A low-energy smelting method for high-alloy building steel

Abstract

This invention belongs to the field of steel metallurgical process innovation. It provides a low-energy smelting method for high-alloy building steel, comprising: collecting and standardizing historical data; constructing a standard set and time series; using an LSTM model to capture time series correlation coefficients; constructing an intelligent control system to output the deoxidation and slag-forming argon flow rate; monitoring the uniformity of steel composition through a LIBS sensor array and calculating adjustment coefficients to determine whether adjustment is needed; if so, fitting a correlation model to extract the flow composition influence coefficient for adjustment, and simultaneously calculating the flow temperature influence coefficient. By analyzing the positive and negative consistency of the flow composition influence coefficient and the flow temperature influence coefficient, the temperature argon flow rate value is calculated. Through multi-stage coordinated control, precise optimization of argon flow rate is achieved, realizing coordinated uniformity of composition and temperature with minimal or no power supply, thereby improving energy utilization efficiency.

Description

Technical Field

[0001] This invention belongs to the field of innovation in iron and steel metallurgical processes, specifically a low-energy smelting method for high-alloy building steel. Background Technology

[0002] With the development of high alloy building steel smelting, how to achieve low energy consumption smelting based on high alloy building steel has become a research hotspot in the field of iron and steel metallurgical process innovation. However, there are still significant technical bottlenecks in the LF refining process. Therefore, it is of great significance to develop a low energy consumption smelting method for high alloy building steel.

[0003] In the refining process of LF (Ladle Refining Furnace), although the state of molten steel can be adjusted by the argon flow rate, on the one hand, the flow rate is set based only on static data at a single point in time, and the argon flow rate value is set by fixed flow rate threshold or manual experience. There is a lack of systematic analysis of historical data of deoxidation and slag formation, which makes it difficult to accurately match the actual working conditions with the argon flow rate setting. This can easily cause fluctuations in the oxidizability of molten steel or uneven slag composition. For example, when a medium-sized steel company was producing 20CrMnTi gear steel, the argon flow rate was always 0. This resulted in insufficient flow rate during the deoxidation stage, making it difficult for inclusions to float. Furthermore, in the later stage of slag formation, the excessive flow rate caused slag to be entangled. During the inspection, all products in this batch were unqualified.

[0004] On the other hand, the lack of consideration for the influence between the introduced argon gas and the composition and temperature of the molten steel, and the lack of feasibility analysis on whether the influence of argon gas on the composition and temperature of the molten steel can simultaneously meet the requirements of both, can easily lead to a situation where when one is adjusted, the other also changes accordingly, and it is always impossible to achieve the simultaneous satisfaction of the working requirements of both, causing control contradictions and frequent work stoppages.

[0005] Therefore, the present invention provides a low-energy smelting method for high-alloy building steel. Summary of the Invention

[0006] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0007] The technical solution adopted by this invention to solve its technical problem is: a low-energy smelting method for high-alloy building steel, comprising:

[0008] Historical data of the deoxidation and slag-forming process is obtained, the temporal correlation coefficient between the standardized values ​​of the historical data is calculated, and an intelligent argon control system is constructed to output the argon flow rate value for deoxidation and slag-forming.

[0009] Argon gas is input according to the deoxidation slag formation argon gas flow rate value for deoxidation slag formation. During the deoxidation slag formation process, it is determined whether the deoxidation slag formation argon gas flow rate value needs to be adjusted. If so, the flow component influence coefficient between the argon gas flow rate value and the component uniformity analysis value is derived, and the component argon gas flow rate value is calculated.

[0010] The argon flow rate for deoxidation and slag formation was initially adjusted based on the argon flow rate of the composition, and the temperature uniformity analysis value was calculated in real time after the adjustment. The flow-temperature influence coefficient between the argon flow rate and the temperature uniformity analysis value was then derived.

[0011] Determine the consistency between the positive and negative values ​​of the flow component influence coefficient and the flow temperature influence coefficient. If they are consistent, calculate the temperature argon flow rate value and make a secondary adjustment to the deoxidation slag formation argon flow rate value according to the temperature argon flow rate value.

[0012] Furthermore, the process of calculating the time-series correlation coefficient between the standardized values ​​of historical data is as follows:

[0013] Historical data on the deoxidation and slag-forming process were obtained, including the amount of deoxidizer, the amount of slag-forming agent, the argon flow rate, the initial state of molten steel oxidation, and the amount of slag conversion. The historical data on the deoxidation and slag-forming process were then standardized.

[0014] The standardized processing results are divided and selected to obtain several 3-step time series sequences;

[0015] Several 3-step time series sequences are input into the LSTM model. The hidden layer captures the temporal correlation between each standardized value and outputs the temporal correlation coefficient between the standardized values ​​of historical data.

