Instantaneous liquid level fluctuation control method, device and equipment of slab continuous casting crystallizer and medium

Abstract

The invention relates to the technical field of crystallizers, and provides an instantaneous liquid level fluctuation control method, device and equipment for a slab continuous casting crystallizer and a medium. The method comprises the steps that parameter data are extracted from the slab continuous casting crystallizer of a target steel mill; based on the Pearson correlation coefficient and the Spearman correlation coefficient, analyzing a first relation and a second relation between the argon blowing amount at the stopper rod and the liquid level fluctuation data and between the argon blowing amount at the sliding plate and the liquid level fluctuation data; determining a fitting linear relation between the argon blowing quantity data and the liquid level fluctuation data; real-time argon blowing quantity data are obtained, real-time liquid level fluctuation data are predicted according to the real-time argon blowing quantity data and the fitting linear relation, and predicted liquid level fluctuation data are obtained; and in response to the situation that the real-time liquid level fluctuation data exceed the preset normal range, the argon blowing amount at the stopper rod and / or the argon blowing amount at the sliding plate are / is adjusted so as to control the real-time liquid level fluctuation data to be maintained within the preset normal range. According to the scheme, the liquid level fluctuation can be quickly and accurately judged, and then the liquid level fluctuation of the crystallizer is controlled to be stable.

Description

Technical Field

[0001] This application relates to the field of crystallizer technology, and in particular to a method, apparatus, equipment and medium for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer. Background Technology

[0002] Continuous casting is a core step in the steel production process, playing a crucial role. However, due to the numerous complex processes involved, surface defects often occur in the produced billets, reducing their quality. Numerous studies have demonstrated that fluctuations in the molten steel level within the crystallizer significantly impact the surface quality of the billets, with abnormal fluctuations drastically increasing the probability of surface defects. In particular, transient abnormal fluctuations in the crystallizer level, lacking periodicity, are difficult to control and eliminate effectively in actual production. Frequent transient abnormal fluctuations in the crystallizer level severely reduce continuous casting efficiency and billet quality, leading to significant economic losses. Currently, with the continuous development of computer technology and data feature mining techniques, utilizing big data feature analysis to explore the potential relationships between anomalies and processes, and then optimizing these processes to eliminate anomalies in production, has become an inevitable trend in eliminating abnormal fluctuations in the crystallizer level. Summary of the Invention

[0003] This application provides a method, apparatus, equipment, and medium for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer, aiming to solve many technical problems existing in related technologies.

[0004] In a first aspect, embodiments of this application provide a method for controlling instantaneous liquid level fluctuations in a slab continuous casting mold, including: Parameter data are extracted from the slab continuous casting crystallizer of the target steel plant, wherein the parameter data includes: liquid level fluctuation data, stopper rod position data and argon blowing volume data at different positions; The first relationship between the argon blowing volume at the stopper rod and the argon blowing volume at the slide plate and the liquid level fluctuation data was analyzed based on the Pearson correlation coefficient, and the second relationship between the argon blowing volume at the stopper rod and the argon blowing volume at the slide plate and the liquid level fluctuation data was analyzed based on the Spearman correlation coefficient. Based on the first relationship and the second relationship, a fitted linear relationship is determined between the argon blowing volume data and the liquid level fluctuation data; The real-time argon blowing volume data in the slab continuous casting crystallizer of the target steel plant is obtained. Based on the real-time argon blowing volume data and the fitted linear relationship, the real-time liquid level fluctuation data is predicted to obtain the predicted liquid level fluctuation data. In response to the real-time liquid level fluctuation data exceeding the preset normal range, the argon blowing amount at the stopper rod and / or the argon blowing amount at the slide plate is adjusted to control the real-time liquid level fluctuation data to remain within the preset normal range.

[0005] In one embodiment, optionally, after extracting the parameter data, the method further includes: The parameter data is preprocessed to obtain preprocessed parameter data, wherein the preprocessing process includes: Abnormal fluctuation data caused by changes in process conditions are removed from the parameter data to obtain the initial screening data; Based on the process stabilization time, all abnormal liquid level fluctuation data within the same time period are obtained from the initial screening data and sorted in descending order; From all the abnormal fluctuation data of liquid level, select the top-ranked preset number of target abnormal fluctuation data, and calculate the average value corresponding to all target abnormal fluctuation data. The average value is used as the degree of liquid level fluctuation during the time period.

[0006] In one embodiment, optionally, analyzing the first relationship between the argon blowing volume at the stopper rod, the argon blowing volume at the slide plate, and the liquid level fluctuation data based on the Pearson correlation coefficient includes: Two sets of variable pairs were obtained by extracting the argon blowing volume at the stopper rod and its corresponding first liquid level fluctuation degree, and the argon blowing volume at the slide plate and its corresponding second liquid level fluctuation degree from the preprocessed parameter data. Based on the Pearson correlation coefficient calculation formula, the first correlation coefficient between the amount of argon blown at the stopper rod and the degree of the first liquid level fluctuation, and the second correlation coefficient between the amount of argon blown at the slide plate and the degree of the second liquid level fluctuation are calculated respectively. The first correlation coefficient, the second correlation coefficient, and the first preset coefficient threshold are compared. Based on the comparison results, the first target variable pair with a correlation coefficient greater than the preset coefficient threshold is selected to form the first effective association dataset.

[0007] In one embodiment, optionally, analyzing the second relationship between the argon blowing volume at the stopper rod, the argon blowing volume at the slide plate, and the liquid level fluctuation data based on the Spearman correlation coefficient includes: The variable values ​​of the two sets of variable pairs are sorted respectively, and a unique rank order is assigned to each variable value. If there are variable values ​​with the same value, the average of the rank orders corresponding to the variable values ​​with the same value in the group is taken as their common rank. For each pair of variables, calculate the grade difference of the corresponding variable values ​​according to the assigned grade order; Based on the grade difference and Spearman correlation coefficient calculation formula, the third correlation coefficient between the argon blowing amount at the stopper rod and the first liquid level fluctuation degree, and the fourth correlation coefficient between the argon blowing amount at the slide plate and the second liquid level fluctuation degree are calculated respectively. The third correlation coefficient, the fourth correlation coefficient, and the second preset coefficient threshold are compared. Based on the comparison results, the second target variable pairs with correlation coefficients greater than the second preset coefficient threshold are selected to form the second effective association dataset.

[0008] In one embodiment, optionally, determining the fitted linear relationship between the argon blowing volume data and the liquid level fluctuation data based on the first relationship and the second relationship includes: The final valid association dataset is determined based on the first valid association dataset and the second valid association dataset; Big data feature analysis was performed on the final effective correlation dataset, and the linear relationship between the argon blowing volume data and the liquid level fluctuation degree was obtained by fitting the least squares method.

