Intelligent control method and system for zinc layer thickness of continuous hot galvanizing

Abstract

The invention provides an intelligent control method and system for the zinc layer thickness of continuous hot galvanizing, and relates to the technical field of data processing, and the method comprises the steps: collecting the historical production data of a production line, and constructing a zinc layer thickness and air knife pressure prediction model; the two models are trained according to historical data, and a setting table of the distance and height of the air knife is obtained through statistical analysis; acquiring real-time production data, determining air knife parameters according to the setting table, and inputting the air knife parameters into the trained air knife pressure prediction model to obtain an air knife pressure set value; and inputting the real-time data and the air knife set value into the zinc layer thickness prediction model to obtain a first zinc layer thickness prediction value. If the predicted value meets a preset standard, current air knife parameters are stored, otherwise, a feedback adjustment mechanism is used for adjusting air knife pressure, and a second zinc layer thickness predicted value is calculated; if the adjusted predicted value meets the requirement, air knife parameters are stored; if the requirement is not met, the adjusting step is repeated; and intelligent control is carried out through the stored air knife parameters.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to an intelligent control method and system for the zinc layer thickness of continuous hot-dip galvanizing. Background Art

[0002] As a protective material widely used in fields such as automobile manufacturing, refrigeration equipment, building structures, and heating facilities, galvanized strip steel is favored for its excellent corrosion resistance and cost-effectiveness. In the process of hot-dip galvanizing production, the precise control of the zinc layer thickness is the core link to ensure product quality. The thickness and uniformity of the zinc layer directly affect the protective performance and service life of the product. Therefore, how to achieve precise thickness control in hot-dip galvanizing production, especially considering multi-variable disturbances and complex physical effects during the production process, has become a technical problem to be solved urgently.

[0003] Traditional methods for controlling the zinc layer thickness mainly rely on the adjustment of air knife parameters. Factors such as air knife pressure, the distance between the air knife and the strip steel, and the speed of the strip steel jointly determine the final zinc layer thickness. With the increase of complex disturbance factors during the production process, such as fluctuations in zinc liquid composition and temperature, changes in strip steel specifications and shapes, and changes in the production line running speed, there are complex non-linear relationships among these multi-variables, posing great challenges to thickness control.

[0004] However, the existing methods for measuring the zinc layer thickness have a large time lag and cannot achieve real-time and precise feedback, resulting in insufficient control accuracy of the zinc layer thickness. Although the existing technology adjusts the air knife parameters through a model trained based on historical data, the problem of spatio-temporal transformation is not considered during the adjustment process, which may lead to unreasonable adjustment of the air knife parameters, thereby exacerbating the fluctuations in the zinc layer thickness. Summary of the Invention

[0005] In order to solve the technical problems that the existing methods for measuring the zinc layer thickness have a large time lag and cannot achieve real-time and precise feedback, resulting in insufficient control accuracy of the zinc layer thickness. Although the existing technology adjusts the air knife parameters through a model trained based on historical data, the problem of spatio-temporal transformation is not considered during the adjustment process, which may lead to unreasonable adjustment of the air knife parameters, thereby exacerbating the fluctuations in the zinc layer thickness, the present invention provides an intelligent control method and system for the zinc layer thickness of continuous hot-dip galvanizing.

[0006] The technical solutions provided by the embodiments of the present invention are as follows:

[0007] First aspect:

[0008] An intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing provided by the embodiments of the present invention includes:

[0009] S1: Collect historical production data of the production line;

[0010] S2: Construct a prediction model for the thickness of the hot-dip galvanized zinc layer based on the Gaussian process regression algorithm and a prediction model for the air knife pressure based on the random forest algorithm;

[0011] S3: Train the prediction model for the thickness of the hot-dip galvanized zinc layer and the prediction model for the air knife pressure respectively according to the historical production data;

[0012] S4: Conduct statistical analysis on the historical production data to obtain an air knife distance and height setting table;

[0013] S5: Obtain the real-time production data of the production line, and select the corresponding air knife distance setting value and air knife height setting value from the air knife distance and height setting table;

[0014] S6: Input the real-time production data, the air knife distance setting value, and the air knife height setting value into the trained air knife pressure prediction model to output the air knife pressure setting value;

[0015] S7: Input the real-time production data, the air knife distance setting value, the air knife height setting value, and the air knife pressure setting value into the trained prediction model for the thickness of the hot-dip galvanized zinc layer to output the first predicted zinc layer thickness value;

[0016] S8: Determine whether the first predicted zinc layer thickness value meets the preset zinc layer thickness value; if so, save the current air knife parameters to the data temporary storage area; otherwise, proceed to the next step;

[0017] S9: Use the feedback adjustment mechanism of the air knife parameters to adjust the air knife pressure setting value;

[0018] S10: Calculate the second predicted zinc layer thickness value according to the adjusted air knife pressure setting value;

[0019] S11: Determine whether the second predicted zinc layer thickness value meets the preset zinc layer thickness value; if so, save the adjusted air knife parameters to the data temporary storage area; otherwise, repeat S9 to S10 until the predicted zinc layer thickness value calculated meets the preset zinc layer thickness value;

[0020] S12: Perform intelligent control of the zinc layer thickness according to the saved air knife parameters.

[0021] Second aspect:

[0022] An intelligent control system for the thickness of the zinc layer in continuous hot-dip galvanizing provided by an embodiment of the present invention includes:

[0023] A processor;

[0024] A memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing as described in the first aspect is implemented.

