KR20200034481A - Coating control device and method - Google Patents

Abstract

In a continuous plating process in which a strip is dipped into a molten metal pot and coated, an apparatus for controlling a plating amount coated on the strip using an air knife disposed along the traveling direction of the strip includes: a prediction model unit including a prediction model which has neural network-learned accumulated operating conditions; and an optimum air knife condition calculation unit for deriving an absolute value of an air knife pressure based on the input operating condition, and deriving an absolute value of an air knife gap using the derived absolute value of the air knife pressure and the prediction model.

Description

Plating amount control device and plating amount control method {COATING CONTROL DEVICE AND METHOD}

The present disclosure relates to a plating amount control device and a plating amount control method.

In the hot-dip plating process, the plating amount control of the steel sheet is controlled by an air knife installed on the front / bottom surface of the steel sheet, respectively. In order to measure the plating amount, the plating layer in the molten state must be completely dried, so the coating weight gauge is located at the rear end of about 200m from the air knife. Therefore, immediate feedback control is not possible, and the air knife pressure, which is the air knife gap, which is the distance between the air knife and the steel sheet, and the pressure of the gas injected from the air knife to the steel sheet, referring to the operator's experience or setting table. Pressure) is properly adjusted to operate.

Conventionally, the air knife pressure is adjusted solely according to the change in the line speed by using the plating amount prediction model ("Coating mass control system design for a continuous galvanizing line", 1976, WJ Edwards, etc.) There are methods to control the plating amount by adjusting the air knife gap of the front / rear solely. However, the conventional methods are methods for deriving and controlling the respective changes in the air knife gap and pressure according to the change in the operating conditions, and there are limitations in directly deriving the air knife gap and pressure suitable for each steel type and plating amount.

Prior Art Document 1: Korean Application No. 10-2013-0100205

Prior Art Document 2: Korean Application No. 10-2014-0190176

It is intended to provide a plating amount control device and a plating amount control method capable of improving the plating amount control accuracy, reducing the plating amount deviation, and improving the surface quality compared to the conventional manual operation.

In a continuous plating process in which a strip according to one aspect of the present invention is immersed and coated in a molten metal port, an apparatus for controlling the amount of plating coated on the strip using an air knife disposed along the traveling direction of the strip accumulates A predictive model unit including a predicted model in which a neural network has been trained for operating conditions, and an absolute value of the air knife pressure is derived based on the input operating condition, and the absolute value of the derived air knife pressure and the predicted model are used. And an optimum air knife condition calculator for deriving the absolute value of the air knife gap.

 The air knife condition calculation unit, the air knife pressure derivation unit for deriving the air knife pressure based on the input operation conditions, and the input operation conditions and the derived using the predicted model neural network learning the accumulated operation conditions And an air knife gap derivation unit for deriving an air knife gap based on the air knife pressure.

The air knife pressure derivation unit, a method for deriving the air knife pressure through a statistical method for the operation conditions corresponding to the input operation conditions in the database, the operating conditions excluding the air knife gap among the accumulated operation conditions One or more methods of learning a neural network, deriving an air knife pressure using the learned neural network, and deriving an air knife pressure corresponding to the input operating condition using a look-up table By using, it is possible to derive the air knife pressure.

The method of deriving the air knife pressure through a statistical method in the database, in the database, one of the mode, average, and median of the data for the air knife pressure corresponding to the input operating conditions, and the data Among them, one or more of the values with the lowest plating amount error between the target plating amount and the measured plating amount may be used.

The air knife pressure derivation unit derives a first air knife pressure on one side of the strip based on the input operating condition, and derives a second air knife pressure on the other side of the strip based on the input operating condition. The air knife gap derivation unit includes a first prediction model for one side of the strip and a second prediction model for the other side of the strip, and at least the input operating conditions for the first prediction model and the The first air knife pressure is applied to derive a first air knife gap for one side of the strip, and at least the input operating condition and the second air knife pressure are applied to the second prediction model to apply the first air knife pressure to the other surface of the strip. The second air knife gap for can be derived.

The air knife gap derivation unit may compare the first air knife gap and the second air knife gap, and correct the first air knife gap and the second air knife gap according to the comparison result.

When the difference between the first air knife gap and the second air knife gap is less than a predetermined threshold, the plating amount control device outputs each of the first air knife gap and the second air knife gap, or the first air If the difference between the knife gap and the second air knife gap is greater than the predetermined threshold, adjust the first air knife gap and the second air knife gap to derive the corrected first air knife gap and second air knife gap can do.

The air knife gap derivation unit performs an operation of deriving the corrected first air knife gap and the second air knife gap by adjusting the difference between the first air knife gap and the second air knife gap to be equal to or less than a predetermined threshold, and , The optimum air knife condition calculating unit, based on the corrected first and second air knife gaps, corrects air knife pressure for deriving the corrected air knife pressure for each of the one side and the other side of the strip using the prediction model It may further include wealth.

The air knife gap derivation unit derives an average of the first air knife gap and the second air knife gap as an optimum air knife pressure, and the optimum air knife condition calculator calculates the strip of the strip based on the optimum air knife gap. It may further include an air knife pressure correction unit for deriving the air knife pressure for each of the one side and the other side again.

The plating amount control device measures the plating amount of the strip, and corrects the prediction model based on a difference between the plating amount measurement value and the predicted plating amount prediction value using the prediction model or the target plating amount included in the input operating condition. can do.

The plating amount control device may correct a target plating amount input to the predicted value or the predicted model of the predicted model based on a difference between the measured amount of plating and the predicted plating amount or the target plating amount after the strip moves by a predetermined distance. You can.

The plating amount control device may further include a memory array that stores each of the plating amount prediction value or the target plating amount and the plating amount measurement value in a corresponding cell while the strip is moved by the predetermined distance.

In a continuous plating process in which a strip according to another aspect of the present invention is immersed in a molten metal port and coated, a method of controlling the amount of plating coated on the strip using an air knife disposed along the traveling direction of the strip is accumulated Learning a neural network based on the operated conditions, and deriving an absolute value of the air knife pressure based on the input operating condition, and using the derived neural network and the learned absolute value of the air knife gap It includes the step of deriving.

The plating amount control method further comprises: constructing a predictive model by learning the neural network of the accumulated operating conditions, and deriving an air knife gap based on the input operating conditions and the air knife pressure using the predictive model. It can contain.

The step of deriving the air knife pressure is a step of deriving the air knife pressure through a statistical method for the operation conditions corresponding to the input operation conditions in the database, excluding the air knife gap among the accumulated operation conditions Of the steps of learning the neural network operating conditions, deriving the air knife pressure using the learned neural network, and deriving the air knife pressure corresponding to the input operating conditions using a look-up table (Look-up table) It may include at least one.

The method for deriving the air knife pressure through a statistical method in the database, in the database, one of the mode, average, and median of the data for the air knife pressure corresponding to the input operating conditions, and the data Among them, one or more of the values with the lowest plating amount error between the target plating amount and the measured plating amount may be used.