[0009] In one embodiment, optionally, the method further includes: Extract the actual real-time liquid level fluctuation data corresponding to the real-time argon blowing data from the slab continuous casting crystallizer; The real-time liquid level fluctuation data and the predicted liquid level fluctuation data are compared, and the fitting linear relationship is updated based on the comparison result.

[0010] In one embodiment, optionally, comparing the real-time liquid level fluctuation data and the predicted liquid level fluctuation data, and determining whether to update the fitted linear relationship based on the comparison result, includes: Calculate the mean absolute error and coefficient of determination between the real-time liquid level fluctuation data and the predicted liquid level fluctuation data; If the mean absolute error value is greater than or equal to a preset error threshold and the coefficient of determination is greater than or equal to a preset coefficient threshold, it is determined not to update the fitted linear relationship; otherwise, it is determined to update the fitted linear relationship.

[0011] Secondly, embodiments of this application provide a device for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer, comprising: The extraction module is used to extract parameter data from the slab continuous casting crystallizer of the target steel plant, wherein the parameter data includes: liquid level fluctuation data, stopper rod position data and argon blowing volume data at different positions; The analysis module is used to analyze the first relationship between the argon blowing volume at the stopper rod and the argon blowing volume at the slide plate and the liquid level fluctuation data based on the Pearson correlation coefficient, and to analyze the second relationship between the argon blowing volume at the stopper rod and the argon blowing volume at the slide plate and the liquid level fluctuation data based on the Spearman correlation coefficient. The relationship determination module is used to determine the fitted linear relationship between the argon blowing volume data and the liquid level fluctuation data based on the first relationship and the second relationship; The prediction module is used to acquire real-time argon blowing data in the slab continuous casting crystallizer of the target steel plant, and predict real-time liquid level fluctuation data based on the real-time argon blowing data and the fitted linear relationship, thereby obtaining predicted liquid level fluctuation data. The control module is used to adjust the argon blowing volume at the stopper rod and / or the argon blowing volume at the slide plate in response to the real-time liquid level fluctuation data exceeding the preset normal range, so as to control the real-time liquid level fluctuation data to be maintained within the preset normal range.

[0012] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for controlling the instantaneous liquid level fluctuation of a slab continuous casting crystallizer.

[0013] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the above-described method for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer.

[0014] In the above-described method, device, equipment, and medium for controlling instantaneous liquid level fluctuations in slab continuous casting molds, parameter data is extracted from the slab continuous casting mold of the target steel plant. This parameter data includes liquid level fluctuation data, stopper rod position data, and argon blowing volume data at different positions. A first relationship between the argon blowing volume at the stopper rod, the argon blowing volume at the slide plate, and the liquid level fluctuation data is analyzed based on Pearson correlation coefficients. A second relationship between the argon blowing volume at the stopper rod, the argon blowing volume at the slide plate, and the liquid level fluctuation data is analyzed based on Spearman correlation coefficients. Based on the first and second relationships, a fitted linear relationship is determined between the argon blowing volume data and the liquid level fluctuation data. Real-time argon blowing volume data in the slab continuous casting mold of the target steel plant is obtained. Based on the real-time argon blowing volume data and the fitted linear relationship, real-time liquid level fluctuation data is predicted to obtain predicted liquid level fluctuation data. In response to the real-time liquid level fluctuation data exceeding a preset normal range, the argon blowing volume at the stopper rod and / or the argon blowing volume at the slide plate is adjusted to control the real-time liquid level fluctuation data to remain within the preset normal range. The present invention firstly proposes a new evaluation standard for the degree of transient abnormal fluctuations in the liquid level of the crystallizer by using data feature mining technology, and then establishes a relationship between the degree of fluctuation and the production process to determine the continuous casting production process that causes the transient abnormal fluctuations in the liquid level of the crystallizer, thereby enabling rapid and accurate judgment of the liquid level fluctuations and thus controlling the stability of the liquid level fluctuations in the crystallizer. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A schematic flowchart of a method for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer according to an embodiment of this application is shown.

[0017] Figure 2 A flowchart of the parameter data preprocessing process in a method for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer according to an embodiment of this application is shown.

[0018] Figure 3 A schematic flowchart of step S102 in a method for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer according to an embodiment of this application is shown.

[0019] Figure 4 A schematic flowchart of step S102 in a method for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer according to another embodiment of this application is shown.

[0020] Figure 5 A schematic flowchart of a method for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer according to another embodiment of this application is shown.

[0021] Figure 6 A liquid level fluctuation diagram according to an embodiment of this application is shown.

[0022] Figure 7 A correlation analysis result graph is shown according to an embodiment of this application.

[0023] Figure 8 A comparison graph of field data and a straight line obtained by fitting a formula is shown according to an embodiment of this application.

[0024] Figure 9 A graph showing the comparison results of different evaluation parameters according to one embodiment of this application is illustrated.

[0025] Figure 10 A block diagram of a device for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer according to an embodiment of this application is shown.

[0026] Figure 11 A block diagram of a computer device according to one embodiment of this application is shown. Detailed Implementation

[0027] To better understand the technical solution of this application, the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0028] It should be understood that the described embodiments are merely some, not all, of the embodiments in this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0029] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0030] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0031] Please see Figure 1 , Figure 1 A schematic flowchart of a method for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer according to an embodiment of this application is shown.

[0032] like Figure 1 As shown, the flow of a method for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer according to an embodiment of this application includes: Step S101: Extract parameter data from the slab continuous casting crystallizer of the target steel plant, wherein the parameter data includes: liquid level fluctuation data, stopper rod position data and argon blowing volume data at different positions; The slab continuous casting crystallizer is the core equipment in continuous casting production. It is a metal container that continuously cools and solidifies molten steel into a slab shape. The stability of the molten steel level inside the crystallizer directly determines the surface quality of the slab. Liquid level fluctuation data is a numerical sequence of changes in the molten steel level over time, collected by level sensors (such as electromagnetic induction or thermocouple types). The unit is usually millimeters (mm), reflecting the dynamic changes in the molten steel level. Stopper rod position data is the real-time displacement data of the stopper rod (a columnar component used to adjust the flow rate of molten steel) installed above the immersion nozzle of the crystallizer, measured in millimeters (mm). Its positional changes directly control the opening and closing degree of the nozzle channel. Argon blowing data at different locations is the flow rate data of protective argon gas introduced into the crystallizer through argon gas pipelines. This mainly includes the argon gas flow rate at the stopper rod (center channel of the stopper rod) and the sliding plate (the gap between the nozzle and the stopper rod), measured in liters per minute (L / min). This is used to prevent nozzle blockage and improve the fluidity of the molten steel.

[0033] In this step, first determine the target steel mill and the corresponding slab continuous casting machine, and extract the parameter data during the continuous production process from the data storage through the automated data acquisition system (such as PLC control system, SCADA system) supporting the mold. The acquisition frequency needs to match the dynamic changes of the process to ensure capturing instantaneous liquid level fluctuations. The extracted data needs to be aligned by time stamps to ensure that the liquid level fluctuations, stopper rod positions, and argon blowing amounts at the same moment correspond one by one.