[0025] The third aspect:

[0026] A computer-readable storage medium provided by an embodiment of the present invention stores a computer program. When the program is executed by a processor, the intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing as described in the first aspect is implemented.

[0027] The beneficial effects brought by the technical solutions provided by the embodiments of the present invention at least include:

[0028] (1) In the embodiment of the present invention, by collecting the production data of the production line in real time and combining with the air knife distance and height setting table, the air knife distance and height are dynamically selected, and the air knife distance and height and the real-time production data are input into the trained air knife pressure prediction model to output the air knife pressure setting value. It ensures that the feedback mechanism can timely reflect the current production status, thereby reducing the influence brought by measurement lag and improving the real-time performance and accuracy of zinc layer thickness control.

[0029] (2) In the embodiment of the present invention, by constructing a prediction model based on Gaussian process regression and random forest algorithm, the system can more accurately adjust the air knife parameters, consider the non-linear relationship in the production process, and reduce the influence brought by the problem of not considering spatio-temporal transformation in historical data training. The feedback adjustment mechanism comprehensively considers the real-time data and the predicted value during the adjustment process to ensure that the air knife parameter adjustment is more reasonable, thereby avoiding the zinc layer thickness fluctuation caused by unreasonable parameter adjustment. Description of the Drawings

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0031] Figure 1 It is a flowchart of an intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing provided by an embodiment of the present invention;

[0032] Figure 2 It is a structural diagram of an intelligent control system for the zinc layer thickness of continuous hot-dip galvanizing provided by an embodiment of the present invention. Detailed Embodiments

[0033] The following describes the technical solutions in the present invention with reference to the drawings.

[0034] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to give examples, illustrations, or explanations. Any embodiment or design described as an "example" in the present invention should not be construed as being more preferred or having more advantages than other embodiments or designs. Rather, the use of the word "example" is intended to present concepts in a specific manner. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either one of the two.

[0035] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when not emphasizing the difference, the meanings they convey are the same. "Of", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be noted that when not emphasizing the difference, the meanings they convey are the same.

[0036] In the embodiments of the present invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When not emphasizing the difference, the meanings they convey are the same.

[0037] To make the technical problems to be solved, technical solutions, and advantages of the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0038] Referring to the attached Figure 1 , a schematic flowchart of an intelligent control method for the zinc layer thickness in continuous hot-dip galvanizing provided by the embodiments of the present invention is shown.

[0039] The embodiments of the present invention provide an intelligent control method for the zinc layer thickness in continuous hot-dip galvanizing. This method can be implemented by an intelligent control device for the zinc layer thickness in continuous hot-dip galvanizing. The intelligent control device for the zinc layer thickness in continuous hot-dip galvanizing can be a terminal or a server. The processing flow of the intelligent control method for the zinc layer thickness in continuous hot-dip galvanizing can include the following steps:

[0040] S1: Collect historical production data of the production line.

[0041] In a possible implementation manner, the historical production data specifically includes:

[0042] Product specifications and control target data, process parameter data, and working condition state data.

[0043] The product specifications and control target data include: steel grade, coating type, strip thickness, strip speed, strip width, and target coating thickness.

[0044] The process parameter data includes: air knife pressure, air knife distance, and air knife height.

[0045] The working condition status data includes: zinc bath temperature, strip speed, nitrogen pressure, nitrogen flow rate, actual value of the leveling roll, strip tension, and air knife angle.

[0046] In a possible implementation, after S1, it further includes:

[0047] Align the process parameter data with the thickness gauge data through spatio-temporal conversion technology.

[0048] Among them, the goal of the spatio-temporal conversion technology is to align data from different sources by calculating and adjusting the time and space differences between the data, ensuring that it can accurately reflect the production status at the same moment and the same position.

[0049] Specifically, set the time of the i-th record as t i , define the relative displacement distance as S, and calculate the displacement from t i to t i+j moment:

[0050]

[0051] Among them, S represents the relative displacement distance, v(t) is the instantaneous speed of the strip in the time interval [t i ,t i+j , t i represents the i-th moment, and t i+1 represents the (i + 1)-th moment.

[0052] Set the lag displacement of the actual thickness gauge relative to the air knife as d meters. When the relative displacement S satisfies d ≤ S ≤ d + 5, determine the thickness gauge value of the (i + j)-th record as the zinc layer thickness value corresponding to the air knife process parameter of the i-th record. At this time, take d + 5 meters as the maximum allowable relative displacement; if the (i + j)-th record that satisfies the condition d ≤ S ≤ d + 5 cannot be retrieved by cyclic search, then eliminate the i-th data record.

[0053] Optionally, the processed data is used for model training.

[0054] In the present invention, the thickness gauge is usually installed at a position far from the air knife. Therefore, the measured zinc layer thickness data may have a time lag relative to the air knife position. Through spatio-temporal conversion, calculate the relative displacement from the air knife to the thickness gauge, and adjust it in combination with the instantaneous speed of the strip, which can eliminate this lag effect, enabling the thickness gauge data to more accurately reflect the actual influence of the air knife process parameters, thereby improving the control accuracy. At the same time, during the hot-dip galvanizing process, the adjustment of the air knife directly affects the formation of the zinc layer thickness. By aligning the thickness gauge data with the air knife process parameters in terms of time and displacement, it can ensure that each process data is correctly matched with its corresponding zinc layer thickness value. This enables the model to make more accurate predictions and adjustments based on real-time production data.

