A method and system for controlling electric furnace carbon injection and oxygen blowing process based on target carbon content

Abstract

The present invention belongs to the technical field of metallurgical industry, and discloses a method and system for controlling an electric furnace carbon injection and oxygen blowing process based on a target carbon content. A real-time prediction model for the molten pool temperature is constructed, and the molten pool temperature is predicted by real-time reading of the amount of lime, power consumption, flue gas composition and temperature, oxygen consumption, carbon powder amount, continuous addition of scrap steel, natural gas amount, and furnace weight. The carbon content of the molten pool is calculated according to the volume fraction of carbon monoxide and carbon dioxide in the flue gas components, the total volume of flue gas, and the total mass of molten steel, and the oxygen content of the molten pool is calculated in combination with the activity coefficients of carbon and oxygen and the current molten pool temperature. According to the difference between the target endpoint carbon content and the current molten pool carbon content, the additional amount of carbon powder or oxygen added is determined to control the carbon and oxygen addition process. The system includes data acquisition, temperature prediction, carbon and oxygen calculation, feeding control, alarm and visualization modules. The present invention can accurately control the carbon and oxygen content in the electric furnace steelmaking process, and improve production efficiency and product quality.

Description

Technical Field

[0001] The invention relates to the technical field of metallurgical industry, and in particular to a method and system for controlling an electric furnace carbon injection and oxygen blowing process based on a target carbon content. Background Art

[0002] Electric furnace smelting is a key process in steel production, in which the precise control of the molten pool temperature, carbon content and oxygen content has a decisive influence on the final product quality and production efficiency. However, traditional control methods mainly rely on the operator's experience and judgment. The operator needs to estimate the state of the molten pool based on observation and limited measurement data, and decide the timing and amount of carbon injection and oxygen blowing. This method is easily affected by human factors and it is difficult to ensure the consistency and accuracy of control. Due to the limitation of high temperature environment, direct measurement of molten pool temperature and composition usually has a large lag. For example, it may take several minutes to determine the composition of the molten pool through sampling analysis, which makes real-time control difficult. Complex physical and chemical reactions occur inside the electric furnace, including redox reactions, gas-liquid reactions, etc. The dynamic changes of these reactions make it extremely difficult to accurately predict the state of the molten pool. Due to insufficient control accuracy, excessive addition of carbon powder or oxygen is often required, resulting in energy waste and increased production costs.

[0003] In recent years, with the development of artificial intelligence and big data technology, some researchers have tried to apply machine learning methods to smelting process control. Although machine learning methods have made some progress, there are still some shortcomings: the existing models are often difficult to accurately capture the complex dynamic process inside the electric furnace, resulting in insufficient prediction accuracy; many methods cannot meet the needs of rapid changes in electric furnace smelting, so the real-time performance is not enough; the model is difficult to adapt to different production processes and equipment state changes, so the adaptability is poor; some "black box" models are difficult to provide intuitive decision-making basis for operators, resulting in insufficient interpretability; there is currently a lack of comprehensive solutions that organically combine multiple links such as temperature prediction, composition control, and feeding optimization. Therefore, the development of an intelligent system that can accurately predict the molten pool temperature in real time and achieve precise control of carbon and oxygen content based on this has important theoretical significance and practical application value for improving the automation level of electric furnace smelting process, improving product quality, and reducing energy consumption. The present invention is aimed at the shortcomings of the prior art and proposes a method and system for controlling the carbon injection and oxygen blowing process of an electric furnace based on target carbon content, aiming to solve the above problems. Summary of the invention

[0004] The purpose of the present invention is to provide an electric furnace carbon injection and oxygen blowing process control method and system based on target carbon content, which realizes precise control of the electric furnace smelting process by predicting the molten pool temperature in real time, calculating the carbon content and oxygen content of the molten pool, and automatically adjusting the amount of carbon powder and oxygen added according to the target carbon content.

