JP2024012077A - Generation method for shape control actuator setting model of rolling equipment, setting method for shape control actuator of rolling equipment, steel plate shape control method, and steel plate manufacturing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a generation method for shape control actuator setting model of rolling equipment that can generate a shape control actuator setting model which can control a shape of a steel plate in a longer direction with good accuracy and can improve stability of the shape of the steel plate.

SOLUTION: A generation method for shape control actuator setting model of rolling equipment includes a setting model generation step at which a shape control actuator setting model is generated by a neural network method with use of a plurality of pieces of learning data in which one or more actual operation data selected from rolling operational data of rolling equipment and actual shape data acquired from a shape detector are so made as to be actual input data, and setting change information as to whether setting change of a shape control actuator is performed or not and a set value of the shape control actuator are so made as to be actual output data.

SELECTED DRAWING: Figure 3

COPYRIGHT: (C)2024,JPO&INPIT

Description

The present invention relates to a method for generating a shape control actuator setting model for rolling equipment, a method for setting a shape control actuator for rolling equipment, a method for controlling the shape of a steel plate, a shape control device for a steel plate, and a method for manufacturing a steel plate.

In a rolling equipment that reduces the thickness of a thin steel plate, the steel plate is wound into a coil and charged into the rolling equipment as a steel strip, and the steel plate is continuously unloaded and rolled while being conveyed. In the rolling process, it is required to control the steel plate to a predetermined size and to appropriately control the shape of the steel plate. The shape of a steel plate refers to its flatness, and typical examples include ear waves and mid-elongation. The shape of a steel plate is such that the elongational strain (longitudinal strain) distributed in the width direction of the steel plate cannot be maintained as an in-plane stress distribution during rolling, so buckling occurs and corrugations (defects in shape) become apparent. be. When a defect in the shape of a steel plate occurs, not only does the quality of the product deteriorate, but the rolling speed (line speed) of the rolling equipment has to be reduced, resulting in a decrease in production efficiency. Furthermore, poor shape of the steel plate may cause operational troubles such as breakage and squeezing of the steel plate. Therefore, in the rolling process, it is required to control the shape of the steel plate to a desired shape over the longitudinal direction.

On the other hand, the rolling mill that constitutes the rolling equipment is equipped with a shape control actuator that controls the elastic deformation of the work roll so that the elongation strain distributed in the width direction of the steel plate has a desired distribution. Different shape control actuators are used depending on the type of rolling mill, but a typical example is a work roll bender that applies bending force to the work roll to adjust the deflection of the work roll. Other methods include a roll shift method in which the work roll and intermediate roll are moved in the axial direction, and a method in which the deflection of the work roll is adjusted using a split backup roll. Further, the rolling equipment includes a shape detector for the steel plate, and a shape control actuator is set so that the shape of the steel plate detected by the shape detector matches a target shape (target shape).

The shape control device that sets the shape control actuator has the function of performing initial settings before starting the rolling pass, and correcting the settings of the shape control actuator by referring to the output from the shape detector after the start of the rolling pass ( (Resetting) is often provided with a feedback control function. By performing appropriate initial settings before starting the rolling pass, the shape of the steel sheet can be improved from the tip end, and the feedback control function can provide a good shape over the entire length of the steel sheet even when disturbances occur. However, it may be difficult to quantitatively predict the influence of the setting value of the shape control actuator on the elastic deformation of the work roll of a rolling mill. For example, in order to quantitatively predict the deflection of work rolls, it is necessary to predict the thermal expansion behavior of the various rolls that make up the rolling mill, and these predictions often contain various errors. . In response to this, a technique has been proposed for appropriately adjusting the set value of the shape control actuator of the rolling equipment.

Specifically, Patent Document 1 describes a method for appropriately initializing a shape control actuator. This is a method of configuring a neural network using influencing factors regarding the flatness of a rolled material as input data. Patent Document 1 discloses that an optimum setting value calculation device calculates the optimum setting value of each operating end for the actual shape value when rolling with a certain setting value, and a pair of influencing factors and each operating end setting value is calculated. It is described that data is accumulated and used as training data for a neural network. In this case, the neural network has a hierarchical structure, and learning is performed by error backpropagation learning with influencing factors as input and each operating terminal setting value as output.

Further, Patent Document 2 also describes a method for appropriately initializing a shape control actuator. This is a method in which rolling conditions for a steel plate, setting values for a shape control actuator, and shape data are accumulated in advance as shape control performance data, and typical data (prototype) is extracted by learning using a neural network. Then, before executing a rolling pass, the degree of similarity with typical data regarding the rolling conditions and target shape of the steel plate is calculated, and based on the calculated degree of similarity, the shape control actuator settings are combined with the shape control actuator based on the typical data. Perform initial settings.

On the other hand, Patent Document 3 exemplifies shape control of a steel plate using rolling equipment, in which shape deviation data, which is the difference between the actual shape and the target shape by a shape detector, is input, and the operation amount of a shape control actuator is output. Describes how to configure neural networks. Further, Patent Document 3 also describes shape feedback control in which a shape control actuator is reset in accordance with a detected shape deviation. Further, Patent Document 3 describes a method of optimizing the control output of feedback control by providing a control result quality determination section that determines the quality of the control output of the shape control actuator.

Japanese Patent Application Publication No. 4-167908 Patent No. 2677964 Japanese Patent Application Publication No. 2018-180799

However, the methods described in Patent Document 1 and Patent Document 2 are related to an initial setting method of a shape control actuator, and are not related to shape feedback control that controls the shape of a steel sheet in the longitudinal direction. For this reason, it may not always be possible to control the shape of the steel plate to a desired shape over the longitudinal direction. On the other hand, the method described in Patent Document 3 can be applied to dynamic shape control (shape feedback control) that controls the shape of a steel plate in rolling equipment, and can control the shape of the steel plate in the longitudinal direction to a target shape. . However, in the method of configuring a neural network in which shape deviation data, which is the difference between the actual shape and the target shape determined by the shape detector, is input and the operation amount of the shape control actuator is output, shape deviation data is generated every control cycle of shape control. changes. Therefore, the amount of operation of the shape control actuator changes from time to time. In other words, unless the actual shape and target shape fall within a predetermined deviation, the set value of the shape control actuator will always fluctuate during rolling of the steel plate, and such fluctuations may cause disturbances when rolling the steel plate. Therefore, the shape of the steel plate in the longitudinal direction is not stable. Furthermore, fluctuations in the setting values of the shape control actuator cause fluctuations in the lubrication and friction conditions at the contact area between the steel plate and the rolling rolls, resulting in appearance defects such as uneven color tone on the surface of the steel plate after rolling. Yield decreases. Moreover, Patent Document 3 describes a control result quality determination unit that determines the quality of the control output of the shape control actuator. However, this is to determine whether the performance data (shape deviation data in this case) obtained as a result of changing the set value of the shape control actuator has improved compared to before the change. For this reason, it is not possible to solve the problem that the set value of the shape control actuator fluctuates during rolling of the steel plate.

