Steel hot rolling control system and method based on digital twin

Abstract

The present invention provides a steel hot rolling control system and method based on digital twins, comprising: a perception module for sensing the physical plant environment, equipment operating conditions, and hot-rolled slab quality attributes; a data module for collecting, storing, fusion, transmitting, and analyzing the perceived environment, equipment, and hot-rolled slabs; a presentation module for representing the production conditions of the actual physical plant using virtual twin images through three-dimensional graphics rendering and augmented reality technology based on the analyzed data; and a control module for generating decision-making control recommendations based on a fault knowledge base and issuing control instructions, according to control objectives and corresponding control rules. The present invention uses a fault diagnosis method based on a multi-scale convolutional neural network to monitor the operating status of the equipment, effectively identifying abnormal conditions in the equipment and matching the corresponding fault knowledge base to provide repair recommendations.

Description

Technical Field

[0001] The present invention relates to the field of simulation technology, and in particular to a steel hot rolling control system and method based on digital twins. Background Art

[0002] The concept of digital twins was first proposed by the US Apollo program, which used virtual twins to monitor various indicators of the space shuttle. Digital twin technology involves digitizing the physical world to create a nearly identical virtual simulation. This process integrates and interacts with different control layers within the production process, integrating artificial intelligence, edge computing, and data mining to achieve a comprehensive, simultaneous mapping of the physical and virtual worlds. Digital twin technology is primarily used in product design, R&D, manufacturing, service, and operations. It is a management and control technology for the entire product lifecycle, reducing costs and improving manufacturing efficiency.

[0003] Since the 21st century, with the vigorous development of industrial internet-related technologies, the level of intelligent industrial manufacturing has been significantly improved. Developed countries, led by the United States and Germany, have proposed the concept of cyber-physical systems (CPSs), aiming to achieve a complete integration of the physical world and the virtual cyber world. Digital twin technology, as a representative supporting technology for CPSs, has attracted widespread attention. It accurately maps all data from the production process into the virtual world in real time, conducts comprehensive analysis and decision-making within the virtual world, and uses the results of these decisions to reversely control the physical production process, achieving lean production, effectively reducing energy and material waste, and mitigating safety hazards in the production process. This significantly improves production efficiency and is of great significance to the advancement of enterprise informationization and intelligence.

[0004] The digital twin factory provides a 3D, visual mirror model of the physical factory. This factory twin, formed entirely through data mapping from the physical factory, monitors the physical factory's status in real time. Combined with the big data optimization capabilities of the factory service system, this optimized factory data is simulated, allowing the digital twin factory to achieve production within the virtual manufacturing process. The 3D visualization of the factory makes simulations more realistic and effective. Direct feedback on issues encountered during the simulation provides guidance on operational planning and eliminates defects in the actual production process.

[0005] However, digital twin technology is currently immature. Most current research focuses solely on process modeling, data fusion, and interactive collaboration, with little investment in service and production applications. In particular, most research is focused solely on the product design phase, serving as the initial stages of product development. Investment in the actual manufacturing process and production control is minimal. Furthermore, while traditional production plant management has largely achieved automation and information-based management, it lacks system simulation analysis and decision-making feedback mechanisms. Information barriers exist between different control levels, preventing the interoperability of all information. Data visualization is also limited, making it impossible to present the entire production process through intuitive data charts or 3D images. Simulation analysis and decision-making models are also lacking to support production. Therefore, understanding the implications of digital twin technology and effectively leveraging it to enhance intelligent plant management and control and achieve virtual manufacturing are crucial issues that need to be addressed.