[0005] The technical solution of the present invention is as follows: A method for controlling an electric furnace carbon injection and oxygen blowing process based on a target carbon content comprises the following steps:

[0006] Step 1, construct a real-time prediction model for the molten pool temperature to predict the molten pool temperature T; read the amount of lime, power consumption, flue gas composition, flue gas temperature, oxygen consumption, carbon powder amount, continuous addition of scrap steel, natural gas amount, and furnace body weight at intervals of t seconds; use a thermocouple to measure the molten pool temperature at intervals of t seconds; use the real-time amount of lime, power consumption, flue gas composition, flue gas temperature, oxygen consumption, carbon powder amount, continuous addition of scrap steel, natural gas amount, and furnace body weight as inputs, and the molten pool temperature measured by the thermocouple as the target, and establish a real-time prediction model for the molten pool temperature based on a time series neural network;

[0007] Step 2, calculating the carbon content of the molten pool using the flue gas composition;

[0008] Step 3, calculating the oxygen content of the molten pool according to the carbon-oxygen product;

[0009] Step 4, adjusting the carbon injection and oxygen blowing processes according to the molten pool carbon content, the molten pool oxygen content and the target endpoint carbon content;

[0010] Step 5: When the carbon content or oxygen content of the molten pool exceeds a preset range, an alarm is issued;

[0011] Step 6: Regularly update the real-time prediction model of the molten pool temperature.

[0012] The step 2 is specifically as follows:

[0013] Obtain the volume fraction of carbon monoxide in the flue gas components and the volume fraction of carbon dioxide Total volume of flue gas 、Total mass of molten steel Then, according to the formula w ( [ C ] ) = ( y C O ⋅ V g a s + 2 ⋅ y C O 2 ⋅ V g a s 2 2 . 4 ⋅ Q s t e e l ) × 1 0 0 % Calculate the carbon content of the molten pool w ( [ C ] ) ;

[0014] The total volume of flue gas is obtained according to the flue gas analyzer; the total mass of molten steel = density of molten steel × geometric volume of furnace at height; the height is the height of slag interface, which is the height of molten steel from furnace bottom and is obtained through industrial camera installed at furnace door.

[0015] The step 3 is specifically as follows:

[0016] The activity coefficient of carbon is obtained according to the activity coefficient calculation method. and oxygen activity coefficient ;

[0017] According to the activity coefficient of carbon , oxygen activity coefficient , current molten pool temperature T, molten pool carbon content w ( [ C ] ) , according to the formula w ( [ O ] ) = 1 . 0 1 3 2 5 × 1 0 5 ⋅ f C f O 1 0 ( 1 1 6 8 T + 2 . 0 7 ) ⋅ w ( [ C ] ) Calculate the oxygen content of the molten pool w ( [ O ] ) = 1 . 0 1 3 2 5 × 1 0 5 ⋅ f C f O 1 0 ( 1 1 6 8 T + 2 . 0 7 ) ⋅ w ( [ C ] ) .

[0018] The activity coefficient calculation method is specifically as follows: according to the current molten pool temperature, by searching the interaction coefficient table corresponding to the current molten pool temperature, the first-order interaction coefficient of element j relative to element i is obtained: and the second-order interaction coefficient , according to the mass fraction of element j [ % j ] , and the formula l g f i = ∑ j = C , S i , M n . . . . . . ( e i j [ % j ] + γ i j [ % j ] 2 ) Calculate the activity coefficient of element i.

[0019] The step 4 is specifically as follows: when the smelting process of the electric furnace foam slag is stable, the amount of carbon powder and oxygen added to the molten pool is kept constant, and the target end point carbon content of the current smelting is obtained at this time w t a r g e t ( [ C ] ) ; When the target end point carbon content w t a r g e t ( [ C ] ) Greater than the current molten pool carbon content w ( [ C ] ) When the additional amount of carbon powder is added, the formula C a d d i t i o n a l = w ( [ O ] ) × 1 2 1 6 + Q s t e e l × ( w t a r g e t ( [ C ] ) − w ( [ C ] ) ) Obtained, no additional oxygen is added; when the target end point carbon content w t a r g e t ( [ C ] ) Less than the current molten pool carbon content w ( [ C ] ) When the additional amount of oxygen is added, the formula O a d d i t i o n a l = ( 4 3 × ( w t a r g e t ( [ C ] ) − w ( [ C ] ) ) × Q s t e e l ) − w ( [ O ] ) get.