The step of deriving the air knife pressure includes: deriving a first air knife pressure on one side of the strip based on the input operating condition, and removing the first air knife pressure on the other side of the strip based on the input operating condition. Deriving 2 air knife pressures, wherein the predictive model includes a first predictive model for one side of the strip and a second predictive model for the other side of the strip, and deriving the air knife gap is , Deriving a first air knife gap for one side of the strip by applying at least the input operating conditions and the first air knife pressure to the first prediction model, and at least the second prediction model Applying the operating conditions and the second air knife pressure to input the second air knife gap to the other side of the strip. Shipments may include the steps.

The plating amount control method may further include comparing the first air knife gap and the second air knife gap, and correcting the first air knife gap and the second air knife gap according to the comparison result. have.

The plating amount control method may include outputting each of the first air knife gap and the second air knife gap when the difference between the first air knife gap and the second air knife gap is less than a predetermined threshold, and the second air knife gap. If the difference between the first air knife gap and the second air knife gap is greater than the predetermined threshold, the first air knife gap and the second air knife gap corrected by adjusting the first air knife gap and the second air knife gap It may further include the step of deriving.

The plating amount control method comprises: adjusting the difference between the first air knife gap and the second air knife gap to be equal to or less than the predetermined threshold to derive the corrected first air knife gap and second air knife gap, and The method may further include deriving the corrected air knife pressure for each of the one side and the other side of the strip using a predictive model based on the corrected first and second air knife gaps.

The step of deriving the optimum air knife gap may include deriving an average of the first air knife gap and the second air knife gap as an optimum air knife gap.

The plating amount control method includes deriving an average of the first air knife gap and the second air knife gap as an optimal air knife gap, and air for each of the one side and the other side of the strip based on the optimum air knife gap The method may further include deriving the knife pressure again.

The plating amount control method includes: predicting a plating amount using the prediction model, measuring a plating amount of the strip, and correcting the prediction model based on a difference between the plating amount measurement value and the plating amount prediction value or the target plating amount It may further include a step.

In the correcting of the predictive model, after the strip moves by a predetermined distance, the predicted value or the target plating amount of the predictive model is corrected based on a difference between the measured plating amount measurement value and the plating amount prediction value or the target plating amount. It may include steps.

The step of correcting the prediction model may further include storing each of the plating amount prediction value or the target plating amount and the plating amount measurement value in a corresponding cell of a memory array while the strip is moved by the predetermined distance. .

The prediction model is a model for predicting the amount of plating, and may be a model that uses the input operating conditions as input and outputs the amount of plating.

The operation conditions may include one or more of line-related operation conditions, the air knife-related operation conditions, and the strip-related operation conditions in which the strip process is performed.

In the strip according to another feature of the present invention produced according to the plating amount control method, when the target plating amount for the strip is changed from the first level to the second level, the measured plating amount is -3 of the target plating amount of the second level The stabilization distance reaching ~ 3% may be less than 50M from the target plating amount change starting point, or the distance at which the measured plating amount converges to -1 to + 1% of the target plating amount of the second level may be less than 250M from the target plating amount changing starting point.

In the strip according to another feature of the present invention produced according to the plating amount control method, in the 200M section from the starting point of the target plating amount change, when the strip is ultra-thin plated, the longitudinal double-sided sum of the strip is before and after the strip. Deviation of 0.25% or less based on the sum of the target plating amount, when the strip is medium thin plating, the lengthwise double-sided sum of the strip is 0.66% or less based on the sum of the target plating amount of the front and rear surfaces of the strip, or the strip is post-plated In this case, the lengthwise double-sided sum of the strips may converge to a deviation of 1% or less based on the sum of the target plating amount of the front and rear surfaces of the strip.

In a strip according to another feature of the present invention produced according to the plating amount control method, a check mark in a diagonal pattern may not be generated on the surface of the strip.

The present invention derives the operating conditions of the air knife for realizing the target plating amount by the plating amount prediction model, thereby improving the plating amount control accuracy compared to the conventional manual operation, thereby reducing the plating amount variation and improving the surface quality.

1 is a view schematically showing a plating apparatus and a plating amount control apparatus according to an embodiment.
2 is a view showing a plating amount control device according to an embodiment.
3 is a view showing an optimum air knife condition derivation unit according to an embodiment.
4 is a diagram illustrating a memory array according to an embodiment.
5 is a graph showing a variation in plating amount according to an embodiment and a variation in plating amount in a conventional manual operation.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art to which the present invention pertains may easily practice. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. In addition, in order to clearly describe the present invention in the drawings, parts irrelevant to the description are omitted, and like reference numerals are assigned to similar parts throughout the specification.

In the continuous galvanizing line, the plating amount of the strip is controlled by two air knives installed on the front and rear surfaces of the strip, respectively.The plating amount is the distance between the air knife and the strip (air knife gap), and the air knife is supported. It is greatly influenced by the pressure (air knife pressure) of the gas sprayed on one side of the strip. Hereinafter, in the present disclosure, a steel sheet will be described as an example of a strip. However, the invention is not limited thereto, and the present invention can be applied to a metal plate such as a magnesium plate in a strip form.

In order to control the plating amount of the steel sheet surface to a target plating amount, a control technique based on a regression model has been applied. For example, by the method proposed by Edwards in 1976, the predicted plating amount (CP) is a function of line speed (V), air knife gap (D), and air knife pressure (P). There is a method using the defined plating amount prediction model of equation (1).

[Equation 1]

CP = V ^a G ^b P ^c

In Equation 1, a, b, and c are constants as model parameters.

Using the plating amount prediction model of Equation 1, when the target plating amount is given, the appropriate air knife gap and air knife pressure can be inverted at a specific line speed. However, there are a myriad of suitable air knife gaps and air knife pressures. When the air knife gap or pressure is too large or too small, surface quality can be deteriorated independently of the plating amount. Air knife gap, air knife pressure) cannot be derived.

The present invention relates to an apparatus for controlling the amount of plating of a steel sheet using an air knife, which implements a predictive model by learning a neural network with accumulated operating conditions including at least an air knife gap and an air knife pressure, and using the predictive model The absolute value of at least one of the air knife gap and the pressure according to the input operating conditions is derived.

In the present invention, by deriving the absolute values of the air knife gap and pressure, initial control after the operating conditions are changed is possible. If the operating conditions such as changing the type of steel sheet, changing the line speed, and changing the target plating amount are changed, it is not suitable as a criterion for deriving the relative change amount of the air knife gap and pressure in the operating conditions where the current air knife gap and pressure are changed. In order to apply the relative changes in the air knife gap and pressure to the air knife gap and pressure control, it is necessary to ensure that the current state (initial state) is optimal. However, this is difficult to guarantee in actual operation. Therefore, according to the conventional method of deriving the relative change amount of the air knife gap and pressure, it is difficult to derive the exact change amount of the air knife gap and pressure for the target plating amount after the operating condition is changed. The present invention can provide more accurate initial control by using the neural network learning to derive the absolute values of the air knife gap and pressure under the changed operating conditions without using the air knife gap and pressure as a reference before the change.