[0006] Patent document CN107423458A (application number: 201710132969.8) discloses a steel production simulation system that integrates multiple functional operating platforms and is organized in a modular manner, including a scene roaming module, a human-computer interaction module, a production plan input module, a production plan execution module, and a simulation system evaluation module, which constitute the framework of the entire system. The present invention introduces virtual environment technology into the entire process of steel production and establishes a physical model of the steel production line, so that it can intuitively understand the equipment structure and the actual performance and production efficiency after production during the design period, and evaluate the plan. This patent is mainly for the implementation of a steel production virtual simulation system, which uses a virtual production line to pre-simulate the production plan to be executed, and its main function is to simulate and evaluate the process of the production line. The present invention mainly embodies a steel hot rolling process control system based on digital twin technology, emphasizing the characteristics of digital twin technology - accurate real-time tracking and analysis and evaluation. The system also includes an equipment monitoring module, and analyzes equipment status trends based on a fault diagnosis algorithm based on a convolutional neural network. Finally, the synchronous tracking or simulation process is presented in the form of augmented reality and large-screen projection. Compared with the above inventions, it has greatly expanded its functions and is more mature in technology and performance. Summary of the Invention

[0007] In view of the defects in the prior art, the purpose of the present invention is to provide a steel hot rolling control system and method based on digital twin.

[0008] The steel hot rolling control system based on digital twin provided by the present invention includes:

[0009] Perception module: Perceives the physical plant environment, equipment operating conditions, and hot-rolled slab quality attributes;

[0010] Data module: performs data collection, data storage, data fusion, data transmission and data analysis on the perceived environment, equipment and hot-rolled plates;

[0011] Performance module: Based on the analyzed data, the system uses 3D graphics rendering and augmented reality technology to represent the production status of the real physical factory using virtual twin images;

[0012] Control module: Based on the transmission and fusion of heterogeneous data, through comprehensive calculation and analysis of data in the information space, it drives the operation of virtual scene equipment and the movement and deformation of slabs based on preset business rules, forming a real-time simulation animation of the entire production line. When an abnormal condition occurs in the equipment, decision-making control suggestions are formed based on the fault knowledge base according to the control objectives and corresponding control rules, and control instructions are issued.

[0013] Preferably, heterogeneous production equipment and material data provide real-time data of each data point in the form of OPC SERVER, and communication is achieved through TCP / UDP communication technology.

[0014] Preferably, the data storage includes: transferring production plan data and equipment status data to a universal database for management, supporting the export of data within any time period, conducting equipment health status trend analysis and historical production plan simulation and reproduction, using cloud center big data analysis platform services for construction, and building the system's data storage layer based on a distributed architecture storage computing service, storing the collected production performance information and quality data information in the distributed database included in the service, and providing external data access services.

[0015] Preferably, the data transfer is based on system integration between different levels, and is used for multi-source data fusion and information interaction.

[0016] Preferably, the data analysis includes parsing of plan instruction driven data and analysis of equipment status trend data, and the analysis results are transmitted to the presentation module.

[0017] Preferably, the rendering module includes: constructing a plant structure and equipment model through three-dimensional point cloud scanning technology and corresponding CAD structure drawings based on the results of data analysis, rendering a proportional virtual simulation scene through graphics rendering technology and environmental map textures, and representing environmental fog, air and water elements based on particle effect rendering technology;

[0018] Particle effects are also applied during signal acquisition, interaction and processing between different layers.

[0019] Preferably, the terminal is displayed through large-screen projection and mobile augmented reality. The large-screen projection projects a virtual image of the entire production line, with the function of graphically displaying complete production data and planning and scheduling data;

[0020] Mobile augmented reality uses small-screen mobile terminals to identify physical world device entities based on plane detection or image feature recognition technology, and simultaneously presents corresponding virtual entities. It supports point-and-click viewing of any device and switching of perspectives, supports partial perspective or cross-sectional viewing of the device, and displays the physical status data and health status of the device in a combination of 2D and 3D graphics.

[0021] The steel hot rolling control method based on digital twin provided by the present invention includes:

[0022] Synchronous tracking steps: Through real-time interaction between the physical and virtual factories, based on the analysis results of drive data signals, real-time responses are made to dynamic changes in the physical world, and synchronous simulation tracking is performed in the virtual scene;

[0023] Planning simulation step: Calling the production plan information within the preset time period, analyzing it and using the data to drive the operation of each device in the virtual production line, guiding production based on the defects of historical or future production plans;

[0024] Equipment monitoring steps: Monitor the equipment's operating status and factors related to potential failures.