[0020] The real-time prediction model for the molten pool temperature adopts a long short-term memory network LSTM or a gated recurrent unit network GRU.

[0021] The training process of the real-time prediction model of molten pool temperature includes:

[0022] Collect historical production data, including lime quantity, electricity consumption, flue gas composition and temperature, oxygen consumption, carbon powder quantity, continuous scrap steel addition, natural gas quantity, furnace weight and corresponding bath temperature;

[0023] Preprocess the collected data, including data cleaning, normalization, and time series alignment;

[0024] Divide the processed data into training set and validation set;

[0025] Use the training set to train the neural network model and use the validation set to evaluate the model performance;

[0026] The model is optimized according to the evaluation results until the predetermined accuracy requirements are met.

[0027] It also includes an alarm mechanism and a data visualization step; when the calculated carbon content or oxygen content of the molten pool exceeds the preset range, an alarm is issued to the operator; the real-time calculated molten pool temperature, carbon content, oxygen content, and the amount of carbon powder and oxygen added are displayed in the form of a chart on the control interface.

[0028] A control system for an electric furnace carbon injection and oxygen blowing process based on a target carbon content, comprising:

[0029] Data acquisition module, used to collect real-time data on lime quantity, power consumption, flue gas composition, flue gas temperature, oxygen consumption, carbon powder quantity, continuous addition of scrap steel, natural gas quantity, and furnace weight;

[0030] A molten pool temperature prediction module is used to predict the molten pool temperature;

[0031] Carbon and oxygen content calculation module, used to calculate the carbon content and oxygen content of the molten pool;

[0032] A feeding control module is used to calculate and control the additional amount of carbon powder and oxygen added;

[0033] Alarm module, used to implement the alarm mechanism;

[0034] Visualization module, used to realize data visualization function.

[0035] A data storage module is also included to store historical production data and model parameters to support regular updating and optimization of the model.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] 1. Precise control of melt pool temperature and composition

[0038] Real-time prediction and control: This invention uses a time series-based neural network (such as LSTM or GRU) to predict the molten pool temperature in real time, which can accurately reflect the temperature changes in the actual production process. Compared with traditional methods, this invention avoids temperature control problems caused by lag and inaccuracy, and improves the prediction accuracy and response speed.

[0039] Accurately calculate composition: By calculating the carbon content and oxygen content of the molten pool through an accurate model, the composition of the molten pool can be better controlled. Traditional methods often rely on experience and have large uncertainties, while the present invention makes composition control more stable and reliable through real-time calculation and adjustment.

[0040] 2. Improve production efficiency

[0041] Automatic adjustment: The present invention automatically adjusts the amount of carbon powder and oxygen added through real-time data collection and model calculation, reducing manual intervention. In traditional processes, operators make adjustments based on experience, which is inefficient and prone to errors. The present invention significantly improves production efficiency and makes the production process more continuous and efficient.

[0042] Reduce resource waste: By accurately controlling the amount of carbon powder and oxygen added in real time, resource waste is avoided. In traditional methods, due to inaccurate control, over- or under-dosage often occurs, resulting in resource waste and increased production costs. Through the precise control of the present invention, production costs can be effectively reduced.

[0043] 3. Improve product quality

[0044] Stable composition control: The present invention makes the composition of molten steel more stable by precisely controlling the carbon and oxygen content of the molten pool, significantly improving the quality and consistency of the product. In traditional processes, due to inaccurate composition control, product quality fluctuates greatly and unqualified products are prone to occur. The present invention improves the yield rate and product consistency.

[0045] Reduce defects: As the composition control is more precise and stable, product defects caused by unstable carbon and oxygen content are reduced, and the product qualification rate and customer satisfaction are improved.

[0046] 4. Strong adaptability

[0047] Model updateability: The prediction model of the present invention can be regularly updated according to new production data, adapting to changes in production processes and equipment performance, and maintaining high accuracy of prediction and control. In traditional methods, the model and control parameters are fixed and difficult to adapt to changing production conditions. The present invention has strong adaptability and flexibility.

