-physical quantity simulation of stamping equipment at the physical layer;

[0061] Step 3: Collect and process real-time data of stamping equipment at the digital layer and build a digital twin engine;

[0062] Step 4: Perform virtual-real integration and real-time linkage processing in the virtual layer to complete stamping equipment data analysis and optimization;

[0063] Step 5: Perform human-computer interaction and visualization processing at the application layer, and provide intelligent decision support through data integration and dynamic adjustment.

[0064] Specifically, in one embodiment of the present invention, the overall structure of the digital twin system of the stamping equipment, such as Figure 1 As shown, it includes: physical layer, digital layer, virtual layer and application layer, which are used to realize high-precision three-dimensional modeling of stamping equipment, multi-physical quantity simulation, data acquisition and processing, virtual-reality mapping and real-time linkage, data analysis and optimization, human-computer interaction and visualization, data integration and dynamic adjustment, and intelligent decision support.

[0065] Specifically, the physical layer includes high-precision three-dimensional modeling and multi-physics simulation.

[0066] High-precision 3D modeling: Utilizing advanced 3D modeling software and combining detailed data from the stamping equipment, we construct a 1:1 realistic 3D model. This ensures the model is highly consistent with the physical equipment in terms of geometry, material properties, and other aspects, providing a precise physical foundation for subsequent simulation and data analysis.

[0067] In this embodiment, the three-dimensional modeling software includes: SolidWorks, which is suitable for parametric modeling and supports the whole process design from sketch to assembly; CATIA, which is suitable for complex surface modeling (such as automobile cover molds) and supports high-precision reverse engineering; AutoCAD Inventor, which is suitable for mechanical structure modeling and seamlessly connects with simulation tools (such as ANSYS); NX (UG), which integrates CAD / CAE functions and supports multidisciplinary collaborative design.

[0068] Multi-physics simulation: Based on the 3D model, we integrate the simulation of multiple physical quantities such as friction, gravity, and resistance. This step can accurately reproduce the operating state of the stamping equipment under different working conditions, providing a reliable basis for equipment optimization design and fault analysis.

[0069] Among them, working conditions include: process parameter conditions, material property conditions, mold and equipment conditions, environmental condition conditions, special scenario conditions, and multi-physical field coupling conditions.

[0070] Specifically, in this embodiment:

[0071] 1. Under the process parameter conditions, the operation process includes:

[0072] Match the material deformation rate according to the preset punching speed (low, medium, high);

[0073] Precisely control the stroke length to ensure consistent contact time between the mold and the workpiece and the forming depth;

[0074] Stabilize the load within the rated range and dynamically balance the hydraulic / mechanical system pressure.

[0075] Abnormal conditions include:

[0076] Speed ​​fluctuation: Delayed response of the motor or servo system, causing the material to tear or rebound uncontrollably;

[0077] Stroke deviation: sensor failure or mechanical wear causes stroke to exceed the limit, causing mold collision;

[0078] Overload failure: The pressure peak exceeds the equipment limit, resulting in hydraulic oil leakage or transmission mechanism breakage; 2. Under material characteristics, the operation process includes:

[0079] Materials (such as high-strength steel and aluminum alloy) are fed stably according to preset parameters (thickness and ductility);

[0080] Annealed / coated materials deform uniformly and have a controllable coefficient of friction.

[0081] Abnormal conditions include:

[0082] Material mismatch: material hardness or thickness exceeds tolerance, resulting in punching burrs or cracks (such as deep drawing cracks on thin plates); Friction loss: coating peeling or sudden change in surface roughness, causing localized adhesive wear (such as mold "hair"); 3. Under the working conditions of mold and equipment, the operation process includes:

[0083] The mold gap accurately matches the material thickness to ensure a smooth shear surface;

[0084] Polishing / coating the mold surface reduces friction and extends life;

[0085] The equipment has high rigidity and the dynamic deformation is within the allowable range.

[0086] Abnormal conditions include:

[0087] Abnormal clearance: The mold is misaligned or worn, resulting in excessive burrs or chipped cutting edges;

[0088] Surface deterioration: coating peeling or scratching, causing workpiece scratches or mold jamming;

[0089] Insufficient rigidity: The vibration of old equipment is aggravated, resulting in reduced forming accuracy (such as out-of-tolerance dimensions of automobile cover parts).

