JP2016074008A - Control device and control method for tandem rolling mill - Google Patents

Abstract

PROBLEM TO BE SOLVED: To materialize a balance maintenance of a rolling load among a plurality of rolling stands with a simple processing.SOLUTION: A load estimation error calculation part 14 estimates a rolling load on a rolling stand 61 by using a rolling actual value obtained when a steel plate 63 is rolled and calculates a load estimation error using the estimated rolling load and the rolling load actual value. A load balance maintenance value calculation part 15 calculates a load balance maintenance value which expresses the degree of a difference between a load estimation error on the rolling stand 61 and a load estimation error on a rolling stand 61 adjacent to the rolling stand. A load correction value calculation part 16 calculates a load correction value using the load estimation error and the load balance maintenance value. Further, a control command set-up part 11 estimates a rolling load for a steel plate 63 to be subsequently rolled, at the same time, corrects the estimated rolling load with the load correction value calculated by the load correction value calculation part 16 and calculates a screw-down position set on each rolling stand 61 by using the corrected rolling load.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a control device and a control method for a tandem rolling mill suitable for improving the quality of a steel sheet.

  Conventionally, in hot tandem rolling control of a steel sheet, the rolling state of the rolled steel sheet is predicted prior to rolling, the rolling position (corresponding to the gap between the upper and lower work rolls) and the roll speed are determined, and the steel sheet tip is controlled, Thereafter, the method of gradually correcting the reduction position and roll speed to appropriate values using the plate thickness obtained from the detector and the steel plate tension between the rolling stands is generally used. In this control method, in order to accurately control the thickness of the tip of the steel plate to the target value and stabilize rolling when the tip is caught in each rolling stand of the finishing mill, the rolling position of each rolling stand and It is necessary to determine the roll speed command value to an appropriate value by predictive calculation. In that case, in order to obtain the target plate thickness on the exit side of the rolling stand in the final stage, the reduction position is particularly important. In order to appropriately determine the reduction position, it is necessary to improve the prediction accuracy of the rolling load.

The rolling control method for improving the estimation accuracy of the rolling load is disclosed in, for example, Patent Documents 1 to 3.
In Patent Document 1, a learning coefficient determined based on a rolling load actual value Pact and a rolling load model calculated value Pcal obtained by substituting the rolling actual value into the rolling load model P is used as an error inherent to the rolling material. An example of a rolling control method is disclosed in which the first learning coefficient Zpk of the component to be learned and the second learning coefficient Zpm of the component to learn the error due to the change of the rolling mill with time are separately learned, and both these components are individually learned. Has been.

  Further, in Patent Document 2, in addition to the process of learning the estimation error of the deformation resistance of the material to be rolled based on the rolling results, the friction coefficient is calculated based on the rolling results regarding the friction phenomenon between the rolling roll and the material to be rolled. An example of a rolling load estimation model in which the prediction accuracy of the deformation resistance and the friction coefficient is improved and the prediction accuracy of the rolling load is improved by executing a process of learning the estimation error is disclosed.

  The above two examples are characterized by introducing the concept of learning into the estimation of the rolling load of each rolling stand. Patent Document 3 focuses on the relationship between the rolling loads among a plurality of rolling stands. A method for improving the prediction accuracy of the rolling load is disclosed. That is, in Patent Document 3, the rolling load error value in each rolling stand is calculated based on the rolling load actual value in each rolling stand and the rolling load model calculated value obtained by substituting the rolling condition actual value. An example of a rolling control method is disclosed in which a change in error from a preceding rolling stand to a subsequent rolling stand is modeled and a rolling load prediction error variation between the rolling stands is suppressed using the model.

Japanese Patent Laid-Open No. 10-263640 JP 2013-226596 A JP 2009-113101 A

However, the rolling control methods disclosed in Patent Documents 1 to 3 have the following problems.
For example, in the rolling load estimation model disclosed in Patent Document 1, a rolling load estimation error (a difference between the rolling load actual value Pact and the rolling load model calculation value Pcal) is converted into an error component (first learning) inherent to the material to be rolled. The error is estimated separately using the coefficient Zpk) and the error component based on the change over time of the rolling mill (the error predicted using the second learning coefficient Zpm). Separation is difficult.

  According to Patent Document 1, it is possible to separate error factors because of the small inclination of the rolling mill with time. However, for example, when the change in the thickness or width of the material to be rolled that is continuously rolled is small, the estimation error of the rolling load does not change greatly. For this reason, the error inherent to the material to be rolled may be erroneously separated as an error due to aging of the rolling mill. Therefore, the first learning coefficient Zpk and the second learning coefficient Zpm in the rolling load prediction model are improperly learned, resulting in a problem that the estimation accuracy of the rolling load is lowered.

  In the method disclosed in Patent Document 2, for example, it is actually difficult to distinguish whether the increase in rolling load is due to an increase in deformation resistance or an increase in friction. . Therefore, similarly to the case of Patent Document 1, the deformation resistance and the friction coefficient used in the rolling load prediction model are improperly learned, resulting in a problem that the estimation accuracy of the rolling load is lowered.

  Moreover, the feature of the rolling load prediction model disclosed in Patent Documents 1 and 2 is that learning used for the prediction is independently performed at each rolling stand based on the respective load results obtained at each rolling stand. There is. For this reason, the learning value at a specific rolling stand increases, and the learning coefficient at an adjacent rolling stand decreases due to the reaction, and the learning value may be greatly different between the two. As a result, the rolling load at each rolling stand is corrected to be large or small, so that a problem arises that the load balance between adjacent rolling stands is lost. When the load balance between adjacent rolling stands is lost, there is a problem that unbalanced tension or looseness occurs in the tension between the rolling stands and the rolling becomes unstable.

  In the technique disclosed in Patent Document 3, a rolling load error change model is constructed based on the rolling load error of each rolling stand, and the learning value of each rolling stand is calculated according to the model. As a result, the learning value of a specific stand is not significantly different from that of other stands, so that the load balance among the rolling stands can be maintained. On the other hand, since processing for constructing a rolling load error change model is required, there is a problem that the amount of calculation increases.

  In general, in a tandem rolling mill, the characteristics of the rolling load prediction error may change in a complicated manner as the process proceeds from the upstream rolling stand to the downstream rolling stand. For example, when the predicted load and the actual load are compared, the actual load may be larger in the upstream and downstream rolling stands, and the predicted load may be larger in the intermediate rolling stand. That is, the prediction error may be biased toward the plus side upstream and downstream of the rolling stand, for example, and toward the minus side at the intermediate portion.

  On the other hand, since the rolling load error change model disclosed in Patent Document 3 is a linear approximation, it cannot cope with such nonlinear rolling load prediction error characteristics. Therefore, when a model that can deal with nonlinearity is introduced, the model becomes complicated and the amount of calculation further increases. In addition, the rolling load prediction results vary depending on the complexity of the model.

  The present invention has been made in order to solve the above-described problems of the prior art, and the object thereof is tandem rolling capable of maintaining the balance of rolling load among a plurality of rolling stands by a simple process. It is providing the control apparatus and control method of a mill.

  In order to achieve the object of the present invention, a control device for a tandem rolling mill according to the present invention is a control device for a tandem rolling mill for continuously rolling a steel plate with a plurality of rolling stands, and the steel plate is rolled. The rolling load value obtained at each rolling stand of the plurality of rolling stands and estimating the rolling load at each rolling stand, and the estimated rolling load at each rolling stand, and the rolling Based on the actual rolling load value at each rolling stand obtained in (1), a load prediction error calculation unit for calculating a load prediction error at each rolling stand, and each of the rolled rollings calculated by the load prediction error calculation unit A load balance maintenance value representing a degree of difference between a load prediction error at a stand and a load prediction error at a rolling stand adjacent to each of the rolling stands is expressed as each rolling stand. The load balance maintenance value calculation unit to calculate the load balance, the load prediction error at each rolling stand calculated by the load prediction error calculation unit, and the load balance maintenance value at each rolling stand calculated by the load balance maintenance value calculation unit From the load correction value calculation unit for calculating the load correction value at each rolling stand, and the steel sheet to be rolled next, the rolling load at each rolling stand is estimated, and at each estimated rolling stand A control command set-up unit that corrects the rolling load of the rolling stand with the load correction value at each rolling stand calculated by the load correction value calculating unit, and calculates the reduction position to be set at each rolling stand using the corrected rolling load. And.