[0055] S2: Construct a prediction model for the thickness of the hot-dip galvanized zinc layer based on the Gaussian process regression algorithm and a prediction model for the air knife pressure based on the random forest algorithm.

[0056] Among them, Gaussian Process Regression (GPR) is a non-parametric regression method based on a probability model. It is widely used in machine learning and statistics, especially performing well when dealing with data with uncertainty and noise. Gaussian process regression performs function estimation and prediction by modeling the function and providing a confidence interval for each predicted value.

[0057] Among them, Random Forest is an ensemble learning method belonging to the category of supervised learning. It consists of multiple decision trees and makes predictions through "voting" or "averaging".

[0058] It should be noted that the inputs of the prediction model for the thickness of the hot-dip galvanized zinc layer based on the Gaussian process regression algorithm include: air knife pressure (mbar), air knife distance (mm), air knife angle (°), air knife height (mm), strip speed (m / min), strip thickness (mm), and zinc liquid temperature, and the output is the value measured by the front thickness gauge.

[0059] The inputs of the prediction model for the air knife pressure based on the random forest algorithm include: target zinc layer thickness (g / m 2 ), air knife distance (mm), air knife height (mm), air knife angle (°), strip speed (m / min), strip thickness (mm), and zinc liquid temperature (°C), and the output is the air knife pressure (mbar).

[0060] In a possible implementation, the construction method of the prediction model for the thickness of the hot-dip galvanized zinc layer based on the Gaussian process regression algorithm specifically includes:

[0061] S201: Collect input parameters related to the zinc layer thickness.

[0062] S202: Determine the kernel function of the prediction model for the thickness of the hot-dip galvanized zinc layer based on the Gaussian process regression algorithm to effectively capture the non-linear relationship between the input data and the zinc layer thickness:

[0063]

[0064] Among them, k(x,x′) represents the kernel function, α represents the weight coefficient of the RBF kernel, ||x - x′|| 2 represents the square of the Euclidean distance between the input data points x and x′, l represents the length scale of the RBF kernel, β represents the weight coefficient of the linear kernel, T represents matrix transpose.

[0065] S203: Based on the kernel function, through the particle swarm optimization algorithm, with the goal of maximizing the log marginal likelihood function, optimize the hot-dip galvanized zinc layer thickness prediction model based on the Gaussian process regression algorithm, and determine the optimal hyperparameter combination of the hot-dip galvanized zinc layer thickness prediction model.

[0066] S204: Based on the posterior distribution Calculate the zinc layer thickness prediction mean μ * and the confidence interval [μ * - 2σ * , μ * + 2σ * :

[0067]

[0068] where μ * represents the prediction mean, x * represents the input data to be predicted by the model, X represents the training data set, K(x * , X) represents the covariance matrix between the input to be predicted and the training data set, K represents the covariance matrix, represents the noise variance, I represents the identity matrix, y represents the output vector of the training data, represents the prediction variance, k(x * , x * ) represents the covariance between the input to be predicted and itself, () -1 represents the inverse matrix.

[0069] In a possible implementation manner, the construction method of the air knife pressure prediction model based on the random forest algorithm specifically includes:

[0070] S205: Collect the input parameters related to the zinc layer thickness.

[0071] S206: Using the input parameters related to the zinc layer thickness as the input, determine the air knife pressure prediction model of the random forest algorithm by taking the average of the prediction results of multiple decision trees:

[0072]

[0073] where represents the final predicted value, T represents the number of decision trees, y i represents the predicted value of the i-th decision tree.

[0074] S207: Through the particle swarm optimization algorithm, with the goal of minimizing the 5-fold cross-validation mean squared error, optimize the air knife pressure prediction model based on the random forest algorithm, and determine the optimal hyperparameters of the air knife pressure prediction model.

[0075] Optionally, the mean square error is specifically:

[0076]

[0077] where MSE represents the mean square error, n represents the number of samples, and y i represents the actual air knife pressure value, and represents the predicted air knife pressure value.

[0078] S208: Use the optimized air knife pressure prediction model based on the random forest algorithm to output the air knife pressure setting value.

[0079] In the present invention, Gaussian process regression is a non-parametric regression method based on a probability model, which can well handle data with uncertainty and noise. It can not only provide predicted values but also give the confidence interval of the prediction, enabling the model to better quantify the uncertainty of the data. During the hot-dip galvanizing process, considering variable factors such as strip speed and temperature, Gaussian process regression can effectively provide more accurate predictions for the zinc layer thickness and give the confidence interval, which helps with risk management. At the same time, by constructing multiple decision trees and combining voting mechanisms or averages, it can effectively handle non-linear and complex input-output relationships. It is particularly suitable for handling large-scale data and can effectively mitigate the overfitting problem. In air knife pressure prediction, random forest uses the "collective wisdom" of multiple decision trees to provide more stable and reliable predictions.

[0080] S3: According to historical production data, train the hot-dip galvanized zinc layer thickness prediction model and the air knife pressure prediction model respectively.

[0081] S4: Conduct statistical analysis on historical production data to obtain the air knife distance and height setting table.

[0082] It should be noted that the air knife distance and height setting table is obtained by statistically analyzing historical air knife process data and classifying it according to the coating thickness specifications and strip speed intervals.