[0090] 4. Under environmental conditions, the operation process includes:

[0091] The lubricant evenly covers the interface between the mold and the workpiece, reducing frictional heat;

[0092] Stable ambient temperature / high temperature environment (such as hot stamping furnace temperature ±5℃ control);

[0093] The basic shock-absorbing device isolates the external vibration interference.

[0094] Abnormal conditions include:

[0095] Lubrication failure: Lubricant contamination or supply interruption causes mold overheating or workpiece adhesion;

[0096] Temperature control out of control: local overheating during hot stamping causes abnormal material phase change (such as incomplete martensite transformation); resonance risk: the superposition of stamping frequencies at multiple stations causes severe vibration of the equipment and even structural damage.

[0097] 5. Under special working conditions, the operation process includes:

[0098] In case of emergency stop, the braking system quickly absorbs the inertial kinetic energy to avoid mold collision;

[0099] Under eccentric load conditions, the adaptive pressure compensation system balances the load distribution;

[0100] Regularly check mold wear and replace or repair it in time.

[0101] Abnormal conditions include:

[0102] Unbalanced load instability: Material placement deviation causes one-sided overload of the mold, causing the guide pillar to break (such as in the stamping of precision connectors);

[0103] Shutdown shock: Brake failure causes the slider to hit the lower die due to inertia, causing die corner collapse;

[0104] Fatigue accumulation: If the mold is not replaced for a long time, the edge will become blunt, resulting in a sudden increase in blanking force (such as excessive burrs on silicon steel sheets).

[0105] 6. Under multi-physics field coupling conditions, the operation process includes:

[0106] Thermal-mechanical coupling: Frictional heat is quickly dissipated through the cooling system to avoid excessive softening of the material.

[0107] Fluid-solid coupling: Lubricant flow and mold movement work together to form a stable lubricating film.

[0108] Control-Structure Coupling: The servo motor adjusts the stamping curve in real time to suppress vibration.

[0109] Abnormal conditions include:

[0110] Uncontrolled thermal softening: The local temperature is too high during high-speed stamping, and the workpiece springback is unpredictable (such as warping of aluminum sheet forming).

[0111] Lubricating film rupture: Lubricant is squeezed out under high pressure, causing boundary friction and die scratches.

[0112] Control delay: The servo system responds laggingly, resulting in motion trajectory deviation (such as failure in forming complex curved parts).

[0113] In this embodiment, multi-physical quantity simulation achieves stable and efficient stamping by accurately matching process parameters, material properties, equipment status and environmental conditions; further analyzes the root cause of abnormal conditions, determines the imbalance of multi-physical quantity coupling (such as thermal-mechanical mismatch, friction-motion imbalance) or external interference beyond the design tolerance; its simulation value lies in: predicting abnormal risks in advance (such as mold overheating, eccentric load fracture), and optimizing parameters to expand the "safe operating window".

[0114] Specifically, the digital layer includes data acquisition and processing, and digital twin engine.

[0115] Data collection and processing: Using industrial internet technologies and radio frequency identification (RFID) technologies, physical layer data is collected in real time and processed through cleaning, correlation, and mining. This processed data forms the data foundation of the digital twin system, supporting subsequent data analysis and model updates.

[0116] Among them, physical layer data includes: equipment operation status data, mold and workpiece interaction data, environment and auxiliary system data, engineering control and abnormal signal data.

[0117] In this embodiment, the specific physical layer data is:

[0118] 1. Equipment operating status data: motion parameters, mechanical parameters, dynamic response;

[0119] 2. Interaction data between mold and workpiece: mold status, workpiece information, friction characteristics;

[0120] 3. Environmental and auxiliary system data: environmental parameters, lubrication status, energy consumption data;

[0121] 4. Engineering control and abnormal signals: control instructions, abnormal characteristics, and time series data.

[0122] Digital Twin Engine: As the core module connecting the physical and virtual layers, the digital twin engine in this embodiment is responsible for real-time data transmission, dynamic model updates, and driving virtual-reality interactions. This engine is the key technology for achieving real-time linkage between physical devices and virtual models.

[0123] Specifically, the virtual layer includes virtual-reality mapping, real-time linkage, and data analysis and optimization.

