Electric furnace tapping control system and method

Abstract

The invention relates to an electric furnace tapping control system and method, and belongs to the technical field of steel smelting systems and methods. According to the technical scheme, liquid level data of a tapping hole and a steel ladle are monitored in real time through a visual identification module, operation parameters of an electric furnace are collected, an artificial intelligence module is used for carrying out data analysis and processing, the state of the tapping hole and the bearing capacity of the steel ladle are judged, then the optimal tapping time is judged in combination with a self-learning mathematical model, and an automatic tapping instruction is sent out; and the control module accurately controls the buggy ladle to move and the tapping hole to open and close so as to realize automatic tapping. The method has the beneficial effects that the operation safety is greatly improved, the safety risk in the automatic tapping process is effectively eliminated, the qualified rate of molten steel is increased, the tapping period is shortened, the working environment is improved, and the labor intensity of workers is reduced.

Description

Technical Field

[0001] The present invention relates to an electric furnace steel tapping control system and method, belonging to the technical field of steel smelting systems and methods. Background Art

[0002] With the rapid development of the steel industry, the market has placed higher demands on the quality and efficiency of steel products. Traditional electric furnace tapping methods rely heavily on manual operation, which presents numerous drawbacks. During manual tapping, tapping time and quantity are primarily controlled by the operator's experience, resulting in poor stability and accuracy during the tapping process. Differences in operator skill levels and operating habits make it difficult to ensure consistent tapping quality, impacting the overall quality of steel products. Manual operation is subject to significant human influence, making it prone to errors during the tapping process. For example, tapping too much or too little not only wastes molten steel but can also impact subsequent refining and continuous casting processes, reducing production efficiency. Furthermore, manual operation poses safety risks. During the tapping process, operators come into close contact with high-temperature molten steel, exposing them to the risk of scalding and burns. Furthermore, manual operation is labor-intensive, and prolonged exposure to high temperatures and noise can also take a toll on the operator's health. In order to overcome the drawbacks of traditional manual steelmaking, improve the production efficiency and quality of electric furnace steelmaking, reduce production costs and safety risks, it has become an inevitable trend to introduce advanced intelligent manufacturing technology and realize the automation and intelligent control of electric furnace steelmaking.

[0003] Traditional electric furnace steel-tapping operations rely on manual experience and judgment, and are generally subject to the following technical defects: (1) Inaccurate timing of steel-tapping, resulting in oxidation of the molten steel or temperature loss; (2) Low precision in controlling the amount of steel tapped, resulting in waste of alloy materials; (3) Significant safety hazards associated with close manual operation; and (4) Differences in skills between operators lead to quality fluctuations.

[0004] The existing technology discloses a steel tapping control method based on weighing feedback, but does not solve the dynamic control problem brought about by the flow characteristics of molten steel; the existing technology proposes an image recognition steel tapping port solution, but lacks a metallurgical process knowledge integration mechanism. Summary of the Invention

[0005] The purpose of the present invention is to provide an electric furnace steel tapping control system and method. By combining a theoretical steel tapping model with on-site operation big data analysis, the control model parameters are continuously optimized, and an advanced image recognition and analysis system based on visual recognition is utilized to monitor the ladle liquid level and the steel tapping situation in real time, thereby rationally optimizing the steelmaking action control. Fully automatic steel tapping replaces manual confirmation of steel tapping, greatly improving the safety of operation, effectively eliminating the safety risks of the automatic steel tapping process, increasing the qualified rate of molten steel, shortening the steel tapping cycle, improving the working environment, reducing manual labor intensity, and effectively solving the above-mentioned problems existing in the background technology.