  ADVANTAGE OF THE INVENTION According to this invention, the tandem rolling mill control apparatus which can implement | achieve the balance maintenance of the rolling load between several rolling stands by a simple process, and its control method are provided.

It is the figure which showed the example of the structure of the tandem rolling mill control apparatus and control object which concern on embodiment of this invention. It is the figure which showed the example of the processing flow of the process which a control command setup part performs. It is the figure which showed the example of the structure of the draft schedule table memorize | stored in a draft schedule memory | storage part. It is the figure which showed the example of the structure of the speed pattern table memorize | stored in a speed pattern memory | storage part. It is the figure which showed the example of the processing flow of the process which an intermediate | middle board thickness calculation part performs. It is the figure which showed the example of the processing flow of the process which a load prediction error calculation part performs. It is the figure which showed the example of the processing flow of the process which a load balance maintenance value calculation part performs. It is the figure which showed the example of the structure of the load balance ratio table memorize | stored in a load balance ratio memory | storage part. It is the figure which showed the example of the processing flow of the process which a load correction value calculation part performs. It is the figure which showed the example of the structure of the load correction performance value table memorize | stored in a load correction performance value storage part. It is the figure which showed the example of the structure of the tandem rolling mill control apparatus which concerns on the 2nd Embodiment of this invention. It is the figure which showed the example of the processing flow of the process which a steel kind similarity calculation part performs. It is the figure which showed the example of the structure of the similarity number table memorize | stored in a similarity number memory | storage part.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a diagram showing an example of the configuration of a tandem rolling mill control device 10 and a control object 50 according to an embodiment of the present invention. As shown in FIG. 1, the tandem rolling mill control device 10 acquires signals representing various states from the control target 50 and outputs various control signals to the control target 50. In FIG. 1, a signal connecting the blocks is represented by a single arrow regardless of the type or number of signals.

  First, the configuration of the controlled object 50 will be described with reference to FIG. In the present embodiment, the controlled object 50 is a hot tandem rolling mill provided with a finishing mill 60. The finishing mill 60 includes a plurality of rolling stands 61, and rolls a rough material 65 having a thickness of, for example, about 30 mm, which has been rolled by a rough rolling machine (not shown) in the previous process, to produce a thin steel plate 63. .

In the example of FIG. 1, the finishing mill 60 has a configuration in which seven rolling stands 61 are continuously arranged, and the steel plate 63 (coarse material 65) is rolled while being moved from left to right. It is supposed to be. Specifically, the steel plate 63 (coarse material 65) is successively thinned by rolling at each rolling stand 61 (F _{1 to} F ₇ ), and is about 1 mm to 15 mm from the exit side of the rolling stand 61 (F ₇ ). The steel plate 63 is paid out.

  In the finishing mill 60, it is the work roll 62 of each rolling stand 61 that directly rolls the rough material 65 and the steel plate 63. The coarse material 65 may be called by a name such as a coarse bar, an incoming bar, or a transfer bar. Further, in this specification, the roll speed means the peripheral speed of the work roll 62.

Further, a multigauge 64 for measuring the thickness, width, temperature, etc. of the steel plate 63 is disposed on the exit side of the final rolling stand 61 (F ₇ ) of the finishing mill 60. Although not shown in FIG. 1, actually, as a detector for grasping the state of the coarse material 65 and the steel plate 63, a thermometer for measuring the temperature of the coarse material 65 and the steel plate 63, the steel plate 63 Various detectors, such as a shape meter that measures the flatness of the steel, a crop profile meter that measures the leading edge shape image of the rough material 65, and a surface wrinkle meter that detects surface flaws on the steel plate 63 are deployed in various places as needed. Has been.

  Then, the structure of the tandem rolling mill control apparatus 10 is demonstrated. As shown in FIG. 1, the tandem rolling mill control device 10 includes a control command setup unit 11, a rolling performance collection unit 12, an intermediate sheet thickness calculation unit 13, a load prediction error calculation unit 14, a load balance maintenance value calculation unit, A load correction value calculation unit 16, a rolling position control unit 17, a roll speed control unit 18, a draft schedule storage unit 21, a speed pattern storage unit 22, a load correction value storage unit 23, and the like are configured.

  The control command setup unit 11 receives information necessary for rolling, such as the steel type, the target plate thickness, and the target plate width of the steel plate 63 to be rolled, which is transmitted from the host computer 40. Then, according to the received information, the rolling load, rolling position (roll gap), roll speed, etc. of each rolling stand 61 are calculated using information obtained from the draft schedule storage unit 21 and the speed pattern storage unit 22. To do. Although details will be described later, in the calculation of the rolling load, the load correction value calculated by the load correction value calculation unit 16 is taken into consideration.

  Further, the control command setup unit 11 sends the calculation results such as the rolling load, the rolling position, and the roll speed of each rolling stand 61 to the rolling position control unit 17 and the roll speed control unit 18 as control commands for the rolling position and the roll speed, respectively. Output. Then, the reduction position control unit 17 outputs a control value for controlling the rolling load and the reduction position to each rolling stand 61 based on this control command. Similarly, the roll speed control unit 18 outputs a control value for controlling the roll speed to each rolling stand 61 based on the roll speed command.

  The rolling record collecting unit 12 is a rolling record value of the steel plate 63 detected via the multigauge 64 or the like, or a control command value (rolling) output to the control object 50 from the rolling position control unit 17 or the roll speed control unit 18. Load, rolling position, roll speed, etc.).

  The intermediate plate thickness calculation unit 13 estimates the intermediate plate thickness of the steel plate 63 between the rolling stands 61 using the data collected by the rolling record collection unit 12. Further, the load prediction error calculation unit 14 predicts the rolling load of each rolling stand 61 using the intermediate plate thickness estimated by the intermediate plate thickness calculation unit 13, and the deviation from the actual load (hereinafter referred to as load prediction error). ).

  Moreover, the load balance maintenance value calculation unit 15 includes, among the load prediction errors calculated by the load prediction error calculation unit 14, a load prediction error of the rolling stand 61 of interest, a load prediction error of the adjacent rolling stand 61, and A load balance maintenance value for maintaining the load balance between the rolling stands 61 is calculated using the ratio acquired from the load balance ratio storage unit 23 (load balance ratio in FIG. 8 and the like).

  Further, the load correction value calculation unit 16 stores the load prediction error calculated by the load prediction error calculation unit 14, the load balance maintenance value calculated by the load balance maintenance value calculation unit 15, and the load correction actual value storage unit 24. A load correction value to be output to the control command setup unit 11 is calculated using the stored load correction value calculated in the past rolling.

  The tandem rolling mill control device 10 having the above configuration is realized as a specific hardware by a computer or a workstation including an arithmetic processing device and a storage device. Then, the control command setup unit 11, the rolling record collection unit 12, the intermediate plate thickness calculation unit 13, the load prediction error calculation unit 14, the load balance maintenance value calculation unit 15, the load correction value calculation unit 16, the reduction position shown in FIG. The functions of each unit such as the control unit 17 and the roll speed control unit 18 are performed by the arithmetic processing unit including a microprocessor executing a predetermined program stored in the storage unit including a semiconductor memory or a hard disk unit. Realized. In addition, storage units such as the draft schedule storage unit 21, the speed pattern storage unit 22, the load balance ratio storage unit 23, and the load correction actual value storage unit 24 store predetermined data in an area allocated to a part of the storage device. It is realized by doing.

  Hereinafter, the operation of each part constituting the tandem rolling mill control device 10 will be described in detail sequentially. First, the operation of the control command setup unit 11 will be described with reference to FIGS. Here, FIG. 2 is a diagram illustrating an example of a processing flow of processing executed by the control command setup unit 11. 3 shows an example of the configuration of the draft schedule table 211 stored in the draft schedule storage unit 21, and FIG. 4 shows an example of the configuration of the speed pattern table 221 stored in the speed pattern storage unit 22. FIG.