[0124] Virtual-reality mapping and real-time linkage: Digital twin technology enables virtual-reality mapping between the physical and cyberspace of stamping equipment, collecting production data in real time and reflecting it in the virtual model. This high degree of synchronization and real-time linkage ensures efficient and accurate production processes.

[0125] Data analysis and optimization: Data analysis and optimization based on digital twin models supports real-time monitoring of the production process, fault warnings, and scheduling planning. This helps improve the efficiency of fault detection and resolution, and reduces the probability of production accidents and downtime.

[0126] In the virtual layer of stamping equipment based on digital twins, virtual-reality interaction achieves high-precision collaboration between physical space and cyberspace through virtual-reality mapping and real-time linkage. This is carried out in conjunction with coordinate system construction, data synchronization, and optimization algorithms. Specifically, it includes the following sub-steps:

[0127] 1. Coordinate system construction and dynamic conversion to achieve physical-virtual coordinate system alignment:

[0128] Deploy laser trackers at key motion nodes of physical equipment (such as sliders and molds) to establish a world coordinate system (physical space reference);

[0129] Define a local coordinate system in the virtual model (e.g. the center of the mold is the origin) and implement coordinate mapping through a homogeneous transformation matrix;

[0130] Dynamic compensation: Kalman filtering is used to fuse multi-sensor data (displacement, acceleration) to correct coordinate system drift caused by mechanical vibration or thermal deformation in real time.

[0131] In particular, in this embodiment, virtual mapping of the stamping slider stroke is performed, and the physical displacement is converted into a virtual displacement after coordinate transformation, which can control the accuracy error within ±0.1 mm.

[0132] 2. Data synchronization and real-time linkage include the following sub-steps:

[0133] 2.1. Timestamp alignment:

[0134] Use IEEE 1588 Precision Time Protocol (PTP) to ensure synchronization between physical sensors and virtual model clocks (microsecond deviation);

[0135] Data streams are cached and sorted at the millisecond level through edge computing nodes to avoid time sequence errors.

[0136] 2.2 Real-time linkage mechanism:

[0137] Physical → Virtual: Punch force sensor data drives the finite element simulation (FEA) of the virtual model in real time, updating the stress field distribution.

[0138] Virtual → Physical: Optimization instructions (such as adjusting punch speed) are generated based on simulation results and fed back to the PLC controller via the OPC UA (Open Platform Communications Unified Architecture) protocol. OPC UA is a communication standard for industrial automation developed by the OPC Foundation.

[0139] In particular, in this embodiment, when high-speed stamping is performed, the virtual model can detect that the mold stress exceeds the limit and immediately trigger the physical equipment to slow down to avoid mold cracking.

[0140] 3. Multi-physics data analysis and optimization, including the following sub-steps:

[0141] 3.1. Construct the objective function and extract data features through gradient descent method and genetic algorithm;

[0142] The objective function is to minimize energy consumption E and die wear W while satisfying the forming quality constraints.

[0143] The optimization algorithm is:

[0144] Gradient descent method: quickly search for local optimal solutions for continuous variables (such as punch speed and lubricant flow);

[0145] Genetic Algorithm (GA): The GA method is used to process the global optimization of discrete variables (such as mold gap level) and adapt to multi-constraint scenarios.

[0146] The optimization process is: data preprocessing and key feature extraction:

[0147] Before optimization, the data contained noise (such as vibration interference), non-steady state (such as lubricant flow fluctuation) and multi-time scale characteristics; data preprocessing was performed, denoising was performed through wavelet transform, and key features (such as the mean value of the impact force and the peak value of the spectrum energy) were extracted.

[0148] 3.2 Model construction and training:

[0149] Build a proxy model (such as Gaussian process regression) based on historical data to establish a mapping relationship between process parameters and objective functions;

[0150] 3.3. Perform online optimization:

[0151] After each stamping cycle is completed, the agent model is updated and the optimal parameter combination is solved.

[0152] In particular, in this embodiment, the optimization comparison is shown in Table 1 below:

[0153] Table 1: Comparison of multi-physics data optimization effects

[0154]

[0155] 4. Abnormal state prediction and self-healing, including the following sub-steps:

[0156] 4.1 Feature Modeling

[0157] Under normal working conditions, the punching pressure curve conforms to the bimodal characteristics, which is divided into the pre-pressing stage t1 and the forming stage t2;

[0158] Under abnormal conditions (such as eccentric loading), the curve is distorted (single peak or abnormal amplitude).

