Digital twinning and AI auxiliary decision-making method and device in production line life cycle management

Abstract

The invention relates to a digital twinning and AI auxiliary decision-making method and device in production line life cycle management. The method comprises the following steps: constructing a digital twinning system comprising a physical production line, a virtual production line, a twinning database, service management and connection; data acquisition equipment of each link of a physical production line is utilized to acquire equipment parameters and production process data according to a preset time interval, and the equipment parameters and the production process data are preprocessed, classified and stored in a twin database. According to physical characteristics of equipment, attributes such as rigid bodies are set for physical and virtual equipment, a mapping relation is established, and key parameters are pushed. And adjusting the position and operation of the virtual production line component based on the mapping relation, and storing and evaluating virtual production line data. The AI algorithm is introduced to analyze data, mine potential rules, generate decision suggestions and display information through a human-computer interaction interface, so that intelligent management of the whole life cycle of the production line is realized, the efficiency is improved, the cost is reduced, and the management progress of the industrial production line is promoted.

Description

Technical Field

[0001] The present invention relates to the field of industrial automation and intelligent manufacturing technology, and in particular to a digital twin and AI-assisted decision-making method and device in production line lifecycle management. Background Art

[0002] In modern industrial production, production line management faces many challenges. Traditional production line management methods mainly rely on manual experience and simple automated control, which has many limitations. On the one hand, the operating status and performance monitoring of physical production lines are often not comprehensive and real-time, resulting in equipment maintenance, which is often carried out after the failure occurs, and it is impossible to predict and prevent it in advance, thus affecting production efficiency and product quality, and increasing equipment downtime and maintenance costs. On the other hand, in terms of production plan adjustment and production line design optimization, due to the lack of effective analysis and mining of large amounts of data, decisions are often conservative and unscientific, and the mutual influence and long-term benefits of various factors cannot be fully considered, resulting in insufficient resource utilization and difficulty in adapting to rapidly changing market demands and product upgrades.

[0003] With the rise of Industry 4.0 and intelligent manufacturing, digital twin technology and artificial intelligence technology have developed rapidly. Digital twin technology provides an accurate virtual mirror for the physical production line, enabling the monitoring and management of the production line to be simulated and analyzed in a virtual environment. However, in some existing digital twin applications, the real-time, accuracy and stability of data synchronization between physical and virtual production lines are still insufficient. Especially in complex industrial environments, how to ensure the integrity and timeliness of data, and how to make full use of the data in the digital twin system for deep mining and decision support are still urgent issues to be solved. Summary of the invention

[0004] The main purpose of the present invention is to provide a digital twin and AI-assisted decision-making method and device in production line life cycle management, so as to realize intelligent management of the entire life cycle of the production line, improve production efficiency, reduce costs, enhance the adaptability of production line management and the effectiveness of decision-making, and promote the development of industrial production line management in a more intelligent, efficient and precise direction.

[0005] To achieve the above object, the present invention provides a digital twin and AI-assisted decision-making method in production line lifecycle management, comprising the following steps: Build a digital twin system including physical production lines, virtual production lines, twin databases, service management and connections. The twin database is the core of the entire digital twin system and is used to manage all data related to the digital twin system. Using data collection equipment deployed at various links of the physical production line, the equipment parameters, production process data, environmental data, and logistics data collected periodically at predetermined time intervals are pre-processed, and the data is classified according to pre-set data classification rules and transmitted to the twin database for storage and management; According to the physical characteristics of each device in the physical production line, set the rigid body, collision body properties and corresponding kinematic pairs and constraints of the physical and virtual production line devices. According to the type, specification and performance of the equipment, set the corresponding property settings of each device, and establish the mapping relationship between the physical and virtual devices. Receive and parse key parameters in the physical production line data, and push the key parameters to corresponding equipment components on the virtual production line, wherein the key parameters include position information and speed parameters of each equipment component of the physical production line; Based on the status information of each equipment component in the virtual production line and the relative position information of each equipment component in the physical production line, according to the established mapping relationship, the corresponding position of each equipment component in the virtual production line is adjusted, and it is operated according to the speed parameters of each equipment component. At the same time, the virtual production line data during the operation of the virtual production line is obtained and stored in the twin database. The operation status of the virtual production line is evaluated based on the virtual production line data. If it is found that the performance of the virtual production line deviates from that of the physical production line, the model parameters of the virtual production line are adjusted; Introduce AI algorithms to analyze the data in the twin database, select appropriate algorithms according to the type of data and the purpose of analysis, and mine potential rules in the data. The potential rules include but are not limited to periodic changes in equipment performance, the impact of process parameters on product quality, the impact of environmental factors on equipment and production processes, and potential bottlenecks in logistics. Generate decision-making suggestions for production line design optimization, production plan adjustment, equipment maintenance prediction and quality control based on the analysis results of AI algorithms; The operating status, data analysis results and decision-making suggestions of the virtual production line are displayed through a human-computer interaction interface. The human-computer interaction interface provides a real-time update function, and according to the feedback decision operations, the specific operation instructions are transmitted to the physical production line and the virtual production line to manage the entire life cycle of the production line.

