Method for smart manufacturing based on digital twin

Abstract

The present invention provides a smart manufacturing system and method based on digital twin technology, applicable to the operation and process control of production machinery or machining equipment. By integrating the Internet of Things (IoT), big data analytics, and digital twin technology, this invention achieves Cyber-Physical integration and dynamic optimization of manufacturing processes. The method comprises the following steps: 1. Scene IoT Integration: Utilizing multi-dimensional sensors embedded in intelligent machining devices, such as "spindle speed," "feed rate," "object position," "axial force, torque, or bending moment of the tool," "vibration values of production machinery or machining equipment," "current," and "temperature," to monitor operational parameters of production machinery or machining equipment in real time. Data is transmitted synchronously to industrial PCs and cloud servers via edge computing to minimize latency. 2. Digital Synchronization: Leveraging digital twin modeling technology to construct real-time synchronized digital twin models reflecting the operations of production machinery or machining equipment. These models simulate dynamic features such as tool wear and fluctuations in cutting forces during machining processes. 3. Decision Feedback: Employing big data analytics for in-depth analysis to achieve anomaly diagnosis, predict machining trends, and generate targeted optimization recommendations. 4. Closed-loop Control: Dynamically adjusting the parameters of production machinery or machining equipment based on analysis results to optimize machining strategies. Simultaneously, the digital twin models are updated in real time to maintain Cyber-Physical synchronization. By employing digital twin technology, this method significantly enhances the efficiency, precision, and resource utilization of production or machining processes, providing an effective solution for building smart manufacturing systems and smart factories. Furthermore, this method improves production or machining efficiency, enhances precision, and extends tool life, offering an optimal approach to advancing smart manufacturing and smart factory development.

Description

[Technical Field]

[0001] A method of manufacturing or processing, particularly a method of applying digital twins to smart manufacturing. [Previous Technology]

[0002] Digital twin technology has become a key driving force for smart manufacturing, especially in the production and processing fields, where scene IoT, digital synchronization, and decision feedback are the core of its application. However, current technological development still faces the following challenges in data collection, model synchronization, and optimization feedback:

[0003] Challenges of Scene IoT: Taking precision machining center equipment as an example, previous monitoring systems mostly relied on single sensors or static data for monitoring, limited to basic parameters (such as tool speed) recording, making it difficult to form a comprehensive monitoring of the entire machining site. For example, the sensors of traditional cutting centers mainly capture vibration data, but lack comprehensive analysis of parameters such as cutting force, torque, and bending moment. This leads to anomaly detection relying heavily on human experience, making it impossible to achieve real-time monitoring and response in the control room.

[0004] Deficiencies of digital synchronization: Delays during data transmission lead to untimely updates of the virtual model. For example, in the tool wear monitoring of a cutting center, incorrect judgments in the virtual model caused by data delays may result in premature tool replacement or overuse, affecting production costs and machining accuracy.

[0005] Limitations of Anomaly Feedback: Cutting center control systems typically rely on historical data or static rules, lacking real-time feedback and intelligent adjustment capabilities. For example, when the cutting parameters of a cutting center are abnormal, existing systems can only issue warnings but cannot automatically generate adjustment suggestions or correct the machining path in real time, leading to increased problem-solving time due to manual intervention. Simultaneously, the maintenance of production machines or machining equipment still relies on fixed schedules, lacking a dynamic adjustment mechanism based on the health status of the machines, resulting in inefficiency and increased unnecessary downtime.

[0006] This invention integrates innovative technologies to provide the following solutions:

[0007] Enhancing Data Collection and Visualization Capabilities in IoT Scenarios: By deploying various sensing modules, such as sensors for "spindle speed," "feed rate," "object position," "tool axial force, torque, or bending moment," "vibration value of production machine or processing equipment," "current," and "temperature," and connecting them in series with the controllers of production machines or processing equipment via communication protocols, real-time collection and comprehensive monitoring of multi-dimensional data from the machining or manufacturing site can be achieved. For example, multiple sensors can be integrated into a cutting center. Data from the sensors and controllers on the cutting center can be collected by an industrial computer or a server with computing capabilities and aggregated to a unified platform, forming a visualized machining control room. These sensors can be built into the production machine or processing equipment itself or used as accessories for peripheral devices.

[0008] Optimize digital synchronization and virtual model updates: Utilize efficient IoT technology to ensure the timeliness and integrity of production or processing data transmission, ensuring accurate updates of the virtual model. For example, production or processing data such as "spindle speed," "feed rate," "object position," "tool axial force, torque, or bending moment," "vibration value of production machine or processing equipment," "current," and "temperature" are imported into the virtual model in real time. Through big data analysis, the production process or processing behavior is simulated, providing visualized simulation results for on-site operators or managers to make relevant decision-making improvements. This virtual model is generated using computer graphics or other digital modeling methods and is simulated synchronously with the imported production or processing data.