[0016] Furthermore, the process of dividing and selecting the standardized results to obtain several 3-step time series sequences is as follows:

[0017] Standardize the historical data of the deoxidation and slag-making process.

[0018] Obtain the standardized value of deoxidizer dosage Standardized values ​​of slag-forming agent dosage Standardized value of initial state of oxidative properties of molten steel Standardized value of slag conversion Standardized value of argon flow rate And the five standardized values ​​within the same time step are recorded as a standard set;

[0019] Obtain a standard set of historical data within several time steps. Group three consecutive time steps together to form a 3×5 time series matrix. This 3×5 time series matrix is ​​a 3-step time series sequence. Perform several selections to obtain several 3-step time series sequences.

[0020] Furthermore, the process of constructing an intelligent argon control system and outputting the argon flow rate value for deoxidation and slag formation is as follows:

[0021] Input layer: The standardized values ​​of deoxidizer dosage, slag-forming agent dosage, initial oxidative properties of molten steel, and slag conversion amount within the same step length in historical data are used as input;

[0022] Hidden layer: The weight matrix between the standardized values ​​output by the LSTM model and the argon flow rate values ​​within the same step size in historical data;

[0023] Output layer: Outputs the predicted argon flow rate for deoxidation and slag formation;

[0024] Obtain the current deoxidizer and slag-forming agent dosages in the current deoxidation and slag-forming process, and calculate the corresponding standardized values ​​for deoxidizer and slag-forming agent dosages.

[0025] The standardized values ​​for the initial state of molten steel oxidation and the standardized value for slag conversion are set to 1.

[0026] The current standardized values ​​of deoxidizer dosage, slag-forming agent dosage, initial oxidative state of molten steel, and slag conversion amount are input into the argon intelligent control model, which outputs the deoxidation and slag-forming argon flow rate.

[0027] Furthermore, the process for determining whether the argon flow rate for deoxidation and slag formation needs to be adjusted is as follows:

[0028] The molten steel after the addition of materials is divided into several unit regions, and the different elements in the unit regions are classified according to category. The composition uniformity analysis value of each unit region is calculated. If the composition uniformity analysis value of any unit region is greater than the composition uniformity analysis threshold, the element composition of that unit region is determined to be non-uniform.

[0029] The number of elements with uneven composition in a given cell region is obtained, and its ratio to the total number of cells in the region is used to calculate an adjustment coefficient. This coefficient is then compared with an adjustment coefficient threshold.

[0030] If the adjustment coefficient is greater than or equal to the adjustment coefficient threshold, it is determined that the argon flow rate value for deoxidation and slag formation needs to be adjusted.

[0031] Conversely, if the result is not satisfactory, then no adjustment is required.

[0032] Furthermore, the process of calculating the compositional uniformity analysis values ​​for each unit region is as follows:

[0033] The standard deviation of different element sets within the cell region is processed, and the average of the standard deviations of all element sets is calculated, which is the component homogeneity analysis value.

[0034] Furthermore, the process of deriving the flow component influence coefficient between the argon flow rate value and the component uniformity analysis value is as follows:

[0035] The process of comparing the calculated coefficient of determination with the threshold of the coefficient of determination, and obtaining the influence coefficient of the flow component based on the comparison result, is as follows:

[0036] If the coefficient of determination is greater than or equal to the threshold of the coefficient of determination, it indicates that there is a linear correlation between the standardized value of argon flow rate and the corresponding homogeneous analysis value of composition.

[0037] Conversely, it indicates that there is no linear relationship.

[0038] If it is a linear correlation, the standardized value coefficient of argon flow rate in the univariate linear regression model formula is the influence coefficient of flow component;

[0039] If the correlation is nonlinear, construct multiple models and calculate the determination coefficient for each model.

[0040] The nonlinear model corresponding to the maximum value of the coefficient of determination is taken as the nonlinear model between the standardized value of argon flow rate and the standardized value of the corresponding component homogeneity analysis;

[0041] Based on the characteristic values ​​of the current standardized argon flow rate of the model, obtain the influence coefficient of the flow rate components;

[0042] The derivation process of the flow rate temperature influence coefficient is the same as that of the flow rate temperature influence coefficient.

[0043] Furthermore, the process of obtaining the influence coefficient of the flow components based on the comparison results is as follows:

[0044] If the coefficient of determination is greater than or equal to the threshold of the coefficient of determination, it indicates that there is a linear correlation between the standardized value of argon flow rate and the corresponding homogeneous analysis value of composition.

[0045] Conversely, it indicates that there is no linear relationship.

[0046] If it is a linear correlation, then the standardized value coefficient of argon flow rate in the univariate linear regression model formula is the influence coefficient of flow rate components;

[0047] If the correlation is nonlinear, construct a binary linear model, a higher-order polynomial model, an exponential model, a power function model, and a trigonometric function model respectively, and calculate the determination coefficient of each model.

