Method for predicting the risk of splashing in a converter and device therefor

Abstract

The present application relates to a converter spitting risk prediction method and device thereof, comprising: monitoring X, Y, Z three-axis orientation oxygen lance vibration acceleration data; determining the spitting risk probability of the converter according to the data; judging whether the spitting risk probability is greater than the set threshold; if yes, predicting that the spitting risk will occur; if no, predicting that the spitting risk will not occur. It gives a new spitting risk prediction mode, taking X, Y, Z three-axis orientation oxygen lance vibration acceleration data as the monitoring object to determine the spitting risk probability of the converter; compared with single direction oxygen lance amplitude, sound intensity change, converter surface image and other parameters, the X, Y, Z three-axis orientation oxygen lance vibration acceleration can reflect the real situation in the furnace in all directions, especially the oxygen lance vibration acceleration feature, which can more truly reflect the pressure in the furnace and further improve the prediction accuracy and timeliness of the spitting risk probability.

Description

Technical Field

[0001] This invention relates to the field of converter steelmaking, and in particular to a method for predicting the risk of splashing in converters. Background Technology

[0002] Oxygen top-blown converters have become a major type of steelmaking equipment due to their high production efficiency and low cost. Slag formation is a critical process in converter steelmaking. The smoothness of the slag formation process directly affects the quality of the steel and the efficiency of steelmaking. Furthermore, if splashing occurs during slag formation, it will not only cause serious waste of raw materials, but may even lead to accidents such as personnel injuries and equipment damage.

[0003] Therefore, to ensure the smooth progress of slag treatment and avoid splashing accidents, existing technologies employ two methods: First, manual observation. However, this method is limited by experience and skill level, which can lead to instability and inaccuracy of the detection results, and is also inefficient and time-delayed. Second, monitoring oxygen lance amplitude, sound intensity changes, image analysis, and the buoyancy of the oxygen lance can be used to predict splashing risks. However, this method reflects the slag thickness of the converter, and the prediction effect is not good and needs improvement. Summary of the Invention

[0004] To address one of the aforementioned technical problems, this invention provides a method for predicting splash risk in a converter, comprising: S1: Monitor the vibration acceleration data of the oxygen lance in the X, Y, and Z axes; S2: Determine the probability of splash risk in the converter based on the oxygen lance vibration acceleration data in the X, Y, and Z axes; S3: Determine if the probability of splash risk is greater than the set threshold; S4: If so, then a splashing risk is predicted; S5: If not, then no risk of splashing is predicted.

[0005] Furthermore, before step S1, step S0 is included: determining the sampling frequency of the oxygen lance vibration acceleration data monitoring the X, Y, and Z axes, including: 1a: Obtain the current oxygen lance vibration acceleration; 1b: Calculate the current vibration frequency of the molten slag foam based on the current oxygen lance vibration acceleration; 1c: Set the sampling frequency for monitoring oxygen lance vibration acceleration data based on the current vibration frequency of the molten slag foam.

[0006] Furthermore, the step of determining the sampling frequency also includes: Set the sampling frequency to determine the period T. After one period T ends, recursively execute steps 1a-1c to redetermine the sampling frequency based on the current oxygen lance vibration acceleration.

[0007] Further, step S2 includes: Based on the oxygen lance vibration acceleration data in the X, Y, and Z axes at each time point, the total vibration acceleration of the oxygen lance at each time point is determined by vector synthesis. Calculate the ratio of the current value to the trend value of the total acceleration of the oxygen lance vibration at each moment, and use this ratio as the probability of splash risk in the converter.

[0008] Further, step S2 includes: A prediction model was constructed and trained, which included the correlation between oxygen lance vibration acceleration data in the X, Y, and Z axes and furnace pressure and slag over-foaming degree. Input the current monitored oxygen lance vibration acceleration data in the X, Y, and Z axes into the prediction model, and output the current furnace pressure and the current degree of slag over-foaming. The probability of splashing risk inside the furnace is determined based on the furnace pressure and the current degree of slag over-foaming.

[0009] Furthermore, the prediction model includes: Input module: Input oxygen lance vibration acceleration data in the X, Y, and Z axes; The overall extraction module extracts the characteristics of furnace pressure and slag over-foaming degree based on the oxygen lance vibration acceleration data in the XYZ orientation. The output module outputs the current furnace pressure and the degree of slag over-foaming based on the characteristics of the furnace pressure and the degree of slag over-foaming. Calculate the splash risk probability according to formula (1).