[0009] Enhancing Decision Feedback and Production / Process Optimization Capabilities: Combining big data analytics and decision feedback enables closed-loop control. For example, the system can automatically adjust feed rate or enhance cooling intensity based on abnormal fluctuations in cutting force or temperature, reducing tool wear and improving machining accuracy. Maintenance of production machines or processing equipment can also be shifted from a fixed schedule to dynamic adjustments based on health status.

[0010] Promoting collaborative innovation in smart manufacturing: Integrating processing data, order information, and production scheduling to support flexible batch production. For example, when order requirements change, the system quickly rearranges production plans based on virtual processing simulation results, improving the utilization rate of production machines or processing equipment and shortening delivery time.

[0011] Through the all-round monitoring of scene IoT, the precise simulation of digital synchronization and the intelligent optimization of decision feedback, the present invention can significantly improve the efficiency, accuracy and resource utilization of processing or manufacturing, and upgrade the traditional processing mode into a flexible, efficient and intelligent modern manufacturing system to meet the market's demand for high-quality and high-efficiency production. [Summary of the Invention]

[0012] This invention proposes a smart manufacturing method based on digital twins, which combines the Internet of Things and big data analysis to improve processing accuracy, production efficiency and resource utilization, and achieve comprehensive optimization of smart factories.

[0013] To achieve the aforementioned objectives, this invention provides a digital twin-based intelligent manufacturing system that achieves dynamic monitoring and optimization of the processing flow through virtual-physical integration. The system includes the following key units: Production machine tools or processing equipment. Taking a precision machining center as an example, its intelligent machining device integrates cutting tools and multiple IoT sensors (such as "spindle speed," "feed rate," "object position," "tool axial force, torque, or bending moment," "vibration value of the production machine tool or processing equipment," "current," and "temperature," etc.) at the front end for real-time collection and transmission of processing data. IPC industrial computer: Integrates the operation and processing status data of multiple cutting centers for real-time monitoring and control. Cloud server: Provides data storage and computing capabilities, supporting the construction of a simulation environment for the digital twin platform. Digital twin platform: Establishes a virtual-physical integrated model to simulate the processing process, perform operational analysis, anomaly detection, and decision optimization suggestions.

[0014] This invention relates to a smart manufacturing method based on digital twins, comprising the following steps: A1. Scene IoT: Sensors monitor the operating parameters and processing status of the cutting center in real time, transmitting the data to an industrial computer and a cloud server. A2. Digital Synchronization: The digital twin platform utilizes big data from the cloud server to construct a virtual-real integrated model, simulating the actual processing process. A3. Decision Feedback: Based on big data, data regression analysis is performed, anomaly diagnosis and trend prediction are conducted, and specific optimization suggestions are generated. A4. Closed-Loop Control: Based on the analysis results, the operator adjusts the cutting center parameters with reference to the expert system, forming closed-loop control, synchronously updating the digital twin model, and achieving continuous optimization.

[0015] The system and method of the present invention achieve virtual-real integration through digital twin technology, and combine IoT data collection and big data analysis, which has the following advantages in production and manufacturing or processing: (1) Improved operating efficiency: Through parameter optimization, tool wear is significantly reduced and processing accuracy is improved. (2) Smart manufacturing collaboration: Supports flexible batch production and can quickly respond to changes in market demand. (3) Cost and time savings: Real-time problem detection and resolution reduce production and manufacturing or processing costs and shorten delays.

[0016] This invention provides an efficient and feasible solution that integrates the virtual and real worlds through digital twin technology to achieve precise simulation and dynamic optimization of the machining process. This not only improves the performance of cutting centers but also effectively reduces production or processing costs, laying the foundation for the construction of smart factories and possessing broad application value and market prospects.

Implementation Method

[0017] Please refer to Figure 1. The various devices and functions of the present invention, "Intelligent Manufacturing System Based on Digital Twin", are described as follows:

[0018] Taking a precision machining center as an example, the intelligent machining device of a production machine or processing equipment integrates cutting tools and multiple IoT sensors (such as "spindle speed", "feed rate", "workpiece position", "axis force, torque or bending moment of the cutting tool", "vibration value of the production machine or processing equipment", "current" and "temperature") at the front end of the cutting center to monitor multi-dimensional data of the machining process. Data collection: The sensors monitor key parameters in real time during the machining process, such as changes in machining force, torque, bending moment and temperature, and collect operational and machining status data in real time.

