Intelligent control system for manufacturing ladle refractory

Abstract

This invention relates to the field of steel ladle refractory materials technology, and discloses an intelligent control system for the production and manufacturing of steel ladle refractory materials. The system includes a raw material intelligent terminal, a forming control terminal, and a firing optimization terminal. By constructing a digital twin model of the raw materials, the system integrates the real-time physicochemical properties of each batch of raw materials, and optimizes the raw material ratio scheme based on historical data and target performance, ensuring the scientific nature of the raw material ratio formulation. This allows the ratio scheme to adapt to raw material fluctuations and diverse product performance requirements, ensuring the accuracy and stability of product design. Through 3D scanning and machine vision technology, the system perceives the mold status and material distribution in real time, and optimizes process parameters based on a digital simulation model, realizing closed-loop feedback control of the forming process. This reduces the problems of poor billet size consistency and high defect rate caused by traditional reliance on fixed procedures and manual intervention, and improves the billet forming accuracy and quality consistency.

Description

Technical Field

[0001] This invention relates to the field of steel ladle refractory materials technology, specifically to an intelligent control system for the production and manufacturing of steel ladle refractory materials. Background Technology

[0002] A ladle, also known as a steel ladle, steel container, or ladle, is a core piece of equipment in the steelmaking process. It consists of a steel shell, a refractory lining, and an insulation layer. Its main function is to receive molten steel from the converter and perform processes such as desulfurization, deoxidation, and composition adjustment.

[0003] Currently, in the traditional production and manufacturing process of steel ladle refractory materials, the evaluation and management of raw material performance mainly rely on periodic sampling inspections and manual experience. It is impossible to grasp the fluctuations in the physicochemical properties of each batch of raw materials in real time and comprehensively. As a result, the formulation of raw material proportioning schemes lacks data support and is difficult to adapt to the fluctuations in raw materials and the diverse product performance requirements.

[0004] Therefore, an intelligent control system for the production and manufacturing of steel ladle refractory materials is proposed to solve the above problems. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent control system for the production and manufacturing of steel ladle refractory materials, which solves the problems mentioned in the background technology, such as the lack of data support for the formulation of raw material proportioning schemes and the difficulty in adapting to raw material fluctuations and diverse product performance requirements.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent control system for the production and manufacturing of steel ladle refractory materials, the system comprising a raw material intelligent terminal, a forming control terminal, and a firing optimization terminal;

[0007] The raw material intelligent terminal is used to collect the physicochemical performance data of the raw materials entering the warehouse in real time through integrated IoT sensing devices, construct a digital twin model of the raw materials, and dynamically optimize the raw material ratio scheme based on historical production big data and target product performance indicators. At the same time, it regulates the mixing and tamping process through the raw material pretreatment control model and outputs the mixed material.

[0008] The molding control terminal is used to receive the mixture output by the raw material intelligent terminal, and to obtain the mold status and mixture filling status in real time through three-dimensional scanning and machine vision technology. It dynamically optimizes the pressing and molding process parameters in combination with digital simulation model, and performs online detection and adaptive calibration of the blank size and appearance after molding.

[0009] The firing optimization terminal is used to collect firing process data in real time through a sensor network deployed in each temperature zone of the kiln, predict the mapping relationship between the firing curve and the final performance of the product based on a machine learning model, dynamically adjust the firing regime, and perform performance simulation evaluation and quality grade determination on the fired refractory products.

[0010] Preferably, the intelligent raw material terminal includes a raw material management module, a proportioning optimization module, and a pretreatment control module;

[0011] The raw material management module includes a raw material warehousing unit, a performance analysis unit, and a digital modeling unit;

[0012] The raw material warehousing unit is used to bind and register the identity information of each batch of refractory raw materials entering the warehouse through identification technology, and to record the type, origin, batch, warehousing time and quantity of raw materials.

[0013] The performance analysis unit is used to automatically acquire the physicochemical performance data of the current batch of raw materials by connecting a chemical composition analyzer, a particle size distribution analyzer, a moisture analyzer, and a high-temperature performance testing device. The physicochemical performance data includes chemical composition content, particle size distribution, bulk density, apparent porosity, and load softening temperature.

[0014] The digital modeling unit is used to create a corresponding raw material digital twin model for each batch of raw materials based on the real-time data obtained by the performance analysis unit. The raw material digital twin model dynamically integrates the static properties of the raw materials with the dynamic parameters that change in subsequent processes.

[0015] Preferably, the ratio optimization module includes a target setting unit, a model calculation unit, and a formula output unit;

[0016] The target setting unit is used to input and select the performance indicators of the target refractory material product according to the production order requirements. The performance indicators include room temperature compressive strength, high temperature flexural strength, thermal shock stability level and expected service life.

[0017] The model calculation unit is used to call the historical high-quality production data database and combine it with the current available raw material digital twin model set. Through multivariate nonlinear regression analysis, it calculates the optimized raw material ratio scheme that meets the target performance index, and predicts the cost and process feasibility of each scheme.