[0010] P = f(at) / f(Ft) (1); Where at is the current vibration acceleration value; Ft is the predicted value of the furnace pressure and slag over-foaming degree at the next time; and P is the probability of splashing risk.

[0011] Furthermore, the prediction model includes: Input module: Input oxygen lance vibration acceleration data in the X, Y, and Z axes; The first extraction module extracts the furnace pressure characteristics based on the oxygen lance vibration acceleration data in the XY direction; the second extraction module extracts the slag over-foaming characteristics based on the oxygen lance vibration acceleration data in the Z direction. The output module outputs the current furnace pressure based on the furnace pressure characteristics; and outputs the degree of slag over-foaming based on the slag over-foaming characteristics. Calculate the splash risk probability according to formula (2); P=K1C1+K2C2(2) Where K1 and K2 are the weighting coefficients of the current furnace pressure and the current degree of slag over-foaming, respectively; C1 and C2 are the current furnace pressure and the current degree of slag over-foaming, respectively; and P is the probability of splashing risk.

[0012] Furthermore, it also includes: S6: shielding step, when a splash risk is predicted, it determines whether events such as feeding or oxygen lance raising or lowering have occurred, and if so, it shields the splash risk determination result.

[0013] Furthermore, it also includes: S7 automatic control steps: based on the splash risk prediction results, if a splash risk is predicted to occur, take any one or more of the following operations: pressurize the spray, reduce the oxygen flow rate of the oxygen lance, and reduce the height of the oxygen lance.

[0014] On the other hand, the present invention also provides a converter splash risk prediction device for performing any of the above-mentioned splash risk prediction methods, comprising: a monitoring module for monitoring oxygen lance vibration acceleration data in the X, Y, and Z axes; a determination module for determining the splash risk probability of the converter based on the oxygen lance vibration acceleration data in the X, Y, and Z axes; a judgment module for judging whether the splash risk probability is greater than a set threshold; and a prediction module for predicting whether a splash risk will occur.

[0015] This invention provides a method and apparatus for predicting splash risk in a converter, presenting a novel splash risk prediction mode. It uses oxygen lance vibration acceleration data in the X, Y, and Z axes as the monitoring object to determine the predicted probability of splash risk in the converter. Compared to parameters such as oxygen lance amplitude, sound intensity changes, and converter surface images in a single direction, the oxygen lance vibration acceleration in the X, Y, and Z axes can comprehensively reflect the actual situation inside the furnace. In particular, the oxygen lance vibration acceleration characteristic can more accurately reflect the pressure inside the furnace, further improving the accuracy and timeliness of splash risk probability prediction. Attached Figure Description

[0016] Figure 1 A flowchart illustrating an embodiment of the converter splash risk prediction method of the present invention; Figure 2 This is a front view of the converter of the present invention; Figure 3 This is a top view of the converter of the present invention; Figure 4 This is an enlarged view of the sensor installation in the converter of the present invention; Figure 5 This is a test diagram of an embodiment of the converter splash risk prediction method of the present invention; Figure 6 This is a schematic diagram of an embodiment of the automatic control steps of the converter splash risk prediction method of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] It should be noted that if the embodiments of the present invention involve directional indications, such as up, down, left, right, front, back, etc., these directional indications are only used to explain the relative positional relationships and movement of the components in a specific posture. If the specific posture changes, the directional indications will also change accordingly. Furthermore, if the embodiments of the present invention involve descriptions such as "first," "second," "S1," "S2," "step one," "step two," etc., these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance, or implicitly indicating the number of technical features indicated or the execution order of the method. Those skilled in the art will understand that anything that does not violate the inventive concept should be included within the scope of protection of the present invention.

[0019] like Figure 1 As shown, the present invention provides a method for predicting splash risk in a converter, characterized in that it includes: S1: Monitor the vibration acceleration data of the oxygen lance in the X, Y, and Z axes; S2: Determine the probability of splash risk in the converter based on the oxygen lance vibration acceleration data in the X, Y, and Z axes; S3: Determine if the probability of splash risk is greater than the set threshold; S4: If so, then a splashing risk is predicted; S5: If not, then no risk of splashing is predicted.