[0083] In the present invention, by analyzing historical air knife process data, the optimal settings of the air knife distance and height under different coating thickness specifications and strip speeds can be found, which enables the air knife parameters to more precisely adapt to different production conditions and ensures the precise control of the galvanized layer thickness. At the same time, the setting table automatically gives the air knife parameters according to different strip specifications and speeds, and this automated adjustment method can reduce delays and errors in operations. The automated air knife parameter adjustment not only improves production efficiency but also makes the entire production process more intelligent.

[0084] S5: Obtain the real-time production data of the production line, and select the corresponding air knife distance setting value and air knife height setting value from the air knife distance and height setting table.

[0085] Specifically, by obtaining the real-time production data of the production line, such as strip steel specifications and production status information, the timely adjustment of air knife parameters is ensured. Based on these real-time data, the system selects the air knife distance and air knife height setting values that best match the current production conditions from the pre-generated air knife distance and height setting table.

[0086] In the present invention, the real-time acquisition of production data (such as strip steel specifications, speed, etc.) enables the air knife parameters to be adjusted according to the current production status. This can ensure a high degree of matching between the air knife parameters and the actual production conditions, thereby optimizing the zinc layer thickness control. At the same time, the rapid matching of real-time data acquisition and the setting table ensures that the air knife parameters can be adjusted in a timely manner, reducing the quality fluctuations caused by delayed adjustment during the production process. This can respond in real time to any changes during production and ensure the stability of the zinc layer thickness.

[0087] S6: Input the real-time production data, air knife distance setting value, and air knife height setting value into the trained air knife pressure prediction model, and output the air knife pressure setting value.

[0088] In a possible implementation manner, S6 specifically includes:

[0089] S601: When the distance between the weld position of the strip steel head and the air knife position is less than a preset distance during the real-time production process, calculate the target value of the zinc layer thickness using the historical data of strip steel of the same specification.

[0090] It should be noted that those skilled in the art can set the size of the preset distance by themselves, and the present invention does not limit this here.

[0091] S602: According to the production data of the current strip steel and the target value of the zinc layer thickness, determine the air knife distance setting value and air knife height setting value by querying the air knife distance and height setting table.

[0092] S603: Input the real-time production data, air knife distance setting value, and air knife height setting value into the trained air knife pressure prediction model, and output the air knife pressure setting value.

[0093] Specifically, monitor the hot-dip galvanizing production line data in real time. When the weld position of the head of the next coil of strip steel is La meters away from the air knife; calculate the thickness control target using the historical data of strip steel of the same specification, further retrieve the air knife distance and height setting table using the production information of the next coil of strip steel to obtain the air knife distance and air knife height, and further obtain the strip steel specifications and real-time production line status data (width and thickness of the next coil of strip steel and the running speed of the current strip steel) required for model calculation and input them into the air knife pressure prediction model to obtain the air knife pressure.

[0094] In the present invention, the target value of the zinc coating thickness calculated using real-time production data (such as strip specifications and speed) and historical data can accurately predict and adjust the air knife pressure, ensuring that the air knife pressure at each moment during the production process can meet the current production requirements, thereby guaranteeing the accuracy and consistency of the zinc coating thickness. At the same time, through automated prediction based on the trained air knife pressure prediction model, the error of manual adjustment of the air knife pressure is reduced. Manual adjustment is often limited by the experience and judgment of the operator, while the automated model can make accurate predictions driven by data, ensuring the stability of control.

[0095] S7: Input the real-time production data, the set value of the air knife distance, the set value of the air knife height, and the set value of the air knife pressure into the trained hot-dip galvanized zinc coating thickness prediction model, and output the first predicted value of the zinc coating thickness.

[0096] Specifically, input the calculated air knife parameters for the next coil, the corresponding strip specifications, and the production line status data into the zinc coating thickness prediction model to verify the zinc coating thickness accuracy under these parameters.

[0097] In the present invention, by inputting the real-time production data and air knife parameters into the trained hot-dip galvanized zinc coating thickness prediction model, accurate prediction of the zinc coating thickness can be performed based on the current production conditions. This ensures that the zinc coating thickness of each coil of strip can meet the preset standard, improving the accuracy during the production process. At the same time, by predicting and verifying the zinc coating thickness after each adjustment, it is ensured that each adjustment of the air knife parameters can make the zinc coating thickness stably approach the target value. This helps to reduce the quality fluctuations of the product and improve the consistency and uniformity of the zinc coating.

[0098] S8: Determine whether the first predicted value of the zinc coating thickness meets the preset zinc coating thickness value. If so, save the current air knife parameters to the data temporary storage area. Otherwise, proceed to the next step.

[0099] It should be noted that those skilled in the art can set the size of the preset zinc coating thickness value by themselves, and the present invention does not make any limitations here.

[0100] It should be noted that the air knife parameters specifically include: air knife pressure, air knife distance, and air knife height.

[0101] In the present invention, by verifying whether the air knife parameters can meet the expected zinc coating thickness target, the system can ensure that the adjustment of each production link is effective. If the predicted value meets the expectation, the air knife parameters can be safely saved and used for subsequent production, avoiding over-adjustment or unnecessary intervention, and guaranteeing the stability of the production process.

[0102] S9: Use the feedback adjustment mechanism of the air knife parameters to adjust the set value of the air knife pressure.

[0103] Among them, the feedback control mechanism is a control strategy that adjusts input parameters according to the feedback information of the system output to enable the system to reach the desired goal or maintain a stable state. In the hot-dip galvanizing production process, the main function of the feedback control mechanism is to automatically adjust process parameters such as air knives according to the deviation between the actually measured zinc layer thickness and the target thickness to ensure that the zinc layer thickness is maintained within a predetermined range.