[0006] The technical solution of the present invention is: an electric furnace tapping control system, comprising an electric furnace, a tapping and liquid level monitoring device, a ladle car device and a control system; The electric furnace comprises an electric furnace body, an electrode, an oxygen supply carbon spray gun, a charging device, an electric furnace tilting control system and a tapping port, the electrode is connected to the electrode power supply control system, the oxygen supply carbon spray gun is connected to the oxygen supply carbon spray control system, the charging device is connected to the charging control system, the tapping port is connected to the electric furnace tapping control system, and the electrode power supply control system, the oxygen supply carbon spray control system, the charging control system, the electric furnace tilting control system and the electric furnace tapping control system are connected to the control system via Ethernet; The steel tapping and liquid level monitoring device includes dual redundant camera 1, dual redundant camera 2, molten steel liquid level monitoring camera and radar rangefinder. The dual redundant camera 1, dual redundant camera 2, molten steel liquid level monitoring camera and radar rangefinder are connected to the control system via Ethernet. The ladle car device includes a ladle, a ladle car, a laser rangefinder, a weighing sensor, a secondary weighing instrument, a tapping position limiter, and a ladle car encoder. The ladle is placed in the middle of the ladle car. The measurement range of the laser rangefinder matches the position of the ladle car 302. The ladle car encoder 307 is set on the ladle car. There are two tapping position limiters, one on each side of the tapping position. The laser rangefinder, weighing sensor, secondary weighing instrument, tapping position limiter, and ladle car encoder are connected to the control system via Ethernet. The control system includes the main control room control client, the steel-outing room control client, the server and the PLC control system. The main control room control client is placed in the main control room, the steel-outing room control client is placed in the steel-outing room, the server is deployed in the computer room, and the PLC control system is deployed in the PLC room; the main control room control client, the steel-outing room control client and the server are connected to the PLC control system through the access switch.

[0007] The control system includes a five-part system architecture, the first part is the data acquisition layer, the second part is the data processing layer, the third part is the model calculation layer, the fourth part is the process control layer, and the fifth part is the intelligent evaluation layer; the data acquisition layer, data processing layer, model calculation layer and process control layer are connected in sequence.

[0008] The control system includes a visual recognition module, an artificial intelligence module, a laser positioning module and a control module. The visual recognition module is connected to the steel tapping and liquid level monitoring device, the visual recognition module is connected to the artificial intelligence module, the laser positioning module is installed on the ladle car, and is connected to the laser rangefinder and the ladle car encoder; the control module is connected to the electric furnace tilting control system and the electric furnace steel tapping control system.

[0009] A method for controlling steel tapping in an electric furnace, characterized by comprising the following steps: S1. Determine the best time to tap steel from the electric furnace based on the electric furnace smelting model; S2. Using the precise positioning model of the ladle car, the ladle car is driven to the tapping position under the electric furnace to wait for steel to be discharged; S3, Ladle car peeling, adopts model control method to realize the peeling function of the ladle car weighing system and gives the electric furnace tapping display; S4. Pull out the vertical support of the electric furnace and wait for steel tapping; S5, the electric furnace automatically swings to the tapping position, starts tapping, and the TBT rotary plate is automatically unlocked, preparing for the next step of automatic tapping; S6, the model issues a steel tapping instruction, and the system starts to tap steel automatically; S7, the precise tapping amount of the electric furnace reaches the planned value, the electric furnace quickly tilts back, and the rotary plate closes to complete automatic tapping; S8. During the electric furnace tapping process, if the slag amount calculated by infrared video monitoring exceeds 100kg, the model will also complete automatic tapping.

[0010] In step S1, the visual recognition module uses dual-redundant camera 1, dual-redundant camera 2 and molten steel level monitoring camera to shoot the tapping port and the ladle liquid level, and extracts characteristic information such as the shape, size, blockage condition of the tapping port and the height and fluctuation of the ladle liquid level through the image recognition algorithm. At the same time, a radar rangefinder is used to monitor, repair and correct the liquid level; the artificial intelligence module uses a machine learning algorithm to train a large amount of collected production data, establish a mathematical model of the tapping process, and combine it with the electric furnace steelmaking endpoint carbon and temperature prediction model to determine the optimal tapping time and tapping amount based on the real-time collected and predicted data, and continuously learn and optimize.

[0011] In step S2, the positional relationship between the ladle car and the tapping port is measured in real time by laser ranging technology to achieve precise positioning of the ladle car. The precise positioning of the ladle car includes the use of an encoder to achieve redundant monitoring of the ladle car distance measurement and the limit correction configuration of the tapping position. The electric furnace tapping position, argon station position, refining position and steel loading position are established according to the laser ranging signal and the relative positions of the electric furnace, refining and continuous casting.

[0012] In step S6, according to the automatic tapping instruction of the model system, the tilting angle of the electric furnace and the movement and positioning of the ladle car are controlled, and the opening and closing of the tapping port are adjusted to achieve precise control of the tapping process.