[0048] The nonlinear model corresponding to the maximum value of the coefficient of determination is taken as the nonlinear model between the standardized value of argon flow rate and the standardized value of the corresponding component homogeneity analysis;

[0049] Based on the characteristic values ​​of the current standardized argon flow rate of the model, the influence coefficients of the flow rate components are obtained.

[0050] Furthermore, the process of determining the consistency between the positive and negative values ​​of the flow component influence coefficient and the flow temperature influence coefficient is as follows:

[0051] If the influence coefficients of flow composition and flow temperature are both positive or both negative, it indicates that their positive and negative values ​​are consistent; otherwise, they are inconsistent.

[0052] Furthermore, the process of adjusting the argon flow rate for deoxidation and slag formation based on the temperature and argon flow rate is as follows:

[0053] If the influence coefficients of flow composition and flow temperature are of the same sign, then obtain the temperature uniformity analysis values ​​of all units and perform averaging to obtain the average value of the temperature uniformity analysis.

[0054] The temperature uniformity analysis value and the average temperature uniformity analysis value of each unit are respectively processed by difference, and the maximum value is taken as the component uniformity adjustment value.

[0055] The component uniformity adjustment value is compared with the flow rate temperature influence coefficient to obtain the component flow rate adjustment value, and then added to the component argon flow rate value to obtain the temperature argon flow rate value.

[0056] The beneficial effects of this invention are as follows:

[0057] Effect 1: Improve the accuracy and dynamic adaptability of argon flow control. By constructing a time-series correlation model (LSTM network) of historical data in the deoxidation and slag-forming process, the dynamic relationship between parameters such as deoxidizer dosage, slag-forming agent dosage, and molten steel oxidizability and argon flow is quantified, achieving precise matching between argon flow and process parameters, effectively reducing the fluctuation of molten steel oxidizability and the problem of uneven slag composition, and improving refining stability.

[0058] Effect 2: Under conditions of low or no power supply, the argon flow rate is dynamically adjusted to balance composition and temperature uniformity. By dynamically deducing the composition and temperature coefficients controlled by the unit argon flow rate, the precise control of composition and temperature uniformity by the argon flow rate is achieved, improving the uniformity control accuracy, avoiding over-stirring or under-stirring, and significantly improving the uniformity of molten steel.

[0059] Effect 3: The dynamic flow adjustment mechanism based on real-time operating conditions reduces unnecessary excessive or insufficient use of argon, lowers energy consumption, and at the same time, preventive adjustments reduce rework caused by substandard quality, significantly shortening the refining cycle and reducing production costs. Attached Figure Description

[0060] The invention will now be further described with reference to the accompanying drawings.

[0061] Figure 1 This is a schematic diagram of the steps in a low-energy smelting method for high-alloy building steel according to an embodiment of the present invention.

[0062] Figure 2 This is a schematic diagram of the logic judgment of a low-energy smelting method for high-alloy building steel according to an embodiment of the present invention. Detailed Implementation

[0063] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0064] Example 1: Please refer to Figure 1 - Figure 2 As shown in the embodiment of the present invention, a low-energy smelting method for high-alloy building steel includes the following steps:

[0065] Step 1: During the LF refining process, acquire historical data of the deoxidation and slag-forming process, calculate the time-series correlation coefficient between the standardized values ​​of the historical data, construct an intelligent argon control model, and output the argon flow rate value for deoxidation and slag-forming.

[0066] It should be noted that the refining process is divided into three stages: deoxidation and slag formation, argon fine-tuning, and stirring control. The process begins after the previous stage is completed. The argon fine-tuning stage involves fine-tuning the uniformity of the molten steel composition and then fine-tuning the uniformity of the molten steel temperature.

[0067] In step one, the process of calculating the time-series correlation coefficient between the standardized values ​​of historical data is as follows:

[0068] Obtain historical data on the deoxidation and slag-forming process;

[0069] Historical data for the deoxidation and slag-forming process includes: deoxidizer dosage, slag-forming agent dosage, argon flow rate, initial oxidizability of molten steel, and slag conversion rate.

[0070] The amount of deoxidizer used at any time step in the historical data is standardized to obtain the standardized value of the deoxidizer usage. :

[0071]

[0072] The standard range for deoxidizer dosage is the minimum value of all historical deoxidizer dosage data. With the maximum value The time step is set by experts based on the actual situation;

[0073] Similarly, calculate the standardized value of slagging agent dosage within any time step in the historical data. Standardized value of initial state of oxidative properties of molten steel Standardized value of slag conversion Standardized value of argon flow rate ;

[0074] Within the same time step , , , , , denoted as a standard set ( , , , , );

[0075] For example, the process of obtaining the standard set is as follows:

[0076] Obtain the amount of aluminum particles added to the molten steel per ton of molten steel within a time step (0s-15s). The standard range for deoxidizer dosage is 0.5 kg / t. The value is (0.2, 1.0) kg / t. Substituting this into the formula, we obtain the standardized value of the deoxidizer dosage. The value is 0.375. Similarly, calculate... It is 0.467. It is 0.556. It is 0.779. If the value is 0.837, then the standard set is (0.375, 0.367, 0.456, 0.779, 0.837).

