Intelligent manufacturing production line life cycle management and control method based on digital twinning and AI cooperation

Abstract

The invention relates to an intelligent manufacturing production line life cycle management and control method based on digital twinning and AI cooperation. The method comprises the following steps: transmitting acquired operation data of a physical production line to a preset digital twinning model; driving the digital twin model to perform simulation operation by using the operation data, and analyzing the operation state and potential problems of the physical production line to obtain a simulation analysis result; the simulation analysis result is input into an AI system, and intelligent decision making and optimization are carried out through the AI system; the optimization decision of the AI system is fed back to the physical production line, the equipment and the production process are controlled and adjusted, and the actual operation condition of the adjusted physical production line is fed back to the digital twin model for data verification and model updating, so that effective management and control of the life cycle of the intelligent manufacturing production line are realized; the problems of data islands, equipment fault prediction and maintenance limitation and the like in traditional production management are solved, the production efficiency and the product quality are improved, and the intelligent manufacturing level of enterprises is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of intelligent manufacturing technology, and in particular to an intelligent manufacturing production line lifecycle management and control method based on digital twin and AI collaboration. Background Art

[0002] In the process of rapid development of today's manufacturing industry, intelligent manufacturing has become the core direction of industry transformation and upgrading. With the continuous advancement of the concept of industrial intelligent manufacturing, enterprises have put forward higher requirements for refined management of production processes, efficient decision-making and stable operation of equipment. Traditional production management methods mainly rely on manual experience and simple automated control, which is difficult to cope with complex and changing production environments and growing market competition pressure. In terms of production line control, although some companies have introduced automated equipment and information systems, there are still many problems in data utilization and collaborative management. For example, there is often a data island phenomenon between the physical production line and the information system, which makes it impossible for production data to be fed back to the management decision-making level in real time and accurately, thereby affecting production efficiency and product quality. At the same time, equipment failure prediction and maintenance mainly rely on post-maintenance or regular maintenance, lacking real-time monitoring and accurate prediction of equipment operation status, which not only increases equipment downtime, but also increases maintenance costs. The emergence of digital twin technology provides a new way to solve these problems. However, pure digital twin technology still has certain limitations in intelligent decision-making and optimization. Summary of the invention

[0003] The main purpose of this invention is to provide a method for lifecycle management of intelligent manufacturing production lines based on the collaboration of digital twins and AI, so as to achieve effective management and control of the lifecycle of intelligent manufacturing production lines, solve problems such as data silos, equipment fault prediction and maintenance limitations existing in traditional production management, improve production efficiency and product quality, and enhance the company's intelligent manufacturing level.

[0004] To achieve the above objectives, the present invention provides a method for lifecycle management of intelligent manufacturing production lines based on collaboration between digital twins and AI, comprising the following steps: The acquired operation data of the physical production line is transmitted to the preset digital twin model, wherein the operation data includes equipment status data, production progress data and product quality data; Using the operation data to drive the digital twin model to perform simulation operation, analyze the operation status and potential problems of the physical production line, and obtain simulation analysis results; Input the simulation analysis results into the AI ​​system, and use the AI ​​system to make intelligent decisions and optimizations, including automatically adjusting production parameters, optimizing production scheduling, and predicting equipment maintenance; The optimization decisions of the AI ​​system are fed back to the physical production line to control and adjust the equipment and production process, and the actual operation of the adjusted physical production line is fed back to the digital twin model for data verification and model updating.

[0005] Furthermore, the step of transmitting the acquired operation data of the physical production line to the preset digital twin model includes: Use temperature sensors, vibration sensors and pressure sensors to collect operating parameters of key parts of equipment in the physical production line, combine counters, position sensors and the data interface of the production management system to obtain the processing progress of the product in each process, and use high-precision measuring instruments and automated testing equipment to obtain the product's dimensional accuracy, surface roughness and physical performance parameters; The collected operating data is preprocessed by data cleaning, data verification and format conversion, and the processed data is transmitted to the preset digital twin model.

[0006] Furthermore, the digital twin model is based on the mechanical properties, operating logic and process flow of the equipment in the actual physical production line, and adds two dimensions of twin data and service to the existing five-dimensional model framework, and uses modeling software to construct a three-dimensional model of the physical entity and the virtual entity. The twin data includes historical data of the entire life cycle of the equipment, data collected in real-time operation, and data generated during the virtual simulation process.