[0013] The beneficial effects of the present invention are: through a method combining a theoretical steel-tapping model with on-site operation big data analysis, the control model parameters are continuously optimized, and an advanced image recognition and analysis system based on visual recognition is used to monitor the ladle liquid level and the steel-tapping situation in real time, thereby rationally optimizing the steelmaking action control; fully automatic steel-tapping replaces manual confirmation of steel-tapping, which greatly improves the safety of operation, effectively eliminates the safety risks of the automatic steel-tapping process, improves the qualified rate of molten steel, shortens the steel-tapping cycle, improves the working environment, and reduces the intensity of manual labor. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a system structure diagram of the present invention; Figure 2 It is a system architecture block diagram of the control system of the present invention; Figure 3 It is a block diagram of the module composition of the control system of the present invention; Figure 4 It is a workflow diagram of the present invention; In the figure: electric furnace 1, electric furnace body 101, electrode 102, oxygen supply carbon injection gun 103, charging device 104, electric furnace tilting control system 105, tapping port 106, tapping and liquid level monitoring device 2, dual redundant camera 1 201, dual redundant camera 2 202, molten steel level monitoring camera 203, radar rangefinder 204, ladle car device 3, ladle 301, ladle car 302, laser rangefinder 303, weighing sensor 304, weighing secondary instrument 305, tapping position limit 306, ladle car encoder 307, control system 4, main control room control client 401, tapping room control client 402, server 403, PLC control system 404, access switch 405, Ethernet 406, main control room 5, tapping room 6, machine room 7, PLC room 8, transformer 9, tapping position 10, argon station 11, refining position 12, loading position 13. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solutions and advantages of the invention implementation cases clearer, the technical solutions in the invention implementation cases will be clearly and completely described below in conjunction with the drawings in the implementation cases. Obviously, the implementation cases described are only a small part of the implementation cases of the present invention, rather than all the implementation cases. Based on the implementation cases in the present invention, all other implementation cases obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0016] An electric furnace tapping control system, comprising an electric furnace 1, a tapping and liquid level monitoring device 2, a ladle car device 3 and a control system 4; The electric furnace 1 includes an electric furnace body 101, an electrode 102, an oxygen supply carbon spray gun 103, a charging device 104, an electric furnace tilting control system 105 and a tapping port 106, the electrode 102 is connected to the electrode power supply control system, the oxygen supply carbon spray gun 103 is connected to the oxygen supply carbon spraying control system, the charging device 104 is connected to the charging control system, the tapping port 106 is connected to the electric furnace tapping control system, and the electrode power supply control system, the oxygen supply carbon spraying control system, the charging control system, the electric furnace tilting control system 105 and the electric furnace tapping control system are connected to the control system 4 via Ethernet 406; The steel tapping and liquid level monitoring device 2 includes a dual redundant camera 1 201, a dual redundant camera 202, a molten steel liquid level monitoring camera 203, and a radar rangefinder 204. The dual redundant camera 1 201, the dual redundant camera 202, the molten steel liquid level monitoring camera 203, and the radar rangefinder 204 are connected to the control system 4 via Ethernet 406. The ladle car device 3 includes a ladle 301, a ladle car 302, a laser rangefinder 303, a weighing sensor 304, a weighing secondary instrument 305, a tapping position limiter 306, and a ladle car encoder 307. The ladle 301 is placed in the middle of the ladle car 302. The measurement range of the laser rangefinder 303 matches the position of the ladle car 302. The ladle car encoder 307 is set on the ladle car 302. There are two tapping position limiters 306, one on each side of the tapping position. The laser rangefinder 303, the weighing sensor 304, the weighing secondary instrument 305, the tapping position limiter 306, and the ladle car encoder 307 are connected to the control system 4 via Ethernet 406. The control system 4 includes a main control room control client 401, a steel-outing room control client 402, a server 403 and a PLC control system 404. The main control room control client 401 is placed in the main control room, the steel-outing room control client 402 is placed in the steel-outing room, the server 403 is deployed in the machine room, and the PLC control system 404 is deployed in the PLC room; the main control room control client 401, the steel-outing room control client 402 and the server 403 are connected to the PLC control system 404 through the access switch 405.

[0017] The control system 4 includes a five-part system architecture, the first part is the data acquisition layer, the second part is the data processing layer, the third part is the model calculation layer, the fourth part is the process control layer, and the fifth part is the intelligent evaluation layer; the data acquisition layer, data processing layer, model calculation layer and process control layer are connected in sequence.