[0020] This embodiment presents a method for predicting the splash risk of a converter according to the present invention. It introduces a novel splash risk prediction mode, using oxygen lance vibration acceleration data in the X, Y, and Z axes as the monitoring object to determine the predicted splash risk probability of the converter. Optionally, but not limited to, sensors such as crystal oscillating accelerometers can be installed on the oxygen lance bracket to detect the three-dimensional vibration acceleration of the oxygen lance. Compared to parameters such as oxygen lance amplitude, sound intensity changes, and converter surface images in a single direction, the oxygen lance vibration acceleration in the X, Y, and Z axes can comprehensively reflect the actual situation inside the furnace. In particular, the oxygen lance vibration acceleration characteristic can more accurately reflect the furnace pressure. Studies have shown that the oxygen lance vibration acceleration is positively correlated with the furnace pressure, which can further improve the prediction accuracy and timeliness of the splash risk probability.

[0021] Specifically: In step S1, preferably, such as Figure 2 As shown, to avoid the impact of production operations such as oxygen lance replacement and equipment maintenance on the detection of oxygen lance vibration acceleration data, one or more acceleration sensors can be optionally, but not limited to, installed on the oxygen lance lifting trolley. This ensures continuous and stable detection of oxygen lance vibration acceleration data in the X, Y, and Z axes. Specifically, as... Figure 2-4 As shown, the vibration acceleration of the oxygen lance can be detected, but is not limited to, by an online vibration test sensor 1, or, but is not limited to, by a protective cover 2 and a welding plate 3 mounted on the oxygen lance lifting trolley platform 4. More preferably, after data acquisition, data preprocessing steps, including but not limited to filtering, may also be performed.

[0022] More preferably, to ensure the accuracy of the oxygen lance vibration acceleration data and avoid interference from noise, thereby enabling real-time analysis of the converter's internal conditions and further improving the accuracy and timeliness of the subsequent early warning function for potential splashing during the smelting process, the monitoring of the oxygen lance vibration acceleration data may optionally, but is not limited to, setting an appropriate sampling frequency. Specifically, before step S1, step S0 is included: determining the sampling frequency for monitoring the oxygen lance vibration acceleration data in the X, Y, and Z axes. Specifically, the step of determining the sampling frequency may optionally, but is not limited to, including: 1a: Obtain the current oxygen lance vibration acceleration; for example, at time t, the current oxygen lance vibration acceleration is collected through the vibration acceleration sensor; more specifically, for ease of calculation, the current oxygen lance vibration acceleration can be the oxygen lance vibration acceleration data in a certain direction, or it can be the vector sum of the oxygen lance vibration acceleration data in the X, Y, and Z axes.

[0023] 1b: Calculate the current vibration frequency of the molten slag foam based on the current oxygen lance vibration acceleration; for example, use Fourier transform to perform frequency analysis on the vibration acceleration signal for the current oxygen lance vibration acceleration, determine the vibration frequency related to the furnace pressure and the degree of slag over-foaming based on the spectrum diagram, and make a time-frequency diagram through inverse transform to obtain the current vibration frequency of the molten slag foam. 1c: Set the sampling frequency for monitoring the oxygen lance vibration acceleration data based on the current vibration frequency of the molten slag foam; for example, to avoid noise interference, the sampling frequency for the oxygen lance vibration acceleration data can be selected, but is not limited to, being 4-5 times the current vibration frequency of the molten slag foam.

[0024] In step S2, two preferred embodiments are given below, but they are not limited thereto. This invention is based on theoretical research on the positive correlation between oxygen lance vibration acceleration and the probability of splashing risk in converters. All technical solutions guided by this technical concept should be included in the protection scope of this invention.