The present invention has been made to solve the above problems, and its purpose is to provide a shape control actuator setting model that can accurately control the shape of a steel plate in the longitudinal direction and improve the stability of the shape of the steel plate. The object of the present invention is to provide a method for generating a shape control actuator setting model for rolling equipment that can generate a shape control actuator setting model for rolling equipment. Another object of the present invention is a method for setting a shape control actuator for rolling equipment, and a method for controlling the shape of a steel plate, which can control the shape of a steel plate with high precision in the longitudinal direction and improve the stability of the shape of the steel plate. An object of the present invention is to provide a shape control device for a steel plate. Furthermore, another object of the present invention is to provide a method for producing a steel plate that can produce a steel plate having a desired shape in the longitudinal direction with a high yield.

A method for generating a shape control actuator setting model for rolling equipment according to the present invention includes a shape control actuator for rolling equipment that includes a shape control actuator that controls the shape of a steel plate, and a shape detector that detects the shape of the steel plate. A method for generating a shape control actuator setting model for a rolling equipment for setting, the method comprising: one or more operation performance data selected from rolling operation parameters of the rolling equipment; and shape performance data acquired from the shape detector; , is the input performance data, setting change information indicating whether to change the settings of the shape control actuator, and the setting value of the shape control actuator are the output performance data, and the neural network uses a plurality of learning data. The method includes a setting model generation step of generating a shape control actuator setting model using a network method.

The rolling operation parameters preferably include target shape data of the steel plate.

The neural network includes, as an output layer, a first output layer that outputs the setting change information, and a second output layer that outputs the setting value of the shape control actuator, and the first output layer and the second output layer output the setting change information. The output layer may be a network structure branching from a common intermediate layer.

A method for setting a shape control actuator in a rolling equipment according to the present invention sets a shape control actuator in a rolling equipment including a shape control actuator that controls the shape of a steel plate, and a shape detector that detects the shape of the steel plate. A method for setting a shape control actuator for a rolling equipment, wherein one or more operation data selected from rolling operation parameters of the rolling equipment and shape data acquired from the shape detector are input data, and the Using a shape control actuator setting model generated by a neural network method, in which setting change information indicating whether to change the settings of the shape control actuator and the setting value of the shape control actuator are output data, If the setting change information is for changing the settings of the shape control actuator, the method includes a resetting step of resetting the shape control actuator to the set value.

The neural network includes, as an output layer, a first output layer that outputs the setting change information, and a second output layer that outputs the setting value of the shape control actuator, and the first output layer and the second output layer output the setting change information. The output layer has a network structure branched from a common intermediate layer, and the resetting step includes changing the configuration of the shape control actuator when the output of the first output layer is to change the configuration of the shape control actuator. It is preferable to reset to the above setting value.

A method for controlling the shape of a steel plate according to the present invention includes a step of resetting the shape control actuator during rolling of a steel plate using the method for setting a shape control actuator for a rolling equipment according to the present invention.

The shape control device for a steel plate according to the present invention includes means for resetting the shape control actuator during rolling of a steel plate using the method for setting the shape control actuator for a rolling equipment according to the present invention.

The method for manufacturing a steel plate according to the present invention includes the step of manufacturing a steel plate using the method for controlling the shape of a steel plate according to the present invention.

According to the method for generating a shape control actuator setting model for rolling equipment according to the present invention, it is possible to generate a shape control actuator setting model that accurately controls the shape of a steel plate in the longitudinal direction and improves the stability of the shape. Further, according to the method for setting a shape control actuator for rolling equipment, the method for controlling the shape of a steel plate, and the shape control device for a steel plate according to the present invention, the shape of the steel plate can be precisely controlled in the longitudinal direction, and the stability of the shape can be improved. can be done. Further, according to the method for manufacturing a steel plate according to the present invention, a steel plate having a desired shape along the longitudinal direction can be manufactured with a high yield.

FIG. 1 is a side view showing the configuration of a rolling facility that is an embodiment of the present invention. FIG. 2 is a front view showing a schematic configuration of a shape control actuator and a rolling mechanism. FIG. 3 is a block diagram showing the configuration of a shape control actuator setting model generating section according to an embodiment of the present invention. FIG. 4 is a diagram showing an example of shape performance data. FIG. 5 is a diagram for explaining a method of specifying setting change information from time-series data of actuator setting values. FIG. 6 is a diagram showing the configuration of a neural network that is an embodiment of the present invention. FIG. 7 is a diagram showing a configuration example of a steel plate shape control system including a shape control actuator setting section in a rolling equipment. FIG. 8 is a diagram showing a comparison result between the output value and the actual value of the shape control actuator setting value in the present example and the comparative example. FIG. 9 is a diagram showing the results of evaluating the shape deviation in the longitudinal direction of the rolled steel sheets for this example, comparative example 2, and comparative example 3.

Hereinafter, one embodiment of the present invention will be described with reference to the drawings.

[Rolling equipment]
First, with reference to FIGS. 1 and 2, the configuration of a rolling facility that is an embodiment of the present invention will be described. In this embodiment, equipment including at least a rolling mill and a shape detector will be referred to as rolling equipment.

FIG. 1 is a side view showing the configuration of a rolling facility that is an embodiment of the present invention. As shown in FIG. 1, a rolling equipment 1 according to an embodiment of the present invention is comprised of a 12-high cluster type rolling mill. The 12 stages of rolls include a work roll 2 consisting of one upper and lower pair, an intermediate roll 3 consisting of two upper and lower pairs, a small backup roll 4 consisting of one upper and lower pair, two upper pairs of upper backup rolls 5, and two lower pairs of lower rolls. It is equipped with a backup roll 6. A steel plate S, which is a rolled material, is supplied from one of the payoff reel 7, the right tension reel 8, and the left tension reel 9, is rolled, and is wound onto the opposite tension reel. The shape detector 10 detects the difference in elongation of the steel plate S in the width direction.

FIG. 2 is a front view showing a schematic configuration of a shape control actuator (operating end) and a rolling mechanism. The steel plate S is rolled and stretched by the work rolls 2 from above and below. The upper backup roll 5 is divided into seven parts in the axial direction, and the roll crown can be adjusted by extruding each of the center 51a, quarter-in 51b, quarter-out 51c, and edge 51d in order from the axial center position. . The lower backup roll 6 is divided into six parts in the axial direction, and the roll crown can be adjusted by extruding the center 61a, quarter 61b, and edge 61c in order from the axial center position.

