Method for cold-rolling steel sheet, method for manufacturing cold-rolled steel sheet, and facility for manufacturing cold-rolled steel sheet

Abstract

A method for cold-rolling a steel plate according to the present invention is a method for cold-rolling a steel plate by controlling the control amount of a shape control actuator of a cold-rolling mill when cold-rolling the steel plate by using a prediction model that predicts the tension at the end of the steel plate in the width direction at the exit side of the cold-rolling mill, the prediction model is a model generated using rolling performance data when a steel plate is cold-rolled in the past as an explanatory variable and using an estimated value of the tension at the end of the steel plate in the width direction on the outlet side of the cold-rolling mill as a target variable, and the rolling performance data includes rolling load performance data appropriately acquired in the whole cold-rolling region. The cold rolling method comprises the following steps: continuously predicting the tension of the width direction end part of the steel plate to be rolled at the outlet side of the cold rolling mill by inputting the rolling condition of the steel plate to be rolled into the prediction model; the control amount of the shape control actuator is controlled so that the predicted tension becomes a preset target tension.

Description

Technical Field

[0001] The present invention relates to a cold rolling method for steel sheets, a manufacturing method for cold rolled steel sheets, and a manufacturing apparatus for cold rolled steel sheets. Background Art

[0002] In the cold rolling process, sometimes minute cracks, namely defects called edge cracks (alias, edge breaks), are generated at both end faces in the width direction of the steel sheet during rolling. In recent years, the demand for high-tensile steel sheets that contribute to the high strength and light weight of components and electromagnetic steel sheets essential for the electrification of automobiles has increased. However, in the rolling process of these steel sheets, in addition to the brittleness of the raw materials, the rolling load is also high, so edge cracks are likely to occur. Edge cracks are major defects that not only lead to a reduction in the quality and yield of the steel sheet, but also lead to a reduction in productivity caused by the fracture of the steel sheet and damage to the equipment. Against this background, a cold rolling method for suppressing edge cracks has been proposed. Specifically, Patent Document 1 describes the plate thickness direction strain ε corresponding to the reduction ratio at the end in the width direction of the steel sheet ye relative to the plate thickness direction strain ε corresponding to the reduction ratio at the center in the width direction of the steel sheet yc ratio ε ye / ε yc becomes equal to or greater than the edge crack limit value that increases with the progress of passes, and a method of setting the reduction ratio at the end in the width direction and the work roll shape in each pass for rolling.

[0003] On the other hand, in the cold rolling process, regardless of the type of rolling mill such as a tandem rolling mill having a plurality of rolling stands or a reversible rolling mill having a single rolling stand, the rolling speed at the front and rear ends of the coil is slower than the rolling speed at the stable part of the coil, and rolling conditions such as the rolling load change according to the rolling time (refer to Figure 7 ). Therefore, in order to suppress edge cracks throughout the cold rolling process, an operation corresponding to the change in the rolling conditions during rolling is required. It should be noted that the entire cold rolling process refers to the rolling area that includes both the stable rolling area where rolling is performed at a constant speed and the non-stable rolling area that is an acceleration / deceleration area, and extends over the entire length of the coil. Therefore, a method of setting the control amount of the shape control unit, the front tension, and the rear tension so that the shape of the steel sheet during rolling is consistent with the target shape has been proposed (refer to Patent Document 2). In this method, a mathematical model representing the elongation difference with the rolling load, the control amount of the shape control unit, the convexity of the raw material, the shape before rolling, and the convexity of the work roll as variables and a mathematical model representing the rolling load with the front tension and the rear tension as variables are used. In addition, in order to suppress edge cracks throughout the cold rolling process, the following method has been proposed: using a prediction formula representing the tension on the exit side of the plate end to adjust the control amount of the shape control actuator so that the tension on the exit side of the plate end becomes equal to or less than the fracture limit value at which the steel sheet fractures (refer to Patent Document 3).

[0004] Prior art documents

[0005] Patent documents

[0006] Patent Document 1: Japanese Patent Laid-Open No. 2015-100796

[0007] Patent Document 2: Japanese Patent Laid-Open No. 2009-22985

[0008] Patent Document 3: Japanese Patent Laid-Open No. 2017-164796 Summary of the invention

[0009] Problems to be solved by the invention

[0010] However, when the method described in Patent Document 1 is applied to a rolling mill without a work roll traverse function, the optimized tapered work roll cannot be applied to steel plates with a specific plate width other than that. In the cold rolling process, even for the same steel grade, there are generally steel plates with various plate widths. Therefore, the method described in Patent Document 1 lacks generality. In addition, when the reduction ratio at the end in the width direction of the steel plate is small, the elongation rate in the length direction of the steel plate at the end in the width direction of the steel plate is smaller than that at the center in the width direction of the steel plate. Moreover, due to the restraint from the center in the width direction of the steel plate, tensile stress is generated at the end in the width direction of the steel plate. If the elongation rate at the end in the width direction of the steel plate caused by the tensile stress exceeds the tensile fracture elongation rate, edge cracks will occur. However, it is considered that for the tensile stress generated at the end in the width direction of the steel plate, not only the elongation rate difference between the center and the end in the width direction of the steel plate has an impact, but also the magnitude of the total tension has an impact. And the rolling conditions such as the rolling load change constantly during rolling. Therefore, it can be said that the method described in Patent Document 1 is not sufficient to stably suppress the generation of edge cracks over the entire length of the coil.

[0011] On the other hand, the method described in Patent Document 2 maintains the target shape over the entire length of the coil by performing control using a mathematical model that uses rolling load, control amount of the shape control unit, convexity of the raw material, shape before rolling, and convexity of the work roll as variables, which may cause changes in the shape after rolling. However, when the purpose is to suppress the generation of edge cracks, it is appropriate to use the tension acting on the end portions in the width direction of the steel sheet (hereinafter referred to as edge tension) as a variable instead of the plate shape. In addition, in the method described in Patent Document 3, regarding the rolling load that constitutes the prediction formula of the edge tension, at least in the non-steady rolling region, a predicted value based on the correlation between the rolling speed and the friction coefficient is used. However, in the non-steady rolling region where the rolling load varies significantly, the variation of the edge tension also becomes large, and edge cracks are likely to occur. Therefore, from the viewpoint of accuracy, it is preferable to use the measured value according to the predicted value as the rolling load used in the prediction of the edge tension in the non-steady rolling region.