[0007] Furthermore, the digital twin model also includes service management, which adopts a structured programming method and creates an electrical control program based on S7-1200 using the LAD\SCL programming language. The communication and analysis module is responsible for processing the data exchange analysis between the PLC and the CNC lathe, and the communication programming with the AGV and industrial robot; the loading and unloading and processing control module realizes the fully automatic process control of loading and unloading and processing through sequential programming, and can switch the manual process as needed; the RFID control and warehouse management module is used for the effective management of materials, and through the human-machine interface operation instructions, it realizes the coordinated control of the physical production line and the virtual production line, as well as the intelligent level of warehouse management and the coordinated work of the overall production line.

[0008] Furthermore, the step of using the operation data to drive the digital twin model to perform simulation operation, analyzing the operation status and potential problems of the physical production line, and obtaining simulation analysis results includes: The collected equipment status data is transmitted to the virtual equipment components corresponding to the digital twin model according to the predetermined data mapping rules, and the actual operation status of the actual physical production line is reflected through the virtual equipment components; Based on the flow time and completion status of each process of the physical production line in the production progress data, the digital twin model is used to simulate the product processing path and the distribution status of the work-in-progress, and display the real-time progress of the production process; Compare and analyze the quality parameters in the product quality data with the preset quality standards, and use the digital twin model to visually mark the equipment links and / or process steps that may actually affect the product quality; Through the built-in physical engine and process logic algorithm in the digital twin model, the virtual production line is dynamically simulated. Based on the kinematic and dynamic models of the equipment, the operating behavior of each device under different working conditions is simulated in the digital twin model, the possible failure modes and failure probability of the equipment are predicted, and the impact of the fluctuation of the process parameters of each device in the production process on product quality and production efficiency is analyzed; Based on the simulation results, data analysis and pattern recognition technology are used to comprehensively evaluate the operating status of the production line, identify potential problems, and generate a detailed simulation analysis report. The report includes a description of the problem, possible cause analysis, and corresponding improvement suggestions.

[0009] Furthermore, when the AI ​​system makes intelligent decisions and optimizes, in the process of automatically adjusting production parameters, the neural network algorithm based on deep learning trains the model through a large amount of historical production data and real-time operation data, predicts the optimal production parameter combination under different product types, equipment conditions and production environments, and realizes automatic adjustment of production parameters through the integrated equipment control system; In production scheduling optimization, we use intelligent optimization algorithms such as mixed integer programming and genetic algorithms to comprehensively consider factors such as order delivery time, equipment capacity, material supply, and human resources to build a production scheduling model to achieve the allocation of production tasks and optimization of production processes. For equipment maintenance prediction, we use time series analysis and machine learning algorithms to deeply mine and analyze equipment operation data, establish an equipment failure prediction model, and predict the time, type, and probability of equipment failure in advance.

[0010] Furthermore, the actual operation status of the adjusted physical production line is fed back to the digital twin model for data verification and model update, including: Real-time monitoring and analysis of newly generated operating data of the adjusted physical production line; Evaluate the effectiveness of the AI ​​system’s optimization decisions on the performance of the physical production line by comparing the key performance indicators before and after the physical production line adjustment, including equipment stability, production efficiency improvement, and product quality fluctuations; If the adjusted physical production line finds data anomalies or the optimization effect does not meet expectations, the digital twin model update program will be started; If new problems or data anomalies still exist after the digital twin model is updated, the new data will be integrated with the historical data, and the digital twin model and AI system will be retrained and optimized using incremental learning and model adaptive adjustment.

[0011] Furthermore, an adaptive adjustment mechanism for models and algorithms is established. For digital twin models, a dynamic model update mechanism is established to automatically identify the parts of the model that need to be adjusted based on equipment updates, process improvements or changes in the production environment of the physical production line. The structure and parameters of the digital twin model are updated in real time through parameter learning of new equipment, logical embedding of new processes, and quantitative analysis of environmental factors to ensure that the actual status of the physical production line can be continuously and accurately reflected. For AI systems, an adaptive learning strategy is adopted to automatically adjust the model structure and hyperparameters of the algorithm based on the characteristics of production data and changes in management and control objectives at different stages. Online learning and transfer learning techniques are used to enable AI algorithms to quickly adapt to new data distributions and production task requirements.

[0012] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the above-mentioned intelligent manufacturing production line lifecycle management method based on the collaboration of digital twins and AI are implemented.

[0013] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the above-mentioned intelligent manufacturing production line lifecycle management method based on the collaboration of digital twins and AI are implemented.