The rolling equipment is equipped with a roll crown adjustment function for the upper backup roll 5 and lower backup roll 6 as shape control actuators, an intermediate roll bender 11, and a work roll bender 12 that applies bending force to the upper and lower work rolls 2. The shape control actuator may include a leveling mechanism that tilts the axial center of the upper backup roll 5 or the lower backup roll 6 in the width direction. In addition to these shape control actuators, there is also an intermediate roll shift in which the upper and lower intermediate rolls 3 are shifted in opposite directions with respect to the axial direction of the rolls, and a work roll in which the upper and lower work rolls 2 are shifted in opposite directions to each other. Shifts may also be provided.

Returning to FIG. 1, the shape detector 10 is a measuring instrument that measures the shape of a steel plate S rolled by a rolling mill. As the shape detector 10, a means commonly used for measuring the shape of the steel plate S may be selected. The shape of the steel plate S is caused by the distribution of the line segment lengths in the longitudinal direction of the steel plate S in the width direction. -unit (1×10 ⁻⁵ )) is usually used for evaluation. Specifically, when the length of the steel plate S is distributed in the width direction, the contact pressure (vertical load to the contacting roll) on the roll that contacts the steel plate S at a constant wrap angle is increased in the width direction. The distribution will be as follows. The shape detector 10 calculates the difference in elongation by measuring the distribution of the contact pressure. At this time, the widthwise distribution of the contact pressure is detected by load detectors that are divided and arranged in the axial direction inside the contact roll (divided load cell method).

However, as the shape detector 10, it is possible to detect the out-of-plane displacement of the steel plate S using a non-contact distance meter (such as a laser distance meter) and convert it into a difference in elongation, or to calculate the difference in elongation from moiré fringes irradiated on the steel plate S. You may also use the method of The difference in elongation rate or the difference in elongation (I-unit) shall be measured from the line segment length in the width direction of the steel plate S using any of the shape measuring instruments. The number of channels (the number of divisions in the width direction) for detecting the difference in elongation in the width direction is preferably greater than the number of shape control actuators used for shape control. Specifically, the shape detector 10 is divided into 24 to 64 parts in the width direction and is capable of measuring the difference in elongation rate or the difference in elongation at each position.

[Shape control system]
Next, with reference to FIG. 1, a shape control system for the rolling equipment 1 will be described.

As shown in FIG. 1, the shape control system for the rolling equipment 1 includes a control computer 100 and a rolling control controller 110. The control computer 100 acquires information on the base material dimensions (thickness, width, length, etc.) of the steel plate S, information on the deformation resistance of the steel plate S, product thickness, etc. from the host computer, and controls the manufacturing of the steel plate S. When rolling is performed for the number of rolling passes set according to the specifications, a predicted value of the rolling load in each rolling pass is calculated. Then, the control computer 100 sends the initial setting value of the shape control actuator to the rolling control controller 110 for each rolling pass based on the predicted value of the rolling load. The rolling control controller 110 operates each device of the shape control actuator based on the acquired initial setting values. The rolling control controller 110 also acquires information on the target shape of the steel plate S in each rolling pass from the control computer 100.

After the initial setting value of the shape control actuator is set and the rolling pass is started, the rolling control controller 110 resets the setting value of the shape control actuator as needed (shape feedback control). Shape feedback control is performed using the actual shape of the steel plate S acquired from the shape detector 10 and the target shape of the steel plate S acquired from the control computer 100. By executing shape feedback control, even if there is a disturbance such as a change in the thickness of the base material in the longitudinal direction of the steel plate S, the shape of the steel plate S can be brought closer to the target shape, and the steel plate S can be made to have a uniform shape in the longitudinal direction. can be manufactured.

[Shape control actuator setting model]
Next, a shape control actuator setting model that is an embodiment of the present invention will be described with reference to FIGS. 3 to 6.

A shape control actuator setting model, which is an embodiment of the present invention, is provided in the rolling control controller 110 and sends appropriate setting values to the shape control actuator with reference to the actual shape of the steel plate S obtained from the shape detector 10. used for As shown in FIG. 3, the shape control actuator setting model M is generated by the shape control actuator setting model generation unit 140. In the present embodiment, the data acquisition unit 120 performs one or more operations selected from the actual shape data of the steel plate S acquired during rolling and the rolling operation parameters of the rolling equipment 1 in the rolling pass of the steel plate S. Acquire performance data, performance data of setting values of the shape control actuator (hereinafter sometimes referred to as actuator setting values), and performance data of setting change information of the shape control actuator.

Shape control actuator setting change information (hereinafter sometimes referred to as "setting change information") refers to the change in shape control actuator settings from the current shape control actuator settings during rolling equipment operation to the next control period (next step). Information indicating whether or not to change the setting value of the shape control actuator. The setting change information can be represented by binary information in which, for example, "1" indicates that the actuator setting value has been changed, and "0" indicates that the actuator setting value has not been changed. Moreover, the performance data of the setting values of the shape control actuators refers to the setting values of at least one or more shape control actuators selected from the shape control actuators described above. The various data acquired by the data acquisition section 120 constitute a set of data and are accumulated in the database section 141 of the shape control actuator setting model generation section 140, and the shape control actuator setting model M is generated by the machine learning section 143. Ru.

The shape performance data acquired by the data acquisition unit 120 is shape data measured by the shape detector 10 (see FIG. 1), and is illustrated in FIG. 4, for example. The shape data is information specified by the width direction position of the steel plate S and the elongation difference rate or elongation difference (I-unit) at that position. The elongation difference ratio is a value expressed in a strain format of the line segment length of the steel plate S at other positions in the width direction based on the position where the line segment length in the width direction of the steel plate S is the shortest. In addition, the difference in elongation (I-unit) is expressed by converting the difference in elongation into a value multiplied by 100,000, since the shape data will be a very small value even if the difference in elongation is expressed as a percentage. It is expressed by a unit called I-unit. In this case, a case where the difference in elongation rate or difference in elongation at the ends of the steel plate S in the width direction is larger than other parts is referred to as an ear wave, and a case in which the difference in elongation ratio or elongation difference in the center part in the width direction is larger than in other parts. is often called medium elongation. Further, a case where the elongation difference rate or elongation difference between the width direction end portion and the width direction center portion is large is called quarter elongation, and a form in which the above shapes are combined is called composite elongation.