  When the steel plate 63 is rolled by the finishing mill 60 (see FIG. 1), in order to obtain a desired plate thickness from the tip of the steel plate 63, the rolling load in each rolling stand 61 and the reduction position of the work roll 62 are appropriate. It is required to be. Further, in order to stabilize the behavior when the steel plate 63 is caught in the downstream rolling stand 61, the roll speed of each rolling stand 61 is disturbed by the mass flow (product of the plate thickness and the plate speed) of the steel plate 63. There is no need to be balanced.

  Therefore, the control command setup unit 11 receives information such as the steel type, target plate thickness, target plate width, and the like of the steel plate 63 to be rolled from now on, which is transmitted from the host computer 40, and rolls the steel plate 63 as intended. Control commands such as the reduction position and roll speed necessary for the calculation are calculated.

  As shown in FIG. 2, the control command setup unit 11 first refers to the draft schedule table 211 (see FIG. 3) stored in the draft schedule storage unit 21 and refers to the rolling target transmitted from the host computer 40. A draft schedule corresponding to the steel type, target plate thickness, and target plate width of the steel plate 63 is acquired, and the rolling reduction of each rolling stand 61 is calculated (step S11).

Here, as shown in FIG. 3, the draft schedule table 211 is configured by a draft schedule stratified by the steel type, target plate thickness, target plate width, and the like of the steel plate 63 to be rolled. The draft schedule is information representing how much the rough material 65 or the steel plate 63 is rolled in each of the rolling stands 61 (F _{1 to} F ₇ ), that is, the rolling reduction (the entry side and the exit side with respect to the entry side plate thickness). The ratio of the plate thickness difference) is expressed in percentage.

For example, suppose that a rough material 65 having a steel type of SS400 and a thickness of 35 mm is rolled into a steel plate 63 having a target plate thickness of 2.5 mm and a target plate width of 900 mm. In the draft schedule table 211 of FIG. 3, the steel plates 63 are stratified into a target plate thickness of 2.0 to 3.0 mm and a target plate width of 1000 mm or less. Accordingly, the coarse material 65 having a plate thickness of 35 mm is rolled by 14 mm corresponding to 40% thereof at the most upstream rolling stand 61 (F ₁ ), resulting in a steel plate 63 having a discharge plate thickness of 21 mm. Similarly, in the rolling stand 61 (F ₂ ), 35% of the steel plate 63 having an inlet side plate thickness of 21 mm is rolled into a steel plate 63 having an outlet plate thickness of 13.65 mm.

Note that a slight deviation occurs between the exit side plate thickness and the target plate thickness of 2.5 mm in the final rolling stand 61 (F ₇ ) thus obtained. The rolling reduction of each rolling stand 61 can be eliminated by correcting according to the rolling reduction.

Returning to the description of the processing flow shown in FIG. The control command setup unit 11 refers to the speed pattern table 221 (see FIG. 4) stored in the speed pattern storage unit 22 following the processing of step S11, and the steel type, target plate thickness, A speed pattern corresponding to the target sheet width is obtained, and the rolling speed (sheet speed) of the steel sheet 63 on the exit side of the final stage rolling stand 61 (F ₇ ) is calculated (step S12).

As shown in FIG. 4, the speed pattern table 221 is configured by speed patterns stratified by the steel type, target plate thickness, target plate width, and the like of the steel plate 63 to be rolled. Here, the speed pattern is speed (plate speed) information when the steel plate 63 to be rolled is paid out from the rolling stand 61 (F ₇ ) at the final stage, for example, initial speed, first acceleration, second acceleration. Refers to information composed of steady speed, deceleration, and final speed.

The initial speed is the speed at which the tip of the steel plate 63 is paid out from the final rolling stand 61 (F ₇ ), and the first acceleration is the speed of the steel plate 63 after the tip of the steel plate 63 is paid out. The acceleration rate and the second acceleration are the acceleration rate and deceleration until the steady speed is reached after the steel plate 63 is bitten by a downcoiler (not shown in FIG. 1) as the subsequent equipment. The deceleration rate when the steel plate 63 stably passes through each rolling stand 61 and decelerates to the final speed, and the final speed are when the tail end of the steel plate 63 is paid out from the final rolling stand 61 (F ₇ ). Say speed.

  In the example of FIG. 4, when the steel type is SS400, the target plate thickness is 1.2 to 1.4 mm, and the target plate width is 1000 mm or less, the initial speed is 650 mpm (meter per minute), and the first acceleration is 2 mpm / s, the second acceleration is 12 mpm / s, the steady speed is 1100 mpm, the deceleration is 6 mpm / s, and the final speed is 700 mpm.

  Returning to the description of the processing flow shown in FIG. The control command setup unit 11 performs a process of estimating the rolling temperature of each rolling stand 61 following the process of step S12 (step S13). At this time, the temperature of the coarse material 65 and the steel plate 63 is the temperature detected by thermometers (not shown in FIG. 1) installed at various locations of the controlled object 50, heat radiation, heat transfer, and further the steel plate 63 by rolling. It is estimated by combining processing heat generation due to deformation, roll contact heat transfer taken by the roll surface during rolling, and the like. A number of temperature estimation methods have been introduced in thermodynamic literature, and the temperature change in rolling of the steel plate 63 is, for example, “the theory and practice of plate rolling” (Japan Steel Association, issued in September 2010). Chapter 6 “Temperature change in rolling” is described in detail, and detailed description thereof is omitted.

  Next, the control command setup unit 11 calculates a deformation resistance that is a value corresponding to the hardness of the steel plate 63 rolled at each rolling stand 61 (step S14). In addition, how to obtain the deformation resistance is described in various documents. For example, the above-mentioned document “Theory and Practice of Sheet Rolling” is described in detail in Chapter 7 (deformation resistance).

Incidentally, the deformation resistance Kf can be calculated by the following equation (1) according to Equation 7.54 in the above-mentioned document “Theory and Practice of Sheet Rolling”.

Kf = κ · ε ⁿ · (dε / dt) ^m · exp (A / T) (1)

Where T: estimated rolling temperature of the steel sheet 63
ε: Strain
dε / dt: Strain rate
κ, n, m, A: Constants depending on steel type

Next, the control command setup unit 11 calculates the roll speed at each rolling stand 61 (step S15). Since the plate speed on the exit side of the rolling stand 61 (F ₇ ) at the final stage is obtained in step S12, here, based on this, first, the following equation (2) is used. The plate speed Vs _i on the exit side of the rolling stand 61 is calculated.

Vs _i = Vs ₇ · h _i / h ₇ (2)

Here, Vs _i : Plate speed on the exit side of the rolling stand (F _i )
Vs ₇ : Plate speed on the exit side of the rolling stand (F ₇ ) (final stage rolling stand)
h _i : Thickness at the exit of the rolling stand (F _i )
h ₇ : Thickness at the exit side of the rolling stand (F ₇ ) (final stage rolling stand)

Subsequently, the control command setup section 11 uses the concept of forward slip, calculating the roll speed at each roll stand 61 from the plate-speed Vs _i at each rolling stand 61 outlet side. Here, the advanced rate is a value corresponding to the ratio between the peripheral speed of the work roll 62 and the exit speed of the steel plate 63 rolled by the work roll 62. For example, the advanced rate f is known to be expressed as a function of a plurality of parameters as shown in the following formula (3) (for details, refer to the above-mentioned literature “Theory and Practice of Sheet Rolling”). reference).

f = F (H, h, R ′, Kf, tb, tf) (3)

Here, H: entry side plate thickness, h: exit side plate thickness, R ′: flat roll diameter,
Kf: deformation resistance, tb: entry side tension, tf: exit side tension

Therefore, the use of forward slip _{f i} at the i-th rolling stand 61 _{(F i),} the roll speed Vr _i of the rolling stand 61 _{(F i),} the rolling stands _{(F i)} plate speed at the delivery side It can be calculated by the following equation (4) using Vs _i .