[0018] The control system 4 includes a visual recognition module, an artificial intelligence module, a laser positioning module and a control module. The visual recognition module is connected to the steel tapping and liquid level monitoring device 2, the visual recognition module is connected to the artificial intelligence module, the laser positioning module is installed on the ladle car, and is connected to the laser rangefinder 303 and the ladle car encoder 307; the control module is connected to the electric furnace tilting control system 105 and the electric furnace steel tapping control system.

[0019] A method for controlling steel tapping in an electric furnace, characterized by comprising the following steps: S1. Determine the best time to tap steel from the electric furnace based on the electric furnace smelting model; S2. Using the precise positioning model of the ladle car, the ladle car is driven to the tapping position under the electric furnace to wait for steel to be discharged; S3, Ladle car peeling, adopts model control method to realize the peeling function of the ladle car weighing system and gives the electric furnace tapping display; S4. Pull out the vertical support of the electric furnace and wait for steel tapping; S5, the electric furnace automatically swings to the tapping position, starts tapping, and the TBT rotary plate is automatically unlocked, preparing for the next step of automatic tapping; S6, the model issues a steel tapping instruction, and the system starts to tap steel automatically; S7, the precise tapping amount of the electric furnace reaches the planned value, the electric furnace quickly tilts back, and the rotary plate closes to complete automatic tapping; S8. During the electric furnace tapping process, if the slag amount calculated by infrared video monitoring exceeds 100kg, the model will also complete automatic tapping.

[0020] In step S1, the visual recognition module uses dual-redundant camera 1, dual-redundant camera 2 and molten steel level monitoring camera to shoot the tapping port and the ladle liquid level, and extracts characteristic information such as the shape, size, blockage condition of the tapping port and the height and fluctuation of the ladle liquid level through the image recognition algorithm. At the same time, a radar rangefinder is used to monitor, repair and correct the liquid level; the artificial intelligence module uses a machine learning algorithm to train a large amount of collected production data, establish a mathematical model of the tapping process, and combine it with the electric furnace steelmaking endpoint carbon and temperature prediction model to determine the optimal tapping time and tapping amount based on the real-time collected and predicted data, and continuously learn and optimize.

[0021] In step S2, the positional relationship between the ladle car and the tapping port is measured in real time by laser ranging technology to achieve precise positioning of the ladle car. The precise positioning of the ladle car includes the use of an encoder to achieve redundant monitoring of the ladle car distance measurement and the limit correction configuration of the tapping position. The electric furnace tapping position, argon station position, refining position and steel loading position are established according to the laser ranging signal and the relative positions of the electric furnace, refining and continuous casting.

[0022] In step S6, according to the automatic tapping instruction of the model system, the tilting angle of the electric furnace and the movement and positioning of the ladle car are controlled, and the opening and closing of the tapping port are adjusted to achieve precise control of the tapping process.

[0023] In practical application, the present invention is composed of an electric furnace 1, a steel tapping and liquid level monitoring device 2, a ladle car device 3 and a control system 4.

[0024] The electric furnace 1 includes a furnace body 101, an electrode 102, an oxygen-supply carbon spray gun 103, a charging device 104, a furnace tilting control system 105, and a tapping port 106. The electrode power supply control system controls the power supply to the electrode 102 to enable smelting; the oxygen-supply carbon spray control system controls the opening and closing of the oxygen-supply carbon spray gun 103 to perform carbon spraying and oxygen blowing; the charging control system controls the operation of the charging device 104 to enable charging operations; the furnace tilting control system 105 controls the furnace tilt angle to enable the furnace to shake and tilt for tapping; and the furnace tapping control system controls the opening of the tapping port 106 to enable automatic tapping. These control systems are connected to the control network 4 via Ethernet 406.

[0025] The tapping and liquid level monitoring device 2 includes dual redundant cameras 1 201 and 202 for monitoring steel flow and slag discharge. These cameras intelligently monitor and control the flow of molten steel during tapping. They also include redundant monitoring via a molten steel level monitoring camera 203 and a radar rangefinder 204, enabling accurate monitoring of the molten steel level during tapping, ensuring accurate tapping volume control, and ensuring the molten steel level remains within a normal range. Signals from these monitoring cameras are connected to the control network 4 via Ethernet 406.