[0025] Preferably, the synchronous tracking step includes: performing a full range of real-time virtual simulation of the production line, mapping the hot rolling process into corresponding micro-units by comprehensively considering the factory layout, production line station equipment combination, production line plan and rolling sequence, establishing a digital model of the production process, and establishing a real-time one-to-one mapping relationship between the digital model and the history, sensing, control and logistics information of the actual production line and equipment, thereby obtaining a multi-physical quantity, multi-scale and multi-dimensional digital twin intelligent body that synchronizes the actual production and virtual production in real time;

[0026] Plan its own response mechanism based on production plan data, process data and disturbance data, and coordinately control and optimize the behavior of each production unit under the goal of global optimization.

[0027] Preferably, the equipment monitoring step includes: setting up multiple data collection points on the production line for health diagnosis and monitoring, and through data collection and analysis, using a multi-scale convolutional neural network fault detection method to input the speed, current, and voltage values ​​of the equipment at a certain moment into a trained model to predict the health status of the equipment in real time.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. The equivalent virtual simulation scene constructed by the present invention based on digital twin technology is more detailed and three-dimensional than the general control system scene. It supports two modes of display: large-screen global business data and mobile augmented reality. It is more user-friendly in interaction and is aimed at on-site management personnel and inspection personnel.

[0030] 2. This invention breaks down the information barriers between the physical and virtual worlds through heterogeneous system integration and data fusion. Based on the results of this information interaction, real-time dynamic data drives the operation of virtual simulation scenarios. Managers no longer need to visit the production site to intuitively view the production status of the entire production line from different perspectives. This solves the problems of traditional monitoring systems with limited field of view and non-intuitive information display, making decision-making more convenient.

[0031] 3. The present invention has the function of simulating historical production plan data and unimplemented production plans. It can switch from precise tracking to plan simulation. The system analyzes the planned production information and drives the pre-run of virtual simulation scenarios based on certain process and business rules, which can timely identify deficiencies in future plans and defects in past plans.

[0032] 4. The fault diagnosis method based on the multi-scale convolutional neural network of the present invention monitors the operating status of the equipment, can effectively determine the abnormal conditions of the equipment and match the corresponding fault knowledge base to provide repair suggestions. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0034] Figure 1 Schematic diagram of structure and function. DETAILED DESCRIPTION

[0035] The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those skilled in the art, several changes and improvements can be made without departing from the scope of the present invention. These all fall within the scope of protection of the present invention.

[0036] Example 1:

[0037] like Figure 1 The specific implementation system of the present invention includes the following 4-layer structure:

[0038] The perception layer is used to perceive the environmental conditions of the physical factory, the operating status of the equipment, and the quality attributes of the hot-rolled slabs, while collecting and fusing heterogeneous data. It is the data source for digital twin information interaction, and the subsequent function implementation and status performance will depend on this data.

[0039] The data layer, the core component of the four layers, primarily encompasses data acquisition, data storage, data transmission, and data analysis. Heterogeneous production equipment and material data is provided in real-time via OPC servers, communicated via TCP / UDP. Data storage facilitates the transfer of production plan and equipment status data to a universal database for management. Data export is supported for any time period, enabling analysis of equipment health trends and the simulation of historical production plans. Built as a cloud-based big data analytics platform service, the system's data storage layer, built on the Hadoop distributed architecture, stores production performance and quality data collected from various system servers in the included distributed HBase database, providing high-speed data access. Data transmission, based on system integration across different layers, primarily facilitates multi-source data integration and information exchange. Data analysis processes various production data, primarily parsing plan-driven data and analyzing equipment status trends. The results are then communicated to the presentation layer.