Vr _i = Vs _i / f _i (4)

Further, the control command setup unit 11 calculates a rolling load predicted value P at each rolling stand 61 (step S16). Here, the formula for calculating the rolling load prediction value P is a function of a plurality of parameters such as the following formula (5), as shown in detail in, for example, the above-mentioned document “Theory and Practice of Sheet Rolling”. It is known to be expressed as

P = G (w, Kf, Qp, tf, tb, R ′, H, h, μ) (5)

Where w: plate width, Kf: deformation resistance, Qp: rolling force function, μ: friction coefficient,
tb: entry side tension, tf: exit side tension, H: entry side plate thickness, h: exit side plate thickness,
R ': Flat roll diameter

  Note that there is a discrepancy between the rolling load predicted value P calculated by Equation (5) and the rolling load obtained by actual rolling. Therefore, in order to reduce the divergence and increase the accuracy of the rolling load prediction value P, as an actual rolling load, an appropriate correction coefficient (in this specification, the rolling load prediction value P calculated by the equation (5) Multiplied by a load correction value). The calculation of the correction coefficient will be described in detail in the description of the processing of the load correction value calculation unit 16 (see FIG. 9).

Finally, the control command setup unit 11 calculates the reduction position (roll gap) of the work roll 62 at each rolling stand 61 (step S17). The reduction position S of the work roll 62 can be basically obtained by the following equation (6). However, in practice, various correction terms are often added to improve calculation accuracy.

S = h−P / K (6)

Here, P: rolling load prediction value, K: mill spring constant, h: outlet side plate thickness

  As described above, the control command setup unit 11 outputs the control command for the rolling position and roll speed calculated for the steel sheet 63 to be rolled to the rolling position control unit 17 and the roll speed control unit 18. The reduction position control unit 17 performs the reduction position control so that the reduction position of the work roll 62 becomes a value according to the control command with respect to the control command of the reduction position output by the control command setup unit 11. Similarly, the roll speed control unit 18 performs speed control so that the roll speed of the work roll 62 becomes a value according to the control command in response to the roll speed control command output by the control command setup unit 11.

  The processing in the control command setup unit 11 described above is performed on the steel plate 63 to be rolled. The intermediate plate thickness calculation unit 13, the load prediction error calculation unit 14, and the load balance maintenance value calculation described below. The processes in the unit 15 and the load correction value calculation unit 16 are performed using various rolling record values acquired by the rolling at the timing when the rolling is completed. Hereinafter, in this specification, the rolling targeted by these treatments is referred to as the rolling, and the steel plate 63 produced by the rolling is referred to as the steel plate 63.

FIG. 5 is a diagram illustrating an example of a processing flow of processing executed by the intermediate plate thickness calculation unit 13. As shown in FIG. 1 above, the multi-gauge 64 is only disposed on the exit side of the final stage rolling stand 61 (F ₇ ), and each rolling stand 61 (F _{1 to} F _7). ) In the middle. Therefore, the intermediate plate thickness calculation unit 13 estimates the plate thickness of the steel plate 63 (hereinafter referred to as the intermediate plate thickness) at the intermediate position of each rolling stand 61 (F _{1 to} F ₇ ).

In the following description, the intermediate plate thickness t _{i at} the intermediate position between the rolling stand 61 (F _i ) and the rolling stand 61 (F _{i + 1} ) (i = 1 to 6) is the input of the rolling stand 61 (F _{i + 1} ). The side plate thickness is assumed to be the same as t _{i + 1} . Also, when the detector for measuring the thickness of the coarse material 65 is not provided, the thickness t ₀ of the coarse material 65 _(i.e., rolling stands 61 (thickness at entrance side of the F ₁₎₎ and also to estimate the Do.

Therefore, as shown in FIG. 5, the intermediate plate thickness calculation unit 13 first uses a multigauge 64 disposed on the exit side of the final rolling stand 61 (F ₇ ) via the rolling record collection unit 12. acquires roll speed _Vr 1 through Vr ₇ work rolls 62 at the detected output side thickness _{t 7} and the rolling stand 61 _(F 1 ~F _{7) (step} S21).

Next, the intermediate plate thickness calculation unit 13 estimates the entry side plate thickness t ₆ from the exit side plate thickness t _{7 of} the rolling stand 61 (F ₇ ) based on the so-called constant mass flow rule (step S22). Here, the mass flow constant law means that the product of the exit side plate thickness t _i and the exit side plate speed Vs _i of the rolling stand 61 (F _i ) is the entry side plate thickness and the entrance side plate speed (that is, the rolling stand 61 ( F _i-1 ) is equal to the product of the exit side plate thickness t _i-1 and the exit side plate speed Vs _i-1 ).

That is, the intermediate plate thickness calculation unit 13 calculates the entry side plate thickness t ₆ of the rolling stand 61 (F ₇ ) based on the following equation (7). Note that the thickness at entrance side _{t 6} corresponds to the intermediate thickness _{t 6} between the rolling stands 61 and _{(F 6)} and the rolling stand 61 <sub>(F 7).

t 6 = t 7 · V 7</sub> · (1 + f 7) / {V 6 · (1 + f 6)} (7)

Here, t ₇ : exit side plate thickness of the rolling stand (F ₇ )
V ₇ : Roll speed (circumferential speed) of the work roll of the rolling stand (F ₇ )
f ₇ : Advanced rate of rolling stand (F ₇ )
V ₆ : Roll speed (circumferential speed) of the work roll of the rolling stand (F ₆ )
f ₆ : Advanced rate of rolling stand (F ₆ )

The advance rates f ₆ and f ₇ are calculated by the control command setup unit 11 according to the equation (3) prior to rolling the steel plate 63. Since the advanced rates f ₆ and f ₇ are estimated and calculated using Expression (3), the calculated values include a certain error. Therefore, an error is also included in the entry side plate thickness t ₆ calculated using the advanced rates f ₆ and f ₇ .

As described above, the intermediate plate thickness t ₆ between the rolling stand 61 (F ₆ ) and the rolling stand 61 (F ₇ ) is estimated. This intermediate plate thickness t ₆ is determined by the rolling stand 61 (F It is also the delivery side thickness _{t 6} of _6). Therefore, the intermediate plate thickness calculator 13 estimates the entry side plate thickness t ₅ of the rolling stand 61 (F ₆ ) from the outlet side plate thickness t ₆ of the rolling stand 61 (F ₆ ) in the same manner as in step S22 (step S22). S23).

Similarly, the intermediate plate thickness calculating unit 13, the rolling stands 61 _{(F 5)} estimates the thickness at entrance side _{t 4} in (step S24), and estimates the thickness at entrance side _{t 3} of the rolling stand 61 _{(F 4)} ( step S25), and the rolling stand 61 _{(F 3)} to estimate the thickness at entrance side _{t 2} of (step S26), to estimate the thickness at entrance side _{t 1} of the rolling stand 61 _{(F 2)} (step S27), further, rolling stand An entry-side plate thickness t ₀ of 61 (F ₁ ) is estimated (step S28). Incidentally, thickness at entrance side _{t 0} corresponds to the thickness _{t 0} of the coarse material 65

Above, in this embodiment, the thickness at entrance side of the rolling stand 61 shows an example of obtaining from the downstream stand in order, the rolling stand 61 which is measured (F _7), forward slip f of the rolling stand 61 (F _{7) 7,} may be calculated at a time from the relationship between the forward slip _{f i} and roll speed Vr _i of the rolling stand 61 _{(F i).}

FIG. 6 is a diagram illustrating an example of a processing flow of processing executed by the load prediction error calculation unit 14. As shown in FIG. 6, the load prediction error calculation unit 14 first selects one rolling stand 61 from among the _seven rolling stands 61 (F _{1 to} F ₇ ) (step S31). In this case, the selection order is not particularly limited. For example, the selection is made in order from the upstream rolling stand 61.

  Next, the load prediction error calculation unit 14 acquires the actual rolling load at the selected rolling stand 61 via the rolling performance collection unit 12 (step S32). Further, the load prediction error calculation unit 14 acquires the entry side plate thickness and the exit side plate thickness at the selected rolling stand 61 from the intermediate plate thickness calculation unit 13, and calculates the estimated rolling load at the rolling stand 61 by the above formula ( 5) is calculated (step S33).