[0025] In the first embodiment, optional, but not limited to, the following may be included: 2a. Based on the oxygen lance vibration acceleration data of the X, Y, and Z axes at each time point, the total vibration acceleration of the oxygen lance at each time point is determined by vector synthesis. 2b. Calculate the ratio of the current value to the trend value of the total vibration acceleration of the oxygen lance at each moment, and use this ratio as the probability of splash risk in the converter. Specifically, for the current value, at any moment t, the current value is obtained by vector summation of the monitored XYZ three-axis azimuth oxygen lance vibration acceleration data at that moment. For the trend value, any method can be selected, but is not limited to, the extended time interval method, the moving average method, or the least squares method to determine the trend value over a period of time. For example, to characterize the severity of splash events and the correlation with vibration acceleration, the trend value is calculated once every 10 sampling points, such as... Figure 5 As shown, the current value curve (real-time line) and trend value curve (trend line) of the total acceleration of the oxygen lance vibration are obtained. Then, the ratio of the two at time t is calculated to reflect the splash risk index, thus determining the splash risk probability of the converter. For example, the ratio of the current value to the trend value of the oxygen lance vibration acceleration at each time point measures the deviation from the normal range. This can be, but is not limited to, vibrations within the upper and lower limits of the trend value. A larger ratio is defined as a larger splash risk index, indicating a higher probability of splashing, thus establishing a splash risk probability model. For example, a splash risk is predicted when the deviation from the normal range is 40%.

[0026] In the second embodiment, optional, but not limited to, the following are included: 2a'. Construct and train a prediction model that includes the correlation between oxygen lance vibration acceleration data in the X, Y, and Z axes and the furnace pressure and slag over-foaming degree. More specifically, the prediction model includes the correlation between oxygen lance vibration acceleration in the X and Y axes and the furnace pressure, and the correlation between oxygen lance vibration acceleration data in the Z axis and the slag over-foaming degree. This method decouples the oxygen lance vibration acceleration data in the X, Y, and Z axes, using the horizontal oxygen lance vibration acceleration in the X and Y axes to reflect the horizontal furnace pressure, and the vertical oxygen lance vibration acceleration in the Z axis to reflect the vertical slag over-foaming degree. This can more accurately reflect the actual situation of the converter and further improve the prediction accuracy and timeliness of splash risk.

[0027] Specifically, the correlation between the data in the prediction model can be determined by fitting data based on 100 smelting batches, but is not limited to this method. Preferably, it can be determined based on any form of neural network model.

[0028] For example, first, a predictive model is constructed. The specific form of this predictive model can be, but is not limited to, any neural network model employing existing techniques. The structure includes: Input module: Input oxygen lance vibration acceleration data in the X, Y, and Z axes; The extraction module extracts furnace pressure characteristics and slag over-foaming characteristics based on the oxygen lance vibration acceleration data in the XYZ axes. Specifically, lightweight backbone networks such as ResNet, MobileNet, and ShuffleNet are optional, but not limited to, a smaller 18-layer ResNet network. More specifically, it may also include, but is not limited to, the following: a first extraction module extracts furnace pressure characteristics based on the oxygen lance vibration acceleration data in the XY axes; and a second extraction module extracts slag over-foaming characteristics based on the oxygen lance vibration acceleration data in the Z axis. The output module outputs the current furnace pressure and slag over-foaming degree values ​​based on the furnace pressure characteristics and slag over-foaming degree characteristics.

[0029] Then, from the smelting data of 100 heats, 80 heats were selected as the training sample set and 20 heats as the validation sample set. These were then sequentially input into the aforementioned neural network-based prediction model. The loss function was iteratively reduced to optimize and correct the various parameters of the prediction model, resulting in the trained neural network model, i.e., the prediction model, which reflects the correlation between the oxygen lance vibration acceleration data in the X, Y, and Z axes and the furnace pressure and the degree of slag over-foaming.

[0030] 2b' Input the currently monitored oxygen lance vibration acceleration data in the X, Y, and Z axes into the prediction model, and output the current furnace pressure and the current degree of slag over-foaming; 2c' Determine the probability of splashing risk inside the furnace based on the furnace pressure and the current degree of slag over-foaming.

[0031] Specifically, if the overall characteristics of the furnace pressure and the degree of slag over-foaming are extracted based on the oxygen lance vibration acceleration in the XYZ three-axis orientation, the splash risk probability is calculated according to formula (1).

[0032] P = f(at) / f(Ft) (1); Where at is the current vibration acceleration value; Ft is the predicted value of the furnace pressure and slag over-foaming degree at the next time; and P is the probability of splashing risk.

[0033] If the furnace pressure characteristics are extracted based on the oxygen lance vibration acceleration data in the XY direction, and the slag over-foaming characteristics are extracted based on the oxygen lance vibration acceleration data in the Z direction, the weighted coefficient summation method can be selected, but is not limited to, to calculate the splash risk probability according to formula (2).