[0034] Through the above technical solution, systematic acquisition of key process parameters is achieved, avoiding the randomness and errors of manual acquisition, ensuring data integrity and timeliness, and providing raw data support for subsequent analysis of the correlation between argon blowing amount and liquid level fluctuations.

[0035] Step S102, analyze the first relationship between the argon blowing amount at the stopper rod, the argon blowing amount at the slide plate, and the liquid level fluctuation data based on the Pearson correlation coefficient, and analyze the second relationship between the argon blowing amount at the stopper rod, the argon blowing amount at the slide plate, and the liquid level fluctuation data based on the Spearman correlation coefficient; The Pearson correlation coefficient is an index in statistics used to measure the linear association strength between two continuous variables, with a value range of [-1, 1]. The closer the absolute value is to 1, the stronger the linear association; positive correlation means that when one variable increases, the other variable increases synchronously, and negative correlation is the opposite.

[0036] The Spearman correlation coefficient is an index used to measure the monotonic association strength between two variables, without requiring the variables to follow a normal distribution, and also has a value range of [-1, 1]. By ranking the variable values and assigning ranks, the association degree is calculated, and it is more tolerant to outliers and can capture linear and non-linear monotonic associations.

[0037] The first relationship is the linear association relationship between the argon blowing amount at the stopper rod / slide plate and the liquid level fluctuation data obtained through Pearson correlation coefficient analysis.

[0038] The second relationship: the monotonic association relationship between the argon blowing amount at the stopper rod / slide plate and the liquid level fluctuation data obtained through Spearman correlation coefficient analysis.

[0039] Such as Figure 2 shown, in one embodiment, optionally, after extracting the parameter data, the method further includes: Preprocess the parameter data to obtain preprocessed parameter data, where the preprocessing process includes: Step S201, eliminate the abnormal fluctuation data caused by process condition changes from the parameter data to obtain preliminarily screened data; Changes in process conditions: Deterministic process operations that cause non-instantaneous abnormal fluctuations in liquid level during continuous casting production mainly include ladle replacement (replacing the molten steel ladle), speed reduction / increase (adjusting the billet casting speed), and nozzle replacement. The fluctuations caused by these operations are unrelated to the argon blowing volume and need to be excluded to avoid interfering with the correlation analysis.

[0040] The initial screening data consists of parameter data that remain after removing abnormal fluctuations caused by changes in process conditions. This data reflects only the influence of controllable factors such as the amount of argon blown, providing a clean data sample for subsequent calculations of the degree of liquid level fluctuation.

[0041] This step eliminates interference from uncontrollable factors (process operations) on liquid level fluctuation data, ensuring that the correlation between argon blowing volume and liquid level fluctuation in subsequent analyses only reflects the true impact of both, avoiding false correlations and improving data quality.

[0042] Step S202: Based on the process stabilization time, obtain all abnormal liquid level fluctuation data within the same time period from the initial screening data and sort them in descending order; The process stabilization time is the period during which process parameters (casting speed, stopper position, and argon blowing rate) remain stable in continuous casting production. This ensures that liquid level fluctuations during this period are mainly affected by the argon blowing rate, rather than process adjustments.

[0043] The abnormal fluctuation data of liquid level is the fluctuation value of the liquid level deviating from the set value in the initial screening data, reflecting the degree of abnormal deviation of the liquid level at that moment.

[0044] In this step, the initial screening data is divided into multiple consecutive time segments, using the process stabilization time (e.g., 30 seconds) as the time unit. All abnormal liquid level fluctuation data (absolute values ​​are taken, focusing only on the magnitude of the fluctuation) within each time segment are sorted in descending order of value.

[0045] Step S203: Select the top-ranked preset number of target liquid level abnormal fluctuation data from all the liquid level abnormal fluctuation data, and calculate the average value corresponding to all target liquid level abnormal fluctuation data. In this process, a preset quantity can be set according to the process requirements. For example, the preset quantity can be set to 10 to ensure that the main abnormal fluctuations within the time period are covered, while avoiding too many small fluctuation data from lowering the representativeness of the average value.

[0046] For example, set a preset quantity (e.g., 10), and select the top 10 maximum values ​​from the descending fluctuation data of each stable process time period as the target data. If there are fewer than 10 fluctuation data in that time period (e.g., only 8), then all of them are selected. Calculate the arithmetic mean of these target data. For example, if the target data are 5mm, 4.8mm, 4.5mm, 4.2mm, 3.9mm, 3.7mm, 3.5mm, 3.3mm, 3.1mm, and 2.9mm, the average value is 3.8mm.

[0047] Step S204: The average value is taken as the degree of liquid level fluctuation during the time period.

[0048] The level fluctuation degree is the average value of the abnormal fluctuation data of the target level within each stable process period, in millimeters (mm), and is used to measure the severity of the instantaneous abnormal level fluctuation within that period.

[0049] This technical solution transforms discrete instantaneous fluctuation data into a quantitative indicator (average value) reflecting the overall fluctuation level over a time period. This facilitates subsequent correlation analysis with argon blowing volume data, thereby avoiding random interference from single-moment data. Simultaneously, a unified standard for calculating fluctuation levels ensures the comparability of fluctuation data from different time periods, improving the accuracy of correlation analysis.

[0050] like Figure 3 As shown, in one embodiment, optionally, step S102 includes: Step S301: Extract the argon blowing amount at the stopper rod and its corresponding first liquid level fluctuation degree, and the argon blowing amount at the slide plate and its corresponding second liquid level fluctuation degree from the preprocessed parameter data to obtain two sets of variable pairs; From the aligned parameter data, two sets of variable pairs were separated: the argon blowing amount at the stopper rod and its corresponding first liquid level fluctuation degree, and the argon blowing amount at the slide plate and its corresponding second liquid level fluctuation degree.

[0051] Step S302: Based on the Pearson correlation coefficient calculation formula, calculate the first correlation coefficient between the amount of argon blown at the stopper rod and the degree of the first liquid level fluctuation, and the second correlation coefficient between the amount of argon blown at the slide plate and the degree of the second liquid level fluctuation. To accurately determine the strength and direction of the correlation between different process parameters and crystallizer level fluctuations in continuous casting production, this study introduces the Pearson correlation coefficient as the core analytical tool. Through quantitative calculation of this coefficient, the degree of linear correlation between process parameter changes and level fluctuations can be objectively measured, providing data support for subsequent screening of key process factors affecting level fluctuations. Specifically, the Pearson correlation coefficient is a classic quantitative indicator in statistics used to analyze the correlation between two continuous random variables. Its core function is to describe the strength of the linear correlation between two random variables in a numerical form, and this coefficient has clear statistical significance in representing the linear relationship between variables, effectively avoiding bias caused by subjective judgment. From a definitional perspective, for any two random variables M and N, the calculation logic of their Pearson correlation coefficient is as follows: first, solve for the covariance of these two variables (used to measure the trend of common change between variables), then divide the covariance result by the product of the standard deviations of the two variables (used to eliminate the influence of differences in the magnitude of the variables on the correlation judgment). Through this standardization process, the final correlation coefficient is strictly limited to the range of -1 to 1, ensuring the comparability of the correlations of variables with different dimensions.

