CN-122024922-A - Method for predicting residual carbon content of molten pool of steelmaking converter

CN122024922ACN 122024922 ACN122024922 ACN 122024922ACN-122024922-A

Abstract

The invention relates to a method for predicting residual carbon content of a molten pool of a steelmaking converter, which comprises the steps of firstly calculating theoretical oxygen consumption, obtaining actual oxygen consumption, setting a decarburization stage dynamic dividing mechanism, then respectively generating corresponding dynamic correction coefficients K according to the difference of decarburization modes of process flows of TSC and TSO at different stages of a smelting period, and finally carrying out dynamic correction optimization on the K value by using an optimization iterative algorithm and predicting and applying the K value in real time. According to the invention, through real-time monitoring of the component parameters of the flue gas at the mouth of the converter at the end of converting and combining a carbon-oxygen reaction mechanism model and a machine learning algorithm, the non-contact continuous inversion of the global carbon content of the molten pool is realized. The method is suitable for accurately controlling the carbon content of molten steel in a converter smelting link, can replace or assist the traditional sublance spot measurement technology, remarkably improves the carbon hit rate of the end point, reduces the over-blowing/under-blowing risk, and provides powerful support for smelting accurate control.

Inventors

GUO FAN
FAN JIA
WU ZHIJIE
HE FANG
DING JIAN
SU ZHENJUN
WU WEI

Assignees

邯郸钢铁集团有限责任公司
河钢股份有限公司邯郸分公司

Dates

Publication Date: 20260512
Application Date: 20260114

Claims (6)

