Abstract: As an important component of steel enterprises' information management, the continuous casting process control system integrates advanced process mathematical models and control technologies, greatly improving the quality of cast slabs, facilitating production management, enhancing enterprise competitiveness, and showing a broad development prospect. Keywords: process control; process mathematical model; high-quality cast slab 0 Introduction With the rapid development of China's steel industry, replacing the old-fashioned, decentralized, and human-based management model with a unified information-based management model has become the goal of various steel enterprises. Currently, information-based management models have been implemented in some large enterprises, fully demonstrating their advantages of easy management, resource sharing, saving manpower and material resources, and greatly improving production efficiency and quality. As a crucial component in achieving the informatization goals of steel enterprises, the continuous casting process control system integrates advanced processes and control technologies, significantly improving billet quality, facilitating production management, and enhancing enterprise competitiveness. The hierarchical structure of steel enterprise information management systems can generally be divided into: ERP resource management system; production management system (Level 3); manufacturing execution system (MES); process control system (Level 2); and basic automation (Level 1). The continuous casting process control system is a process control system built on the continuous casting production line, connecting the continuous casting basic automation (Level 1) system and the steelmaking production management (Level 3) system, thus integrating the production information system and the production basic automation system. It is a vital and irreplaceable link in improving product quality. 1. Hardware and Software Composition Because the continuous casting process control system needs to ensure the security of production information data storage, the high concurrency and real-time performance of model calculations, and the ease of operation, manageability, and stability of the entire system, the rational selection and configuration of hardware and software are extremely important. 1.1 Hardware Structure The hardware system mainly consists of one database server (SVR_DBS), one application server (VR_APP), two client machines (CLIENT1 and CLIENT2), one 3COM switch, one CD burner, one network printer, and several network connection devices (network cards, network cables). The database server is an HP ProLiant DL360 G5 rack-mount server with a redundant ET200 RAID card (RAID 5) and four large-capacity hot-swappable SCIS hard drives. This allows for hot-swapping of any failed hard drive to replace it and automatically recover data, ensuring system security. The CD burner can periodically burn historical data for archiving. The application server is an HP ProLiant DL360 G5 rack-mount server. Its Xeon-level dual-core CPU processing power and large-capacity L1 and L2 caches and memory ensure real-time model calculations and high-concurrency processing capabilities. The application server uses dual network cards, one connected to the basic automation system and the other to an L2-level 3COM switch, serving as the isolation and interface point between L1 and L2 networks. Client computers are HP DC7700 commercial computers. One client is located in the control room, allowing process engineers and operators to monitor the production process status and model calculation results, and to provide reasonable production suggestions to operators (such as suggesting stable production speeds). The other client is located in the cutting room, primarily calculating the optimal cutting length based on the collected production process data using a cutting optimization model, providing this information to operators for reference. An HP 1020n network printer is used to print production reports. The database server, application server, clients, and network printer are connected to the 3COM switch via network connectivity devices, forming a 100M Ethernet network. A simplified hardware structure diagram is shown below: 1.2 Software Structure The software structure adopts a server/client architecture. The entire system is centered around the database. The communication software in the application server collects real-time data from the basic automation system and saves it to the Oracle database on the database server. The calculation model extracts data from the database, performs calculations, and then stores the results back in the database for client monitoring and display, or downloads them to the basic automation system via the communication program to guide production. The database server runs on Windows 2003 Server and uses Oracle 10g database software for storing, maintaining, and managing all real-time and historical data. The application server runs on Windows 2003 Server and uses Visual Studio 2005. The communication software is Siemens Industrial Ethernet SoftNet-S7 for Windows (supporting OPC). The process calculation mathematical model runs as a service in the background. The client machines run on Windows XP and use CCM_PCS application software. These clients primarily display production plans, production information, process information, equipment management, and model calculation results; they also provide interfaces requiring manual input and intervention, and manage various subsystems. The simplified software logic structure diagram is as follows: 2. Functional Description The continuous casting secondary process control system mainly performs the following functions: maintenance and management of the metallurgical database; maintenance and management of continuous casting production process information; information management of the use and maintenance of major equipment; collection, storage, display, analysis of on-site process data and log recording of important operations; tracking of the production process; operation and monitoring of the process mathematical model; guidance for process personnel; access control for operation and maintenance personnel; and generation of production reports. 2.1 Maintenance and Management of the Metallurgical Database The metallurgical database is the information data warehouse of the continuous casting process control system. All real-time data, historical data, model calculation results, production plan information, steel grade information, etc., collected from the production process are stored in it. The database is automatically backed up once periodically, and maintenance personnel can periodically burn the backup files to CDs for archiving. Maintenance and management personnel can also operate the database through the human-computer interaction interface on the client to complete the maintenance and management of the database. 