KR20220161169A - Plant control apparatus, plant control method and program - Google Patents

Abstract

According to the present invention, occurrence of an operational abnormality in a plant to be controlled is predicted so as to appropriately operate a control operation end to prevent the operational abnormality from occurring, thereby improving a control effect and operational efficiency. A plant to be controlled is subjected to first operation processing in which the response speed with respect to operation is a predetermined response speed and second operation processing in which the response speed with respect to operation is less than that of the first operation processing. The first operation processing is performed by a first operation end, and the second operation processing is performed by a second operation end. In this regard, provided is a safe operation range determination unit for determining a safe operation range of the first operation processing by the first operation end, based on performance of the first operation end. Moving a performance position of the first operation processing by the first operation end to a performance position in which occurrence of an operational abnormality is not estimated, by correcting or changing instructions of the second operation processing when the first operation processing by the first operation end is not within the safe operation range based on the determination by the safe operation range determination unit is performed in units of production materials.

Description

Plant control device, plant control method and program {PLANT CONTROL APPARATUS, PLANT CONTROL METHOD AND PROGRAM}

The present invention relates to a plant control device, a plant control method, and a program.

When performing rolling mill control, which is one of the plant controls, conventionally, fuzzy control or neuro-fuzzy control has been applied at the time of shape control for controlling the bending state of a sheet. Fuzzy control is applied to shape control using a coolant, and neuro-fuzzy control is applied to shape control of a Genjimere rolling mill.

In shape control applying neuro-fuzzy control, as disclosed in Patent Literature 1, a process of obtaining a similarity ratio between a difference between an actual shape pattern detected by a shape detector and a target shape pattern and a preset reference shape pattern is performed, have. And, in the shape control to which neuro-fuzzy control is applied, the control output amount for the manipulation stage is obtained from the obtained similarity ratio by the control rule expressed according to the manipulation amount of the control manipulation stage for the preset reference shape pattern.

The shape control has a plurality of control operation stages, and controls are executed by differences in characteristics of the plurality of control operation stages. The shape is a bent state of the sheet in the sheet width direction, and the control operation stage can change the shape of a specific region in the sheet width direction. For example, the AS-U roll can change the shape near the saddle position to be operated, and the intermediate roll shift can change the shape of the plate end, respectively. When performing shape control, each control operating stage is combined and operated according to the actual shape so that shape deviation is suppressed.

When the rolling mill performs rolling, a coolant (hereinafter referred to as a coolant) is required for lubrication between the material to be rolled and the rolls of the rolling mill and cooling of heat generated by rolling. This coolant serves as an operating stage for shape control, and it is possible to change the shape in the entire region in the plate width direction by adjusting the coolant injection amount in the plate width direction. In the 6-high rolling mill, as shown in Patent Literature 2, shape control is performed by having a coolant spraying amount adjustment mechanism in the sheet width direction and changing the spraying amount using the actual shape. However, in the Genjimere rolling mill, the rolls of the rolling mill are in a state of being submerged in the coolant during rolling, and the coolant takes longer than AS-U or intermediate roll shift until the effect on the shape is revealed. Moreover, since adjustment of the coolant injection amount cannot be performed automatically, the case where it is necessary for an operator to perform, such as operation of a flow control valve, is conceivable.

In addition, depending on the mechanical configuration of the rolling mill, there are cases where it is impossible to adjust the coolant flow rate in the sheet width direction, but even in that case, it is possible to change the actual sheet width direction spraying amount by replacing the coolant injection nozzle. The Genjimere rolling mill has a mechanism for setting a mechanical device in which a plurality of coolant injection nozzles are set in the sheet width direction in order to circulate the coolant to the vicinity of the work rolls. For this reason, the Genjimere rolling mill can change the coolant flow rate in the sheet width direction by replacing them before rolling by preparing a plurality of machines in which coolant injection nozzles having different flow rates are set.

During implementation of rolling in a rolling mill, an operational abnormality in which a material to be rolled breaks may occur. Breakage of the material to be rolled is often caused by the material to be rolled, but in some cases, the material to be rolled is meandering. Meandering of the material to be rolled means that the material to be rolled is shifted to one side of the rolling mill, and rolling is usually performed at the center of the rolling mill.

Such meandering is expected to occur depending on the position of the AS-U roll or intermediate roll shift. In shape control, an AS-U or an intermediate roll shift is operated to maintain the shape of a material to be rolled in a target shape, but as a result, the mechanical state may become a state in which meandering of the material to be rolled is likely to occur.

Conventionally, in normal rolling mills such as 4-high rolling mills and 6-high rolling mills, in addition to mechanical operating units such as a bender and leveling, an operation unit for changing the coolant injection amount in the sheet width direction is used for shape control. Unlike the Genjimere rolling mill, in a normal rolling mill, the rolls of the rolling mill are not immersed in the coolant, and the effect of the coolant on the shape is obtained in a time equal to that of mechanical operation means. Therefore, shape control in a normal rolling mill treats a mechanical operation means and a coolant equally, and uses a coolant as a control operation stage. Even in that case, since the coolant has an effect over the entire sheet width direction, it competes with the mechanical operation unit, and even if the performance shape of the material to be rolled is the same, the performance position of the mechanical operation unit is often different.

In addition, even in a 4-high rolling mill or a 6-high rolling mill, depending on the mechanical configuration of the coolant device, it may be impossible to adjust the coolant flow rate in the sheet width direction during rolling. Even in this case, the operator can change the coolant flow rate in the sheet width direction before rolling starts by manual operation.

Fig. 19 shows a schematic configuration of a control device of a conventional Genjimere rolling mill.

First, in the calculator 901, the difference between the target shape d1 and the shape performance d2 of the rolled material obtained in the rolling mill 990 is taken, and this difference is given to the first shape control unit 902. The first shape control unit 902 controls a mechanical operation unit 904 that is a mechanical operation unit.

The rolling mill 990 executes the mechanical operation process by the mechanical operation stage 904 and the operation process of the coolant injection amount by the coolant operation stage 905 to perform the rolling process of the material to be rolled, The shape result d2 and the rolling result d3 are obtained. In this case, in the mechanical operation process by the mechanical operation stage 904, the shape of the material to be rolled is changed with a relatively high-speed response by controlling the operation amount.

In the case of the structure shown in this FIG. 19, there was a problem that it was difficult to appropriately control the mechanical operation process by the mechanical operation stage 904. For example, when it is in a state close to the upper limit of the operating range of the high-speed response mechanical operating stage 904, it is not possible to stably control the shape performance d2 or the rolling performance d3, and vibration occurs, causing the plant to be controlled. There are cases in which the phosphorus rolling mill 990 cannot operate stably.

As a prior art for appropriately controlling the shape of such a Genjimere rolling mill, as described in Patent Document 3, for example, the relationship between the performance shape deviation of the material to be rolled and the operation amount of the control operation stage using machine learning from performance data. There is a skill to learn and control. In the technology disclosed in this Patent Literature 3, shape control is performed in which a control output is issued according to a shape deviation and a control operation stage is operated.

Japanese Patent No. 2804161 Japanese Patent No. 2515028 Japanese Unexamined Patent Publication No. 2018-005544

The technology disclosed in Patent Literature 3 performs shape control by learning a control operation method for shape deviation, and does not consider the position of the control operation end. In particular, in conventional shape control, since control is based on shape deviation, there has been a problem that it is not possible to limit the performance positions of mechanical operating stages leading to operational abnormalities such as plate breakage.

In addition, in the description so far, the problem of shape control of the Genjimere rolling mill has been explained. However, in various plant control devices, when a control operation with a fast response and a control operation with a slow response are simultaneously performed, both control operations are performed. There is a similar problem when doing both properly at the same time.

An object of the present invention is a plant control device capable of improving control effects and operational efficiency by predicting the occurrence of an operational abnormality in a controlled plant and appropriately operating a control operating stage so that an operational abnormality does not occur. It is to provide a control method and program.

In order to solve the said subject, the structure described in the claim is employ|adopted, for example.

