Intelligent refining LF furnace end point temperature calculation method

Abstract

The invention relates to an intelligent refining LF (ladle furnace) terminal temperature calculation method, which comprises the following steps of: performing interval division on refining LF terminal temperature historical data, molten steel tank turnover time, standing time and LF terminal temperature according to a technological process, wherein the number nlt of LF refining production furnaces from one month to one year; the method comprises the following steps of: setting a qualified temperature interval Taim of the tundish temperature when the time is 20, calculating an average value Ttund of acquisition points as the tundish temperature, calculating a tank number H of which Ttud is less than Taim and a tank number C of which Ttud is greater than Taim, calculating probabilities that the tundish temperature Ttund in each temperature interval falls into Taim, Ttud is less than Taim and Ttud is greater than Taim, calculating an index Mm of each set interval, and calculating a minimum value M alpha of the temperature indexes, the refining LF furnace is moved out according to the calculated temperature interval result, and the subsequent actual tundish temperature result is returned to the next round of calculation; according to the method, a calculation model of the final temperature of the refining LF furnace is simplified in a probability calculation mode, the model calculation speed is increased by 80% on the original basis, and the tundish temperature calculation accuracy is improved by 60%.

Description

Technical Field

[0001] The present invention relates to the technical field of iron and steel metallurgy, and particularly relates to a method for calculating the end-point temperature of a smart refining LF furnace. Background Art

[0002] The tundish temperature of molten steel is affected by various factors, such as the refining discharge temperature, the turnover cycle of the ladle for holding molten steel, the continuous casting order number of molten steel, the tundish baking time, the ladle transfer time, the ambient temperature, etc. Technicians have been looking for a relational expression that can represent the relationship between the tundish temperature and various measurable parameters.

[0003] After the development of big data science and mathematical fitting model science, technicians input various measured parameters of the steel mill for calculation. Although a fitting model between the tundish temperature and each parameter can be obtained, the error of the model is too large, resulting in a very low reliability of the model. Often, the error ratio in the model exceeds 50%, which cannot meet the production requirements. The possible reasons for this result are that the errors of the collected parameters themselves are relatively large. For example, the tundish temperature is collected as the point value at half of the ladle weight, but the actual tundish temperature is the continuous value during the entire casting cycle of molten steel. The temperature value at a single point cannot evaluate the discharge temperature of the refining LF. Or there are parameters that have a great influence on the tundish temperature but are not recognized or cannot be measured. Whatever the reason, it leads to the inability to calculate the discharge temperature of the refining LF furnace by mathematical methods, thus hindering the progress of intelligent models and unmanned operations.

[0004] According to the calculation results of the model, it is found that the most significant contributor to the tundish temperature is the end-point temperature of the refining LF furnace. There is an order-of-magnitude difference between the contribution values of the remaining measurable parameters and this parameter of the end-point temperature of the refining LF furnace. Therefore, this patent uses the expectation of the qualified probability of the tundish temperature to calculate and proposes a method for calculating the end-point temperature of a smart refining LF furnace to improve the control accuracy of the tundish temperature. Summary of the Invention

[0005] The present invention provides a method for calculating the end-point temperature of a smart refining LF furnace, which improves the control accuracy of the tundish temperature, effectively controls the production and smelting costs, ensures the stable and smooth progress of the production process, and improves the product quality.

[0006] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0007] A method for calculating the end-point temperature of a smart refining LF furnace includes the following steps:

[0008] S1. When the process is converter smelting → LF refining → continuous casting or ingot casting, or converter smelting → RH refining or CAS refining → LF refining → continuous casting or ingot casting, select the steel grade;

[0009] S2. Classify the ladles according to the historical data of the turnover time as L1 , L 2 , L 3 , …, L x , …L i ;

[0010] S3. Divide the historical data of the calming time into intervals of 10 min to 20 min, t 1 , t 2 , t 3 , …t y , …t n ;

[0011] S4. Divide the historical data of the end-point temperature of the refining LF furnace into intervals of 5 °C to 20 °C, and divide them into a total of m temperature intervals T 1 , T 2 , …, T m ;

[0012] S5. Combine the ladle turnover time, calming time, and LF furnace end-point temperature intervals to form a set M {L x t y T z};