  In the calculation of the equation (5), the measured results are used as they are for the output side tension tf and the input side tension tb that can be directly measured. The calculation formulas for the sheet width w, the deformation resistance Kf, the rolling force function Qp, the flat roll diameter R ′, and the friction coefficient μ are shown in the above-mentioned document “Theory and Practice of Sheet Rolling”. Accordingly, calculation is performed based on the actual values of various detectors acquired via the rolling performance collecting unit 12. At this time, in order to calculate the plate width w and the deformation resistance Kf, further rolling temperature is required. The rolling temperature is based on the value detected by a temperature detector (not shown in FIG. 1). The one estimated in consideration of the distance from the installation position of the temperature detector to the rolling stand 61 is used.

Next, the load prediction error calculation unit 14 calculates a load prediction error that is a ratio of the actual rolling load acquired in step S32 and the estimated rolling load calculated in step S33 (step S34). That is, the load prediction error Zpl _i of the rolling stand 61 (F _i ) is calculated according to the following equation (8).

Zpl _i = Pa _i / Ps _i (8)

Here, Pa _i : Actual rolling load at the rolling stand (F _i )
Ps _i : Estimated rolling load based on various actual values at the rolling stand (F _i )

Next, the load prediction error calculation unit 14 determines whether or not the process of calculating the load prediction error Zpl _i has been completed for all the rolling stands 61 (F _{1 to} F ₇ ) (step S35), and the process ends. If not (No in step S35), the processes in and after step S31 are repeated. Further, when the process of calculating the load prediction error Zpl _i is completed for all the rolling stands 61 (F _{1 to} F ₇ ) (Yes in step S35), the load prediction error Zpl _i for the steel plate 63 is determined. The calculation process is terminated.

FIG. 7 is a diagram illustrating an example of a processing flow of processing executed by the load balance maintenance value calculation unit 15. By the way, although the rolling load prediction value P in each rolling stand 61 is calculated with respect to the steel plate 63 rolled next time by the control command setup part 11, in order to raise the prediction precision, the calculated rolling load prediction value P is used. Correction processing is performed. At that time, it is possible to use the load prediction error Zpl _i as a load correction value as it is. However, since the load prediction error Zpl _i is calculated for each rolling stand 61, there is considerable variation among the plurality of rolling stands 61. May occur. As a result, the degree of correction of the rolling load predicted value P at each rolling stand 61 also varies.

  The corrected rolling load prediction value P is actually used for calculating the rolling load at each rolling stand 61, that is, the reduction position. Therefore, even if the rolling load prediction value P is corrected, if a variation remains between the rolling stands 61 adjacent to each other, there arises a problem that the rolling load balance may be lost due to the variation. Therefore, in the processing of the load balance maintenance value calculation unit 15 described below, a value (hereinafter referred to as a load balance maintenance value) for suppressing the collapse of the rolling load balance between the adjacent rolling stands 61 is calculated.

As shown in FIG. 7, the load balance maintenance value calculation unit 15 first selects one rolling stand 61 (F _i ) from among the _seven rolling stands 61 (F _{1 to} F ₇ ) (step S41). In this case, the selection order is not particularly limited. For example, the selection is made in order from the upstream rolling stand 61.

Next, the load balance maintenance value calculation unit 15 acquires the load prediction error Zpl _i of the selected rolling stand 61 (F _i ) from the processing result in the load prediction error calculation unit 14 (step S42), and rolling stands 61 _{(F i)} rolling stand ₆₁ adjacent to _{(F i-1, F i} + 1) to obtain a load prediction error _{Zpl i-1, Zpl i +} 1 of the (step S43). Here, the adjacent rolling stands _{61 (F i-1, F} i + 1), a rolling stand 61 of the upstream side of the rolling stand _{61 (F i) (F i} -1) and rolling stand 61 on the downstream side _{(F i + 1} ).

Of course, the rolling stand 61 adjacent to the most upstream rolling stand 61 (F ₁ ) is only the rolling stand 61 (F ₂ ), and the rolling adjacent to the most downstream rolling stand 61 (F ₇ ). The stand 61 is only the rolling stand 61 (F ₆ ). However, in this specification, although there are such exceptions at the most upstream and downstream, the rolling stand 61 adjacent to the rolling stand 61 (F ₁ ) is the rolling stand 61 (F _i−1 , F _{i + 1} ). write.

Next, the load balance maintenance value calculation unit 15 calculates the load balance maintenance value Zplb _i of the selected rolling stand 61 according to, for example, the following equation (9) (step S44).

When i = 2 to 6:
Zplb _i = α · {(Zpl _i−1 −Zpl _i ) + (Zpl _{i + 1} −Zpl _i } / 4
When i = 1:
Zplb _i = α · (Zpl _{i + 1} −Zpl _i ) / 3
When i = 7:
Zplb _i = α · (Zpl _i−1 −Zpl _i ) / 3 (9)

Where Zpl _i : Load prediction error
α: Load balance ratio

According to equation (9), the load balance maintained value Zplb _i is the load prediction error Zpl rolling stands 61 _{(F i)} of the load prediction error Zpl _i the rolling stand ₆₁ adjacent to it _{(F i-1, F i} + 1) _This corresponds to the average difference between _i−1 and Zpl _{i + 1} . Therefore, when a load prediction error Zpl _i rolling stands 61 _{(F 1)} is larger than the load prediction error _{Zpl i-1, Zpl i +} 1 of the rolling stand ₆₁ adjacent _{(F i-1, F i} + 1) is a load balance maintained The value Zplb _i is a negative value. In addition, the, the load balancing when the load prediction error Zpl _i is smaller than the load prediction error _{Zpl i-1, Zpl i +} 1 of the rolling stand ₆₁ adjacent _{(F i-1, F i} + 1) of the rolling stand 61 _{(F 1)} The maintenance value Zplb _i is a positive value. Further, when the load prediction error _{Zpl i-1, Zpl i +} 1 of the rolling stand ₆₁ to the load prediction error Zpl _i rolling stands 61 _{(F 1)} is adjacent _{(F i-1, F i} + 1) is generally the same, the load balance The maintenance value Zplb _i is a value close to 0 (zero).

Next, the load balance maintenance value calculation unit 15 determines whether or not the process of calculating the load balance maintenance value Zplb _i has been completed for all the rolling stands 61 (F _{1 to} F ₇ ) (Step S45). If not completed (No in step S45), the processes in and after step S41 are repeatedly executed. In addition, when the process of calculating the load balance maintenance value Zplb _i is completed for all the rolling stands 61 (F _{1 to} F ₇ ) (Yes in step S45), the load balance maintenance value Zplb for the steel plate 63 is completed. _The process of calculating _i ends.

In equation (9), the load balance ratio α is a parameter indicating how much the compensation of the load prediction error Zpl _i and the maintenance of the load balance are compatible, and is stored in the load balance ratio storage unit 23. . This load balance ratio α takes a value between 0 and 1, and when it is 0, it indicates that maintenance of the load balance is not considered. On the other hand, when α is 1, the maintenance of the load balance is considered to the maximum. At this time, the load balance maintenance value Zplb _i is considered at a ratio equivalent to the load prediction error Zpl _i .

  FIG. 8 is a diagram illustrating an example of the configuration of the load balance ratio table 231 stored in the load balance ratio storage unit 23. As shown in FIG. 8, the load balance ratio table 231 is configured with values of the load balance ratio α stratified by the steel type, target plate thickness, target plate width, and the like of the steel plate 63 to be rolled. For example, when the steel type is SS400, the target plate thickness is 2.0 to 3.0 mm, and the target plate width is 1000 mm or less, the load balance ratio α is 1.0. When the steel type is SS400, the plate thickness is 12.0 mm or more, and the plate width is 1400 mm or more, the load balance ratio α is 0.4.

FIG. 9 is a diagram illustrating an example of a processing flow of processing executed by the load correction value calculation unit 16. Load correction value calculation unit 16 uses the load prediction error Zpl _i calculated by the load prediction error calculating unit 14, and a load balance maintained value calculated by the load balance maintained value calculating section 15 Zplb _i, control command setup A load correction value (a first load correction value and a second load correction value referred to below) used to increase the accuracy of the predicted load value predicted by the unit 11 is calculated.