[0019] IPC Industrial Computer Data Integration and Monitoring: The industrial computer can receive sensor data from multiple (e.g., 1 to 10) production machines or processing equipment, such as cutting centers, and perform data integration and monitoring. Production Machine or Processing Equipment Operation Control: Performs preliminary data analysis and issues control commands based on the analysis results to achieve real-time monitoring and adjustment of the operation of production machines or processing equipment.

[0020] Cloud Server Data Storage and Computation: Supports large-scale data storage and provides efficient data processing and computing capabilities, serving as the basic computing resource for digital twin platforms. Supports Digital Twin Platforms: Provides a simulation environment for digital twin platforms, supporting data modeling and dynamic computation.

[0021] Digital Twin Platform Virtual-Real Integration Model Processing Simulation: Based on big data from cloud servers, a virtual model is established to simulate the processing flow, accurately reflect the operating status of physical equipment, and perform operational analysis. Anomaly Diagnosis and Optimization Suggestions: Anomaly diagnosis is performed through model analysis, processing trends are predicted, and specific optimization suggestions are generated.

[0022] Closed-loop control system decision feedback: Optimization suggestions and adjustment parameters generated by the digital twin platform are transmitted to the IPC industrial computer via a cloud server. Parameter adjustment: The IPC adjusts the operating parameters of the production machine or processing equipment in real time based on the optimization suggestions to ensure processing efficiency and accuracy. Model update: After the parameters of the production machine or processing equipment are adjusted, the new processing data is transmitted to the cloud again to update the digital twin model and maintain the virtual-real synchronization state.

[0023] Efficient bidirectional communication is achieved between various devices within the system, supporting smooth exchange of data transmission and control commands, as detailed below: (1). Production machine or processing equipment - IPC industrial computer: bidirectional communication, the production machine or processing equipment will transmit real-time sensing data to the IPC, and the IPC will adjust the processing parameters according to the commands. (2). IPC industrial computer - cloud server: bidirectional communication, the IPC will upload the integrated data to the cloud server and receive the analysis results and optimization suggestions generated by the cloud server. (3). Cloud server - digital twin platform: internal communication, the cloud server provides data and computing resources to support the model and simulation analysis of the digital twin platform. (4). Digital twin platform - IPC industrial computer: through the cloud server, the optimization suggestions generated by the digital twin platform are fed back to the IPC, realizing the adjustment and closed-loop control of the parameters of the production machine or processing equipment.

[0024] Please refer to Figure 2. The overall operation flow of the "Smart Manufacturing Method Based on Digital Twin" of this invention is summarized as follows: A1. Scene IoT: Data collection and transmission; Sensors on production machines or processing equipment monitor and collect data in real time, and transmit it to IPCs and cloud servers. A2. Digital synchronization: Virtual model construction and simulation; The digital twin platform uses big data to construct a virtual model and simulate the actual processing process. A3. Decision feedback: Data regression analysis and decision support; Anomaly diagnosis, prediction of processing trends, and generation of optimization suggestions are performed. A4. Closed-loop control: Parameter adjustment and closed-loop control; The expert system adjusts the parameters of the production machines or processing equipment, updates the virtual model, forms closed-loop control, and continuously optimizes the processing flow. Table 1 Flowchart of the "Smart Manufacturing Method Based on Digital Twin" flow Procedure step illustrate A1 Scene IoT Real-time monitoring Install sensors on production machines or processing equipment, including sensors for "spindle speed", "feed rate", "workpiece position", "tool axial force or torque or bending moment", "vibration value of production machine or processing equipment", "current" and "temperature", to monitor key parameters during the processing. Data transmission The collected data is synchronously transmitted to IPC industrial computers and cloud servers to ensure data integrity and timeliness. A2 Digital synchronization Big data processing The cloud server stores and processes the large amount of sensor data received. Virtual-Real Integration Model Digital twin platforms use processed data to create virtual models that accurately simulate the actual processing of production machines or processing equipment. Process simulation Dynamically simulate processing scenarios and predict possible anomalies to provide basic support for subsequent decision analysis. A3 Decision Feedback Data Regression Analysis By utilizing big data and regression analysis techniques, we can conduct in-depth analysis of machining parameters and monitoring data such as cutting force, torque, and vibration. Abnormal diagnosis Diagnose abnormal conditions during the machining process, such as abnormal vibration and temperature fluctuations, determine the causes and their impact range, machining parameter deviations, or tool wear. Trend Forecast Predict future processing trends and potential problems, and generate a future trend report. Optimize decision generation Based on the analysis results, specific optimization suggestions are generated, such as parameter adjustment or tool replacement schemes. A4 Closed-loop control Expert system dynamic adjustment Based on the optimization suggestions of the digital twin platform expert system, operators can automatically or semi-automatically adjust parameters of production machines or processing equipment (such as spindle speed (RPM), feed rate (mm / sec), cutting fluid flow rate, etc.) to improve processing efficiency and accuracy. Virtual model iterative optimization The adjusted processing data is then transmitted back to the cloud server to dynamically update the virtual model, maintaining synchronization between the virtual and real systems and forming a closed-loop control system for continuous optimization.