[0018] The formula output unit is used to convert the optimized raw material ratio scheme recommended by the model calculation unit into specific raw material weighing instructions and send them to the automatic batching system for execution, while synchronizing the formula data to the preprocessing control module.

[0019] Preferably, the pretreatment control module includes a mixing control unit, a material trapping monitoring unit, and a material property evaluation unit;

[0020] The mixing control unit is used to control the feeding sequence, feeding speed, mixing time and mixing intensity of the mixing equipment according to the instructions issued by the formula output unit, and to monitor the uniformity and state of the mixture in real time through an online viscometer and temperature sensor, so as to realize closed-loop feedback control of the mixing process.

[0021] The material trapping monitoring unit is used to monitor the temperature, humidity and volatile content of the material trapping environment in real time through temperature and humidity sensors and gas composition sensors installed in the material trapping silo during the material trapping process, and dynamically adjust the material trapping time.

[0022] The material property evaluation unit is used to perform rapid sampling and testing of the mixture after the pretreatment process. It obtains the final working performance data of the mixture through a plasticity tester and a bulk density meter, and compares it with the expected value of the model. When the deviation exceeds the tolerance, it is fed back to the proportion optimization module for fine adjustment.

[0023] Preferably, the molding control terminal includes a mold management module, a pressing and molding module, and a dimension verification module;

[0024] The mold management module includes a mold status unit and a material placement guidance unit;

[0025] The mold status unit is used to automatically detect the dimensional accuracy, surface wear and cleanliness of the mold cavity before each molding operation using a three-dimensional laser scanner, compare the detection data with the standard mold digital model, determine the usability of the mold based on the comparison results, and generate corresponding maintenance instructions.

[0026] The fabric guiding unit is used to plan the fabric path and weight distribution of the mixture in the mold according to the product model and mold status, guide the automatic fabric spreading equipment to execute, and confirm the uniformity of the fabric spreading through the vision system.

[0027] Preferably, the compression molding module includes a parameter optimization unit and a pressure monitoring unit;

[0028] The parameter optimization unit is used to call the finite element analysis digital simulation model based on the material property evaluation data of the current batch of mixture, the shape and density requirements of the target product, to simulate the material flow and stress distribution during the pressing process, and optimize the initial pressing pressure, holding time, pressing curve and demolding speed parameter set.

[0029] The pressure monitoring unit is used to collect the actual pressure and displacement data of the press at each stage in real time through pressure and displacement sensors during the actual pressing process, and to compare and control the data in real time with the set of optimized parameters set by the parameter optimization unit.

[0030] Preferably, the size verification module includes an online detection unit and an error compensation unit;

[0031] The online inspection unit is used to perform online inspection of the key dimensions and appearance defects of the billet after demolding, through a machine vision inspection system and a contact measurement probe. The key dimensions include length, width, thickness, and hole position deviation, and the appearance defects include cracks, missing edges, and chipped corners.

[0032] The error compensation unit is used to analyze the batch detection data obtained by the online detection unit. When a systematic deviation in dimensional error is found, it automatically generates mold compensation amount and pressing parameter compensation value, and sends them to the mold management module and pressing module for adaptive calibration of subsequent products, thereby achieving continuous optimization of the molding process.

[0033] Preferably, the firing optimization module includes a firing management module, a quality prediction module, and a performance evaluation module;

[0034] The firing management module includes a data acquisition unit and a dynamic adjustment unit;

[0035] The data acquisition unit is used to collect real-time data on the temperature field distribution, pressure gradient, atmosphere concentration, and surface condition of products in the kiln through thermocouples, pressure sensors, flue gas composition analyzers, and video monitoring deployed in various temperature zones of the tunnel kiln and shuttle kiln.

[0036] The dynamic adjustment unit is used to dynamically fine-tune the set temperature, heating rate, holding time and kiln pressure of each temperature zone based on the preset basic firing curve and the real-time data fed back by the data acquisition unit, in order to cope with disturbance factors such as raw material fluctuations and differences in green body state.

[0037] Preferably, the quality prediction module includes a model training unit and a real-time prediction unit;

[0038] The model training unit is used to collect complete firing process data and corresponding batch final performance test reports from historical production. It uses machine learning algorithms to train a firing process and product performance prediction model, and establishes the correlation between firing curve characteristics and key performance indicators such as product strength, thermal shock count, and microstructure.

[0039] The real-time prediction unit is used to input the real-time process data acquired by the data acquisition unit into the firing process and product performance prediction model during the firing process, and to predict the final key performance indicators of the product currently being fired in real time. When the predicted value deviates from the target range, it sends process parameter adjustment suggestions to the dynamic adjustment unit to achieve feedforward and feedback composite control.

[0040] Preferably, the performance evaluation module includes a simulation unit and an automatic rating unit;

[0041] The simulation unit is used to simulate the stress distribution, thermal spalling trend and erosion rate of the finished product under the actual steel ladle usage environment based on the final firing data and the product's digital twin, after the product has been fired and cooled but before destructive physical testing, and to generate a virtual performance evaluation report.