The present application includes a plurality of means for solving the above problems. As an example thereof, a plant control device includes a first operation process in which a response speed to an operation is a predetermined response speed for a plant to be controlled; A plant control device that performs a second operation process with a slower response speed to the operation than the operation process,

an operation stage control unit that obtains a target state quantity of the plant to be controlled and gives instructions for a first operation process;

a second operation stage setting unit for instructing a second operation process;

A first operating stage for executing the first operation process of the plant to be controlled in response to instructions from the operating stage control unit;

A second operating stage for executing a second operation process of the plant to be controlled in response to instructions from the operating stage setting unit;

a safe operation range determination unit for determining a safe operation range of a first operation process by the first operation stage based on the performance of the first operation stage;

A learning unit that learns the position of the first operating stage and the target state quantity by the first operation process of the plant to be controlled in units of production materials from operation performance in the plant to be controlled;

An operating end position predicting unit for predicting a position of a first operating end during production of a tea production product using a learning result in the learning unit,

The operating stage setting unit determines the operation range of the position of the first operating stage in the difference product in the operating stage position prediction unit and the judgment in the safety operation range determining unit, so that the first operating stage is operated during operation of the production material. The position of the second control end is determined so that the position does not deviate from the safe operation range.

ADVANTAGE OF THE INVENTION According to this invention, the operating state in a control object plant can be controlled appropriately, and the performance position of the operation stage which becomes an operation abnormality can be suppressed. As a result, improvement in the control accuracy of the control target plant, improvement in operational efficiency, and suppression of occurrence of operational abnormalities can be expected.

Subjects, configurations, and effects other than those described above will be clarified by the description of the embodiments below.

1 is a block diagram showing a configuration example of a plant control device according to an embodiment of the present invention.
Fig. 2 is a block diagram showing a configuration example when a plant control device according to an embodiment of the present invention is applied to a rolling mill.
3 is a configuration diagram showing an example of a Genjimere rolling mill.
Fig. 4 is a configuration diagram showing an example of rolling equipment of a single stand rolling mill.
5 is a diagram showing an outline of a mechanical operating stage according to an embodiment of the present invention.
6 is a block diagram showing an example of a configuration of a safe operation range determination unit according to an embodiment of the present invention.
7 is a diagram showing an example of the configuration of a neural network of a mechanical manipulation end safety manipulation range determination unit according to an embodiment of the present invention.
8 is a diagram showing an example of the configuration of a neural network management table according to an example of an embodiment of the present invention.
9 is a diagram showing a configuration example of a learning database according to an embodiment of the present invention.
Fig. 10 is a block diagram showing an example of a configuration of a mechanical manipulation end position prediction unit according to an embodiment of the present invention.
11 is a diagram showing an example of the configuration of a neural network of a mechanical manipulation end position prediction unit according to an embodiment of the present invention.
12 is a diagram showing a configuration example of a neural network management table according to an example of an embodiment of the present invention.
13 is a diagram showing an example of a configuration of a learning database according to an example of an embodiment of the present invention.
14 is a diagram showing an example of a configuration of a coolant operation stage setting unit according to an embodiment of the present invention.
15 is a diagram showing an example of a predicted value of a mechanical operating stage according to an embodiment of the present invention.
16 is a diagram showing an example of determination of a mechanical operating stage according to an embodiment of the present invention.
17 is a diagram showing an example of a configuration of a coolant operating stage control output calculating unit and a coolant operating stage control output selecting unit according to an embodiment of the present invention.
Fig. 18 is a block diagram showing an example of a hardware configuration in the case where a plant control device according to an embodiment of the present invention is configured with a computer.
Fig. 19 is a block diagram showing a configuration example of a conventional rolling mill control device.

Hereinafter, a plant control device according to an embodiment of the present invention (hereinafter referred to as “this example”) will be described with reference to FIGS. 1 to 18 .

[Overall configuration of plant control device]

1 shows an example of the overall configuration of the plant control device of this example.

The plant control device shown in FIG. 1 controls a control target plant 190, and as control of the control target plant 190, operation processing by a high-speed operation stage (first operation stage) 103 and low-speed operation An operation process by the operating end (second operating end) 104 is executed.

The plant control device of this example acquires the first state quantity target d11, and acquires the difference from the first state quantity d12 with the calculator 101. The first state quantity d12 is obtained from the control target plant 190 as a result of operation processing by the high-speed operation stage 103 and the low-speed operation stage 104 . Further, in the high-speed operation stage 103 and the low-speed operation stage 104, the second state quantity d13 is obtained as a result of the operation processing of the other operation stages applied to the control target plant 190.

The control unit 110 of the plant control device has a high-speed operation stage controller 111 and a low-speed operation stage setting unit 112 . The high-speed operation stage controller 111 controls the operation process of the high-speed operation stage 103 . The low-speed operation stage setting unit 112 controls the operation process of the low-speed operation stage 104 .

The difference between the first state amount target d11 obtained by the calculator 101 and the first state amount d12 is supplied to the high-speed operation stage controller 111, and control is performed to bring the first state amount d12 closer to the difference. .

The control output of the high-speed operation stage controller 111 is directly supplied to the high-speed operation stage 103 to control the operation process of the high-speed operation stage 103 .

In addition, the plant control device of this example includes a safe operation range determining unit 105 . The safe operation range determination unit 105 acquires the operation results of the high-speed operation stage 103, determines whether or not the acquired operation results have a margin in the safe operation range, and determines the safety operation range d14 as a result of the determination. The data is supplied to the low-speed operation stage setting unit 112 .

In addition, the safe operation range determination unit 105 acquires the first state quantity d12 and the second state quantity d13 of the control target plant 190 . Then, the safe operation range determination unit 105 refers to the first state amount d12 and the second state amount d13 to determine whether the current operation performance of the high-speed operation stage 103 has margin in the safe operation range. decide whether

Here, the safe operation range determination unit 105 detects the occurrence of an operation abnormality, learns the first state quantity d12 and the second state quantity d13 at that time, and the actual position of the high-speed operation stage 103, A safe manipulation range d14, which is an operable range, is determined so that abnormal operation does not occur in the high-speed manipulation stage 103. Since operational abnormality is not a phenomenon that occurs frequently, it is preferable that the safety operation range determining unit 105 continuously collects performance data and performs learning using machine learning or the like to obtain the safe operation range d14. do.

Then, the safe operation range determination unit 105 supplies data of the determined safe operation range d14 to the low-speed operation stage setting unit 112 .

The operating end position prediction unit 102 learns the operation range of the high-speed operating end 103 in units of production materials from past operation performance data, and based on the result, the high-speed manipulation end 103 at the time of producing car production materials predict the amount of motion of

The low-speed operation stage setting unit 112 determines whether the high-speed operation stage 103 is within a safe operation range ( d14) Determines the set value of the low-speed operation stage 104 to be within, and sets the low-speed operation stage 104. The setting of the low-speed operation stage 104 is performed mechanically according to a command from the low-speed operation stage setting unit 112 if automatic setting is possible, and when automatic setting is not possible, the setting of the low-speed operation stage 104 by the operator. A value is notified by means such as being displayed on a screen, and the operator sets it manually.

According to the plant control device having the structure shown in FIG. 1, it becomes possible to efficiently operate the high-speed operating stage 103 within a range in which operational abnormality does not occur, and in addition to improvement in operational efficiency, improvement in control precision can be expected. have.

[Overall configuration when applied to the control device of Genjimere rolling mill]

Next, the overall configuration in the case of applying the plant control device of this example to a Genjimere rolling mill will be described.

Fig. 2 shows the configuration of the plant control device of this example when applied to a Genjimere rolling mill.

The plant control apparatus shown in FIG. 2 acquires the target shape d21 of the material to be rolled, and acquires the difference with the shape result d23 after rolling with the calculator 201.

The plant control device of this example executes the operation process by the mechanical operation stage 203 and the setting operation process by the coolant operation stage 204 as control of the rolling mill 301 . The operation process by the mechanical operation stage 203 mechanically changes the roll gap or the like for performing the rolling process, and the response shown in the shape result d23 of the rolled material becomes high-speed. On the other hand, the setting operation processing by the coolant operation stage 204 is to change the coolant injection amount, and the response shown in the rolling performance d24 of the material to be rolled is slower than the operation by the mechanical operation stage 203. .