As shown in FIG. 9, the load correction value calculation unit 16 first selects one rolling stand 61 from among the _seven rolling stands 61 (F _{1 to} F ₇ ) (step S51). In this case, the selection order is not particularly limited. For example, the selection is made in order from the upstream rolling stand 61.

Next, the load correction value calculation unit 16 acquires the load prediction error Zpl _i of the selected rolling stand 61 (F _i ) from the processing result in the load prediction error calculation unit 14 (step S52), and further the load The load balance maintenance value Zplb _i of the selected rolling stand 61 (F _i ) is acquired from the processing result in the balance maintenance value calculation unit 15 (step S53).

Further, the load correction value calculation unit 16 calculates a first load correction value Zpn _i corresponding to the steel plate 63 of the selected rolling stand 61 (F _i ) according to the following equation (10) (step S54).

Zpn _i = Zpl _i + Zplb _i (10)

Here, Zpn _i : First load correction value
Zpl _i : Load prediction error
Zplb _i : Load balance maintenance value

As described above, the load balance maintenance value Zplb _i is equal to the load prediction error Zpl _i of the rolling stand 61 (F _i ) and the load prediction error Zpl _{i− of} the rolling stand 61 (F _i−1 , F _{i + 1} ) adjacent thereto _{. 1} , corresponding to the average of the differences from Zpl _{i + 1} . Therefore, when a load prediction error Zpl _i rolling stands 61 _{(F i)} is greater than the load prediction error _{Zpl i-1, Zpl i +} 1 of two adjacent rolling stands _{61 (F i-1, F} i + 1) is The load balance maintenance value Zplb _i is a negative value. Then, according to equation (10), the first load correction value Zpn _i, since obtained by adding the load balance maintained value Zplb _i to the load prediction error Zpl _i, is smaller than the load prediction error Zpl _i .

On the other hand, when calculating the first load correction values Zpn _i-1 and Zpn _{i + 1} of the adjacent rolling stands 61 (F _i−1 , F _{i + 1} ), the relatively large load of the rolling stand 61 (F _i ). Since the prediction error Zpl _i acts, the first load correction values Zpn _i−1 and Zpn _{i + 1} of the rolling stand 61 (F _i−1 , F _{i + 1} ) are larger than the load prediction errors Zpl _i−1 and Zpl _{i + 1.} Become.

That is, the value of the first load correction value Zpn _i of the rolling stand 61 (F _i ) having a relatively large load prediction error Zpl _i relative to the adjacent rolling stand 61 (F _i−1 , F _{i + 1} ) decreases. On the other hand, the values of the first load correction values Zpn _i and Zpn _{i + 1} of the adjacent rolling stand 61 (F _i−1 , F _{i + 1} ) having relatively small load prediction errors Zpl _i−1 and Zpl _{i + 1} increase. As a result, the first load correction values Zpn _i−1 , Zpn _i , and Zpn _{i + 1} vary between the rolling stand 61 (F _i ) and the adjacent rolling stand 61 (F _i−1 , F _{i + 1} ). Reduced. This means that the load balance between a certain rolling stand 61 (F _i ) and the adjacent rolling stand 61 (F _i−1 , F _{i + 1} ) is maintained.

Although the case load prediction error Zpl _i rolling stands 61 _{(F 1)} is greater than the average of the load prediction error _{Zpl i-1, Zpl i +} 1 of the rolling stand ₆₁ adjacent _{(F i-1, F i} + 1), they Although it has been described that the load balance between the rolling stands 61 (F _i−1 , F ₁ , F _{i + 1} ) is improved, the same applies to the reverse case. That, also when the load prediction error Zpl _i rolling stands 61 _{(F 1)} is smaller than the load prediction error _{Zpl i-1, Zpl i +} 1 of the rolling stand ₆₁ adjacent _{(F i-1, F i} + 1), rolling The load balance between the stand 61 (F ₁ ) and the adjacent rolling stand 61 (F _i−1 , F _{i + 1} ) is maintained.

Further, when the load prediction error Zpl _i rolling stands 61 _{(F i)} is approximately the same as the load prediction error _{Zpl i-1, Zpl i +} 1 of the rolling stand ₆₁ adjacent _{(F i-1, F i} + 1) is a load balance Since the maintenance value Zplb _i is a value close to 0, the first load correction value Zpn _i is substantially the same value as the load prediction error Zpl _i . In this case, it can be said that the load balance between the rolling stand 61 (F ₁ ) and the adjacent rolling stand 61 (F _i−1 , F _{i + 1} ) is originally maintained.

Subsequently, returning to the description of the processing flow of FIG. 9, FIG. 10 will be described. In FIG. 9, the load correction value calculation unit 16 performs load correction corresponding to the past rolling record of the rolling stand 61 (F ₁ ) selected in step S51 from the load correction record value storage unit 24 as the next process of step S54. The actual value Zpp _i is acquired (step S55).

FIG. 10 is a diagram illustrating an example of the configuration of the load correction actual value table 241 stored in the load correction actual value storage unit 24. As shown in FIG. 10, the load correction actual value table 241 includes load correction actual value Zpp _i of each rolling stand 61 (F _i ) stratified by the steel type, target plate thickness, and the like of the steel plate 63. Here, the load correction result values Zpp _i, each rolling stand 61 outputted from the load correction value calculating unit 16 to the control command setup section 11 when the steel plate 63 belonging to different respective layers are rolled in the past (F _i ) the second load correction value Zp _i (see step S56) is obtained by storing. For example, in the example of FIG. 10, when the steel type is SS400 and the steel plate 63 has a target plate thickness of 1.6 mm or less, the load correction actual value Zpp _i corresponding to each of the rolling stands 61 (F _{i to} F ₇ ) is 1. 11, 1.08, 0.94, 0.98, 1.04, 0.99, and 1.03.

Further, as shown in FIG. 9, the load correction value calculation unit 16 uses the first load correction value Zpn _i calculated in step S54 and the load correction actual value Zpp _i acquired in step S54 as processing subsequent to step S55. used, according to the following equation (11), second calculating a load correction value Zp _i, second load correction value Zp _i a control command setup section that the calculated relative to the rolling stand 61 selected at step S51 _{(F 1)} 11 (step S56).

Zp _i = β · Zpn _i + (1−β) · Zpp _i (11)

Here, Zp _i : Second load correction value of the rolling stand (F _i )
Zpn _i : First load correction value of the rolling stand (F _i )
Zpp _i : Actual load correction value of rolling stand (F _i )
β: Allocation coefficient

In calculating the second load correction value Zp _i of the rolling stand 61 (F _i ), the distribution coefficient β is a load correction actual value Zpp based on past rolling results stored in the load correction actual value storage unit 24. _This is a coefficient that determines the distribution ratio between _i and the first load correction value Zpn _i calculated in the most recent rolling of the steel sheet 63 (hereinafter, the latest means the closest past), and takes a value of 0 to 1. . That is, when the distribution coefficient β is 0, the first load correction value Zpn _i calculated by rolling the steel sheet 63 is ignored, and the load correction based on the past rolling record stored in the load correction record value storage unit 24 is performed. The second load correction value Zp _i is determined according to the actual value Zpp _i .

On the other hand, when the distribution coefficient β is 1, the load correction actual value Zpp _i based on the past rolling actual result is ignored, and the second load correction value is calculated according to the first load correction value Zpn _i calculated by rolling the steel plate 63. Zp _i is determined. When the distribution coefficient β is an intermediate value between 0 and 1 (0 <β <1), the first load correction value Zpn _i and the load correction actual value Zpp _i at a ratio according to the value of the distribution coefficient β. There are apportioned, second load correction value Zp _i is determined. For example, when β = 0.5, the first load correction value Zpn _i and the load correction actual value Zpp _i are equally distributed to determine the second load correction value Zp _i .

Furthermore, as shown in FIG. 9, the load correction value calculation unit 16 calculates in step S56, the second load correction value Zp _i outputted toward the control command setup section 11, with stratification in which the steel plate 63 belongs Then, it is stored in the load correction actual value storage unit 24 as the load correction actual value Zpp _i of the rolling stand 61 (F _i ) selected in step S51 (step S57).