In addition, when calculating the relative amount of change, it is difficult to guarantee accuracy when various operating conditions are changed at the same time. For example, when the target plating amount and the line speed are changed at the same time, the result of calculating the pressure change for the changed target plating amount and then deriving the pressure change for the changed line speed may be different from the sum and calculated in the reverse order.

In addition, when controlling with a relative change amount, it is difficult to reflect operating know-how such as using a specific gap to secure the surface quality in a specific steel type / condition. The present invention can solve this problem by first deriving an air knife gap suitable for the operating conditions input from the accumulated operating conditions.

The operating conditions have a plurality of types, and the plurality of types may include, for example, steel sheet related data, air knife related data, line related data, and plating amount data. Steel sheet related data may include, for example, steel type, thickness, width, vibration, and the like of the steel sheet. The air knife related data may include, for example, air knife gap, air knife pressure, air knife angle and / or air knife height. The air knife gap may be a gap between the air knife and the center line to which the center of the steel plate is to be moved. The air knife pressure can be the pressure of the air used in the air knife. The air knife height may be the height of the air knife based on the hot water surface of the plating port. The air knife angle may be an angle of the air knife relative to the horizontal plane. Line-related data may include, for example, line speed, tension, and the like. The line speed is the speed at which the steel sheet moves along the traveling direction, and the tension may be a tension for moving the steel sheet in the traveling direction. In addition, operating conditions may include a target plating amount.

An embodiment relates to a plating amount control device and a plating amount control method for deriving an air knife gap and pressure suitable for input operation conditions using a plating amount prediction model and accumulated operating conditions. In one embodiment, the error due to uncertainty may be corrected by using the amount of plating measured at a time when a predetermined distance for measuring the amount of plating of the steel sheet has been moved.

For example, an air knife pressure suitable for the input operation condition from the accumulated operation condition may be first derived, and an air knife gap may be derived using the derived air knife pressure and plating amount prediction model. Since the accumulated operating conditions are used, it is possible to derive the air knife pressure reflecting the driver's know-how, and through this, it is possible to automatically control the plating amount to improve the surface quality. And, in one embodiment, the air knife gap for the front and rear surfaces of the steel sheet derived from the predictive model is the same, or the difference between the front and rear air knife gaps is adjusted to be within a predetermined range, and the air knife pressure is corrected according to the adjusted air knife gap can do. In addition, an embodiment may correct an error of a plating amount prediction model by using a plating amount measurement value.

Expanding the above-mentioned plating amount prediction model of Equation 1, when the plating amount prediction model is expressed as a function including operation conditions other than line speed, air knife gap, and air knife pressure as inputs, the plating amount prediction model can be expressed as follows.

[Equation 2]

CP = F (V, G, P,…)

Other operating conditions in Equation 2 may include steel grade, thickness, width, vibration, and tension of the steel sheet, air knife height and angle, and the like.

The plating amount prediction model of Equation 2 may be applied separately to each of the front and back surfaces of the steel sheet, and the function of Equation 2 satisfies the condition of Equation 3 below, where the inverse functions for G and P are satisfied under other input variables. Must exist.

[Equation 3]

In the plating amount prediction, using the plating amount prediction model of Equation 2, when the target plating amount CP is given, the air knife gap and pressure for realizing it under the corresponding operating conditions (steel plate speed, etc.) can be found. There are countless.

In one embodiment, one optimal solution can be found among a myriad of solutions using a database in which accumulated operating conditions are stored and a neural network learning the accumulated operating conditions. To this end, the plating amount control device according to one embodiment, the operation data [time, coil number, steel type, thickness, width, vibration, and tension, line speed, air knife gap (Top / Bottom), air knife pressure (Top / Bottom), target plating amount, plating amount measurement value (Top / Bottom), air knife height, air knife angle (Top / Bottom),… .] Can be stored in the database in real time.

In one embodiment, a neural network having an operation condition as an input and a plating amount as an output learns an operation condition stored in a database, and the plating amount prediction model may be implemented as a learned neural network. However, the invention is not limited thereto, and the input and output of the predictive model may vary according to design. In one embodiment, the plating amount prediction model implemented through neural network learning may find an optimal air knife gap based on the operating conditions including the first derived air knife pressure and at least the target plating amount.

The plating amount control device may derive the air knife pressure through a statistical method for the operation conditions corresponding to the operation conditions input from the database. The operation conditions corresponding to the input operation conditions include the same operation conditions input and similar operation conditions within a predetermined range. For example, the plating amount control device is one of the mode, average, and median values of the data for the air knife pressure, and among the data for the air knife pressure, the air knife with the lowest plating amount error between the target plating amount and the measured plating amount. The air knife pressure can be derived using at least one of the pressures.

Alternatively, the plating amount control device may learn the neural network of the operating conditions excluding the air knife gap among the accumulated operating conditions, and derive the air knife pressure in the input operating conditions using the learned neural network.

Alternatively, the plating amount control device may derive the air knife pressure corresponding to the input operation condition using a look-up table related to the accumulated operation condition. The lookup table may include high-accuracy operation conditions obtained from operations performed by skilled operators among accumulated operation conditions. Alternatively, the lookup table may include operating conditions reflecting the unique characteristics of the line on which the plating operation is performed. Since the line is a combination of various facilities for plating operations, characteristics may differ for each line. By reflecting this in the look-up table, the accuracy of plating amount control can be improved.

The plating amount control apparatus according to an embodiment first derives an air knife pressure, and derives an air knife gap through a predictive model implemented through neural network learning. In addition, in one embodiment, since the plating amount prediction model can be separately applied to each of the front and rear surfaces of the steel sheet, the air knife is spaced apart by a distance according to the air knife gap for each of the front and rear surfaces of the steel sheet, thereby affecting air knife pressure. Follow the gas injection. That is, since the plating processes for each of the front and rear surfaces of the steel sheet are separated, a model for predicting the plating amount for each of the front and rear surfaces can be separately applied. An air knife gap for realizing the target plating amount can be derived for each of the front / rear surfaces, and an optimal air knife gap can be determined based on the derived front / rear air knife gap.

For example, the plating amount control device may compare the front / rear air knife gap and correct the front / rear air knife gap according to the comparison result. When the difference between the air knife gaps of the front and rear surfaces is large, vibration or tilting of the steel sheet may be caused by the gap difference. Accordingly, one embodiment may correct the difference between the front and rear air knife gaps so as not to deviate from a predetermined threshold that may cause vibration or tilt of the steel plate.