[0042] The automatic grading unit is used to comprehensively analyze the final key performance indicators of the real-time prediction unit and the virtual evaluation report of the simulation unit to determine the quality level of the product, and bind the determination result to the product identity information to guide subsequent sorting, warehousing and delivery. At the same time, the complete data package is fed back to the system database to optimize the raw material ratio and molding process model.

[0043] Compared with the prior art, the present invention provides an intelligent control system for the production and manufacturing of steel ladle refractory materials, which has the following beneficial effects:

[0044] 1. In this invention, by constructing a digital twin model of raw materials to integrate the real-time physicochemical properties of each batch of raw materials, and by optimizing the raw material ratio scheme based on historical data and target performance, the scientific nature of the raw material ratio formulation is ensured, so that the ratio scheme can adapt to raw material fluctuations and diverse product performance requirements, thus ensuring the accuracy and stability of product design.

[0045] 2. In this invention, the mold status and material distribution are perceived in real time through three-dimensional scanning and machine vision technology, and the process parameters are optimized based on the digital simulation model, realizing closed-loop feedback control of the forming process. This reduces the problems of poor blank size consistency and high defect rate caused by traditional reliance on fixed procedures and manual intervention. It can detect blank size and appearance in real time and perform adaptive calibration, improving blank forming accuracy and quality consistency.

[0046] 3. In this invention, the product performance is predicted in real time and the firing process is dynamically adjusted by using sensor networks and machine learning models. Combined with thermo-coupling simulation technology, virtual performance evaluation is carried out, which realizes dynamic optimization of the firing process and forward-looking judgment of product quality. The product quality level is also determined, thereby improving the stability of product performance and realizing intelligent decision-making and quality control in the production process. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the intelligent control system for the production and manufacturing of steel ladle refractory materials according to the present invention. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] For specific implementation examples, please refer to: Figure 1 A smart control system for the production and manufacturing of steel ladle refractory materials, the system includes a raw material intelligent terminal, a molding control terminal and a firing optimization terminal;

[0050] The raw material intelligent terminal is used to collect the physicochemical performance data of raw materials entering the warehouse in real time through integrated IoT sensing devices, build a digital twin model of raw materials, and dynamically optimize the raw material ratio scheme based on historical production big data and target product performance indicators. At the same time, the raw material pretreatment control model is used to regulate the mixing and tamping process and output the mixture.

[0051] The molding control end is used to receive the mixture output from the raw material intelligent end. It uses 3D scanning and machine vision technology to obtain the mold status and mixture filling status in real time. It combines digital simulation model to dynamically optimize the pressing and molding process parameters, and performs online detection and adaptive calibration of the blank size and appearance after molding.

[0052] The firing optimization end is used to collect firing process data in real time through a sensor network deployed in each temperature zone of the kiln, predict the mapping relationship between the firing curve and the final performance of the product based on machine learning models, dynamically adjust the firing regime, and conduct performance simulation evaluation and quality grade determination of the fired refractory products.

[0053] The intelligent raw material terminal includes a raw material management module, a proportioning optimization module, and a pretreatment control module;

[0054] The raw material management module includes a raw material warehousing unit, a performance analysis unit, and a digital modeling unit;

[0055] The raw material warehousing unit is used to bind and register the identity information of each batch of refractory raw materials entering the warehouse through identification technology, and to record the type, origin, batch, warehousing time and quantity of raw materials;

[0056] The performance analysis unit is used to automatically acquire the physicochemical performance data of the current batch of raw materials by connecting to a chemical composition analyzer, particle size distribution analyzer, moisture analyzer and high temperature performance testing equipment. The physicochemical performance data includes chemical composition content, particle size distribution, bulk density, apparent porosity and load softening temperature.

[0057] The digital modeling unit is used to create a corresponding digital twin model of raw materials for each batch of raw materials based on the real-time data obtained by the performance analysis unit. The raw material digital twin model dynamically integrates the static properties of the raw materials with the dynamic parameters that change in subsequent processes.

[0058] The proportioning optimization module includes a target setting unit, a model calculation unit, and a recipe output unit;

[0059] The target setting unit is used to input and select the performance indicators of the target refractory material products according to the production order requirements. The performance indicators include room temperature compressive strength, high temperature flexural strength, thermal shock stability level and expected service life.

[0060] The model calculation unit is used to access a historical high-quality production data database and, combined with a set of digital twin models of currently available raw materials, calculates the optimal raw material ratio scheme that meets the target performance indicators through multivariate nonlinear regression analysis. It also predicts the cost and process feasibility of each scheme, specifically including:

[0061] The raw material ratio data corresponding to each historical batch in the historical high-quality production data database were used as independent variables, and the performance index data of each historical batch of finished products obtained by testing were used as dependent variables.