The operation performance data acquired by the data acquisition unit 120 together with the shape performance data of the steel sheet S is one or more operation performance data selected from the rolling operation parameters. Rolling operation parameters are parameters that represent the rolling state when executing a rolling pass, and are parameters that represent factors that may affect the shape of the steel sheet S, excluding the actual shape data of the steel sheet S and the actual setting data of the shape control actuator. It is. As the rolling operation parameters, for example, as dimensional information of the steel plate S, information on the thickness and width of the base material held by a host computer that gives instructions to the control computer 100 may be used. Alternatively, the inlet side plate thickness, outlet side plate thickness, and plate width in the rolling pass for obtaining shape performance data may be used. Another example of rolling operation parameters is attribute information of base material. The attribute information of the base material may be an index representing the characteristics of the base material, such as the deformation resistance of the steel plate S, the steel type, the composition, the surface roughness of the base material, the distribution state of surface oxides, etc. Further, as the attribute information of the base material, manufacturing conditions in a step (pre-step) before executing the rolling step using the rolling equipment 1 may be used. For example, if the previous process is an annealing process, information such as the annealing temperature, soaking time, cooling rate, etc., and if the previous process is a hot rolling process, the coiling temperature on the hot rolling line, etc. is also included in the attribute information of the base material. May be included. This is because these are parameters that indirectly affect the rolling load when shape control is executed in the rolling equipment 1.

Another example of the rolling operation parameters is rolling condition information in a rolling pass for acquiring shape performance data. Examples of rolling condition information include rolling load, rolling torque, rolling speed, entry tension, exit tension, and work roll diameter. Moreover, parameters related to lubrication conditions during rolling, such as coolant flow rate, coolant temperature, and coolant concentration, may be used as the rolling condition information. Each parameter of the dimensional information of the steel plate S, the attribute information of the base material, and the rolling condition information may be set using either set values set in advance by the host computer or control computer 100 or actual values actually measured by the rolling equipment 1. good. When the rolling equipment 1 is equipped with measurement equipment such as a plate thickness gauge for measuring the thickness of the steel plate S, a load cell for measuring rolling load, and a tension meter for measuring the tension applied to the steel plate S, the actual measured values thereof. It is preferable to use

It is preferable that the rolling operation parameters include data on the target shape of the steel sheet S set in the rolling controller 110. The target shape of the steel plate S may be a target value of the elongation difference rate at each position in the width direction of the steel plate S, as illustrated as the actual shape data in FIG. However, as the target shape of the steel plate used in the rolling operation parameters of this embodiment, a polynomial approximation may be performed using the target elongation difference as a function of the position in the width direction, and the values of the coefficients of the polynomial may be used. For example, as a target shape of the steel plate S, the width direction distribution of the target elongation difference (I-unit) is approximated by a quartic function and expressed as in the following equation (1). However, z in formula (1) is a position in the width direction where the board width is normalized in the range of -1 to +1. The coefficients λ1 to λ4 in this case may be included in the rolling operation parameters as data on the target shape of the steel plate.

As the operation performance data of this embodiment, at least one parameter is selected from the above-mentioned rolling operation parameters, and operation performance data regarding this parameter is used. On the other hand, the rolling operation parameters may include identification data of an operator who manually operated the shape control actuator of the rolling equipment 1. When the shape of the steel plate S is controlled by an operator's manual operation, numbers, codes, names, etc. may be included in the rolling operation parameters as identification data that can identify the operator who performed the operation. Shape control of the steel plate S exhibits unique control performance depending on the characteristics of the shape control actuator of the rolling equipment 1. This is because there is a difference in the superiority and inferiority of Therefore, the shape control performance of the steel plate S is separately evaluated based on the difference between the target shape and the actual shape, etc., and a shape control actuator setting model M is generated that selectively outputs the operator's operation data with good evaluation results. Because it can be done.

The setting performance data of the shape control actuator of the rolling equipment 1 refers to the setting information of the shape control actuator specified according to the rolling mill used. When the rolling mill includes a plurality of shape control actuators, any shape control actuator may be selected from the shape control actuators and used as the shape control actuator setting performance data of this embodiment. As the setting information of the shape control actuator, a leveling setting value for controlling the tilting state of the axial center of the upper backup roll 5 or the lower backup roll 6 can be used. Further, set values for the bender force of the intermediate roll bender 11 and the work roll bender 12 may be used. Furthermore, a roll shift setting value representing the shift amount of the intermediate roll 3 or work roll 2 in the axial direction may be used. Further, the 12-high cluster type rolling mill shown in FIG. 1 is equipped with a shape control actuator that adjusts the position of the divided backup rolls. Therefore, the set position information of each backup roll, such as the center position, quarter-in position, quarter-out position, and edge position of the upper backup roll 5, and the center position, quarter position, and edge position of the lower backup roll 6, is transferred to the shape control actuator. May be included in configuration information. As the setting performance data of the shape control actuator, the setting values (command values) of each operating end illustrated above may be used, or actual values may be used when actually measured. Furthermore, if the bending force or the like can be calculated from a measured value such as hydraulic pressure, it is preferable to use the actual measured value instead of the set command value.

Regarding the selection of the shape control actuator, it is preferable to select the shape control actuator used for dynamic control during rolling of the steel plate from among all the shape control actuators. Thereby, a steel plate having a desired shape in the longitudinal direction can be manufactured with high yield. The performance data of the configuration change information of the shape control actuator is information representing the actual performance of whether or not the actuator setting value has been changed in the next control cycle or later (next step) from the actuator setting value at the current point in time when the rolling equipment is in operation. It is. It can be represented by binary information in which "1" indicates that the setting value of the shape control actuator has been changed, and "0" indicates that the setting value has not been changed. The performance data of the configuration change information of the shape control actuator may be acquired as performance data indicating whether the operator of the rolling equipment changed the shape control actuator using an operation switch connected to the rolling control controller 110 or the like.

On the other hand, it is also possible to specify the performance data of the setting change information after acquiring the performance data of the actuator setting values as time-series data at regular intervals (for example, every control cycle). FIG. 5 is a diagram for explaining a method of specifying setting change information from time-series data of actuator setting values. FIG. 5 shows actual data of actuator setting values at times t1, t1+Δt, t1+2Δt, and t1+3Δt. However, time t1 is an arbitrary time during rolling of the steel plate, and Δt is a control period (for example, 0.2 to 10 seconds) for performing shape control. In this example, actual data of actuator setting values at time t1 and time t1+Δt are compared. As a result of the comparison, if any actuator setting value has changed, it is determined that the setting value has changed, and if any actuator setting value has not changed, it is determined that the setting value has not changed. Then, the determination result is assigned as performance data of the setting change information at time t1. For example, in the example shown in FIG. 5, since none of the actuator setting values has changed between time t1 and time t1+Δt, the performance data of the setting change information at time t1 is set to "0". On the other hand, since one of the actuator setting values changes between time t1+Δt and time t1+Δt, the performance data of the setting change information at time t1+Δt is set to “1”. In this way, the performance data of the setting change information can be specified based on the performance data of the actuator setting values that are continuously acquired. Therefore, the data acquisition unit 120 shown in FIG. 3 acquires the performance data of the setting change information as continuous data at regular intervals, and the data processing unit specifies the performance data of the setting change information as described above. may be provided.