[0048] Applicable to various furnace types: The method and system of the present invention are applicable to various types of electric furnaces and have wide adaptability. Traditional methods are often optimized for specific furnace types and have poor adaptability. The present invention has a wider application prospect.

[0049] 5. Improved security

[0050] Real-time monitoring and early warning: By real-time monitoring of the temperature and composition of the molten pool, when the temperature or composition exceeds the preset range, the system can promptly alarm to avoid production accidents and improve production safety. In traditional methods, the early warning mechanism is imperfect and potential safety hazards are prone to occur. This invention significantly improves production safety.

[0051] Data visualization: Through the visualization of data, operators can intuitively understand the changes in various parameters in the production process and better manage and control production. In traditional methods, data display is not intuitive, and it is difficult for operators to fully understand the production situation. This invention improves the convenience of operation and management level.

[0052] 6. Reduce environmental impact

[0053] Optimize resource use: The present invention reduces unnecessary resource consumption and environmental pollution by precisely controlling the amount of carbon powder and oxygen added. In traditional methods, there is a lot of waste of resources and a heavy burden on the environment. The present invention optimizes resource use and has significant environmental benefits.

[0054] Reduce energy consumption: By improving production efficiency and reducing resource waste, energy consumption is reduced, which has significant environmental benefits. In traditional methods, energy consumption is high and the impact on the environment is large. The present invention reduces production energy consumption through optimization control.

[0055] 7. Improve operational convenience

[0056] Automated system: The system of the present invention can automatically collect data, predict temperature, calculate ingredients and adjust the amount of addition. The operator only needs to monitor and manage, which makes the operation easier. In the traditional method, the operator needs to manually adjust and control, which is complicated and inefficient. The present invention significantly improves the convenience of operation.

[0057] Friendly user interface: The present invention provides an intuitive and easy-to-use user interface, allowing operators to easily view and manage various parameters and data in the production process. In traditional methods, the interface is complex and unfriendly, making it difficult for operators to quickly master it. The present invention improves user experience and operational efficiency.

[0058] 8. Data-driven decision support

[0059] Historical data analysis: By analyzing historical production data, potential problems and optimization opportunities in production can be discovered, providing data support for production decisions. In traditional methods, data utilization is low and decisions rely on experience. This invention provides a scientific basis for production optimization through data analysis.

[0060] Continuous optimization: Based on the feedback of data and models, the present invention continuously optimizes the production process and parameter settings, and continuously improves production performance and product quality. In traditional methods, the optimization process is slow and unsystematic. The present invention achieves continuous optimization of the production process.

[0061] In summary, the present invention significantly improves the accuracy and efficiency of the electric furnace carbon injection and oxygen blowing process through advanced prediction and control technology, reduces production costs and environmental impact, improves product quality and production safety, and has significant economic and social benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 The present invention is a flow chart of the electric furnace carbon injection and oxygen blowing process control method based on the target carbon content.

[0063] Figure 2 It is a structural block diagram of the electric furnace carbon injection and oxygen blowing process control system based on the target carbon content of the present invention.

[0064] Figure 3 It is a schematic diagram of real-time data collection of an electric furnace in the present invention.

[0065] Numbers in the figure: 1. Lime weighing sensor; 2. Electric energy meter; 3. Flue gas analyzer; 4. Oxygen flow sensor; 5. Carbon powder flow sensor; 6. Scrap steel weighing sensor; 7. Natural gas flow meter; 8. Furnace weighing sensor; 9. Thermocouple. DETAILED DESCRIPTION

[0066] In the present invention, the relationship between the real-time lime amount, electricity consumption, flue gas composition and temperature, oxygen consumption, carbon powder amount, continuous addition of scrap steel, natural gas amount, furnace body weight and the molten pool temperature is taken into consideration. The above input quantities are all key factors affecting the molten pool temperature and none of them can be missing. Therefore, the above input quantities are taken into consideration when establishing the real-time prediction model for the molten pool temperature.