The control unit 220 of the plant control device includes a first shape controller 211 that controls the operation of the mechanical operating stage 203 and a coolant operating stage setting unit that sets the coolant operating stage 204 ( 212).

The 1st shape control part 211 performs feedback control so that a to-be-rolled material may become the target shape d21. The target shape d21 is set in advance according to the characteristics of the material to be rolled and the like.

The coolant operation stage setting unit 212 sets the coolant operation stage 204 . The meaning of setting here means that the flow rate distribution in the sheet width direction of the coolant operating stage 204 is performed in advance before performing the manufacturing operation for each production material unit, which is the production unit of the rolled material (product material) as a product manufactured by the rolling mill. means to determine During the manufacturing operation of the material to be rolled, the flow distribution in the sheet width direction is not changed. A manufacturing unit of a material to be rolled means one rolling operation, and is also referred to as a production material unit. More specifically, in the Genjimere rolling mill, the same rolled material is subjected to a plurality of rolling operations with different exit plate thicknesses (sequentially decreasing the exit plate thickness) while changing the rolling direction, so as to achieve the target product plate thickness. The operation to obtain is performed. The entirety of one of these multiple times of rolling operation is the above-mentioned manufacturing unit or production material unit.

The control output of the first shape controller 211 is directly supplied to the mechanical manipulation stage 203, and controls the operation process of the mechanical manipulation stage 203.

The coolant operation stage setting unit 212 sets the coolant operation stage 204, but when direct operation from the control unit 220 of the plant control device is impossible, the operator has the coolant operation stage 204 ) is displayed, and the operator performs a setting operation. The case where direct operation from the control unit 220 here is impossible means, for example, when the control unit 220 is realized by an electronic computer, the electronic computer directly operates the coolant operating stage 204 via the input/output unit. say in case you can't.

In addition, the plant control device of this example includes a mechanical operation end safe operation range determination unit 205 .

The mechanical manipulation end safety operation range determining unit 205 acquires the mechanical manipulation end position performance d22, which is the manipulation result of the mechanical manipulation end 203, and the obtained mechanical manipulation end position performance d22 is margined in the safe manipulation range. determine whether you have Then, the mechanical manipulation end safe operation range determination unit 205 supplies the data of the mechanical manipulation end safe operation range d25 as the determination result to the coolant operation end setting unit 212 .

Further, the mechanical operation end safe operation range determination unit 205 acquires the shape performance d23 and the rolling performance d24 of the rolling mill 301 . Then, the mechanical operation end safety operation range determining unit 205 refers to the shape results d23 and the rolling results d24, and determines whether or not the acquired mechanical operation end position results d22 have a margin in the safe operation range. judge

The mechanical operating end safety operation range determination unit 205 detects the occurrence of an operation abnormality and serves as a learning unit for learning the shape record d23 and rolling record d24 and the mechanical operating end position record d22 at that time. do the processing Based on this learning, the mechanical manipulation end safe manipulation range determination unit 205 determines the mechanical manipulation end safe manipulation range d25, which is an operable range, so that operational abnormality does not occur in the mechanical manipulation end 203. Here, the mechanical manipulation end safe operation range determining unit 205 continuously collects performance data and performs learning using machine learning or the like to obtain the mechanical manipulation end safe manipulation range d25.

Then, the mechanical manipulation stage safe operation range determination unit 205 supplies the determined mechanical manipulation stage safe operation range d25 data to the coolant operation stage setting unit 212 .

Further, a detailed configuration in which the mechanical manipulation end safety manipulation range determination unit 205 performs learning of an operation abnormality will be described later with reference to FIG. 6 .

The mechanical operation end position prediction unit 210 calculates the mechanical operation end position variation amount ( d26). The amount of change in the position of the mechanical operating end d26 is obtained by classifying the manufacturing units according to product specifications and learning within which range the performance value of the mechanical operating end 203 varied for each manufacturing unit.

Here, classification according to product specifications refers to setting a standard in advance so that the performance values of the mechanical operation stage 203 are almost the same according to the steel type, sheet width, sheet thickness, etc. of the material to be rolled, and classifying according to the standard. .

If the classification is the same product specification, the movement range of the actual value of the mechanical operating stage 203 becomes the same, so that the product specification of the manufacturing unit starting the manufacturing work is found before the manufacturing work for each manufacturing unit of the material to be rolled is started. do. For this reason, the mechanical manipulation end position prediction unit 210 can obtain the learning result of the mechanical manipulation end position variation amount d26 of the same classification of product specifications. At the same time, the mechanical operating end position prediction unit 210 can also predict the position of the coolant operating end 204 in the case of the mechanical manipulation end position variation amount d26, and calculates the coolant operating end position prediction d28. It is output to the coolant operation stage setting unit 212.

The coolant operating stage setting unit 212 receives, from the mechanical operating stage position predicting unit 210, the amount of change in the position of the mechanical operating stage d26 and the prediction of the position of the coolant operating stage at that time d28 in each manufacturing unit. Then, the coolant operation end setter 212 compares the mechanical manipulation end safe manipulation range d25 received from the mechanical manipulation end safe manipulation range determining part 205 with the mechanical manipulation end position change amount d26. In this way, the coolant operating stage setting unit 212 determines whether or not the position performance of the mechanical operating stage 203 is within the mechanical operating stage safe operation range d25 when the manufacturing operation of the car manufacturing unit is performed. .

As a result of this determination, when it is within the mechanical operating end safe operation range d25, the coolant operating end setting unit 212 gives a coolant setting instruction in which the coolant operating end position prediction d28 is the coolant flow rate. It is output to the coolant operation stage 204.

In addition, as a result of the determination, when it is out of the mechanical manipulation end safe operation range d25, the coolant manipulation end setting unit 212 sets the position performance of the mechanical manipulation end 203 to the mechanical manipulation end safe manipulation range d25. A coolant setting instruction corrected for the coolant operating stage position prediction d28 is output to the coolant operating stage 204 at a coolant flow rate in the plate width direction that becomes

As an output instruction to the coolant operating stage 204 in this case, in addition to direct output from the coolant operating stage setting unit 212 to the coolant operating stage 204, the coolant operating stage 204 to the operator This includes the case where the set value of is displayed and the operator performs work to change the actual coolant flow rate.

[Configuration of Genjimere Rolling Mill]

Here, a configuration example of the Genjimere rolling mill will be described.

Fig. 3 shows a schematic configuration in the case of performing shape control in the Genjimere rolling mill.

In the Genzimere rolling mill, the actual shape of the rolled material is detected by the shape detector 14.

The actual shape detected by the shape detector 14 is assigned to one of the reference shape patterns preset in the pattern recognition unit 12 after the pattern recognition preprocessing is performed in the shape detection preprocessing unit 11 of the control device 10. The nearest edge is calculated. Then, based on the calculated reference shape pattern, the control operation unit 13 determines the operation stage and the operation amount to be operated, and the processing of controlling the Genjimere rolling mill with the operation stage and operation amount to be operated is executed.

Fig. 4 shows an example of rolling equipment of a single stand rolling mill. The Genjimere rolling mill is a type of single stand rolling mill.

The rolling equipment shown in FIG. 4 includes a rolling mill 301, an entry tension reel (hereinafter referred to as “TR”) 302, and an exit TR 303, and a material to be rolled drawn from the entry TR 302 ( 300) passes through the rolling mill 301 and is wound by the output TR 303.

The rolling mill 301 rolls the material to be rolled 300 . Rolling here is a process of thinning the plate thickness of the material to be rolled 300 to a predetermined plate thickness.

The rolling mill 301 is provided with a mill speed control unit 304 for adjusting the rolling speed and a roll gap control unit 307 for adjusting the roll gap of the rolling mill 301 . In addition, in the entry TR 302 and the exit TR 303, an entry TR control section 305 and an exit TR control section 306 are provided to adjust the tension generated by each.

In the rolling process, by adjusting the gap between the upper and lower rolls of the rolling mill 301 by the roll gap control unit 307, pressure is applied to crush the material to be rolled 300, and the material to be rolled 300 is output by the mill speed control unit 304. It is carried out by sending it to the side. At this time, processing for applying tension to the material to be rolled 300 using the entry TR 302 and the exit TR 303 is also performed on the entry and exit sides of the rolling mill 301 .