[0062] Based on the least squares method, the independent and dependent variables are fitted to establish a multiple nonlinear regression mathematical model with raw material ratio as input and predicted performance index as output. The core of this model is to construct a regression equation of the following form:

[0063] ;

[0064] in, Indicates the first Predicted values ​​of each performance indicator express The proportions of the raw materials, To fit the obtained nonlinear function, This is the random error term;

[0065] The target performance index is input into a multivariate nonlinear regression mathematical model, and the inverse solution is used to calculate the optimal raw material ratio scheme that meets the target performance index.

[0066] The formula output unit is used to convert the optimized raw material ratio scheme recommended by the model calculation unit into specific raw material weighing instructions and send them to the automatic batching system for execution. At the same time, the formula data is synchronized to the preprocessing control module.

[0067] The pretreatment control module includes a mixing control unit, a material trapping monitoring unit, and a material property assessment unit;

[0068] The mixing control unit is used to control the feeding sequence, feeding speed, mixing time and mixing intensity of the mixing equipment according to the instructions issued by the formula output unit. It also monitors the uniformity and state of the mixture in real time through online viscometers and temperature sensors to achieve closed-loop feedback control of the mixing process.

[0069] The material trapping monitoring unit is used to monitor the temperature, humidity and volatile content of the material trapping environment in real time through temperature and humidity sensors and gas composition sensors deployed in the material trapping silo during the material trapping process, and dynamically adjust the material trapping time.

[0070] The material property assessment unit is used to perform rapid sampling and testing of the mixture after the pretreatment process. It obtains the final working performance data of the mixture using a plasticity tester and a bulk density meter, and compares it with the model's expected values. When the deviation exceeds the tolerance, it feeds back to the proportioning optimization module for fine-tuning. Specifically, this includes:

[0071] In the material property evaluation unit, a standard range of mixture performance corresponding to the current proportioning scheme and pretreatment process parameters, obtained from historical data statistics, is pre-stored as the model's expected value. The final performance data is denoted as... Its standard range is ,in, This represents the lower limit of the standard range for the performance of the mixture. The upper limit of the standard range for the working performance of the mixture is determined by comparing the final working performance data obtained from the plasticity tester and the bulk density meter with the standard range to determine whether the data falls within the standard range. If the condition is not met, the deviation is determined to exceed the tolerance.

[0072] The molding control module includes a mold management module, a compression molding module, and a dimension verification module.

[0073] The mold management module includes a mold status unit and a material placement guidance unit;

[0074] The mold status unit is used to automatically detect the dimensional accuracy, surface wear and cleanliness of the mold cavity before each molding operation using a 3D laser scanner, compare the detection data with the standard mold digital model, determine the usability of the mold based on the comparison results, and generate corresponding maintenance instructions.

[0075] The material guiding unit is used to plan the material distribution path and weight distribution of the mixture in the mold according to the product model and mold status, and guide the automatic material distribution equipment to execute the process, and confirm the uniformity of the material distribution through the vision system.

[0076] The compression molding module includes a parameter optimization unit and a pressure monitoring unit;

[0077] The parameter optimization unit is used to optimize the initial pressing pressure, holding time, pressing curve, and demolding speed parameter set based on the material property evaluation data of the current batch of mixture and the target product shape and density requirements. It calls the finite element analysis digital simulation model to simulate material flow and stress distribution during the pressing process, specifically including:

[0078] Based on the shape and density requirements of the target product, a three-dimensional finite element mesh model corresponding to the blank and mold is established in the simulation software.

[0079] The material property evaluation data of the current batch of mixture is converted into the constitutive model parameters of the material in the finite element mesh model. The Drucker-Prager model, which is suitable for powder materials, is usually used, and its yield function is:

[0080] ;

[0081] in, As the first stress invariant, For the second deviatoric stress invariant, and These are parameters related to the material's cohesion and friction angle.

[0082] In the simulation software, set the boundary conditions and load the initial parameters of initial pressing pressure, holding time, pressing curve and demolding speed. Run the simulation calculation to obtain the density distribution and stress distribution cloud map of each region of the blank after simulated pressing.

[0083] Analyze the density distribution and stress distribution cloud maps. If there are areas of uneven density and stress concentration, adjust the initial parameters and re-simulate until the simulation results meet the preset uniformity criteria. Then, confirm the parameter set corresponding to this point as the optimized parameter set.

[0084] The pressure monitoring unit is used to collect the actual pressure and displacement data of the press at each stage in real time through pressure and displacement sensors during the actual pressing process, and compares and controls the data with the optimized parameter set set by the parameter optimization unit in real time.

[0085] The dimension verification module includes an online detection unit and an error compensation unit;

[0086] The online inspection unit is used to inspect the key dimensions and appearance defects of the billet after demolding through a machine vision inspection system and a contact measuring probe. The key dimensions include length, width, thickness, and hole position deviation, and the appearance defects include cracks, missing edges, and chipped corners.