[0012] The present invention has been completed to solve the above problems, and an object thereof is to provide a cold rolling method for a steel sheet, a method for manufacturing a cold rolled steel sheet, and a manufacturing apparatus for a cold rolled steel sheet that can stably suppress the generation of edge cracks over the entire length of the coil regardless of the type of rolling mill and shape control actuator.

[0013] Means for Solving the Problems

[0014] The cold rolling method for a steel sheet of the present invention controls the control amount of the shape control actuator of the cold rolling mill when cold rolling the steel sheet by using a prediction model that predicts the tension at the end portions in the width direction of the steel sheet at the exit side of the cold rolling mill, and cold rolls the steel sheet. The prediction model is a model generated by using the rolling performance data when the steel sheet was cold rolled in the past as explanatory variables and the estimated value of the tension at the end portions in the width direction of the steel sheet at the exit side of the cold rolling mill as the target variable. The rolling performance data includes the performance data of the rolling load appropriately obtained in the entire cold rolling region. The cold rolling method for the steel sheet includes the following steps: continuously predicting the tension at the end portions in the width direction of the steel sheet to be rolled at the exit side of the cold rolling mill by inputting the rolling conditions of the steel sheet to be rolled into the prediction model, and controlling the control amount of the shape control actuator so that the predicted tension becomes a preset target tension.

[0015] The rolling performance data may include at least one of the change data of the hot rolling convexity in the length direction of the steel sheet and the time-dependent change data of the thermal convexity of the work roll.

[0016] The tension at the end portions in the width direction of the steel sheet at the exit side of the cold rolling mill predicted by the prediction model may be the tension at a preset position within a range of 2 to 15 mm inside from the end portions in the width direction of the steel sheet.

[0017] The method for manufacturing a cold-rolled steel sheet according to the present invention includes the step of manufacturing a cold-rolled steel sheet by using the cold rolling method of the steel sheet according to the present invention.

[0018] The manufacturing apparatus for a cold-rolled steel sheet according to the present invention includes: a cold rolling mill for cold rolling a steel sheet; and a control device for controlling the control amount of a shape control actuator of the cold rolling mill by using the cold rolling method of the steel sheet according to the present invention.

[0019] Advantageous Effects of the Invention

[0020] According to the cold rolling method of a steel sheet, the method for manufacturing a cold-rolled steel sheet, and the manufacturing apparatus for a cold-rolled steel sheet of the present invention, generation of edge cracks can be stably suppressed over the entire length of the coil regardless of the type of rolling mill and shape control actuator. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram showing the structure of a manufacturing apparatus for a cold-rolled steel sheet according to an embodiment of the present invention.

[0022] Figure 2 It is a block diagram showing the structure of an arithmetic unit.

[0023] Figure 3 It is for explaining Figure 1 the function of the prediction model shown.

[0024] Figure 4 It is a diagram showing an example of the hot rolling crown shape at the front end / tail end part and the stable part of a steel sheet.

[0025] Figure 5 It is a diagram showing the change in thermal crown accompanying the change in rolling time.

[0026] Figure 6 It is a flowchart showing the flow of a bending machine condition control process according to an embodiment of the present invention.

[0027] Figure 7 It is a diagram showing an example of the change in rolling speed and rolling load accompanying the change in rolling time. DETAILED DESCRIPTION OF THE INVENTION

[0028] Hereinafter, with reference to the drawings, a cold rolling method of a steel sheet, a method for manufacturing a cold-rolled steel sheet, and a manufacturing apparatus for a cold-rolled steel sheet according to an embodiment of the present invention will be described in detail. It should be noted that the constituent elements in the following-described embodiments include elements that can be replaced by those skilled in the art and are easily replaceable, or substantially the same elements.

[0029] 〔Structure〕

[0030] First, with reference to Figure 1The structure of the manufacturing equipment for cold-rolled steel sheets according to an embodiment of the present invention will be described.

[0031] Figure 1 It is a schematic diagram showing the structure of the manufacturing equipment for cold-rolled steel sheets according to an embodiment of the present invention. As Figure 1 shown, the manufacturing equipment for cold-rolled steel sheets according to an embodiment of the present invention (hereinafter simply referred to as the manufacturing equipment) is a continuous tandem rolling production line having a plurality of rolling stands. The manufacturing equipment includes an uncoiler 1, a joining device 2, a loop former 3, a cold tandem mill 4, a cutter 5, a tension reel 6, and an arithmetic unit 7. In the present embodiment, the arithmetic unit 7 and a work roll bending machine control device described later function as the control device of the present invention.

[0032] The uncoiler 1 is a device for feeding out the steel sheet S. It should be noted that the manufacturing equipment may also include a plurality of uncoilers 1. In this case, the plurality of uncoilers 1 feed out different steel sheets S respectively.

[0033] The joining device 2 is a device for joining the tail end portion of the steel sheet (preceding material) fed out from the uncoiler 1 first and the front end portion of the steel sheet (succeeding material) fed out from the uncoiler 1 later to form a joined steel sheet. As the joining device 2, a laser welding machine is preferably used.

[0034] The loop former 3 is a device for storing the steel sheet S during the period until the end portions of the steel sheet S are joined by the joining device 2 (until the joining is completed) so that the cold rolling of the steel sheet S by the cold tandem mill 4 can continue.

[0035] The cold tandem mill 4 is a device for cold rolling the steel sheet S so that the thickness of the steel sheet S becomes the target thickness. In the present embodiment, the cold tandem mill 4 sequentially includes five rolling stands, namely, the first rolling stand to the fifth rolling stand (#1std to #5std), from the inlet side of the steel sheet S (toward the Figure 1 right side of the paper surface of Figure 1 the paper surface) to the outlet side (toward the

[0036] left side of the paper surface of

[0037] the paper surface). In this cold tandem mill 4, tension rolls, deflector rolls, and thickness gauges (not shown) are appropriately provided between adjacent rolling stands.

[0038] The tension coiler 6 is a device for coiling the steel sheet S cut by the cutting machine 5. The form of the tension coiler 6 is not limited. For example, it can also be a disk-type tension coiler. In addition, the manufacturing equipment may also be provided with a plurality of tension coilers 6. In this case, the plurality of tension coilers 6 continuously coil a plurality of steel sheets S.