The shape performance data, operation performance data of rolling operation parameters, performance data of actuator setting values, and performance data of setting change information acquired by the data acquisition unit 120 are handled based on the manufacturing control number and rolling pass information of the steel plate S. The data are assigned and stored in the database unit 141 as a set of data. A plurality of data sets accumulated in the database section 141 can be generated along the longitudinal direction of the steel plate S even in the same rolling pass of the same steel plate S. For example, when performance data is acquired at predetermined intervals (for example, 100 ms) during rolling, the number of data sets acquired per rolling pass becomes enormous. Therefore, when accumulating data sets in the database unit 141, the number of data sets accumulated in the database unit 141 can be reduced by performing a screening process to extract only the data sets necessary for generating the shape control actuator setting model M. You can reduce it.

For example, it takes about 0.2 to 10 seconds for the steel plate S to reach the shape detector 10 after the shape control actuator of the rolling equipment 1 is reset. Therefore, the sampling period for acquiring data sets may be set in the range of 0.2 to 10 seconds, and only the data sets extracted in this manner may be stored in the database unit 141. In this case, the sampling period may be changed depending on the rolling speed. In addition, in areas where the plate thickness and deformation resistance of the base material vary greatly, such as near the tip and tail ends of the steel plate S, the shape control actuator of the rolling equipment 1 must be reset before the steel plate S detects the shape. Even before reaching the container 10, the rolling state may change significantly over time. In this case, the correspondence between the set value of the shape control actuator and the actual shape may not be clear. In such a case, processing may be performed to exclude the data set acquired at the tip or tail end of the steel plate S from the database section 141. Furthermore, if the output of the shape detector 10 regarding the actual shape changes greatly in a short period of time, there is a possibility that a detection error of the shape detector 10 has occurred. Therefore, if the actual shape changes in a short period, screening processing may be performed to exclude it from the database section 141. On the other hand, there is a data set in which the performance data of the setting change information is "1" (shape control actuator settings changed), and the data set where the performance data of the setting change information is "0" (shape control actuator settings not changed). Screening may be performed to preferentially delete part of the data set.

The shape control actuator setting model generation unit 140 generates a shape control actuator setting model M in the machine learning unit 143 using the learning data accumulated in the database unit 141. The shape control actuator setting model generation unit 140 may be located inside the control computer 100 that controls the rolling process of the steel plate S by the rolling equipment 1, or may be configured by separate hardware from the control computer 100. . Alternatively, it may be placed in the shape control actuator setting section 150, which will be described later.

The number of data sets stored in the database unit 141 is 10,000 or more, preferably 100,000 or more, and more preferably 1,000,000 or more. Alternatively, the data set may be divided based on the attribute information of the steel sheet S or the rolling pass number, and the shape control actuator setting model M may be generated according to the classification. The number of data sets stored in the database unit 141 is limited to a certain number of data sets, and the data sets stored in the database unit 141 may be updated as appropriate within the upper limit.

After the learning data is accumulated in the database unit 141, the preliminary processing unit 142 performs processing on the data set accumulated in the database unit 141 as necessary before executing machine learning to generate the shape control actuator setting model M. Perform preliminary processing on the Examples of the preliminary processing performed by the preliminary processing unit 142 include standardization of actual data of shape data and setting values of shape control actuators, data shuffling, and the like.

The preliminary processing unit 142 performs standardization processing on the shape performance data that is the input performance data accumulated in the database unit 141, the operation performance data selected from the rolling operation parameters, and the performance data of the setting values of the shape control actuator that is the output performance data. You may do so. The normalization process is a process for efficiently updating weighting coefficients during network learning, and is effective in avoiding the vanishing gradient problem when learning a neural network. When standardizing input variables and output variables, it is preferable to specify the minimum and maximum values of each variable and standardize them within the range of 0 to 1. Alternatively, the average value of the variables may be calculated and normalized within the range of -1 to +1.

The database section 141 contains data consisting of shape performance data as input performance data, operation performance data selected from rolling operation parameters, and performance data of shape control actuator setting performance data and setting change information as output performance data. Sets are accumulated in the order in which data is retrieved. On the other hand, data shuffling may be performed to randomly rearrange the order of the data sets accumulated in the database unit 141. Changing the order of data sets acquired in time series is effective in improving the generalization performance of the neural network and suppressing overfitting.

The data set accumulated in the database unit 141 may be divided into learning data and test data. The training data is 70 to 80% of the total data set and is used for training the neural network. On the other hand, the remaining data set is used to confirm whether the neural network obtained by completing the learning by the machine learning unit 143 can exhibit good performance even on test data.

The machine learning unit 143 uses the data set accumulated in the database unit 141 to generate a shape control actuator setting model M by a neural network method using a plurality of learning data (setting model generation step). The learning data includes one or more operation performance data and shape performance data selected from rolling operation parameters as input performance data, and shape control actuator setting change information and actuator setting values as output performance data. As the machine learning model for generating the shape control actuator setting model M, any machine learning model can be used as long as a practically sufficient estimation accuracy of the shape control actuator setting values can be obtained. Examples include commonly used neural networks, decision tree learning, random forests, support vector regression, etc. Alternatively, an ensemble model that combines a plurality of models may be used. Further, this embodiment uses a neural network as a machine learning method, and includes a first output layer that outputs setting change information as an output layer, a second output layer that outputs actuator setting values, and a first output layer that outputs setting change information. Preferably, the layer and the second output layer have a network structure branching from a common intermediate layer.

The neural network of this embodiment will be described with reference to FIG. 6. In the neural network of this embodiment, shape data acquired from the shape detector 10 and operation data selected from rolling operation parameters of the rolling equipment 1 are input from the input layer. The neural network of this embodiment includes one or more intermediate layers and two output layers branching from a common intermediate layer for these input data. The configuration is such that the first output layer outputs setting change information of the shape control actuator, and the second output layer outputs the setting value of the shape control actuator. The first output layer is a classifier that outputs as setting change information whether to change the settings of the shape control actuator ("1") or not ("0"), and uses a sigmoid function as the activation function. good. On the other hand, the second output layer is a regressor that outputs the setting value of the shape control actuator used as output performance data when generating the shape control actuator setting model M, and an identity function may be used as the activation function. . However, if the actuator setting values are standardized when generating the shape control actuator setting model M, it is preferable to convert the output from the second output layer into a physical quantity that is the setting value of the shape control actuator. . The intermediate layer of the neural network preferably ranges from 1 to 10 layers, and the number of nodes in the intermediate layer preferably ranges from 32 to 2048. This is because if the number of intermediate layers or nodes is too small, the setting accuracy of the shape control actuator setting model M will decrease, and if there are too many, overfitting will become noticeable and the prediction accuracy may actually decrease. From the viewpoint of prediction accuracy, it is preferable to use the ReLU function as the activation function in the intermediate layer.