[0026] The ladle car device 3 includes a ladle 301 for holding smelted molten steel, and a ladle car 302. The ladle car 302 can handle both empty and molten steel. It includes a laser rangefinder 303 for ladle distance measurement, a ladle car encoder 307, and two tapping position limiters 306. The laser rangefinder 303 monitors the ladle car's position, and the distance between the ladle car encoder 307 and the laser rangefinder 303 is compared and analyzed to achieve precise control of the ladle car's distance measurement. Monitoring the tapping position limiters 306 corrects and resets distance measurement errors, ensuring the accuracy of the ladle car's distance measurement and the safety of automatic tapping. The device also includes weighing and metering equipment, such as a load cell 304 and a secondary weighing instrument 305, which are connected to the control system 4 via Ethernet 406.

[0027] In control system 4, server 403 is deployed in the computer room, performing data processing and model algorithm development, and accessing PLC control system 404 via switch 405. Tapping room control client 402 is located in the tapping room and accesses PLC control system 404 via switch 405, enabling automatic tapping monitoring. Main control room control client 401 is located in the main control room and accesses PLC control system 404 via switch 405, enabling automatic tapping monitoring.

[0028] The system architecture of the control system 4 consists of five parts: the first part is the data acquisition layer, the second part is the data processing layer, the third part is the model calculation layer, the fourth part is the process control layer, and the fifth part is the intelligent evaluation layer.

[0029] The data acquisition layer is used to collect data such as the shape, size, blockage of the taphole, and the height and fluctuation of the ladle liquid level in real time. It also collects the operating parameters of the electric furnace, such as power supply, charging, furnace temperature, ladle distance measurement, ladle car weighing, etc. Taphole data collection: Dual redundant cameras 1 201 and 202 perform redundant data collection of the taphole. For example, during heat 250201435, produced at 06:58:13 on April 11, 2025, the taphole shape captured by dual redundant camera 1 201 and the shape and size captured by dual redundant camera 2 202 were essentially identical. Both cameras indicated the taphole was normal and unobstructed.

[0030] Ladle level data collection: Redundant dual detection is performed using the molten steel level monitoring camera 203 and the radar rangefinder 204. In this heat, the ladle level height calculated by the molten steel level monitoring camera 203 was 400 mm, while the radar rangefinder 204 recorded 410 mm, with a deviation of 10 mm, which is within the error range, indicating that the automatic liquid level height collection is highly accurate.

[0031] Electric furnace operation parameter collection, power-on start time: 2025-04-11 06:17:40, power-on end time: 2025-04-11 06:54:03, Power consumption: 67230.000 Kwh. Feedings: 4029.87 kg of activated calcium ash, 700 g of graphite electrode, 0.15 ton, 3032.95 kg of light-burned dolomite, 127.5 ton of scrap steel, 33 ton of DRI, and 3661.46 kg of carbon powder.

[0032] The data processing layer performs real-time analysis and processing on the data collected by the data acquisition layer, and uses artificial intelligence algorithms and metallurgical process knowledge to determine the state of the tapping hole and the theoretical tapping volume; Real-time analysis of collected data revealed that the tapping port for heat 250201435 was normal and unobstructed. Based on the input of 127.5 tons of scrap steel, 33 tons of DRI, 8 tons of flux and other materials, and 7 tons of slag discharge, the AI ​​model, combined with AI algorithms and metallurgical process knowledge, combined with the furnace steel retention and smelting losses, predicted a theoretical tapping yield of 169.5 tons.

[0033] The model calculation layer, combined with the self-learning mathematical model, automatically determines the optimal tapping time based on the analysis results of the data processing layer and the electric furnace steelmaking endpoint carbon and temperature prediction model; Integrating a self-learning mathematical model, the optimal tapping timing is determined based on the results of the data processing layer. For this heat, the model calculated and analyzed the current taphole status, theoretical tapping volume, and real-time furnace temperature, molten steel composition, smelting conditions, and other parameters, determining the optimal tapping timing to be 2025-04-11 06:55:00, plus or minus 1 minute.

[0034] The process control layer issues automatic steel tapping control instructions based on the judgment results of the model calculation layer, controls the tilting angle of the electric furnace and the movement and positioning of the ladle car, adjusts the opening and closing of the steel tapping port, and realizes automatic steel tapping.

[0035] After determining the optimal tapping time, the process control layer issues a command to the furnace automation control system. It controls the furnace's tilt angle to 4 degrees, ensuring smooth molten steel flow. Simultaneously, it controls the ladle car to move to the precise position, with a positioning error within 5 mm. The taphole opens, enabling automatic tapping and automatically resetting the tilting mechanism upon completion.