[0006] Furthermore, in the digital twin system, the physical production line includes multiple interconnected production equipment, and the virtual production line corresponds one-to-one to the production equipment in the physical production line through a predefined mapping relationship. The twin database adopts a distributed storage architecture to store various types of data collected from the physical production line, data generated during the operation of the virtual production line, and historical operation data of the physical production line and the virtual production line. The services include data processing services, decision services and communication services, which are used to realize data analysis and processing, decision generation and information interaction between system components. The service management adopts a structured programming method, based on S7-1200 using the LAD\SCL programming language, to create an electrical control program, debug and run it in the physical entity and the digital twin system to achieve virtual and real synchronization. The electrical control program includes communication and data analysis modules, loading and unloading and processing flow control modules, and RFID read-write control and warehouse management modules. The connection performs network node access to each device in the production line, and uses respective private protocols to realize real-time communication within the entire digital twin system through industrial Ethernet.

[0007] Furthermore, the digital twin system also includes, by adopting a machine learning algorithm, using historical data and real-time data to train and optimize the neural network model, adjusting the parameters and structure of the model according to the actual production situation, obtaining a virtual production line model for simulating the operating status and performance of the physical production line, and realizing real-time data interaction and status synchronization between the physical production line and the virtual production line through connection.

[0008] Furthermore, a dynamic adjustment mechanism for the mapping relationship between physical devices and virtual devices is established, and when the physical device is updated or replaced, the mapping relationship is automatically updated according to the information of the new device.

[0009] Furthermore, during the data synchronization process between the physical production line and the virtual production line, when data delay or data loss occurs, the automatic error correction mechanism is triggered. According to the degree of data delay and the scale of data loss, data backup is used for recovery first. When the backup data is unavailable, data retransmission is used. For a small amount of data loss, a data interpolation algorithm is used to compensate. The performance of the automatic error correction mechanism is monitored and evaluated, and optimized through the set timeliness and accuracy indicators.

[0010] Furthermore, the AI ​​algorithm is an integrated algorithm set, including but not limited to cluster analysis algorithms, association rule mining algorithms, and time series prediction algorithms. The corresponding algorithms are selected to analyze data according to different analysis tasks.

[0011] Furthermore, after generating the decision proposal, based on the various factors affecting the production line, the impact of different decision plans on the short-term and long-term benefits of the production line is comprehensively considered, and each decision plan is comprehensively evaluated and prioritized. The various factors include but are not limited to production line production goals, cost-effectiveness, resource conditions, and product quality.

[0012] Furthermore, the human-computer interaction interface uses visualization technology to display the operating status of the production line and data analysis results in the form of charts, 3D models, and animations, while providing interactive operation elements to customize the display content and operation interface layout according to the user's own needs.

[0013] Furthermore, the data twin system has automatic learning and self-adaptation functions. According to the long-term operation data of the production line, it automatically updates and optimizes the parameters and decision-making rules of the AI ​​algorithm, and automatically switches the analysis and decision-making modes according to the operation stage of the production line. The production line operation stage includes the startup stage, the stable operation stage and the fault stage.

[0014] The present invention also provides a digital twin and AI-assisted decision-making device in production line lifecycle management, including: System building module, used to build a digital twin system including physical production line, virtual production line, twin database, service management and connection; The data acquisition module is used to use the data acquisition equipment deployed in various links of the physical production line to pre-process the collected physical production line data according to the equipment parameters, production process data, environmental data and logistics data collected periodically at predetermined time intervals, and classify the data according to the pre-set data classification rules and transmit it to the twin database for storage and management; The mapping establishment module is used to set the rigid body and collision body properties of the physical production line equipment and the virtual production line equipment and the corresponding kinematic pairs and constraints according to the physical characteristics of each device in the physical production line, set the corresponding property settings of each device according to the type, specification and performance of the device, and establish the mapping relationship between the physical device and the virtual device; A data analysis module, used to receive and analyze key parameters in the physical production line data, and push the key parameters to corresponding equipment components on the virtual production line; The virtual operation module is used to adjust the corresponding positions of the equipment components in the virtual production line based on the status information of each equipment component in the virtual production line and the relative position information of each equipment component in the physical production line according to the established mapping relationship, and operate according to the speed parameters of each equipment component. At the same time, the virtual production line data during the operation of the virtual production line is obtained and stored in the twin database. The operation status of the virtual production line is evaluated based on the virtual production line data. If it is found that the performance of the virtual production line deviates from that of the physical production line, the model parameters of the virtual production line are adjusted; The AI ​​analysis module is used to introduce AI algorithms to analyze the data in the twin database, select appropriate algorithms according to the type of data and analysis purpose, and mine potential rules in the data; A decision-making support module, which is used to generate decision-making suggestions for production line design optimization, production plan adjustment, equipment maintenance prediction and quality control based on the analysis results of AI algorithms; The production line management module is used to display the operating status, data analysis results and decision-making suggestions of the virtual production line through a human-computer interaction interface. The human-computer interaction interface provides a real-time update function. According to the feedback decision operations, specific operation instructions are transmitted to the physical production line and the virtual production line to manage the entire life cycle of the production line.