What is important in the rolling operation is the sheet thickness of the material to be rolled 300 as a product (exit sheet thickness of the rolling mill), and the roll gap, the entry tension, and the exit tension are determined in advance so that the material to be rolled 300 has a predetermined sheet thickness. is set

The input tension current conversion unit 315 uses the input tension set in the input tension setting unit 311 to obtain the current required to obtain the set input tension, and passes it to the input TR 302 through the input TR control unit 305. By applying, the input side tension is obtained.

Similarly, the output tension current conversion unit 316 uses the output tension set by the output tension setting unit 312 to obtain a current required to obtain the set output tension, and through the output TR control unit 306, the output TR 303 ) to obtain the output tension.

The roll gap set by the roll gap setting unit 319 is given to the roll gap control unit 307, and the roll gap control unit 307 sets the roll gap.

The rolling speed setting unit 310 determines the speed of the rolling mill 301 according to the instructions of the rolling mill operator, and the mill speed control unit 304 sets the speed of the rolling mill 301 .

An entry tension meter 308 and an exit tension meter 309 are installed on the entry and exit sides of the rolling mill 301, and the entry tension control section 313 and the exit tension control section ( 314) executes the control. Further, on the exit side of the rolling mill 301, an exit thickness gauge 317 is installed, and an exit thickness controller 318 controls so that the actual sheet thickness measured there coincides with the set sheet thickness.

In addition to the above configuration, as shown in FIG. 3 previously described, a shape detector 14 for detecting the shape of the material to be rolled is installed on the exit side of the rolling mill, and the shape is controlled so that the detected shape matches a preset target shape. is executed

As already explained, the shape is the degree of curvature of the metal plate as a material to be rolled. Therefore, a target shape, which is a target shape, is set in advance from the workability in the process under the rolling mill and the efficiency of the rolling operation in the rolling mill. In general, since tension is applied to the material to be rolled, if there is a crack or other flaw at the edge of the plate, cracks are generated therefrom, and the material to be rolled is often divided in the sheet width direction (plate breakage). . For this reason, in many cases, the plate end is in a bent state so that the tension is not concentrated.

The bending of the material to be rolled does not materialize because tension is actually applied to the material to be rolled, and the tension distribution changes in the sheet width direction, although there is no bending in appearance.

Here, the shape detector 14 shown in FIG. 3 measures the tension distribution in the sheet width direction to estimate the curvature of the sheet and detect it as a shape result.

[Configuration and processing of shape control mechanical control unit]

Fig. 5(a) shows a configuration when an operation process is performed by the mechanical operation end 203 of the Genjimere rolling mill. In FIG. 5, the cross section of the plate width direction of the material to be rolled 300 is shown, and only the upper structure of the material to be rolled 300 is shown, and the lower structure is omitted.

5(b) and (c) show operation waveforms when the shape of the material to be rolled 300 is changed, respectively.

As shown in Fig. 5 (a), the Genjimere rolling mill has a work roll 401, a first intermediate roll 402, a second intermediate roll 403 with a material to be rolled 300 therebetween, AS-U roll 404.

The first intermediate roll 402 is provided with a taper on the opposite side in the vertical direction, and can affect the shape of the plate end of the material to be rolled 300 by shifting in the plate width direction.

The AS-U roll 404 has a structure in which a saddle 406 is inserted between a plurality of split rolls 405, and by changing the position of the saddle 406 (position in the vertical direction in FIG. 5), the AS-U roll 404 The warp of the U roll 404 can be varied in the sheet width direction.

For example, as shown in Fig. 5(b), when the central saddle 406 is lowered, the shape of the central portion of the material to be rolled 300 may be affected.

Here, the operation waveforms shown in the lowermost stages of (b) and (c) of FIG. 5 are changes in the sheet thickness distribution of the material to be rolled 300 when the saddle 406 or the first intermediate roll 402 is operated to shift. indicates The shape change is inverse to the plate thickness distribution.

The shape is the distribution of the degree of waviness in the sheet width direction, and a large waviness indicates that the material to be rolled 300 is elongated. This is because "the exit sheet thickness becomes thinner", "the elongation of the rolled material in the thinned portion is large", and "the shape of the rolled material becomes larger" are equal.

Regarding operational abnormalities, plate breakage of the material to be rolled 300 is a major problem. When a plate fracture occurs, the material to be rolled 300 after fracture damages the work roll 401 and the first intermediate roll 402 of the rolling mill. In some cases, the second intermediate roll 403 and the AS-U roll 404 are also damaged due to plate breakage. When such a breakage occurs, replacement of these rolls is required, and time is required for the removal treatment of the material to be rolled 300 remaining in the rolling mill, resulting in extremely low operating efficiency.

Since the AS-U roll 404 presses the split roll 405 against the second intermediate roll 403 in the form of pressing it with the saddle 406, depending on the position of the saddle 406, the split roll 405 ) and the second intermediate roll 403 do not come into contact. In such a state, the force applied to the material to be rolled 300 from the work roll 401 of that portion is rapidly reduced, the material to be rolled 300 is not stretched, and the tension applied to the material to be rolled 300 of that portion increase

When this state occurs at the end of the plate of the material to be rolled 300, the plate is broken from the determination unit. In addition, since the tension applied to both ends of the material to be rolled 300 in the sheet width direction changes, a phenomenon occurs in which the center of the material to be rolled 300 in the sheet width direction is shifted from the center of the sheet width direction of the rolling mill, and mechanical equipment before and after the rolling mill is shifted. In some cases, the plate may be broken by colliding with the plate. In this way, depending on the actual position of the mechanical operating stage 203, there is a case where an operation abnormality may occur.

The actual position where operational abnormality occurs is not obtained by calculation from the machine configuration, but the sheet thickness distribution in the width direction of the rolled material 300, the sheet thickness at the entry and exit side, the rolling state such as tension and rolling load, and other shape control Since the positional relationship with the mechanical operating end also changes, it is difficult to predict in advance.

Therefore, in this example, the mechanical operation end safety operation range determination unit 205 stores these conditions when rolling abnormality occurs as performance data and compares it with the performance data during normal times, so that rolling abnormality is likely to occur. The actual position of the shape control mechanical operating stage is obtained.

The mechanical operation end safe operation range determination unit 205 of the present example determines the mechanical operation end safe operation range using machine learning. For the performance data when machine learning is performed, the entry/exit sheet thickness, rolling conditions such as tension and rolling load, and the performance position of the mechanical operating stage 203 are used, and rolling abnormal occurrence information is used for the teacher data.

As the rolling abnormal occurrence information, information on sheet breakage and emergency stop of the rolling mill is used. Plate breakage can be determined by reducing the entry/exit tension, and for emergency stop, information on an operation switch operated by an operator when an abnormality occurs in the rolling state and the operation is stopped is used. The information on the operation switch can be detected by a calculator constituting the control device that controls the rolling mill, and can be used as one of performance information. The mechanical operation end safety operation range determining unit 205 creates a neural network (N.N.) that determines whether or not an operation abnormality has occurred using these performance data and teacher data.

[Configuration of the safety operation range determination unit of the mechanical control unit and configuration of the neural network]

Fig. 6 shows a configuration in the case where the mechanical manipulation end safe operation range determination unit 205 is realized by machine learning.

7 shows the configuration of the neural network 502 included in the mechanical manipulation end safety manipulation range determining unit 205 .

As shown in Fig. 7, the neural network 502 obtains the rolling results d24 and the mechanical operation end position results d22 from the input data creation unit 501 at the input stage 502a, and from the output stage 502b. An operation abnormality determination value d32 is output. The operation abnormality determination value d32 is information of plate breakage and emergency stop information, which are rolling abnormal occurrence information. The neural network 502 learns from combinations of these input data and output data.

Describing the mechanical manipulation end safety operation range determination unit 205 shown in FIG. 6 , the input data creation unit 501 collects the mechanical manipulation end position results d22 and shape results d23. In addition, the teacher data creation unit 505 collects the operation abnormality determination value d32 determined by the operation abnormality determination unit 506 . Data collection by these input data creation unit 501 and teacher data creation unit 505 is performed at regular time intervals under the control of the neural network learning control unit 503, and one set of learning data is obtained for each operation cycle. The obtained learning data is sequentially stored in the learning data database 511 .