[0067] The present invention starts from a perspective of considering the temperature of the molten pool, thereby adjusting the carbon injection and oxygen blowing processes by calculating the carbon content and oxygen content of the molten pool.

[0068] The preferred implementation is as follows:

[0069] 1. Real-time prediction model of molten pool temperature: using long short-term memory network (LSTM) or gated recurrent unit (GRU) network structure.

[0070] 2. Model training process: Collect historical production data, including lime quantity, electricity consumption, flue gas composition and temperature, oxygen consumption, carbon powder quantity, continuous scrap steel addition, natural gas quantity, furnace weight and corresponding molten pool temperature; pre-process the collected data, including data cleaning, normalization and time series alignment; divide the processed data into training set and validation set; use the training set to train the neural network model, and use the validation set to evaluate the model performance; optimize the model according to the evaluation results until the predetermined accuracy requirements are met.

[0071] 3. Regular model update: This includes regular updates of the real-time prediction model of the molten pool temperature to adapt to changes in production processes and equipment performance.

[0072] 4. Alarm mechanism: When the calculated carbon content or oxygen content of the molten pool exceeds the preset range, an alarm is issued to the operator.

[0073] 5. Data visualization: The real-time calculated molten pool temperature, carbon content, oxygen content, and the amount of carbon powder and oxygen added are displayed in graphical form on the control interface.

[0074] The present invention also provides an electric furnace carbon injection and oxygen blowing process control system based on target carbon content, comprising:

[0075] 1. Data acquisition module: used to collect real-time data on lime quantity, power consumption, flue gas composition and temperature, oxygen consumption, carbon powder quantity, continuous addition of scrap steel, natural gas quantity, and furnace weight;

[0076] 2. Molten pool temperature prediction module: used to predict the molten pool temperature according to the above method;

[0077] 3. Carbon and oxygen content calculation module: used to calculate the carbon content and oxygen content of the molten pool according to the above method;

[0078] 4. Feeding control module: used to calculate and control the additional amount of carbon powder and oxygen added according to the above method;

[0079] 5. Alarm module: used to implement the above alarm mechanism;

[0080] 6. Visualization module: used to realize the above data visualization function;

[0081] 7. Data storage module: used to store historical production data and model parameters to support regular updating and optimization of the model.

[0082] Through the above technical scheme, the present invention can predict the molten pool temperature in real time, accurately calculate the carbon and oxygen contents of the molten pool, and control the amount of carbon powder and oxygen added according to the target carbon content, thereby realizing precise control of the electric furnace carbon injection and oxygen blowing process and improving production efficiency and product quality.

[0083] Example 1

[0084] As a basic embodiment of the present invention, the present invention includes a method for controlling a carbon injection and oxygen blowing process in an electric furnace based on a target carbon content, referring to Figure 1 , including the following steps:

[0085] 1. Construct a real-time prediction model for the molten pool temperature: read the lime quantity, power consumption, flue gas composition and temperature, oxygen consumption, carbon powder quantity, continuous scrap steel addition, natural gas quantity, and furnace weight data at 10-second intervals, and use thermocouples to measure the molten pool temperature at 10-second intervals. Using the real-time lime quantity, power consumption, flue gas composition and temperature, oxygen consumption, carbon powder quantity, continuous scrap steel addition, natural gas quantity, and furnace weight as inputs, and the molten pool temperature measured by thermocouples as the target, a real-time prediction model for the molten pool temperature is established based on a time series neural network.

[0086] The model training process is as follows:

[0087] Data collection and preprocessing: Collect historical production data, including lime quantity, electricity consumption, flue gas composition and temperature, oxygen consumption, carbon powder quantity, continuous scrap steel addition, natural gas quantity, furnace weight and corresponding molten pool temperature. Preprocess the collected data, including data cleaning, normalization and time series alignment.

[0088] Data partitioning: The processed data is divided into a training set and a validation set. The training set is used to train the neural network model, and the validation set is used to evaluate the model performance.

[0089] Model training and evaluation: Use the training set to train the neural network model, and use the validation set to evaluate the model performance. Optimize the model based on the evaluation results until the predetermined accuracy requirements are met.