[0015] The digital twin and AI-assisted decision-making method and device in the production line lifecycle management provided by the present invention have the following beneficial effects: by collecting physical production line data and introducing AI algorithm analysis, the present invention can detect periodic changes in equipment performance and potential failures in advance, realize predictive maintenance, reduce downtime and maintenance costs, and extend equipment life. The physical production line and the virtual production line achieve precise virtual-real synchronization through mapping and real-time synchronization mechanisms, set rigid body and other properties and real-time data updates to ensure that the virtual production line reflects the physical state truthfully. At the same time, the virtual environment simulation test can be performed through the visual interface to avoid the risks of physical experiments. Integrate a variety of AI algorithms, select algorithms to mine data on demand, generate decision recommendations, and then conduct comprehensive evaluation and ranking. By integrating multiple factors, the decision is made more scientific and forward-looking. The data twin system in the present invention can automatically update and optimize algorithm parameters and decision rules according to the long-term data of the production line, switch analysis and decision-making modes according to stages, and enhance the adaptability and effectiveness of decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of a digital twin and AI-assisted decision-making method in production line lifecycle management in one embodiment of the present invention; Figure 2 It is a structural block diagram of a digital twin and AI-assisted decision-making device in production line lifecycle management in one embodiment of the present invention; The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0018] Reference Figure 1, which is a flow chart of a digital twin and AI-assisted decision-making method in production line lifecycle management proposed by the present invention, comprising the following steps: S1, building a digital twin system including physical production lines, virtual production lines, twin databases, service management and connections. The twin database is the core of the entire digital twin system and is used to manage all data related to the digital twin system; S2, using the data collection equipment deployed in each link of the physical production line, pre-processing the collected physical production line data according to the equipment parameters, production process data, environmental data and logistics data collected periodically at predetermined time intervals, and classifying the data according to the pre-set data classification rules and transmitting it to the twin database for storage and management; S3, according to the physical characteristics of each device in the physical production line, set the rigid body, collision body properties and corresponding kinematic pairs and constraints of the physical production line equipment and virtual production line equipment, set the corresponding property settings of each device according to the type, specification and performance of the equipment, and establish a mapping relationship between the physical equipment and the virtual equipment; S4, receiving and parsing key parameters in the physical production line data, and pushing the key parameters to corresponding equipment components on the virtual production line, wherein the key parameters include position information and speed parameters of each equipment component of the physical production line; S5, based on the status information of each equipment component in the virtual production line and the relative position information of each equipment component in the physical production line, according to the established mapping relationship, adjust the corresponding position of each equipment component in the virtual production line, and operate according to the speed parameters of each equipment component. At the same time, obtain the virtual production line data during the operation of the virtual production line and store it in the twin database. Evaluate the operation status of the virtual production line based on the virtual production line data. If it is found that the performance of the virtual production line deviates from that of the physical production line, adjust the model parameters of the virtual production line. S6, introduce AI algorithms to analyze the data in the twin database, select appropriate algorithms according to the type of data and the purpose of analysis, and mine potential rules in the data. The potential rules include but are not limited to periodic changes in equipment performance, the impact of process parameters on product quality, the impact of environmental factors on equipment and production processes, and potential bottlenecks in logistics. S7, generates decision-making suggestions for production line design optimization, production plan adjustment, equipment maintenance prediction and quality control based on the analysis results of AI algorithms; S8, displays the operation status, data analysis results and decision suggestions of the virtual production line through the human-computer interaction interface. The human-computer interaction interface provides real-time update function, and transmits specific operation instructions to the physical production line and the virtual production line according to the feedback decision operation, so as to manage the whole life cycle of the production line.

[0019] As described in step S1 above, step S1 builds the infrastructure for the lifecycle management of the entire production line. In the digital twin system, the physical production line covers multiple interconnected production equipment and is the part that actually performs production tasks; the virtual production line corresponds to the production equipment in the physical production line one by one through a predefined mapping relationship, which is a digital mirror of the physical production line and can be used for simulation, analysis and optimization. The twin database is at the core and adopts a distributed storage architecture. It not only stores various types of data (equipment parameters, environmental data, logistics data, etc.) collected from the physical production line, but also stores the data generated during the operation of the virtual production line and the historical operation data of both, providing data support for the data analysis and decision-making of the system. Service management is the functional implementation unit of the system. It adopts a structured programming method and creates an electrical control program based on S7-1200 using the LAD\SCL programming language. The program includes communication and data analysis modules, loading and unloading and processing flow control modules, and RFID read-write control and warehouse management modules. It can be debugged and run in the physical entity and digital twin system to achieve virtual-real synchronization. The connection part is responsible for network node access for each device in the production line, and uses respective private protocols to achieve real-time communication within the entire digital twin system through industrial Ethernet, ensuring efficient transmission of data within the system and information interaction between system components, and ensuring close connection and collaborative work between physical and virtual production lines.