The operation abnormality determination unit 506 determines whether or not there is a plate breakage and an emergency stop of the rolling mill, which is an operation abnormality, from the rolling results d24. The operation abnormality determination value d32 that is the determination result is information on plate breakage and emergency stop.

By the way, a rolling mill rolls various to-be-rolled materials 300 according to specifications, and obtains a product. For this reason, the rolling mill sets the material to be rolled 300 according to the specifications, the specifications of the work roll 401 (diameter distribution in the sheet width direction), the taper specification of the first intermediate roll 402, and the AS-U roll, which are mechanical configurations. It is common to respond by changing the combination of the split rolls 405 of (404). Also, the material to be rolled 300 is not uniform in width or material. Therefore, efficient learning becomes possible when the neural network 502 is divided according to the mechanical configuration or the specifications of the material to be rolled 300.

For this reason, the mechanical operating end safety operation range determination unit 205 of the present example has a plurality of types of neural networks 502 and uses the control rule database 512 and the neural network selection unit 504 so that switching is possible. provide

8 shows an example of the configuration of the control rule database 512.

In the control rule database 512, as shown in FIG. 8(a), a plurality of neural networks learned using learning data including a combination of input data and teacher data are stored.

Then, the neural network learning control unit 503 designates a neural network No. that requires learning. The neural network selection unit 504 receives the designation of the neural network No. that needs to be learned by the neural network learning control unit 503, retrieves the neural network from the control rule database 512, and sets it in the neural network 502.

The neural network selector 504 extracts the neural network of the corresponding neural network No. from the control rule database 512 in accordance with the rolling conditions and machine configuration from the current rolling performance d24, and obtains a mechanical neural network as the control neural network d33. It is set in the operating end position suppression control unit 202.

8(b) shows the configuration of the neural network management table stored in the control rule database 512. The management table is classified according to (B1) sheet width, (B2) steel type, and machine configuration (A). (B1) As the board width, for example, four divisions of a 3-foot width, a meter width, a 4-foot width, and a 5-foot width are used. (B2) As the steel type, about 10 types of steel types (1) to (10) are used. About (A), it is divided into (A1) and (A2) according to the length of the taper part which is a taper specification of the 1st intermediate|middle roll 402, for example.

The above table classification is an example, and it is necessary to set it in a timely manner according to the type of rolling equipment or the rolled material to be produced.

The mechanical operation end safe operation range determination unit 205 uses these neural networks separately according to rolling conditions and machine configuration.

The neural network learning control unit 503 assigns learning data, which is a combination of input data and teacher data shown in FIG. 8(a), to the corresponding neural network No. and It is connected and stored in the learning data database 511.

9 shows an example of learning data stored in the learning data database 511 .

As shown in Fig. 9, the learning data database 511 stores corresponding learning data for each neural network No..

The neural network learning control unit 503 performs retrieval from the learning data database 511 from a management table of input data and teacher data corresponding to the neural network by the input data creation unit 501 and the teacher data creation unit 505 direct to The neural network 502 performs learning using these. Various learning methods for neural networks have been proposed in the past, and any learning method may be used.

Machine learning requires a large amount of sets of learning data, and when a certain amount (for example, 10000 trillion) is stored in the learning data database 511, the neural network 502 executes learning.

When learning of the neural network 502 is completed, the neural network learning control unit 503 writes back the neural network 502 as the learning result to the position of the corresponding neural network No. in the control rule database 512, so that the learning is completed. It is done.

The learned neural network 502 outputs an operation abnormality determination value by inputting the results of rolling (d24) and the results of mechanical operation end positions (d22). For this reason, the neural network 502 can predict whether or not an operation abnormality has occurred by giving the predicted future shape results d23 and the mechanical operation end position results d22, and sets the mechanical manipulation end safe operation range d25. it is possible to explore

[Configuration of mechanical control end position prediction unit and neural network configuration]

Fig. 10 shows a configuration in the case where the mechanical operating end position prediction unit 210 is realized by machine learning.

11 shows the configuration of the neural network 702 included in the position predicting unit 210 of the mechanical manipulation end.

As shown in Fig. 11, the neural network 702 obtains the rolling performance d24 from the input data creation unit 701 at the input end 702a, and the coolant operation end position prediction d28 from the output end 702b, and The amount of change in the position of the mechanical control end (d26) is output. The neural network 702 learns from the combination of these input data and output data.

If the mechanical operation end position prediction part 710 shown in FIG. 10 is demonstrated, the input data creation part 701 collects rolling performance d24. In addition, the teacher data creation unit 705 collects the mechanical operation end position variation d26 and the coolant operation end position prediction d28 for each manufacturing unit created by the production unit data creation unit 706 .

The production unit data preparation unit 706 determines the production unit of the material to be rolled from the rolling speed of the rolling performance d24. As already explained, since the production unit of the material to be rolled means one rolling operation, one rolling operation (manufacture of the production unit) of the material to be rolled is in a state where the rolling speed is not 0 from the rolling speed = 0, and that It can be judged that it is until post-rolling speed = 0.

Therefore, the manufacturing unit data creation unit 706 collects the maximum and minimum values of the mechanical operation end position results d22 and the coolant operation end position results d27 until the rolling speed reaches 0, and mechanical operation The stage position variation amount d26 and the coolant operating end position prediction d28 are created.

When one rolling operation (manufacturing unit) of the material to be rolled is completed and creation of the mechanical manipulation end position variation d26 and the coolant manipulation end position prediction d28 is completed, the learning data necessary for learning the neural network 702 Since one is created, the production unit data creation unit 706 notifies the neural network learning control unit 703. The neural network learning control unit 703 receives a notification from the production unit data creation unit 706, and collects data in the input data creation unit 701 and the teacher data creation unit 705. The obtained learning data is sequentially stored in the learning data database 711 .

By the way, a rolling mill rolls various to-be-rolled materials 300 according to specifications, and obtains a product. For this reason, the rolling mill sets the material to be rolled 300 according to the specifications, the specifications of the work roll 401 (diameter distribution in the sheet width direction), the taper specification of the first intermediate roll 402, and the AS-U roll, which are mechanical configurations. It is common to respond by changing the combination of the split rolls 405 of (404). Also, the sheet width and material of the material to be rolled 300 are not uniform. Therefore, efficient learning becomes possible when the neural network 702 is divided according to the mechanical configuration or the specification of the material to be rolled 300.

For this reason, the mechanical operating end position predicting unit 201 of this example includes a control rule database 712 and a neural network selection unit 704 so that a plurality of types of neural networks 702 can be used and switched. .

Fig. 12 shows an example of the configuration of the control rule database 712.

In the control rule database 712, as shown in Fig. 8(a), a plurality of neural networks learned using learning data including a combination of input data and teacher data are stored.

Then, the neural network learning control unit 703 designates a neural network No. that requires learning. The neural network selection unit 704 receives the designation of the neural network No. that needs to be learned by the neural network learning control unit 703, retrieves the neural network from the control rule database 712, and sets it in the neural network 702.

The neural network selector 704 extracts the neural network of the corresponding neural network No. from the control rule database 712 according to the rolling conditions and machine configuration from the current rolling performance d24, and sets it as the neural network for control d33. .

The rolling performance d24 from the input data creation unit 701 is input to the control neural network d33, and the mechanical operation end position variation d26 and the coolant operation end position are input to the coolant operation end setting part 212 The prediction (d28) is output. At this time, since the rolling result d24 includes manufacturing information of the next manufacturing unit of the rolled material, such as set values such as sheet thickness, steel type, tension, load, etc., based on these manufacturing information, the amount of change in the position of the mechanical operating end ( d26) and the estimation of the position of the coolant control end (d28) are estimated.

Then, the amount of change in the position of the mechanical manipulation end d26 and the prediction of the position of the coolant manipulation end d28 obtained by the control neural network d33 are output to the coolant manipulation end setting unit 212 .