[0034] P=K1C1+K2C2(2) Wherein, K1 and K2 are weighting coefficients, C1 and C2 are the current furnace pressure and the current degree of slag over-foaming, and P is the probability of splashing risk. Specifically, the weighting coefficients K1 and K2 can be arbitrarily set by those skilled in the art based on the actual conditions of the converter.

[0035] In the above preferred embodiment, a specific implementation method for predicting splash risk in the converter of the present invention is given. It takes the oxygen lance vibration acceleration data in the X, Y, and Z axes as a whole, and vector synthesizes the total radial vector of oxygen lance vibration acceleration to reflect the three-dimensional spatial state inside the furnace in all directions. The ratio of the current value to the trend value of the oxygen lance vibration acceleration at each moment is used as the evaluation standard, i.e., the splash risk probability, rather than simply using the current value of the oxygen lance vibration acceleration as the direct evaluation standard. It can fully reflect the rapid change process of the state inside the furnace, such as the change process of furnace pressure and slag over-foaming degree. Based on this, it is determined whether the splash risk probability is greater than a set threshold. If it is, the splash risk is predicted to occur; if not, the splash risk is predicted not to occur. It can predict splashing phenomena in advance, achieve early warning, and has higher prediction accuracy.

[0036] More preferably, based on several smelting data analyses, a positive correlation was found between the oxygen lance vibration acceleration signal and the furnace pressure and slag over-foaming degree. A correlation model between these two factors can be constructed, but is not limited to, to accurately reflect the furnace pressure and slag over-foaming degree. Compared to methods using oxygen lance amplitude or converter surface images, this approach more fully reflects the root cause of splashing. Specifically, the fundamental cause of splashing is that the expansion rate of the gas inside the furnace exceeds the discharge rate, leading to a continuous increase in furnace pressure and pushing the slag upwards. The phenomenon is that the slag rises rapidly, exceeding the furnace opening height, resulting in splashing. If the expansion rate and discharge rate of the gas inside the furnace are equal, the furnace pressure and slag height will be in equilibrium, and splashing will not occur even if the slag height increases. Therefore, existing technologies assume that increased slag height leads to splashing (thick slag has high damping, resulting in low sound intensity, small amplitude, and high buoyancy). Using methods that detect sound intensity, amplitude, and buoyancy, and taking slag thickness (i.e., slag height) as the detection object to predict splashing, this approach is actually inaccurate. In reality, a thick layer of slag doesn't necessarily cause splashing. It's like cooking noodles at home: when the pot boils, gas is produced, and foam rises. If the gas flame is suitable, the rate of bubble formation and the rate of gas release are equal, so even if the foam reaches the rim of the pot after boiling, it won't overflow. Similarly, a thin layer of slag doesn't necessarily prevent splashing. Again, like cooking noodles at home, if the flame isn't well controlled, a large number of bubbles will be produced after the pot boils, and it will overflow the moment it boils. Therefore, using slag height as a detection target can only provide an alarm, not a warning, and thus cannot allow for timely control measures to prevent splashing. The innovation of this invention lies in using the vibration acceleration of the oxygen lance as the monitoring object to indirectly reflect the furnace pressure and slag over-foaming. It addresses the root cause of splashing and reflects the probability of splashing risk. In a preferred embodiment, the actual value and trend value of the oxygen lance vibration acceleration are compared. More preferably, the probability of splashing risk is used as the evaluation index to establish a positive correlation between vibration acceleration, furnace pressure, and the degree of slag over-foaming. By detecting and calculating vibration acceleration, the trend of changes in furnace pressure and slag over-foaming can be predicted, allowing for a splashing risk forecast 15 to 60 seconds in advance, giving users sufficient time to take measures to curb splashing.

[0037] More preferably, in steps S3-S5, it is determined whether the probability of splashing risk is greater than a set threshold. Specifically, the set threshold can be determined based on the size, model, and amount of slag in the converter. If it is, then it is predicted that splashing risk will occur; if not, then it is predicted that splashing risk will not occur.

[0038] More preferably, the converter splash risk prediction method of the present invention may also include, but is not limited to, step S6: shielding step, in which, when a splash risk is predicted to occur, it is determined whether events such as feeding or oxygen lance raising or lowering occur, and if so, the splash risk determination result is shielded.