[0077] Obtain a standard set of historical data over several time steps;

[0078] Based on a standard set of historical data within several time steps, the deoxidation and slagging process of the same LF refining of molten steel is divided into three time steps to obtain a three-step time sequence, which is then obtained several times.

[0079] For example, the process of constructing a 3-step time series sequence is as follows:

[0080] Set the time step to 15 seconds and obtain the standard set of historical data within three adjacent time steps.

[0081] Time step 1 (0-15 seconds) ,

[0082] Time step 2 (15-30 seconds) ,

[0083] Time step 3 (30-45 seconds) ,

[0084] A 3×5 time series matrix is ​​formed, which is a 3-step time series sequence;

[0085] Input layer: Several 3-step time series sequences are used as input features and fed into the LSTM model;

[0086] Hidden layer: Learn long-term dependencies through gating mechanism, use ReLU activation function to enhance nonlinear expressive power and accelerate convergence, and capture temporal correlation between parameters;

[0087] Output layer: Outputs the temporal correlation coefficients between standardized values ​​of historical data, in the form of a weight matrix between each standardized value;

[0088] An intelligent argon control model is constructed, taking the standardized values ​​of deoxidizer dosage, slag-forming agent dosage, initial oxidative state of molten steel, and slag conversion amount within the same step length in historical data as inputs. Combining the weight matrix between the various standardized values ​​output by the LSTM model and the argon flow rate within the same step length in historical data, the predicted deoxidation and slag-forming argon flow rate is output.

[0089] It should be noted that the process of outputting the predicted argon flow rate for deoxidation and slag formation, by combining the weight matrix between the standardized values ​​of historical data output by the LSTM model, is as follows:

[0090] The standard value parameters of the input layer are weighted and the corresponding weights are calculated to obtain the predicted argon flow rate. The mean square error (MSE) is used to calculate the mean square error between the predicted argon flow rate and the argon flow rate within the same step length in the historical data. The predicted argon flow rate is adjusted by combining the flow rate weights in the weight matrix. The final adjusted predicted argon flow rate is output to obtain the predicted deoxidation slag formation argon flow rate.

[0091] In step one, the process of obtaining the argon flow rate value for deoxidation and slag formation is as follows:

[0092] Obtain the current deoxidizer and slag-forming agent dosages in the current deoxidation and slag-forming process, and calculate the corresponding standardized values ​​for deoxidizer and slag-forming agent dosages.

[0093] The standardized values ​​for the initial state of molten steel oxidation and the standardized value for slag conversion amount are set to 1;

[0094] It should be noted that the standardized value of the initial oxidative state of molten steel and the standardized value of slag conversion amount are set to 1 because: when the standardized value of the initial oxidative state of molten steel and the standardized value of slag conversion amount are set to 1, the initial oxidative state of molten steel and the slag conversion amount meet the standards of the initial oxidative state and the standard of slag conversion amount.

[0095] The standardized values ​​of deoxidizer dosage, slag-forming agent dosage, initial state of molten steel oxidation, and slag conversion amount are input into the argon intelligent control model, and the deoxidation and slag-forming argon flow rate is output.

[0096] Step 2: Input argon gas according to the deoxidation slag forming argon gas flow rate value for deoxidation slag forming, and determine whether the deoxidation slag forming argon gas flow rate value needs to be adjusted during the deoxidation slag forming process. If so, derive the flow component influence coefficient between the argon gas flow rate value and the component uniformity analysis value, and calculate the component argon gas flow rate value.

[0097] In step two, the process of determining whether the argon flow rate for deoxidation and slag formation needs to be adjusted is as follows:

[0098] By uniformly installing LIBS sensors at different heights and angles on the furnace wall of the refining furnace, a laser-induced breakdown spectrum (LIBS) array is formed, and the molten steel after the addition of materials is divided into several unit regions.

[0099] Based on all LIBS sensors in any unit area, the different elements in the unit area are divided into categories, and the content of different categories of elements at each monitoring point in the unit area is obtained respectively.

[0100] The content of elements in the same category is summarized into an element set, and the element set of each element category is obtained;

[0101] Perform uniform elemental composition analysis on any unit region;

[0102] Within each cell region, the standard deviation of different element sets is processed, and the average of the standard deviations of all element sets is calculated. This average is the component homogeneity analysis value, which is then compared with the component homogeneity analysis threshold.