[0090] Model optimization and update: Regularly collect new historical production data, update and optimize the model to adapt to changes in production processes and equipment performance.

[0091] 2. Input real-time data and predict the molten pool temperature: Input the current amount of lime, power consumption, flue gas composition and temperature, oxygen consumption, carbon powder amount, continuous addition of scrap steel, natural gas amount, and furnace weight information into the real-time prediction model of the molten pool temperature, and output the current molten pool temperature T.

[0092] 3. Calculate the carbon content of the molten pool: obtain the volume fraction of carbon monoxide in the flue gas composition and carbon dioxide volume fraction Total volume of flue gas 、Total mass of molten steel , according to the formula w ( [ C ] ) = ( y C O ⋅ V g a s + 2 ⋅ y C O 2 ⋅ V g a s 2 2 . 4 ⋅ Q s t e e l ) × 1 0 0 % Calculation of melt pool carbon content w ( [ C ] ) .

[0093] 4. Calculate the activity coefficients of carbon and oxygen: According to the current molten pool temperature, by looking up the interaction coefficient table, obtain the first-order interaction coefficient of element j relative to i and the second-order interaction coefficient , and then according to the formula The activity coefficients of carbon and oxygen were calculated.

[0094] 5. Calculate the oxygen content of the molten pool: Based on the activity coefficient of carbon and oxygen activity coefficient , and the formula w ( [ O ] ) = 1 . 0 1 3 2 5 × 1 0 5 ⋅ f C f O 1 0 ( 1 1 6 8 T + 2 . 0 7 ) ⋅ w ( [ C ] ) Calculation of melt pool oxygen content w ( [ O ] ) .

[0095] 6. Determine the target endpoint carbon content: After the electric furnace foam slag smelting process is stable, the amount of carbon powder and oxygen added to the molten pool remains constant, and the target endpoint carbon content of the current smelting is obtained. w t a r g e t ( [ C ] ) .

[0096] 7. Adjust the amount of carbon powder and oxygen added: When the target end point carbon content w t a r g e t ( [ C ] ) Greater than the current molten pool carbon content w ( [ C ] ) When the additional amount of carbon powder is added, the formula C a d d i t i o n a l = w ( [ O ] ) × 1 2 1 6 + Q s t e e l × ( w t a r g e t ( [ C ] ) − w ( [ C ] ) ) Calculation, oxygen does not need to be added additionally. When the target end point carbon content w t a r g e t ( [ C ] ) Less than the current molten pool carbon content w ( [ C ] ) When the additional amount of oxygen is added, the formula O a d d i t i o n a l = ( 4 3 × ( w t a r g e t ( [ C ] ) − w ( [ C ] ) ) × Q s t e e l ) − w ( [ O ] ) calculate.

[0097] 8. Alarm mechanism and data visualization: During the calculation process, when the carbon content or oxygen content of the molten pool exceeds the preset range, the system automatically issues an alarm to the operator. At the same time, the real-time calculated molten pool temperature, carbon content, oxygen content, and the amount of carbon powder and oxygen added are displayed in a graphical form on the control interface, which is convenient for the operator to monitor and adjust in real time.

[0098] 9. Regularly update and optimize the model: The real-time prediction model of the molten pool temperature is regularly updated to adapt to changes in production processes and equipment performance. This includes collecting new historical production data, preprocessing the data, retraining the model, and evaluating the model performance through the validation set. The model is optimized based on the evaluation results until the predetermined accuracy requirements are met.

[0099] Example 2

[0100] As the best embodiment of the present invention, the present invention provides an electric furnace carbon injection oxygen blowing process control system based on the target carbon content, referring to Figure 2 , including the following modules:

[0101] 1. Data acquisition module: used to collect real-time data on lime quantity, power consumption, flue gas composition and temperature, oxygen consumption, carbon powder quantity, continuous addition of scrap steel, natural gas quantity, furnace weight, and molten pool temperature. Figure 3 The lime weighing sensor 1, electric energy meter 2, flue gas analyzer 3, oxygen flow sensor 4, carbon powder flow sensor 5, scrap steel weighing sensor 6, natural gas flow meter 7, furnace weighing sensor 8, and thermocouple 9 are connected. The data acquisition module is connected to various sensors and measuring instruments of the electric furnace. It can read various data at intervals of 10 seconds and transmit them to the central processing unit of the system through the network.