[0104] In a possible implementation, the feedback control mechanism specifically includes: feedforward regulation and feedback regulation.

[0105] In a possible implementation, the feedforward regulation specifically includes:

[0106] S901: Read the strip speed during the production process and set the strip speed as the initial locked speed.

[0107] S902: After a first preset time when the current strip passes through the air knife, collect the actual speed of the current strip.

[0108] It should be noted that those skilled in the art can set the size of the first preset time by themselves, and the present invention does not limit it here.

[0109] S903: Calculate the speed change amount according to the initial locked speed and the actual speed.

[0110] S904: Determine whether the speed change amount exceeds the threshold. If so, trigger the feedforward calculation, adjust the air knife pressure, and update the locked speed to the strip speed at the calculation moment. Otherwise, do not adjust the air knife pressure.

[0111] Specifically, set the strip speed read during the control calculation as the initial locked speed v fix . When the strip head passes through the air knife for t ff seconds, start the feedforward calculation, collect the actual speed every s f seconds, and subtract it from the locked speed v fix to obtain Δv. If the speed change amount |Δv| exceeds the threshold Δv min , then trigger the feedforward calculation and adjust the air knife pressure. After triggering the air knife parameter adjustment, update the locked speed to the strip speed at the calculation moment.

[0112] In the present invention, the air knife pressure is adjusted by perceiving the change in the strip speed in advance. This means that when the strip speed changes, the system can react in advance to avoid fluctuations in the zinc layer thickness. This can effectively reduce the negative impact caused by the change in the strip speed, thereby improving the accuracy of zinc layer thickness control. At the same time, through feedforward regulation, the system can immediately perform predictive adjustment when the strip speed changes, thereby eliminating the lag in the feedback process and improving the response speed and stability of the production process.

[0113] In a possible implementation manner, the feedback regulation specifically includes:

[0114] S905: After a second preset time after the current strip passes through the air knife, determine the predicted value of the zinc layer thickness through the hot-dip galvanized zinc layer thickness prediction model.

[0115] It should be noted that those skilled in the art can set the size of the second preset time by themselves, and the present invention does not make a limitation here.

[0116] S906: Calculate the thickness deviation value according to the predicted value of the zinc layer thickness and the target value of the zinc layer thickness.

[0117] S907: Determine whether the thickness deviation value is less than the first preset deviation value. If so, determine that the preset value of the zinc layer thickness meets the control accuracy requirement, and do not adjust the current air knife parameters. Otherwise, determine that the preset value of the zinc layer thickness does not meet the control accuracy requirement, and enter S908.

[0118] S908: Determine whether the thickness deviation value is greater than or equal to the first preset deviation value and less than or equal to the second preset deviation value. If so, calculate the adjusted air knife pressure by reading the efficacy coefficient and combining it with the self-learning value of the air knife pressure, and send the adjusted air knife pressure to the production line. Otherwise, calculate the adjusted air knife distance by reading the efficacy coefficient and combining it with the self-learning value of the air knife distance, and send the adjusted air knife distance to the production line.

[0119] It should be noted that those skilled in the art can set the sizes of the first preset deviation value and the second preset deviation value by themselves, and the present invention does not make a limitation here.

[0120] Specifically, when the strip head passes through the air knife for t ff seconds, start the feedback calculation, call the thickness prediction model and the self-learning module to calculate the predicted value Th of the zinc layer thickness after deviation correction pri , and compare this value with the target Th of the zinc layer thickness tarSubtract to obtain the deviation ΔTh. When |ΔTh| < ΔTh1, it is determined that the zinc layer thickness meets the control accuracy requirements, and no adjustment of the air knife parameters is required. When ΔTh1 < |ΔTh| < ΔTh2, the zinc layer thickness does not meet the control accuracy requirements, and the deviation is small. Start the feedback calculation adjustment. At this time, adopt the strategy of adjusting the air knife pressure: read the efficacy coefficient, calculate the adjusted air knife pressure, and input the adjusted air knife parameters, etc. into the prediction model to obtain a predicted value, and further add the self-learning value to verify whether the predicted thickness after parameter adjustment meets the accuracy requirements; if it meets the requirements, the air knife parameters are issued. When ΔTh > ΔTh2, the zinc layer thickness does not meet the control accuracy requirements, and the deviation is large. Start the feedback calculation adjustment. At this time, adopt the strategy of adjusting the air knife distance: read the efficacy coefficient, calculate the adjusted air knife distance, and input the adjusted air knife parameters, etc. into the prediction model to obtain a predicted value, and further add the self-learning value to verify whether the predicted thickness after parameter adjustment meets the accuracy requirements; if it meets the requirements, the air knife parameters are issued.

[0121] In the present invention, by calculating the thickness deviation value and judging whether it meets the preset accuracy requirements, it can be ensured that the system makes precise adjustments according to the error between the actually measured zinc layer thickness and the target thickness. If the deviation is small, the current air knife parameters are maintained; if the deviation is large, adjustments are made according to the prediction model and the self-learning value, thereby improving the control accuracy and ensuring that each adjustment is based on real production data.

[0122] S10: Calculate the predicted value of the second zinc layer thickness according to the set value of the adjusted air knife pressure.