[0103] If the composition uniformity analysis value is greater than the composition uniformity analysis threshold, the element composition of the unit region is determined to be non-uniform.

[0104] Conversely, it is judged to be uniform;

[0105] It should be noted that the threshold for component homogeneity analysis is set by experts based on the current work requirements;

[0106] Similarly, several unit regions are processed to obtain the number of units with uneven element composition.

[0107] The adjustment coefficient is obtained by comparing the number of units with uneven element composition to the total number of units.

[0108] The adjustment factor is compared with the adjustment factor threshold.

[0109] If the adjustment coefficient is greater than or equal to the adjustment coefficient threshold, it is determined that the argon flow rate value for deoxidation and slag formation needs to be adjusted.

[0110] If the adjustment coefficient is less than the adjustment coefficient threshold, it is determined that no adjustment of the argon flow rate for deoxidation and slag formation is required.

[0111] It should be noted that the adjustment coefficient threshold is set by experts based on the current work needs;

[0112] In step two, if it is necessary to adjust the argon flow rate value for deoxidation and slag formation, the process for calculating the influence coefficient of the argon flow rate value on the flow composition of the composition uniformity analysis value is as follows:

[0113] Obtain the argon flow rate values ​​for the fine-tuning process of different historical LF refining processes, and calculate the corresponding component homogeneity analysis values;

[0114] All argon flow rate values ​​and corresponding component homogeneity analysis values ​​were standardized to obtain several standardized argon flow rate values ​​and corresponding standardized component homogeneity analysis values.

[0115] The standardized argon flow rate and the corresponding homogeneous composition analysis value were fitted using the least squares method.

[0116] Constructing standardized values ​​for argon flow rate Standardized values ​​of the corresponding component homogeneous analysis values The univariate linear regression model between them:

[0117]

[0118] in, Here, k is a constant term, and k is the influence factor. This is the random error term;

[0119] Introducing the coefficient of determination The standardized values ​​of the actual component homogeneous analysis values ​​were repeatedly obtained. Standardized values ​​of component homogeneity analysis predicted by the univariate linear regression model Enter the following formula to calculate the coefficient of determination for discriminant fitting:

[0120]

[0121] in, Standardized values ​​of actual component homogeneous analysis values The mean;

[0122] Compare the calculated coefficient of determination with the coefficient of determination threshold:

[0123] If the coefficient of determination is greater than or equal to the threshold, it indicates a linear correlation between the standardized argon flow rate value and the corresponding homogeneous composition analysis value. The influence coefficient of flow component;

[0124] Conversely, this indicates that there is no linear relationship;

[0125] If there is no linear correlation between the standardized value of argon flow rate and the standardized value of the corresponding component homogeneous analysis,

[0126] Models such as the bivariate linear model, higher-order polynomial model, exponential model, power function model, and trigonometric function model are constructed respectively.

[0127] The determination coefficient of each model is calculated by fitting the model using the least squares method.

[0128] The nonlinear model corresponding to the maximum value of the coefficient of determination is taken as the nonlinear model between the standardized value of the argon flow rate and the standardized value of the corresponding component homogeneous analysis value.

[0129] Based on the characteristic values ​​(such as derivatives and function values) of the current standardized value of argon flow rate in the model, the control value that reflects the effect of unit argon flow rate on composition uniformity is obtained, and the characteristic value is the flow rate composition influence coefficient.

[0130] For example, the method for obtaining the flow component influence coefficient based on the eigenvalues ​​of the model's current standardized argon flow rate is as follows:

[0131] If the nonlinear model corresponding to the maximum value of the coefficient of determination is a quadratic function model ( The influencing factor is the derivative (2ax+b) of the current argon flow rate standardized value. Its value changes dynamically with the flow rate range, reflecting the intensity of regulation at different stages. Therefore, the derivative (2ax+b) of the current argon flow rate standardized value is the flow component influence coefficient.

[0132] It should be noted that the physical meaning of the influence coefficient of the flow rate on the compositional uniformity analysis value of argon flow rate is as follows:

[0133] The flow composition influence coefficient is obtained by least squares fitting of the standardized value of argon flow rate and the standardized value of the corresponding component uniformity analysis value. In essence, it is the average change of the standardized value of the steel component uniformity analysis value when the standardized value of argon flow rate changes by 1 unit. It directly reflects the rate at which the stirring intensity of molten steel with argon flow rate affects the component uniformity.

[0134] In step two, the process of calculating the argon flow rate is as follows:

[0135] The component homogeneity analysis values ​​of each unit are averaged to obtain the average value of the component homogeneity analysis.