[0020] As described in step S2 above, data acquisition equipment is deployed in each link of the physical production line to collect equipment parameters (such as equipment operating speed, power, pressure, etc.), production process data (processing time, temperature, pressure and other process conditions), environmental data (temperature, humidity, vibration and other environmental indicators) and logistics data (material transportation and storage information) at fixed time intervals. The collected data must first be preprocessed, including cleaning, screening, denoising and other operations to ensure data quality. Then it is classified according to the set data classification rules, such as dividing equipment parameters into power equipment parameters, transmission equipment parameters, etc., and then the processed data is stored in the twin database, providing an orderly and accurate data basis for subsequent analysis and decision-making, and realizing the comprehensive, systematic collection and preliminary organization of physical production line information.

[0021] As described in step S3 above, for the unique physical characteristics of each device in the physical production line, such as device type (lathe, milling machine, robot, etc.), specifications (size, load capacity, etc.) and performance (maximum processing accuracy, operating speed range, etc.), set the corresponding rigid body, collision body attributes, kinematic pairs and constraints. For virtual production line equipment, set the corresponding attributes to ensure that there is an accurate mapping relationship between the two. For example, for a physical lathe, set the corresponding rigid body range and kinematic pairs according to its processing range, accuracy, etc., and the virtual lathe also sets the corresponding attributes, so that the virtual device can accurately simulate the physical behavior and motion state of the physical device in the virtual environment, establish a one-to-one correspondence between the physical device and the virtual device, and ensure that in subsequent data synchronization and simulation operations, the virtual device can truly reflect the behavior of the physical device.

[0022] As described in step S4 above, data collected from the physical production line and stored in the twin database is received, and key parameters are parsed from the data, including the position information and speed parameters of the equipment components. The key parameters are transmitted to the corresponding equipment components of the virtual production line, so that the position and speed information of the virtual equipment are synchronized with the physical equipment. For example, when the position and conveying speed of the goods on the physical conveyor belt change, the information is parsed and passed to the corresponding components on the virtual conveyor belt to ensure that the virtual production line is synchronized with the physical production line in real time in terms of position and speed, so that the virtual production line can immediately reflect the dynamic operation status of the physical production line.

[0023] As described in step S5 above, the status information and relative position information of the equipment components in the virtual production line and the physical production line are used to adjust the position of the equipment components of the virtual production line according to the mapping relationship established previously. For example, the position of the virtual device is adjusted according to the position of the physical device, and the virtual device is operated at the corresponding speed parameters to simulate the operation of the physical production line. During the operation of the virtual production line, corresponding data will be generated, and these data will be stored in the twin database. In addition, the operating status of the virtual production line will be evaluated, and its performance will be compared with the physical production line. If there is a performance deviation (such as the operating efficiency of the virtual device is inconsistent with the physical device), the model parameters of the virtual production line (such as the dynamic parameters, kinematic parameters, etc. of the model) will be adjusted to ensure that the virtual production line can accurately reflect the operating status of the physical production line, so that the virtual production line becomes a reliable digital mirror of the physical production line.

[0024] As described in step S6 above, select a suitable algorithm from the integrated AI algorithm set (including cluster analysis algorithm, association rule mining algorithm, time series prediction algorithm, etc.), and analyze the data according to the data type (time series data, classification data, etc.) and analysis purpose (such as performance analysis, quality association analysis, etc.) stored in the twin database. Through analysis, the potential rules in the data can be mined. For example, the time series prediction algorithm can be used to find the periodic change law of equipment performance over time, and the association rule mining algorithm can be used to find the potential relationship between process parameters and product quality, how environmental factors affect equipment performance and production process, and possible bottleneck problems in logistics. These potential rules can provide an in-depth understanding of the inherent mechanisms and potential problems of production line operation, and provide a basis for subsequent decision-making.

[0025] As described in step S7 above, based on the analysis results of the data by the AI ​​algorithm, decision suggestions are generated for different aspects of the production line. For production line design optimization, suggestions such as equipment layout adjustment and equipment update can be proposed based on the equipment performance analysis results and process data analysis; in terms of production plan adjustment, the arrangement and time planning of production tasks can be adjusted based on logistics data and equipment status data; according to the periodic changes in equipment performance and environmental impacts, the maintenance needs of equipment are predicted and maintenance plans are arranged in advance; based on the analysis of factors affecting product quality, quality control strategies are formulated, such as adjusting process parameters, replacing raw materials, etc., to make decisions more scientific, reasonable and forward-looking, avoiding the blindness and experience of traditional manual decision-making.