[0014] The intelligent manufacturing production line lifecycle management method based on digital twin and AI collaboration provided by the present invention has the following beneficial effects: by real-time monitoring of physical production line operation data to drive digital twin model simulation, production process bottlenecks and potential problems can be found in time; the digital twin model analyzes product quality data and combines AI algorithm to accurately control process parameters, which can greatly reduce product defect rates; according to real-time monitoring of equipment operation data and maintenance prediction function of AI system, potential equipment failures can be detected in advance and maintenance can be arranged, reducing high maintenance costs and production losses caused by sudden equipment failures; at the same time, the digital twin model provides managers with an intuitive virtual production line display, which helps to grasp production progress, equipment status, product quality and other information in real time. With the help of human-machine interface operation instructions, physical and virtual production lines can be conveniently and collaboratively managed to achieve transparent and precise management of the production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a flow chart of a method for controlling the life cycle of an intelligent manufacturing production line based on collaboration between digital twins and AI in one embodiment of the present invention; Figure 2It is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0016] 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 method for intelligent production line lifecycle management and control based on digital twin and AI collaboration proposed by the present invention, comprising the following steps: S1, transmitting the acquired operation data of the physical production line to the preset digital twin model, wherein the operation data includes equipment status data, production progress data and product quality data; S2, using the operation data to drive the digital twin model to perform simulation operation, analyze the operation status and potential problems of the physical production line, and obtain simulation analysis results; S3, inputting the simulation analysis results into the AI ​​system, and making intelligent decisions and optimizations through the AI ​​system, including automatically adjusting production parameters, optimizing production scheduling, and predicting equipment maintenance; S4 feeds back the optimization decisions of the AI ​​system to the physical production line, controls and adjusts the equipment and production process, and feeds back the actual operating conditions of the adjusted physical production line to the digital twin model for data verification and model updating.

[0019] As described in step S1 above, in the production line environment of intelligent manufacturing, it is necessary to collect various types of data from the physical production line. Equipment status data may include physical parameters such as temperature, pressure, vibration, etc. of the equipment. These data can be collected through devices such as temperature sensors, vibration sensors, and pressure sensors. They reflect the current operation of the equipment, such as whether the equipment is in normal operation, whether there are problems such as overheating and overload. Production progress data covers the progress of each process of the product on the production line. For example, the processing progress of the product in each process is obtained by using counters, position sensors, and data interfaces of the production management system. It can help us grasp the overall progress of the production process and determine whether there are problems such as process delays or backlogs. Product quality data is obtained through high-precision measuring instruments and automated testing equipment, including product dimensional accuracy, surface roughness, and physical performance parameters. These data are crucial for evaluating whether the product meets quality standards. After collecting these data, they need to be transmitted to a pre-set digital twin model to provide a basis for subsequent analysis and decision-making. The digital twin model is a virtual mapping of the physical production line. By receiving these data, the actual situation of the physical production line can be reflected in the virtual environment.

[0020] As described in step S2 above, in the digital twin model, the operating data obtained and transmitted from the physical production line is used as input to drive it to perform simulation operation. First, the collected equipment status data is transmitted to the virtual equipment components corresponding to the digital twin model according to the predetermined data mapping rules, so that the status of the virtual equipment components can accurately reflect the status of the actual physical equipment. For example, the temperature, pressure and other parameters of the virtual equipment will be updated according to the sensor data of the physical equipment. According to the flow time and completion status of the product in each process of the physical production line in the production progress data, the digital twin model is used to simulate the processing path of the product and the distribution status of the work-in-progress, and the real-time progress of the production process is displayed in a visual way, so that the management personnel can intuitively understand whether the production progress meets expectations. At the same time, the quality parameters in the product quality data are compared and analyzed with the preset quality standards, and the equipment links and / or process steps that may affect the product quality are marked in a visual way, so as to quickly locate the quality problems. Through the built-in physical engine and process logic algorithm, the virtual production line is dynamically simulated and run. Based on the kinematic and dynamic models of the equipment, the operating behavior of each device under different working conditions is simulated to predict the possible failure mode and failure probability of the equipment. It can also analyze the impact of the fluctuation of process parameters of each device during the production process on product quality and production efficiency. Finally, using data analysis and pattern recognition technology, a comprehensive assessment of the operating status of the production line is conducted to identify potential problems such as equipment performance degradation and process bottlenecks, and a detailed simulation analysis report is generated. The report contains a description of the problem, possible cause analysis, and corresponding improvement suggestions, providing a basis for subsequent intelligent decision-making.

