JPH03188501A - System driving device - Google Patents

Abstract

PURPOSE:To stably maintain the operating state of a plant by selecting plant states exceeding first thresholds to operate degress of importance and obtaining the synthesized degree of importance based on these degrees of importance and driving a control system or the like at the time when this synthesized degree of importance exceeds a second threshold. CONSTITUTION:Based on plant data obtained from a prescribed plant, the degree of conviction of each of one or more kinds of preliminarily classified plant state is operated by a conviction degree calculating means 8. An importance degree arithmetic means 9 compares degrees of conviction with the first threshold set for each plant state to select the plant states related to degrees of conviction exceeding the first thresholds and operates their degrees of importance. A control means 11 obtains the synthesized degree of importance, where respective degrees of importance are synthetically taken into consideration, in accordance with degrees of importance of respective plant states by a function having each degree of importance as the variable, and the control system or the like for the plant is automatically driven when the synthesized degree of importance exceeds the second threshold. Thus, the control system or the like is properly driven at any time in accordance with the operating state of the plant, and the plant is stably operated.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is a system for timely driving a control system or alarm system (hereinafter referred to as a control system, etc.) that controls a predetermined plant or issues an alarm to the plant. Connect to system drive.

[Prior art]

Fig. 19 shows a roll rolling mill 2 for aluminum foil rolling, which is an example of the background of the present invention.
The raw aluminum foil 51 with a thickness of about 300 to 120 m to several hundred μm
It is rolled by a pair of rolling rolls 52 at a speed of 0 m/win to reduce its thickness to about %-4. The rolled aluminum foil 53 is conveyed in the direction of the arrow by a constant tension generated by rotation of the drive shaft of the output coil 64 (FIG. 1), and wound around the output coil 64.

For example, when rolling raw aluminum foil 51 with a thickness of several hundred μm into aluminum foil 53 with a thickness of several μm, the rolling process is repeated several times, and the number of rolling times cannot be called the number of passes. It will be done.

In the above-mentioned metal rolling in micron units, the aluminum foil 53 has a part that is "stretched" in the width direction of the foil in the arrow L), as shown in FIG. 20, even though the thickness is the same. There are clearly areas that are "tight".

That is, the stretched portion 54 has a peak portion 56 and a valley portion 57 formed along the conveying direction (arrow K) of the aluminum foil 53, and the stretched portion 55 has a generally flat shape. Therefore, the aluminum shown in the diagram? I53 is in a state where the central part in the foil width direction (arrow L) is elongated and the ra is stretched.

The distribution of elongation and tension in the foil width direction (arrow L) will hereinafter be referred to as the surface shape or actual shape of the aluminum foil 53. The surface shape has a great influence on the quality of the foil product, and in some cases, a large tension is applied to the stretched portion 55, causing the foil to break. Further, the stretched portion 54 causes wrinkles.

Regarding the aluminum foil 53 as a final product, it is natural that a flat shape with uniform elongation and tension is desired, but it is not necessarily a flat actual shape for each bus, and there are The aluminum foil 53 in the bus comes in a wide variety of shapes.

The surface shape of the aluminum foil 53 as described above is similar to that of the rolling roll 5.
can be controlled by changing the shape of 2. As shown in FIGS. 19 to 21, the rolling roll 52 is
The example shown in FIG. 21, in which a bulge called a thermal crown occurs due to heat generation during rolling and its heat conduction characteristics, is a case where the quarter portion a swells. Such a bulge, ie, a thermal crown, changes the surface shape of the aluminum foil 53 depending on its appearance location and degree of swell. That is, the aluminum foil 53 that has been rolled at a portion of the rolling roll 52 where the degree of expansion of the thermal crown is large is in an elongated state at that rolled portion. Therefore, the surface shape of the aluminum ti53 is the same as that of the rolling roll 5.
It is controlled by changing the temperature or the amount of coolant 58 (Fig. 1) injected toward the rolling roll 52 in the width direction of the aluminum foil 53 (arrow L (Fig. 20)) to cool the aluminum foil 53. obtain.

Such shape control of the aluminum foil 53 is performed by roll rolling 11.
This is done by a shape control section 3 adjacent to 2. That is, the shape control section 3 is rotatably provided on the exit side of the rolling roll 52, and controls the aluminum foil 53 from the inspection roll 4, which is made up of 4° elements divided into 36 pieces in the foil width direction (arrow L).
Actual shape data of elongation and tension are input. One piezoelectric element (not shown) is embedded in each element 4°, and serves as a sensor for detecting the pressure force applied to the outer peripheral surface of the element 4°.