On the other hand, in the neural network of this embodiment, a batch normalization layer that performs batch normalization of data immediately before the activation function calculation in the intermediate layer may be provided. Batch normalization refers to an operation that normalizes data using the mean and variance calculated for each batch. This makes learning the weighting coefficients more efficient. Additionally, a dropout layer may be provided downstream of the activation function in the intermediate layer. The dropout layer is an operation that invalidates the output with a certain probability (for example, 0.1 to 0.4). This is because overfitting can be suppressed.

The reason why the output of the shape control actuator setting model M is divided into the setting change information and the actuator setting value as described above is because the following shape control system for rolling equipment can be configured. In other words, if the setting change information, which is the first output of the shape control actuator setting model, is "0" (no change in shape control actuator settings), the current settings are maintained without changing the shape control actuator settings. Control to maintain the value. On the other hand, if the first output of the shape control actuator setting model, which is the setting change information, is "1" (shape control actuator settings changed), the second output, which is the shape control actuator setting value (actuator setting (value), change the settings of the shape control actuator of the rolling equipment. This can solve the problem of the prior art in which the setting value of the shape control actuator is frequently changed during rolling, and fluctuations in the rolling state cause disturbance and the shape of the steel plate in the longitudinal direction is not stable. Further, it is possible to suppress fluctuations in the lubrication and friction conditions at the contact portion between the steel plate and the rolling roll, thereby suppressing a decrease in yield due to poor appearance and the like. In particular, when an operator with extensive rolling equipment operation experience manually operates the shape control actuator, the shape control actuator must be actively operated in response to small fluctuations in rolling operation parameters or shape performance data during operation. is not carried out. On the other hand, if rolling operation parameters or actual shape data change significantly during operation, the settings of the shape control actuator are actively corrected to prevent the shape of the steel sheet from being significantly disturbed. In the above embodiment, the shape control actuator operation is performed in the longitudinal direction by specifying the actual data of the shape control actuator operation performed based on the rolling operation actual data along with other operating conditions, similar to the operation by an operator who has extensive operating experience of rolling equipment. Achieves shape control of rolling equipment with stable shape. Therefore, by including the identification data of the operator who manually operated the shape control actuator in the input data of the shape control actuator setting model M, it becomes possible to set the shape control actuator to reproduce the operations of an excellent operator.

On the other hand, the network structure in which the first output layer and the second output layer are branched from a common intermediate layer is such that, even if the operator decides whether or not to change the settings of the shape control actuator, it is necessary to change the settings. This is because it is associated with the setting change amount. In other words, when changing the settings of the shape control actuator, it is likely that the shape of the steel plate will change to a certain degree, and in that case, the shape control actuator is often changed to a certain extent. Therefore, since it is difficult to evaluate the setting change information and the actuator setting value separately, a common intermediate layer is used. Since it is preferable that the first output layer is configured as a classifier and the second output layer is configured as a regressor, it is preferable to adopt a neural network structure having two different output layers as described above.

By using the shape actual data acquired from the shape detector 10 as the input actual data of the shape control actuator setting model M, the settings of the shape control actuator can be changed significantly when the shape of the steel plate S is greatly disturbed. . In addition, by using the target shape data as the input performance data of the shape control actuator setting model M, it is possible to determine whether or not to change the shape control actuator settings and to change the actuator settings depending on the deviation between the shape performance data and the target shape data. The amount of change is set. This makes it possible to achieve highly accurate shape control.

The reason why the input of the shape control actuator setting model M includes not only the shape data of the steel sheet S but also the operation performance data selected from the rolling operation parameters is as follows. In other words, since the rolling operation parameters affect the rolling load during rolling of the steel plate S, and the bending deformation of the work roll changes in a complicated manner depending on the rolling load, there is a certain correlation between the rolling operation parameters and the actual shape data. Because it exists. In addition, when controlling the shape of the rolling equipment 1, even if the setting values of the shape control actuator are the same, if the rolling conditions such as the thickness, width, rolling reduction, deformation resistance, etc. of the steel plate S are different, the rolling This is because it is known that the shapes of the steel plates S are different.

By performing learning by dividing the data set stored in the database unit 141 into training data and test data, it is possible to improve the setting accuracy of shape control actuator setting change information and actuator setting values. For example, the weighting coefficients of the neural network are learned using training data, and the structure of the neural network (number of intermediate layers and number of nodes) is adjusted so that the correct answer rate for setting change information and actuator setting values in test data is high. The shape control actuator setting model M may be obtained while changing it as appropriate. An error propagation method can be used to update the weighting coefficients.

The shape control actuator setting model M may be updated to a new model by relearning, for example, every month or every year. This is because the more data stored in the database section 141, the more accurate shape control actuator setting becomes possible. Furthermore, by updating the shape control actuator setting model M based on the latest data, it is possible to generate a shape control actuator setting model M that reflects changes in operating conditions over time.

[Steel plate shape control method]
Next, with reference to FIG. 7, a method for controlling the shape of a steel plate, which is an embodiment of the present invention, will be described.

In the method for controlling the shape of a steel plate, which is an embodiment of the present invention, the shape control actuator of the rolling equipment 1 is set using the shape control actuator setting model M generated by the machine learning section 143. FIG. 7 shows a configuration example of a steel plate shape control system including a shape control actuator setting section 150 in the rolling equipment 1. The shape control actuator setting section 150 shown in FIG. 7 may be provided, for example, in a general-purpose computer such as a workstation or a personal computer. Further, the shape control actuator setting section 150 includes an input section for acquiring the shape control actuator setting model M generated by the shape control actuator setting model generation section 140 via the network, and a storage section for storing it. However, the shape control actuator setting section 150 may be configured inside the control computer 100 or the rolling control controller 110 for controlling the rolling equipment 1.