[0087] The error compensation unit analyzes the batch inspection data acquired by the online inspection unit. When a systematic deviation in dimensional error is detected, it automatically generates mold compensation amount and pressing parameter compensation value, and sends them to the mold management module and pressing module for subsequent adaptive calibration of products, achieving continuous optimization of the molding process. Specific operation steps include:

[0088] Data collection and preprocessing: Continuously collect key dimension detection data output by the online detection unit after detecting multiple blanks, form the detection dataset for the current batch, and filter outliers in the dataset;

[0089] Systematic bias assessment: Calculate the average deviation value of each key dimension in the detection dataset and compare it with a preset random error threshold. The calculation formula is:

[0090] ;

[0091] in, This represents the number of samples analyzed in the current batch. For the first The measured dimensions of each sample. This is the standard value for this size;

[0092] when The absolute value of the error continuously exceeds the preset random error threshold. If so, the dimensional error is determined to be a systematic deviation;

[0093] Compensation value calculation: Based on the critical dimensions with systematic deviations and their average deviation values, and according to the pre-stored mapping relationships between dimensions and compensation amounts and between deviations and process parameters, the corresponding mold compensation amounts and pressing parameter compensation values ​​are calculated respectively. A linear function is used as an example for the mapping relationships. Mold compensation amount... The calculation formula is:

[0094] ;

[0095] in, The compensation coefficient is determined based on experience with mold and size variations;

[0096] Compensation command issuance: The calculated mold compensation amount is encapsulated as a mold adjustment command and issued to the mold management module to guide mold maintenance and CNC correction; at the same time, the calculated pressing parameter compensation value is encapsulated as a process parameter update command and issued to the pressing module to replace the corresponding initial parameter set.

[0097] The firing optimization module includes a firing management module, a quality prediction module, and a performance evaluation module.

[0098] The firing management module includes a data acquisition unit and a dynamic adjustment unit;

[0099] The data acquisition unit is used to collect real-time data on temperature field distribution, pressure gradient, atmosphere concentration and product surface condition images in the kiln through thermocouples, pressure sensors, flue gas composition analyzers and video monitoring distributed in each temperature zone of the tunnel kiln and shuttle kiln.

[0100] The dynamic adjustment unit is used to dynamically fine-tune the set temperature, heating rate, holding time, and kiln pressure of each temperature zone based on a preset basic firing curve and real-time data fed back by the data acquisition unit, in order to cope with disturbances such as raw material fluctuations and differences in green body condition. Its specific operation steps include:

[0101] Benchmark establishment: Obtain and load the preset basic firing curve corresponding to the current firing product. The basic firing curve defines the initial set temperature, heating rate, holding time and kiln pressure parameters for each temperature zone.

[0102] Real-time monitoring and disturbance identification: Receive real-time feedback from the data acquisition unit on the kiln temperature field distribution, pressure gradient and atmosphere concentration data, compare the real-time data with the expected value of the basic firing curve at the same time, and identify the presence of process disturbance caused by raw material fluctuations and differences in green body state when the deviation between the real-time data and the expected value continues to exceed the preset process tolerance zone.

[0103] Fine-tuning decision generation: For the temperature zone where a disturbance is identified, based on the direction and magnitude of the deviation and according to the pre-stored process parameter adjustment rule library, a fine-tuning instruction is generated for the corresponding parameters in the set temperature, heating rate, holding time and kiln pressure of the temperature zone. The fine-tuning instruction aims to pull the subsequent real-time process parameters back to the process tolerance zone.

[0104] Command execution and closed-loop verification: The generated fine-tuning command is sent to the kiln control system for execution, and the process parameters after execution are monitored through the data acquisition unit to form a closed-loop control until the process parameters in this temperature zone are stable within the process tolerance range.

[0105] The quality prediction module includes a model training unit and a real-time prediction unit;

[0106] The model training unit collects complete firing process data from historical production and the final performance test reports of corresponding batches of finished products. It uses machine learning algorithms to train a firing process and product performance prediction model, establishing the correlation between firing curve characteristics and key performance indicators such as product strength, thermal shock cycles, and microstructure. Specifically, this includes:

[0107] The collected historical firing process data is preprocessed, and the time and temperature curve features and time and atmosphere concentration curve features of each temperature zone are extracted as input feature vectors.

[0108] Use the key performance index data from the final performance test report of the corresponding batch of finished products as the output label;

[0109] The input feature vector and the output label form a sample set, which is then divided into a training set and a test set according to a certain ratio.

[0110] The random forest regression model was trained using a training set, which integrates multiple decision trees. The final prediction is obtained from the prediction results. :

[0111] ;

[0112] in, This represents the predicted output values ​​of the random forest model for the final key performance indicators of the product. For the input feature vector, For the number of decision trees, For the first The predicted output of each tree is calculated, and the model hyperparameters are optimized using a grid search method.

[0113] The training model's prediction accuracy is verified using a test set. When the accuracy reaches a preset threshold, the model training is complete, and a prediction model for the firing process and product performance is obtained.