[0052] In the interpretation of the Pearson correlation coefficient, a coefficient of 0 indicates that there is no linear relationship between the two variables, meaning that changes in process parameters do not cause regular changes in liquid level fluctuations. Conversely, a Pearson correlation coefficient of -1 or 1 indicates a completely monotonic linear relationship between the two variables. Specifically, a coefficient of 1 indicates a perfectly positive linear correlation (increased process parameters lead to a synchronous increase in liquid level fluctuation amplitude), and a coefficient of -1 indicates a perfectly negative linear correlation (increased process parameters lead to a synchronous decrease in liquid level fluctuation amplitude). Both cases imply a very strong linear linkage between the variables.

[0053] Based on the above definition, the mathematical expression of the Pearson correlation coefficient is shown in formula (1). This formula directly outputs the degree of linear correlation between the variables M and N by quantifying their statistical characteristics. It is the core calculation basis for subsequent correlation analysis between process parameters and liquid level fluctuations.

[0054] (1) Where Me is the average of variable Mi, and Ne is the average of variable Ni. The correlation coefficient rMN ranges from -1 to 1.

[0055] The correlation coefficient between two sets of variables is calculated based on the Pearson correlation coefficient formula: first, calculate the mean of each set of variables, then solve for the covariance (which reflects the common trend of change of the variables), and finally divide by the product of the standard deviations of the two variables (to eliminate the difference in magnitude) to obtain the quantitative value of linear association.

[0056] Step S303: Compare the first correlation coefficient, the second correlation coefficient, and the first preset coefficient threshold, and select the first target variable pair with a correlation coefficient greater than the preset coefficient threshold based on the comparison result to form the first effective association dataset.

[0057] The first preset coefficient threshold is a critical value used to determine whether the Pearson correlation coefficient is significant. It is set based on the characteristics of continuous casting process data. A coefficient greater than this threshold indicates a significant linear association and can be included in subsequent analysis. The first target variable pair is a pair of variables whose Pearson correlation coefficient is greater than the first preset coefficient threshold (e.g., argon blowing amount at the stopper rod - fluctuation degree). This reflects that the linear association of this set of variables is statistically significant and can be used to construct an effective association dataset. The first effective association dataset consists of sample data from all first target variable pairs, providing a basis for subsequent cross-validation with the Spearman coefficient screening results.

[0058] For example, the first preset coefficient threshold is set to 0.6. The calculated first correlation coefficient (0.72) and second correlation coefficient (0.65) are compared with the threshold respectively. Both are greater than 0.6. Therefore, the argon blowing amount at the stopper rod and the argon blowing amount at the slide plate are both determined as the first target variable pair. All sample data of these two variable pairs (such as the argon blowing amount and fluctuation value of each time period) are integrated to form the first effective correlation dataset.

[0059] In the above technical solution, pairs of variables with insignificant linear associations are eliminated, the interference of invalid data on model fitting is reduced, the analysis efficiency is improved, and a dataset focusing on strongly linearly related variables is formed.

[0060] like Figure 4 As shown, in one embodiment, optionally, step S102 includes: Step S401: Sort the variable values ​​of the two sets of variable pairs respectively, and assign a unique level order to each variable value. If there are variable values ​​with the same value, take the average of the level orders corresponding to the variable values ​​with the same value in the group as their common level. The ranking order is a sequence number assigned to each variable value after sorting them by size (e.g., variable values ​​[2.3, 2.5, 2.1] are sorted as [2.1, 2.3, 2.5], with ranking orders of 1, 2, and 3 respectively), which is used to calculate the Spearman correlation coefficient.

[0061] Step S402: For each pair of variables, calculate the grade difference of the corresponding variable values ​​according to the assigned grade order; The grade difference is the difference between the argon blowing volume grade and the fluctuation level grade in the same data sample (e.g., if the argon blowing volume grade of a sample is 1.5 and the fluctuation level grade is 2, the grade difference is 1.5-2=-0.5), which is the core intermediate variable for calculating the Spearman correlation coefficient.

[0062] The difference reflects the degree of deviation between the two groups of levels. The smaller the difference, the stronger the monotonic correlation between the amount of argon blown and the degree of fluctuation, providing an intuitive intermediate result for coefficient calculation.

[0063] Step S403: Based on the grade difference and Spearman correlation coefficient calculation formula, calculate the third correlation coefficient between the argon blowing amount at the stopper rod and the first liquid level fluctuation degree, and the fourth correlation coefficient between the argon blowing amount at the slide plate and the second liquid level fluctuation degree. When analyzing the correlation between continuous casting process parameters and transient abnormal fluctuations in the liquid level of the crystallizer, in addition to the Pearson correlation coefficient mentioned earlier, the Spearman correlation coefficient is also an important quantitative analysis tool. The core function of this coefficient is to assess the strength and direction of the correlation between two variables. Unlike the Pearson correlation coefficient, which focuses on linear relationships, the Spearman correlation coefficient describes the association between variables through a monotonic function. That is, it does not require a strict linear relationship; as long as the variables exhibit a monotonic trend of "when one variable increases, the other variable continuously increases or continuously decreases," this coefficient can effectively capture such a correlation.

[0064] In terms of applicability, Spearman correlation coefficient is more flexible. It can be used not only for continuous variables but also for discrete variables. It has a significant advantage in correlation analysis of ordinal variables. This characteristic allows it to cover more types of process parameters and liquid level fluctuation data combinations in continuous casting production, making up for the shortcomings of some parameters that cannot be analyzed by Pearson correlation coefficient due to data type limitations.

[0065] The Spearman correlation coefficient is calculated as follows: The values ​​of the two random variables A and B are sorted from smallest to largest (or from largest to smallest), that is, each variable value is assigned a rank order according to its position relative to other variable values ​​within its group. When there are no two A values ​​or two B values ​​with the same rank order, the Spearman correlation coefficient is defined by formula (2).

[0066] (2) Where d represents the difference between the two order ranks A and B, and n is the number of ordered pairs. In terms of the meaning of the coefficient values, the Spearman correlation coefficient and the Pearson correlation coefficient have similar interpretive logic: when the coefficient is 0, it indicates that there is no monotonic relationship between the two variables, meaning that the change in one variable cannot predict the trend of the other variable through a monotonic function; conversely, when the coefficient is -1, it represents a completely negative monotonic relationship between the two variables (when one variable increases, the other variable continuously decreases); when the coefficient is 1, it represents a completely positive monotonic relationship between the two variables (when one variable increases, the other variable continuously increases). Both extreme values ​​imply a very strong monotonic linkage between the variables.