[0159] 4.2 Self-healing strategy:

[0160] Real-time detection: Based on the dynamic time warping (DTW) algorithm, the similarity S between the current pressure curve and the standard template is calculated. If S<0.8, an alarm is triggered.

[0161] Root cause analysis: Classify the type of anomaly (e.g., insufficient lubrication, mold misalignment) through convolutional neural networks (CNN);

[0162] Parameter adjustment: automatically switch to the backup process parameter library.

[0163] In particular, in this embodiment, after the mold is detected to be unevenly loaded, the virtual layer dynamically adjusts the slider pressure distribution in real time, and can quickly restore balanced stamping.

[0164] 5. Virtual-reality interaction closed-loop verification includes the following sub-steps:

[0165] 5.1. Digital Twin Confidence Assessment

[0166] Define consistency indicators between the virtual model and the physical device (such as displacement error, force error), quantify uncertainty propagation through Monte Carlo simulation, and ensure that the confidence level is greater than or equal to 95%.

[0167] 5.2. Closed-loop iteration:

[0168] The operating data of physical equipment is continuously input into the virtual layer to update the model parameters (such as the coefficients of the material constitutive equation), forming a "perception-simulation-optimization-feedback" closed loop.

[0169] It can be seen that the present invention can achieve the following by performing virtual-reality mapping, real-time linkage, and data analysis optimization through the virtual layer:

[0170] (1) High-precision coordinate system dynamic mapping: sub-millimeter level virtual-real space alignment is achieved through laser tracking and Kalman filtering;

[0171] (2) Fusion of multi-objective optimization algorithms: combining gradient descent and genetic algorithms to balance efficiency and global optimality;

[0172] (3) Abnormal self-healing mechanism: real-time diagnosis and parameter adaptive adjustment based on DTW and CNN;

[0173] (4) Closed-loop verification system: Monte Carlo simulation and confidence assessment ensure the reliability of digital twins.

[0174] Furthermore, through the virtual layer of this embodiment, the process controllability, abnormal response speed and energy efficiency level of the stamping equipment can be significantly improved, providing core support for intelligent manufacturing.

[0175] Specifically, the application layer includes: human-computer interaction and visualization, data integration and dynamic adjustment, and intelligent decision support.

[0176] Human-machine interaction and visualization: Provides a complete human-machine interaction interface and visualization effects, allowing users to intuitively understand the structure and operating status of the stamping equipment. This improves the level of operation and maintenance management and enables operators to monitor and manage equipment more conveniently.

[0177] Data integration and dynamic adjustment: Leveraging IoT technology and sensor networks, we collect real-time operational data from stamping equipment and integrate it with the digital twin model. Through data analysis, we can promptly identify potential equipment issues and performance bottlenecks, dynamically adjust production plans and equipment parameters, and ensure stable operation and efficient production.

[0178] In the data integration and dynamic adjustment of stamping equipment based on digital twins, data analysis methods, potential problems and performance bottlenecks, and dynamic adjustment strategies are the core technical links, as follows:

[0179] 1. Data analysis, including the following methods:

[0180] 1.1. Time series signal analysis: Through wavelet transform, time-frequency analysis of non-stationary signals such as vibration and pressure is performed to extract transient features

[0181] In particular, this embodiment can be used to detect abnormal impact signals (such as die collision spikes) during the stamping process.

[0182] 1.2. Multivariate statistical modeling: Through principal component analysis (PCA), dimensionality reduction is performed on high-dimensional data (such as temperature, pressure, and vibration) to extract the main eigenvectors.

[0183] In particular, this embodiment can be used to identify the hidden variable relationship between energy consumption and punching speed, and optimize energy efficiency.

[0184] 1.3. Machine Learning Classification and Prediction: Using the random forest method and multiple decision tree ensembles, we classify abnormality types (such as insufficient lubrication vs. mold overload). We use the short-term memory network (LSTM) to predict the remaining life of the mold (the input data is the historical wear time series data, and the output data is the RUL value).

[0185] 1.4. Optimization modeling: Multi-stage production planning optimization is performed through dynamic programming methods, with the goal of minimizing the total delay time.