[0021] As described in step S3 above, the simulation analysis results of the digital twin model are input into the system, and the system's powerful computing and learning capabilities are used to make intelligent decisions and optimizations. For automatic adjustment of production parameters, a neural network algorithm based on deep learning is used to train the model using a large amount of historical production data and real-time operation data to predict the optimal production parameter combination under different product types, equipment conditions and production environments. For example, the optimal cutting speed, feed rate and other parameters of engine processing under the current equipment status and production tasks can be predicted, and these parameters can be automatically adjusted through the integrated equipment control system to achieve the purpose of improving production efficiency and product quality. In terms of production scheduling optimization, intelligent optimization algorithms such as mixed integer programming and genetic algorithms are used to comprehensively consider factors such as order delivery time, equipment capacity, material supply and human resources, build a production scheduling model, and reasonably allocate production tasks. For example, for different types of engine orders, the processing sequence of the engine cylinder and the allocation of processing equipment are optimized according to factors such as the urgency of the order, the idle time of the equipment, and the inventory of the required materials, so as to optimize the production process. For equipment maintenance prediction, time series analysis and machine learning algorithms are used to conduct in-depth mining and analysis of equipment operation data, establish an equipment failure prediction model, and predict the time, type and probability of equipment failure in advance, so that enterprises can arrange equipment maintenance in advance and reduce equipment downtime.

[0022] As described in step S4 above, the optimization decision obtained by the system is fed back to the physical production line, so that the equipment and production process of the physical production line are controlled and adjusted accordingly according to these decisions, such as applying the optimized production parameters to the actual equipment controller, adjusting the operating parameters of the equipment, or adjusting the production sequence of the order and the task allocation of the equipment according to the production schedule, and performing maintenance operations on the equipment. After the physical production line is adjusted, its actual operation needs to be monitored and analyzed in real time, and the newly generated operation data is fed back to the digital twin model. By comparing the key performance indicators before and after the physical production line is adjusted, the effectiveness of the system's optimization decision on the performance of the physical production line is evaluated. These key performance indicators include equipment stability, production efficiency improvement, and product quality fluctuations. If data anomalies are found or the optimization effect does not meet expectations, the digital twin model update program is started to update the digital twin model to make it more accurately reflect the actual situation of the physical production line. If there are still new problems or data anomalies after the update, the new data is merged with the historical data, and the digital twin model and system are retrained and optimized using incremental learning and model adaptive adjustment to improve the performance and adaptability of the entire management and control system.

[0023] In one embodiment, a car engine production line is constructed, including various equipment, such as CNC machine tools, industrial robots, conveyor belts, testing equipment, storage equipment, and AGV carts. Sensors are installed on key equipment, such as temperature sensors, vibration sensors, and pressure sensors on key parts such as the spindle and motor of the CNC machine tool to monitor the operating status of the equipment. Counters and position sensors are installed at each process point on the production line, and RFID tag readers are equipped in the storage area. At the same time, information in the production process is obtained through the data interface of the production management system.

[0024] In this embodiment, NX MCD is used to achieve modeling, reduce the complexity of model making, and improve the simulation effect and fluency. According to the mechanical characteristics, operation logic and process flow of the equipment in the actual physical production line, the framework is designed based on the existing five-dimensional model to build a data-driven intelligent production line digital twin system. Among them, the five-dimensional model framework includes: MDT=(PE,VE,Ss,DD,CN) In the formula: VE and PE correspond to virtual and physical entities respectively, Ss represents service, DD represents twin data, and CN represents the connection between the components. The twin data in the framework of the digital twin system of the intelligent production line is the core of the entire digital twin intelligent control system. Any manufacturing data and virtual simulation data of the system can be collected into the twin data module to form big data, and the control, diagnosis and prediction of the production line and equipment can be achieved through data model driving; the service system is based on the virtual workshop, and the current operating status of the production line and each equipment is monitored and predicted in real time. The staff can view the system visualization interface at any time. The digital twin model creates corresponding virtual equipment for each device, including engine block processing machine tools, crankshaft processing machine tools, assembly robots, material transfer devices, etc., to ensure that the appearance, size and action logic of the virtual equipment match the physical equipment. For example, the spindle, tool holder, tool, and workbench inside the machine tool need to set the corresponding rigid body and collision body properties, and add the corresponding kinematic pairs and constraints; finally, define the position and speed of the kinematic pairs (such as the X-axis, Y-axis, Z-axis, etc. in the machine tool) to make them actuators, which can reach the specified position according to the target position and specified speed and then feedback information to the PLC control unit.