[0025] In A3. Decision Feedback - Data Regression Analysis, regression analysis, as the core tool of data analysis, helps the system establish a mathematical relationship model between operating parameters and processing results. It can not only diagnose anomalies in the processing process, but also predict processing trends and provide parameter optimization suggestions. Its application scope includes, but is not limited to, the analysis and prediction of influencing factors of tool wear, processing accuracy, and energy efficiency. The application steps of regression analysis are explained as follows: (1). Establish a mathematical model: Use a multiple linear regression model to establish a correlation between different independent variables (such as cutting speed, feed rate, and tool wear degree) and target dependent variables (such as surface roughness). For example: Y = β0 + β1 X1 + β2 X2 + β3 X3 + ϵ Y: surface roughness (µm) X1: cutting speed (m / min) X2: feed rate (mm / s) X3: tool wear (mm) ϵ: error caused by unconsidered factors (2). Model training and verification: train the regression model through historical machining data, and verify the model using on-site experiments or machining data. (3). Application and optimization: predict the best combination of machining parameters based on the regression model, and provide intelligent parameter optimization suggestions or early warning of abnormal situations.

[0026] Taking a machining center using high-hardness cutting tools to manufacture precision parts as an example, the sensor collects multidimensional data and obtains the following mathematical model through regression analysis (Y = 5 + 0.02X1 + 0.1X2 + 0.5X3). This model indicates that the surface roughness (Y) of the machined surface will increase significantly with the feed rate (X2) and the degree of tool wear (X3).

[0027] Analysis revealed that when tool wear exceeds 0.3 mm or feed rate exceeds 0.5 mm / s, the surface roughness will exceed the allowable range (>10 µm). The system provides optimization suggestions such as reducing the feed rate to 0.4 mm / s or lower. When tool wear approaches 0.25 mm, the system issues an alarm and recommends tool replacement.

[0028] In the A4. Closed-loop control - Expert System, the knowledge base and rule base of the expert system are used to transform machining data into actual control commands, realize closed-loop iteration of parameter adjustment, and ensure the efficiency and stability of production or processing. The operating principle is explained as follows: (1). Knowledge Base: contains data on material properties, tool performance, machining conditions and common abnormal situations, and is constructed based on expert experience and historical machining data. (2). Inference Engine: uses condition-action rules in the rule base, combined with real-time data analysis, to diagnose abnormal situations in the machining process and propose solutions. (3). Closed-loop control: generates control commands and provides operators with automatic or semi-automatic real-time adjustment of cutting parameters (such as cutting speed and feed rate), synchronously updates the digital twin model, and maintains the continuity of virtual and real integration.

[0029] Taking the machining of high-hardness materials—nickel-based alloys—in a factory as an example, the cutting tool is subjected to extremely high cutting forces, which may lead to rapid tool wear or surface roughness exceeding specifications. The sensor detects in real time that the cutting force exceeds the specified threshold (e.g., 40 Nm), accompanied by a sharp rise in machining temperature.

[0030] The expert system's reasoning engine generates an automatic adjustment solution based on rules in the knowledge base (e.g., "When the cutting force is too high, reduce the feed rate and increase the coolant flow rate"). The system prompts the operator to reduce the feed rate by 20%, increase the coolant flow rate by 15%, and update the virtual model to simulate the adjusted machining effect.

[0031] Actual results show that the machining roughness has recovered to the specified range (e.g., Ra = 0.8μm), and the tool life is estimated to be extended by 20%. The system records the data of this adjustment to the knowledge base, optimizes the machining suggestions for nickel-based alloys in the rule base, and improves the response efficiency for similar machining tasks in the future.

[0032] This system has the following three core features to achieve real-time monitoring, synchronous optimization and efficient resource utilization of the processing flow: (1). Real-time: The system fully emphasizes the ability to collect and feedback data in real time, ensuring that any changes in parameters during the processing (such as tool status, "spindle speed", "feed rate", "object position", "axis force or torque or bending moment of the tool", "vibration value of production machine or processing equipment", "current" and "temperature") can be quickly captured and processed. Real-time monitoring reduces processing errors caused by delayed response to abnormal situations and ensures the stability of the processing flow. (2). Synchronization: The virtual model and physical equipment maintain real-time synchronous operation to ensure that the virtual simulation results accurately reflect the physical processing status. Synchronization improves the accuracy and effectiveness of the decision-making process and helps operators quickly grasp the overall picture of the processing site. (3). Optimization goals: Through parameter adjustment and closed-loop control, the system effectively improves processing efficiency and accuracy, while reducing tool wear and processing energy consumption. It further realizes efficient resource utilization, significantly reduces processing failures and production or processing costs, and creates greater benefits for the smart manufacturing environment.