[0036] The intelligent evaluation layer conducts intelligent statistical analysis based on the automatic and non-automatic steel tapping production conditions of the electric furnace model, realizes intelligent evaluation of the standardized operation of electric furnace smelting steel tapping control, improves the standardization of the electric furnace smelting steel tapping process, and thus realizes precise steel tapping, ensures the stability of steel tapping volume, and improves the quality of molten steel.

[0037] The intelligent evaluation layer performs statistical analysis on both automatic and non-automatic tapping. For this particular furnace, for example, during the automatic tapping process, the tapping volume was 170 tons, within 0.5 tons of the planned volume, resulting in an A rating for the automatic tapping performance of the furnace. Since this furnace received an A rating, the model's automatic tapping control level was exceptionally high, meeting the requirements for standardized operation and stable tapping volume. The tapping performance was considered excellent. Therefore, the furnace data was subjected to self-learning to calibrate the automatic tapping control model parameters. However, the automatic tapping performance rating was considered B. The model's automatic tapping control level was high, meeting the requirements for standardized operation and stable tapping volume. The tapping performance was considered good. Therefore, the furnace data was not subjected to self-learning to calibrate the automatic tapping control model parameters, and the original tapping control model parameters were retained. The automatic tapping performance evaluation is C. The model's automatic tapping volume control level is low, which basically meets the requirements of standardized operation and the stability of molten steel tapping volume. The tapping operation performance is medium. The furnace data will not be self-learned, and the automatic tapping control model parameters will be corrected. The original tapping control model parameters will be maintained, and the reasons for the furnace will be analyzed. It is recommended to pay attention to the endpoint. The automatic tapping performance evaluation is D. The model's automatic tapping volume control level is too low, which does not meet the requirements of standardized operation and the stability of molten steel tapping volume. The tapping operation performance is poor. The reasons for the furnace will be analyzed, and it is recommended to pay attention to the endpoint. The automatic tapping performance evaluation is E. The model's automatic tapping volume control fails, which does not meet the requirements of standardized operation and the stability of molten steel tapping volume. The tapping operation performance is unqualified. The reasons for the furnace will be analyzed and assessed. Unqualified automatic tapping operations are strictly prohibited. In such cases, manual tapping should be switched to in a timely manner during operation, and the reasons should be analyzed.

[0038] Table 1 - Evaluation table of the advantages and disadvantages of automatic steel tapping in electric furnaces

[0039] The modules of control system 4 include: The visual recognition module uses a camera to capture images of the taphole and ladle liquid level. Using an image recognition algorithm, it extracts characteristic information such as the taphole's shape, size, and blockage, as well as the ladle liquid level's height and fluctuations. It also uses radar ranging to monitor, inspect, and correct liquid level deviations. The artificial intelligence module uses machine learning algorithms to train a large amount of collected production data, establish a mathematical model of the steel-making process, and automatically determine the optimal steel-making time and amount based on the real-time collected data, and continuously learn and optimize.

[0040] The laser positioning module, installed on the ladle car, uses laser ranging technology to measure the positional relationship between the ladle car and the tapping hole in real time, enabling precise positioning of the ladle car. This laser positioning module achieves precise positioning of the ladle car, including redundant monitoring of ladle car distance measurement using the ladle car encoder 307 and configuration of limit corrections for the tapping position. Based on the laser ranging signals and the relative positions of the electric furnace, refining, and continuous casting, the electric furnace tapping position, argon station position, refining position, and loading position are established.

[0041] The control module controls the tilting angle of the electric furnace and the movement and positioning of the ladle car according to the automatic tapping instructions, adjusts the opening and closing of the tapping port, and realizes precise control of the tapping process.

[0042] Its visual recognition module uses dual redundant camera 1 201 and dual redundant camera 2 202 to detect the steel flow at the tapping port. Through the image recognition algorithm, it captures the electric furnace tapping information in real time, determines whether slag is released, and calculates the slag release amount in real time. Once the predicted slag release amount reaches the set value or the tapping amount reaches the planned tapping amount, the tapping is automatically terminated.

[0043] The ladle liquid level control is monitored by a control model that combines the molten steel liquid level monitoring camera 203 monitoring and the radar rangefinder 204 liquid level monitoring.

[0044] Its artificial intelligence module uses machine learning algorithms to train a large amount of production data collected through the server to establish a mathematical model of the steel output and slag weight data during the steelmaking process.