12(b) shows the configuration of the neural network management table stored in the control rule database 712. The management table is classified according to (B1) sheet width, (B2) steel type, and machine configuration (A). (B1) As the board width, for example, four divisions of a 3-foot width, a meter width, a 4-foot width, and a 5-foot width are used. (B2) As the steel type, about 10 types of steel type (1) to steel type (10) are used. About (A), it is divided into (A1) and (A2) according to the length of the taper part which is a taper specification of the 1st intermediate|middle roll 402, for example.

The above table classification is an example, and it is necessary to set it in a timely manner according to the type of rolling equipment or the rolled material to be produced.

The mechanical manipulation end position predictor 210 uses these neural networks separately according to rolling conditions and machine configuration.

The neural network learning control unit 703 assigns learning data, which is a combination of input data and teacher data shown in FIG. 12(a), to the corresponding neural network No. and It is connected and stored in the learning data database 711.

13 shows an example of learning data stored in the learning data database 711 .

As shown in Fig. 13, the learning data database 711 stores corresponding learning data for each neural network No.

The neural network learning control unit 703 performs retrieval from the learning data database 711 from the management table of input data and teacher data corresponding to the neural network by the input data creation unit 701 and the teacher data creation unit 705 direct to The neural network 702 performs learning using these. Various learning methods for neural networks have been proposed in the past, and any learning method may be used.

Machine learning requires a large amount of sets of learning data, and when a certain amount (for example, 10000 trillion) is stored in the learning data database 711, the neural network 702 executes learning.

When the learning of the neural network 702 is completed, the neural network learning control unit 703 writes back the neural network 702 as the learning result to the position of the corresponding neural network No. in the control rule database 712, and the learning is completed.

When the rolling performance d24 is input from the input data creation unit 701, the learning completed neural network 702 provides the neural network selection unit 704 with the mechanical operation end position variation d26 and the coolant operation end position prediction ( d28) is output. For this reason, the neural network 702, in the case of manufacturing the same or similar manufacturing unit as the next manufacturing unit, the variation range of the position of the mechanical operating end and the distribution of the coolant injection amount in the sheet width direction at that time are recorded in the past. can be predicted from the performance data of

[Configuration of coolant control stage setting unit]

14 shows the configuration of the coolant operation stage setting unit 212 .

The coolant operation end setting unit 212 includes a mechanical operation end position abnormality area determination unit 610 and a mechanical operation end position abnormal position suppression control unit 620 .

The mechanical manipulation end position abnormality area determination unit 610 uses the neural network 502 described in FIG. 7 to estimate the mechanical manipulation end 203 where the occurrence of an operational abnormality is predicted. The neural network 502 used here is the control neural network d33 received from the mechanical manipulation end safety manipulation range determination unit 205 (FIG. 6).

The mechanical operation end position abnormality suppression control unit 620 creates a setting command for the coolant operation end 204 based on the determination result in the mechanical operation end position abnormality area determining unit 610 .

The mechanical operating end position prediction unit 210 (FIG. 2) predicts how much the actual position of the mechanical operating end 203 will fluctuate during the rolling operation of the material to be rolled, which is a manufacturing unit to be manufactured next, and It is transmitted to the coolant operation stage setting unit 212 as the position change amount d26.

Here, since the amount of change in position of the mechanical operating end d26 is the difference between the maximum value and the minimum value of each mechanical operating stage 203 within the production unit, the content is as shown in FIG. 15 described below.

A neural network capable of determining whether or not a combination of various positions within each motion range of the mechanical manipulation end becomes an operational anomaly or not, the operational anomaly learned by the mechanical manipulation end safety range determination unit 205 by dividing the manipulation range into several parts. It can be known by inputting in (d33) and determining the possibility of an operation abnormality. However, since there are a plurality of mechanical operation stages 203 (seven types in the case of the present embodiment), the number of cases requiring judgment is enormous, which is not practical.

It is considered that the actual position of the mechanical operating end that causes the operation abnormality is near the upper or lower limit of the operation of the mechanical operating end 203, and is not located in the center of the operating range. Therefore, the mechanical manipulation end position predictor 210 simply determines the safe manipulation range by taking a combination of the maximum value and the minimum value of each mechanical manipulation end 203 and inputting the result to the control neural network d33 (see Fig. 10). it becomes possible to carry out

15 shows an example of the mechanical manipulation end position variation amount. In Fig. 15, the horizontal axis represents the type of the mechanical manipulation stage 203, and the vertical axis represents the predicted position value.

In the example of FIG. 15, when the mechanical manipulation stage 203 is of n types (n is an integer), the maximum and minimum values of the mechanical manipulation stage position variation d26 are:

POSMAX(k), POSMIN(k), k=1, 2, … , n

to be.

For example, the operating range of the manufacturing unit of the mechanical operating stage 3 shown in FIG. 15 is the range represented by the maximum value POSMAX(3) and the minimum value POSMIN(3).

Here, the number of n types of mechanical operation stages 203 means, for example, the sum of the number of saddles 406 of the AS-U roll 404 and the number of shiftable first intermediate rolls 402 in the sheet width direction. this applies For example, in the case of the example shown in FIG. 5, n=7 because the number of saddles is 5 and the first intermediate roll is two up and down. In the following description, when indicated by the mechanical operation stage 203 (k), each of n types (1 to n) of the mechanical operation stage 203 is indicated.

As a result, 2 ⁿ kinds of estimated position results of each mechanical operation stage 203 can be created. For example, in the case of n = 7, 128 estimated position results can be created. These estimated position results are sequentially output to the input data creation unit 612 (see Fig. 7).

The input data creation unit 612 creates input data to the neural network 502 from the rolling results d24 and the estimated position d31 and outputs it to the neural network 502 .

The neural network 502 outputs the operation abnormal determination value d32 shown in FIG. The operation abnormality determination value d32 is the degree to which plate breakage and emergency stop occur. Here, the operation abnormality judgment value d32 output from the neural network 502 is received, the output data determination part 613 weights and adds both degrees, and sets it as the operation abnormality evaluation value d52. In general, when an operation abnormality occurs, an operator performs an emergency stop, but the operator also performs an emergency stop when a symptom leading to plate breakage occurs. Here, as a symptom leading to plate breakage, for example, a case where the material to be rolled 300 meanders can be considered.

In the case where there is no emergency stop and the plate breakage occurs, the plate breakage occurs without signs of the plate breakage. In this case, the priority of suppressing the plate breakage is higher. Accordingly, the weight of the degree of plate breakage is increased.

The mechanical operation end position abnormal region search unit 611 shown in FIG. 14 stores the output estimated position d31 and the returned abnormal operation evaluation value d52, and the abnormal operation evaluation value d52 has the maximum value. search for an estimate of As a result of the search, when the maximum value of the operational abnormality evaluation value d52 exceeds a predetermined threshold, the mechanical operating end position abnormal region search unit 611 determines the mechanical operating end position at the estimated position d31 in that case. Any combination of the maximum value POSMAX(i) and the minimum value POSMIN(i) of each operating stage of the fluctuation amount d26 is stored as POSEST(i). Then, in order to confirm the increase or decrease in the operation abnormality evaluation value d52 due to correcting POSEST(i), POSEST(i) is increased or decreased by a predetermined ΔPOS(i).

That is, the calculation shown below is performed.

POSEST(i)+=POSEST(i)+ΔPOS(i)

POSEST(i)-=POSEST(i)-ΔPOS(i)

POSEST(i)0=POSEST(i)+0 (ΔPOS(i)=0)

Then, again, the estimated position d31 of each mechanical manipulation stage 203 is set by combining the estimated values of the positions of the three kinds of mechanical manipulation stages of each mechanical manipulation stage 203, print out The input data creation unit 612 creates input data to the neural network 502 from the rolling results d24 and the estimated position d31 and outputs it to the neural network 502 .

Here, the mechanical operation end position anomaly area search unit 611 searches for a combination of estimated positions d31 in which the operation abnormality evaluation value d52 is the smallest (the degree of abnormality is the smallest), and returns the result to actual results. It is output as the judgment value d51 (ΔPOS(i), -ΔPOS(i)) of the position abnormality area operation end. Moreover, the mechanical operation end position abnormality area search part 611 also includes the operation abnormality evaluation value d52 at this time as the maximum operation abnormality evaluation value in the actual position abnormality area operation end judgment value d51.