[0114] The real-time prediction unit is used to input the real-time process data acquired by the data acquisition unit into the firing process and product performance prediction model during the firing process. It predicts the final key performance indicators of the fired product in real time. When the predicted value deviates from the target range, it sends process parameter adjustment suggestions to the dynamic adjustment unit, achieving a combined feedforward and feedback control. Specifically, this includes:

[0115] The real-time prediction unit is pre-set with target ranges for key performance indicators and mapping adjustment rules for process parameters in each temperature zone.

[0116] When the final key performance indicators are predicted in real time Beyond the target range At that time, among them, This represents the lower limit of the target range. As the upper limit of the target range, based on the mapping adjustment rules, the following linear adjustment rule is used to generate the set temperature for a specific temperature range. Recommended adjustment amount :

[0117] ;

[0118] in, The target range is defined by the median and a set point. Similarly, based on the preset adjustment coefficients, suggestions for adjusting the heating rate and holding time are generated.

[0119] Suggestions containing specific adjustment parameters will be sent to the dynamic adjustment unit.

[0120] The performance evaluation module includes a simulation unit and an automatic rating unit;

[0121] The simulation unit is used to simulate the stress distribution, thermal spalling trend and erosion rate of the finished product in the actual ladle use environment after the product has been fired and cooled, and before destructive physical testing. Based on the final firing data and the product's digital twin, it uses thermo-mechanical coupling simulation technology to simulate the finished product's stress distribution, thermal spalling trend and erosion rate, conduct virtual performance evaluation, and generate a virtual evaluation report.

[0122] The automatic grading unit integrates the final key performance indicators of the real-time prediction unit and the virtual evaluation report of the simulation unit to determine the product's quality level. It then links the determination result to the product's identity information to guide subsequent sorting, warehousing, and shipping. Simultaneously, it feeds back the complete data package to the system database to optimize raw material ratios and molding process models. Specifically, this includes:

[0123] The automatic grading unit pre-stores information for different quality levels. Correspondingly, the grading thresholds for each evaluation item in the final key performance indicators and virtual evaluation report;

[0124] The final key performance indicators provided by the real-time prediction unit The results in the virtual evaluation report provided by the simulation unit. Each is compared with the grading threshold to determine... and Which level will I fall into? threshold range Inside;

[0125] Based on the threshold range of all comparison results, the final product quality level is obtained according to the predetermined comprehensive judgment logic.

[0126] The operation steps of the intelligent control system for the production and manufacturing of steel ladle refractory materials are as follows:

[0127] Step 1: Intelligent Raw Material Processing Stage

[0128] The system first digitally manages incoming raw materials through the raw material management module: the raw material warehousing unit uses identification technology to bind identity information to each batch of raw materials; the performance analysis unit connects to various testing equipment to automatically acquire physicochemical performance data; and the digital modeling unit creates a dynamic digital twin model of each batch of raw materials based on this data. Next, the proportioning optimization module begins operation: the target setting unit sets product performance indicators based on orders; the model calculation unit calls up the historical database and combines it with the current raw material digital twin model set to calculate and recommend the optimal raw material proportioning scheme that meets the target through multivariate nonlinear regression analysis; and the formula output unit converts this scheme into weighing instructions and issues them. Finally, the preprocessing control module executes the preparation: the mixing control unit controls the mixing process according to instructions and monitors it in real time; the material trapping monitoring unit monitors the material trapping environment parameters; and the material property evaluation unit performs rapid testing of the mixture and compares the final working performance data with the model's expected values. If the deviation exceeds the limit, feedback is provided for fine-tuning, thereby outputting a mixture with stable quality.

[0129] Step 2: Intelligent Control Stage for Blank Forming

[0130] The system operates in a coordinated manner through the molding control terminal: First, before each molding process, the mold status unit of the mold management module uses a 3D laser scanner to detect the internal state of the mold cavity and compares it with a standard model to generate maintenance instructions; the material placement guidance unit plans and guides automatic material placement. Next, the pressing molding module starts: the parameter optimization unit calls a finite element analysis digital simulation model to simulate the pressing process based on the performance of the mixture and product requirements, optimizing the pressing parameter set; the pressure monitoring unit collects data in real time during actual pressing and performs closed-loop control with the optimized parameter set. After the billet is demolded, the dimension verification module immediately starts working: the online detection unit performs online detection of the key dimensions and appearance of the billet; the error compensation unit analyzes the batch detection data, and when a systematic deviation in dimensions is found, it automatically calculates the mold compensation amount and the pressing parameter compensation value, and sends them to the corresponding modules to achieve adaptive calibration and continuous process optimization.