[0039] In this embodiment, a splash risk prediction method for the converter of the present invention is provided. Considering that changes in the amount of molten steel and slag, oxygen lance raising and lowering, and charging events can affect the detection results of the oxygen lance vibration acceleration data, the monitoring of oxygen lance vibration and velocity data based on these events is not accurate. Therefore, the splash risk prediction results obtained based on these events may not be accurate. Thus, a shielding step is added. When events such as charging or oxygen lance raising and lowering occur, even if a splash risk is predicted, it is shielded to avoid false warnings and subsequent incorrect operations. This further improves the accuracy of the splash risk prediction method, enabling early warning of splashing 15 to 50 seconds in advance.

[0040] More preferably, the converter splash risk prediction method of the present invention may also optionally include, but is not limited to, the following: S7 automatic control step: based on the splash risk prediction result, if a splash risk is predicted to occur, take any one or more of the following operations: pressurized material injection, reduce oxygen lance flow rate, reduce oxygen lance height, etc.

[0041] Specifically, but not limited to, further setting control levels based on the probability of splashing risk, and dynamically adjusting processes such as oxygen lances and charging, can mitigate splashing risk and prevent splashing from occurring in the converter. For example, ... Figure 6 As shown, a preferred embodiment of automatic control is provided. It determines whether a splash warning has occurred. If no splash risk warning has occurred, everything proceeds as normal, and monitoring and judgment continue. If a splash warning has occurred, the oxygen lance can be lowered for low-risk warnings, material can be added while the oxygen lance is lowered for medium-risk warnings, and the flow rate can be reduced for high-risk warnings. The specific values ​​can be arbitrarily set by those skilled in the art based on the actual situation of the converter. The system continuously monitors and judges whether the warning value is normal. If not, it continues to adjust. If it is, the lance position is returned to its original position.

[0042] On the other hand, the present invention also provides a converter splash risk prediction device to execute the above-mentioned splash prediction method, comprising: a monitoring module for monitoring oxygen lance vibration acceleration data in the X, Y, and Z axes; a determination module for determining the splash risk probability of the converter based on the oxygen lance vibration acceleration data in the X, Y, and Z axes; a judgment module for judging whether the splash risk probability is greater than a set threshold; and a prediction module for predicting whether a splash risk will occur. Preferably, the monitoring module may be, but is not limited to, a vibration acceleration sensor, and is preferably installed on the oxygen lance lifting trolley. Compared with placing the sensor under the oxygen lance or bracket, it has a longer service life, is not affected by oxygen lance replacement, and has higher monitoring accuracy.

[0043] Specifically, the prediction device is created based on the above-mentioned prediction method, and its combination of technical effects and technical features will not be elaborated here.

[0044] On the other hand, the present invention also provides a converter that employs any of the above-mentioned splash risk control methods or includes any of the above-mentioned splash risk control devices.

[0045] On the other hand, the present invention also provides a computer storage medium storing executable program code; the executable program code is used to execute any of the above-mentioned splash risk control methods.

[0046] On the other hand, the present invention also provides a terminal device, including a memory and a processor; the memory stores program code that can be executed by the processor; the program code is used to execute any of the above-mentioned splash risk control methods.

[0047] For example, the program code can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the program code in the terminal device.

[0048] The terminal device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that the terminal device may also include input / output devices, network access devices, buses, etc.

[0049] The processor can be a Central Processing Unit (CPU), or 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. A general-purpose processor can be a microprocessor or any conventional processor.

[0050] The memory can be an internal storage unit of the terminal device, such as a hard drive or RAM. The memory can also be an external storage device of the terminal device, such as a plug-in hard drive, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory can include both internal and external storage units of the terminal device. The memory is used to store the program code and other programs and data required by the terminal device. The memory can also be used to temporarily store data that has been output or will be output.