[0067] The correlation coefficient between two sets of variables is calculated based on the Spearman correlation coefficient formula: First, sort each set of variable values ​​in ascending / descending order and assign them to ranks (if there are identical values, take the average of the ranks), then calculate the sum of squares of the rank differences, and substitute them into the formula to obtain the monotonic correlation quantification value.

[0068] Step S404: Compare the third correlation coefficient, the fourth correlation coefficient, and the second preset coefficient threshold. Based on the comparison result, select the second target variable pair with a correlation coefficient greater than the second preset coefficient threshold to form the second effective association dataset.

[0069] The second preset coefficient threshold is the critical value for judging whether the Spearman correlation coefficient is significant. It is usually consistent with the first preset threshold to ensure that the dual-coefficient screening standard is consistent.

[0070] The second target variable pair is the pair of variables whose Spearman correlation coefficient is greater than the second preset coefficient threshold, reflecting that the monotonic association of this group of variables is significant.

[0071] The second effective association dataset consists of sample data for all pairs of the second target variables and is used for cross-validation with the first effective association dataset.

[0072] For example, the second preset coefficient threshold is set to 0.6. The third correlation coefficient (0.7) and the fourth correlation coefficient (0.63) are compared with the threshold respectively. Both are greater than 0.6. Therefore, the amount of argon blown at the stopper rod and the degree of fluctuation, as well as the amount of argon blown at the slide plate and the degree of fluctuation, are determined to be the second target variable pair. Their sample data are integrated to form the second effective association dataset.

[0073] In the above technical solution, the Pearson coefficient focuses on linear correlations, while the Spearman coefficient covers monotonic correlations. By performing complementary analysis on the two coefficients, potential correlations are avoided from being overlooked by a single coefficient (e.g., when the argon blowing rate and liquid level fluctuation have a non-linear monotonic relationship, the Pearson coefficient may underestimate the correlation strength). Quantifying the impact of the argon blowing rate on liquid level fluctuations provides data for subsequent screening of key control variables, avoiding blind adjustments to process parameters.

[0074] Although both Pearson and Spearman correlation coefficients can be used to characterize the correlation between variables, and their results show some consistency in certain linear correlation scenarios, their core applicable scenarios and calculation logic differ fundamentally. Pearson correlation coefficient focuses on linear relationships and has high requirements for preconditions such as data normality and the absence of outliers; Spearman correlation coefficient focuses on monotonic relationships, has no strict requirements on data distribution, and is more tolerant of outliers.

[0075] Therefore, to ensure the comprehensiveness and reliability of the correlation analysis between continuous casting process and transient crystallizer level anomalies in this invention, and to avoid the limitations of single-coefficient analysis, the study simultaneously selected Pearson correlation coefficient and Spearman correlation coefficient as dual reference indicators for joint analysis. By mutually verifying and supplementing the calculation results of the two coefficients, the study comprehensively judges whether there is a correlation between continuous casting process parameters and transient crystallizer level anomalies, what type of correlation (linear or monotonic), and the strength of the correlation. This provides more comprehensive data analysis support for subsequent screening of key influencing process parameters and optimization of level control strategies.

[0076] Step S103: Determine the fitted linear relationship between the argon blowing volume data and the liquid level fluctuation data based on the first relationship and the second relationship; Fitting a linear relationship is a linear function model between the argon blowing volume and the liquid level fluctuation data, constructed based on a large amount of sample data and mathematical methods (such as the least squares method). The form is usually y=kx+b (single variable) or y=k1x1+k2x2+b (bivariate), where y is the liquid level fluctuation data, x1 / x2 is the argon blowing volume at the stopper rod / slide plate, k1 / k2 is the influence coefficient, and b is a constant term.

[0077] In the above technical solution, the analysis results of the first relationship (linear correlation) and the second relationship (monotonic correlation) are combined: for example, if the argon blowing amount at a certain location simultaneously satisfies "the absolute value of the Pearson correlation coefficient ≥ the first preset threshold (e.g., 0.5)" and "the absolute value of the Spearman correlation coefficient ≥ the second preset threshold (e.g., 0.5)", then the argon blowing amount at that location is determined as the key variable. Based on the key variable and the effective sample of liquid level fluctuation data, the coefficients (k, b) of the linear function are solved using the least squares method to obtain the fitted linear relationship. For example, if only the argon blowing amount at the stopper rod is the key variable, the fitting formula might be: .

[0078] In one embodiment, optionally, step S103 includes: The final valid association dataset is determined based on the first valid association dataset and the second valid association dataset; The final effective association dataset can be the intersection of the first effective association dataset and the second effective association dataset, that is, the sample data of variable pairs that exist in both datasets, ensuring that the data samples simultaneously meet the requirements of linear association and significant monotonic association.

[0079] Big data feature analysis was performed on the final effective correlation dataset, and the linear relationship between the argon blowing volume data and the liquid level fluctuation degree was obtained by fitting the least squares method.

[0080] The least squares method is a mathematical approach used to find the optimal straight line for fitting data. By minimizing the sum of squared errors between the actual data points and the fitted line, it ensures that the fitted model is closest to the true data pattern.

[0081] Step S104: Obtain real-time argon blowing data in the slab continuous casting crystallizer of the target steel plant; predict real-time liquid level fluctuation data based on the real-time argon blowing data and the fitted linear relationship; and obtain predicted liquid level fluctuation data. Real-time argon flow data is collected in real time from the argon flow rate at the stopper rod / slide plate by flow meters (such as vortex flow meters and mass flow meters) installed on the argon pipeline. The collection frequency is matched with the control requirements to reflect the current argon supply status of the process.

[0082] The predicted liquid level fluctuation data is obtained by substituting the real-time argon blowing data into the fitted linear relationship to calculate the predicted value of the current liquid level fluctuation in the crystallizer, in millimeters (mm), which is used to predict the trend of liquid level fluctuation.

[0083] Step S105: In response to the real-time liquid level fluctuation data exceeding the preset normal range, adjust the argon blowing amount at the stopper rod and / or the argon blowing amount at the slide plate to control the real-time liquid level fluctuation data to remain within the preset normal range.

[0084] The preset normal range is the allowable range of crystallizer liquid level fluctuations set according to the slab quality requirements (such as surface cracks and subcutaneous bubble control standards).

[0085] Argon blowing volume adjustment is achieved by controlling the regulating valve (such as electric regulating valve or pneumatic regulating valve) on the argon pipeline to change the argon flow rate at the stopper rod / slide plate. The adjustment range needs to be determined based on the deviation between the predicted liquid level fluctuation and the normal range.