[0102] 2. Molten pool temperature prediction module: used to predict the molten pool temperature based on real-time collected data. This module uses a long short-term memory network (LSTM) or a gated recurrent unit (GRU) network structure. First, a real-time prediction model for the molten pool temperature is constructed. The model inputs are the real-time amount of lime, power consumption, flue gas composition and temperature, oxygen consumption, carbon powder, continuous addition of scrap steel, natural gas, and furnace weight. The output is the molten pool temperature. The model is trained using historical production data, and the model is optimized and updated regularly.

[0103] 3. Carbon and oxygen content calculation module: used to calculate the carbon content and oxygen content in the molten pool. By obtaining the volume fraction of carbon monoxide and carbon dioxide in the flue gas components, the total volume of flue gas, and the total mass of molten steel, the formula

[0104] w ( [ C ] ) = ( y C O ⋅ V g a s + 2 ⋅ y C O 2 ⋅ V g a s 2 2 . 4 ⋅ Q s t e e l ) × 1 0 0 % Calculation of melt pool carbon content w ( [ C ] ) Then according to the obtained carbon activity coefficient and oxygen activity coefficient , through the formula w ( [ O ] ) = 1 . 0 1 3 2 5 × 1 0 5 ⋅ f C f O 1 0 ( 1 1 6 8 T + 2 . 0 7 ) ⋅ w ( [ C ] ) Calculation of melt pool oxygen content w ( [ O ] ) .

[0105] 4. Feeding control module: according to the calculation results of carbon content and oxygen content in the molten pool and the target carbon content w t a r g e t ( [ C ] ) , calculate the additional amount of carbon powder and oxygen added. When the target carbon content is greater than the current carbon content, calculate the additional amount of carbon powder added; when the target carbon content is less than the current carbon content, calculate the additional amount of oxygen added.

[0106] 5. Alarm module: When the calculated carbon content or oxygen content of the molten pool exceeds the preset range, an alarm is issued to the operator.

[0107] 6. Visualization module: The real-time calculated molten pool temperature, carbon content, oxygen content, and the amount of carbon powder and oxygen added are displayed in the form of graphs on the control interface, which is convenient for operators to monitor and adjust in real time.

[0108] 7. Data storage module: used to store historical production data and model parameters to support regular updating and optimization of the model.

[0109] Through the above-mentioned implementation mode, the present invention can realize the precise control of the electric furnace carbon injection and oxygen blowing process based on the target carbon content, improve the efficiency and quality of the smelting process, and reduce the production cost.

Claims

1. A method for controlling an electric furnace carbon injection and oxygen blowing process based on a target carbon content, characterized in that: The following steps are involved: Step 1, construct a real-time prediction model for the molten pool temperature to predict the molten pool temperature T; read the amount of lime, power consumption, flue gas composition, flue gas temperature, oxygen consumption, carbon powder amount, continuous addition of scrap steel, natural gas amount, and furnace body weight at intervals of t seconds; use a thermocouple to measure the molten pool temperature at intervals of t seconds; use the real-time amount of lime, power consumption, flue gas composition, flue gas temperature, oxygen consumption, carbon powder amount, continuous addition of scrap steel, natural gas amount, and furnace body weight as inputs, and the molten pool temperature measured by the thermocouple as the target, and establish a real-time prediction model for the molten pool temperature based on a time series neural network; Step 2, calculating the carbon content of the molten pool using the flue gas composition; Get the volume fraction y of carbon monoxide in the flue gas components CO and the volume fraction of carbon dioxide y CO2 、Total volume of flue gas V gas 、Total mass of molten steel Q steel Then, according to the formula Calculate and obtain the carbon content w([C]) of the molten pool; The total volume of flue gas is obtained according to the flue gas analyzer; the total mass of molten steel = density of molten steel × geometric volume of furnace at height; the height is the height of the slag interface, which is the height of molten steel from the furnace bottom and is obtained through an industrial camera installed at the furnace door; Step 3: Calculate the oxygen content of the molten pool according to the carbon-oxygen product; obtain the carbon activity coefficient f according to the activity coefficient calculation method. C and the activity coefficient of oxygen f O ; According to the carbon activity coefficient f C , oxygen activity coefficient f O , the current molten pool temperature T, the molten pool carbon content w([C]), according to the formula Calculate the oxygen content of the molten pool Step 4, adjusting the carbon injection and oxygen blowing processes according to the molten pool carbon content, the molten pool oxygen content and the target endpoint carbon content; Step 5: When the carbon content or oxygen content of the molten pool exceeds a preset range, an alarm is issued; Step 6: Regularly update the real-time prediction model of the molten pool temperature.