[0136] The difference between the homogeneous analysis value of each unit and the average homogeneous analysis value of the components is processed to obtain several adjustment values ​​for the homogeneous analysis value of the components to be determined. The maximum value among them is taken as the adjustment value for the homogeneous analysis value of the components.

[0137] It should be noted that the reason for using the maximum value as the adjustment value for the argon flow rate in deoxidation and slag formation is as follows:

[0138] The adjustment value of the argon flow rate for deoxidation and slag forming is obtained by the difference between the component uniform analysis value and the component uniform analysis average value of each unit. The maximum value reflects the maximum degree. If the maximum adjustment value of the argon flow rate for deoxidation and slag forming is met, then the other adjustment values ​​of the argon flow rate for deoxidation and slag forming will definitely be met.

[0139] The ratio of the component uniformity analysis value adjustment value to the flow component influence coefficient is used to obtain the deoxidation slag forming argon flow adjustment value.

[0140] The argon flow rate adjustment value for deoxidation and slag formation is added to the argon flow rate value for deoxidation and slag formation to obtain the component argon flow rate value.

[0141] Step 3: Under the refining requirements of low or no power supply, the argon flow rate of deoxidation and slag formation is adjusted for the first time according to the argon flow rate of the composition. After the adjustment, the temperature uniformity analysis value is calculated in real time, and the flow-temperature influence coefficient between the argon flow rate value and the temperature uniformity analysis value is derived.

[0142] In step three, the process of calculating the temperature uniformity analysis value in real time after adjustment is as follows:

[0143] Multiple thermocouple sensors are evenly embedded in key locations such as the inner wall and bottom of the steelmaking furnace or steel holding container to form a thermocouple sensor array.

[0144] After inputting the argon flow rate value, the molten steel is divided into several unit regions. Based on the thermocouple sensor in each unit region, the real-time temperature values ​​of multiple monitoring points in each unit region are obtained.

[0145] Perform temperature uniformity analysis on any unit region;

[0146] By processing the standard deviation of temperature values ​​from multiple monitoring points within a unit area, a real-time temperature uniformity analysis value is obtained.

[0147] All unit regions are processed to obtain several real-time temperature uniformity analysis values;

[0148] In step three, the process of deriving the flow-temperature influence coefficient of argon flow rate on the uniformity of molten steel temperature is as follows:

[0149] Obtain the argon flow rate and corresponding temperature uniformity analysis values ​​for the fine-tuning process of different LF refining processes in history.

[0150] By standardizing the argon flow rate values ​​and corresponding temperature uniformity analysis values ​​in several historical data, we obtain the standardized values ​​of argon flow rate and corresponding temperature uniformity analysis values.

[0151] Combining the process of the influence coefficient of argon flow rate value on the flow component of the composition uniformity analysis value in step two, the uniform value of molten steel temperature is used to replace the uniform value of composition, and the standardized value of temperature uniformity analysis value is used to replace the standardized value of composition uniformity analysis value. Following the derivation process of the flow component influence coefficient, the flow temperature influence coefficient is derived.

[0152] Step 4: Determine the consistency between the positive and negative values ​​of the flow component influence coefficient and the flow temperature influence coefficient. If they are consistent, calculate the temperature argon flow rate value and make a secondary adjustment to the deoxidation slag formation argon flow rate value according to the temperature argon flow rate value.

[0153] In step four, the process of determining the consistency between the positive and negative values ​​of the flow component influence coefficient and the flow temperature influence coefficient is as follows:

[0154] The influence coefficients of flow composition and flow temperature are compared with the threshold 0:

[0155] If the influence coefficients of flow composition and flow temperature are of the same sign, it can be determined that the argon flow rate can be adjusted to simultaneously meet the requirements of uniform steel temperature and composition.

[0156] It should be noted that if the positive and negative values ​​are consistent, the determination can be made by adjusting the argon gas flow rate. The reason why the requirements for uniform steel temperature and composition are met is as follows:

[0157] The argon flow rate value for composition is adjusted based on the difference between the set composition uniformity analysis threshold and the composition uniformity analysis value. Among them, the adjustment amount of the argon flow rate value for deoxidation and slag removal is the minimum value. If the influence coefficient of flow rate composition and the influence coefficient of flow rate temperature are consistent, the temperature argon flow rate value is obtained by adjusting the argon flow rate value for composition. This satisfies the requirement of uniform steel temperature while the adjustment amount of the argon flow rate value for deoxidation and slag removal is still within the range that satisfies composition uniformity, so that the temperature argon flow rate value also meets the requirement of uniform steel composition.

[0158] If the influence coefficients of flow composition and flow temperature are not consistent, it is determined that the flow rate cannot be adjusted by the argon flow rate value to simultaneously meet the requirements of uniform steel temperature and composition.