For example, if the difference between the front / rear air knife gap is smaller than a predetermined threshold, each of the derived front / rear air knife gaps can be used as they are. If the difference between the front / rear air knife gap is greater than the threshold, the plating amount control device can adjust the front / rear air knife gap to derive an optimum air knife gap. Specifically, the optimum air knife gap can be derived by adjusting the plating amount control device so that the difference between the front and rear air knife gaps is equal to or less than a predetermined threshold. As an example, the average of the front / rear air knife gap can be determined as the optimum air knife gap.

Alternatively, the plating amount control device may determine whether the difference between the air knife gaps is equal to or less than a threshold, and determine the average of the front / rear air knife gaps as the optimum air knife gap.

When the difference between the air knife gaps for each of the front and rear surfaces is large, vibration or tilting of the steel sheet may be caused by the difference. To prevent this, one embodiment may calculate the average of two air knife gaps derived from each of the front / rear models, and apply the same air knife gap to the front / rear surfaces. In this case, one embodiment derives the air knife pressure of each front / rear surface using the plating amount prediction model again to correspond to the optimum air knife gap.

In one embodiment, an error in a plating amount prediction model may be corrected by measuring a plating amount prediction error. For example, the air knife gap monitored in actual operation is only the mechanical position of the air knife and may not accurately represent the distance between the actual air knife and the steel plate. This is because the actual position of the steel sheet changes depending on the thickness, tension, and vibration of the steel sheet.

In order to correct the error caused by the uncertainty, the plating amount control device receives feedback of the actual plating amount measured at a rear end by a predetermined distance from the plating operation position (eg, air knife position). For example, the plating amount control device may receive a feedback of a plating amount measurement value measured at a rear end of approximately 200 m, and correct an error of a plating amount prediction model.

Conventionally, since the distance of the steel sheet from the plating operation position to the actual measurement position of the plating amount for feedback is long, the plating amount control device has limited gain that reflects an error between the measured value and the predicted value. By correcting the error of the plating amount prediction model through feedback according to an embodiment, feedback control may be performed by the corrected plating amount prediction model without performing separate feedback control. Then, due to the limited gain, the moving distance of the steel sheet required for the predicted value to converge to the measured value can be improved.

Hereinafter, an embodiment will be described with reference to the drawings. The embodiment description with reference to FIGS. 1 to 4 is an example of implementing the invention, and the invention is not limited thereto.

1 is a view schematically showing a plating amount control apparatus and a plating apparatus according to an embodiment.

As shown in FIG. 1, the plating apparatus 100 operates under the control of the plating amount control apparatus 200. In one embodiment, a data communication device 300 for transmitting and receiving information between the plating device 100 and the plating amount control device 200 may be provided. The invention is not limited to this, and the plating apparatus 100 and the plating amount control apparatus 200 may be provided with apparatuses for transmitting and receiving data.

The plating apparatus 100 includes a plating port 110, an air knife 120 and a cooling unit 130. In one embodiment, the plating apparatus 100 may be a hot-dip plating apparatus.

The plating port 110 is for hot-dipping the steel plate SS, and the molten metal is contained in the plating port, and the steel plate SS guided to the plating port 110 is a sink roll disposed in the plating port 110 ( While passing through the sink roll 111, the hot dip plating process proceeds by being immersed in the molten metal 112, so that the surface of the steel sheet SS is coated.

The steel plate SS is shifted by the sink roll 111 and moves to the top of the plating port 110. The steel plate SS having a surface plated by the molten metal 112 in the plating port 110 is drawn out to the upper portion of the plating port 110. The steel plate SS is made of a plated steel plate through the air knife 120 and the cooling unit 130, which are sequentially arranged along the traveling direction. The steel plate SS cooled through the cooling unit 130 proceeds to the process through the tension roll 140.

In one embodiment, the plating solution may be zinc, zinc alloy, aluminum and / or aluminum alloy, and the like.

The air knife 120 is disposed on one side or both sides of the steel plate at the rear end of the plating port 110 along the traveling direction of the steel plate SS to control the plating amount of the steel plate. The air knife 120 includes air knives 121 and 122, and the air knives 121 and 122 are air knife pressures at a distance spaced by an air knife gap on a plated layer attached to the steel plate SS surface. The amount of plating is controlled by spraying gas with. For example, the air knife 120 has a body that extends in the width direction of the steel plate SS and a cryogenic liquid is circulated inside, and a tip inclined by an air knife angle with respect to the plated layer of the steel plate SS at the body tip (not shown) May not be formed).

Each of the air knives 121 and 122 may control the air knife gap and pressure according to the control signals AFC1 and AFC2. In FIG. 1, the control signals AFC1 and AFC2 are illustrated as being transmitted to the air knives 121 and 122 through the data communication device 300, but the invention is not limited thereto. Each of the air knives 121 and 122 may receive control signals AFC1 and AFC2 directly from the plating amount control device 200.

The cooling unit 130 directly contacts the plated layer on the surface of the steel plate SS to cool the steel plate SS. For example, the cooling bodies 131 and 132 may include a cooling roll (not shown) that extends in the width direction of the steel sheet and circulates a cryogenic liquid therein and pressurizes the plating layer on the surface of the steel sheet to apply cool air. A plurality of such cooling rolls may be arranged in multiple stages at intervals along the traveling direction of the steel plate SS.

The plating amount control device 200 automatically controls the operating conditions used for the plating operation of the plating device 100. The plating amount control apparatus 200 according to an embodiment derives an air knife pressure based on an input operation condition among accumulated operation conditions, and inputs an operation condition including the derived air knife pressure and at least a target plating amount into a prediction model By doing so, the air knife gap can be derived. In addition, the plating amount control device 200 may derive an optimum air knife gap based on the front / rear air knife gap, and correct the front / rear air knife pressure based on the optimum air knife gap.

The plating amount measuring apparatus 400 scans the front and rear surfaces of the steel sheet, measures the front and rear plating amounts, and generates the measured values FB1 and FB2. The plating amount measuring apparatus 400 includes a front plating amount measuring unit 401 and a rear plating amount measuring unit 402, and the front plating amount measuring unit 401 generates a measurement value FB1 of the front plating amount, and a rear plating amount measuring unit 402 generates a measured value (FB2) of the back plating amount.

The data communication device 300 includes plating amount measurement values (FB1, FB2), which are values obtained by measuring the plating amount of the plated steel sheet according to the operating conditions used for the plating operation in the plating apparatus 100 and the operating conditions in the plating apparatus 100. The operation data to be collected may be collected and transmitted to the plating amount control device 200. In addition, the data communication device 300 may receive data regarding an operation instruction from the plating amount control device 200 and transmit a corresponding control signal to the plating device 100. The data communication device 300 may be implemented as a computing device capable of communicating with the plating device 100 and the plating amount control device 200.

2 is a view showing a plating amount control device according to an embodiment.

As illustrated in FIG. 2, the plating amount control apparatus 200 includes a plating amount prediction unit 210, an optimum air knife condition derivation unit 220, a feedback model correction unit 230, a plating amount prediction model unit 240, and It includes a database 250.