[0026] As described in step S8 above, the human-computer interaction interface is used to display the operating status of the virtual production line (the operating position and speed of the equipment, etc.), data analysis results (potential rules mined, performance indicators, etc.) and decision-making recommendations using visualization technology (such as charts, 3D models, and animations). The interface provides real-time update functions to ensure that users can keep up to date with the latest information. At the same time, users can customize the display content and operation interface layout according to their own needs for easy operation and viewing. The decision-making operations fed back by users on the interface will be converted into specific operation instructions and passed to the physical production line and the virtual production line to manage the entire life cycle of the production line, including startup, operation, maintenance, and optimization, and realize intelligent operation and control of the production line. For example, based on the displayed equipment maintenance prediction information, users can issue equipment maintenance instructions, which will be passed to the physical and virtual production lines to update the equipment maintenance plan.

[0027] In one embodiment, an automobile engine production line includes various processing equipment, such as CNC machine tools, casting equipment, assembly equipment, conveying equipment, etc., which are interconnected to form a physical production line. Using 3D modeling software, a corresponding virtual production line is created based on the equipment specifications, layout and process path in the physical production line. Each physical device has its corresponding virtual device, such as a physical CNC machine tool that is accurately modeled as a virtual CNC machine tool with the same processing range and motion trajectory in a virtual environment. A distributed database system is established. In this embodiment, Hadoop Distributed File System (HDFS) is used to store various types of data. This database is the core to store information about physical production lines and virtual production lines. Using structured programming methods, the LAD\SCL programming language is used to create electrical control programs on Siemens S7-1200 PLC. Among them, the communication and data analysis module is responsible for the analysis and communication of data within the digital twin system; the loading and unloading and processing process control module controls the fully automatic loading and unloading process in the physical production line and the virtual production line, and can be switched to a manual process as needed; the RFID read-write control and warehouse management module manages the storage and tracking of materials, such as tracking the position and status of engine parts on the production line through RFID tags, and guiding the production of the virtual production line through user interface operation instructions, thereby completing the management and control of the physical production line. Through the Profinet industrial Ethernet protocol, the devices in the physical production line are connected to the network node to achieve real-time communication between physical devices, virtual devices and databases.

[0028] Sensors are installed on each CNC machine tool in the physical production line to collect equipment parameters, such as spindle speed, cutting force, processing accuracy, etc.; temperature sensors and pressure sensors are installed on casting equipment to collect production process data; temperature and humidity sensors are installed at different locations in the workshop to collect environmental data; position sensors, speed sensors, etc. are installed on logistics transportation equipment to collect logistics data. These sensors collect data according to a predetermined 1-minute time interval. For the collected data, preprocessing is first performed: data cleaning, removing obviously erroneous data, such as spindle speeds that exceed the normal range (such as abnormal values ​​where the spindle speed exceeds the maximum speed of the equipment); data classification, dividing equipment parameters into machine tool parameters, casting equipment parameters, etc.; classifying production process data into casting process data, processing process data, etc.; storing environmental data according to different areas of the workshop; and classifying logistics data according to the type of transportation equipment. The processed data is stored in the twin database.

[0029] For each device in the physical production line, set properties according to its physical characteristics. For CNC machine tools, set the rigid body and collision body properties of virtual CNC machine tools according to their models, processing range, maximum load, etc. For example, set the motion range of the virtual machine tool according to the processing stroke of the machine tool; set the rigid body properties according to its weight. Set kinematic pairs and constraints for physical and virtual devices, such as setting the guide kinematic pairs of the machine tool so that the motion trajectory of the virtual machine tool is the same as that of the physical machine tool; set the constraints of the assembly equipment to ensure that the assembly actions in the assembly process conform to the actual assembly logic. Finally, establish a one-to-one mapping relationship between physical devices and virtual devices, so that the virtual device can accurately simulate the actions and states of the physical device. When a CNC machine tool in the physical production line is updated from an old model to a new model, according to the parameters of the new machine tool (such as a larger processing range, a higher spindle speed), the properties and mapping relationship of the virtual machine tool are automatically updated, and the model of the virtual machine tool is adjusted so that it can accurately simulate the physical characteristics and motion behavior of the new machine tool.

[0030] Parse the key parameters in the physical production line data from the twin database. When the spindle speed, tool position and speed parameters collected from the physical machine tool are updated, use this information as the key parameters. Push these key parameters to the corresponding components on the virtual machine tool to ensure that the spindle speed, tool position and speed of the virtual machine tool are updated in real time to synchronize them with the physical machine tool. Adjust the position and operation of the virtual production line components based on the mapping relationship. Adjust the position of the virtual machine tool according to the position information of the physical machine tool and the relative position information of the virtual machine tool; run the virtual machine tool according to the speed parameters of the physical machine tool. During the operation of the virtual production line, obtain the virtual production line data, such as the processing simulation data of the virtual machine tool and the assembly action data of the virtual assembly equipment, and store them in the twin database. Evaluate the performance of the virtual production line. If it is found that the simulated processing accuracy of the virtual machine tool is lower than the actual processing accuracy of the physical machine tool, adjust the model parameters of the virtual machine tool, such as adjusting the wear model of the virtual tool, so that the performance of the virtual production line is consistent with that of the physical production line. During the data synchronization process, if data delay occurs, such as the delayed arrival of the position data of the physical machine tool, data backup is used for recovery first, and the backed-up position data is sent to the virtual machine tool; if the backup is unavailable, the data is retransmitted; if a small amount of position data is lost, a data interpolation algorithm is used to compensate for the lost data based on the data from the previous and next moments.

