Converter control method based on static model

Abstract

The invention relates to a converter control method based on a static model, and belongs to the technical field of ferrous metallurgy. The method is used for solving the problems of low control precision and insufficient self-learning ability caused by the fact that a small-tonnage converter depends on artificial experience or a fixed parameter static model. The method comprises the steps that molten iron, waste steel and terminal point target information is obtained, and the target molten steel amount, the final slag amount and the furnace feeding slag remaining amount are calculated based on historical data; the magnesium oxide saturation content, the phosphorus distribution ratio, the heat balance and the oxygen balance in the slag are combined to respectively calculate the adding amount and the oxygen blowing amount of the magnesium-containing auxiliary material, lime and the cooling agent / heating agent; and by comparing model prediction and production actual performance data, similar reference heat is screened to realize model parameter self-learning. The calculation precision of the auxiliary materials and the oxygen blowing amount is improved, the hit rate exceeds 80%, and reliable support is provided for automatic control of the converter.

Description

Technical Field

[0001] The present invention belongs to the technical field of iron and steel metallurgy, and relates to a converter control method based on a static model Background Art

[0002] As a core device in iron and steel metallurgy, the converter undertakes key chemical reactions such as oxidation and decarbonization during the steelmaking process. Due to the complex and dynamic internal reactions, accurately controlling the addition amount of auxiliary materials and the oxygen blowing amount is crucial for the quality of molten steel. Currently, large-tonnage converters (over 120 tons) generally use a sublance device to real-time monitor the composition and temperature of molten steel. However, this device requires a large installation space, making it difficult to apply in small-tonnage converters below 100 tons. Therefore, small-tonnage converters mostly rely on manual experience or static models based on fixed parameters for control. However, manual experience is greatly affected by subjective factors of operators and has poor stability; although conventional static models can partially replace manual work, their parameters are mostly preset or manually updated regularly and cannot be dynamically optimized according to the real-time furnace conditions, resulting in weak self-learning ability and insufficient calculation accuracy of the models. In addition, traditional models have poor dynamic adaptability to key parameters such as slag amount and heat balance, and it is difficult to meet the requirements of high-precision steelmaking, seriously restricting the realization of converter automatic control. Therefore, there is an urgent need for a static model control method that can optimize parameters in real time and improve the self-learning ability to solve the technical bottlenecks of low control accuracy and dependence on manual experience in small-tonnage converters Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a converter control method based on a static model

[0004] To achieve the above purpose, the present invention provides the following technical solutions

[0005] A converter control method based on a static model, comprising the following steps

[0006] S1. Obtain the hot metal information, scrap information, auxiliary material information and end-point target information of this heat

[0007] S2. Calculate the target molten steel amount, final slag amount and slag left from the previous heat of this heat according to historical data

[0008] S3. Calculate the amount of magnesium-containing auxiliary materials added based on the slag amount and the saturated content of magnesium oxide in the slag

[0009] S4. Calculate the amount of lime added according to the phosphorus distribution ratio

[0010] S5. Calculate the amount of coolant or heating agent added based on heat balance

[0011] S6. Calculate the total oxygen blowing amount through oxygen balance

[0012] S7. Compare the model prediction value with the actual production value, and screen the qualified data for model self-learning.

[0013] Furthermore, the calculation of the slag remaining amount in the previous furnace in S2 includes:

[0014] Fit the slag volume V according to the tilting angle x at the end of slag tapping from the converter ReSlag , and the formula is V ReSlag = 7×10 11 ×e -0.275x , and calculate the density ρ through the slag composition ReSlag , and finally obtain the slag remaining amount W in the previous furnace ReSlag = ρ ReSlag ×V ReSlag .

[0015] Furthermore, the calculation formula for the saturated content of magnesium oxide in the slag in S3 is:

[0016]

[0017] where T AimSteel is the target temperature at the end of this heat, and %TFe, %CaO, and %SiO2 are the corresponding component contents in the slag of the previous heat.

[0018] Furthermore, in S4, the lime addition amount is determined by the relationship between the target phosphorus distribution ratio L P，目标 and the theoretical phosphorus distribution ratio L P，理论 , where:

[0019]

[0020] and it satisfies L P，理论 ≥k.L P，目标 , where k is a correction coefficient, k≥1, and B0 to B4 are the regression parameters of the reference heats.