[0086] In this step, the predicted liquid level fluctuation data is compared with the preset normal range: if the predicted value is higher than the upper limit of the normal range (e.g., predicted 827 mm, upper limit 825 mm), it is determined that the argon blowing rate needs to be reduced (e.g., the argon blowing rate at the stopper rod is reduced from 2.5 L / min to 2.0 L / min); if the predicted value is lower than the lower limit of the normal range (e.g., predicted 813 mm, lower limit 815 mm), the argon blowing rate needs to be increased. The adjustment command is sent to the argon regulating valve through the control system. The valve response time is ≤1 second to ensure rapid adjustment of the argon blowing rate, so that the subsequent actual liquid level fluctuation remains within the normal range.

[0087] The above technical solution enables proactive control of instantaneous liquid level fluctuations, significantly reducing the probability of fluctuations exceeding normal ranges, thereby reducing the generation of surface defects on slabs. It also reduces the frequency of manual intervention, lowers the workload of operators, and avoids the lag and errors of manual adjustments, thus improving the stability of continuous casting production.

[0088] like Figure 5 As shown, in one embodiment, optionally, the method further includes: Step S501: Extract the actual real-time liquid level fluctuation data corresponding to the real-time argon blowing data from the slab continuous casting crystallizer; The actual real-time liquid level fluctuation data is the fluctuation value of the molten steel level deviating from the set value at the current moment, which is collected in real time by the liquid level sensor of the crystallizer. It is different from the predicted value based on the fitting model and is the real basis for verifying the accuracy of the model.

[0089] Step S502: Compare the real-time liquid level fluctuation data with the predicted liquid level fluctuation data, and determine whether to update the fitted linear relationship based on the comparison result.

[0090] By comparing predicted and actual values, the accuracy of the fitted model can be intuitively judged, avoiding the model from deviating from actual production conditions. This provides a trigger for subsequent model updates, ensuring that the linear model can be dynamically optimized as production conditions change and maintain high prediction accuracy in the long term.

[0091] In one embodiment, optionally, step S502 includes: Calculate the mean absolute error and coefficient of determination between the real-time liquid level fluctuation data and the predicted liquid level fluctuation data; If the mean absolute error value is greater than or equal to a preset error threshold and the coefficient of determination is greater than or equal to a preset coefficient threshold, it is determined not to update the fitted linear relationship; otherwise, it is determined to update the fitted linear relationship.

[0092] To more intuitively demonstrate the fitting effect, mean absolute error (MAE) and coefficient of determination (COP) were selected. The MAE (mean absolute error) is used to quantitatively evaluate the effectiveness of linear fitting. It represents the average absolute error between predicted and actual values, directly reflecting the average level of prediction error and exhibiting low sensitivity to outliers. [The text then abruptly shifts to a different topic:] ...determine... It is used to measure the proportion of the variance of the dependent variable explained by the model. Its value is between 0 and 1. The closer it is to 1, the better the model fits.

[0093] The above technical solution of the present invention will be described in detail below with reference to a specific embodiment.

[0094] Step 1: Collect on-site continuous casting production data, including liquid level fluctuation data, stopper rod position data, and argon blowing volume at different locations. The steel composition is shown in Table 1, and the relevant parameters of the continuous casting machine are shown in Table 2. The liquid level fluctuation data and related process data are collected at a frequency of 10Hz.

[0095] Table 1. Steel Composition (%)

[0096] Table 2 Parameters of Continuous Casting Machine

[0097] Step 2: Data on crystallizer liquid level fluctuations collected on-site, such as... Figure 6 As shown, the on-site liquid level fluctuation setting is 820mm. (From...) Figure 6 It is known that the fluctuation range of the crystallizer level at certain times far exceeds the normal range. Combined with actual production processes, it is understood that process operations such as ladle changes and speed adjustments can cause significant abnormal fluctuations in the crystallizer level. Therefore, it is necessary to combine the level fluctuation data with the on-site process data for preliminary data cleaning. Furthermore, considering that the one-to-one correspondence between abnormal crystallizer level fluctuations and continuous casting process data is not clear, data processing is performed on the crystallizer level fluctuation data and different process data to explore the relationship between the two. Using the process stabilization time as a benchmark, the average of the ten largest values ​​in the instantaneous abnormal crystallizer level fluctuation data within the same time period is taken as the degree of crystallizer level fluctuation within that time period.

[0098] Step 3: Analyze the relationship between the argon blowing volume at the stopper rod and the slide plate and the crystallizer liquid level fluctuation data based on Pearson correlation coefficient.

[0099] Step 4: To ensure the comprehensiveness and rigor of the research process, Spearman's correlation coefficient was used to analyze the relationship between the argon blowing rate at the stopper rod and the slide plate and the crystallizer liquid level fluctuation data. The results of the Pearson correlation coefficient and Spearman's correlation coefficient were used together as the analytical standard. The analysis results are as follows: Figure 7 As shown.

[0100] Step 5: Based on data collected from the steel plant, a linear formula was fitted to express the relationship between the argon blowing rate and the crystallizer level fluctuation. This data clearly shows how changes in the argon blowing rate affect the crystallizer level fluctuation. Specifically, according to this formula, the maximum fluctuation of the crystallizer level can be calculated from the argon blowing rate at the submersible nozzle. This method provides a powerful tool for predicting and controlling crystallizer level fluctuations. The calculation formula for when the submersible nozzle is blocked is shown in formula (3), where Ml is the maximum fluctuation of the crystallizer level and a is the argon blowing rate.

[0101] (3) Step 6: The field data and the straight line obtained by fitting the formula are as follows: Figure 8 As shown. From Figure 8 It can be seen that the maximum fluctuation of the crystallizer liquid level corresponding to different argon blowing rates differs little from the fitted results, and the field data are evenly distributed around the fitted line, indicating a good fitting effect. This shows that the relationship between argon blowing rate and crystallizer liquid level fluctuation is relatively stable under the condition of immersion nozzle blockage, and can be well described by a linear model. Furthermore, this stability also indicates that in actual production, monitoring the argon blowing rate can accurately predict the fluctuation of the crystallizer liquid level, thus providing a basis for process adjustment.

[0102] Step 7: To more intuitively demonstrate the fitting effect, the mean absolute error (MAE) and coefficient of determination (COP) were selected. To quantitatively evaluate the effectiveness of linear fitting, the comparison results of the evaluation parameters are as follows: Figure 9 As shown. MAE represents the average absolute error between the predicted and actual values. It directly reflects the average level of prediction error and has low sensitivity to outliers. (Decision) It is used to measure the proportion of the variance of the dependent variable explained by the model. Its value is between 0 and 1. The closer it is to 1, the better the model fits.

[0103] Figure 10 A block diagram of a device for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer according to an embodiment of this application is shown.