The database 250 may receive and store data regarding operating conditions. The optimum air knife condition derivation unit 220 may use data stored in the database 250 in deriving the optimum air knife gap. In FIG. 2, although it is illustrated that the database 250 is included in the plating amount control device 200, the invention is not limited thereto, and the database 250 is implemented as a separate device and the plating amount control device 200 and Data can be transmitted and received from each other.

The plating amount prediction model unit 240 includes separate plating amount prediction models for each of the front / rear plating amounts, and each plating amount prediction model may be implemented as a neural network learning the accumulated operating conditions.

The plating amount prediction unit 210 may predict the plating amount using the plating amount prediction model of the plating amount prediction model unit 240. For example, the plating amount prediction unit 210 receives the operation conditions, and the front / rear plating amount prediction values (ECP1, ECP2) according to the operation conditions input through the front / rear plating amount prediction model of the plating amount prediction model unit 240 are input. Predict and output.

The air knife condition derivation unit 220 may derive the optimum air knife gap and the air knife pressure of the front and rear surfaces under the current operating conditions using the plating amount prediction model. The air knife condition derivation unit 220 may input one of the air knife gap and the air knife pressure and a target plating amount into the plating amount prediction model, and inversely calculate the plating amount prediction model to derive one of the air knife gap and the air knife pressure. .

For example, the optimum air knife condition derivation unit 220 derives the front / rear air knife pressure, and the derived front / rear air knife pressure and target plating amount are predicted by the front / rear plating amount prediction model of the plating amount prediction model unit 240 Enter to to derive the front / rear air knife gap, derive the optimum air knife gap based on the derived front / rear air knife gap, and input the optimum air knife pressure and target plating amount into the front / rear plating amount prediction model. / Rear optimum air knife pressure can be derived.

3 is a view showing an air knife condition derivation unit according to an embodiment.

As shown in FIG. 3, the air knife condition derivation unit 220 includes an air knife pressure derivation unit 221, an air knife gap derivation unit 222, and an air knife pressure correction unit 223.

The air knife pressure relief part 221 includes a front air knife pressure relief part 2211 and a rear air knife pressure relief part 2212. The front / rear air knife pressure deriving units 2211 and 2212 may receive operating conditions and derive front / rear air knife pressures P1 and P2.

Each of the front / rear air knife pressure derivation units 2211 and 2212 may derive the air knife pressure through a statistical method for the operation conditions corresponding to the operation conditions input from the database 250. For example, each of the front / rear air knife pressure derivation units 2211 and 2212 may derive one of a mode, an average value, and a median value of air knife pressures of operation conditions similar to input operation conditions. Each of the front / rear air knife pressure derivations 2211 and 2212 each uses an air knife pressure using an air knife pressure having the smallest plating amount error between a target plating amount and a measured plating amount among air knife pressures of operating conditions similar to the input operating conditions. Can be derived.

Alternatively, each of the front / rear air knife pressure derivation units 2211 and 2212 learns the neural network of the operating conditions excluding the air knife gap among the accumulated operating conditions, and uses the learned neural network to determine the air knife pressure at the input operating conditions. Can be derived.

Alternatively, each of the front / rear air knife pressure deriving units 2211 and 2212 may derive the air knife pressure corresponding to the input operating condition using a look-up table related to the accumulated operating condition. In one embodiment, the look-up table may include high-accuracy operation conditions obtained from operations performed by skilled operators among accumulated operation conditions. Alternatively, the lookup table may include operating conditions reflecting the unique characteristics of the line on which the plating operation is performed. Since the line is a combination of various facilities for plating operations, characteristics may differ for each line. By reflecting this in the look-up table, the accuracy of plating amount control can be improved.

The air knife gap lead part 222 includes a front air knife gap lead part 2221, a rear air knife gap lead part 2222, and an optimal air knife gap lead part 2223.

The front / rear air knife gap derivation units 2221 and 2222 load the front / rear plating amount prediction model from the plating amount prediction model unit 240. The front / rear air knife gap derivation units 2221 and 2222 input at least a target plating amount and a front / rear air knife pressure P1 and P2 among operating conditions into a front / rear plating amount prediction model, and the front / rear plating amount prediction model is The output creates front / rear air knife gaps G1 and G2.

The optimum air knife gap derivation unit 2223 derives the optimum air knife pressure G by averaging the front / rear air knife gaps G1 and G2. The method for deriving the optimum air knife gap is not limited to the average. The average is one example, one of various ways of deriving the same optimal air knife gap using the front / rear air knife gaps G1 and G2 is applied to one embodiment, or the front / rear air knife gaps G1 and G2 It can be applied to one embodiment by selecting any one of them.

Alternatively, the optimum air knife gap derivation unit 2223 may compare the front / rear air knife gaps G1 and G2, and correct the front / rear air knife gap according to the comparison result.

For example, if the difference between the front / rear air knife gaps G1 and G2 is smaller than a predetermined threshold, the optimal air knife gap derivation unit 2223 maintains each of the derived front / rear air knife gaps G1 and G2 as it is. Can be used. In this case, the air knife pressure compensator 223 is bypassed, so that "P1" and "P2" are output as the optimum air knife pressure, and "G1" and "G2" can be output as the optimum air knife gap. have.

If the difference between the front / rear air knife gaps is greater than the threshold, the optimal air knife gap derivation unit 2223 can derive the optimum air knife gap by adjusting the front / rear air knife gaps G1 and G2. For example, the optimum air knife gap derivation unit 2223 may derive the average of the front / rear air knife gaps G1 and G2 as the optimum air knife gap G.

Alternatively, the optimum air knife gap derivation unit 2223 may derive the optimum air knife gap G by adjusting the difference between the front / rear air knife gaps G1 and G2 to be equal to or less than a predetermined threshold. For example, the optimum air knife gap derivation unit 2223 may determine the average of the front / rear air knife gaps as the optimum air knife gap G.

The air knife pressure compensator 223 includes a front air knife pressure compensator 2223 and a rear air knife pressure compensator 2232.

The front / rear air knife pressure correcting units 2231 and 2232 input operating conditions including at least a target plating amount and an optimum air knife gap G in the model for predicting the amount of plating in the front and rear of the plating amount prediction model unit 240, and the plating amount The predictive model predicts and outputs optimal air knife pressures P3 and P4.

Finally, the optimum air knife condition derivation unit 220 may output “P3”, “P4”, and “G” as the optimum air knife gap and pressure. The optimum air knife gap and pressure are transmitted to the data communication device 300, and the data communication device 300 may generate control signals AFC1 and AFC2 based thereon and transmit them to the air knives 121 and 122.

The feedback model correction unit 230 receives the front / rear plating amount measurement values (FB1, FB2) and the front / rear plating amount prediction values (ECP1, ECP2), and the front / rear plating amount measurement values (FB1, FB2) and the front / rear plating amount The plating amount prediction model can be corrected based on the difference between the predicted values (ECP1, ECP2). The present invention is not limited to this, and the plating amount prediction model may be corrected using the target plating amount instead of the front / rear plating amount prediction values (ECP1, ECP2). That is, the feedback model correction unit 230 may correct the plating amount prediction model based on the difference between the input target plating amount and the front / rear plating amount measurement values FB1 and FB2.