[0021] Furthermore, the calculation formula for the total oxygen blowing amount V O2 in S6 is:

[0022] V O2 = a(∑V OSS ) + bV Oyc + cV Olq + dV Oks

[0023] where ∑V OSS is the oxygen amount required for the burning loss of each element, V Oyc is the oxygen consumption of the dust, V Olq is the free oxygen amount in the furnace gas, V Oks is the oxygen supply from the decomposition of the coolant or heating agent; a to d are correction coefficients obtained by regression of the classified reference heats.

[0024] Further, in S7, the conditions for screening reference furnace heats include:

[0025] The hot metal weight ranges from 60t to 75t, the scrap weight ranges from 7t to 25t, the C content of the hot metal is 3.7% to 4.8%, the Si content of the hot metal is 0.1% to 0.8%, and the hot metal temperature is 1220°C to 1470°C; and the parameter fluctuation ranges of the current heat meet the requirements of ±5t for hot metal weight, ±20°C for hot metal temperature, and ±4t for scrap weight; the fluctuation range of hot metal C is ±0.2%, and the fluctuation range of hot metal Si is ±0.1%.

[0026] Further, in S7, the specific method of model self-learning includes:

[0027] Compare the model parameters B0 to B4, a to d in S4 to S6 with the parameters deduced from production performance. If the relative deviation of the parameters is less than or equal to 5%, directly store and call them; if the relative deviation is greater than 5%, store and call them after parameter compensation according to the deviation amount.

[0028] Further, in S5, the calculation of the heat surplus is based on the difference between the heat input and output of the reference furnace heat, and the addition amount of the coolant or the heating agent is determined through heat balance.

[0029] Further, the data for model self-learning in S7 is selected from the furnace heats similar to the current heat in the production performance of the last 100 furnaces.

[0030] Further, the specific method of the parameter compensation is as follows:

[0031] If the relative deviation of the parameter is greater than 5%, the compensated parameter value P 新 is calculated according to the following formula:

[0032]

[0033] where P 模型 is the model prediction parameter value, ΔP = P 实绩 - P 模型 , and P 实绩 is the parameter value deduced from production performance.

[0034] The beneficial effects of the present invention are as follows:

[0035] (1) By optimizing the model algorithm (such as the calculation of the magnesia saturation content of the magnesium-containing auxiliary material, the matching of the phosphorus distribution ratio, and the multi-factor oxygen blowing amount formula), the problem of fixed parameters in the traditional static model is overcome, so that the hit rates of lime, magnesium-containing auxiliary materials, and oxygen blowing amount all exceed 80%, significantly reducing the manual intervention error.

[0036] (2) By comparing the model prediction values with the actual production data in real time, the parameters (such as B0 - B4 for lime calculation and a - d for oxygen blowing volume) are dynamically corrected, and automatic compensation is performed based on the deviation threshold (≤5%), so that the model parameters are continuously optimized to adapt to the working condition fluctuations of different furnace heats.

[0037] (3) Aiming at the pain point that converters below 100 tons lack sublance monitoring, a high-precision static model is provided to replace manual experience, providing a technical basis for the unmanned operation of the entire converter process.

[0038] (4) By screening similar reference furnace heats to optimize the model input data, the interference of abnormal working conditions is reduced, the fluctuation range of molten steel composition is narrowed, and the consistency of product quality is improved.

[0039] Other advantages, objectives and features of the present invention will be described to some extent in the subsequent specification, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be described in detail preferably with reference to the accompanying drawings, where:

[0041] Figure 1 is the flow chart of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0042] The following specific examples illustrate the embodiments of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the drawings provided in the following examples only illustrate the basic concept of the present invention in a schematic manner. Without conflict, the following examples and the features in the examples can be combined with each other.

[0043] Among them, the drawings are only for illustrative purposes, showing only schematic diagrams, not physical diagrams, and should not be construed as a limitation to the present invention; in order to better illustrate the embodiments of the present invention, some components in the drawings will be omitted, enlarged or reduced, and do not represent the dimensions of the actual product; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0044] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "rear", etc. indicating the orientation or positional relationship, it is based on the orientation or positional relationship shown in the accompanying drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, the terms describing the positional relationship in the accompanying drawings are only for illustrative purposes and should not be construed as a limitation of the present invention. For those of ordinary skill in the art, the specific meanings of the above terms can be understood according to specific circumstances.

[0045] A converter control system based on a static model includes an oxygen blowing amount calculation module, a lime calculation module, a magnesium-containing auxiliary material calculation module, a coolant (heating agent) calculation module, and a self-learning module. The control method of the system is as follows:

[0046] S1. After the smelting of the previous heat is completed, obtain the iron, scrap steel information, auxiliary material information, and end-point target information of this heat.