Subsequently, the load correction value calculation unit 16, all of the rolling stands 61 _(F 1 ~F ₇₎ determines whether the processing of steps S51~ step S57 is finished for (step S58), the processing is ended If not (No in step S58), the processes in and after step S51 are repeatedly executed. Moreover, when the process from step S51 to step S57 is completed for all the rolling stands 61 (F _{1 to} F ₇ ) (Yes in step S58), the second load correction value Zp _i shown in FIG. The process of calculating the above is terminated.

Calculated as described above, the second load correction value Zp _i outputted to the control command setup section 11, the control command setup section 11 is used for the prediction accuracy of the rolling load prediction value P. That is, the control command setup section 11, in step S16 in the processing flow in FIG. 2, to calculate the rolling load prediction value P according to Equation (5), receiving a second load correction value Zp _i from the load correction value calculating section 16 When corrects formula the calculated rolling load prediction value P with (5) in the second load correction value Zp _i.

That is, the control command setup unit 11 calculates the rolling load set value Pset _i for calculating the rolling position command for each rolling stand 61 (F _i ) according to the following equation (12).

_{Pset i = Zp i · G (} w i, Kf i, Qp i, tf i, tb i,
R ′ _i , H _i , h _i , μ _i ) (12)

Here, w _i , Kf _i , Qp _i , tf _i , tb _i , R ′ _i , H _i , h _i , and μ _i are the plate width and deformation of the steel plate 63 in each rolling stand 61 (F _i ), respectively. Resistance, rolling force function, entry side tension, exit side tension, thickness flat roll diameter, entry side plate thickness, exit side plate, friction coefficient.

Therefore, even if the rolling load predicted value P _i calculated by the equation (5) varies, the rolling load predicted value P _i is supplied to the actual rolling stand 61 (F _i ) via the reduction position control unit 17 (see FIG. 1). The rolling load set value Pset _i is reduced in variation between adjacent rolling stands 61 (F _i−1 , F _{i + 1} ). Therefore, the load balance among the plurality of rolling stands 61 (F _i ) is prevented from being lost, that is, the load balance is maintained.

In the embodiment described above, the load balance ratio α is used in the calculation of the load balance maintenance value calculation unit 15 (see Expression (9)), but after removing α from Expression (9). The equation (10) may be replaced with the following equation (13) and used in the processing of the load correction value calculation unit 16.

Zpn _i = Zpl _i + α · Zplb _i (13)

Here, Zpn _i : First load correction value
Zpl _i : Load prediction error
Zplb _i : Load balance maintenance value

As described above, according to the embodiment of the present invention, the load balance maintenance value Zplb _i of each rolling stand 61 (F _i ) is adjacent to the load prediction error Zpl _{i in} the rolling stand 61 (F _i ). It is calculated according to the magnitude relationship with the load prediction errors Zpl _i−1 and Zpl _{i + 1 at i−1} , F _{i + 1} ) (see equation (9)). In that case, the load balance maintained value Zplb _i rolling stands 61 _{(F i),} in its rolling stands 61 _{(F i)} rolling stand ₆₁ a load prediction error Zpl _i are adjacent in _{(F i-1, F i} + 1) When it is larger than the load prediction errors Zpl _i−1 and Zpl _{i + 1} , it becomes a negative value, and when it is opposite, it becomes a positive value. That is, the first load correction value Zpn _i obtained by adding the load balance maintained value Zplb _i to the load prediction error Zpl _i (see equation (10)) is such as to suppress variation of the load prediction error Zpl _i.

Furthermore, a learning process is performed on the first load correction value Zpn _i to obtain a second load correction value Zp _i (see formula (11)), and the rolling stand 61 is obtained using the second load correction value Zp _i. A rolling load set value Pset _i that is actually set to (F _i ) is calculated (see formula (12)). Therefore, even if the rolling load predicted value P _i varies, the rolling load set value P set _i suppresses the variation. Therefore, in the embodiment of the present invention, it is possible to prevent the occurrence of a problem that the load balance among the plurality of rolling stands 61 (F _i ) is lost.

In the embodiments of the present invention, the load correction value of the rolling stand 61 _{(F i)} (second load correction value Zp _i) is basically a load prediction error Zpl _i rolling stands 61 _{(F i)} It is determined only by the magnitude relationship with the load prediction errors Zpl _i-1 and Zpl _{i + 1} of the adjacent rolling stands 61 (F _i−1 , F _{i + 1} ). Therefore, for example, there is no need to construct a rolling load error change model as disclosed in Patent Document 3. That is, in the embodiment of the present invention, it is possible to expect an effect that it is possible to maintain the load balance among the plurality of rolling stands 61 (F _i ) by a simple process. Further, for example, the rolling load error change model disclosed in Patent Document 3 is a linear model, but the fluctuation of the load prediction error Zpl _{i in} the embodiment of the present invention is not limited to the linear model, and various fluctuations are possible. It can correspond to the way.

Therefore, in the tandem rolling mill control device 10 according to the embodiment of the present invention, when predicting the rolling load at each rolling stand 61 (F _i ) for rolling to a desired plate thickness prior to rolling the steel plate 63, The rolling load can be estimated with a simple calculation in consideration of the load balance between the plurality of rolling stands 61 ( _Fi ). As a result, a steel plate 63 having a more accurate thickness can be obtained and the rolling can be stabilized.

(Second Embodiment)
FIG. 11 is a diagram showing an example of the configuration of a tandem rolling mill control device 10a according to the second embodiment of the present invention. The configuration of the tandem rolling mill control device 10a according to the second embodiment is that a steel type similarity calculation unit 31, a distribution coefficient calculation unit 32, and a similarity number storage unit 35 are newly added, and the tandem of FIG. This is different from the configuration of the rolling mill control device 10. Only this difference will be described below. In addition, the same code | symbol is attached | subjected about the same component as the component of the tandem rolling mill control apparatus 10 of FIG.

  In the tandem rolling mill control device 10a according to the second embodiment, a function is added in which the distribution coefficient β used in the equation (11) is variable according to the similarity number of the steel plate 63 rolled back and forth. That is, the steel type similarity calculation unit 31 refers to the similarity number storage unit 35 for the steel type of the steel plate 63 that was rolled most recently and the steel type of the steel plate 63 that will be rolled next time, acquires the respective similarity number, and distributes them. It outputs to the coefficient calculation part 32. The distribution coefficient calculation unit 32 calculates the distribution coefficient β based on the similarity numbers of the steel plates 63 before and after acquired from the steel type similarity calculation unit 31 and outputs the distribution coefficient β to the load correction value calculation unit 16. The load correction value calculation unit 16 calculates Equation (11) using the distribution coefficient β received from the distribution coefficient calculation unit 32 in the process of step S56.

  FIG. 12 is a diagram illustrating an example of a processing flow of processing executed by the steel type similarity calculation unit 31. FIG. 13 is a diagram showing an example of the configuration of the similarity number table 351 stored in the similarity number storage unit 35. As shown in FIG. 13, the similarity number table 351 is configured by defining a similarity number for each steel type of the steel plate 63. Here, the closer the similarity number of the steel types of the steel plate 63 is, the more similar the properties of the steel plates 63 are such as deformation resistance. The farther the similarity number is, the more the properties of the steel plates 63 are. Means different.

  For example, in the example of FIG. 13, the similarity number of the steel plate 63 whose steel type is SS400 is 4, and the similarity number of the steel plate 63 whose steel type is SS490 is 5. Therefore, it can be said that these steel plates 63 have a large similarity as a steel type. On the other hand, the similarity number of the steel plate 63 whose steel type is SS400 is 4, and the similarity number of the steel plate 63 whose steel type is SPHC is 2. Therefore, the similarity as a steel type of these steel plates 63 is not as great as the similarity between the steel plate 63 of SS400 and the steel plate 63 of SS490. Furthermore, since the similarity number of the steel plate 63 having the steel type SS400 is 4 and the similarity number of the steel plate 63 having the steel type DP is 25, the similarity of these steel plates 63 as the steel types is greatly different. Therefore, these steel plates 63 should no longer be called foreign steel plates 63.