In the shape control of the steel plate S using the shape control actuator setting section 150 shown in FIG. After that, the rolling pass is started. Then, during execution of the rolling pass, the data acquisition unit 120 acquires operation performance data selected from the rolling operation parameters of the rolling equipment 1 and shape performance data acquired from the shape detector 10. The data acquired by the data acquisition section 120 is sent to the shape control actuator setting section 150. The shape control actuator setting section 150 inputs the input data obtained by the data acquisition section 120 to the shape control actuator setting model M obtained from the shape control actuator setting model generation section 140 and stored in the storage section. Accordingly, setting change information and actuator setting values are calculated. Then, based on the calculated setting change information and actuator setting value, a new shape control actuator setting command value is generated and sent to the rolling control controller 110. When the configuration change information indicates that the configuration of the configuration control actuator is to be changed, the configuration control actuator setting command value sets the calculated actuator setting value as the new configuration control actuator configuration command value. If represents that the setting of the shape control actuator is not changed, the change to the new shape control actuator setting command value is not performed so as to maintain the current shape control actuator setting value. Thereby, the rolling control controller 110 operates the shape control actuator of the rolling equipment 1 by updating or maintaining the current shape control actuator setting based on the shape control actuator setting command value acquired from the shape control actuator setting section 150. Can be done.

By acquiring the shape control actuator setting command value every preset shape control control cycle and sending it to the rolling control controller 110, the shape control actuator setting value is updated in the longitudinal direction of the steel plate S, and the shape control actuator setting value is stabilized. The shape of the steel plate S can be controlled. The control cycle of the shape control is approximately 0.2 to 10 seconds, since it takes a certain amount of time from updating the shape control actuator settings of the rolling equipment 1 until the steel plate S reaches the shape detector 10. It is preferable to do so. Further, in the conventional automatic shape feedback control, shape control actuator setting values are updated at any time based on shape actual data detected by the shape detector 10 using a shape control actuator setting model M configured by a physical model or the like. There are many. Therefore, the shape control of a steel plate using the shape control actuator setting section 150 of this embodiment can be realized by replacing a part of the conventional system for executing automatic shape feedback control. Further, the shape control actuator setting section 150 of this embodiment may be provided in parallel with the conventional automatic shape feedback control system, and the shape control of the rolling equipment may be executed by switching as necessary. For example, when rolling a steel plate using multiple rolling passes, there may be large variations in the material and plate thickness of the base material in the first pass, and it may be necessary to operate the shape control actuator frequently. In the initial rolling pass, shape control may be performed using a conventional automatic shape feedback control system, and in subsequent rolling passes, shape control of the steel plate may be performed using the shape control actuator setting unit 150 of this embodiment. . In addition, at the tip and tail ends of the steel plate, the shape at the entrance of the rolling mill may be disordered, and the shape control actuator may need to be operated frequently. Shape control may be performed using a conventional automatic shape feedback control system, and shape control of the steel plate may be performed in the steady state using the shape control actuator setting section 150 of this embodiment.

Examples of the present invention are shown below. In this example, first, a shape control actuator setting model was generated for a 12-high cluster rolling mill. The rolling mill used in this example has a work roll diameter of 70 to 120 mm and a maximum rolling speed of 600 m/min. The steel plate that is the material to be rolled is high carbon steel, and has a base material plate thickness of 1.2 to 2.3 mm, a plate width of 400 to 1070 mm, and a product plate thickness of 0.5 mm by rolling passes of 1 to 15. A steel plate of 4 to 1.8 mm was rolled. In addition, the rolling mill used was a lever type rolling mill, and a shape detector was placed between the two tension reels on the left and right sides of the rolling mill. You can obtain shape data of steel plates. The shape detector is a divided load cell type, and has 44 built-in load cells in the roll axis direction, and the detection width per channel is 25.4 mm. The shape detector outputs the value of I-unit at each position in the width direction of the steel plate as the difference in elongation in the width direction of the steel plate. In addition, in this example, 420 coils were rolled on the above steel plate, and during that time, an operator with extensive operating experience of this rolling mill manually operated the shape control actuator, and the learning data was stored in the database section. accumulated in

In generating the shape control actuator setting model M, the inlet side plate thickness, outlet side plate thickness, and plate width were selected as the dimensional information of the steel plate from among the rolling operation parameters. In addition, the steel type code of the steel plate was selected as the attribute information of the base material. Furthermore, the rolling pass number, work roll diameter, rolling speed, coolant flow rate, and coolant temperature were selected from the rolling condition information. In this example, in addition to these rolling operation parameters, the coefficients λ1 to λ4 when the shape of the steel plate shown in Equation (1) is functionally approximated by a quartic equation were selected as the target shape of the steel plate. On the other hand, the I-unit output from the shape detector as described above was used to obtain actual shape data when a steel plate was rolled with the above rolling operation parameters. However, the position in the width direction of the steel plate was standardized by the maximum width that could be rolled by this rolling mill, and actual data of I-unit at each position was collected.

On the other hand, the shape control actuator used for the actuator setting values that are the output data of the shape control actuator setting model M has six positions: work roll bender, intermediate roll bender, leveling, lower backup roll edge position, quarter position, and center position. selected and obtained actual data for these settings. In this example, performance data of setting change information was identified by the method shown in FIG. 5 from performance data of actuator setting values acquired as time-series data. Then, a data set consisting of shape performance data, operation performance data of rolling operation parameters, performance data of shape control actuator setting change information, and performance data of shape control actuator setting values was accumulated in the database section 141. When the number of data sets accumulated in the database unit 141 reaches 100,000, the shape control actuator setting model generation unit 140 performs data division through preliminary processing, and uses 70% of the total data sets as learning data. was used to train the neural network, and the remaining dataset was used as test data to evaluate prediction accuracy.

As shown in Figure 6, the neural network with two output layers used in this example has a network structure consisting of an input layer, seven intermediate layers each having 512 nodes, a first output layer, and a second output layer. Using. The ReLU function was used as the activation function. In this embodiment, the output of the activation function is input to a dropout layer with a probability of 0.3. The number of nodes in the input layer was 64, the first output layer was 1 node, the second output layer was 6 nodes, and the setting values of the shape control actuators were standardized.

As the loss function in the learning calculation, we used the mean squared error calculated from the predicted value and actual measured value of the network, and selected ADAM optimization as the optimization calculation method. In addition, mini-batch learning (mini-batch gradient descent method) was used for machine learning, with a batch size of 512 and a number of epochs of 500. However, if the loss function for the test data did not improve over 30 epochs or more, the learning rate was lowered to 1/10 and learning proceeded. Using the shape control actuator setting model M generated as described above, the output value and the actual value of the shape control actuator setting value were compared with respect to the test data. In this case, the output value of the shape control actuator setting value is updated to the value output by the second output layer if the output of the first output layer is to change the settings of the shape control actuator. . On the other hand, when the output of the first output layer was such that the setting of the shape control actuator was not changed, the output value of the shape control actuator setting value was maintained at the value of the previous step.