[0039] The arithmetic unit 7 uses a prediction model to predict the tension (edge tension) acting on the widthwise ends of the steel sheet S to be rolled at the exit side of the cold rolling mill 4, and controls the work roll bending machine conditions of the tandem cold rolling mill 4 when cold rolling the steel sheet S to be rolled based on the predicted edge tension. The edge tension predicted using the prediction model may be the tension at the widthwise ends of the steel sheet S, but may be the tension at a preset position within the range of 2 to 15 mm inside from the widthwise ends of the steel sheet S, or may also be within the range of 5 to 10 mm inside from the widthwise ends of the steel sheet S. Originally, it was considered that the tension affecting edge cracking is the tension acting on the widthwise ends (the outermost edges) of the steel sheet S, but sometimes the analysis result of the outermost edges in the shape analysis for estimating the tension becomes unstable. Therefore, the reliability of the tension within the range of 2 to 15 mm inside from the widthwise ends of the steel sheet S is higher. In the case of further seeking the reliability of the tension, considering the instability of the analysis result near the widthwise ends of the steel sheet S and that the predicted edge tension variation is smaller than the actual edge tension variation due to the longer distance from the widthwise ends of the steel sheet S, it is preferable to use the tension within the range of 5 to 10 mm inside from the widthwise ends of the steel sheet S. In addition, it is preferable to use the tension at a position as close as possible to the widthwise ends within the range where the data of the shape analysis is reliable as the edge tension.

[0040] Figure 2 is a block diagram showing the structure of the arithmetic unit 7. As Figure 2 shown, the arithmetic unit 7 includes an arithmetic device 71, an input device 72, a storage device 73, and an output device 74.

[0041] The arithmetic device 71 is wired-connected to the input device 72, the storage device 73, and the output device 74 via a bus wiring 75. However, the connection method of the arithmetic device 71, the input device 72, the storage device 73, and the output device 74 is not limited to wired connection, and may also be wireless connection, or may be connected in a combination of wired and wireless ways.

[0042] The input device 72 has an input port function for inputting control information of the work roll bender control device and information from the operation information device 8 and the HOT information device 9. Among the information from the operation information device 8, it includes information related to the steel plate S to be rolled (steel grade, dimensions), cold rolling condition information (numerical information, text information, and image information) set by the process computer or the operator before cold rolling, and rolling condition information (numerical information, text information, and image information) during cold rolling. The HOT information device 9 includes information on the hot rolling crown (plate crown) of the steel plate S to be rolled. The information on the plate crown is composed of measurement data at multiple positions in the length direction of the steel plate S, and it is preferably that the measurement interval is shorter.

[0043] The storage device 73 is composed of well-known storage devices such as a hard disk drive, a semiconductor drive, and an optical drive, and is a device for storing information required in this system (information required to implement the functions of the arithmetic processing unit 78 described later).

[0044] The output device 74 functions as an output port for outputting a control signal from the arithmetic device 71 to the work roll bender control device.

[0045] The arithmetic device 71 includes a RAM 76, a ROM 77, and an arithmetic processing unit 78. The RAM 76, the ROM 77, and the arithmetic processing unit 78 are connected to the input device 72, the storage device 73, and the output device 74 via the bus wiring 75.

[0046] The RAM 76 is a working main memory used when certain processing is performed by the arithmetic device 71.

[0047] The ROM 77 stores a prediction model 77A and a prediction model execution program 77B. As Figure 3 shown, when making the prediction model 77A, first, using a shape analysis model, the edge tension (target variable) is estimated based on the rolling actual performance data (explanatory variable) of the steel plate S to be rolled. As the shape analysis model, a segmentation model that divides the roll into small intervals and assumes the same rolling load distribution and roll gap contact pressure distribution in each interval can be used, or other models can also be used. Then, as much rolling actual performance data as possible is used to estimate the edge tension, and the set of the rolling actual performance data and the estimated value of the edge tension is used as learning data for machine learning, thereby generating a prediction model 77A for predicting the edge tension based on the rolling conditions of the steel plate S to be rolled. As the machine learning method, a neural network model such as deep learning, a regression tree model such as a random forest, a gradient boosting model such as XGBOOST, etc. can be used, but there is no particular limitation, and other well-known machine learning methods can also be adopted.

[0048] In the present embodiment, since a neural network is adopted as a machine learning method, the prediction model 77A becomes a neural network model. The neural network model is represented by a function, for example. Specifically, hyperparameters used in the machine learning model are set, and learning is performed based on the neural network model using these hyperparameters. As an optimization calculation of the hyperparameters, neural network models in which several of the hyperparameters are changed step by step for the learning data are generated, and the hyperparameters with the highest prediction accuracy for the verification data are selected. As the hyperparameters, the number of hidden layers, the number of neurons in each hidden layer, the dropout rate (cutting off the transmission of neurons with a certain probability) in each hidden layer, the activation function in each hidden layer, and the number of outputs are generally set, but are not limited thereto. In addition, the optimization method of the hyperparameters is not particularly limited, and grid search for changing the parameters step by step, random search for randomly selecting the parameters, or search based on Bayesian optimization can be used. Note that, in the present embodiment, the prediction model 77A is pre-stored in the ROM 77, but the method is not limited thereto. For example, the arithmetic processing unit 78 may be provided with a function for creating the prediction model 77A to create the prediction model 77A online.

[0049] Here, research was conducted on the parameters of the explanatory variables used in the prediction model 77A. Specifically, in addition to the parameters conventionally used when estimating the edge tension such as the reduction rate distribution, rolling load, tension, roll diameter, and control amount of the shape control actuator, the effects of changes in the hot rolling convexity in the longitudinal direction of the steel sheet and the temporal change in the thermal convexity of the work roll were also evaluated. First, the effect of the change in the hot rolling convexity in the longitudinal direction of the steel sheet on the change in the edge tension was evaluated. The shape of the hot rolling convexity was investigated at multiple positions in the longitudinal direction of the steel sheet S. As a result, it was confirmed that in the coils investigated, there were many coils in which the shape of the hot rolling convexity was different between the leading and trailing ends and the stable part. An example is shown in Figure 4Therefore, an example of the hot rolling crown shape of the leading and trailing ends and the stabilizing part of the coil is used as a parameter for shape analysis, and the exit side edge tension of each rolling stand of the tandem cold rolling mill 4 is estimated. As a result, according to the difference in the hot rolling crown shape, an edge tension change of about 30% at most can be confirmed. Based on the experience so far, this edge tension change cannot be ignored as an influence on the generation of edge cracks. Therefore, it is considered that by adding the change in the hot rolling crown in the length direction of the steel plate to the explanatory variable of the prediction model 77A, a prediction model more effective for suppressing the generation of edge cracks can be obtained. In the present embodiment, the information on the hot rolling crown in the steel plate S is collected from the HOT information device 9 that stores the data measured in the hot rolling production line, but the data collection method is not limited to this. For example, the data measured by the shape measuring device provided at the entrance side of the tandem cold rolling mill 4 can also be used. When using the data measured in the hot rolling production line, by considering the difference in the plate width of the measured steel plate (exit side of hot rolling finish rolling) and the plate width of the steel plate at the entrance side of the tandem cold rolling mill 4 (trimming amount), the edge tension can be estimated with higher accuracy.