For example, the feedback model correcting unit 230 averages the difference between the two values while the steel plate moves through a predetermined error calculation section (for example, 10 to 15 m) to predict the average errors ER1 and ER2 as the plating amount prediction model. And transmit it to the unit 240.

The plating amount prediction model unit 240 may correct the front / rear plating amount prediction model by reflecting the average errors ER1 and ER2.

The feedback model correction unit 230 includes a memory array 231 to accurately match the positions of the front / rear plating amount prediction values ECP1 and ECP2 or the input target plating amount and the plating amount measurement values FB1 and FB2.

4 is a diagram illustrating a memory array according to an embodiment.

As illustrated in FIG. 4, each of the front / rear plating amount prediction values ECP1 and ECP2 or the input target plating amount and the plating amount measurement values FB1 and FB2 are respectively stored in the memory arrays 2311 and 2312 corresponding to the corresponding locations. For example, the steel plates are divided at predetermined intervals (for example, 1 m) along the direction in which the steel plate travels, and the front / rear plating amount prediction values (ECP1, ECP2) in each region or the input target plating amount and the plating amount measurement value ( Each of FB1 and FB2) is stored in the corresponding cell of each of the memory arrays 2311 and 2312.

The memory array cells are shifted according to the steel plate speed so as to match the positions of the steel plates from the air knives 121 and 122 to the plating amount measuring devices 401 and 402.

As shown in Fig. 4, the steel plate reaches the air knife position A1 and is synchronized with the front / rear plating amount prediction values (ECP1, ECP2) or the input target plating amount stored in the corresponding cell of the memory array 2311. Can be. Whenever the steel sheet passes through the air knife position A1 by a predetermined interval (1m), the cell is shifted, and the target plating amount to be inputted or predicted by the front / rear plating amount to the cells of the corresponding memory array 2311 (ECP1, ECP2) This is saved.

When the steel sheet reaches the plating amount measuring device position A2, the front / rear plating amount measurement values FB1 and FB2 are stored in the corresponding cells 2312_2 of the memory array 2312. Each time the steel sheet passes through the air knife position A2 by a predetermined interval (1m), the cells of the memory array 2312 are shifted, and the front / rear plating amount measurement value FB1, in the cells of the corresponding memory array 2312, FB2) is stored.

When the steel sheet area corresponding to the cell 2311_1 reaches the plating amount measuring device position A2, and the front / rear plating amount measurement values FB1 and FB2 are stored in the cell 2312_1, the feedback model correcting unit 230 displays the cell ( The difference between the front / rear plating amount measurement values FB1 and FB2 stored in 2312_1) and the front / rear plating amount prediction values ECP1 and ECP2 stored in the cell 2311_1 is calculated.

In this way, the feedback model correcting unit 230 averages the differences between the two values calculated while the steel sheet passing through the plating amount measuring device position A2 moves a predetermined error calculation section, and the average error (ER1, ER2) Can be calculated. ER1 is the average error obtained by averaging FB1-ECP1 in the error calculation section, and ER2 is the average error obtained by averaging FB2-EPC2 in the error calculation section.

Then, the plating amount prediction model unit 240 corrects the plating amount prediction model based on the average error ER1 or ER2. For example, the plating amount prediction model unit 240 may reflect the average error in the target plating amount or the average error in the predicted plating amount.

For example, the target plating amount may be decreased by the average error ER1 or ER2, or the prediction amount of the plating amount by the average error ER1 or ER2 may be increased.

After the plating amount prediction model unit 240 corrects, it waits until the portion predicted by the corrected plating amount prediction model passes through the plating amount measuring apparatus position A2. At this time, the period of the feedback model correction is expressed by Equation (4).

[Equation 4]

Feedback model correction period = [time when the steel plate moves the distance A1-A2] + [time when the steel plate moves the error calculation section] + [scan time of the plating amount measuring device]

The correction period of the plating amount prediction model must be longer than the feedback model correction period, and if it is short, hunting may occur. The corrected plating amount prediction model is used for inversion in the air knife condition derivation unit again, so that an effect of feedback control can be provided.

In an embodiment of the present invention, the air knife pressure is first derived and the air knife gap is derived based on the derived air knife pressure. The embodiment may be applied when the variation width of the plating amount between the coils continuous in the steel sheet is greater than or equal to a predetermined value.

5 is a graph showing a variation in plating amount according to an embodiment and a variation in plating amount in a conventional manual operation.

As shown in FIG. 5, it can be seen that, in each of the target plating amounts CW _A , CW _B , and CW _C , the plating amount variation according to an embodiment is significantly improved compared to the plating amount variation of the conventional manual operation.

Specifically, when the plating amount deviation by the conventional manual operation for the target plating amount CW _A , CW _B , and CW _C is SD _A , SD _B , and SD _C , when the plating amount is controlled according to an embodiment, the plating amount deviation is 0.33 SD _A , 0.17SD _B , and 0.34SD _C.

When the plating amount variation is reduced, the actual amount of the plating attachment to achieve the customer ordered plating amount can be significantly reduced compared to the prior art. For example, in order to achieve a customer order plating amount in the prior art, if the actual plating attachment instruction amount was an overprohibition order of 3.22% on both sides, in one embodiment, the overprohibition order can be reduced to 0.499% on both sides. Then, it is possible to provide an effect in which the overplating amount for achieving the customer order plating amount is lower than in the prior art.

Edge build-up, flow pattern, and check mark generated by the conventional manual operation may not be revealed to the naked eye after the SPM process. However, it is revealed in all cases where surface inspection (grindstone inspection) is performed for at least 5m in length at least once for the entire width using a grindstone on the vertical or horizontal inspection table.

For example, an edge build-up is generated by a conventional manual operation for a plated steel sheet having a raw material cold rolled steel sheet of 0.4 t or less and a double-sided sum of 140 g / m ² or more. The plating amount control apparatus and method according to an embodiment may prevent edge build-up by learning air knife conditions (eg, pressure and gap) in which edge build-up does not occur, and automatically controlling them.

The temperature of the steel sheet in the region immediately after the plating amount is controlled by the pressure and flow rate of the fluid in the air knife is, for example, before / after 430 ° C. Then, the area immediately after the plating amount is controlled is an initial state of solidification, but an oxide film is formed on the surface of the area, but the shear stress generated when the fluid from the air knife strikes the steel sheet and exits is physically attached to the unsolidified oxide layer. It applies a force. At this time, when the force exceeds the critical stress required for the rupture of the oxide film, a fine wave is generated, and a wave mark, which is a dark plating flow mark in the form of a wave pattern, is generated on the front and rear surfaces of the plated steel sheet. do.