Therefore, as shown in FIG. 12, the steel type similarity calculation unit 31 refers to the similarity number storage unit 35 and each of the steel type of the steel plate 63 rolled most recently and the steel type of the steel plate 63 rolled next time, respectively. The similarity number corresponding to the steel type is acquired (step S51). Next, the similarity V is calculated from the difference between the similarity numbers of the two steel plates 63 according to the following equation (14) (step S52). However, the similarity V defined by the equation (14) is larger as the value is smaller, and the similarity is smaller as the value is larger.

V = | Vn _i −Vn _j | (14)

Here, Vn _i: steels similarity number of most recently rolled steel sheet 63
Vn _j : Similarity number of the steel type of the steel plate 63 to be rolled next time

  Next, the steel type similarity calculation unit 31 outputs the calculated similarity V to the distribution coefficient calculation unit 32 (step S53), and ends the process.

Subsequently, the distribution coefficient calculation unit 32 calculates the distribution coefficient β according to the following equation (15), for example.

β = 1− (V / Vc) (15)

However, when Vc <V, V = Vc, and Vc is a similarity corresponding to β = 0.

  According to equation (15), the greater the similarity between the steel plates 63 rolled back and forth (the smaller the similarity V), the closer the distribution coefficient β is to 1 and the smaller the similarity (the similarity V is large). ), The distribution coefficient β becomes a value close to zero. When the similarity V is greater than Vc, the distribution coefficient β is 0. The load correction value calculation unit 16 calculates Equation (11) using the distribution coefficient β.

  The second embodiment described above shows an example in which the distribution coefficient β used in Expression (11) is specifically calculated, and the effect is almost the same as the effect of the above-described embodiment. It is.

  In the embodiment and the second embodiment of the present invention described above, the tandem rolling mill control devices 10 and 10a are applied to hot rolling, but can also be applied to cold rolling. It is.

  In addition, this invention is not limited to embodiment described above, Furthermore, various modifications are included. The above-described embodiment has been described in detail for easy understanding of the present invention, and is not necessarily limited to one having all the configurations described. Further, a part of the configuration of an embodiment can be replaced with a part of the configuration of another embodiment, and further, a part or all of the configuration of the other embodiment is added to the configuration of the certain embodiment. Is also possible.

DESCRIPTION OF SYMBOLS 10,10a Tandem rolling mill control apparatus 11 Control command setup part 12 Rolling performance collection part 13 Intermediate sheet thickness calculation part 14 Load prediction error calculation part 15 Load balance maintenance value calculation part 16 Load correction value calculation part 17 Rolling position control part 18 Roll Speed control unit 21 Draft schedule storage unit 22 Speed pattern storage unit 23 Load balance ratio storage unit 24 Load correction actual value storage unit 31 Steel type similarity calculation unit 32 Distribution coefficient calculation unit 35 Similarity number storage unit 40 Host computer 50 Control target 60 Finishing mill 61 Rolling stand 62 Work roll 63 Steel plate 64 Multi gauge 65 Rough material 211 Draft schedule table 221 Speed pattern table 231 Load balance ratio table 241 Load correction actual value table 351 Similarity number table

Claims (7)

A control device for a tandem rolling mill that continuously rolls steel sheets with a plurality of rolling stands,
When the steel plate is rolled, the rolling load at each rolling stand is estimated using the rolling record value obtained at each rolling stand of the plurality of rolling stands, and rolling at each estimated rolling stand is performed. Based on the load and the rolling load actual value at each rolling stand obtained by the rolling, a load prediction error calculating unit that calculates a load prediction error at each rolling stand,
The load balance maintenance value representing the degree of difference between the load prediction error at each rolling stand calculated by the load prediction error calculation unit and the load prediction error at the rolling stand adjacent to each rolling stand, A load balance maintenance value calculator for calculating the stand,
The load correction at each rolling stand is calculated from the load prediction error at each rolling stand calculated by the load prediction error calculating unit and the load balance maintaining value at each rolling stand calculated by the load balance maintaining value calculating unit. A load correction value calculation unit for calculating a value;
Next, with respect to the steel sheet to be rolled, the rolling load at each rolling stand is estimated, and the estimated rolling load at each rolling stand is calculated by the load correction value calculation unit at each rolling stand. And a control command setup unit for calculating a rolling position to be set in each rolling stand using the corrected rolling load,
A control device for a tandem rolling mill. A load balance ratio storage unit that stores a load balance ratio, which is a constant not less than 0 and not more than 1 representing the importance of the load balance maintenance value, in association with at least one of the steel type, target plate thickness, and target plate width of the steel plate. Further comprising
The load correction value calculation unit
The load balance ratio corresponding to the steel type, target plate thickness and target plate width of the rolled steel sheet is acquired from the load balance ratio storage unit, and the value obtained by multiplying the load balance maintenance value by the acquired load balance ratio and the The control apparatus for a tandem rolling mill according to claim 1, wherein the load correction value is calculated by adding a load prediction error value. A load correction value storage unit that stores the load correction value calculated in the past rolling record in association with the steel type, target plate thickness, and target plate width of the steel plate;
The load correction value calculation unit
Obtained from the load prediction error value calculated by the load prediction error calculation unit, the load correction value calculated using the load balance maintenance value calculated by the load balance maintenance value calculation unit, and the load correction value storage unit The tandem rolling mill control device according to claim 1, wherein a load correction value output to the control command setup unit is calculated from the load correction value calculated in the past rolling. The plate thickness of the steel plate detected by the plate thickness measuring means provided on the exit side of the final rolling stand of the tandem rolling mill, and the roll speed that is the peripheral speed of each work roll of the plurality of rolling stands An intermediate plate thickness calculation unit that sequentially calculates the intermediate plate thickness, which is the plate thickness on the entry side of each of the plurality of rolling stands, from the rolling stand at the final stage toward the upstream,
The load prediction error calculation unit
When estimating the rolling load in the rolling based on the rolling performance of the rolled steel plate, the rolling load in each of the plurality of rolling stands is estimated using the intermediate plate thickness calculated by the intermediate plate thickness calculation unit. The control apparatus of the tandem rolling mill according to claim 1, wherein the control apparatus is a tandem rolling mill. A similarity number storage unit that stores a similarity number corresponding to a steel type of the steel plate, as the similarity of the properties of the steel plate is greater,
A steel sheet similarity calculation unit that acquires the similarity number of each of the rolled steel sheet and the next rolled steel sheet from the similarity number storage unit, and calculates a difference between the two acquired similarity numbers as a similarity degree; ,
A distribution coefficient calculation unit that calculates a distribution coefficient that is a constant between 0 and 1 based on the calculated similarity;
The load correction value calculation unit
The calculated load correction value is further corrected to a value obtained by apportioning and adding the load correction value and the load correction actual value acquired from the load correction actual value storage unit by the calculated distribution coefficient. The control apparatus of the tandem rolling mill according to claim 1. The distribution coefficient calculation unit
The hot rolling mill according to claim 5, wherein the distribution coefficient is calculated such that the distribution coefficient increases as the similarity decreases, and the distribution coefficient decreases as the similarity increases. Control device. A computer that controls a tandem rolling mill that continuously rolls steel sheets with a plurality of rolling stands.
When the steel plate is rolled, the rolling load at each rolling stand is estimated using the rolling record value obtained at each rolling stand of the plurality of rolling stands, and rolling at each estimated rolling stand is performed. Based on the load and the actual rolling load value at each rolling stand obtained by the rolling, a load prediction error calculating step for calculating a load prediction error at each rolling stand,
A load balance maintenance value representing a degree of difference between a load prediction error at each rolling stand calculated at the load prediction error calculation step and a load prediction error at a rolling stand adjacent to each rolling stand, A load balance maintenance value calculating step for calculating the stand;
From the load prediction error at each rolling stand calculated in the load prediction error calculation step and the load balance maintenance value at each rolling stand calculated at the load balance maintenance value calculation step, load correction at each rolling stand is performed. A load correction value calculating step for calculating a value;
Next, with respect to the steel sheet to be rolled, the rolling load at each rolling stand is estimated, and the estimated rolling load at each rolling stand is calculated at the load correction value calculating step. And a control command setup step for calculating a rolling position to be set in each rolling stand using the corrected rolling load,
The control method of the tandem rolling mill characterized by performing this.

Images