[0159] In step four, the process of calculating the temperature argon flow rate and adjusting the composition argon flow rate is as follows:

[0160] If the influence coefficient of flow composition and the influence coefficient of temperature per unit flow are of the same sign, obtain the temperature uniformity analysis value for all units;

[0161] The temperature uniformity analysis values ​​of all units are averaged to obtain the average temperature uniformity analysis value.

[0162] The temperature uniformity analysis value of each unit is compared with the average temperature uniformity analysis value to obtain several undetermined component uniformity adjustment values. The maximum value among them is taken as the component uniformity adjustment value.

[0163] The component uniformity adjustment value is compared with the flow rate temperature influence coefficient to obtain the component flow rate adjustment value, and then added to the component argon flow rate value to obtain the temperature argon flow rate value.

[0164] If the influence coefficients of flow composition and flow temperature are not consistent, the flow rate of argon gas can be adjusted by other methods, such as flow field optimization design.

[0165] The technical solution of this invention is as follows: Historical data of the deoxidation and slag-forming process is acquired, the temporal correlation coefficient between standardized values ​​of the historical data is calculated, and an intelligent argon control system is constructed. This system outputs the argon flow rate value for deoxidation and slag-forming. Argon is input according to the argon flow rate value for deoxidation and slag-forming. During the deoxidation and slag-forming process, it is determined whether the argon flow rate value needs adjustment. If so, the flow-component influence coefficient between the argon flow rate value and the component uniformity analysis value is derived, and the component argon flow rate value is calculated. The argon flow rate value for deoxidation and slag-forming is adjusted for the first time according to the component argon flow rate value. After adjustment, the temperature uniformity analysis value is calculated in real time, and the flow-temperature influence coefficient between the argon flow rate value and the temperature uniformity analysis value is derived. The positive and negative consistency of the flow-component influence coefficient and the flow-temperature influence coefficient is determined. If they are consistent, the temperature argon flow rate value is calculated, and the argon flow rate value for deoxidation and slag-forming is adjusted a second time according to the temperature argon flow rate value.

[0166] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A low-energy smelting method for high-alloy building steel, characterized in that: include: Historical data of the deoxidation and slag-forming process is obtained, the temporal correlation coefficient between the standardized values ​​of the historical data is calculated, and an intelligent argon control system is constructed to output the argon flow rate value for deoxidation and slag-forming. Argon gas is input according to the deoxidation slag formation argon gas flow rate value for deoxidation slag formation. During the deoxidation slag formation process, it is determined whether the deoxidation slag formation argon gas flow rate value needs to be adjusted. If so, the flow component influence coefficient between the argon gas flow rate value and the component uniformity analysis value is derived, and the component argon gas flow rate value is calculated. The process for determining whether the argon flow rate for deoxidation and slag formation needs adjustment is as follows: After the initial adjustment of the argon flow rate for deoxidation and slag formation, the molten steel area is divided into several unit areas, and the different elements in each unit area are classified according to category. The content of elements of the same category is summarized into an element set, and the composition uniformity analysis value of each unit area is calculated. If the composition uniformity analysis value of any unit area is greater than the composition uniformity analysis threshold, it is determined that the element composition of that unit area is non-uniform. The number of elements with uneven composition in a given cell region is obtained, and its ratio to the total number of cells in the region is used to calculate an adjustment coefficient. This coefficient is then compared with an adjustment coefficient threshold. If the adjustment coefficient is greater than or equal to the adjustment coefficient threshold, it is determined that the argon flow rate value for deoxidation and slag formation needs to be adjusted. Conversely, if no adjustment is required, then no adjustment is necessary. The argon flow rate for deoxidation and slag formation was initially adjusted based on the argon flow rate of the composition, and the temperature uniformity analysis value was calculated in real time after the adjustment. The flow-temperature influence coefficient between the argon flow rate and the temperature uniformity analysis value was then derived. The process of deriving the flow component influence coefficient between the argon flow rate value and the component uniformity analysis value is as follows: Obtain the component uniformity analysis value and argon flow rate value of different historical LF refining processes, and perform standardization processing to obtain the standardized values ​​of the two. Then, perform least squares fitting processing on the two. Constructing standardized values ​​for argon flow rate Standardized values ​​of homogeneous analysis of corresponding components A univariate linear regression model between them; Calculate the coefficient of determination of the fitting result of the univariate linear regression model; The calculated coefficient of determination is compared with the threshold of the coefficient of determination, and the influence coefficient of the flow component is obtained based on the comparison result; The derivation process of the flow rate temperature influence coefficient is the same as that of the flow rate temperature influence coefficient. Determine the consistency between the positive and negative values ​​of the flow component influence coefficient and the flow temperature influence coefficient. If they are consistent, calculate the temperature argon flow rate value and make a secondary adjustment to the deoxidation slag formation argon flow rate value according to the temperature argon flow rate value. The process of making a secondary adjustment to the argon flow rate for deoxidation and slag formation based on the temperature and argon flow rate is as follows: If the influence coefficients of flow composition and flow temperature are of the same sign, then obtain the temperature uniformity analysis values ​​of all units and perform averaging to obtain the average value of the temperature uniformity analysis. The temperature uniformity analysis value and the average temperature uniformity analysis value of each unit are respectively processed by difference, and the maximum value is taken as the component uniformity adjustment value. The component uniformity adjustment value is calculated by ratioing the flow rate temperature influence coefficient to obtain the component flow rate adjustment value, which is then added to the component argon flow rate value to obtain the temperature argon flow rate value.