[0051] The above-mentioned splash risk prediction device, computer storage medium and terminal equipment are created based on the above-mentioned splash risk prediction method. Their technical functions and beneficial effects will not be repeated here. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0052] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for predicting splash risk in a converter, characterized in that, include: S1: Monitor the vibration acceleration data of the oxygen lance in the X, Y, and Z axes; S2: Determine the probability of splash risk in the converter based on the oxygen lance vibration acceleration data in the X, Y, and Z axes; S3: Determine whether the probability of splash risk is greater than the set threshold. S4: If yes, then a splash risk is predicted; S5: If no, then a splash risk is predicted. Step S2 includes: constructing and training a prediction model, which includes the correlation between oxygen lance vibration acceleration data in the X, Y, and Z axes and the furnace pressure and slag over-foaming degree; inputting the currently monitored oxygen lance vibration acceleration data in the X, Y, and Z axes into the prediction model, and outputting the current furnace pressure and the current slag over-foaming degree; determining the probability of splashing risk in the furnace based on the furnace pressure and the current slag over-foaming degree. The prediction model includes: an input module for inputting oxygen lance vibration acceleration data in the X, Y, and Z axes; a total extraction module for extracting furnace pressure and slag over-foaming characteristics based on the oxygen lance vibration acceleration data in the X, Y, and Z axes; an output module for outputting the current furnace pressure and slag over-foaming characteristics based on the furnace pressure and slag over-foaming characteristics; and a splash risk probability calculated according to formula (1). P = f(at) / f(Ft) (1) Where at is the current vibration acceleration value; Ft is the predicted value of the furnace pressure and slag over-foaming degree at the next time; and P is the probability of splashing risk. Alternatively, the prediction model includes: an input module for inputting oxygen lance vibration acceleration data in the X, Y, and Z axes; a first extraction module for extracting furnace pressure characteristics based on the oxygen lance vibration acceleration data in the X and Y axes; a second extraction module for extracting slag over-foaming characteristics based on the oxygen lance vibration acceleration data in the Z axis; an output module for outputting the current furnace pressure based on the furnace pressure characteristics; outputting the degree of slag over-foaming based on the slag over-foaming characteristics; and calculating the splash risk probability according to formula (2). P=K1C1+K2C2(2) Where K1 and K2 are the weighting coefficients of the current furnace pressure and the current degree of slag over-foaming, respectively; C1 and C2 are the current furnace pressure and the current degree of slag over-foaming, respectively; and P is the probability of splashing risk.

2. The splatter risk prediction method of claim 1, wherein, Before step S1, step S0 is also included: determining the sampling frequency of the oxygen lance vibration acceleration data monitoring the X, Y, and Z axes, including: 1a: Obtain the current oxygen lance vibration acceleration; 1b: Calculate the current vibration frequency of the molten slag foam based on the current oxygen lance vibration acceleration; 1c: Set the sampling frequency for monitoring oxygen lance vibration acceleration data based on the current vibration frequency of the molten slag foam.

3. The splatter risk prediction method of claim 2, wherein, The steps for determining the sampling frequency also include: Set the sampling frequency to determine the period T. After one period T ends, recursively execute steps 1a-1c to redetermine the sampling frequency based on the current oxygen lance vibration acceleration.

4. The splatter risk prediction method of claim 1, wherein, Step S2 includes: Based on the oxygen lance vibration acceleration data in the X, Y, and Z axes at each time point, the total vibration acceleration of the oxygen lance at each time point is determined by vector synthesis. Calculate the ratio of the current value to the trend value of the total acceleration of the oxygen lance vibration at each moment, and use this ratio as the probability of splash risk in the converter.

5. The splatter risk prediction method of claim 1, wherein, It also includes: S6: shielding step, when a splash risk is predicted, it determines whether events such as feeding or oxygen lance raising or lowering have occurred, and if so, it shields the splash risk assessment result.

6. The splatter risk prediction method according to any one of claims 1 to 5, characterized in that Also includes: S7 Automatic Control Procedure: Based on the splash risk prediction results, if a splash risk is predicted to occur, take any one or more of the following actions: pressurize the spray, reduce the oxygen flow rate of the oxygen lance, or reduce the height of the oxygen lance.

7. A spitting risk prediction device for a converter, characterized by The method for performing the splash risk prediction method according to any one of claims 1-6 includes: a monitoring module for monitoring oxygen lance vibration acceleration data in the X, Y, and Z axes; a determination module for determining the splash risk probability of the converter based on the oxygen lance vibration acceleration data in the X, Y, and Z axes; a judgment module for judging whether the splash risk probability is greater than a set threshold; and a prediction module for predicting whether a splash risk will occur.

Images