[0123] S11: Judge whether the predicted value of the second zinc layer thickness meets the preset zinc layer thickness value. If so, save the adjusted air knife parameters in the data temporary storage area. Otherwise, repeat S9 to S10 until the predicted value of the zinc layer thickness calculated meets the preset zinc layer thickness value.

[0124] It should be noted that if the predicted value of the zinc layer thickness still cannot meet the preset zinc layer thickness value after repeatedly adjusting the set value of the air knife pressure multiple times, it is determined by the process personnel through inspection.

[0125] In the present invention, the practice of repeatedly adjusting and verifying ensures the precise control of the zinc layer thickness and the stability of the production process by ensuring that each adjustment of the air knife parameters is verified. The way of gradually optimizing the air knife parameters avoids over-adjustment and improves production efficiency and resource utilization.

[0126] S12: Perform intelligent control of the zinc layer thickness according to the saved air knife parameters.

[0127] Specifically, the system first extracts the verified and adjusted air knife parameters from the data buffer area, and then applies these parameters to the production line to automatically adjust the working state of the air knife. By real-time monitoring the production state of the strip steel and the thickness of the zinc layer, the system ensures that the air knife parameters after each adjustment can stably maintain the zinc layer thickness within the preset target range.

[0128] In the present invention, by real-time monitoring the production state of the strip steel and the thickness of the zinc layer, it is ensured that the air knife parameters after each adjustment can accurately control the zinc layer thickness and maintain it within the preset target range. The fluctuations are reduced, ensuring the consistency of the zinc layer thickness and the stability of the product quality. At the same time, through real-time adjustment and monitoring, the intelligent control system can optimize according to the actual effect after each adjustment, thereby gradually improving the reliability of quality control and ensuring the qualified rate and consistency of the product.

[0129] In a possible implementation manner, after S12, it further includes:

[0130] Before any strip steel enters hot-dip galvanizing, determine whether the layer corresponding to the strip steel specification size has switched. If so, read the inherited self-learning value of the previous coil of strip steel. Otherwise, continue to use the short-term self-learning value of the current layer.

[0131] Obtain the short-term self-learning value and the long-term self-learning value corresponding to the current layer, combine the short-term self-learning value and the long-term self-learning value, and correct and optimize the hot-dip galvanized zinc layer thickness prediction model.

[0132] Specifically, after the head of any strip steel is galvanized, calculate the short-term self-learning value and store the short-term self-learning value in the corresponding layer. After the last coil of strip steel in any galvanizing unit is galvanized, calculate the long-term self-learning value and store the long-term self-learning value in the corresponding layer, and at the same time clear the short-term self-learning value. Before any strip steel enters hot-dip galvanizing, determine whether the strip steel layer has switched. If the layer has switched, read the inherited self-learning value of the previous coil of strip steel. Before any strip steel enters hot-dip galvanizing, obtain the short-term self-learning value and the long-term self-learning value from the corresponding layer, and combine them with the inherited self-learning value to correct and optimize the zinc layer thickness prediction.

[0133] In the present invention, the introduction of the self-learning mechanism enables the system to better adapt to the changes in the strip steel layer and the uncertain factors in production. By combining the short-term and long-term self-learning values, the system can continuously optimize the prediction model, improve the accuracy and stability of zinc layer thickness control. At the same time, this method reduces manual intervention, optimizes the automation and intelligence level of the production process, and improves the overall production efficiency and product quality.

[0134] The beneficial effects brought by the technical solution provided by the embodiment of the present invention at least include:

[0135] (1) In the embodiment of the present invention, by collecting the production data of the production line in real time and combining with the air knife distance and height setting table, the air knife distance and height are dynamically selected, and the air knife distance, height and real-time production data are input into the trained air knife pressure prediction model to output the air knife pressure setting value. This ensures that the feedback mechanism can timely reflect the current production state, thereby reducing the influence caused by measurement lag and improving the real-time performance and accuracy of zinc layer thickness control.

[0136] (2) In the embodiment of the present invention, by constructing a prediction model based on Gaussian process regression and random forest algorithm, the system can more accurately adjust the air knife parameters, consider the non-linear relationship in the production process, and reduce the influence caused by the problem of not considering spatio-temporal transformation in historical data training. The feedback adjustment mechanism comprehensively considers real-time data and predicted values during the adjustment process to ensure that the air knife parameter adjustment is more reasonable, thereby avoiding fluctuations in zinc layer thickness caused by unreasonable parameter adjustment.

[0137] Refer to the attached Figure 2 figures, which show the structural schematic diagram of an intelligent control system for zinc layer thickness in continuous hot-dip galvanizing provided by the present invention.

[0138] The present invention also provides an intelligent control system 20 for zinc layer thickness in continuous hot-dip galvanizing, which is applied to the above-mentioned intelligent control method for zinc layer thickness in continuous hot-dip galvanizing, and includes:

[0139] A processor 201.

[0140] A memory 202, on which computer-readable instructions are stored. When the computer-readable instructions are executed by the processor 201, the intelligent control method for zinc layer thickness in continuous hot-dip galvanizing as in the method embodiment is implemented.

[0141] The intelligent control system 20 for zinc layer thickness in continuous hot-dip galvanizing provided by the present invention can execute the above-mentioned intelligent control method for zinc layer thickness in continuous hot-dip galvanizing and achieve the same or similar technical effects. To avoid repetition, the present invention will not elaborate further.