2. The electric furnace carbon injection and oxygen blowing process control method based on target carbon content according to claim 1 is characterized in that: The activity coefficient calculation method is specifically as follows: according to the current molten pool temperature, by looking up the interaction coefficient table corresponding to the current molten pool temperature, the first-order interaction coefficient of element j relative to i is obtained: and the second-order interaction coefficient According to the mass fraction of element j [% j], and the formula Calculate the activity coefficient of element i.

3. The electric furnace carbon injection and oxygen blowing process control method based on target carbon content according to claim 1, characterized in that: The step 4 is specifically as follows: when the smelting process of the electric furnace foam slag is stable, the amount of carbon powder and oxygen added to the molten pool is kept constant, and the target end point carbon content w of the current smelting is obtained at this time. target ([C]); When the target endpoint carbon content w target When ([C]) is greater than the current molten pool carbon content w([C]), the additional amount of carbon powder added is calculated according to the formula Obtained, no additional oxygen is added; when the target end point carbon content w target When ([C]) is less than the current carbon content of the molten pool w([C]), the additional amount of oxygen added is calculated according to the formula get.

4. The electric furnace carbon injection and oxygen blowing process control method based on target carbon content according to any one of claims 1 to 3, characterized in that: The real-time prediction model for the molten pool temperature adopts a long short-term memory network LSTM or a gated recurrent unit network GRU.

5. The electric furnace carbon injection and oxygen blowing process control method based on target carbon content according to claim 4 is characterized in that: The training process of the molten pool temperature real-time prediction model includes: Collect historical production data, including lime quantity, electricity consumption, flue gas composition and temperature, oxygen consumption, carbon powder quantity, continuous scrap steel addition, natural gas quantity, furnace weight and corresponding bath temperature; Preprocess the collected data, including data cleaning, normalization, and time series alignment; Divide the processed data into training set and validation set; Use the training set to train the neural network model and use the validation set to evaluate the model performance; The model is optimized according to the evaluation results until the predetermined accuracy requirements are met.

6. The electric furnace carbon injection and oxygen blowing process control method based on target carbon content according to claim 5, characterized in that: It also includes an alarm mechanism and a data visualization step; when the calculated carbon content or oxygen content of the molten pool exceeds the preset range, an alarm is issued to the operator; the real-time calculated molten pool temperature, carbon content, oxygen content, and the amount of carbon powder and oxygen added are displayed in the form of a chart on the control interface.

7. A control system for an electric furnace carbon injection and oxygen blowing process based on a target carbon content, using the control method of claim 1, characterized in that: include: Data acquisition module, used to collect real-time data on lime quantity, power consumption, flue gas composition, flue gas temperature, oxygen consumption, carbon powder quantity, continuous addition of scrap steel, natural gas quantity, and furnace weight; The molten pool temperature prediction module is used to predict the molten pool temperature; Carbon and oxygen content calculation module, used to calculate the carbon content and oxygen content of the molten pool; A feeding control module is used to calculate and control the additional amount of carbon powder and oxygen added; Alarm module, used to implement the alarm mechanism; Visualization module, used to realize data visualization function.

8. The electric furnace carbon injection and oxygen blowing process control system based on target carbon content according to claim 7, characterized in that: A data storage module is also included to store historical production data and model parameters to support regular updating and optimization of the model.

Images