[0047] S2. Obtain the steel material consumption, slag amount unit consumption, and converter-related parameters according to historical data, and calculate the target molten steel amount, final slag amount, and retained slag amount of the previous heat of this heat.

[0048] S3. Calculate the amount of magnesium-containing auxiliary material added according to parameters such as the slag amount and the saturated magnesium oxide content in the slag.

[0049] S4. Calculate the amount of lime added according to the phosphorus distribution ratio.

[0050] S5. Calculate the amount of coolant (heating agent) added according to the heat balance.

[0051] S6. Calculate the oxygen blowing amount according to the oxygen balance.

[0052] S7. After the smelting of this heat is completed, compare the model prediction value with the production actual value, and the data that meets the conditions is included in the reference heat for model self-learning.

[0053] Preferably, in the converter control method based on a static model, the calculation of the retained slag amount of the previous heat in S2 is obtained by fitting the relationship between the volume of the slag in the furnace at the end of the slag tapping of the converter and the tilting angle of the converter, and obtaining

[0054] V ReSlag =7×10 11 ×e -0.275x

[0055] In the formula, x is the tilting angle at the end of the slag tapping of the converter in the previous heat.

[0056] Calculate the density from the slag composition

[0057]

[0058] Obtain the slag retention amount of the upper furnace

[0059] W ReSlag = ρ ReSlag ×V ReSlag

[0060] Preferably, in the converter control method based on a static model, it is characterized in that: in S3, the saturated magnesia is calculated

[0061]

[0062] In the formula, (%TFe) is the TFe content in the slag of the previous heat, (%CaO), (%SiO2) are the CaO and SiO2 components in the slag of the previous heat, and T AimSteel is the target temperature at the end point

[0063] Preferably, in the converter control method based on a static model, it is characterized in that: in S4, the CaO addition amount is calculated through the phosphorus distribution ratio

[0064] L P，理论 ≥ k·L P，目标

[0065] In the formula, L P，目标 is the target phosphorus distribution ratio; L P，理论 is the theoretical phosphorus distribution ratio; k is a correction coefficient, k≥1, obtained by regression from historical heats

[0066] Calculation of the target phosphorus distribution ratio

[0067]

[0068] In the formula, (%P2O5) is the P2O5 content in the slag of the previous heat; [P] 目标 : Set according to the steelmaking process regulations of the steel grade

[0069] Calculation of the theoretical phosphorus distribution ratio

[0070]

[0071] In the formula, (%MnO) is the MnO content in the slag of the previous heat; T AimSteel is the target temperature of the molten steel at the end point of this heat; [%C] is the target C content at the end point of the converter for this heat; B0, B1, B2, B3, B4 are obtained by regression from the classified reference heats

[0072] Preferably, for the converter control method based on a static model, it is characterized in that: in S5, by comparing the heat input-output differences of the reference heats, the heat surplus is calculated using heat balance, and then the addition amount of the coolant (heating agent) is calculated.

[0073] Preferably, for the converter control method based on a static model, it is characterized in that: in S6, by comprehensively considering factors such as element burnout, fume consumption, furnace gas input, and decomposition of the coolant (heating agent), the total oxygen blowing amount is calculated.

[0074] V O2 = a(∑V OSS ) + bV Oyc + cV Olq + dV Oks

[0075] In the formula, ∑V OSS is the oxygen amount required for the burnout of each element, V Oyc is the oxygen consumption of the fume, V Olq is the free oxygen amount in the furnace gas, V Oks is the oxygen supply from the decomposition of the coolant (heating agent). a, b, c, and d are correction coefficients obtained by regression of the classified reference heats.

[0076] Preferably, for the converter control method based on a static model, it is characterized in that: in S4 - S7, the model prediction value is compared with the production actual value. If the corresponding parameter of the production actual value meets the conditions, then this heat is included in the reference heats. The valid data range of the heat includes 60t < hot metal weight ≤ 75t, 7t < scrap weight ≤ 25t, 3.7% < hot metal C content ≤ 4.8%, 0.1% < hot metal Si content ≤ 0.95%, 0.15% < hot metal Mn content ≤ 1%, 0.09% < hot metal P content ≤ 0.16%, 1220°C < hot metal temperature ≤ 1470°C.

[0077] Preferably, for the converter control method based on a static model, it is characterized in that: for the reference heats in S4 - S7, the selection conditions are the production actuals of the nearest 100 heats, and the heats similar to this heat, including the hot metal weight fluctuation range of ±5t, the hot metal temperature fluctuation range of ±20°C, the scrap weight fluctuation range of ±4t, the hot metal C fluctuation range of ±0.2%, and the hot metal Si fluctuation range of ±0.1%.