[0040] The presentation layer, primarily based on 3D graphics rendering and augmented reality (AR), uses virtual twins to represent the physical factory's production conditions. 3D point cloud scanning and corresponding CAD structural drawings are used to construct models of the factory's main structures and equipment. Graphics rendering and environmental textures are used to render a full-scale virtual simulation scene. Particle effects are used to represent environmental elements such as fog, air, and water. To enhance data visualization, particle effects are also applied to signal acquisition, interaction, and processing at different levels. The system primarily uses large-screen projection and mobile AR to present the system on the terminal. Large-screen projection projects a virtual image of the entire production line, providing a comprehensive graphical display of production data and scheduling plans. Mobile AR, primarily using small, portable screens, uses plane detection or image feature recognition to identify physical equipment entities and simultaneously present the corresponding virtual entities. The system supports point-and-click viewing of any device, switching perspectives, and viewing partial perspectives or sections of equipment. It also presents the physical status and health of the equipment using a combination of 2D and 3D graphics.

[0041] The control layer is mainly based on the transmission and fusion of heterogeneous data. Through comprehensive calculation and analysis of data in the information space, it drives the operation of virtual scene equipment and the movement and deformation of slabs based on certain business rules, forming a real-time simulation animation of the entire production line. In terms of equipment monitoring, when an abnormal condition occurs in the equipment, the corresponding control objectives and control rules are considered, and decision-making control suggestions are formed based on the fault knowledge base, and control instructions are issued.

[0042] In addition, the specific implementation system of the present invention includes the following three major functions:

[0043] Precise tracking function: that is, all-round real-time virtual simulation of the production line, mainly for the hot rolling process, comprehensively considering the factory layout, production line station equipment combination, production line plan and rolling sequence and other data, mapping them into corresponding micro-units, and establishing a digital model of the production process. In addition, a real-time one-to-one mapping relationship is established between the digital model and the history, sensing, control, logistics and other information of the actual production line and equipment, realizing a multi-physical quantity, multi-scale and multi-dimensional digital twin intelligent body that synchronizes actual production and virtual production in real time. Through the real-time interaction between the physical factory and the virtual factory, based on the analysis results of the drive data signal, it can respond to the dynamic changes of the physical world in real time and achieve synchronous simulation tracking in the virtual scene. This function also has the ability to plan its own response mechanism based on production plan data, process data and disturbance data, and to coordinate and optimize the behavior of each production unit under the goal of global optimization.

[0044] Planning simulation function: The execution is the same as the precise tracking function, but the difference is that it mainly targets the historical production plans and unimplemented production plan instruction information stored in the data layer. By calling the production plan information within a specified time period, and parsing and conveying it to the system execution module based on certain rules, the operation of each device in the virtual production line is driven by data, thereby observing the defects of historical or future production plans and guiding production.

[0045] Equipment monitoring monitors factors related to equipment operating status and potential failures. Production line health diagnosis and monitoring requires multiple data collection points. By collecting and analyzing data and applying multi-scale convolutional neural network fault detection methods, the speed, current, voltage, and other values ​​of the equipment at a specific moment are fed into a trained model. This allows for real-time prediction of equipment health, prompt identification of faults, and feedback on solutions. This effectively addresses the issues of manual inspections, which often lack pertinence, accuracy, and effectiveness, and the timeliness of responses to abnormal status events.

[0046] Example 2:

[0047] According to the present invention, a steel hot rolling control system and method based on digital twin technology is provided. It mainly focuses on the steel hot rolling process scenario, constructs a virtual simulation scenario equivalent to the physical factory, and drives the operation of the virtual simulation entity based on technologies such as digital twin and big data computing. It realizes the virtual twin precise tracking of the physical factory production situation, pre-simulation of historical and future production plans, and monitoring and health prediction of key production equipment. It also presents two-dimensional or three-dimensional visual data elements of business and production status through large-screen projection or mobile augmented reality, thereby obtaining an immersive interactive experience of the factory. This embodiment mainly includes four layers: perception layer, data layer, presentation layer, and control layer. In terms of function, it is mainly divided into precise tracking, plan simulation, and equipment monitoring.