And element 4. The aluminum foil 53, which is pressed upward and pulled in the conveying direction (arrow K) with a constant tension, has an elongated portion 54 of the element 4 when it passes over the element 4. The pressure contact force is small, and on the other hand, when the tensioned portion 55 passes, it is detected to be large.

Therefore, the actual shape of the aluminum foil 53 is expressed as the distribution in the width direction of the elongation rate (actual shape data) converted from the pressure contact force data detected from each element 4°, as shown in FIG. In this case, the quarter part a of the rolling roll 52 is too swollen, so cooling of that part is encouraged and heat is stored in the central part of the rolling roll 52 and both ends thereof.
A target shape has been set.

The shape control unit 3 compares and calculates the actual shape data with target shape data input in advance, and directs the actual shape data to the part of the rolling roll corresponding to the element 46 having a higher elongation rate. Increase the amount of Ruriichi Land 58. The above-mentioned coolant 5 is injected from an injection pipe 59 which is disposed on the inlet side of the rolling roll 52 and is injected in parts in the foil width direction (arrow L).

Thereby, the thermal crown of the rolling roll 52 is alleviated,
The portion of the aluminum foil 53 corresponding to the quarter portion a deforms toward the tensioned state. Furthermore, if the elongation rate of the actual shape data is lower than that of the actual shape data, the opposite operation is performed. In addition, in the target shape data, the elongation rate obtained from the element 46 corresponding to the quarter part a of the rolling roll 52 is often set to O so as to promote cooling of the quarter part a of the rolling roll 52.

As described above, in foil rolling, the final product is obtained by repeating multiple rolling passes. The plan for how many passes and how much thickness or surface shape to obtain the final product is called the operation policy.This operation policy is used to finish the final product into a flat shape with uniform elongation and tension. , the target shape for elongation and tension in intermediate passes is determined together with the target thickness.

In actual operation, the distribution of elongation and tension in the intermediate passes as described above differs. This is because, as described above, operating conditions such as thermal deformation (thermal crown) of the rolling rolls differ for each pass. Therefore, the above-mentioned operating policy is one in which the target shape of elongation and tension in each pass is determined by taking into account the expected operating conditions to some extent among such actual operating conditions.

However, the target shape of elongation based on the operation policy as described above and the target shape of actual elongation/tension by the shape control section 3, that is, the target shape data, are often indistinguishable. In a rolling mill, a certain material is rolled into a second
An example of this is when rolling with a target shape in the operation policy as indicated by the broken line in Figure 2, and obtaining good results by setting the actual target shape data in the shape control section 3 as indicated by the solid line. It is.

In this way, the target shape based on the operation policy and the target shape based on actual control are always consistent, even if the amount of coolant to the rolling roll 52 is adjusted to the element 4 of the inspection roll 4. Even if it is divided and adjusted according to the temperature, the thermal crown moves due to the heat conduction of the rolling roll, and the mode of movement also depends on the type of material, the surface shape of the rolling roll, heat balance, temperature, and rolling speed.

This is because it changes from time to time depending on operating conditions such as the shape of the foil base. The effects of operating conditions and material types are often based on know-how that cannot be expressed using mathematical models.
Therefore, in order to get closer to the operating policy, the target shape data must be adjusted in response to the various actual shapes that appear along the way of each bus.

[Problem to be solved by the invention]

However, with the above-mentioned target shape data, it is up to the operator 5 to determine whether or not it is necessary to change the target shape, such as when to change the target shape, and to start the target shape change adjustment process. Automatic processing of the change adjustment was not possible. Therefore, it has not been possible to adjust the target shape appropriately and timely to follow the slightly changing operating conditions of aluminum foil rolling.

Furthermore, even if the determination of the necessity of changing the target shape and the activation of the target shape change adjustment process are automatically performed based on a threshold value determination for the detected actual shape data, the actual shape data is determined to be between O and 1 after the threshold value determination. Since it is processed as 200 million logics, it is difficult to precisely grasp the degree of abnormality of the actual shape. If the amount exceeds the above, the process will not be started, and the state A will be changed to a different state J.
When IB and IB were abnormal at the same time, it was not possible to make a determination that the process would be activated if the actual shape data related to state A exceeded the threshold even by a small amount.