For example, in order to control the plating amount for the post-plating material, when using a long air knife gap (for example, 10 mm or more), the variation of the air knife gap is changed to at least 2 mm or more to match the plating amount. Then, a flow pattern is generated on the surface of the steel sheet. The plating amount control apparatus and method according to an embodiment may prevent a flow pattern by learning a boundary condition in which the flow pattern occurs.

In addition, under certain conditions, such as line speed, wiping pressure, and viscosity of molten metal, a flow of molten metal attached to the strip and molten metal descending due to wiping occurs, resulting in strip width. The pattern occurs in a semicircular shape in the direction. That is, the molten metal flow phenomenon under the air knife may generate a diamond-shaped check pattern in a predetermined pattern to generate a check mark.

For example, when the line speed is constant in the conventional medium-thin plating amount control, a check mark may occur. The plating amount control apparatus and method according to an embodiment may prevent a checkmark by learning a boundary condition in which a checkmark occurs, based on a result determined through a surface defect detection device (SDD).

In one embodiment of the present invention, the air knife gap is first derived and the air knife pressure is derived based on the derived air knife gap. An embodiment may be applied when the variation width of the plating amount between the coils continuous in the steel sheet is greater than or equal to a predetermined value.

The process of the hot-dip plating facility is a continuous process due to the operation characteristics, and when the air knife gap and pressure are continuously derived and changed at the same time, despite the goal of converging to the target plating amount, it is erroneous due to an incorrect answer (Local Minima) during the optimization process. The air knife gap and pressure are derived, so that the actual plating amount may not converge to the target plating amount.

For example, in the result of feedback after scanning by the plating amount measuring device, the airknife pressure is optimized while the airknife gap and pressure are simultaneously derived, even though the airknife pressure has already been found to be optimal. Change together. Then, it is impossible to find the correct answer (Global Minima) of the air knife gap and pressure to achieve the target plating amount, and continues to fall into the wrong answer, and the plating amount control is repeatedly repeated to continuously find the optimum value.

In this situation, if the operator of the hot-dip plating facility does not respond appropriately, the strip may pass through the air knife in an over-plated and under-plated state, and the plating amount may not be controlled.

The present invention derives a relatively accurate value of the air knife gap and pressure from the accumulated operating conditions or first fixes the accumulated operating conditions using a neural network, and then controls the remaining factors through the plating amount prediction model, resulting in incorrect answers (Local Minima) can significantly reduce the probability of falling.

According to one embodiment, when the target plating amount is changed from the first level to the second level, the stabilization distance at which the measured plating amount reaches -3% to 3% of the target plating amount of the second level is the first plating amount is removed from the first level. It may be less than 50M from the base point changed to 2 levels.

In addition, when the target plating amount is changed from the first level to the second level, the distance at which the measured plating amount converges to -1% to% of the target plating amount of the second level is the starting point at which the target plating amount is changed from the first level to the second level. From less than 250M.

According to one embodiment, in the 200M section from the starting point of the change in the target plating amount, when the strip is ultra-thin plated (for example, 100 g / m ² or less product), the lengthwise double-sided sum of the strip is the front / rear target of the strip. Convergence with a deviation of 0.25% or less based on the sum of the plating amounts, or when the strip is medium-plated (101 to 180 g / m ² products), the longitudinal double-sided sum of the strips is 0.66% based on the sum of the target plating amount of the front and rear sides of the strip If the convergence is as follows, or if the strip is a post-plated product of 180 to 300 g / m ² ), the lengthwise double-sided sum of the strips may converge to a deviation of 1% or less based on the sum of the target plating amount of the front and back sides of the strip. .

For reference, the ratio (CW1 / CW2) of the target plating amount (CW2) of the trailing coil to the target plating amount (CW1) of the preceding coil may be 0.29 to 3.43.

Although the embodiments of the present invention have been described in detail above, the scope of rights of the present invention is not limited thereto, and various modifications and improvements of those skilled in the art using the basic concept of the present invention defined in the following claims are also provided. It belongs to the scope of rights.

100: plating device
200: plating amount control device
300: data communication device
210: plating amount prediction unit
220: optimum air knife condition deriving unit
230: feedback model correction unit
240: plating amount prediction model
250: database

Claims (32)