[0131] Step 3: Intelligent Optimization Stage of Firing Process

[0132] The firing management module at the firing optimization end is responsible for process monitoring and adjustment: the data acquisition unit collects temperature, pressure, and atmosphere concentration data in real time through a sensor network deployed in each temperature zone of the tunnel kiln and shuttle kiln; the dynamic adjustment unit dynamically fine-tunes the process parameters of each temperature zone based on the preset basic firing curve and real-time data to cope with abnormal disturbances such as raw material fluctuations and green body conditions. Simultaneously, the quality prediction module works in parallel: the model training unit uses historical data to train a firing process and product performance prediction model; the real-time prediction unit inputs real-time process data into this model during firing to predict the final key performance indicators of the current product. When the predicted value deviates from the target range, it immediately sends process parameter adjustment suggestions to the dynamic adjustment unit, forming a feedforward and feedback composite control.

[0133] Step Four: Intelligent Product Performance Evaluation Phase

[0134] The performance evaluation module's simulation unit is activated first. Based on the final firing data and the product's digital twin, it uses thermo-coupling simulation technology to simulate the stress distribution, thermal spalling trend, and erosion rate of the finished product under actual ladle usage conditions, performing a virtual performance evaluation and generating a virtual evaluation report. Subsequently, the automatic grading unit integrates the final key performance indicators from the real-time prediction unit and the virtual evaluation report from the simulation unit to automatically determine the product's quality level. The determination result is linked to the product's identity information to guide sorting, warehousing, and shipping. Simultaneously, the complete data package is fed back to the system database for continuous optimization of raw material ratios and molding process models, forming a self-iteratory and continuously improving intelligent production closed loop.

[0135] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0136] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent control system for the production and manufacturing of steel ladle refractory materials, characterized in that: The system includes a raw material intelligent terminal, a molding control terminal, and a firing optimization terminal; The raw material intelligent terminal is used to collect the physicochemical performance data of the raw materials entering the warehouse in real time through integrated IoT sensing devices, construct a digital twin model of the raw materials, and dynamically optimize the raw material ratio scheme based on historical production big data and target product performance indicators. At the same time, it regulates the mixing and tamping process through the raw material pretreatment control model and outputs the mixed material. The molding control terminal is used to receive the mixture output by the raw material intelligent terminal, and to obtain the mold status and mixture filling status in real time through three-dimensional scanning and machine vision technology. It dynamically optimizes the pressing and molding process parameters in combination with digital simulation model, and performs online detection and adaptive calibration of the blank size and appearance after molding. The firing optimization terminal is used to collect firing process data in real time through a sensor network deployed in each temperature zone of the kiln, predict the mapping relationship between the firing curve and the final performance of the product based on a machine learning model, dynamically adjust the firing regime, and perform performance simulation evaluation and quality grade determination on the fired refractory products.

2. The intelligent control system for the production and manufacturing of steel ladle refractory materials according to claim 1, characterized in that: The intelligent raw material terminal includes a raw material management module, a proportioning optimization module, and a pretreatment control module. The raw material management module includes a raw material warehousing unit, a performance analysis unit, and a digital modeling unit; The raw material warehousing unit is used to bind and register the identity information of each batch of refractory raw materials entering the warehouse through identification technology, and to record the type, origin, batch, warehousing time and quantity of raw materials. The performance analysis unit is used to automatically acquire the physicochemical performance data of the current batch of raw materials by connecting a chemical composition analyzer, a particle size distribution analyzer, a moisture analyzer, and a high-temperature performance testing device. The physicochemical performance data includes chemical composition content, particle size distribution, bulk density, apparent porosity, and load softening temperature. The digital modeling unit is used to create a corresponding raw material digital twin model for each batch of raw materials based on the real-time data obtained by the performance analysis unit. The raw material digital twin model dynamically integrates the static properties of the raw materials with the dynamic parameters that change in subsequent processes.

3. The intelligent control system for the production and manufacturing of steel ladle refractory materials according to claim 2, characterized in that: The ratio optimization module includes a target setting unit, a model calculation unit, and a formula output unit; The target setting unit is used to input and select the performance indicators of the target refractory material product according to the production order requirements. The performance indicators include room temperature compressive strength, high temperature flexural strength, thermal shock stability level and expected service life. The model calculation unit is used to call the historical high-quality production data database and combine it with the current available raw material digital twin model set. Through multivariate nonlinear regression analysis, it calculates the optimized raw material ratio scheme that meets the target performance index, and predicts the cost and process feasibility of each scheme. The formula output unit is used to convert the optimized raw material ratio scheme recommended by the model calculation unit into specific raw material weighing instructions and send them to the automatic batching system for execution, while synchronizing the formula data to the preprocessing control module.

4. The intelligent control system for the production and manufacturing of steel ladle refractory materials according to claim 3, characterized in that: The pretreatment control module includes a mixing control unit, a material trapping monitoring unit, and a material property evaluation unit; The mixing control unit is used to control the feeding sequence, feeding speed, mixing time and mixing intensity of the mixing equipment according to the instructions issued by the formula output unit, and to monitor the uniformity and state of the mixture in real time through an online viscometer and temperature sensor, so as to realize closed-loop feedback control of the mixing process. The material trapping monitoring unit is used to monitor the temperature, humidity and volatile content of the material trapping environment in real time through temperature and humidity sensors and gas composition sensors installed in the material trapping silo during the material trapping process, and dynamically adjust the material trapping time. The material property evaluation unit is used to perform rapid sampling and testing of the mixture after the pretreatment process. It obtains the final working performance data of the mixture through a plasticity tester and a bulk density meter, and compares it with the expected value of the model. When the deviation exceeds the tolerance, it is fed back to the proportion optimization module for fine adjustment.