[0013] S6. Count the number of LF refining production furnaces n within each set interval in the time interval from one month to one year. If n < 20, no calculation is performed;

[0014] S7. Set the qualified temperature interval T aim ;

[0015] S8. Set the sampling points for the tundish temperature during the casting process of each ladle of molten steel in minutes, and calculate the average value T tund of the sampling points as the tundish temperature;

[0016] S9. Calculate the number of ladles H where T tund < T aim , the number of ladles C where T tund > T aim . According to the values of H, C, and n, calculate the probability that the tundish temperature T tund falls within T aim , T tund < T aim , T tund > T aim . The calculation formula is as follows:

[0017]

[0018]

[0019]

[0020] Among them, UCL is the upper control limit of the temperature in each interval, and LCL is the lower control limit of the temperature in each interval;

[0021] S10. Calculate the index M of each set interval m ;

[0022]

[0023] Among them, UCL m is the upper control limit of the temperature in the m-th set interval, and LCL m is the lower control limit of the temperature in the m-th set interval;

[0024] S11. According to the calculation results of each temperature index, calculate the minimum value M of the temperature index α , M α corresponds to the best LF end-point temperature interval T under the ladle turnover time L X and the holding time t y ;

[0025] M α = min{M 1 , M 2 , …, M m} (5)

[0026] S12. The refining LF furnace is removed according to the calculated temperature interval results, and the results of the subsequent actual tundish temperature are fed back into the next round of calculations.

[0027] Furthermore, the formula for setting the qualified temperature interval of the tundish temperature in step S7 is as follows:

[0028] T aim = {T liq + T suheat - T CL , T liq + T supheat + T CL} (6)

[0029] Among them, T liq is the liquidus temperature of the steel grade, T supheat is the superheat of the steel grade set by the production plant, and T CL is the temperature control accuracy set by the production plant according to its own equipment process level.

[0030] Compared with the prior art, the beneficial effects of the present invention are:

[0031] By using the method of probability calculation, the calculation model of the end-point temperature of the refining LF furnace is simplified. On the original basis, the calculation speed of the model is increased by 80%, the calculation accuracy of the tundish temperature is improved by 60%, the product quality is improved, and the production and smelting cost is effectively controlled. Different production enterprises can adjust the temperature control accuracy T CL in such a way that, according to the historical data under their own equipment and tooling conditions, the tundish temperature is adjusted purposefully, and the adjustment accuracy makes the result flexible and reliable. Specific implementation mode

[0032] The following further describes the specific implementation mode of the present invention:

[0033] A method for calculating the end-point temperature of an intelligent refining LF furnace according to the present invention includes the following steps:

[0034] S1. The process is converter smelting → LF refining → continuous casting. The steel grade Q235B is selected, and the following is all for the steel grade Q235.

[0035] S2. The turnover time of the ladle is 132 min, and it is queried that the computer setting falls within L 3 inside.

[0036] S3. The holding time is 45 min, and it is queried that the computer setting falls within the turnover time t of the ladle 3 in the ladle turnover time t 3 .

[0037] S4. The tapping temperature of the refining LF starts from 1560 °C and ends at 1590 °C. It is divided into 6 temperature intervals T 1 , T 2 , …, T 6 according to a 5 °C temperature interval.

[0038] S5. The ladle turnover time, the holding time, and the end-point temperature interval of the LF furnace form a set M {L 3 t 3 T z}.

[0039] S6. Statistically, within a certain period of time, the number of LF refining production furnaces n in each set interval is counted. When n < 20, no calculation is performed;

[0040] S7. Set the qualified temperature interval of the tundish temperature:

[0041] T aim = {T liq + T suheat - T CL , T liq + T supheat + T CL} (7)

[0042] T liq is the liquidus temperature of the steel grade, 1518 °C, T supheat is the superheat of the steel grade set by the production plant, 25 °C, T CL is the temperature control accuracy set by the production plant according to its own equipment and process level, 1 °C. Therefore, the qualified temperature range of the tundish temperature T aim = {1542 °C, 1544 °C}.