2.2 Continuous Casting Production Management Continuous casting production management includes continuous casting production planning, steel grade management and maintenance, smelting rhythm, and material tracking. 2.2.1 Continuous Casting Production Plan According to the production process requirements, the production plan must be issued before the ladle arrives at the rotary table. Once the ladle arrives, it cannot be modified. The production plan data processing system has the following functions: 1) Receive and save plan data (including casting cycle plan, heat cycle plan, and cutting plan) received from the manufacturing execution system or manually input the plan through the client's human-machine interface. 2) Maintain and manage the plan through the client's human-machine interface (including viewing, adding, modifying, and deleting). 3) Generate casting commands to guide production based on the plan. 2.2.2 Steel Grade Management and Maintenance 1) Receive or manually input the composition standards (maximum value, target value, minimum value) of various steel grades from the previous level, generally divided into national standards and plant standards (internal control standards), and save them to the metallurgical database. 2) Receive or manually input the actual composition of molten steel detected by the laboratory from the previous level and save it to the metallurgical database. 3) Various steel grade standards and molten steel sampling compositions can be maintained and managed through the client's human-machine interface (including viewing, adding, modifying, and deleting). 2.2.3 Smelting Rhythm To ensure continuous casting, the steelmaking process and the casting machine process should maintain a matching furnace rhythm. The smelting rhythm provides the following information to support the matching process: 1) The estimated time for the current ladle to finish pouring molten steel. 2) Display of the ladle tapping time at the previous station (refining) (input from the previous station). 3) Calculate the recommended stable casting speed and recommended maximum casting speed based on the current casting steel grade, ladle weight, temperature, the weight of the remaining molten steel in the ladle, and the tapping time of the ladle at the previous station to match the smelting rhythm. This information is periodically refreshed and displayed on the HMI. 2.2.4 Material Tracking Process The computer system will track each ladle of molten steel from its arrival at the rotary table until it is cut into fixed-length billets. The computer will record the historical information of the state of each ladle of molten steel at different positions during the continuous casting process. Material tracking mainly includes furnace tracking, casting flow tracking, and slab tracking in the billet exit area. 2.2.4.1 Furnace Tracking Furnace tracking mainly includes information on each ladle of molten steel (casting number, furnace number, composition, steel grade, etc.); information collection on the molten steel's arrival and departure from the turret (arrival and departure time, weight, temperature, etc.); and information on the slabs generated from each furnace of molten steel (number of slabs, specifications, etc.). All this data is stored in the database for operator querying, analysis, and report generation. 2.2.4.2 Casting Flow Tracking Casting flow tracking provides production information throughout the entire process from the tundish, crystallizer, casting flow body to slab cutting: automatically calculating the usage of media (such as water, gas, coal gas, etc.) and storing it in the database; automatically tracking furnace joints and responding to abnormal events; the system divides the entire area of ​​casting slabs into many logical 'segments', linking each event to the specific location of each segment, tracking the process information and events of each segment, and using the collected historical information of each segment's casting slabs as the basis for judging casting slab quality. 2.2.4.3 Slab Tracking: Tracks the slab position in the output area (from the cutting machine to when the slab leaves the output roller conveyor), collecting processing information for each slab (including marking, deburring, weighing, etc.); it also collects information on slabs going to the next process. 2.2.5 Production Information Query and Management: Key field data collected in real-time is displayed on the human-machine interface (current heat, steel grade, specifications, cooling water volume, etc.). Saved production information (casting information, heat information, slab information, etc.) can be queried, added, modified, and deleted. Important data and operational information are saved to a database to form historical data and log files, and historical curves can be generated or exported to analysis software for analysis, providing process engineers with a basis and means to find faults and analyze processes. 2.3. Main Equipment Information Management: The usage and maintenance information (lifespan, specific information on each maintenance, manufacturer, materials, etc.) of the ladle, intermediate ladle, crystallizer, and sector section is stored in the database, and this information can be queried, edited, and maintained. 2.4 Mathematical Model for Process Control 2.4.1 Mathematical Model for Secondary Cooling Water Control The process computer derives the mathematical control model for secondary cooling water based on heat conduction theory and empirical formulas, according to different steel grades, cross-sectional dimensions, and other process parameters. The process computer calculates the cooling water flow rate of each secondary cooling zone based on the collected actual billet drawing speed. However, the calculated functional relationship between cooling water volume and drawing speed is discrete, which inevitably brings a large amount of complex calculation work to water volume control. Due to the discontinuity of water volume control, the surface quality of the billet is inevitably affected. Therefore, the least squares method is used for fitting to form a quadratic equation functional relationship between cooling water volume and drawing speed. The quadratic equation can be expressed as: Qi = Ai＊Vg↑2 + Bi＊Vg + Ci Qi: (l/min) Vg: (m/min) The billet pulling speed Ai, Bi, Ci: Corresponding to the water volume setting value of a certain section of the secondary cooling system. Based on the actual billet pulling speed and the water volume calculated by the quadratic equation, dynamic compensation and correction are also performed based on factors such as the actual tundish temperature and secondary cooling water temperature before being downloaded to the basic automation system. 