[0031] From the integrated AI algorithm set, select the algorithm according to the data type and analysis purpose. For the periodic change analysis of equipment performance, the time series prediction algorithm is used to analyze the historical data of the machine tool spindle speed, and it is found that the performance will decline every Friday afternoon due to the long-term operation of the equipment. For the impact of process parameters on product quality, the association rule mining algorithm is used to analyze the relationship between process parameters such as casting temperature and pressure and the quality of the engine cylinder body, and it is found that excessive casting temperature will cause sand holes on the surface of the cylinder body. For the impact of environmental factors, the cluster analysis algorithm is used to find that the failure rate of equipment increases significantly under high temperature and high humidity environments. For the logistics link, the transportation speed and location data of the transportation equipment are analyzed through the time series prediction algorithm, and it is found that 9 to 10 am every day is the transportation bottleneck period.

[0032] According to the periodic changes in equipment performance, it is recommended to arrange maintenance of some machine tools or reduce the processing tasks on Friday afternoons; according to the impact of process parameters on quality, it is recommended to optimize the casting temperature control strategy and adjust the casting temperature to the optimal range. According to the influence of environmental factors, it is recommended to add cooling and dehumidification devices to the equipment in high temperature and high humidity environments. For the logistics bottleneck period, adjust the production plan and adjust the transportation task of the engine block to other time periods to avoid congestion. According to the equipment performance analysis, it is predicted that some machine tools may experience performance degradation every Friday afternoon, and maintenance personnel are arranged in advance to perform preventive maintenance. According to the results of process parameter analysis, the casting process in the production process is controlled in quality, and the monitoring and adjustment of casting temperature and pressure are strengthened.

[0033] Through the human-computer interaction interface, the virtual engine production line is displayed in a 3D model, and the operator can intuitively see the operating status of the machine tools and assembly equipment. The data analysis results are displayed in charts, such as the periodic change curve of machine tool performance and the correlation chart between process parameters and quality. The transportation process and speed of logistics transportation equipment are displayed in animation. The interface also displays the decision suggestions generated above, such as the prompt "Schedule maintenance of some machine tools on Friday afternoon". According to the interface information, the operator can issue decision operations through the operation interface, such as confirming the machine tool maintenance plan. The operation instruction will be passed to the physical production line and the virtual production line. In the physical production line, the maintenance personnel will receive the maintenance task notification and maintain the corresponding machine tool; in the virtual production line, the maintenance status of the virtual machine tool will be updated, and the simulation data of the maintenance plan will be updated at the same time.

[0034] Reference Figure 2 , is a structural block diagram of a digital twin and AI-assisted decision-making device in production line lifecycle management in one embodiment of the present invention, including: System building module, used to build a digital twin system including physical production line, virtual production line, twin database, service management and connection; The data acquisition module is used to use the data acquisition equipment deployed in various links of the physical production line to pre-process the collected physical production line data according to the equipment parameters, production process data, environmental data and logistics data collected periodically at predetermined time intervals, and classify the data according to the pre-set data classification rules and transmit it to the twin database for storage and management; The mapping establishment module is used to set the rigid body and collision body properties of the physical production line equipment and the virtual production line equipment and the corresponding kinematic pairs and constraints according to the physical characteristics of each device in the physical production line, set the corresponding property settings of each device according to the type, specification and performance of the device, and establish the mapping relationship between the physical device and the virtual device; A data analysis module, used to receive and analyze key parameters in the physical production line data, and push the key parameters to corresponding equipment components on the virtual production line; The virtual operation module is used to adjust the corresponding positions of the equipment components in the virtual production line based on the status information of each equipment component in the virtual production line and the relative position information of each equipment component in the physical production line according to the established mapping relationship, and operate according to the speed parameters of each equipment component. At the same time, the virtual production line data during the operation of the virtual production line is obtained and stored in the twin database. The operation status of the virtual production line is evaluated based on the virtual production line data. If it is found that the performance of the virtual production line deviates from that of the physical production line, the model parameters of the virtual production line are adjusted; The AI ​​analysis module is used to introduce AI algorithms to analyze the data in the twin database, select appropriate algorithms according to the type of data and analysis purpose, and mine potential rules in the data; A decision-making support module, which is used to generate decision-making suggestions for production line design optimization, production plan adjustment, equipment maintenance prediction and quality control based on the analysis results of AI algorithms; The production line management module is used to display the operating status, data analysis results and decision-making suggestions of the virtual production line through a human-computer interaction interface. The human-computer interaction interface provides a real-time update function. According to the feedback decision operations, specific operation instructions are transmitted to the physical production line and the virtual production line to manage the entire life cycle of the production line.

[0035] For the specific implementation of each device in the above device example, please refer to the above method embodiment, which will not be repeated here.