[0104] like Figure 10 As shown, in a second aspect, embodiments of this application provide a device 1000 for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer, comprising: Extraction module 1001 is used to extract parameter data from the slab continuous casting crystallizer of the target steel plant, wherein the parameter data includes: liquid level fluctuation data, stopper rod position data and argon blowing volume data at different positions; Analysis module 1002 is used to analyze the first relationship between the amount of argon blown at the stopper rod, the amount of argon blown at the slide plate and the liquid level fluctuation data based on Pearson correlation coefficient, and to analyze the second relationship between the amount of argon blown at the stopper rod, the amount of argon blown at the slide plate and the liquid level fluctuation data based on Spearman correlation coefficient. The relationship determination module 1003 is used to determine the fitted linear relationship between the argon blowing volume data and the liquid level fluctuation data based on the first relationship and the second relationship; Prediction module 1004 is used to acquire real-time argon blowing data in the slab continuous casting crystallizer of the target steel plant, and predict real-time liquid level fluctuation data based on the real-time argon blowing data and the fitted linear relationship to obtain predicted liquid level fluctuation data. The control module 1005 is used to adjust the argon blowing amount at the stopper rod and / or the argon blowing amount at the slide plate in response to the real-time liquid level fluctuation data exceeding the preset normal range, so as to control the real-time liquid level fluctuation data to be maintained within the preset normal range.

[0105] In one embodiment, optionally, after extracting the parameter data, the apparatus further includes: A preprocessing module is used to preprocess the parameter data to obtain preprocessed parameter data. The preprocessing process includes: Abnormal fluctuation data caused by changes in process conditions are removed from the parameter data to obtain the initial screening data; Based on the process stabilization time, all abnormal liquid level fluctuation data within the same time period are obtained from the initial screening data and sorted in descending order; From all the abnormal fluctuation data of liquid level, select the top-ranked preset number of target abnormal fluctuation data, and calculate the average value corresponding to all target abnormal fluctuation data. The average value is used as the degree of liquid level fluctuation during the time period.

[0106] In one embodiment, optionally, the analysis module includes: The extraction unit is used to extract the amount of argon blown at the stopper rod and its corresponding first liquid level fluctuation degree, and the amount of argon blown at the slide plate and its corresponding second liquid level fluctuation degree from the preprocessed parameter data, respectively, to obtain two sets of variable pairs; The first calculation unit is used to calculate, based on the Pearson correlation coefficient calculation formula, the first correlation coefficient between the amount of argon blown at the stopper rod and the degree of the first liquid level fluctuation, and the second correlation coefficient between the amount of argon blown at the slide plate and the degree of the second liquid level fluctuation. The first comparison unit is used to compare the first correlation coefficient, the second correlation coefficient, and the first preset coefficient threshold, and to select the first target variable pair with a correlation coefficient greater than the preset coefficient threshold based on the comparison result, so as to form the first effective association dataset.

[0107] In one embodiment, optionally, the analysis module further includes: The sorting unit is used to sort the variable values ​​of the two sets of variable pairs respectively, and assign a unique level order to each variable value. If there are variable values ​​with the same value, the average of the level orders corresponding to the variable values ​​with the same value in the group is taken as their common level. The second calculation unit is used to calculate the grade difference of the corresponding variable values ​​for each pair of variables according to the assigned grade order. The third calculation unit is used to calculate the third correlation coefficient between the amount of argon blown at the stopper rod and the first liquid level fluctuation degree, and the fourth correlation coefficient between the amount of argon blown at the slide plate and the second liquid level fluctuation degree, respectively, based on the grade difference and the Spearman correlation coefficient calculation formula. The second comparison unit is used to compare the third correlation coefficient, the fourth correlation coefficient, and the second preset coefficient threshold, and to select the second target variable pair with a correlation coefficient greater than the second preset coefficient threshold based on the comparison result, so as to form the second effective association dataset.

[0108] In one embodiment, optionally, the relationship determination module includes: The first determining unit is configured to determine the final valid association dataset based on the first valid association dataset and the second valid association dataset. The feature analysis unit is used to perform big data feature analysis on the final effective association dataset and obtain a linear relationship between the argon blowing volume data and the liquid level fluctuation degree by fitting the least squares method.

[0109] In one embodiment, optionally, the apparatus further includes: The extraction module is used to extract the actual real-time liquid level fluctuation data corresponding to the real-time argon blowing data from the slab continuous casting crystallizer; The update module is used to compare the real-time liquid level fluctuation data and the predicted liquid level fluctuation data, and determine whether to update the fitted linear relationship based on the comparison result.

[0110] In one embodiment, optionally, the update module includes: The fourth calculation unit is used to calculate the average absolute error and the coefficient of determination between the real-time liquid level fluctuation data and the predicted liquid level fluctuation data; The second determining unit is configured to determine not to update the fitted linear relationship in response to the mean absolute error value being greater than or equal to a preset error threshold and the determination coefficient being greater than or equal to a preset coefficient threshold; otherwise, it determines to update the fitted linear relationship.

[0111] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for controlling the instantaneous liquid level fluctuation of a slab continuous casting crystallizer.

[0112] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the above-described method for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer.

[0113] It should be noted that those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the instantaneous liquid level fluctuation control device and each module of the slab continuous casting crystallizer described above can be referred to the corresponding process in the aforementioned embodiment of the instantaneous liquid level fluctuation control method for the slab continuous casting crystallizer, and will not be repeated here.

[0114] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the model training device and each module described above can be referred to the corresponding process in the aforementioned embodiment of the instantaneous liquid level fluctuation control method for slab continuous casting crystallizer, and will not be repeated here.

[0115] The aforementioned instantaneous liquid level fluctuation control device for slab continuous casting molds can be implemented as a computer program, which can, for example... Figure 11 It runs on the computer device shown.

[0116] Figure 11 A block diagram of a computer device according to one embodiment of this application is shown.

[0117] See Figure 11 The computer device includes a processor, memory, and network interface connected via a system bus, wherein the memory may include storage media and internal memory.

[0118] The storage medium may store an operating system and a computer program. The computer program includes program instructions that, when executed, cause the processor to perform any of the instantaneous liquid level fluctuation control methods for multi-source data slab continuous casting crystallizers provided in the embodiments of this application.

[0119] The processor provides computing and control capabilities, supporting the operation of the entire computer device.

[0120] Internal memory provides an environment for the execution of computer programs stored in the storage medium. When executed by a processor, these programs can enable the processor to perform methods for analyzing the transmission paths of any infectious disease or training predictive neural networks. The storage medium can be non-volatile or volatile.

[0121] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0122] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0123] In addition, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions for performing the steps described in the first aspect embodiment.

[0124] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or electronic device described above can be referred to the relevant descriptions in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0125] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0126] It should be understood that although the terms "first," "second," etc., may be used to describe the setting units in the embodiments of this application, these setting units should not be limited to these terms. These terms are only used to distinguish the setting units from each other. For example, without departing from the scope of the embodiments of this application, the first setting unit may also be referred to as the second setting unit, and similarly, the second setting unit may also be referred to as the first setting unit.