[0025] Collect historical data of the entire life cycle of equipment on the production line, including equipment maintenance records and operating parameter records (such as temperature, pressure, and vibration data at different times) of previous engine production lines, and store them in the database. These data will serve as the initial twin data to provide a basis for subsequent analysis and model training. In the digital twin model, set up a data acquisition and storage mechanism to receive real-time operating data from the physical production line, such as transmitting data collected by sensors (temperature sensors, vibration sensors, etc.) on physical equipment to the digital twin model through a data transmission link (CN) to form real-time operating collected data. Create an electrical control program based on S7-1200 using the LAD\SCL programming language. The program is responsible for processing data exchange analysis between PLC and CNC lathes, and communication programming with AGV and industrial robots. For example, PLC will continuously receive data from sensors of various devices, such as temperature, vibration, and other information. The module will parse and classify these data, store them in the corresponding data storage area, and perform preliminary data analysis to determine whether the data is within the normal range. The fully automatic process control of loading and unloading processing is achieved through sequential programming. In the virtual environment, the whole process of the engine cylinder body being grabbed by the robot arm from the blank warehouse, placed on the processing equipment for processing, and then the processed cylinder body being removed can be simulated. At the same time, the module supports the switching of manual processes so that in special cases, the operator can manually control the loading and unloading process. In terms of warehouse management, RFID tags are attached to each engine component. Through the RFID control and warehouse management module, the location and status of the components can be tracked synchronously in the virtual environment and the physical environment. When the parts in the physical warehouse are taken or replenished, the corresponding parts in the virtual environment will also be updated. Through the human-machine interface operation instructions, the operator can easily view and manage the warehouse information, realize the coordinated control of the physical production line and the virtual production line, and improve the intelligent level of warehouse management.

[0026] Using temperature sensors, vibration sensors and pressure sensors, the operating parameters of key parts of the engine processing equipment in the physical production line are collected at certain time intervals (such as once per second). For example, on the engine cylinder processing machine, the temperature of the spindle, the vibration frequency and pressure of the motor are collected. Combined with the counter, position sensor and the data interface of the production management system, the processing progress of the product in each process is obtained. For example, the number of engine cylinders that have been processed is recorded by the counter, and the position of the cylinder on the production line is tracked by the position sensor to determine which processing process it is in. At the same time, the estimated completion time and actual completion time of each process are obtained from the production management system. Use high-precision measuring instruments and automated testing equipment to obtain the dimensional accuracy, surface roughness and physical performance parameters of the product. After the engine cylinder is processed, the dimensional accuracy of the cylinder is measured by a three-coordinate measuring machine, the surface roughness is measured by a roughness meter, and the physical performance parameters such as the hardness of the cylinder are tested by special equipment. The collected operating data is cleaned to remove abnormal values ​​(such as erroneous data caused by instantaneous failure of the sensor), data verification (checking the integrity and accuracy of the data) and format conversion (unifying the data formats of different sensors into a format acceptable to the digital twin model) and other pre-processing operations. The processed data is transmitted to the preset digital twin model through the data transmission link (CN). For example, the collected temperature data is converted into the temperature attribute value corresponding to the virtual device in the digital twin model, and the processing progress data is mapped to the process progress bar of the virtual production line.

[0027] The collected equipment status data is transmitted to the virtual equipment components corresponding to the digital twin model according to the predetermined data mapping rules. For example, the spindle temperature, motor vibration frequency and other data of the actual engine processing equipment are transmitted to the corresponding components of the virtual equipment, so that the status of the virtual equipment is synchronized with the physical equipment. Among them, the data mapping rules are formulated according to the corresponding relationship between the physical equipment and the virtual components, data types, measurement ranges, process logic and other factors. According to the flow time and completion status of the engine cylinder in each process of the physical production line in the production progress data, the digital twin model is used to simulate the processing path of the product and the distribution status of the work-in-progress. In the virtual environment, the movement and processing process of the cylinder between different processing equipment and the number of cylinders currently in each process are displayed in the form of animation. The quality parameters in the product quality data are compared and analyzed with the preset quality standards. For the measured cylinder dimensional accuracy, surface roughness and other data, they are compared with the preset engine cylinder quality standards, and the equipment links and / or process steps that may actually affect the product quality are marked in a visual way through the digital twin model. If the dimensional accuracy of the cylinder exceeds the tolerance range, the corresponding processing equipment and processing procedures are marked in the virtual model to quickly locate the problem.