[0186] 2. Analysis of potential problems and performance bottlenecks at the equipment and process levels. Specific methods include:

[0187] 2.1. Mould related

[0188] Wear accumulation: blunting of the cutting edge leads to increased blanking force;

[0189] Thermal fatigue cracks: local temperature cycling causes micro cracks on the die surface (common in hot stamping).

[0190] 2.2. Dynamic characteristics of equipment

[0191] Insufficient rigidity: deformation of the fuselage under high-speed stamping results in poor forming accuracy;

[0192] Vibration resonance: The punching frequency overlaps with the natural frequency of the equipment.

[0193] 2.3 Process parameter mismatch

[0194] Speed-material mismatch: The springback of high-strength steel is out of control under high-speed stamping;

[0195] Lubrication failure: Boundary friction causes scratches on the workpiece surface.

[0196] 2.4 System Collaboration Bottleneck

[0197] Production rhythm disorder: the multi-station stamping rhythm is not synchronized, resulting in waste of production capacity;

[0198] Low energy utilization: No-load energy consumption accounts for a high proportion.

[0199] 3. Dynamically adjust production plans and equipment parameters to implement real-time optimization frameworks. Specific methods include:

[0200] 3.1. Data fusion is performed at the input layer, where multi-source data streams include: equipment status data (pressure, temperature, vibration), production order information (priority, delivery date), and environmental parameters (temperature, humidity, lubricant inventory).

[0201] 3.2. In the decision-making optimization engine, a dynamic scheduling model is constructed through mixed integer programming (MIP), model predictive control (MPC) is performed, and the parameters of the stamping equipment are continuously optimized.

[0202] 3.3. Closed-loop control is performed at the execution layer, and parameter adaptive adjustment includes: stamping speed adjustment, dynamic speed reduction based on LSTM-predicted mold life; lubrication strategy adjustment, adjusting the lubricant injection amount according to the real-time friction coefficient; production plan rescheduling, and when urgent orders are inserted, quickly reallocating workstation tasks using the Hungarian algorithm.

[0203] In particular, in this embodiment, the example scenario is: an abnormal increase in mold temperature (T>300°C) is detected, and the corresponding dynamic adjustment is:

[0204] 1. Virtual layer simulation predicts the thermal softening effect and generates instructions: reduce speed and increase cooling water flow;

[0205] 2. The production planning system automatically extends the batch cycle time and triggers preparation for mold switching.

[0206] The comparison and effects of the method of the present invention and traditional technology are shown in Table 2 below:

[0207] Table 2: Technology comparison and results

[0208]

[0209] It can be seen that the present invention integrates and dynamically adjusts stamping equipment data based on digital twins. In data analysis, multi-dimensional feature extraction and prediction are realized through wavelet transform, PCA, LSTM, etc.; the potential problems analyzed include: covering key bottlenecks such as mold wear, thermal fatigue, vibration resonance, and parameter mismatch; the dynamic adjustments performed include: based on MIP, MPC and adaptive control algorithms, to achieve real-time closed-loop optimization of production plans and equipment parameters; its innovative value lies in: deeply integrating data-driven decision-making with physical mechanism models to significantly improve the intelligence level and economic benefits of stamping equipment.

[0210] Finally, intelligent decision support is provided: Based on the digital twin model and data analysis results, an intelligent decision support system is built. This system can provide fault warnings, maintenance recommendations, and production optimization strategies based on the equipment's operating status and production needs, assisting managers in making informed decisions and improving production efficiency and equipment utilization.

[0211] In an embodiment of the present invention, an electronic device is also provided, including: one or more processors; a storage device on which one or more programs are stored; when the one or more programs are executed by the one or more processors, the one or more processors implement the digital twin-based stamping equipment construction method described in any of the above embodiments.

[0212] In an embodiment of the present invention, a computer-readable storage medium is further provided, on which a computer program is stored. When the program is executed by a processor, any one of the digital twin-based stamping equipment construction methods in the above embodiments is implemented.

[0213] It can be seen that the digital twin-based stamping equipment construction method of the present invention realizes the efficiency and intelligence of stamping equipment construction through digital twin technology, significantly shortens the design cycle, improves production efficiency and equipment performance, and at the same time reduces costs and operation and maintenance risks, which effectively promotes the development of intelligent manufacturing and enhances the market competitiveness of enterprises.