[0045] The machine learning algorithm includes a dynamic learning algorithm module, which specifically includes the following sub-modules: a time series prediction sub-module that uses an LSTM network to process continuous smelting data; an image recognition sub-module that applies a convolutional neural network to analyze the steel-mouth image; and a multi-objective optimization sub-module that combines the NSGA-II algorithm to balance steel-making efficiency and quality indicators.

[0046] It also includes a live tapping module. According to the prediction of the electric furnace model, when the molten steel temperature is expected to reach the tapping temperature in 1 minute, tapping will begin. After the electric furnace angle reaches 4°, the electrode will automatically shut down and move to the tapping position. The electric furnace will continue to complete the remaining tapping actions until tapping is completed, completing automatic tapping control.

[0047] The method implementation process of the present invention is: S1. Determine the best time to tap steel from the electric furnace based on the electric furnace smelting model.

[0048] S2. The ladle car precise positioning model enables the ladle car to automatically drive to the tapping position under the electric furnace and wait for steel to be unloaded.

[0049] S3. Automatic tare removal of ladle car. The model automatic control method is adopted to realize the automatic tare function of the ladle car weighing system and provide accurate display of steel tapping from the electric furnace.

[0050] S4. Pull out the vertical support of the electric furnace and wait for steel to be tapped.

[0051] S5. The electric furnace automatically swings to the tapping position and starts tapping. At the same time, the TBT rotary plate is automatically unlocked, preparing for the next step of automatic tapping.

[0052] S6. The model issues a steel-out instruction, and the system starts to automatically output steel.

[0053] S7. The precise tapping amount of the electric furnace reaches the planned value, the electric furnace quickly tilts back, and the rotary plate closes to complete automatic tapping.

[0054] S8. During the electric furnace tapping process, if the slag amount calculated by infrared video monitoring exceeds 100kg, the model will also complete automatic tapping.

[0055] This invention significantly reduces tapping waiting time, shortens the smelting cycle, achieves standardized molten steel tapping, and enables precise steel metering and tapping. It reduces steel consumption by 0.60 kg / t. After implementation, the project reduced steelmaking costs by 1.68 yuan / t, generating an annual economic benefit of 3.36 million yuan.

[0056] The present invention can achieve standardized management of electric furnace tapping at the production site, helping to achieve efficient production of electric furnaces. It uses a method that combines theoretical tapping models with on-site operation big data analysis to continuously optimize control model parameters. It utilizes an advanced image recognition and analysis system based on visual recognition to monitor the ladle liquid level and real-time forecast of tapping conditions, thereby rationally optimizing steelmaking operation control. Fully automatic tapping replaces manual confirmation of tapping, greatly improving operational safety, effectively eliminating safety risks in the automatic tapping process, increasing the qualified rate of molten steel, shortening the tapping cycle, improving the working environment, and reducing labor intensity. Specifically, it is embodied in: (1) improving the stability and consistency of electric furnace tapping; (2) reducing the burden on operators and human errors; (3) saving tapping time; (4) facilitating safe tapping; and (5) reducing consumption, increasing the qualified rate, and shortening the cycle.

Claims

1. An electric furnace tapping control system, characterized by: It comprises an electric furnace (1), a steel tapping and liquid level monitoring device (2), a ladle car device (3) and a control system (4); The electric furnace (1) comprises an electric furnace body (101), an electrode (102), an oxygen supply carbon spray gun (103), a charging device (104), an electric furnace tilting control system (105) and a steel tapping port (106), wherein the electrode (102) is connected to the electrode power supply control system, the oxygen supply carbon spray gun (103) is connected to the oxygen supply carbon spraying control system, the charging device (104) is connected to the charging control system, and the steel tapping port (106) is connected to the electric furnace steel tapping control system, and the electrode power supply control system, the oxygen supply carbon spraying control system, the charging control system, the electric furnace tilting control system (105) and the electric furnace steel tapping control system are connected to the control system (4) via Ethernet (406); The steel tapping and liquid level monitoring device (2) includes a dual redundant camera 1 (201), a dual redundant camera 2 (202), a molten steel liquid level monitoring camera (203) and a radar range finder (204), and the dual redundant camera 1 (201), the dual redundant camera 2 (202), the molten steel liquid level monitoring camera (203) and the radar range finder (204) are connected to the control system (4) via Ethernet (406); The ladle car device (3) comprises a ladle (301), a ladle car (302), a laser rangefinder (303), a weighing sensor (304), a weighing secondary instrument (305), a steel tapping position limiter (306) and a ladle car encoder (307), wherein the ladle (301) is placed in the middle of the ladle car (302), the measuring range of the laser rangefinder (303) matches the position of the ladle car (302), the ladle car encoder (307) is arranged on the ladle car (302), and there are two steel tapping position limiters (306), which are respectively arranged on both sides of the steel tapping position; the laser rangefinder (303), the weighing sensor (304), the weighing secondary instrument (305), the steel tapping position limiter (306) and the ladle car encoder (307) are connected to the control system (4) via Ethernet (406); The control system (4) includes a main control room control client (401), a steel-outlet room control client (402), a server (403) and a PLC control system (404), wherein the main control room control client (401) is placed in the main control room, the steel-outlet room control client (402) is placed in the steel-outlet room, the server (403) is deployed in the machine room, and the PLC control system (404) is deployed in the PLC room; The main control room control client (401), the steel-outlet room control client (402) and the server (403) are connected to the PLC control system (404) via the access switch (405).