[0127] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0128] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0129] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in a combination of hardware and software functional units.

[0130] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0131] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer, characterized in that, The method includes: Parameter data are extracted from the slab continuous casting crystallizer of the target steel plant, wherein the parameter data includes: liquid level fluctuation data, stopper rod position data and argon blowing volume data at different positions; The first relationship between the argon blowing volume at the stopper rod and the argon blowing volume at the slide plate and the liquid level fluctuation data was analyzed based on the Pearson correlation coefficient, and the second relationship between the argon blowing volume at the stopper rod and the argon blowing volume at the slide plate and the liquid level fluctuation data was analyzed based on the Spearman correlation coefficient. Based on the first relationship and the second relationship, a fitted linear relationship is determined between the argon blowing volume data and the liquid level fluctuation data; The real-time argon blowing volume data in the slab continuous casting crystallizer of the target steel plant is obtained. Based on the real-time argon blowing volume data and the fitted linear relationship, the real-time liquid level fluctuation data is predicted to obtain the predicted liquid level fluctuation data. In response to the real-time liquid level fluctuation data exceeding the preset normal range, the argon blowing amount at the stopper rod and / or the argon blowing amount at the slide plate is adjusted to control the real-time liquid level fluctuation data to remain within the preset normal range.

2. The method according to claim 1, characterized in that, After extracting the parameter data, the method further includes: The parameter data is preprocessed to obtain preprocessed parameter data, wherein the preprocessing process includes: Abnormal fluctuation data caused by changes in process conditions are removed from the parameter data to obtain the initial screening data; Based on the process stabilization time, all abnormal liquid level fluctuation data within the same time period are obtained from the initial screening data and sorted in descending order; From all the abnormal fluctuation data of liquid level, select the top-ranked preset number of target abnormal fluctuation data, and calculate the average value corresponding to all target abnormal fluctuation data. The average value is used as the degree of liquid level fluctuation during the time period.

3. The method according to claim 2, characterized in that, The first relationship between the argon blowing rate at the stopper rod, the argon blowing rate at the slide plate, and the liquid level fluctuation data was analyzed based on the Pearson correlation coefficient, including: Two sets of variable pairs were obtained by extracting the argon blowing volume at the stopper rod and its corresponding first liquid level fluctuation degree, and the argon blowing volume at the slide plate and its corresponding second liquid level fluctuation degree from the preprocessed parameter data. Based on the Pearson correlation coefficient calculation formula, the first correlation coefficient between the amount of argon blown at the stopper rod and the degree of the first liquid level fluctuation, and the second correlation coefficient between the amount of argon blown at the slide plate and the degree of the second liquid level fluctuation are calculated respectively. The first correlation coefficient, the second correlation coefficient, and the first preset coefficient threshold are compared. Based on the comparison results, the first target variable pair with a correlation coefficient greater than the preset coefficient threshold is selected to form the first effective association dataset.

4. The method according to claim 3, characterized in that, The second relationship between the argon blowing rate at the stopper rod and the argon blowing rate at the slide plate and the liquid level fluctuation data was analyzed based on Spearman correlation coefficient analysis, including: The variable values ​​of the two sets of variable pairs are sorted respectively, and a unique rank order is assigned to each variable value. If there are variable values ​​with the same value, the average of the rank orders corresponding to the variable values ​​with the same value in the group is taken as their common rank. For each pair of variables, calculate the grade difference of the corresponding variable values ​​according to the assigned grade order; Based on the grade difference and Spearman correlation coefficient calculation formula, the third correlation coefficient between the argon blowing amount at the stopper rod and the first liquid level fluctuation degree, and the fourth correlation coefficient between the argon blowing amount at the slide plate and the second liquid level fluctuation degree are calculated respectively. The third correlation coefficient, the fourth correlation coefficient, and the second preset coefficient threshold are compared. Based on the comparison results, the second target variable pairs with correlation coefficients greater than the second preset coefficient threshold are selected to form the second effective association dataset.

5. The method according to claim 4, characterized in that, The step of determining the fitted linear relationship between the argon blowing volume data and the liquid level fluctuation data based on the first relationship and the second relationship includes: The final valid association dataset is determined based on the first valid association dataset and the second valid association dataset; Big data feature analysis was performed on the final effective correlation dataset, and the linear relationship between the argon blowing volume data and the liquid level fluctuation degree was obtained by fitting the least squares method.

6. The method according to claim 1, characterized in that, The method further includes: Extract the actual real-time liquid level fluctuation data corresponding to the real-time argon blowing data from the slab continuous casting crystallizer; The real-time liquid level fluctuation data and the predicted liquid level fluctuation data are compared, and the fitting linear relationship is updated based on the comparison result.

7. The method according to claim 6, characterized in that, Comparing the real-time liquid level fluctuation data with the predicted liquid level fluctuation data, and determining whether to update the fitted linear relationship based on the comparison result, includes: Calculate the mean absolute error and coefficient of determination between the real-time liquid level fluctuation data and the predicted liquid level fluctuation data; If the mean absolute error value is greater than or equal to a preset error threshold and the coefficient of determination is greater than or equal to a preset coefficient threshold, it is determined not to update the fitted linear relationship; otherwise, it is determined to update the fitted linear relationship.

8. A device for controlling instantaneous liquid level fluctuations in a slab continuous casting crystallizer, characterized in that, include: The extraction module is used to extract parameter data from the slab continuous casting crystallizer of the target steel plant, wherein the parameter data includes: liquid level fluctuation data, stopper rod position data and argon blowing volume data at different positions; The analysis module is used to analyze the first relationship between the argon blowing volume at the stopper rod and the argon blowing volume at the slide plate and the liquid level fluctuation data based on the Pearson correlation coefficient, and to analyze the second relationship between the argon blowing volume at the stopper rod and the argon blowing volume at the slide plate and the liquid level fluctuation data based on the Spearman correlation coefficient. The relationship determination module is used to determine the fitted linear relationship between the argon blowing volume data and the liquid level fluctuation data based on the first relationship and the second relationship; The prediction module is used to acquire real-time argon blowing data in the slab continuous casting crystallizer of the target steel plant, and predict real-time liquid level fluctuation data based on the real-time argon blowing data and the fitted linear relationship, thereby obtaining predicted liquid level fluctuation data. The control module is used to adjust the argon blowing volume at the stopper rod and / or the argon blowing volume at the slide plate in response to the real-time liquid level fluctuation data exceeding the preset normal range, so as to control the real-time liquid level fluctuation data to be maintained within the preset normal range.

9. A computer device, characterized in that, include: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores instructions executable by the at least one processor, the instructions being configured to perform the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The device stores computer-executable instructions for performing the method as described in any one of claims 1 to 7.

Images