[0033] The following will describe the actual application scenarios and technical details of each device in the system in sequence:

[0034] Please refer to Figure 3 for the "intelligent machining device" of the production machine or processing equipment. It integrates high-precision sensing technology, dynamic analysis and real-time communication functions to provide solutions for the core pain points in traditional cutting center machining (such as cutting force variation detection, tool wear monitoring and machining efficiency improvement). This device is suitable for fields such as precision manufacturing, aerospace industry and large-scale mass production. By monitoring the tool force and machining status in real time, it realizes automated and data-driven intelligent manufacturing processes.

[0035] The technical details of the "intelligent processing device" are as follows:

[0036] Core sensing module: The intelligent machining device has embedded sensors that can monitor the force changes of the tool in real time during the machining process, including cutting force and torque measurement. The tool force change range is 35~45 Nm, with a resolution of 1~5%, which can capture the strain characteristics of the highly dynamic machining process; it supports center runout and cutting edge balance analysis, calculates runout and cutting edge balance based on the tool force distribution, and is used to evaluate tool stability.

[0037] Data processing and transmission: Sensor data is transmitted to the machine-side IPC industrial computer via the built-in wireless module, and key processing indicators (such as torque change rate and runout) are generated in real time and updated to the control system in a synchronous manner to realize automatic parameter adjustment.

[0038] Application Integration: When integrated with the controller of production machine or processing equipment, it can provide real-time feedback on changes in cutting force, support dynamic optimization of feed rate, and avoid excessive tool wear; when used with offline detection system, it provides multi-view tool wear images, establishes a comparative analysis of tool wear before and after processing, and further optimizes processing parameters and tool life model.

[0039] The actual application scenarios of the "intelligent processing device" are as follows:

[0040] High-precision aerospace parts machining: When machining high-strength materials such as nickel-based alloys and titanium alloys, intelligent machining devices can monitor cutting force and tool stress uniformity in real time; improve the machining hole diameter accuracy to ±10 µm range, reduce the workpiece scrap rate by about 5~15%, and effectively reduce tool breakage and hole diameter runout problems.

[0041] Large-scale production or processing of automotive parts: In mass production processing, real-time tool wear diagnosis is performed by combining the machine-side unit, and tool life is estimated through vibration and current sensing; Taking a certain factory's application as an example, when processing high-strength steel (FDAC), the system estimation error is less than 5~10%, downtime inspection time is reduced by 10~30%, and processing efficiency is increased by 1%~5%.

[0042] Tool life management: In conjunction with the detection system, wear quantitative analysis of high-priced tools is performed with an accuracy of 10~20 µm, and wear conditions at multiple angles are recorded, providing a more accurate basis for replacement; after application in a certain processing plant, the tool cost can be reduced by about 15~25% per year.

[0043] The intelligent machining equipment will further integrate cloud servers and digital twin platforms, using big data and analytical models to achieve tool life prediction and anomaly analysis, thus constructing a comprehensive intelligent tool life management system. Its modular design can meet diverse machining needs, such as high-precision scenarios like medical devices and electronic components, promoting the comprehensive upgrade and application expansion of intelligent manufacturing.

[0044] Please refer to Figure 4, "Digital Twin Platform Operation Flowchart." The digital twin platform is an intelligent system architecture that integrates sensing, data transmission, simulation analysis, application interaction, and control functions to achieve synchronous operation of physical processing equipment and virtual models, thereby improving processing efficiency, reducing costs, and ensuring processing quality. The following is a layered analysis of the platform's functions:

[0045] B1. Sensing Layer: Data Collection and Real-time Monitoring. Multiple high-precision sensors (such as "spindle speed", "feed rate", "object position", "tool axial force, torque, or bending moment", "vibration value of production machine or machining equipment", "current", and "temperature") are installed in the "intelligent machining device" to collect multi-dimensional machining data, including cutting force, vibration, and temperature. During machining on the production machine or machining equipment, the force changes on the tool are monitored in real time, effectively preventing tool breakage or abnormal machining conditions.

[0046] B2. Network Layer: Data Transmission and Security. Sensing data is transmitted via wired (Ethernet) or wireless (Wi-Fi, Bluetooth, or RF) networks to the on-site IPC industrial computer or cloud platform. This ensures secure data transmission, supports real-time analysis and anomaly notifications, and guarantees low latency and high stability.