[0028] Through the built-in physical engine and process logic algorithm in the digital twin model, the virtual production line is dynamically simulated. Simulate the operation of the engine production line under different working conditions, taking into account different processing speeds, loads and other factors. Based on the kinematic and dynamic models of the equipment, simulate the operating behavior of each device under different working conditions in the digital twin model to predict the possible failure modes and failure probabilities of the equipment. For example, simulate the stress and wear of the spindle when the engine processing equipment runs at high speed for a long time, and predict possible fatigue wear failures. Analyze the impact of process parameter fluctuations of each device on product quality and production efficiency during the production process. When the cutting speed of the processing equipment fluctuates, observe the impact on the processing accuracy and processing time of the engine cylinder block, and present these impacts in a visual way, such as generating a trend chart. Based on the simulation results, use data analysis and pattern recognition technology to comprehensively evaluate the operating status of the production line, identify potential problems, and generate a detailed simulation analysis report. The report includes an overall evaluation of the engine production line, pointing out problems such as "the processing accuracy of the engine cylinder block processing process is unstable, which may be caused by tool wear", as well as possible cause analysis and corresponding improvement suggestions, such as "it is recommended to regularly check and replace the tool."

[0029] In the process of automatically adjusting production parameters, the neural network algorithm based on deep learning uses a large amount of historical production data (data stored in DD) and real-time operation data to train the model. For the engine production line, the product quality and production efficiency data under different processing parameters (such as cutting speed, feed rate, and speed) in the past are used as training sets, and the optimal production parameter combination under different product types (different models of engines), equipment conditions (current temperature, vibration, etc. of the equipment) and production environment is predicted through the neural network. Through the integrated equipment control system, the optimal production parameters predicted by the system (such as adjusting the cutting speed of a certain model of engine cylinder processing to 1200 revolutions per minute and the feed rate to 0.4 mm / rev) are automatically adjusted to the equipment controller of the physical production line to achieve automatic adjustment of production parameters. Using the intelligent optimization algorithm of mixed integer programming and genetic algorithm, the production scheduling model is constructed by comprehensively considering multiple factors such as order delivery time, equipment capacity, material supply and human resources. For example, when receiving multiple engine orders, optimize the allocation of production tasks and the optimization of production processes based on the engine models, delivery time requirements, current equipment capacity (which equipment is idle, equipment processing efficiency), material supply (inventory of engine parts) and human resources (worker scheduling and skill level). Allocate orders to the most suitable equipment for processing to avoid idle equipment or order delays. Use time series analysis and machine learning algorithms to deeply mine and analyze equipment operation data (such as the time series of temperature and vibration data of engine processing equipment) and establish an equipment failure prediction model. Based on the changing trend of historical data and real-time data, predict the time, type and probability of equipment failure in advance. For example, it is predicted that the spindle of an engine processing equipment may have a wear failure in the next 3 days, the type is fatigue wear, and the probability is 70%, so that maintenance work can be arranged in advance to avoid production delays caused by sudden equipment shutdown.

[0030] Feedback the system's optimization decisions to the physical production line to control and adjust the equipment and production process. Modify the parameters of the equipment controller according to the production parameters adjusted by the system; reallocate production tasks according to the production scheduling optimization results; and arrange corresponding maintenance plans for the equipment maintenance prediction results. For example, according to the system's decision, adjust the cutting speed of the engine block processing equipment to the optimized 1200 revolutions per minute, assign an urgent high-performance engine order to an idle and suitable device for production, and arrange maintenance personnel to perform preventive maintenance on the equipment predicted to be faulty. Monitor and analyze the newly generated operating data of the adjusted physical production line in real time. Continue to collect new data from the engine production line through sensors and data acquisition systems, including equipment status data, production progress data, and product quality data. Evaluate the effectiveness of the system's optimization decisions on the performance of the physical production line by comparing the key performance indicators before and after the physical production line adjustment. Compare the equipment stability (such as whether the equipment downtime is reduced), the improvement in production efficiency (whether the number of engines produced per hour is increased), and the product quality fluctuation (whether the standard deviation of the dimensional accuracy and performance indicators of the engine block is reduced) of the engine production line before and after the adjustment. If the adjusted physical production line finds data anomalies or the optimization effect does not meet expectations, the digital twin model update program is started. If it is found that the scrap rate of the engine is still high, the fault prediction model or production parameter optimization model in the digital twin model is updated. If there are still new problems or data anomalies after the digital twin model is updated, the new data is merged with the historical data, and the digital twin model and system are retrained and optimized using incremental learning and model adaptive adjustment. For example, the newly collected engine production line data is merged with the previously stored historical data, and the deep learning neural network, time series analysis model, etc. are retrained so that the digital twin model and system can better adapt to new production situations and problems.