[0078] Preferably, in the converter control method based on a static model, it is characterized in that in S7, the model prediction value is compared with the production actual value, and the data that meets the conditions is included in the reference furnace for model self-learning. The specific method includes: comparing the parameters B0, B1, B2, B3, B4 selected by the model in S4 - S6 for calculating the lime addition amount, the parameters a, b, c, d for calculating the oxygen blowing amount, and the relevant parameters for calculating the coolant (heating agent) with the parameters obtained by back-calculation by substituting the production actual value into the calculation formula. The values with similar parameters are stored for model calculation and call. If the relative deviation of the parameters is less than or equal to 5%, it is directly stored and called; if the relative deviation is greater than 5%, it is stored and called after parameter compensation according to the deviation amount.

[0079] The results show that after adopting the static control method of this embodiment, the hit rates of indicators such as oxygen blowing amount, lime, and magnesium-containing auxiliary materials are greater than 80%, which can better guide the converter production and provide support for the subsequent unmanned operation of the converter.

[0080] Example 1: Dynamically adjust the addition amount of magnesium-containing auxiliary materials based on the saturated content of magnesium oxide in the slag.

[0081] (1) Data collection: After the smelting of the previous furnace is completed, obtain the molten iron composition (C, Si, Mn, P content), scrap weight, and end-point target temperature T of this furnace AimSteel , as well as the slag composition (%TFe, %CaO, %SiO2) of the previous furnace.

[0082] (2) Saturated magnesium oxide calculation: Substitute into the formula:

[0083]

[0084] Calculate the saturated content of magnesium oxide in the slag.

[0085] (3) Determination of the addition amount of magnesium-containing auxiliary materials: According to the target slag amount W 渣 and the saturated magnesium oxide content, calculate the addition amount of magnesium-containing auxiliary materials (such as magnesium balls):

[0086]

[0087] Among them, (%MgO) 辅料 is the mass percentage of MgO in the magnesium-containing auxiliary material.

[0088] Add the magnesium-containing auxiliary materials according to the calculation result. After the smelting is completed, detect the actual MgO content in the slag. If the deviation ≤ 3%, include the data of this furnace in the reference library for model self-learning.

[0089] By dynamically calculating the saturated content of magnesium oxide, it avoids over-addition or insufficiency caused by traditional fixed parameters, and the hit rate of magnesium-containing auxiliary materials is increased to 80%.

[0090] Example 2: Optimize the lime addition amount by dynamically matching the theoretical phosphorus distribution ratio with the target value.

[0091] (1) Input the target phosphorus content [P] at the end of this heat, 目标 the P2O5 content (%) in the slag of the previous heat, 历史 the slag composition (%MnO, %CaO) and the target temperature T of the molten steel, AimSteel and the carbon content [%C].

[0092] (2) Calculate the target phosphorus distribution ratio:

[0093]

[0094] (3) Calculate the theoretical phosphorus distribution ratio:

[0095]

[0096] Among them, B0 to B4 are dynamic parameters updated by the self-learning module.

[0097] (4) Calculate the lime amount: By iteratively adjusting the lime addition amount to make L P,理论 ≥ k·L P,目标 , where k = 1.2, and finally determine the lime addition amount.

[0098] (5) Result feedback: After smelting, detect the actual phosphorus content. If it hits (deviation ≤ 0.005%), update the B0 to B4 parameters; otherwise, mark it as abnormal data and trigger manual review.

[0099] The calculation error of the lime addition amount is controlled within ±10%, and the hit rate of the lime addition amount is increased to 90%.

[0100] Example 3: Calculate the total oxygen blowing amount by comprehensively considering dynamic factors such as element burn-off and oxygen consumption of dust.

[0101] (1) Obtain the hot metal composition (C, Si, Mn, P), scrap steel composition, coolant type and addition amount.

[0102] (2) Calculate the oxygen demand:

[0103] Oxygen consumption for element burn-off ∑V OSS : Calculate according to the oxidation reaction equations of each element;

[0104] Oxygen consumption of dust V Oyc : Calculate by multiplying the dust generation amount × FeO content × oxygen conversion coefficient;

[0105] Free oxygen in furnace gas V Olq : Obtain according to the real-time data of the furnace gas composition analyzer;

[0106] Oxygen supply by coolant V Oks: Calculated according to the decomposition reaction of the coolant (such as ore).