[0050] Next, the influence of the change over time of the thermal crown of the work roll on the change in the edge tension is evaluated. The thermal crown refers to the phenomenon that the roll R thermally expands due to the heat generated during rolling as shown in Figure 5 For the growth of the thermal crown, the lower the temperature of the roll R, that is, the earlier after the start of rolling, the more rapidly it grows. Therefore, a method of using a work roll in which the thermal crown is saturated by performing a trial rolling in advance is sometimes adopted, but due to the breakage of the steel plate during rolling, etc., rolling may sometimes start from a state where it has never been used. The change in the edge tension during rolling at this time is estimated by shape analysis. Assuming that the conditions other than the thermal crown are constant and performing the calculation, it is confirmed that the maximum change in the edge tension during rolling is about 60%. Based on the experience so far, this edge tension change cannot be ignored as an influence on the generation of edge cracks. Therefore, it is considered that by adding the change over time of the thermal crown of the work roll to the explanatory variable of the prediction model 77A, a prediction model more effective for suppressing the generation of edge cracks can be obtained. In the present embodiment, the information on the thermal crown in the rolled steel plate S is collected from the thermal crown information device 10 that stores the thermal crown prediction model based on the measured data, but the prediction method of the thermal crown is not limited to this, and numerical analysis or a prediction model made based on the measured data and the numerical analysis results can also be used. In addition, when predicting the thermal crown, the thermal crown can be predicted with higher accuracy by considering the roughness of the work roll.

[0051] Return Figure 2。The arithmetic processing unit 78 has an arithmetic processing function and is wired to the RAM 76 and the ROM 77 via the bus wiring 75. The arithmetic processing unit 78 determines the rolling conditions (work roll bending machine conditions) of the steel plate S to be rolled. To perform the above processing, when the arithmetic processing unit 78 receives a signal notifying that cold rolling is being performed from the work roll bending machine control device via the input device 72, it functions as an information reading unit 78A, a data preprocessing unit 78B, a rolling state prediction unit 78C, a bending machine condition determination unit 78D, and a result output unit 78E by executing the prediction model execution program 77B stored in the ROM 77.

[0052] The information reading unit 78A reads in the control information of the work roll bending machine control device obtained from the input device 72 and the information from the operation information device 8.

[0053] The data preprocessing unit 78B performs generation processing on the data input to the rolling state prediction unit 78C. Specifically, the data preprocessing unit 78B performs processing such as unit conversion, deletion of coil data containing abnormal data or unacquired data, or data filling in order to enable the prediction model 77A to read in the rolling performance data and the rolling conditions of the steel plate S to be rolled.

[0054] The rolling state prediction unit 78C inputs the input data generated by the data preprocessing unit 78B into the prediction model 77A to predict the thickness distribution and tension distribution of the steel plate S.

[0055] The bending machine condition determination unit 78D performs the following processing: changing the work roll bending machine conditions in the rolling conditions of the steel plate S to be rolled, and repeatedly returning to the execution of the processing of the above information reading unit 78A, data preprocessing unit 78B, and rolling state prediction unit 78C until it becomes below the edge crack generation threshold that varies according to the steel type and dimensions of the steel plate S set in advance. That is, the bending machine condition determination unit 78D continuously predicts the edge tension of the steel plate S to be rolled based on the rolling conditions of the steel plate S to be rolled using the prediction model 77A, and controls each unit in such a way as to determine the work roll bending machine conditions such that the predicted edge tension is below the edge crack generation threshold. The edge crack generation threshold is a value that can be quantitatively represented using the edge tension, which is one of the target variables of the prediction model 77A, and is determined based on the experimental results in the tandem cold rolling mill 4.

[0056] The result output unit 78E operates when the edge tension of the steel plate S to be rolled predicted by the prediction model 77A is below the edge crack generation threshold, and outputs the determined work roll bending machine conditions to the work roll bending machine control device.

[0057] 〔Bending Machine Condition Control Processing〕

[0058] Next, refer toFigure 6 , the process of the bending machine condition control process as an embodiment of the present invention will be described.

[0059] Figure 6 It is a flowchart showing the process of the bending machine condition control process as an embodiment of the present invention. Figure 6 The flowchart shown starts the bending machine condition control process to enter the process of step S1 at the timing when a signal notifying that cold rolling is being performed is input from the work roll bending machine control device via the input device 72.

[0060] In the process of step S1, the information reading unit 78A reads the prediction model 77A stored in the ROM 77. Thus, the process of step S1 is completed, and the bending machine condition control process enters the process of step S2.

[0061] In the process of step S2, the information reading unit 78A reads the data of the edge crack generation threshold of the steel plate S to be rolled stored in the storage device 73. Thus, the process of step S2 is completed, and the bending machine condition control process enters the process of step S3.

[0062] In the process of step S3, the information reading unit 78A reads the current rolling conditions of the steel plate S to be rolled via the input device 72. Thus, the process of step S3 is completed, and the bending machine condition control process enters the process of step S4.

[0063] In the process of step S4, the rolling state prediction unit 78C predicts the edge tension of the steel plate S to be rolled by inputting the rolling conditions read in the process of step S3 into the prediction model 77A read in the process of step S1. Thus, the process of step S4 is completed, and the bending machine condition control process enters the process of step S5.

[0064] In the process of step S5, the bending machine condition determination unit 78D determines whether the edge tension predicted in the process of step S4 is below the edge crack generation threshold read in the process of step S2. If the result of the determination is that the edge tension is below the edge crack generation threshold (step S5: yes), the bending machine condition determination unit 78D ends a series of bending machine condition control processes. On the other hand, if the edge tension is greater than the edge crack generation threshold (step S5: no), the bending machine condition determination unit 78D causes the bending machine condition control process to enter the process of step S6.