2.4.2 Optimal Cutting Calculation Model The optimal cutting optimization model includes optimal tail billet cutting and optimal length cutting during continuous casting with tundish replacement. The purpose of the optimal cutting optimization model is to minimize the loss of billet quantity and reduce scrap to a minimum. Control method: The normal cutting length of the billet is determined by the process and issued by the process computer to the PLC of the flame cutting machine for execution. When at the end of the casting process, the process computer can provide an immediate shut-off prompt based on the length of the billet in the casting stream. When the molten steel in the ladle is poured and the stopper is closed, or when different steel grades are continuously poured, the process computer can instantly calculate the optimal cutting size combination based on the remaining length of the slab in the casting flow, the target length, the maximum length, and the minimum length, to ensure maximum utilization of the slab and maximize the steel yield. 2.4.3 Quality Judgment Mathematical Model The quality judgment model analyzes and judges the slab quality based on the abnormal production information recorded in each segment of the casting flow tracking and the degree of impact of the abnormal information on the slab, and evaluates the slab quality grade. Control Method: First, based on the degree of impact of various events that may occur during the casting process on the slab's microstructure and properties, each event is quantitatively classified into a quality grade number. During the slab formation process, the computer divides the slab into several "segments" of a certain length. Combining the casting flow tracking model and tracking the segments, when each segment passes through a real-time process parameter monitoring point, the parameters are continuously compared with the critical values, and the abnormal event parameters are stored as evaluation parameters for the slab quality. Each segment can record four most severe events, but only the highest priority level is used for judgment. After the slab is cut, the quality information (level number) of the corresponding segment is extracted. The quality of the slab and the overall quality level are determined based on the level of harmful events in each segment and their proportion in all segments. This information is then printed and recorded, and the final determination is as follows: scrap, retained (requiring further evaluation or processing), or good. Many factors affect the quality of the slab, such as: excessive steel composition, excessive tundish overheating, nozzle damage, cooling water valve malfunction causing abnormal cooling, and abnormal vibration system. 2.4.4 The process control system of the casting speed optimization model automatically calculates the appropriate current recommended stable casting speed and recommended maximum casting speed based on the current steel composition (mainly carbon content), width, thickness, tundish steel temperature, remaining steel volume, expected arrival time of the bottom ladle, crystallizer cooling, secondary cooling, and other information. This information is then displayed to the operator to adjust the production rhythm, prevent steel leakage, or appropriately increase the speed to improve production efficiency. 2.4.5 The hydraulic vibration model process control system selects suitable vibration modes and process parameters based on the steel grade being cast. It mainly consists of two forms: sinusoidal (frequency, amplitude) and non-sinusoidal (frequency, amplitude, and skew angle). The frequency and amplitude can be preset, vary with casting speed, or be determined by the target speed, negative slip ratio, speed, and amplitude. Generally, the vibration amplitude is less than 6mm, and the vibration frequency does not exceed 380 Hz. A starting vibration frequency greater than 60 Hz ensures system stability and avoids wear on the vibrating machinery. Simultaneously, the model automatically calculates the vibration process parameters: positive slip time, negative slip time, negative slip speed ratio, and negative slip ratio, allowing operators and process engineers to monitor the process effect in real time and adjust parameters accordingly to achieve optimal demolding and reduce vibration marks. 2.3.6 The dynamic light reduction model process control system calculates the billet shell thickness and surface temperature at each roller coordinate based on factors such as steel grade, slab thickness and width, crystallizer water volume, secondary cooling water volume, gas volume, casting speed, tundish molten steel temperature, and production environment parameters during the casting process, using heat conduction theory. This helps identify the solidification endpoint of the slab's liquid core. Based on the different shrinkage rates of different steel grades, the dynamic light reduction model calculates a set value for the roll gap reduction within a certain range before the solidification endpoint of the billet and transmits it to the remote roll gap adjustment system for execution. Generally, the light reduction position is within the slab solids fraction range of 60%–80%, with a unit light reduction not exceeding 1.5 mm/m. Light reduction can significantly improve segregation defects in the internal structure at the solidification endpoint of the slab. 2.5 Production Report System generates and prints reports according to customer requirements through a human-machine interface system based on historical production information stored in the database. The main production reports are as follows: 1) Casting cycle report: summarizes the number of furnaces, steel consumption, slab information, and medium consumption in a casting cycle, providing a basis for cost accounting and output accounting in the factory. 2) Furnace report: data collected during the casting process of a furnace of molten steel: furnace steel weight, production time, steel composition, slab condition, etc. 3) Shift report: a summary of production for a shift, providing a basis for shift output assessment. 4) Daily production report: a summary of production for a specified day. 2.6. Communication Interface The process computer and external computer systems are connected via Ethernet, and the basic automation system is connected via industrial Ethernet, while providing standard software support and communication application software. The process computer has communication capabilities with the following computer systems: 1) Level 3 computer 2) Converter Level 2 computer 3) Refining Level 2 computer 4) Rolling Mill Level 2 computer 3. Conclusion Currently, the Ministry of Information of China has proposed the strategic goal of building information-based enterprises, and establishing information-based enterprises has become an inevitable trend for the development of steel enterprises. As a crucial component in achieving the informatization goals of steel enterprises, the continuous casting process control system employs advanced process models and control technologies, which greatly improves the quality of cast billets, facilitates production management, and enhances enterprise competitiveness, thus demonstrating a very promising future.