[0048] Among them, the perception layer described in this embodiment requires a large number of sensor components, data acquisition equipment and signal transmission equipment. The communication between the devices at each layer is realized through TCP / IP and field bus, so as to perceive the environmental conditions of the physical factory, the operating status of the equipment and the quality attributes of the hot-rolled slab, etc., and at the same time collect heterogeneous data and fuse information.

[0049] The data layer described in this embodiment includes four links: data acquisition, data storage, data transmission and data analysis. Heterogeneous production equipment and material data provide real-time data of each data point in the form of OPCSERVER, and communication is achieved through TCP / UDP communication technology; the data storage link transmits production plan data and equipment status data to a universal database for management, supports the export of data within any time period, and is used as an analysis of equipment health status trends and simulation and reproduction of historical production plans. It is constructed in the form of a cloud center big data analysis platform service, and the storage computing service based on the Hadoop distributed architecture builds the system's data storage layer. The production performance information and quality data information collected from each system server are stored in the distributed HBase database included in the service, providing high-speed data access services to the outside world. The data transmission link performs multi-source data fusion and information exchange based on system integration between different levels; the data analysis link parses the plan instruction driven data and analyzes the equipment status trend data, and transmits the results of the processing and analysis to the presentation layer.

[0050] The presentation layer described in this embodiment is mainly based on three-dimensional graphics rendering technology and augmented reality technology, and uses virtual twin images to represent the production situation of real physical factories. It mainly has the following three points: First, the main structure and equipment model of the factory area is constructed through three-dimensional point cloud scanning technology and corresponding CAD structure drawings, and the proportional virtual simulation scene is rendered through graphics rendering technology and environmental map textures. The environmental fog, air and water elements are represented based on particle effect presentation technology. At the same time, in order to improve the intuitiveness of the data, particle effects are also applied to the signal acquisition, interaction and processing processes between different levels. Second, the optimized digital model, optimized process parameters and business processes are used in the virtual twin scene to make it the fundamental driving force of the scene. Third, it needs to be supported by the business model, and through the collection and processing of actual data, the problems existing in the object to be solved are analyzed.

[0051] The presentation layer primarily uses the Unity graphics engine for project development. Equipment models constructed in 3dMax modeling software are imported into the engine to create a 3D model of the factory. Actual production process parameters and related data are introduced into the virtual world and assembled into a virtual production line within the levels. Water and mist effects are created within the engine to enhance the production process. Triggers are used within the levels to trigger the water and mist effects when a slab passes through the rolling equipment, simulating a real-world production environment. Using C# as the primary development language, C# scripts are written to retrieve production status information from the cloud server for each piece of equipment. A user interface is then designed to display real-time production data for each piece of equipment.

[0052] The system mainly uses large-screen projection and mobile augmented reality on the display terminal. The large-screen projection mainly projects the virtual image of the entire production line, and has the function of graphically displaying complete production data and planning and scheduling data; mobile augmented reality mainly relies on movable small-screen terminals to identify physical world equipment entities based on image feature recognition technology, and synchronously presents the corresponding virtual entities. It supports the switching of any device and the partial perspective or cross-sectional viewing of the device, and displays the physical status data and health status of the equipment in a combination of 2D graphics and 3D graphics.

[0053] The control layer described in this embodiment is mainly based on the transmission and fusion of heterogeneous data. Through comprehensive calculation and analysis of data in the information space, it drives the operation of virtual scene equipment and the movement and deformation of slabs based on certain business rules, forming a real-time simulation animation of the entire production line; in terms of equipment monitoring, when an abnormal condition occurs in the equipment, the corresponding control objectives and control rules are considered, and decision-making control suggestions are formed based on the fault knowledge base, and control instructions are issued.