[0043] S8. Set the sampling points of the tundish temperature during the casting process of each ladle of molten steel in minutes, and calculate the average value of the sampling points T tund is the tundish temperature;

[0044] S9. Calculate T tund < T aim of the number of ladles H, T tund > T aim of the number of ladles C. According to the values of H, C, and n, see Table 1;

[0045] Table 1

[0046]

[0047]

[0048] Calculate the tundish temperature T in each temperature range tund falls within T aim , T tund < T aim , T tund > T aim probability,

[0049]

[0050]

[0051]

[0052] For the first group and the sixth group, n < 20, no calculation is performed,

[0053]

[0054] UCL 2 = 0.582, UCL 3 = 0.709, UCL 4 = 0.630, UCL 5 = 0.663;

[0055] LCL 2 = -0.01, LCL 3 = 0.148, LCL 4= 0.203, LCL 5 = 0.170.

[0056] S10. Calculate the index M for each set interval m ;

[0057]

[0058] M 2 = 0.048, M 3 = 0.103, M 4 = 0.074, M 5 = 0.086,

[0059] M α = min{M 1 , M 2 , …, M m} (12)

[0060] M α = M 4 = {L 3 t 3 T 4}

[0061] S11. Obtain the turnover time L of the ladle for steel grade Q235B 3 , the holding time t 3 Under the condition, the optimal tapping temperature range of the refining LF furnace is [1575°C, 1580°C).

[0062] S12. The refining LF furnace is tapped according to the calculated temperature range result [1575°C, 1580°C), and the result of the tundish temperature in the subsequent process is regressed into the next round of calculation.

[0063] The above embodiments are implemented on the premise of the technical solution of the present invention, and the detailed implementation manners and specific operation processes are given, but the protection scope of the present invention is not limited to the above embodiments. The methods used in the above embodiments are all conventional methods unless otherwise specified.

Claims

1. A method for calculating the endpoint temperature of an intelligent refining LF furnace, characterized in that: The steps include: S1. When the process is converter smelting → LF refining → continuous casting or mold casting, or converter smelting → RH refining or CAS refining → LF refining → continuous casting or mold casting, select the steel grade; S2. Classify the steel ladle into L1, L2, L3, ..., L according to the historical data of turnover time. x ,…L i ; S3. Divide the historical data of sedation time into intervals of 10 minutes to 20 minutes, t1, t2, t3, ...t y ,…t n ; S4, divide the historical data of the final temperature of the refining LF furnace into m temperature intervals T1, T2, ..., T m ; S5, the steel ladle turnover time, calming time, LF furnace end temperature interval into a set M{L x t y T z }; S6. Count the number of LF refining production furnaces n in each collection interval within a period of one month to one year. No calculation is performed when n<20; S7. Set the qualified temperature range T of the tundish temperature aim ; S8. Set the temperature of the tundish in each tank of molten steel casting process in minutes as the unit of collection points, and calculate the average value T of the collection points tund is the tundish temperature; S9. Calculate T tund <T aim Number of tanks H, T tund >T aim The number of tanks C, according to the values ​​of H, C, n, calculate the temperature T of the tundish in each temperature range tund Falling on T aim 、T tund <T aim 、T tund >T aim The probability of; the calculation formula is as follows: Among them, UCL is the upper control limit of the temperature in each interval, and LCL is the lower control limit of the temperature in each interval; S10. Calculate the index M of each set interval m ; Among them, UCL m is the upper control limit of the temperature in the mth set interval, LCL m is the lower control limit of the temperature in the mth set interval; S11. Calculate the minimum value M of the temperature index based on the calculation results of each temperature index α , M α The corresponding is the steel tank turnover time L X , sedation time t y Under this condition, the optimal LF endpoint temperature range is T; M α =min{M1,M2,…,M m } (5) S12. The refining LF furnace is moved out according to the calculated temperature range results, and the results of the subsequent actual tundish temperature are returned to the next round of calculation.

2. The method for calculating the endpoint temperature of a smart refining LF furnace according to claim 1, characterized in that: The formula for setting the qualified temperature range of the tundish temperature in step S7 is as follows: T aim ={T liq +T suheat -T CL ,T liq +T supheat +T CL } (6) Among them, T liq is the liquidus temperature of the steel grade, T supheat Set the superheat of steel grade for the production plant, T CL The temperature control accuracy is set by the manufacturer according to the process level of its own equipment.

Images