[0065] In the process of step S6, the bending machine condition determination unit 78D changes the work roll bending machine condition in the rolling conditions read in the process of step S3. Specifically, the bending machine condition determination unit 78D calculates an appropriate work roll bending machine value based on the difference between the edge tension predicted in the process of step S4 and the edge crack generation threshold read in the process of step S2. Thus, the process of step S6 is completed, and the bending machine condition determination unit 78D returns the bending machine condition control process to the process of step S3.

[0066] If the process returns to step S3, the rolling state prediction unit 78C reads the rolling conditions in which the work roll bending machine condition has been changed. If the rolling conditions are read, in the process of step S4, the rolling state prediction unit 78C predicts the edge tension by inputting the rolling conditions in which only the work roll bending machine value has been changed to the prediction model 77A. If the edge tension is predicted, in the process of step S5, the bending machine condition determination unit 78D determines whether the predicted edge tension is below the edge crack generation threshold. Then, the series of processes of steps S3 to S6 are repeatedly executed until the predicted edge tension becomes below the edge crack generation threshold. In this way, the arithmetic processing unit 78 continuously predicts the edge tension of the steel plate S to be rolled according to the rolling conditions of the steel plate S to be rolled using the prediction model 77A, and determines the work roll bending machine value in such a manner that the predicted edge tension becomes below the edge crack generation threshold.

[0067] From the above description, it can be seen that in the present embodiment, first, a prediction model 77A is created with the rolling actual result data during past cold rolling of the steel plate S as the explanatory variable and the edge tension as the target variable. Then, the arithmetic processing unit 78 continuously predicts the edge tension of the steel plate S to be rolled according to the rolling conditions of the steel plate S to be rolled using the prediction model 77A. Then, the arithmetic processing unit 78 determines the work roll bending machine condition in such a manner that the predicted edge tension becomes below the edge crack generation threshold set in advance based on the experimental results. Thereby, rolling control that satisfies various restrictions in the rolling operation without relying on the experience and subjectivity of the operator is implemented, and the generation of edge cracks during rolling can be stably suppressed.

[0068] 〔Modification Example〕

[0069] As described above, the embodiments of the present invention have been explained, but the present invention is not limited thereto, and various changes and improvements can be made. For example, the devices provided in the manufacturing equipment are not limited to the above-mentioned devices. Therefore, the rolling mill may not be a tandem rolling mill but a reversing rolling mill. In addition, it is also possible to make the cold rolling process and the pickling process as its previous process continuous, and a pickling device for pickling the steel sheet S may be arranged between the loopers 3 and the tandem cold rolling mill 4. In addition, in the present embodiment, the shape control actuator is targeted at a rolling mill with only a work roll bender, so the edge tension control based on the work roll bender control is implemented. However, in the case of a rolling mill equipped with other shape control actuators, they can also be used to perform the edge tension control. In addition, the tandem cold rolling mill 4 is not limited to 4Hi, and it may also be a multi-high rolling mill such as 6Hi, and the number of rolling stands is not particularly limited. In addition, it may also be a multi-roll rolling mill or Sendzimir rolling mill.

[0070] In addition, when the control amount calculated by the arithmetic unit 7 for the work roll bender change amount exceeds the upper and lower limits of the equipment specifications or when the control amount cannot be calculated, a warning screen can be displayed on the monitor and an alarm can be issued, and this implementation is not performed. In addition, in the present embodiment, the work roll bender control is automatically implemented, but an appropriate work roll bender change amount output from the arithmetic device 71 can also be displayed on a monitor or the like, and the change of the work roll bender is manually implemented by the operator. In addition, the present invention can also be used only for setting, and in this case, the work roll bender value during rolling is constant. The data used as the explanatory variable can be the cold rolling condition information set by the process computer or the operator before cold rolling. For the hot rolling crown information, the data of the stable part (near the central part in the length direction) of the hot rolled coil can be used, and for the hot crown information, the value in the state of hot crown saturation can be used.

[0071] [Examples]

[0072] In this example, use Figure 1The cold rolling mill 4 shown in the figure performs a cold rolling experiment on steel type A (electromagnetic steel sheet) with a base material thickness of 2.2 mm, a finished thickness of 0.35 mm, and a plate width of 1300 mm. The prediction model used and the edge crack generation status are shown in Tables 1-1 and 1-2. In this embodiment, as a prior learning, learning based on a machine learning model (deep learning) is implemented using learning data (rolling performance data of about 3,000 past steel plates) to generate a learning model that predicts edge tension according to the rolling conditions of the steel plate to be rolled. This learning model is used in the invention example and comparative example shown below. In addition, the edge crack generation threshold of steel type A was studied based on the rolling performance data of steel type A rolled in the past. As a result, it is estimated that if the edge tension in the rolling stand after #4std is less than 180 MPa, the edge crack can be suppressed. Therefore, the edge crack generation threshold is 180 MPa (#4std, #5std output side).

[0073] In Example 1, as explanatory variables, i.e., rolling performance data, the deformation resistance, plate thickness, plate width, tension, load, work roll diameter, work roll shape, work roll bender value of steel type A, and the shape change coefficient (extension of the steel plate in the width direction) previously incorporated by experiments were used. The work rolls were new, and the surface temperature of the work rolls immediately after the start of rolling was the same as the room temperature. As a result of rolling 20 coils, tiny edge cracks (~1 mm) were confirmed at the leading and trailing ends of 3 coils, but no edge cracks occurred in the remaining 17 coils.

[0074] In Inventive Example 2, as the explanatory variable, i.e., the rolling performance data, in addition to the rolling performance data used in Inventive Example 1, the change in the hot rolling crown in the longitudinal direction of the steel plate was used. The work rolls were new, and the work roll surface temperature immediately after the start of rolling was the same as the room temperature. As a result of rolling 20 coils, a small edge crack (~1 mm) was confirmed in one coil immediately after the start of rolling, but no edge crack occurred in the remaining 19 coils.

[0075] In Inventive Example 3, in addition to the rolling performance data used in Inventive Example 1, the change in the hot rolling crown in the longitudinal direction of the steel plate was used as the explanatory variable, i.e., the rolling performance data. The work rolls used were work rolls saturated with hot crowns by trial rolling. That is, the conditions of the work rolls used in Inventive Example 2 and Inventive Example 3 were different. As a result of rolling 20 coils, edge cracks were suppressed over the entire length of all the coils.