[0214] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for constructing stamping equipment based on digital twin, characterized in that: The steps include: Step 1: Build a digital twin system for stamping equipment, including the physical layer, digital layer, virtual layer, and application layer. Step 2: Perform three-dimensional modeling and multi-physical quantity simulation of stamping equipment at the physical layer; Step 3: Collect and process real-time data of stamping equipment at the digital layer to build a digital twin engine; Step 4: Perform virtual-real integration and real-time linkage processing in the virtual layer to complete stamping equipment data analysis and optimization, and achieve coordination between physical space and cyber space; Step 5: Perform human-computer interaction and visualization processing at the application layer, make intelligent decisions through data integration and dynamic adjustment, and provide fault warnings, maintenance suggestions, and production optimization strategies based on the operating status of the stamping equipment and production needs.

2. The method for constructing stamping equipment based on digital twin according to claim 1, characterized in that: In step 2, the three-dimensional modeling of the stamping equipment is performed by parametric modeling using three-dimensional modeling software and combined with reverse engineering to build a 1:1 true restoration three-dimensional model based on the stamping equipment data, which is consistent with the physical stamping equipment in terms of geometric dimensions and material properties; The three-dimensional modeling software includes: SolidWorks, CATIA, AutoCAD Inventor, and NX; The stamping equipment data includes: geometric data, material properties, and process parameters.

3. The method for constructing stamping equipment based on digital twin according to claim 1, characterized in that: In step 2, the multi-physical quantity simulation integrates multiple physical quantities on the basis of the three-dimensional model, including friction, gravity, and resistance, and performs simulation to restore the operating state of the stamping equipment under different working conditions, analyze the root cause of the abnormal state, determine that the abnormality is an imbalance in the coupling of multiple physical quantities or external interference exceeds the design tolerance, predict the abnormality risk in advance and perform parameter optimization; The working conditions include: process parameter conditions, material property conditions, mold and equipment conditions, environmental condition conditions, special scenario conditions, and multi-physical field coupling conditions.

4. The method for constructing stamping equipment based on digital twin according to claim 1, characterized in that: In step 3, the data collection and processing is to collect physical layer data in real time through the industrial Internet and radio frequency identification methods, and perform cleaning, correlation and mining to form the data foundation of the digital twin system; The physical layer data includes: equipment operation status data, mold and workpiece interaction data, environment and auxiliary system data, engineering control and abnormal signal data; The digital twin engine is used to connect the physical layer and the virtual layer to transmit real-time data, dynamically update the three-dimensional model, and drive the virtual-reality interaction in the virtual layer.

5. The method for constructing stamping equipment based on digital twin according to claim 4, characterized in that: In the virtual layer described in step 4, virtual-reality interaction includes virtual-reality mapping and real-time linkage. The virtual-reality mapping of the physical space and cyberspace of the stamping equipment is performed through the digital twin method to achieve the collaboration between the physical space and cyberspace, including the following sub-steps: Step 4.1, coordinate system construction and dynamic conversion: Laser trackers are deployed at key motion nodes of the physical equipment to establish a world coordinate system. In the virtual 3D model, a local coordinate system is defined with the mold center as the origin, and coordinate mapping is performed using a homogeneous transformation matrix. Kalman filtering is used to fuse multi-sensor data to correct coordinate system drift caused by mechanical vibration or thermal deformation in real time and perform dynamic compensation. Step 4.2: Data synchronization and real-time linkage; Step 4.3: Multi-physics data analysis and optimization; Step 4.4: Abnormal state prediction and self-healing: The punching pressure curve is obtained through feature modeling. Under normal working conditions, the punching pressure curve has a bimodal characteristic. Under abnormal working conditions, the punching pressure curve has a single peak or abnormal amplitude. The dynamic time warping algorithm is used to calculate the similarity between the current pressure curve and the standard template. When the similarity is less than the preset threshold, an alarm is triggered. The convolutional neural network is used to determine the abnormality type and adjust the parameters. Step 4.5: Virtual-reality interaction closed-loop verification: Conduct confidence assessment of digital twins, define consistency indicators between virtual models and physical equipment, quantify uncertainty propagation through Monte Carlo simulation, and continuously obtain stamping equipment operation data to update operating parameters.