[0036] In summary, a digital twin system including physical production lines, virtual production lines, twin databases, service management and connections is constructed; using data acquisition equipment deployed in various links of the physical production line, the equipment parameters, production process data, environmental data and logistics data collected periodically at predetermined time intervals are preprocessed on the collected physical production line data, and the data is classified according to the pre-set data classification rules and transmitted to the twin database for storage and management; according to the physical characteristics of each device in the physical production line, the rigid body, collision body properties and corresponding kinematic pairs and constraints of the physical production line equipment and virtual production line equipment are set; according to the type, specification and performance of the equipment, the corresponding property settings of each device are set, and a mapping relationship between the physical device and the virtual device is established; receiving and parsing the key parameters in the physical production line data, and pushing the key parameters to the corresponding equipment components on the virtual production line; Based on the status information of each equipment component in the virtual production line and the relative position information of each equipment component in the physical production line, the corresponding position of each equipment component in the virtual production line is adjusted according to the established mapping relationship, and the operation is carried out according to the speed parameters of each equipment component. At the same time, the virtual production line data during the operation of the virtual production line is obtained and stored in the twin database, and the operation status of the virtual production line is evaluated according to the virtual production line data. If it is found that the performance of the virtual production line deviates from that of the physical production line, the model parameters of the virtual production line are adjusted; the AI ​​algorithm is introduced to analyze the data in the twin database, and the appropriate algorithm is selected according to the type of data and the purpose of analysis to mine the potential rules in the data; based on the analysis results of the AI ​​algorithm, decision-making recommendations are generated for production line design optimization, production plan adjustment, equipment maintenance prediction and quality control; the operation status, data analysis results and decision recommendations of the virtual production line are displayed through the human-computer interaction interface to realize the intelligent management of the entire life cycle of the production line, improve production efficiency, reduce costs, enhance the adaptability of production line management and the effectiveness of decision-making, and promote the development of industrial production line management in a more intelligent, efficient and accurate direction.

[0037] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0038] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A digital twin and AI-assisted decision-making method in production line lifecycle management, characterized in that: The following steps are involved: Build a digital twin system including physical production lines, virtual production lines, twin databases, service management and connections. The twin database is the core of the entire digital twin system and is used to manage all data related to the digital twin system. Using data collection equipment deployed at various links of the physical production line, the equipment parameters, production process data, environmental data, and logistics data collected periodically at predetermined time intervals are pre-processed, and the data is classified according to pre-set data classification rules and transmitted to the twin database for storage and management; According to the physical characteristics of each device in the physical production line, set the rigid body, collision body properties and corresponding kinematic pairs and constraints of the physical and virtual production line devices. According to the type, specification and performance of the equipment, set the corresponding property settings of each device, and establish the mapping relationship between the physical and virtual devices. Receive and parse key parameters in the physical production line data, and push the key parameters to corresponding equipment components on the virtual production line, wherein the key parameters include position information and speed parameters of each equipment component of the physical production line; Based on the status information of each equipment component in the virtual production line and the relative position information of each equipment component in the physical production line, according to the established mapping relationship, the corresponding position of each equipment component in the virtual production line is adjusted, and it is operated according to the speed parameters of each equipment component. At the same time, the virtual production line data during the operation of the virtual production line is obtained and stored in the twin database. The operation status of the virtual production line is evaluated based on the virtual production line data. If it is found that the performance of the virtual production line deviates from that of the physical production line, the model parameters of the virtual production line are adjusted; Introduce AI algorithms to analyze the data in the twin database, select appropriate algorithms according to the type of data and the purpose of analysis, and mine potential rules in the data. The potential rules include but are not limited to periodic changes in equipment performance, the impact of process parameters on product quality, the impact of environmental factors on equipment and production processes, and potential bottlenecks in logistics. Generate decision-making suggestions for production line design optimization, production plan adjustment, equipment maintenance prediction and quality control based on the analysis results of AI algorithms; The operating status, data analysis results and decision-making suggestions of the virtual production line are displayed through a human-computer interaction interface. The human-computer interaction interface provides a real-time update function, and according to the feedback decision operations, the specific operation instructions are transmitted to the physical production line and the virtual production line to manage the entire life cycle of the production line.

2. The digital twin and AI-assisted decision-making method in production line lifecycle management according to claim 1, characterized in that: In the digital twin system, the physical production line includes multiple interconnected production equipment, and the virtual production line corresponds one-to-one to the production equipment in the physical production line through a predefined mapping relationship. The twin database adopts a distributed storage architecture to store various types of data collected from the physical production line, data generated during the operation of the virtual production line, and historical operation data of the physical production line and the virtual production line. The services include data processing services, decision services and communication services, which are used to realize data analysis and processing, decision generation and information interaction between system components. The service management adopts a structured programming method, based on S7-1200 using the LAD\SCL programming language, to create an electrical control program, debug and run it in the physical entity and the digital twin system to achieve virtual-real synchronization. The electrical control program includes communication and data analysis modules, loading and unloading and processing flow control modules, and RFID read-write control and warehouse management modules. The connection accesses the network nodes of each device in the production line, and uses their own private protocols to realize real-time communication within the entire digital twin system through industrial Ethernet.