[0142] The beneficial effects brought by the technical solution provided by the embodiment of the present invention at least include:

[0143] (1) In the embodiment of the present invention, by collecting the production data of the production line in real time and combining with the air knife distance and height setting table, the air knife distance and height are dynamically selected, and the air knife distance, height and real-time production data are input into the trained air knife pressure prediction model to output the air knife pressure setting value. This ensures that the feedback mechanism can timely reflect the current production state, thereby reducing the influence caused by measurement lag and improving the real-time performance and accuracy of zinc layer thickness control.

[0144] (2) In the embodiment of the present invention, by constructing a prediction model based on Gaussian process regression and random forest algorithm, the system can more accurately adjust the air knife parameters, take into account the non-linear relationship in the production process, and reduce the impact brought by the problem of not considering spatio-temporal transformation in historical data training. The feedback adjustment mechanism comprehensively considers real-time data and predicted values during the adjustment process to ensure that the adjustment of the air knife parameters is more reasonable, thereby avoiding fluctuations in the zinc layer thickness caused by unreasonable parameter adjustment.

[0145] It should be understood that the processor in the embodiment of the present invention may be a central processing unit (CPU), and this processor may 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. The general-purpose processor may be a microprocessor or this processor may also be any conventional processor, etc.

[0146] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0147] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0148] It should be understood that the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. Additionally, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship. Specific understanding can be made by referring to the context before and after.

[0149] In the present invention, "at least one" means one or more, and "a plurality" means two or more. "At least one of the following" or its similar expressions refer to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c can be single or plural.

[0150] It should be understood that in various embodiments of the present invention, the magnitudes of the sequence numbers of the above processes do not imply the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0151] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0152] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the devices, apparatuses, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0153] In several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0154] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0155] In addition, the functional units in each embodiment of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0156] When the above-mentioned functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0157] An embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing as described in the method embodiment.

[0158] The computer-readable storage medium provided by the present invention can implement the steps and effects of the intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing in the above method embodiment. To avoid repetition, the present invention will not elaborate further.

[0159] The beneficial effects brought by the technical solution provided by the embodiments of the present invention at least include:

[0160] (1) In the embodiment of the present invention, by collecting the production data of the production line in real time and combining with the air knife distance and height setting table, the air knife distance and height are dynamically selected, and the air knife distance, height, and real-time production data are input into the trained air knife pressure prediction model to output the air knife pressure setting value. It ensures that the feedback mechanism can timely reflect the current production status, thereby reducing the influence brought by measurement lag and improving the real-time performance and accuracy of zinc layer thickness control.

[0161] (2) In the embodiment of the present invention, by constructing a prediction model based on Gaussian process regression and random forest algorithm, the system can more accurately adjust the air knife parameters, consider the non-linear relationship in the production process, and reduce the influence brought by the problem of not considering spatio-temporal transformation in historical data training. The feedback adjustment mechanism comprehensively considers real-time data and prediction values during the adjustment process to ensure that the air knife parameter adjustment is more reasonable, thereby avoiding fluctuations in the zinc layer thickness caused by unreasonable parameter adjustment.

[0162] As described above, this is only a specific implementation of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

[0163] The following points need to be explained:

[0164] (1) The accompanying drawings of the embodiments of the present invention only relate to the structures involved in the embodiments of the present invention, and other structures can refer to the common design.

[0165] (2) For clarity, in the accompanying drawings used to describe the embodiments of the present invention, the thickness of the layer or region is enlarged or reduced, that is, these drawings are not drawn according to the actual scale. It can be understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element can be "directly" on or under the other element or there can be an intermediate element.

[0166] (3) Without conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other to obtain new embodiments.

[0167] As described above, this is only a specific implementation of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. An intelligent control method for the thickness of zinc coating in continuous hot-dip galvanizing, characterized in that, Including: S1: Collect historical production data of the production line; S2: Construct a hot-dip galvanized zinc layer thickness prediction model based on the Gaussian process regression algorithm and an air knife pressure prediction model based on the random forest algorithm; S3: Train the hot-dip galvanized zinc layer thickness prediction model and the air knife pressure prediction model respectively according to the historical production data; S4: Conduct statistical analysis on the historical production data to obtain an air knife distance and height setting table; S5: Obtain the real-time production data of the production line, and select the corresponding air knife distance setting value and air knife height setting value from the air knife distance and height setting table; S6: Input the real-time production data, the air knife distance setting value, and the air knife height setting value into the trained air knife pressure prediction model to output the air knife pressure setting value; S7: Input the real-time production data, the air knife distance setting value, the air knife height setting value, and the air knife pressure setting value into the trained hot-dip galvanized zinc layer thickness prediction model to output the first zinc layer thickness prediction value; S8: Judge whether the first zinc layer thickness prediction value meets the preset zinc layer thickness value; if so, save the current air knife parameters to the data buffer; otherwise, go to the next step; S9: Use the feedback adjustment mechanism of the air knife parameters to adjust the air knife pressure setting value; S10: Calculate the second zinc layer thickness prediction value according to the adjusted air knife pressure setting value; S11: Judge whether the second zinc layer thickness prediction value meets the preset zinc layer thickness value; if so, save the adjusted air knife parameters to the data buffer; otherwise, repeat S9 to S10 until the calculated zinc layer thickness prediction value meets the preset zinc layer thickness value; S12: Perform intelligent control of the zinc layer thickness according to the saved air knife parameters.