[0054] The precise tracking function and plan simulation function described in this embodiment are all-round virtual simulations of the production line, which are mainly aimed at the hot rolling process, and establish a one-to-one mapping relationship between the digital model and the history, sensing, control, logistics and other information of the actual production line and equipment, so as to realize the multi-physical quantity, multi-scale and multi-dimensional digital twin intelligent body that synchronizes the actual production and virtual production in real time. Through the information interaction between the physical factory and the virtual factory, based on the analysis results of the driving data signal, timely real-time response is made to the dynamic changes of the physical world, and synchronous simulation tracking is achieved in the virtual scene. Among them, the plan simulation function is that it mainly targets the historical production plans and unimplemented production plan instruction information stored in the data layer, and by calling the production plan information within the specified time period, and parsing and conveying it to the system execution module based on certain rules, the operation of each device of the virtual production line is driven by data, thereby observing the defects of historical or future production plans and guiding production.

[0055] The equipment monitoring function described in this embodiment monitors factors related to equipment operating status and potential faults. Production line health diagnosis and monitoring requires multiple data collection points. By collecting and analyzing data and applying a multi-scale convolutional neural network fault detection method, the equipment's rolling force, speed, current, roll gap, and other values ​​at a specific moment are fed into a trained model. This allows for real-time prediction of equipment health, prompt identification of faults, and feedback on solutions for reference.

[0056] First, build the equipment model in 3dMax or other modeling software, export it as an .fbx file, and import it into the Unity engine to generate a prefab for easy access. When writing C# code, treat each prefab file as an object. Store all used model prefab files in a list or dictionary container to form an object pool. When one or more models are needed, they are removed from the pool and returned to the pool after use. Although this approach consumes memory, it improves application response time and operational efficiency. The online data monitoring function requires access to specific data from factory equipment. Real-time production data from factory equipment is transmitted to a cloud server via the Internet of Things. Using the AR device's camera as the image acquisition device, the ASIFT device local feature recognition method is used to identify specific equipment components. The corresponding C# code is then called to generate a 3D model in virtual space from the imported model file. A 3D interactive scene of the component is then constructed and displayed on the AR device's display. Using real-time positioning and mapping technologies, the surrounding environment is constructed in real time to obtain the world coordinates of the object. Spatial anchor technology is used to anchor the virtual scene in space, creating the illusion of virtual equipment superimposed on real production equipment. Specific production data from scanned equipment locations is acquired through the cloud server's communication protocol and displayed on a virtual world dashboard. This acquired data is then compared with normal ranges derived from big data analysis. If a value falls outside this range, an alarm sounds in the virtual world, prompting the user to inspect the equipment. The system uses C# code to compile common equipment anomaly handling procedures and the equipment's technical manual into a behavior tree, creating an interactive technical manual. If a value is abnormal, the system matches the anomaly handling steps in the technical manual, providing the user with the fastest and most comprehensive troubleshooting solution.

[0057] Those skilled in the art will appreciate that, in addition to implementing the system, device, and various modules provided by the present invention in purely computer-readable program code, it is entirely possible to implement the same program in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, embedded microcontrollers, and the like by logically programming the method steps. Therefore, the system, device, and various modules provided by the present invention can be considered a hardware component, and the modules included therein for implementing various programs can also be considered structures within the hardware component; the modules for implementing various functions can also be considered both software programs for implementing the method and structures within the hardware component.

[0058] The above describes specific embodiments of the present invention. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. The embodiments of this application and the features in the embodiments may be combined with each other in any manner unless there is a conflict.