3. The digital twin and AI-assisted decision-making method in production line lifecycle management according to claim 1, characterized in that: The digital twin system also includes using machine learning algorithms to train and optimize the neural network model using historical data and real-time data, adjusting the parameters and structure of the model according to actual production conditions, and obtaining a virtual production line model for simulating the operating status and performance of the physical production line, and realizing real-time data interaction and status synchronization between the physical production line and the virtual production line through connection.

4. The digital twin and AI-assisted decision-making method in production line lifecycle management according to claim 1, characterized in that: A dynamic adjustment mechanism for the mapping relationship between physical devices and virtual devices is established. When a physical device is updated or replaced, the mapping relationship is automatically updated according to the information of the new device.

5. The digital twin and AI-assisted decision-making method in production line lifecycle management according to claim 1, characterized in that: During the data synchronization process between the physical production line and the virtual production line, when data delay or data loss occurs, the automatic error correction mechanism is triggered. According to the degree of data delay and the scale of data loss, data backup is used for recovery first. When the backup data is unavailable, data retransmission is used. For a small amount of data loss, a data interpolation algorithm is used to compensate. The performance of the automatic error correction mechanism is monitored and evaluated, and optimized through the set timeliness and accuracy indicators.

6. The digital twin and AI-assisted decision-making method in production line lifecycle management according to claim 1, characterized in that: The AI ​​algorithm is an integrated algorithm set, including but not limited to cluster analysis algorithm, association rule mining algorithm, and time series prediction algorithm. The corresponding algorithm is selected to analyze the data according to different analysis tasks.

7. The digital twin and AI-assisted decision-making method in production line lifecycle management according to claim 1, characterized in that: After generating the decision proposal, based on the various factors affecting the production line, the impact of different decision plans on the short-term and long-term benefits of the production line is comprehensively considered, and each decision plan is comprehensively evaluated and prioritized. The various factors include but are not limited to production line production goals, cost-effectiveness, resource conditions, and product quality.

8. The digital twin and AI-assisted decision-making method in production line lifecycle management according to claim 1, characterized in that: The human-computer interaction interface uses visualization technology to display the operating status of the production line and data analysis results in the form of charts, 3D models, and animations. It also provides interactive operation elements to customize the display content and operation interface layout according to the user's own needs.

9. The digital twin and AI-assisted decision-making method in production line lifecycle management according to claim 1, characterized in that: The data twin system has automatic learning and self-adaptation functions. It automatically updates and optimizes the parameters and decision-making rules of the AI ​​algorithm based on the long-term operation data of the production line, and automatically switches the analysis and decision-making mode according to the operation stage of the production line. The production line operation stage includes the startup stage, stable operation stage and fault stage.

10. A digital twin and AI-assisted decision-making device in production line lifecycle management, characterized in that: include: System building module, used to build a digital twin system including physical production line, virtual production line, twin database, service management and connection; The data acquisition module is used to use the data acquisition equipment deployed in various links of the physical production line to pre-process the collected physical production line data according to the equipment parameters, production process data, environmental data and logistics data collected periodically at predetermined time intervals, and classify the data according to the pre-set data classification rules and transmit it to the twin database for storage and management; The mapping establishment module is used to set the rigid body and collision body properties of the physical production line equipment and the virtual production line equipment and the corresponding kinematic pairs and constraints according to the physical characteristics of each device in the physical production line, set the corresponding property settings of each device according to the type, specification and performance of the device, and establish the mapping relationship between the physical device and the virtual device; A data analysis module, used to receive and analyze key parameters in the physical production line data, and push the key parameters to corresponding equipment components on the virtual production line; The virtual operation module is used to adjust the corresponding positions of the equipment components in the virtual production line based on the status information of each equipment component in the virtual production line and the relative position information of each equipment component in the physical production line according to the established mapping relationship, and operate according to the speed parameters of each equipment component. At the same time, the virtual production line data during the operation of the virtual production line is obtained and stored in the twin database. The operation status of the virtual production line is evaluated based on the virtual production line data. If it is found that the performance of the virtual production line deviates from that of the physical production line, the model parameters of the virtual production line are adjusted; The AI ​​analysis module is used to introduce AI algorithms to analyze the data in the twin database, select appropriate algorithms according to the type of data and analysis purpose, and mine potential rules in the data; A decision-making support module, which is used to generate decision-making suggestions for production line design optimization, production plan adjustment, equipment maintenance prediction and quality control based on the analysis results of AI algorithms; The production line management module is used to display the operating status, data analysis results and decision-making suggestions of the virtual production line through a human-computer interaction interface. The human-computer interaction interface provides a real-time update function. According to the feedback decision operations, specific operation instructions are transmitted to the physical production line and the virtual production line to manage the entire life cycle of the production line.

Images