[0031] Establish a dynamic model update mechanism to automatically identify the parts of the model that need to be adjusted based on the equipment updates, process improvements, or changes in the production environment of the physical production line. For example, when new engine processing equipment is introduced, the digital twin model automatically updates the corresponding parts of the virtual entity, learns the parameters and characteristics of the new equipment, and updates the structure and parameters of the digital twin model in real time through parameter learning of the new equipment, logical embedding of the new process, and quantitative analysis of environmental factors. If the engine production line adopts a new processing technology, update the corresponding process logic in the digital twin model to ensure that it can continuously and accurately reflect the actual status of the physical production line. For the system, an adaptive learning strategy is adopted to automatically adjust the model structure and hyperparameters of the algorithm according to the characteristics of production data and changes in control objectives at different stages. For example, as the product type and quantity of engine orders change, the AI ​​system automatically adjusts the optimization objectives and constraints of the production scheduling model. Use online learning and transfer learning techniques to enable the algorithm to quickly adapt to new data distribution and production task requirements. When the production task changes from producing standard engines to producing high-performance engines, the AI ​​system uses transfer learning to migrate the original model knowledge to the new task, and updates the model parameters based on new data through online learning, ensuring that efficient intelligent decision-making and optimization can still be achieved under the new production task.

[0032] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0033] In summary, the acquired operation data of the physical production line is transmitted to the preset digital twin model; The operation data is used to drive the digital twin model to perform simulation operation, analyze the operation status and potential problems of the physical production line, and obtain simulation analysis results; the simulation analysis results are input into the AI ​​system, and intelligent decision-making and optimization are performed through the AI ​​system; the optimization decision of the AI ​​system is fed back to the physical production line to control and adjust the equipment and production process, and the actual operation status of the adjusted physical production line is fed back to the digital twin model for data verification and model updating, so as to achieve effective management and control of the life cycle of the intelligent production line, solve the problems of data islands, equipment fault prediction and maintenance limitations in traditional production management, improve production efficiency and product quality, and enhance the company's intelligent manufacturing level.

[0034] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0035] 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.

[0036] 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 method for lifecycle management of intelligent production lines based on digital twins and AI collaboration, characterized in that: The following steps are involved: The acquired operation data of the physical production line is transmitted to the preset digital twin model, wherein the operation data includes equipment status data, production progress data and product quality data; Using the operation data to drive the digital twin model to perform simulation operation, analyze the operation status and potential problems of the physical production line, and obtain simulation analysis results; Input the simulation analysis results into the AI ​​system, and use the AI ​​system to make intelligent decisions and optimizations, including automatically adjusting production parameters, optimizing production scheduling, and predicting equipment maintenance; The optimization decisions of the AI ​​system are fed back to the physical production line to control and adjust the equipment and production process, and the actual operation of the adjusted physical production line is fed back to the digital twin model for data verification and model updating.

2. According to claim 1, the intelligent production line life cycle management method based on digital twin and AI collaboration is characterized in that: The step of transmitting the acquired operation data of the physical production line to the preset digital twin model includes: Use temperature sensors, vibration sensors and pressure sensors to collect operating parameters of key parts of equipment in the physical production line, combine counters, position sensors and the data interface of the production management system to obtain the processing progress of the product in each process, and use high-precision measuring instruments and automated testing equipment to obtain the product's dimensional accuracy, surface roughness and physical performance parameters; The collected operating data is preprocessed by data cleaning, data verification and format conversion, and the processed data is transmitted to the preset digital twin model.

3. The intelligent manufacturing production line lifecycle management and control method based on digital twin and AI collaboration according to claim 1 is characterized in that: The digital twin model is based on the mechanical characteristics, operating logic and process flow of the equipment in the actual physical production line, and adds two dimensions of twin data and service to the existing five-dimensional model framework. The modeling software is used to construct a three-dimensional model of the physical entity and the virtual entity. The twin data includes historical data of the entire life cycle of the equipment, data collected in real-time operation, and data generated during the virtual simulation process.

4. The intelligent manufacturing production line lifecycle management and control method based on digital twin and AI collaboration according to claim 3 is characterized in that: The digital twin model also includes service management, which adopts a structured programming method and creates an electrical control program based on S7-1200 using the LAD\SCL programming language. The communication and analysis module is responsible for processing the data exchange analysis between the PLC and the CNC lathe, and the communication programming with the AGV and industrial robot; the loading and unloading and processing control module realizes the fully automatic process control of loading and unloading and processing through sequential programming, and can switch the manual process as needed; the RFID control and warehouse management module is used for the effective management of materials, and through the human-machine interface operation instructions, it realizes the coordinated control of the physical production line and the virtual production line, as well as the intelligent level of warehouse management and the coordinated work of the overall production line.