Claims

1. A steel hot rolling control system based on digital twin, characterized by: include: Perception module: Perceives the physical plant environment, equipment operating conditions, and hot-rolled slab quality attributes; Data module: performs data collection, data storage, data fusion, data transmission and data analysis on the perceived environment, equipment and hot-rolled plates; Performance module: Based on the analyzed data, the system uses 3D graphics rendering and augmented reality technology to represent the production status of the real physical factory using virtual twin images; Control module: Based on the transmission and integration of heterogeneous data, through comprehensive calculation and analysis of data in the information space, it drives the operation of virtual scene equipment and the movement and deformation of slabs based on preset business rules, forming a real-time simulation animation of the entire production line. When an abnormal condition occurs in the equipment, decision-making control suggestions are formed based on the fault knowledge base according to the control objectives and corresponding control rules, and control instructions are issued; Heterogeneous production equipment and material data provide real-time data of each data point in the form of OPC SERVER, and communication is achieved through TCP / UDP communication technology; The data storage includes: transferring production plan data and equipment status data to a universal database for management, supporting the export of data within any time period, conducting equipment health status trend analysis and historical production plan simulation and reproduction, using cloud center big data analysis platform services for construction, and building the system's data storage layer based on a distributed architecture storage computing service. The collected production performance information and quality data information are stored in the distributed database included in the service, and data access services are provided to the outside world.

2. The steel hot rolling control system based on digital twin according to claim 1 is characterized in that: The data transmission is based on system integration between different levels and is used for multi-source data fusion and information interaction.

3. The steel hot rolling control system based on digital twin according to claim 1 is characterized in that: The data analysis includes parsing the plan instruction driven data and analyzing the equipment status trend data, and the analysis results are transmitted to the presentation module.

4. The steel hot rolling control system based on digital twin according to claim 1 is characterized in that: The rendering module includes: constructing a plant structure and equipment model based on the results of data analysis through 3D point cloud scanning technology and corresponding CAD structural drawings, rendering a proportional virtual simulation scene through graphics rendering technology and environmental mapping textures, and representing environmental fog, air and water elements based on particle effect rendering technology; Particle effects are also applied during signal acquisition, interaction and processing between different layers.

5. The steel hot rolling control system based on digital twin according to claim 1 is characterized in that: The terminal is displayed through large-screen projection and mobile augmented reality. The large-screen projection projects a virtual image of the entire production line, with the function of graphically displaying complete production data and planning and scheduling data; Mobile augmented reality uses small-screen mobile terminals to identify physical world device entities based on plane detection or image feature recognition technology, and simultaneously presents corresponding virtual entities. It supports point-and-click viewing of any device and switching of perspectives, supports partial perspective or cross-sectional viewing of the device, and displays the physical status data and health status of the device in a combination of 2D and 3D graphics.

6. A steel hot rolling control method based on digital twin, characterized in that: The steel hot rolling control system based on digital twin according to claim 1 comprises: Synchronous tracking steps: Through real-time interaction between the physical and virtual factories, based on the analysis results of drive data signals, real-time responses are made to dynamic changes in the physical world, and synchronous simulation tracking is performed in the virtual scene; Planning simulation step: Calling the production plan information within the preset time period, analyzing it and using the data to drive the operation of each device in the virtual production line, guiding production based on the defects of historical or future production plans; Equipment monitoring steps: Monitor the equipment's operating status and factors related to potential failures; The synchronization tracking step includes: conducting a full range of real-time virtual simulation of the production line, mapping the hot rolling process into corresponding micro-units, taking into account the factory layout, production line station equipment combination, production line plan and rolling sequence, and establishing a digital model of the production process. The digital model is then mapped in real time to the historical, sensory, control and logistics information of the actual production line and equipment, thereby obtaining a multi-physical, multi-scale and multi-dimensional digital twin intelligent entity that synchronizes the actual production with the virtual production in real time; Plan its own response mechanism based on production plan data, process data and disturbance data, and coordinately control and optimize the behavior of each production unit under the goal of global optimization.

7. The steel hot rolling control method based on digital twin according to claim 6, characterized in that: The equipment monitoring step includes setting up multiple data collection points on the production line for health diagnosis and monitoring. By collecting and analyzing the data and using a multi-scale convolutional neural network fault detection method, the speed, current, and voltage values ​​of the equipment at a certain moment are input into a trained model to predict the health status of the equipment in real time.

Images