[0047] B3. Computational Layer: Data Processing and Simulation Analysis. IPC industrial computers perform data denoising and feature extraction, while the cloud platform handles large-scale historical data regression analysis and model optimization. It detects tool wear and updates the tool life model, providing accurate data support to optimize machining parameters.

[0048] B4. Platform Layer: Digital Twin Modeling and Management. Real-time import of sensor data into the virtual model, through integration of computer graphics or other digital modeling methods, establishes a digital twin virtual model synchronized with the physical equipment. This model utilizes big data analytics to simulate production processes or machining behaviors, achieving dynamic simulation of tool life and workpiece real-time status under cutting conditions, and providing predictive indicators for machining quality. Simultaneously, the system displays simulation results through a real-time visual interface, providing decision-making support and improvement suggestions for on-site operators and managers, thereby optimizing the machining process and improving production efficiency.

[0049] B5. Application Layer: Human-Machine Interface and Decision Feedback. Through a web-based user interface supporting interactive functions, the processing status (such as workpiece status, tool stress changes, and wear indicators) can be displayed on industrial computers, intelligent mobile vehicles, or XR mixed reality display devices, and maintenance suggestions and operation reports can be generated. Tool wear data is displayed in real-time on the interface, assisting operators in real-time monitoring and operation, quickly determining when to replace tools, and improving decision-making efficiency.

[0050] B6. Control Layer: Closed-Loop Control and Real-Time Adjustment. Based on digital twin simulation results, adjustment commands are sent semi-automatically or automatically to the production machine or machining equipment system to achieve dynamic optimization of machining parameters. When excessive tool center runout or other abnormal conditions are detected, the expert system suggests that the operator reduce the feed rate or adjust the cooling strategy to ensure machining accuracy and stability.

[0051] The following is a layered architecture and functional description of the "Digital Twin Platform", combined with practical application examples to demonstrate its operation and benefits: Table 2 Layered Architecture, Functional Description and Application Examples of Digital Twin Platform Layer Function Description Application Cases Perception layer Using sensors (such as "spindle speed", "feed rate", "object position", "tool axial force or torque or bending moment", "vibration value of production machine or machining equipment", "current" and "temperature") to monitor data in real time, providing dynamic monitoring of tool status. Real-time cutting force and vibration monitoring: predicts tool wear or breakage, issues early warnings, and ensures machining safety and stability. Network layer Data is transmitted to edge devices or the cloud via wired or wireless networks, ensuring low latency and high stability in data transmission. Secure data transmission and notification: Enables immediate notification of processing anomalies and data integrity protection, ensuring data accuracy and reliability. Computation layer The machine-side IPC is responsible for data preprocessing, while the cloud performs large-scale data analysis and model optimization to generate tool life and wear indicators. Tool life prediction: Based on historical and real-time data, update the tool life model to provide accurate wear prediction and reduce maintenance costs. Platform layer Integrate data and models to build digital twins, simulate processing procedures, and provide an instant visualization interface. Machining simulation and quality monitoring: Simulates the force on the cutting tool and the machining process, providing predictions and indicators of workpiece machining quality, and reducing the defect rate. Application layer It displays machining data and tool wear indicators, provides maintenance suggestions and decision feedback, and supports operators in making real-time judgments and actions. Operation optimization suggestions: Real-time display of tool wear data and suggestions in the operation interface of production machines or processing equipment, enabling operators to quickly respond to abnormal situations. control layer Based on the results of digital twin simulation, machining parameters (such as cutting speed and feed rate) are adjusted semi-automatically or automatically to achieve closed-loop control. Dynamic parameter optimization: When excessive tool center runout is detected, the expert system suggests that the operator reduce the feed rate to ensure machining accuracy and extend tool life.

[0052] The following are examples of different manufacturing or processing scenarios, demonstrating the specific effectiveness of digital twin technology in optimizing tool life, improving processing quality, and stabilizing manufacturing or processing: Table 3: Description and Benefits of Digital Twin Platform Cases Scene Case Description Achieve benefits Aerospace parts processing Cutting force variation data was collected and transmitted to a computing device for anomaly analysis. The results showed that tool wear was approaching a critical value. The digital twin system simulated tool life under different machining conditions, provided optimal tool change recommendations, and successfully prevented tool breakage. Avoid machining losses: predict tool wear, replace tools in advance, and reduce the risk of tool breakage; Improve precision: Ensure that the machined surface meets high precision requirements and reduce the defect rate. Milling of automotive engine parts Simulations of the effects of different feed rates on cutting temperature and tool life showed that reducing spindle speed can extend tool life. When applied to real-world machining scenarios, this successfully reduced tool change frequency and improved manufacturing or machining efficiency. Extend tool life: reduce tool replacement frequency and lower consumable costs; Stable production or processing: Optimize cutting parameters, improve processing stability, and reduce downtime. High-precision mold processing Due to variations in material hardness, the cutting force abnormally increased. The system automatically reduced the feed rate to stabilize cutting conditions, and feedback data showed that the surface roughness of the machined surface returned to normal. At the same time, the case was recorded to improve the predictive ability of the big data model. Improve machining quality: Control cutting force to ensure surface roughness meets standards; Data-driven optimization: Record case studies to improve model accuracy and achieve more efficient processing parameter recommendations.