5. The intelligent manufacturing production line lifecycle management and control method based on digital twin and AI collaboration according to claim 1 is characterized in that: The step of using the operation data to drive the digital twin model to perform simulation operation, analyzing the operation status and potential problems of the physical production line, and obtaining simulation analysis results includes: The collected equipment status data is transmitted to the virtual equipment components corresponding to the digital twin model according to the predetermined data mapping rules, and the actual operation status of the actual physical production line is reflected through the virtual equipment components; Based on the flow time and completion status of each process of the physical production line in the production progress data, the digital twin model is used to simulate the product processing path and the distribution status of the work-in-progress, and display the real-time progress of the production process; Compare and analyze the quality parameters in the product quality data with the preset quality standards, and use the digital twin model to visually mark the equipment links and / or process steps that may actually affect the product quality; Through the built-in physical engine and process logic algorithm in the digital twin model, the virtual production line is dynamically simulated. Based on the kinematic and dynamic models of the equipment, the operating behavior of each device under different working conditions is simulated in the digital twin model, the possible failure modes and failure probability of the equipment are predicted, and the impact of the fluctuation of the process parameters of each device in the production process on product quality and production efficiency is analyzed; Based on the simulation results, data analysis and pattern recognition technology are used to comprehensively evaluate the operating status of the production line, identify potential problems, and generate a detailed simulation analysis report. The report includes a description of the problem, possible cause analysis, and corresponding improvement suggestions.

6. The intelligent manufacturing production line lifecycle management and control method based on digital twin and AI collaboration according to claim 1 is characterized in that: When making intelligent decisions and optimizations through the AI ​​system, in the process of automatically adjusting production parameters, the neural network algorithm based on deep learning trains the model through a large amount of historical production data and real-time operation data, predicts the optimal combination of production parameters under different product types, equipment conditions and production environments, and realizes automatic adjustment of production parameters through the integrated equipment control system; In production scheduling optimization, we use intelligent optimization algorithms such as mixed integer programming and genetic algorithms to comprehensively consider factors such as order delivery time, equipment capacity, material supply, and human resources to build a production scheduling model to achieve the allocation of production tasks and optimization of production processes. For equipment maintenance prediction, we use time series analysis and machine learning algorithms to deeply mine and analyze equipment operation data, establish an equipment failure prediction model, and predict the time, type, and probability of equipment failure in advance.

7. The intelligent manufacturing production line lifecycle management and control method based on digital twin and AI collaboration according to claim 1 is characterized in that: The steps of feeding back the actual operation status of the adjusted physical production line to the digital twin model for data verification and model update include: Real-time monitoring and analysis of newly generated operating data of the adjusted physical production line; Evaluate the effectiveness of the AI ​​system’s optimization decisions on the performance of the physical production line by comparing the key performance indicators before and after the physical production line adjustment, including equipment stability, production efficiency improvement, and product quality fluctuations; If the adjusted physical production line finds data anomalies or the optimization effect does not meet expectations, the digital twin model update program will be started; If new problems or data anomalies still exist after the digital twin model is updated, the new data will be integrated with the historical data, and the digital twin model and AI system will be retrained and optimized using incremental learning and model adaptive adjustment.

8. The intelligent manufacturing production line lifecycle management and control method based on digital twin and AI collaboration according to claim 1 is characterized in that: Establish an adaptive adjustment mechanism for models and algorithms. For digital twin models, establish a dynamic model update mechanism to automatically identify the parts of the model that need to be adjusted based on equipment updates, process improvements or changes in the production environment of the physical production line. Through parameter learning of new equipment, logical embedding of new processes, and quantitative analysis of environmental factors, the structure and parameters of the digital twin model are updated in real time to ensure that the actual status of the physical production line can be continuously and accurately reflected. For AI systems, adopt an adaptive learning strategy to automatically adjust the model structure and hyperparameters of the algorithm according to the characteristics of production data and changes in management and control objectives at different stages. Use online learning and transfer learning techniques to enable AI algorithms to quickly adapt to new data distributions and production task requirements.

9. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the intelligent manufacturing production line lifecycle management method based on the collaboration of digital twins and AI as described in any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the intelligent manufacturing production line lifecycle management method based on digital twin and AI collaboration described in any one of claims 1 to 8 are implemented.

Images