[0076] Based on the experimental results of Invention Example 2 and Invention Example 3, it was confirmed that the growth of thermal crown during rolling could be the cause of edge crack generation. Therefore, it was confirmed that when it was possible to use a new work roll, using the time-dependent change of thermal crown as an explanatory variable of the learning model was effective for suppressing edge crack generation.

[0077] In Invention Example 4, as the explanatory variable, i.e., rolling performance data, in addition to the rolling performance data used in Invention Example 1, the time-dependent change of the thermal crown of the work roll was also used. A new work roll was used, and the surface temperature of the work roll just after the start of rolling was the same as the room temperature. In addition, as the explanatory variable, the change in the hot rolling crown in the length direction of the steel sheet was not used. The results of rolling 20 coils showed that minor edge cracks (~1 mm) were confirmed in 2 coils. Edge cracks were confirmed only at the front and tail ends of 1 coil, and edge cracks were confirmed over the entire length of 1 coil. In the post-rolling investigation, the 2 coils with edge cracks had significantly different hot rolling crown shapes at the front and tail ends and significantly different hot rolling crown shapes over the entire length of the coil compared to the coils without edge cracks.

[0078] In Invention Example 3, it was possible to suppress edge cracks in all coils rolled using the change in the hot rolling crown in the length direction of the steel sheet as an explanatory variable. Therefore, in order to cope with the deviation of the hot rolling crown shape and stably suppress edge cracks, it was confirmed that it was preferable to use the change in the hot rolling crown in the length direction of the steel sheet as an explanatory variable.

[0079] In Invention Example 5, as the explanatory variable, i.e., rolling performance data, in addition to the rolling performance data used in Invention Example 1, the change in the hot rolling crown in the length direction of the steel sheet and the time-dependent change of the thermal crown of the work roll were also used. A new work roll was used, and the surface temperature of the work roll just after the start of rolling was the same as the room temperature. The results of rolling 20 coils showed that it was possible to suppress edge cracks over the entire length of all coils.

[0080] In Invention Examples 6 to 12, only the definition of the edge tension in the rolling performance data of Invention Example 5 was changed. Specifically, in Invention Example 6, the tension at the end of the steel plate in the width direction (edge 0 mm) was defined as the edge tension. It should be noted that in Invention Examples 1 to 5, the tension at a position 10 mm inside the end of the steel plate in the width direction was defined as the edge tension. As a result of rolling 20 coils, edge cracks of about 3 mm at most occurred in 3 coils. In Invention Example 5, the amount of change in the work roll bending machine during rolling was 13 ton / chock (ton / bearing block), while in Invention Example 6, it was 15 ton / chock. The reason is considered to be that in the shape analysis model used to create the prediction model, the end of the steel plate in the width direction is an unstable area when predicting the tension, and a larger tension change than the actual one is predicted.

[0081] In Invention Example 7, the tension at a position 2 mm inside the end of the steel plate in the width direction was defined as the edge tension. As a result of rolling 20 coils, a slight edge crack (~1 mm) was confirmed in 1 coil.

[0082] In Invention Example 8, the tension at a position 5 mm inside the end of the steel plate in the width direction was defined as the edge tension. As a result of rolling 20 coils, edge crack suppression was possible throughout the entire length of all coils.

[0083] In Invention Example 9, the tension at a position 7 mm inside the end of the steel plate in the width direction was defined as the edge tension. As a result of rolling 20 coils, edge crack suppression was possible throughout the entire length of all coils.

[0084] In Invention Example 10, the tension at a position 13 mm inside the end of the steel plate in the width direction was defined as the edge tension. As a result of rolling 20 coils, a slight edge crack (~1 mm) was confirmed in 1 coil.

[0085] In Invention Example 11, the tension at a position 15 mm inside the end of the steel plate in the width direction was defined as the edge tension. As a result of rolling 20 coils, a slight edge crack (~1 mm) was confirmed in 1 coil.

[0086] In Invention Example 12, the tension at a position 17 mm inside the end of the steel plate in the width direction was defined as the edge tension. As a result of rolling 20 coils, an edge crack of up to 2 mm was confirmed in 1 coil. There was 1 coil with an edge crack, but the maximum depth of the edge crack was 2 mm, so it could be judged that the edge crack suppression effect was lower compared to Invention Examples 10 and 11.

[0087] Based on the experimental results of Invention Examples 5 to 12, it was confirmed that if the position of the tension used as the edge tension is too close to the width direction end, in the shape analysis model used this time, the predicted value of the tension becomes an unstable value and appropriate bending machine control cannot be performed. On the other hand, it was confirmed that the farther the position of the tension used as the edge tension is from the width direction end, the smaller the variation is than the actual edge tension variation, and appropriate bending machine control cannot be performed. Under the conditions of this time, it was confirmed that the position of the tension used as the edge tension is preferably at a position 2 to 15 mm (more preferably 5 to 10 mm) inward from the width direction end of the steel plate.

[0088] In Comparative Example 1, the edge tension control of the present invention was used only in the stable rolling region and not in the unstable rolling region (acceleration and deceleration parts). As a result of rolling 20 coils, edge cracks up to 2 mm were confirmed in 13 coils. Generally, in the rolling of cold-rolled steel sheets, rolling is performed at a low speed at the beginning of rolling, then accelerated and performed at a high speed. Then, if approaching the coil end part, it decelerates and is performed at a low speed again. There is a correlation between the rolling speed and the rolling load, and the rolling load varies greatly in the acceleration and deceleration parts. If the rolling load varies, the tension distribution in the width direction of the steel plate also changes. After rolling, the longitudinal position of the edge crack of the coil with the edge crack was investigated, and the result was consistent with the unstable rolling region. Therefore, in the case where the edge tension control is not used in the unstable rolling region, it is difficult to maintain an appropriate edge tension in the unstable rolling region, and it can be speculated that edge cracks are generated.

[0089] In Comparative Example 2, the edge tension control of the present invention was used only in the unstable rolling region and not in the stable rolling region. As a result of rolling 20 coils, edge cracks up to 2 mm were confirmed in 14 coils. For the work roll bending machine value at the start of rolling, an appropriate bending machine value was predicted and determined based on the past rolling performance of Steel Grade A. The longitudinal position of the edge crack was confirmed from the rolled coil, and there were cases where it occurred only at the front end of the coil and cases where it also occurred in the stable part. It can be considered that the reason for the former is that the predicted bending machine value based on the past rolling performance is inappropriate, and the reason for the latter is that it cannot cope with the variation of the rolling load in the longitudinal direction of the coil.