In a continuous plating process in which a strip is immersed and coated in a molten metal port, the apparatus for controlling the amount of plating coated on the strip using an air knife disposed along the traveling direction of the strip,
Prediction model unit including a predicted model of neural network learning the accumulated operating conditions, and
And an optimum air knife condition calculator for deriving an absolute value of the air knife pressure based on the input operating condition, and deriving an absolute value of the air knife gap using the derived absolute value of the air knife pressure and the prediction model Plating amount control device, characterized in that. According to claim 1,
The air knife condition calculation unit,
Air knife pressure deriving unit for deriving the air knife pressure based on the input operating conditions, and
And an air knife gap derivation unit that derives an air knife gap based on the input operation condition and the derived air knife pressure using a predictive model that has learned the accumulated operation conditions using a neural network. According to claim 2,
The air knife pressure deriving unit,
A method of deriving the air knife pressure through a statistical method for the operating conditions corresponding to the input operating conditions in the database,
Method of deriving an air knife pressure using the learned neural network by learning a neural network of the operating conditions excluding the air knife gap among the accumulated operating conditions, and
By using a look-up table (Look-up table) using one or more of the methods for deriving the air knife pressure corresponding to the input operation conditions,
Plating amount control device, characterized in that for deriving the air knife pressure. According to claim 3,
The method for deriving the air knife pressure through a statistical method in the database,
In the database, one or more of the mode, average, and median values of the data for the air knife pressure corresponding to the input operating condition, and one or more of the least plating value error between the target plating amount and the measured plating amount among the data Plating amount control device characterized in that using. According to claim 2,
The air knife pressure deriving unit,
Deriving a first air knife pressure on one side of the strip based on the input operating conditions,
Deriving a second air knife pressure on the other side of the strip based on the input operating condition,
The air knife gap lead portion,
A first prediction model for one side of the strip and a second prediction model for the other side of the strip,
Deriving a first air knife gap for one side of the strip by applying at least the input operating conditions and the first air knife pressure to the first predictive model,
And a second air knife gap with respect to the other surface of the strip by applying at least the input operating conditions and the second air knife pressure to the second predictive model. The method of claim 5,
The air knife gap lead portion,
A plating amount control apparatus characterized in that the first air knife gap and the second air knife gap are compared, and the first air knife gap and the second air knife gap are corrected according to the comparison result. The method of claim 5,
If the difference between the first air knife gap and the second air knife gap is less than a predetermined threshold, output each of the first air knife gap and the second air knife gap, or
If the difference between the first air knife gap and the second air knife gap is greater than the predetermined threshold, the first air knife gap and the second air corrected by adjusting the first air knife gap and the second air knife gap Plating amount control device, characterized in that for deriving the knife gap. The method of claim 5,
The air knife gap lead portion,
Performing an operation to derive the corrected first air knife gap and the second air knife gap by adjusting the difference between the first air knife gap and the second air knife gap to be equal to or less than a predetermined threshold,
The optimum air knife condition calculation unit,
Plating amount characterized in that it further comprises an air knife pressure correcting unit for deriving the corrected air knife pressure for each of the one surface and the other surface of the strip using the prediction model based on the corrected first and second air knife gaps. controller. The method of claim 5,
The air knife gap lead portion,
The average of the first air knife gap and the second air knife gap is derived as an optimum air knife pressure,
The optimum air knife condition calculation unit,
And an air knife pressure compensator for deriving air knife pressure for each of the one side and the other side of the strip again based on the optimum air knife gap. According to claim 2,
A plating amount characterized in that the plating amount of the strip is measured and the prediction model is corrected based on a difference between the plating amount measurement value and the predicted plating amount prediction value using the prediction model or the target plating amount included in the input operating condition. controller. The method of claim 10,
After the strip is moved by a predetermined distance, the plating amount control, characterized in that for correcting the target plating amount input to the predicted value or the predicted model of the predicted model based on the difference between the plating amount measurement value and the predicted plating amount or the target plating amount Device. The method of claim 11,
And the memory array storing each of the plating amount prediction value or the target plating amount and the plating amount measurement value in a corresponding cell while the strip is moved by the predetermined distance. According to claim 1,
The prediction model is a plating amount prediction model,
A plating amount control device that is a model that outputs by predicting and outputting the plating amount by using the input operating conditions as an input. The method according to any one of claims 1 to 13,
The operating conditions include a line-related operating condition in which the strip process is performed, the air knife-related operating condition, and the strip-related operating condition, characterized in that it comprises at least one of the strip-related operating conditions. In a continuous plating process in which a strip is immersed and coated in a molten metal port, in the method of controlling the amount of plating coated on the strip using an air knife disposed along the traveling direction of the strip,
Neural network learning the accumulated operating conditions, and
And deriving an absolute value of the air knife pressure based on the input operating condition, and deriving an absolute value of the air knife gap using the derived neural network and the derived absolute value of the air knife pressure. Plating amount control method. The method of claim 15,
Constructing a predictive model by learning a neural network on the accumulated operating conditions, and
And deriving an air knife gap based on the input operating conditions and the air knife pressure using the predictive model. The method of claim 16,
The step of deriving the air knife pressure,
Deriving the air knife pressure through a statistical method for the operation conditions corresponding to the input operation conditions in the database,
Learning the operating conditions excluding the air knife gap among the accumulated operating conditions, and deriving an air knife pressure using the learned neural network, and
Deriving an air knife pressure corresponding to the input operating condition using a look-up table
Plating amount control method comprising at least one of the. The method of claim 17,
The method for deriving the air knife pressure through a statistical method in the database,
In the database, one or more of the mode, average, and median values of the data for the air knife pressure corresponding to the input operating condition, and one or more of the least plating value error between the target plating amount and the measured plating amount among the data Plating amount control method characterized in that using. The method of claim 16,
The step of deriving the air knife pressure,
Deriving a first air knife pressure on one side of the strip based on the input operating conditions, and
Deriving a second air knife pressure on the other surface of the strip based on the input operating condition;
The prediction model includes a first prediction model for one side of the strip and a second prediction model for the other side of the strip,
The step of deriving the air knife gap,
Deriving a first air knife gap for one side of the strip by applying at least the input operating conditions and the first air knife pressure to the first prediction model, and
And deriving a second air knife gap with respect to the other surface of the strip by applying at least the input operating condition and the second air knife pressure to the second predictive model. The method of claim 19,
And comparing the first air knife gap and the second air knife gap, and correcting the first air knife gap and the second air knife gap according to the comparison result. . The method of claim 19,
If the difference between the first air knife gap and the second air knife gap is less than a predetermined threshold, outputting each of the first air knife gap and the second air knife gap, and
If the difference between the first air knife gap and the second air knife gap is greater than the predetermined threshold, the first air knife gap and the second air corrected by adjusting the first air knife gap and the second air knife gap A method of controlling a plating amount, further comprising deriving a knife gap. The method of claim 19,
Adjusting the difference between the first air knife gap and the second air knife gap to be below the predetermined threshold to derive a corrected first air knife gap and a second air knife gap, and
And deriving the corrected air knife pressure for each of the one side and the other side of the strip using a predictive model based on the corrected first and second air knife gaps. The method of claim 22,
The step of deriving the optimum air knife gap,
And deriving an average of the first air knife gap and the second air knife gap as an optimal air knife gap. The method of claim 20,
Deriving an average of the first air knife gap and the second air knife gap as an optimal air knife gap, and
And deriving the air knife pressure for each of the one side and the other side of the strip again based on the optimum air knife gap. The method of claim 16,
Predicting the plating amount using the prediction model,
Measuring the plating amount of the strip, and
And correcting the prediction model based on a difference between the plating amount measurement value and the plating amount prediction value or the target plating amount. The method of claim 25,
The step of correcting the prediction model may include:
And after the strip is moved by a predetermined distance, correcting the predicted value of the predicted model or the targeted plating amount based on a difference between the measured plating amount measurement value and the plating amount prediction value or the target plating amount. Control method. The method of claim 26,
The step of correcting the prediction model may include:
And while the strip is moved by the predetermined distance, storing the plating amount prediction value or the target plating amount and each of the plating amount measurement values in a corresponding cell of a memory array. The method of claim 16,
The prediction model is a plating amount prediction model,
A plating amount control method, which is a model that uses the input operating conditions as an input and outputs a plating amount. The method according to any one of claims 15 to 28,
The operating conditions include a line-related operating condition, the air knife-related operating conditions, and the strip-related operating conditions, wherein the strip process is performed. A strip produced according to the method for controlling the plating amount of any one of claims 15 to 28,
When the target plating amount for the strip is changed from the first level to the second level,
The stabilization distance at which the measured plating amount reaches -3 to 3% of the target plating amount of the second level is less than 50M from the starting point of the target plating amount change, or
A strip characterized in that the distance at which the measured plating amount converges to -1 to + 1% of the target plating amount of the second level is less than 250M from the starting point of the target plating amount change. A strip produced according to the method for controlling the plating amount of any one of claims 15 to 28,
In the 200M section from the starting point of the target plating amount change,
When the strip is ultra-thin plated, the lengthwise double-sided sum of the strips is less than or equal to 0.25% based on the sum of the target plating amount of the front and back sides of the strip
When the strip is medium thin plated, the lengthwise double-sided sum of the strips is a deviation of 0.66% or less based on the sum of the target plating amount of the front and rear surfaces of the strip, or
When the strip is post-plated, the strip, characterized in that the lengthwise double-sided sum of the strips converges with a deviation of 1% or less based on the sum of the target plating amount of the front and rear surfaces of the strip. A strip produced according to the method for controlling the plating amount of any one of claims 15 to 28,
A strip characterized in that a check mark in a diagonal pattern does not occur on the surface of the strip.

Images