6. The method for constructing stamping equipment based on digital twin according to claim 5, characterized in that: In step 4.2, the data synchronization uses the IEEE 1588 precision time protocol to synchronize the physical sensor with the virtual model clock, and the edge computing node caches and sorts the data stream to align the timestamps. The real-time linkage includes: physical to virtual linkage and virtual to physical linkage; The physical-to-virtual linkage uses the sensor data of the stamping equipment to drive the finite element simulation of the virtual three-dimensional model in real time to update the stress field distribution; The virtual-to-physical linkage generates optimization instructions based on simulation results, calculates real-time stamping speed, and feeds back to the PLC controller via the OPC UA protocol.

7. The method for constructing stamping equipment based on digital twin according to claim 5, characterized in that: In step 4.3, the multi-physical quantity data analysis and optimization includes the following sub-steps: Step 4.3.

1. Construct an objective function to minimize energy consumption and mold wear while satisfying the forming quality constraints; use the gradient descent method to search for the local optimal solution for continuous variables and the genetic algorithm to handle the global optimization of discrete variables; Step 4.3.2: De-noise the input data using wavelet transform, extract key data features, and obtain the mean value of the impact force and the peak value of the spectrum energy; Step 4.3.3: Build a proxy model based on historical data and establish a mapping relationship between process parameters and objective functions; Step 4.3.4: Online optimization of the proxy model. After each stamping cycle, the proxy model is updated and the optimal parameter combination is solved.

8. The method for constructing stamping equipment based on digital twin according to claim 5, characterized in that: In step 5, the data integration and dynamic adjustment includes the following sub-steps: Step 5.1: Utilize the Internet of Things and sensor networks to collect real-time operating data of the stamping equipment, integrate it with the digital twin 3D model, perform data analysis, and implement multi-dimensional feature extraction and prediction through wavelet transform, PCA, and LSTM methods; Step 5.2: Identify potential problems and performance bottlenecks in stamping equipment, including: mold wear and thermal fatigue at the process level, stamping equipment stiffness and vibration resonance characteristics, process parameter mismatch, and production cycle imbalance and energy utilization bottlenecks at the system coordination level; Step 5.3: Dynamically adjust the production plan and equipment parameters. Based on MIP, MPC and adaptive control algorithms, achieve real-time closed-loop optimization of the production plan and equipment parameters, and optimize the digital twin system structure in real time.

9. The method for constructing stamping equipment based on digital twin according to claim 8, characterized in that: In step 5.1, the data analysis method is as follows: Step 5.1.1, Time Series Signal Analysis: Perform time-frequency analysis on the non-stationary signal using the wavelet transform method to extract the transient characteristics of vibration and pressure; Step 5.1.2, Multivariate Statistical Modeling: Perform principal component analysis to reduce the dimensionality of high-dimensional temperature, pressure, and vibration data and extract the main eigenvectors; Step 5.1.3, machine learning classification and prediction: Use the random forest method to determine the type of anomaly based on a multi-decision tree ensemble; use the long short-term memory network to predict the remaining life of the mold, with the input being the historical wear time series data and the output being the RUL data value; Step 5.1.4, Optimization modeling: Dynamically optimize the multi-stage production plan with the goal of minimizing the total delay time.

10. The method for constructing stamping equipment based on digital twin according to claim 8, characterized in that: In step 5.3, the method for dynamically adjusting the production plan and equipment parameters is as follows: Step 5.3.

1. Perform multi-source data fusion at the input layer. The multi-source data streams include: stamping equipment pressure, temperature, vibration status data, production order information, and environmental parameters. Step 5.3.2: In the decision-making optimization engine, a dynamic scheduling model is constructed through mixed integer programming to perform predictive control and rolling optimization of stamping equipment parameters. Step 5.3.3: Perform closed-loop control at the execution layer and adaptively adjust parameters, including: stamping speed adjustment, lubrication strategy adjustment, and production plan rescheduling; The stamping speed adjustment is based on LSTM to predict the mold life and dynamically reduce the speed; the lubrication strategy adjusts the lubricant injection amount according to the real-time friction coefficient; the production plan rescheduling inserts orders according to the emergency order status and reallocates workstation tasks using the Hungarian algorithm.

Images