[0090] It was confirmed from Comparative Example 1 and Comparative Example 2 that by applying the edge tension control of the present invention to the entire cold rolling region (the entire length of the coil), edge cracks can be stably suppressed over the entire length of the coil.

[0091] In Comparative Example 3, in the work roll bender control, a machine learning model was not used, and the shape control described in Patent Document 2 was used. As a result of rolling 20 coils, edge cracks of up to 2 mm occurred in 10 coils. It is speculated that this is because the shape control could not cope with the changes in the plate thickness and the hot rolling crown shape of the steel plate during rolling, resulting in edge cracks.

[0092] In Comparative Example 4, in the work roll bender control, a machine learning model was not used, and the shape control described in Patent Document 3 was used. The actual value was used for the rolling load in the stable rolling region, and the value calculated by the prediction formula described in Patent Document 3 was used for the rolling load in the unstable rolling region. As a result of rolling 20 coils, edge cracks of up to 2 mm occurred in 7 coils. Investigating the longitudinal position of the edge cracks from the rolled coils, most of them occurred in the unstable rolling region. It is speculated that this is because the rolling load in the unstable rolling region was not the actual value but the predicted value, so the edge tension could not be accurately predicted, resulting in edge cracks.

[0093] In Comparative Example 5, in the work roll bender control, a machine learning model was not used, and the setting value of the bender was determined by visually checking the shape of the steel plate by the operator. The value of the work roll bender during rolling was constant. As a result of rolling 20 coils, edge cracks of up to about 3 mm occurred in 18 coils.

[0094] In Comparative Example 6, in the work roll bender control, a machine learning model was not used, and the setting value of the work roll bender was determined based on the numerical calculation results. The value of the work roll bender during rolling was constant. As a result of rolling 20 coils, edge cracks of up to 4 mm occurred in all the coils. It is speculated that this is because, under various assumptions, the results were calculated before rolling, so there were also points different from the actual rolling conditions, and the generation of edge cracks could not be suppressed.

[0095] In Comparative Example 7, in the work roll bender control, a machine learning model was not used, and the appropriate setting value of the work roll bender was predicted and determined based on the past rolling performance of Steel Grade A. The value of the work roll bender during rolling was constant. As a result of rolling 20 coils, edge cracks of up to about 2 mm occurred in 15 coils. It is speculated that this is because, similar to Comparative Example 6, there were also points different from the actual rolling conditions, and the generation of edge cracks could not be suppressed.

[0096] In Comparative Example 8, in addition to the work roll bending machine setting method of Comparative Example 7, load interlock control was used during rolling. As a preliminary study, the amount of work roll bending machine correction that could offset the shape change caused by the load variation during rolling of Steel Grade A was estimated through a shape analysis model and used for load interlock control. As a result of rolling 20 coils, edge cracks up to 2 mm maximum occurred in 12 coils. It is considered that this is because the set value of the work roll bending machine is a predicted value based on past rolling performance. Therefore, when the set value of the work roll bending machine is not appropriate for the steel plate to be rolled, edge cracks occur. In addition, since load interlock control is performed based on this value, if the set value is incorrect, load interlock control cannot be used properly. Moreover, the bending machine value for correcting the shape change caused by the rolling load change, which constitutes the load interlock control, is the estimated result under various assumptions. Therefore, it is considered that the accuracy is poor compared with the prediction model of the present invention.

[0097] Here, the edge crack suppression effect of the prediction model of the present invention was verified for coils with a very large amount of change in the length of the hot rolling crown that occurs with a probability of about 1%. The coils described in Tables 1-1 and 1-2 do not include such rarely occurring coils with a specific shape. Table 1-3 shows the prediction models used and the occurrence status of edge cracks. Invention Examples 3-A, 3-B, 5-A, and 5-B shown below used the prediction models used in the rolling experiments described in Tables 1-1 and 1-2.

[0098] In Invention Example 3-A, as explanatory variables, i.e., rolling performance data, in addition to the rolling performance data used in Invention Example 1, the change in the hot rolling crown in the length direction of the steel plate (change in the length of the hot rolling crown) was also used. Here, as the data of the change in the length of the hot rolling crown, the value predicted for the plate crown shape on the inlet side of the cold tandem mill was used, assuming that the difference between the average value of the side plate width hot rolled out and the process width on the inlet side of the cold tandem mill was the trimming amount. It is the same model as the prediction model used in Invention Example 3 shown in Table 1-1. In addition, as the work roll (WR), a work roll that saturated the hot crown through trial rolling was used. As a result of rolling 100 coils, a minute edge crack (~1 mm) was confirmed in 1 coil. If the hot rolling crown shapes of the 100 coils rolled were investigated, only the 1 coil in which the edge crack occurred was a coil with a very large amount of change in the length of the rarely occurring hot rolling crown.

[0099] In Invention Example 3-B, as the explanatory variables, i.e., rolling performance data, in addition to the rolling performance data used in Invention Example 1, the length change of the hot rolling crown was also used. Here, for the trimming performed between the hot rolling finishing exit side and the cold tandem mill entry side, data on the length change of the hot rolling crown was used, and values that could accurately grasp the trimming amount in the length direction of the coil and highly accurately estimate the shape of the coil at the cold tandem mill entry side were obtained. In Invention Example 3-A, the average value of the strip width at the hot rolling finishing exit side was used. However, the strip width at the hot rolling finishing exit side varies by several tens of millimeters in length, and using the average value of the strip width at the hot rolling finishing exit side is insufficient for highly accurately predicting the shape of the strip crown at the cold tandem mill entry side. In this embodiment, based on the strip width measured at the hot rolling finishing exit side, the strip width measured at the cold tandem mill entry side, and the data on the cutting lengths of the front and tail ends of the coil between the hot rolling finishing exit side and the cold tandem mill entry side, the shape of the strip crown at the cold tandem mill entry side was predicted. However, the method for predicting the shape of the strip crown at the cold tandem mill entry side is not limited to this. For example, the shape of the strip crown can also be actually measured at the cold tandem mill entry side. Additionally, work rolls that had saturated their thermal crowns through trial rolling were used. As a result of rolling 100 coils, edge cracks could be suppressed in all the coils. If the hot rolling crown shapes of the 100 coils that were rolled were investigated, for one coil, it was a coil with a very large length variation in the hot rolling crown that rarely occurred. However, by highly accurately predicting the shape of the strip crown at the cold tandem mill entry side, it was also possible to respond to this situation and suppress edge cracks.