1. A method for predicting residual carbon content of a molten pool of a steelmaking converter is characterized by comprising the following steps: S1, calculating theoretical oxygen consumption, acquiring actual oxygen consumption, and setting a decarburization stage dynamic dividing mechanism; s2, respectively generating corresponding dynamic correction coefficients K according to the difference of decarburization modes of the process flows of the TSC and the TSO at different stages of the smelting period; And S3, carrying out dynamic correction optimization on the K value by using an optimization iterative algorithm, and predicting and applying the K value in real time.
2. The method for predicting the residual carbon content of a molten pool of a steelmaking converter as set forth in claim 1, wherein in step S1, the method specifically includes the following steps: S11, setting the current accumulated actual oxygen consumption as Qactual O 2 , setting the theoretical oxygen consumption as Qtheory O 2 , and calculating the theoretical oxygen consumption Qtheory O 2 by using the initial carbon content [ C ] 0 of molten iron, the weight W HM of molten iron, the weight W FG of scrap steel and the carbon contribution delta C scrap of the scrap steel through material balance MB and heat balance HB: Qtheory O 2 =f([C] 0 , W HM , W FG ,ΔC scrap , molten iron temperature T 0 , and molten iron composition S12, defining a decarburization stage in the smelting process according to a ratio eta of the actual oxygen consumption to the theoretical oxygen consumption: ; According to the method, when eta <30% of smelting in the earlier stage is started, silicon/manganese oxidation is taken as a main material, at the moment, the decarburization reaction is weaker, and the carbon reduction rate is lower but gradually increased; When eta is more than or equal to 30 percent and less than or equal to 80 percent in the middle smelting period, the decarburization reaction is taken as the leading part, and the rapid carbon reduction rate of the reaction is the main stage of the decarburization reaction; When eta >80%, the decarburization period is already a deceleration decarburization period, and the blowing stopping time is controlled by slow carbon descent.
3. The method for predicting the residual carbon content of a molten pool of a steelmaking converter according to claim 2, wherein the ratio eta= (30% and 80%) can be dynamically adjusted by +/-5% to +/-10% according to actual requirements so as to define the time range of a main decarburization period.
4. The method for predicting the residual carbon content of a molten pool of a steelmaking converter as claimed in claim 2, wherein in step S2, the method specifically comprises the following steps: S21, the phase of the smelting period is TSC, and a dynamic correction coefficient K TSC is constructed, the phase of the smelting period is TSO, and the treatment process for constructing the dynamic correction coefficients K TSO ,K TSC and K TSO is the same; S22, taking K TSC as an example, the calculation optimizing logic of K TSC is as follows: The time series data of the historical heat is obtained, namely flue gas flow Q g (t), carbon monoxide concentration C CO (t), carbon dioxide concentration C CO2 (t), oxygen flow F O2 (t) and static parameter molten iron, and the initial carbon content [ C ] 0 of the molten pool can be calculated by combining the basic information with static parameter material conservation calculation: [C] 0 =((W HM ·[C] 0 ) + (W FG ·[C] FG )) ·10 -2 ·10 6 In the whole smelting process of one furnace, the time sequence data of the carbon monoxide concentration C CO (t) and the carbon dioxide concentration C CO2 (t) are combined, and the accumulated decarburization quantity delta C decarb under each sampling frequency can be obtained from the beginning to the end of smelting through mass conservation calculation: ; At the time of the sublance under the TSC, a decarburization accumulation amount DeltaC decarb can be obtained in the above manner, wherein DeltaC decarb represents the mass of carbon contained in carbon monoxide and carbon dioxide in flue gas in the converter smelting process, the unit is ton, and the residual carbon content [ C pred of the molten pool can be calculated by the difference between the initial carbon content [ C ] 0 of the molten pool and the total decarburization amount: ; Here, the The calculated content is a percentage, which indicates that the carbon removed accounts for the carbon content in the whole molten steel; the calculation at this time is based on mass conservation completely, but there are hysteresis and nonlinear interference factors of the data in actual production, so a dynamic correction coefficient K is introduced to balance the data anomaly caused by the error: ; Through iterative calculation of historical time sequence data, a dichotomy is adopted to solve a dynamic correction coefficient K, so that the balanced residual carbon content [ C ] K pred of a molten pool can be within +/-5% of the real carbon content [ C ] target in the molten pool measured by a sublance, and the dynamic correction coefficient K is the optimal dynamic correction coefficient K of the current smelting heat: ; The dynamic correction factor K calculated at this stage is referred to herein as K TSC , since it occurs during the TSC stage of the smelting cycle.
5. The method for predicting the residual carbon content of a molten pool of a steel-making converter as set forth in claim 4, wherein in step S22, the static parameters molten iron include an initial carbon content [ C ] 0 , a molten iron weight W HM , a scrap weight W FG , and a scrap integrated carbon content [ C ] FG .
6. The method for predicting the residual carbon content of a molten pool of a steelmaking converter according to claim 1, wherein in the step S3, a K value dynamic correction optimization and real-time prediction mechanism is specifically: s31, when converting is started, collecting real-time measurement data acquired from a flue gas analyzer, and closely focusing on the change of an oxygen blowing stage by combining a decarburization stage dividing method; S32, when the oxygen blowing progress is 75%, starting an online optimizing program, extracting a characteristic vector value, searching an optimal value through Euclidean distance matching, and optimizing a dynamic correction coefficient K; Starting an online optimizing program when eta=75%, extracting curve characteristics in the main decarburization time range of the current heat, carrying out sectional fitting on a DeltaC decarb -t curve, extracting a slope beta and an intercept DeltaC 0 , and calculating Euclidean distance between a current curve characteristic vector V current =[β,ΔC 0 and an ith heat V i in a historical characteristic library: ; K opt corresponding to the heat of D min is selected as a dynamic correction coefficient of the heat, and D min represents the minimum Euclidean distance, wherein K opt is a correction coefficient for compensating or correcting the system measurement error and the process dynamic control error of the online detection equipment in the normal smelting process; When the oxygen blowing amount eta is more than or equal to 80 percent, starting to correct and predict the carbon content of TSC at the end of decarburization, and calibrating the accumulated decarburization amount C' decard in eta <80 percent stage by using K opt : C'decard=K opt ·ΔC decarb at this point, the corrected bath instantaneous carbon content [ C ] t can be calculated: ; wherein, the initial carbon content of molten iron [ C ] 0 , the weight of molten iron W HM and the weight of scrap steel W FG ; fitting the carbon drop curve of the carbon content in the main decarburization period, and predicting the carbon content track of the future delta t seconds by taking the current d [ C ]/dt as a slope: 。