2. The intelligent control method for zinc layer thickness of continuous hot-dip galvanizing according to claim 1, characterized in that, The historical production data specifically includes: Product specifications and control target data, process parameter data, and working condition data; The product specifications and control target data include: steel type, coating type, strip thickness, strip speed, strip width, and target coating thickness; The process parameter data includes: air knife pressure, air knife distance, and air knife height; The working condition data includes: zinc liquid temperature, strip speed, nitrogen pressure, nitrogen flow rate, actual value of the leveling roll, strip tension, and air knife angle.

3. The intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing according to claim 2, characterized in that, After S1, it further includes: Align the process parameter data with the thickness gauge data through the spatio-temporal conversion technology.

4. The intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing according to claim 1, wherein, The specific construction method of the hot-dip galvanized zinc layer thickness prediction model based on the Gaussian process regression algorithm specifically includes: S201: Collect input parameters related to the zinc layer thickness; S202: Determine the kernel function of the hot-dip galvanized zinc layer thickness prediction model based on the Gaussian process regression algorithm to effectively capture the non-linear relationship between the input data and the zinc layer thickness; S203: Based on the kernel function, use the particle swarm algorithm to optimize the hot-dip galvanized zinc layer thickness prediction model based on the Gaussian process regression algorithm with the goal of maximizing the log marginal likelihood function, and determine the optimal hyperparameter combination of the hot-dip galvanized zinc layer thickness prediction model; S204: Based on the posterior distribution, calculate the predicted mean and confidence interval of the zinc layer thickness according to the input parameters.

5. The intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing according to claim 1, characterized in that, The construction method of the air knife pressure prediction model based on the random forest algorithm specifically includes: S205: Collect the input parameters related to the zinc layer thickness; S206: Use the input parameters related to the zinc layer thickness as the input, and determine the air knife pressure prediction model of the random forest algorithm by taking the average of the prediction results of multiple decision trees; S207: Through the particle swarm algorithm, with the goal of minimizing the mean square error of 5-fold cross-validation, optimize the air knife pressure prediction model based on the random forest algorithm to determine the optimal hyperparameters of the air knife pressure prediction model; S208: Use the optimized air knife pressure prediction model based on the random forest algorithm to output the air knife pressure setting value.

6. The intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing according to claim 1, wherein The specific content of S6 includes: S601: When, during the real-time production process, the position of the strip head weld is less than the preset distance from the position of the air knife, calculate the target value of the zinc layer thickness using the historical data of the same-specification strip; S602: According to the production data of the current strip and the target value of the zinc layer thickness, determine the air knife distance setting value and the air knife height setting value by querying the air knife distance and height setting table; S603: Input the real-time production data, the air knife distance setting value, and the air knife height setting value into the trained air knife pressure prediction model to output the air knife pressure setting value.

7. The intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing according to claim 6, characterized in that, The feedback adjustment mechanism specifically includes: feedforward adjustment and feedback adjustment; The specific content of the feedforward adjustment includes: S901: Read the strip speed during the production process and set the strip speed as the initial locked speed; S902: After the first preset time when the current strip passes through the air knife, collect the actual speed of the current strip; S903: Calculate the speed change amount according to the initial locked speed and the actual speed; S904: Determine whether the speed change amount exceeds the threshold; if so, trigger feedforward calculation, adjust the air knife pressure, and update the locked speed to the strip speed at the calculation moment; otherwise, do not adjust the air knife pressure.

8. The intelligent control method for the zinc layer thickness of continuous hot-dip galvanizing according to claim 7, wherein The specific content of the feedback adjustment includes: S905: After the second preset time when the current strip passes through the air knife, determine the predicted value of the zinc layer thickness through the hot-dip galvanized zinc layer thickness prediction model; S906: Calculate the thickness deviation value according to the predicted value of the zinc layer thickness and the target value of the zinc layer thickness; S907: Determine whether the thickness deviation value is less than the first preset deviation value; if so, determine that the preset value of the zinc layer thickness meets the control accuracy requirements and do not adjust the current air knife parameters; otherwise, determine that the preset value of the zinc layer thickness does not meet the control accuracy requirements and enter S908; S908: Determine whether the thickness deviation value is greater than or equal to the first preset deviation value and less than or equal to the second preset deviation value; if so, calculate the adjusted air knife pressure by reading the efficacy coefficient and combining it with the air knife pressure self-learning value, and send the adjusted air knife pressure to the production line; otherwise, calculate the adjusted air knife distance by reading the efficacy coefficient and combining it with the air knife distance self-learning value, and send the adjusted air knife distance to the production line.

9. The intelligent control method for the thickness of the zinc layer in continuous hot-dip galvanizing according to claim 1, characterized in that, After the S12, it further includes: After the galvanization of the head of any strip steel is completed, calculate the short-term self-learning value through the self-learning model, and store the short-term self-learning value in the layer corresponding to the strip steel specification size; After the galvanization of the last coil of strip steel in any galvanizing unit is completed, calculate the long-term self-learning value through the self-learning model, store the long-term self-learning value in the layer corresponding to the strip steel specification size, and clear the short-term self-learning value.

10. An intelligent control system for the thickness of zinc coating in continuous hot-dip galvanizing, characterized in that, It includes: A processor; A memory, on which computer-readable instructions are stored. When the computer-readable instructions are executed by the processor, the intelligent control method for the zinc layer thickness of continuous hot-dip galvanization as described in any one of claims 1 to 9 is implemented.

Images