2. The electric furnace tapping control system according to claim 1, characterized in that: The control system (4) comprises a five-part system architecture, the first part being a data acquisition layer, the second part being a data processing layer, the third part being a model calculation layer, the fourth part being a process control layer, and the fifth part being an intelligent evaluation layer; The data acquisition layer, data processing layer, model calculation layer and process control layer are connected in sequence.

3. The electric furnace tapping control system according to claim 1, characterized in that: The control system (4) comprises a visual recognition module, an artificial intelligence module, a laser positioning module and a control module, wherein the visual recognition module is connected to the steel tapping and liquid level monitoring device (2), the visual recognition module is connected to the artificial intelligence module, the laser positioning module is installed on the ladle car, and is connected to the laser rangefinder (303) and the ladle car encoder (307); and the control module is connected to the electric furnace tilting control system (105) and the electric furnace steel tapping control system.

4. A method for controlling steel tapping in an electric furnace, characterized in that The following steps are involved: S1. Determine the best time to tap steel from the electric furnace based on the electric furnace smelting model; S2. Using the precise positioning model of the ladle car, the ladle car is driven to the tapping position under the electric furnace to wait for steel to be discharged; S3, Ladle car peeling, adopts model control method to realize the peeling function of the ladle car weighing system and gives the electric furnace tapping display; S4. Pull out the vertical support of the electric furnace and wait for steel tapping; S5, the electric furnace automatically swings to the tapping position, starts tapping, and the TBT rotary plate is automatically unlocked, preparing for the next step of automatic tapping; S6, the model issues a steel tapping instruction, and the system starts to tap steel automatically; S7, the precise tapping amount of the electric furnace reaches the planned value, the electric furnace quickly tilts back, and the rotary plate closes to complete automatic tapping; S8. During the electric furnace tapping process, if the slag amount calculated by infrared video monitoring exceeds 100kg, the model will also complete automatic tapping.

5. The method for controlling steel tapping in an electric furnace according to claim 4, wherein: In step S1, the visual recognition module uses dual-redundant camera 1, dual-redundant camera 2 and molten steel level monitoring camera to shoot the tapping port and the ladle liquid level, and extracts characteristic information such as the shape, size, blockage condition of the tapping port and the height and fluctuation of the ladle liquid level through the image recognition algorithm. At the same time, a radar rangefinder is used to monitor, repair and correct the liquid level; the artificial intelligence module uses a machine learning algorithm to train a large amount of collected production data, establish a mathematical model of the tapping process, and combine it with the electric furnace steelmaking endpoint carbon and temperature prediction model to determine the optimal tapping time and tapping amount based on the real-time collected and predicted data, and continuously learn and optimize.

6. The method for controlling steel tapping in an electric furnace according to claim 4, wherein: In step S2, the positional relationship between the ladle car and the tapping port is measured in real time by laser ranging technology to achieve precise positioning of the ladle car. The precise positioning of the ladle car includes the use of an encoder to achieve redundant monitoring of the ladle car distance measurement and the limit correction configuration of the tapping position. The electric furnace tapping position, argon station position, refining position and steel loading position are established according to the laser ranging signal and the relative positions of the electric furnace, refining and continuous casting.

7. The method for controlling steel tapping in an electric furnace according to claim 4, wherein: In step S6, according to the automatic tapping instruction of the model system, the tilting angle of the electric furnace and the movement and positioning of the ladle car are controlled, and the opening and closing of the tapping port are adjusted to achieve precise control of the tapping process.

Images