[0053] Please refer to the “Actual Machining Scene” shown in Figure 5 and the “Digital Twin Simulation” shown in Figure 6 to show the comparison between the machining site (Figure 5) and the virtual model (Figure 6) generated by digital twin technology. Through virtual simulation driven by real-time sensing data, the dynamic interaction process between the tool and the workpiece is reconstructed.

[0054] Figure 7, "Digital Twin Platform and Application Control," depicts the overall architecture of the digital twin platform from sensor data collection to application-layer decision-making, showcasing the data flow and control loop in a smart factory. The following examples illustrate the functional descriptions and corresponding value. Table 4: Examples of Digital Twin Platform Functions and Application Values Interface Project Function Description Application value Energy Status Monitor the energy consumption of production machines or processing equipment to achieve real-time energy management and analysis. Ensure energy efficiency in the processing, reduce operating costs, and support sustainable manufacturing. Temperature Change It displays temperature changes during the processing and can promptly detect abnormal heat sources or overheating problems in production machines or processing equipment. Improve processing stability, prevent workpiece quality problems caused by temperature fluctuations, and ensure the long-term stable operation of production machines or processing equipment. Coordinate Information Provides machining space coordinates (X, Y, Z) to accurately record the relative position of the tool and the workpiece. It provides an accurate foundation for digital synchronization and virtual models, supports the construction of integrated virtual and physical models, and improves the accuracy of machining positioning. Realtime Information Torque: Real-time monitoring of tool torque changes. Xbend, Ybend: Displays the bending moment data along the X and Y axes. It enables full monitoring of the load status of production machines or processing equipment, optimizes the adjustment of processing parameters, and improves tool life and processing efficiency. Product Information Product: Processed product identification (name code). Status: Processing status updated. Alarm Level: Alarm level. Improve production management efficiency, support immediate response and handling of abnormal situations, and avoid production or processing interruptions and cost losses. Handler Information Remaining Usage Count: Displays the remaining number of times the tool can be used. It supports dynamic management of tool life and replacement recommendations to ensure the continuity of production or machining processes and the stability of machining accuracy. Handler Detail Handler Number: Identifies the tool number. Radius Comp. & length Comp.: Displays the radius compensation and length compensation values. Radius Wear: The degree of wear on the cutting tool. Switch Handler: Tool switching. It provides detailed tool information to support tool management and maintenance strategies, supports multi-tool operation needs, and improves machining flexibility and efficiency. Control and Adjustment Based on data and simulation results, machining parameters such as cutting speed and feed rate are adjusted in real time. It enables real-time feedback and adjustment of closed-loop control, improves machining accuracy and efficiency, and reduces tool and equipment wear, meeting the optimization needs of smart manufacturing. Digital Twin Simulation and Feedback By utilizing a digital twin platform to simulate anomalies and provide optimization suggestions, a decision-making system that combines virtual and real elements is formed. Support the full operation of smart factories, realize data-driven efficient production or processing modes, and meet the needs of complex processing scenarios by combining real-time monitoring and prediction.

[0055] The digital twin platform provides users with comprehensive decision support and processing control through data integration and real-time display via the aforementioned interface. Its main advantages include: (1) Immediate and accurate processing data: helping operators grasp the processing status and respond promptly to abnormal situations. (2) Intelligent tool management: enabling full-process tracking and wear analysis of tool life, reducing production or processing costs. (3) Systematic decision feedback: optimizing production or processing parameters based on synchronous analysis and display of multi-dimensional data, improving production efficiency and processing quality. This platform provides an efficient solution for data-driven and closed-loop control in production or processing scenarios, and is the core of smart manufacturing and digital twin technology applications.

[0056] This invention patent proposes a smart manufacturing method based on digital twins. By integrating production machines or processing equipment, intelligent processing devices, IoT sensors, industrial computers, cloud servers, and a digital twin platform, it achieves virtual-physical integration, real-time monitoring, and dynamic optimization of the processing flow. The system uses multi-dimensional sensing technology for scene IoT and combines cloud computing for data analysis and simulation to generate an accurate digital twin model, thereby performing anomaly diagnosis, processing trend prediction, and optimization decision feedback.