On the other hand, as a comparative example, a normal neural neural Learning was performed using a network, and the output values of the generated model and actual values were compared. In this case, the output value of the shape control actuator setting value was updated in accordance with the update of the input data. Note that the input data to the neural network of the comparative example was the same as that of the present example.

FIG. 8 shows a comparison result between the output value and the actual value of the shape control actuator setting value in the present example and the comparative example. FIG. 8 shows the calculated differences between the output values and actual values of the six shape control actuators for the present example and the comparative example, and the mean square error (MSE) thereof is expressed as a loss function. From FIG. 8, it can be seen that the loss function of the prediction result using the neural network having two output layers of the present example is significantly improved compared to the comparative example. This indicates that the shape control actuator setting value reset by the shape control actuator setting method according to the present example matches the result manually set by the operator relatively well. In other words, it was shown that it was possible to suppress the setting value of the shape control actuator from changing during rolling of the steel plate.

Subsequently, in Comparative Example 2, rolling was carried out using conventional automatic shape feedback control to perform shape control on the same rolled material as in the example. In the conventional automatic shape feedback control of Comparative Example 2, the shape control actuator setting value was updated as needed according to shape deviation data, which is the difference between the actual shape and the target shape detected by the shape detector 10. Furthermore, in Comparative Example 3, in the conventional automatic feedback control of Comparative Example 2, the shape control actuator setting value is updated only when the shape deviation data exceeds a preset threshold value. Reduced update frequency. In this case, the threshold value for updating the setting value of the shape control actuator was set to 10 (I-unit).

FIG. 9 shows the results of evaluating the shape deviation in the longitudinal direction of the rolled steel sheets, that is, the deviation between the target shape and the actual shape, using the absolute value error MAE for this example, comparative example 2, and comparative example 3. The larger the absolute value error MAE is, the greater the deviation between the target shape and the actual shape is. In Comparative Example 2, the setting value of the shape control actuator was frequently changed during rolling, and fluctuations in the rolling state caused disturbances, so that the shape of the steel plate in the longitudinal direction was not stable. Therefore, as shown in FIG. 9, in Comparative Example 2, the absolute value error MAE was 8.71 (I-unit). On the other hand, in Comparative Example 3, a threshold value was set for the deviation between the target shape and the actual shape for determining whether or not the shape control actuator can be operated, so the degree to which the shape control actuator setting value is frequently changed during rolling is alleviated. The shape of the steel plate in the longitudinal direction was stabilized to some extent. However, since the provision of the threshold decreased the responsiveness to the shape deviation data, the absolute value error MAE remained at 9.2 (I-unit). Further, when the shape deviation data is close to a set threshold value, the shape control actuator setting value may be updated frequently, and the shape of the steel plate is unstable in some parts in the longitudinal direction. On the other hand, in the example, when a large variation with respect to the target shape occurred, by actively operating the shape control actuator, it was possible to suppress the large disturbance in the shape of the steel plate. In addition, control has been realized in which the shape control actuator is not actively operated in response to small fluctuations in rolling operation parameters and shape performance data during operation. Therefore, the absolute value error MAE was improved to 7.51 (I-unit).

Although the embodiments applying the invention made by the present inventors have been described above, the present invention is not limited to the description and drawings that form part of the disclosure of the present invention by the present embodiments. That is, all other embodiments, examples, operational techniques, etc. made by those skilled in the art based on this embodiment are included in the scope of the present invention.

1 Rolling equipment 2 Work roll 3 Intermediate roll 4 Small backup roll 5 Upper backup roll 6 Lower backup roll 7 Payoff reel 8 Right tension reel 9 Left tension reel 10 Shape detector 11 Intermediate roll bender 12 Work roll bender 100 Control computer 110 Rolling Control controller 120 Data acquisition section 140 Shape control actuator setting model generation section 141 Database section 142 Preliminary processing section 143 Machine learning section M Shape control actuator setting model S Steel plate

Claims (8)

A method for generating a shape control actuator setting model for rolling equipment for setting a shape control actuator in a rolling equipment comprising a shape control actuator that controls the shape of a steel plate and a shape detector that detects the shape of the steel plate. hand,
Input one or more operation performance data selected from the rolling operation parameters of the rolling equipment and the shape performance data obtained from the shape detector, and determine whether or not to change the settings of the shape control actuator. a setting model generation step of generating a shape control actuator setting model by a neural network method using a plurality of learning data, in which setting change information expressed and setting values of the shape control actuator are output performance data. , A method for generating a shape control actuator setting model for rolling equipment. The method for generating a shape control actuator setting model for rolling equipment according to claim 1, wherein the rolling operation parameters include target shape data of the steel plate. The neural network includes, as an output layer, a first output layer that outputs the setting change information, and a second output layer that outputs the setting value of the shape control actuator, and the first output layer and the second output layer output the setting change information. 3. The method for generating a shape control actuator setting model for rolling equipment according to claim 1, wherein the output layer has a network structure branched from a common intermediate layer. A method for setting a shape control actuator of a rolling equipment for setting a shape control actuator in a rolling equipment comprising a shape control actuator that controls the shape of a steel plate, and a shape detector that detects the shape of the steel plate, the method comprising:
One or more operation data selected from the rolling operation parameters of the rolling equipment and the shape data acquired from the shape detector are input data, and a setting change indicating whether to change the settings of the shape control actuator is performed. The setting change information changes the settings of the shape control actuator using a shape control actuator setting model generated by a neural network method using information and setting values of the shape control actuator as output data. A method for setting a shape control actuator for a rolling facility, comprising a resetting step of resetting the shape control actuator to the set value. The neural network includes, as an output layer, a first output layer that outputs the setting change information, and a second output layer that outputs the setting value of the shape control actuator, and the first output layer and the second output layer output the setting change information. The output layer is a network structure branching from a common intermediate layer,
The resetting step includes resetting the shape control actuator to the setting value when the output of the first output layer changes the setting of the shape control actuator.
A method for setting a shape control actuator for rolling equipment according to claim 4. A method for controlling the shape of a steel plate, the method comprising the step of resetting the shape control actuator for a rolling equipment during rolling of the steel plate using the method for setting the shape control actuator for a rolling equipment according to claim 4 or 5. A shape control device for a steel plate, comprising means for resetting the shape control actuator during rolling of a steel plate using the method for setting a shape control actuator for a rolling equipment according to claim 4 or 5. A method for manufacturing a steel plate, comprising the step of manufacturing a steel plate using the method for controlling the shape of a steel plate according to claim 6.

Images