FIG. 16 shows a state in which the amount of change in position of the mechanical operating end shown in FIG. 15 has been corrected. The range indicated by drawing the oblique line of the operating range shown in FIG. 16 is a corrected part.

For example, the operating range of the manufacturing unit of the mechanical operating stage 3 shown in FIG. 16 is POSEST(3)+ or POSEST(3) where POSEST(3) is increased or decreased by a predetermined ΔPOS(3). The range indicated by - is modified. In addition, there are cases where the range is not corrected, as in the case of POSEST(i)0=POSEST(i)+0 (ΔPOS(i)=0) in the above expression.

In the example described above, the mechanical manipulation end position abnormality region search unit 611 created the estimated position d31 using the maximum value and minimum value of each mechanical manipulation end position variation amount d26. On the other hand, if there is room in the processing capacity of the computer, it is also possible to create the estimated position d31 by further subdividing it.

In addition, although the mechanical operation end position abnormality area search unit 611 created the estimated position d31 with three variations from POSEST(i), it may be created by other processing. For example, the area search unit 611 for abnormal position of the mechanical operating end may finely control the amount of change in the estimated position d31. In addition, the mechanical operating end position abnormality area search unit 611 may be changed at the right time depending on the situation, such as not performing a search in a direction in which an operation abnormality does not occur obviously. Here, as for the direction in which operational abnormality does not occur obviously, it is possible to consider, for example, a case of moving toward the center of the mechanical operation end performance position.

The control unit 620 for suppressing the position abnormality of the mechanical operation end determines the resultant position abnormality area operation end determination value d51, which is the output of the mechanical operation end abnormal position area determination unit 610, and the coolant operation of the second shape control unit 212. From the control command to the stage 204, the coolant setting d41, which is the setting output to the coolant operation stage 204, is created.

When it is determined that operation abnormality does not occur even when the position of the mechanical manipulation end changes according to the amount of change in the position of the mechanical manipulation end d26, the position prediction unit 210 of the mechanical manipulation end simultaneously with the amount of change in the position of the mechanical manipulation end d26 The predicted coolant operating end position prediction d28 may be used as it is.

In the coolant control rule database 623, the correspondence between the mechanical operation stage 203 and the influence coolant operation stage 204 for each manufacturing unit is determined in advance. This correspondence can also be obtained by actually operating the mechanical operation stage 203 and the coolant operation stage 204 during rolling operation, and can also be obtained from performance data by machine learning. Here, the case where correspondence is obtained from the result of actual operation and registered in the coolant control rule database 623 is considered.

[Configuration and operation of coolant control stage control output calculation unit]

17 shows the configuration and operation of the coolant operating stage control output calculation unit 621 .

In the coolant control rule database 623 (FIG. 14), the coolant flow rate change required amount for which the effect equivalent to operating each mechanical operation stage 203 (k) is obtained is registered. The database search unit 631 shown in (a) of FIG. 17 searches the coolant control rule database based on the actual position abnormal area operation end judgment value d27 obtained by the mechanical operation end position abnormal area determination unit 610. From 623, the coolant flow rate change required amount corresponding to the actual position change amount of the mechanical operating stage 203(k) where the operation abnormality occurs is taken out.

Then, the output synthesizing unit 632 performs addition processing of the coolant flow rate change required for each mechanical operating stage 203 (k) taken out, and obtains the abnormality suppression output d42 shown in FIG. 17(b). .

The coolant operating stage control output selector 622 adds the abnormal suppression output d42 and the coolant operating stage position prediction d28 with the adder 622a, and also performs upper and lower limit processing in the upper and lower limit processing unit 622b. After implementation, it is output as coolant setting (d41).

The coolant setting (d41) is output as a guide to the operator on the man machine interface screen, in addition to being directly set in the coolant operating stage 204, so that the operator operates the coolant operating stage 204 according to the guidance You can do it.

Based on the above, when it is judged that there is no operational abnormality due to the position of the mechanical operating stage 203 even if the rolling operation of the next manufacturing unit is performed based on past performance, the coolant operating stage position prediction (d28) This is output as the coolant setting d41. In addition, when the occurrence of an operational abnormality is predicted, the corrected coolant setting d41 in the direction of suppressing the occurrence of an operational abnormality is output.

The coolant setting of each manufacturing unit and the change in the position of the mechanical operating end are constantly learned by the mechanical operating end position predictor 210 and used when the rolling operation of the same manufacturing unit is performed next time, so that operation abnormality occurs. It is possible to realize coolant setting that suppresses

As described above, according to the plant control device of this example, it is possible to perform good shape control while preventing operational abnormality resulting from the mechanical operating end position performance d22 of the mechanical operating end 203 .

[modified example]

In addition, this invention is not limited to the example of embodiment mentioned above, and various modified examples are included. For example, the embodiments described above have been described in detail in order to easily understand the present invention, and are not necessarily limited to those having all the described configurations.

For example, in the above-described embodiment, the mechanical operation end safe operation range determination unit 205 is realized by machine learning, but the mechanical operation end safe operation range determination unit 205 is expressed by a formula based on the operator's experience. can be realized. Alternatively, the rolling state at the time of operation abnormality may be made into a database, and the presence or absence of a corresponding case may be determined to realize the mechanical operation end safe operation range determining unit 205 .

In addition, the mechanical operation end safety operation range determination unit 205 may create a numerical model or a preference ethics model based on the knowledge of an operator or an operation technician, and may use it at the time of machine learning.

In addition, in the above-described embodiment, the control unit 620 for suppressing the position abnormality of the mechanical operating end stores the result obtained by experiment or the like in advance in the coolant control rule database 623 and uses it. On the other hand, the control unit 620 for suppressing abnormal position of the mechanical operating end may create a rule base from performance data using machine learning.

Further, in the above-described embodiments, the shape control of the rolling mill was targeted, but the present invention can also be applied to general plant control.

In the block diagrams of FIG. 1 and the like, control lines and information lines are shown only as necessary for explanation, and it cannot be said that all control lines and information lines are necessarily shown on a product. In practice, you may think that almost all components are connected to each other.

Further, the processing units such as the control unit described in the above-described embodiment examples may each include dedicated hardware, but the functions of each processing unit described in the above-described embodiment examples may be realized by implementing a program (application) in a computer.

Fig. 18 shows an example of a hardware configuration in the case where the plant control device is constituted by a computer.

A plant control device (computer) 100 shown in FIG. 18 includes a CPU (Central Processing Unit) 100a, a ROM (Read Only Memory) 100b, and a RAM (Random Access Memory) each connected to a bus. ) (100c). In addition, the plant control device 100 includes a nonvolatile storage 100d, a network interface 100e, an input/output device 100f, and an output device 100g.

The CPU 100a is an arithmetic processing unit that reads program codes of software for realizing the functions performed by the plant control device 100 from the ROM 100b and executes them.

In the RAM 100c, variables and parameters generated during arithmetic processing are temporarily written.

For the nonvolatile storage 100d, a large-capacity information storage medium such as a hard disk drive (HDD) or solid state drive (SSD) is used, for example. In the non-volatile storage 100d, a program (plant control program) that executes processing functions performed by the plant control device 100 is recorded. In addition, data necessary for performing machine learning is recorded in the nonvolatile storage 100d.

The network interface 100e transmits and receives various types of information with the outside via a LAN (Local Area Network), a dedicated line, or the like.

The input/output device 100f inputs various types of information from the control target plant 190 (rolling mill 301) and outputs information for instructing the operation stages 103, 104 (203, 204). .

The display device 100g displays the control state of the control target plant 190 (rolling mill 301).

In addition, program information for realizing each processing function performed by the plant control device 100 can be stored in a recording medium such as a semiconductor memory, IC card, SD card, optical disk, or the like, in addition to non-volatile storage such as HDD or SSD. .

In addition, when a part or all of each processing part of a plant control apparatus is comprised by hardware, you may use FPGA (Field Programmable Gate Array) or ASIC (Application Specific Integrated Circuit).