[0100] In Invention Example 5-A, as the explanatory variables, i.e., rolling performance data, in addition to the rolling performance data used in Invention Example 1, the length change of the hot rolling crown and the temporal change of the thermal crown of the work roll were also used. Additionally, the method of Invention Example 3-B was used to predict the shape of the strip crown at the cold tandem mill entry side. New work rolls were used, and the surface temperature of the work roll immediately after rolling started was the same as the room temperature. The prediction of the hot pinch was the same as that used in Invention Example 5. As a result of rolling 100 coils, minor edge cracks (~1 mm) were confirmed in one coil. If the hot rolling crown shapes of the 100 coils that were rolled were investigated, the only coil with edge cracks was a coil with a very large length variation in the hot rolling crown that rarely occurred.

[0101] In Invention Example 5-B, as explanatory variables, i.e., rolling performance data, in addition to the rolling performance data used in Invention Example 1, the length change of the hot rolling crown and the temporal change of the thermal crown of the work roll were also used. Additionally, the method of Invention Example 3-B was used to predict the plate crown shape on the entry side of the tandem cold rolling mill. New work rolls were used, and the surface temperature of the work roll just after the start of rolling was the same as the room temperature. In the prediction of the thermal crown, a model that can consider the roughness of the work roll to predict the thermal crown was used. The model used in Invention Example 5-A did not consider the roughness of the work roll when predicting the thermal crown. The results of rolling 100 coils were that edge cracks could be suppressed in all the coils. If the hot rolling crown shapes of the 100 coils rolled were investigated, for one coil, it was a coil with a very large length change amount of the hot rolling crown that rarely occurred. However, by accurately predicting the thermal crown, the edge tension could be accurately predicted, and appropriate work roll bender control could be implemented. As a result, edge cracks could be suppressed.

[0102] From the above, it was confirmed that by accurately predicting the plate crown shape and thermal crown on the entry side of the tandem cold rolling mill using the above method, edge cracks could be suppressed even for coils with a very large length change amount of the hot rolling crown that rarely occurred.

[0103] From the above, it was confirmed that the manufacturing method and manufacturing equipment of the cold-rolled steel sheet of the present invention were useful for suppressing edge cracks by appropriately predicting the edge tension of the steel sheet during rolling and controlling the work roll bender so as to be below the edge crack generation threshold. Additionally, it was confirmed that by applying the present invention, not only could the reduction in productivity and damage to equipment caused by the fracture of the steel sheet during rolling be prevented, but also the improvement of quality and yield could be greatly contributed to. Moreover, the part where the edge crack occurred was cut off on another production line before the next process, and by suppressing the edge crack, the reduction of the energy used in the manufacturing process could also be contributed to.

[0104] As described above, embodiments applying the invention completed by the present inventor have been described, but the present invention is not limited to the description and drawings that are part of the disclosure of the present invention constituting this embodiment. That is, all other embodiments, examples, and application techniques and the like completed by those skilled in the art based on this embodiment are included in the scope of the present invention.

[0105] [Table 1-1]

[0106]

[0107] [Table 1-2]

[0108]

[0109] [Table 1-3]

[0110]

[0111] [Industrial Applicability]

[0112] According to the present invention, a cold rolling method for steel sheets, a manufacturing method for cold rolled steel sheets, and a manufacturing apparatus for cold rolled steel sheets can be provided, which can stably suppress the generation of edge cracks throughout the entire length of the coil regardless of the type of rolling mill and shape control actuator.

[0113] Description of Reference Numerals

[0114] 1 Uncoiler

[0115] 2 Joining device

[0116] 3 Looper

[0117] 4 Cold tandem mill

[0118] 5 Cutting machine

[0119] 6 Tension reel

[0120] 7 Arithmetic unit

[0121] 8 Operation information device

[0122] 9 HOT information device

[0123] 10 Thermal crown information device

[0124] 71 Arithmetic device

[0125] 72 Input device

[0126] 73 Storage device

[0127] 74 Output device

[0128] 75 Bus wiring

[0129] 76 RAM

[0130] 77 ROM

[0131] 77A Prediction model

[0132] 77B Prediction model execution program

[0133] 78 Arithmetic processing unit

[0134] 78A Information reading unit

[0135] 78B Data preprocessing unit

[0136] 78C Rolling state prediction unit

[0137] 78D Bending machine condition determination unit

[0138] 78E result output section.

Claims

1. A cold rolling method for steel plates, which controls the control amount of the shape control actuator of a cold rolling mill when cold rolling a steel plate by using a prediction model for predicting the tension at the end in the width direction of the steel plate at the outlet side of the cold rolling mill, thereby cold rolling the steel plate, wherein, the prediction model is a model generated by taking the rolling performance data when cold rolling a steel plate in the past as an explanatory variable and taking the estimated value of the tension at the end in the width direction of the steel plate at the outlet side of the cold rolling mill as an objective variable, the rolling performance data includes the performance data of the rolling load appropriately obtained in the entire cold rolling region, the cold rolling method for the steel plate includes the following steps: by inputting the rolling conditions of the steel plate to be rolled into the prediction model, continuously predicting the tension at the end in the width direction of the steel plate to be rolled at the outlet side of the cold rolling mill, and controlling the control amount of the shape control actuator in such a way that the predicted tension becomes a preset target tension.

2. The cold rolling method for steel plates according to claim 1, wherein, the rolling performance data includes at least one of the change data of the hot rolling crown in the length direction of the steel plate and the time-dependent change data of the thermal crown of the work roll.

3. The cold rolling method for steel plates according to claim 1 or 2, wherein, the tension at the end in the width direction of the steel plate at the outlet side of the cold rolling mill predicted by the prediction model is the tension at a preset position within the range of 2 to 15 mm inside from the end in the width direction of the steel plate.

4. A method for manufacturing cold rolled steel plates, which includes the step of manufacturing cold rolled steel plates by using the cold rolling method for steel plates according to claim 1.

5. A manufacturing apparatus for cold rolled steel plates, comprising: a cold rolling mill for cold rolling a steel plate; and a control device that controls the control amount of the shape control actuator of the cold rolling mill by using the cold rolling method for steel plates according to claim 1.

Images