2. The low-energy smelting method for high-alloy building steel according to claim 1, characterized in that: The process of calculating the time-series correlation coefficient between standardized values ​​of historical data is as follows: Historical data on the deoxidation and slag-forming process were obtained, including the amount of deoxidizer, the amount of slag-forming agent, the argon flow rate, the initial state of molten steel oxidation, and the amount of slag conversion. The historical data on the deoxidation and slag-forming process were then standardized. The standardized processing results are divided and selected to obtain several 3-step time series sequences; Several 3-step time series sequences are input into the LSTM model. The hidden layer captures the temporal correlation between each standardized value and outputs the temporal correlation coefficient between the standardized values ​​of historical data.

3. The low-energy smelting method for high-alloy building steel according to claim 2, characterized in that: The process of dividing and selecting the standardized results to obtain several 3-step time series sequences is as follows: Historical data on the deoxidation and slag formation process were standardized to obtain standardized values ​​for deoxidizer dosage. Standardized values ​​for slag-forming agent dosage Standardized value of initial state of oxidative properties of molten steel Standardized value of slag conversion Standardized value of argon flow rate And the five standardized values ​​within the same time step are recorded as a standard set; Obtain a standard set of historical data within several time steps, group three consecutive time steps together to form a 3×5 time series matrix, which is a 3-step time series sequence. Perform several selections to obtain several 3-step time series sequences.

4. The low-energy smelting method for high-alloy building steel according to claim 2, characterized in that: The process of constructing an intelligent argon control system and outputting the argon flow rate value for deoxidation and slag formation is as follows: Input layer: The standardized values ​​of deoxidizer dosage, slag-forming agent dosage, initial oxidative properties of molten steel, and slag conversion amount within the same step length in historical data are used as input; Hidden layer: The weight matrix between the standardized values ​​output by the LSTM model and the argon flow rate values ​​within the same step size in historical data; Output layer: Outputs the predicted argon flow rate for deoxidation and slag formation; Obtain the current deoxidizer and slag-forming agent dosages in the current deoxidation and slag-forming process, and calculate the corresponding standardized values ​​for deoxidizer and slag-forming agent dosages. The standardized values ​​for the initial state of molten steel oxidation and the standardized value for slag conversion are set to 1. The current standardized values ​​of deoxidizer dosage, slag-forming agent dosage, initial oxidative state of molten steel, and slag conversion amount are input into the argon intelligent control model, which outputs the deoxidation and slag-forming argon flow rate.

5. A low-energy smelting method for high-alloy building steel according to claim 1, characterized in that: The process of calculating the compositional uniformity analysis values ​​for each unit region is as follows: The standard deviation of different element sets within the cell region is processed, and the average of the standard deviations of all element sets is calculated, which is the component homogeneity analysis value.

6. A low-energy smelting method for high-alloy building steel according to claim 5, characterized in that: The process of obtaining the influence coefficient of flow components based on the comparison results is as follows: If the coefficient of determination is greater than or equal to the threshold of the coefficient of determination, it indicates that there is a linear correlation between the standardized value of argon flow rate and the corresponding homogeneous analysis value of composition. Conversely, this indicates that there is no linear relationship; If it is a linear correlation, then the coefficient of the standardized value of argon flow rate in the univariate linear regression model formula is the influence coefficient of the flow rate component. If the correlation is nonlinear, construct a binary linear model, a higher-order polynomial model, an exponential model, a power function model, and a trigonometric function model respectively, and calculate the determination coefficient of each model. The nonlinear model corresponding to the maximum value of the coefficient of determination is taken as the nonlinear model between the standardized value of argon flow rate and the standardized value of the corresponding component homogeneity analysis; Based on the characteristic values ​​of the current standardized argon flow rate of the model, the influence coefficients of the flow rate components are obtained.

7. A low-energy smelting method for high-alloy building steel according to claim 1, characterized in that: The process for determining the consistency between the positive and negative values ​​of the flow component influence coefficient and the flow temperature influence coefficient is as follows: If the influence coefficients of flow composition and flow temperature are both positive or both negative, it indicates that their positive and negative values ​​are consistent; otherwise, they are inconsistent.

Images