[0107] Calculation of the total oxygen blowing volume:

[0108] V O2 = a∑V OSS + bV Oyc + cV Olq + dV Oks

[0109] Among them, a to d are dynamic correction coefficients provided by the self-learning module.

[0110] (3) Supply oxygen according to the calculation result. After the smelting is completed, compare the actual oxygen consumption. If the deviation > 5%, trigger the self-learning module to correct a to d.

[0111] The calculation accuracy of the oxygen blowing volume is controlled within ±5%, and the hit rate of the oxygen blowing volume is increased to 90%.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the purpose and scope of the present technical solution, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A converter control method based on a static model, characterized in that: The following steps are involved: S1. Obtain the hot metal information, scrap steel information, auxiliary material information and end point target information of this furnace; S2. Calculate the target molten steel volume, final slag volume and slag remaining in the furnace according to historical data; S3. Calculate the amount of magnesium-containing auxiliary material based on the amount of slag and the saturated content of magnesium oxide in the slag; S4. Calculate the amount of lime to be added according to the phosphorus distribution ratio; S5. Calculate the amount of coolant or exothermic agent added based on heat balance; S6. Calculate the total oxygen blowing amount through oxygen balance; S7. Compare the model prediction value with the actual production performance value, and select qualified data for model self-learning.

2. The converter control method based on static model according to claim 1, characterized in that: The calculation of the slag amount in the upper furnace in S2 includes: Fitting the slag volume V according to the tilting angle x at the end of converter slag tapping ReSlag , the formula is V ReSlag =7×10 11 ×e -0.275x , and calculate the density ρ through the slag composition ReSlag , and finally the slag amount W is obtained ReSlag =ρ ReSlag ×V ReSlag .

3. The converter control method based on static model according to claim 1, characterized in that: The calculation formula for the saturated content of magnesium oxide in the slag in S3 is: Among them, T AimSteel is the target temperature at the end of this furnace, and %TFe, %CaO, and %SiO2 are the corresponding component contents in the slag of the previous furnace.

4. The converter control method based on static model according to claim 1, characterized in that: In S4, the target phosphorus distribution ratio L is used. P，目标 Compared with the theoretical phosphorus distribution ratio L P , 理论 The amount of lime added is determined by the relationship between And satisfy L p , 理论 ≥kL P，目标 , where k is the correction coefficient, k≥1, and B0~B4 are the reference heat regression parameters.

5. The converter control method based on static model according to claim 1, characterized in that: The total oxygen blowing amount V in S6 O2 The calculation formula is: V o2 =a(∑V OSs )+bV oyc +cV Olq +dV Oks Among them, ∑V OSS is the amount of oxygen required for burning of each element, V Oyc is the smoke oxygen consumption, V Olq is the free oxygen content in the furnace gas, V Oks is the oxygen supply for the decomposition of the coolant or exothermic agent; a~d are correction coefficients obtained by regression of the classified reference furnaces.

6. The converter control method based on static model according to claim 1, characterized in that: In S7, the conditions for screening the reference heats include: The weight range of molten iron is 60t~75t, the weight range of scrap steel is 7t~25t, the C content of molten iron is 3.7%~4.8%, the Si content of molten iron is 0.1%~0.8%, and the temperature of molten iron is 1220℃~1470℃; and the parameter fluctuation range of this furnace meets the requirements of molten iron weight ±5t, molten iron temperature ±20℃, and scrap steel weight ±4t; the fluctuation range of molten iron C is ±0.2%, and the fluctuation range of molten iron Si is ±0.1%.

7. The converter control method based on static model according to claim 1, characterized in that: In S7, the specific method of model self-learning includes: Compare the model parameters B0~B4, a~d in S4~S6 with the reverse parameters of production performance. If the relative deviation of the parameters is less than or equal to 5%, they are directly stored and called; if the relative deviation is greater than 5%, the parameters are compensated according to the deviation and then stored and called.

8. The converter control method based on static model according to claim 1, characterized in that: The calculation of the surplus heat in S5 is based on the difference in heat input and output of the reference heat, and the amount of coolant or heat-generating agent added is determined through heat balance.

9. The converter control method based on static model according to claim 1, characterized in that: The data for the self-learning of the model in S7 is selected from the production results of the last 100 furnaces with conditions similar to the current furnace.

10. The converter control method based on static model according to claim 7, characterized in that: The specific method of parameter compensation is: If the relative deviation of the parameter is greater than 5%, the compensated parameter value P 新 Calculate as follows: Among them, P 模型 is the model prediction parameter value, ΔP=P 实绩 -P 模型 , P 实绩 Reverse the parameter values ​​based on actual production performance.

Images