[0057] This invention possesses three core characteristics: real-time performance, ensuring immediate data collection and feedback during the processing; synchronization, maintaining dynamic synchronization between the virtual model and physical equipment; and optimization objectives, continuously improving processing efficiency, accuracy, and resource utilization through closed-loop control. Its technological innovation lies in the data-driven dynamic model and real-time optimization, solving problems such as insufficient monitoring range, model update delays, and feedback limitations in existing technologies. It is particularly suitable for precision manufacturing, high-precision machining, and smart factory applications, meeting the demands of the smart manufacturing field for high efficiency, high flexibility, and sustainability. [Simplified Explanation of the Diagram]

[0058] Figure 1: Block diagram of "Smart Manufacturing System Based on Digital Twin". Figure 2: Flowchart of "Smart Manufacturing Method Based on Digital Twin". Figure 3: Schematic diagram of intelligent processing device for production machines or processing equipment. Figure 4: Operation flowchart of digital twin platform. Figure 5: Actual processing scenario. Figure 6: Digital twin simulation diagram. Figure 7: Digital twin platform and application control.

Claims

1. A smart manufacturing method based on digital twins, applied to the operation and process control of production machines or machining equipment, wherein the decision feedback includes the following steps: Data regression analysis: Based on real-time and historical multidimensional data collected during the machining process, a correlation model between machining parameters and results is established using big data and regression analysis techniques to obtain mathematical relationships describing changes such as cutting speed, feed rate, tool wear, and surface roughness; Anomaly diagnosis: Based on the established regression model, the operating data of the production machine or machining equipment is analyzed to diagnose abnormal states and determine the causes and scope of impact of the anomalies; Trend prediction: Based on the regression model output and prediction results, the tool wear trend, changes in machining efficiency, and equipment health status are analyzed to generate a future trend report of machining behavior; Decision generation: Based on the trend prediction results, suggestions for optimizing machining parameters are provided, including but not limited to spindle speed, feed rate, cooling strategy, or tool replacement scheme, for use in subsequent control stages; Each step uses a digital twin platform as the computing core and synchronizes data and updates models through industrial computers and cloud servers.

2. A smart manufacturing method based on digital twins, applied to parameter adjustment and model iteration of production machines or processing equipment, wherein closed-loop control includes the following steps: Expert system support: Combining the digital twin platform and expert system database to provide optimal solutions for abnormal situations or analysis results; Real-time parameter adjustment: Based on expert system suggestions and data analysis results, operators can semi-automatically or automatically dynamically adjust the processing parameters of production machines or processing equipment to stabilize the processing status and improve processing accuracy; Virtual model update: Closed-loop control realizes bidirectional transmission of real-time data, sending processing data back to the cloud server to update the digital twin virtual model, keeping it synchronized with the physical processing status; Iterative optimization: Through multiple model updates and iterative adjustments, continuous improvement in processing efficiency and stability is achieved, forming a continuously optimized closed-loop control process.

3. According to the smart manufacturing method described in Request 1, the results of decision generation feedback include multiple benefits: Machining parameter optimization: Based on regression analysis and expert advice, optimal parameters such as spindle speed and feed rate are generated to maximize machining efficiency; Data-driven real-time diagnostics: Based on data analysis, real-time anomaly handling suggestions are provided, such as increasing cutting fluid flow rate and speed, or duration, to reduce machining delays caused by equipment anomalies; Machining trend reports: Predictive machining trend data is provided, such as machining path optimization parameter adjustment strategies, to shorten machining time or pursue higher product quality, assisting operators in pre-adjusting production or machining plans and improving the responsiveness of the smart manufacturing environment; Anomaly handling and maintenance suggestions: Using regression analysis results, maintenance plans for production machines or machining equipment and tools are generated, including wear replacement suggestions and health check schedules; 4. According to the intelligent manufacturing method described in Request 2, the closed-loop control achieves the following specific results: Improved machining quality stability: Semi-automatic or automatic adjustment of machining parameters reduces the error amplitude during the cutting process, enabling machining accuracy to reach the target range; Extended tool life: Through real-time parameter optimization and cooling strategy adjustment, tool wear and breakage are reduced, extending tool life; Improved production or machining efficiency: Closed-loop control shortens anomaly handling time, improves machining stability, and reduces the risk of production or machining interruptions; Iterative updates of the data model: Data results from the closed-loop control process are transmitted back to the digital twin platform to continuously optimize the accuracy and adaptability of the virtual model.