11: shape detection pre-processing unit 12: pattern recognition unit
13: control operation unit 14: shape detector
50: control unit 100: plant control unit (computer)
101: calculator 102: third control unit
103: high-speed operation stage 104: low-speed operation stage
105: safe operation range determining unit 110: control unit
111: high-speed operation stage control unit 112: low-speed operation stage setting unit
190: control target plant 201: calculator
202: mechanical control end position suppression control unit
203: mechanical operation stage 204: coolant operation stage
205: mechanical control end safe operation range determining unit
210: mechanical control end position prediction unit 211: shape control unit
212: coolant control stage setting unit 220: control unit
300: material to be rolled 301: rolling mill
302: entrance tension reel (entrance TR) 303: exit tension reel (output TR)
304: mill speed controller 305: inlet TR controller
306: exit tension reel control unit 307: roll gap control unit
308: entry tensiometer 309: exit tensiometer
310: rolling speed setting unit 311: entry tension setting unit
312: output tension setting unit 313: input tension control unit
314: output tension controller 315: input tension current converter
316: output tension current converter 317: output plate thickness meter
318: exit plate thickness control unit 319: roll gap setting unit
401 work roll 402 first intermediate roll
403: second intermediate roll 404: AS-U roll
405: split roll 406: saddle
501: input data preparation unit 502: neural network
503: neural network learning control unit 504: neural network selection unit
505: Teacher data preparation unit 506: Operation abnormal determination unit
511: learning data database 512: control rule database
610: mechanical control end position abnormal area determining unit
611: mechanical control end position abnormal area search unit
612: input data creation unit 613: output data determination unit
620: mechanical control end position abnormal suppression control unit
621: coolant control stage control output calculation unit
622: coolant control stage control output selection unit
623 Coolant control rule database 701 Input data creation unit
702: neural network 703: neural network learning control unit
704: neural network selection unit 705: teacher data creation unit
706: manufacturing unit data creation unit 711: learning data database
712 Control rule database 901 Calculator
902: first shape controller 903: second shape controller
904: mechanical operation stage 905: coolant operation stage
990: rolling mill d11: first state amount target
d12: first state quantity d13: second state quantity
d14: Safe operation range d21: Target shape
d22: mechanical operating end position results d23: shape results
d24: Rolling performance d25: Safe operating range of mechanical operating end
d26: Change in mechanical control end position d27: Coolant control end position performance
d28: Prediction of coolant control end position d30: Coolant control output
d31: Estimated position d32: Operation abnormal judgment value
d33: control neural network d42: anomaly suppression output
d51: judgment value of the operation end of the area with an abnormal performance position
d52: operation abnormal evaluation value

Claims (7)

A plant control device that performs, with respect to the plant to be controlled, a first operation process in which the response speed to the operation is a predetermined response speed, and a second operation process in which the response speed to the operation is slower than the first operation process,
an operating stage controller for obtaining a target state quantity of the plant to be controlled and instructing the first operation process;
an operation stage setting unit for instructing the second operation process;
A first operating stage for executing the first operation process of the plant to be controlled in response to instructions from the operating stage controller;
a second operating stage for executing the second operation process of the control target plant in response to an instruction by the operating stage setting unit;
a safe operation range determination unit for determining a safe operation range of the first operation process by the first operation stage based on the performance of the first operation stage;
A learning unit that learns the position of a first operating stage and a target state amount by the first operation process of the plant to be controlled in units of production materials from operation results in the plant to be controlled;
An operating end position predicting unit for predicting a position of the first operating end at the time of production of a tea production product using a learning result in the learning unit,
The operating stage setting unit determines the operation range of the position of the first operating stage in the difference product in the operating stage position prediction unit and the judgment in the safety operation range determining unit during operation of the production material. Determining the position of the second manipulation end so that the position of the first manipulation end does not deviate from the safe manipulation range.
plant control unit. The method according to claim 1, wherein the first operation processing by the first operating stage has a control time response faster than that of the second operation processing, and has a limited influence on the state quantity to be controlled of the plant to be controlled.
The second operation process by the second operating stage has a control time response that is slower than that of the first operation process, and has an effect on the entire region of the state quantity to be controlled of the plant to be controlled. . The method according to claim 1, wherein the safety operation range determination unit recognizes the occurrence of an operation abnormality using the performance data of the control target plant, and uses the performance data at the time of an operation abnormality as teacher data to establish a relationship between the performance data and the operation abnormality. learn to determine the safe operation range,
The operating end position prediction unit sets the position results of the first operating end and the second operating end at the production material unit of the control target plant and product performance data including product information as teacher data, so that at the product material unit A plant control device that predicts positional changes of the first operating stage and the second operating stage of The plant control device according to claim 3, wherein the relationship between the performance data and the operational abnormality is obtained through machine learning based on the collected performance data. The method according to any one of claims 1 to 4, wherein the plant to be controlled is a rolling mill,
The first operation process is a mechanical shape operation process for changing the shape with a mechanical structure,
The second operation process is a coolant shape operation process for changing the shape by changing the spray amount of the coolant in the plate width direction,
The safe operation range determination unit determines the safe operation range of the first operation process based on the actual value of the mechanical position of the first operation stage where the operation abnormality does not occur,
The position of the first operating end and the position of the second operating end of the rolling mill in the production material unit are learned from the operational performance in the rolling mill, which is the plant to be controlled, and the learning result is used in the production of secondary production materials Has an operating end position prediction unit for predicting the position of the first operating end of
The operating stage setting unit determines the operation range prediction value of the position of the first operating stage in the automobile manufacturing product in the operating stage position prediction unit and the judgment in the safety operation range determining unit during operation of the automobile manufacturing product. wherein the position of the second operating stage is determined so that the position of the first operating stage does not deviate from the safe operating range. Regarding the plant to be controlled, in the first manipulation process where the response speed to the manipulation by the first manipulation stage is a predetermined response speed, and the operation by the second manipulation stage whose response speed to the manipulation is slower than that of the first manipulation process, A plant control method in which the arithmetic processing unit performs the second operation process of
a control procedure in which the arithmetic processing unit obtains a target state quantity of the plant to be controlled, and instructs the first operation process;
an operation stage setting procedure for instructing the second operation process;
A first operation execution procedure in which the arithmetic processing unit executes the first operation process of the control target plant in response to an instruction by the control procedure;
a second operation execution procedure for executing the second operation process of the plant to be controlled by the calculation processing unit in response to an instruction by the operation stage setting procedure;
a safe operation range determination procedure for determining, by the arithmetic processing unit, a safe operation range of the first operation processing by the first operation execution procedure, based on a result of the first operation processing;
A learning procedure for learning the control operation stage 1 position and target state quantity of the control target plant in units of production materials from the operation performance in the control target plant;
and an operating end position prediction procedure for predicting a position of the first operating end at the time of production of a tea production product by using a learning result in the learning procedure,
The operation stage setting procedure is based on the operation range predicted value of the position of the first operating stage in the vehicle production product in the operation stage position prediction procedure and the judgment in the safety operation range determination procedure, determining the position of the second manipulation end so that the position of the first manipulation end does not deviate from the safe manipulation range when
Plant control method. Regarding the plant to be controlled, in the first manipulation process where the response speed to the manipulation by the first manipulation stage is a predetermined response speed, and the operation by the second manipulation stage whose response speed to the manipulation is slower than that of the first manipulation process, As a computer program that causes a computer to execute the second operation process of
a control procedure of acquiring a target state quantity of the plant to be controlled and instructing the first operation process;
an operation stage setting procedure for instructing the second operation process;
A first operation execution procedure for executing the first operation process of the control target plant in response to an instruction by the control procedure;
a second operation execution procedure for executing the second operation process of the control target plant in response to an instruction by the operation stage setting procedure;
a safe operation range determination procedure for determining a safe operation range of the first operation process by the first operation execution procedure based on a result of the first operation process;
A learning procedure for learning the control operation stage 1 position and target state quantity of the control target plant in units of production materials from the operation performance in the control target plant;
A program for causing the computer to execute an operating end position prediction procedure for predicting a position of the first operating end at the time of production of a tea production product using a learning result in the learning procedure,
The operation stage setting procedure is based on the operation range predicted value of the position of the first operating stage in the vehicle production product in the operation stage position prediction procedure and the judgment in the safety operation range determination procedure, determining the position of the second manipulation end so that the position of the first manipulation end does not deviate from the safe manipulation range when
A computer program stored on a computer-readable recording medium.

Images