Description

Method for predicting residual carbon content of molten pool of steelmaking converter Technical Field The patent application belongs to the technical field of intelligent detection in a ferrous metallurgy process, and particularly relates to a method for predicting residual carbon content of a molten pool of a steelmaking converter. Background In the converter smelting process of steel smelting, the carbon content of a terminal molten pool is a key technological parameter affecting molten steel quality, alloy yield, furnace lining service life and product qualification rate. The accurate hit of the carbon content of the end point is realized, and the method has decisive effects on reducing the production cost, improving the production efficiency, ensuring the purity of molten steel and meeting the performance index of high-end steel. At present, the control of the carbon content of the smelting end point of the converter mainly comprises sublance detection technology, bullet measurement and a flue gas analysis method. Mechanical probe insertion measurement (sublance technology) is used as a common measurement method, and the probe is directly inserted into a molten pool to obtain point measurement data such as temperature, carbon content and the like. This has the advantage that the measurement is intuitive and the chemical composition at a specific location of the bath can be read directly. However, the technology has obvious space-time limitation, and single measurement only reflects transient local data and cannot capture the dynamic change of the whole molten pool. The operation level needs to be strictly dependent on experience to judge the inserting time, the early or late end hit rate can be reduced, and the frequent contact of the probe with molten steel can accelerate equipment loss. The bullet throwing probe is used for measuring that the detection bullet is thrown into a molten pool through the automatic throwing device, the data such as temperature, carbon oxygen content and the like can be obtained within a few seconds, and the effective measurement success rate can reach more than 95%. Compared with a mechanical probe, the device has the advantages that the device does not need to pour into a furnace to stop blowing or reduce oxygen blowing flow, and shortens smelting period. But still belongs to spot measurement technology in essence, has similar limitation as a sublance, data still represent transient state, and the warhead needs to be completely immersed into a molten pool to obtain accurate reading, so that measurement stability can be influenced when the turbulence of the molten pool is severe, and meanwhile, the warhead belongs to a disposable consumable, and single measurement cost is increased. In recent years, the requirements of the molten steel smelting industry on high real-time performance, continuous stability and low cost operation of the production process are increasingly urgent. Under the background, the smoke analysis method capable of being accurately adapted to the requirements is continuously improved in application popularity, and a series of optimized application schemes are also provided by related technical researches. The key logic of the schemes is that a flue gas analyzer is deployed at a key position of a converter flue to monitor the components of the furnace gas in real time and continuously so as to reversely deduce the dynamic reaction state in the molten pool. As a typical non-contact measurement technology, the flue gas analysis method can accurately capture core technological parameters such as the CO/CO 2 proportion and the like in the whole process, provides key technical support for realizing continuous monitoring of the whole flow of molten steel smelting, effectively balances production efficiency and cost control requirements in the power-assisted industry, but has a certain technical defect in the actual application scene: The Chinese patent (application number: CN 202411272108.6) discloses a steelmaking decarburization end point judging method based on flue gas component calculation, which is characterized in that in the steelmaking process, the volume fraction mole number of CO 2 and CO in the converter gas is monitored in real time, and the carbon content of molten steel is dynamically calculated and predicted by utilizing the linear function relation between the data and the carbon content value of the molten steel, so as to judge the decarburization end point. The core logic of the method only depends on the principle of conservation of mass in a theoretical layer, namely the total carbon molar mass of carbon-containing gases (mainly CO and CO 2) in the flue gas at a unit moment is equal to the carbon molar mass reduced in the synchronous molten steel, but the flow of the flue gas in actual production is not considered to be subject to various problems such as fluctuation of various process conditions, equipment acquisition errors, detection tim