5. The intelligent control system for the production and manufacturing of steel ladle refractory materials according to claim 1, characterized in that: The molding control terminal includes a mold management module, a pressing and molding module, and a dimension verification module; The mold management module includes a mold status unit and a material placement guidance unit; The mold status unit is used to automatically detect the dimensional accuracy, surface wear and cleanliness of the mold cavity before each molding operation using a three-dimensional laser scanner, compare the detection data with the standard mold digital model, determine the usability of the mold based on the comparison results, and generate corresponding maintenance instructions. The fabric guiding unit is used to plan the fabric path and weight distribution of the mixture in the mold according to the product model and mold status, guide the automatic fabric spreading equipment to execute, and confirm the uniformity of the fabric spreading through the vision system.

6. The intelligent control system for the production and manufacturing of steel ladle refractory materials according to claim 5, characterized in that: The compression molding module includes a parameter optimization unit and a pressure monitoring unit; The parameter optimization unit is used to call the finite element analysis digital simulation model based on the material property evaluation data of the current batch of mixture, the shape and density requirements of the target product, to simulate the material flow and stress distribution during the pressing process, and optimize the initial pressing pressure, holding time, pressing curve and demolding speed parameter set. The pressure monitoring unit is used to collect the actual pressure and displacement data of the press at each stage in real time through pressure and displacement sensors during the actual pressing process, and to compare and control the data in real time with the set of optimized parameters set by the parameter optimization unit.

7. The intelligent control system for the production and manufacturing of steel ladle refractory materials according to claim 5, characterized in that: The size verification module includes an online detection unit and an error compensation unit; The online inspection unit is used to perform online inspection of the key dimensions and appearance defects of the billet after demolding, through a machine vision inspection system and a contact measurement probe. The key dimensions include length, width, thickness, and hole position deviation, and the appearance defects include cracks, missing edges, and chipped corners. The error compensation unit is used to analyze the batch detection data obtained by the online detection unit. When a systematic deviation in dimensional error is found, it automatically generates mold compensation amount and pressing parameter compensation value, and sends them to the mold management module and pressing module for adaptive calibration of subsequent products, thereby achieving continuous optimization of the molding process.

8. The intelligent control system for the production and manufacturing of steel ladle refractory materials according to claim 1, characterized in that: The firing optimization module includes a firing management module, a quality prediction module, and a performance evaluation module; The firing management module includes a data acquisition unit and a dynamic adjustment unit; The data acquisition unit is used to collect real-time data on the temperature field distribution, pressure gradient, atmosphere concentration, and surface condition of products in the kiln through thermocouples, pressure sensors, flue gas composition analyzers, and video monitoring deployed in various temperature zones of the tunnel kiln and shuttle kiln. The dynamic adjustment unit is used to dynamically fine-tune the set temperature, heating rate, holding time and kiln pressure of each temperature zone based on the preset basic firing curve and the real-time data fed back by the data acquisition unit, in order to cope with disturbance factors such as raw material fluctuations and differences in green body state.

9. The intelligent control system for the production and manufacturing of steel ladle refractory materials according to claim 8, characterized in that: The quality prediction module includes a model training unit and a real-time prediction unit; The model training unit is used to collect complete firing process data and corresponding batch final performance test reports from historical production. It uses machine learning algorithms to train a firing process and product performance prediction model, and establishes the correlation between firing curve characteristics and key performance indicators such as product strength, thermal shock count, and microstructure. The real-time prediction unit is used to input the real-time process data acquired by the data acquisition unit into the firing process and product performance prediction model during the firing process, and to predict the final key performance indicators of the product currently being fired in real time. When the predicted value deviates from the target range, it sends process parameter adjustment suggestions to the dynamic adjustment unit to achieve feedforward and feedback composite control.

10. The intelligent control system for the production and manufacturing of steel ladle refractory materials according to claim 9, characterized in that: The performance evaluation module includes a simulation unit and an automatic rating unit; The simulation unit is used to simulate the stress distribution, thermal spalling trend and erosion rate of the finished product under the actual steel ladle usage environment based on the final firing data and the product's digital twin, after the product has been fired and cooled but before destructive physical testing, and to generate a virtual performance evaluation report. The automatic grading unit is used to comprehensively analyze the final key performance indicators of the real-time prediction unit and the virtual evaluation report of the simulation unit to determine the quality level of the product, and bind the determination result to the product identity information to guide subsequent sorting, warehousing and delivery. At the same time, the complete data package is fed back to the system database to optimize the raw material ratio and molding process model.

Images