On the other hand, it is not possible to express all rolling conditions using only the current actual shape data and the shape characteristics derived by calculation from the actual shape data. Characteristic information that cannot be obtained without using actual shape data at time, such as trends in collected data, average variance, and correlation between data.

There are many types of 3D pattern recognition. Such characteristic information is also important in adjusting the surface shape of the aluminum foil 53. However, judgment based only on the latest so-called actual shape data is not sufficient for driving the shape control section 3.

Therefore, it is an object of the present invention to provide a system drive device that can drive a control system and the like appropriately and timely depending on the operating state of a plant.

[Means to solve the problem]

In order to achieve the above object, the main means adopted by the present invention is that, in a system driving device that drives a control system etc. for a predetermined plant, plant data relating to one or more types of plant states is used. a sensor group for detecting a sensor group, a storage means for storing the plant data, a reliability calculating means for calculating a reliability for each plant state from the plant data, and a first storage means for each state regarding the reliability for each plant state. Compared with the threshold value, the first
an importance calculating means for selecting a plant state that exceeds a threshold and calculating its importance; and calculating a composite importance from each of the aforementioned importance, and when the composite importance exceeds a predetermined second threshold, the control system The present invention is typically applicable to a product shape adjusting device in metal rolling. In addition, the system drive timing can be adjusted using statistical data, and it is even more advantageous to judge the plant status using pattern recognition. From the obtained plant data, the certainty factor for each of one or more pre-classified plant states is calculated by the certainty factor calculation means.
Next, the importance calculation means compares the certainty factor with a first threshold value set for each plant state, selects a plant state with a certainty factor exceeding the first threshold value, and calculates its importance degree. . Further, in the control means, a composite importance level that comprehensively considers each importance level is determined from the importance level for each plant state using a function that uses the respective importance levels as variables, and the composite importance level is determined as a predetermined second level. When the threshold is exceeded, the plant control system etc. are automatically activated.

Moreover, if the characteristic information of the plant state within a predetermined period of time in the past is calculated from the plant data, and the steps after the above-mentioned confidence calculation means are executed based on this, instantaneous It is possible to judge the trending plant state and drive the control system in an appropriate and timely manner without being influenced by noise data.

〔Example〕

Next, embodiments embodying the present invention will be described with reference to the accompanying drawings to provide an understanding of the present invention.

Here, FIG. 1 is a schematic diagram showing a system arrangement of an aluminum foil rolling target shape adjusting device according to an embodiment of the present invention, and FIG. 2 is a configuration diagram showing a processing flow of the aluminum foil rolling target shape adjusting device. Figure 3 (a) is an explanatory diagram showing the main parts of the actual shape data expressed by elongation weight distribution in the foil width direction, and Figure 3 (b) is an explanatory diagram showing the actual shape classification items of pattern-classified aluminum foil. 4(a) is a flowchart showing the processing procedure for determining the change tendency of actual shape data, and FIG. An explanatory diagram showing FIG. 5(a)
and the same figure (b) is a graph showing the correlation between the number of levels of two real shape classification items, and FIG. 6 is a three-dimensional graph showing a three-dimensional pattern due to changes in real shape data of aluminum foil over time.
Figure 7 is a schematic diagram conceptually showing a neural network, Figure 8 is an explanatory diagram showing an example of the relationship between shape change targets for real shape classification items and corresponding axilan candidates, and Figure 9 is an important diagram of shape change targets. Figure 10 shows the relationship between the degree of certainty and the degree of certainty when the actual shape classification item is edge-bound. FIG. 11 is an explanatory diagram showing the processing procedure of a routine that checks the validity of the axilan to be applied using a check tree, and FIG. 12(a) shows the target used to change the target shape. Explanatory diagram showing shape adjustment parameters, same figure (b)
is a state diagram showing a change in the situation of the parameter a), and (C) is a state diagram showing a situation where the center part of the target shape adjusted by the parameter a4 is a forward pattern.
M, the same figure is a state diagram showing a situation where the central part is a reverse pattern, FIG. 13 is a schematic explanatory diagram showing an example of inference execution for changing the target shape, and FIG. 14 is a processing flow for adjusting the target shape. Fig. 15 is a flow sheet showing the incoming camera input displayed on the screen of the terminal on the rolling mill side, Fig. 16
Figure 17 is a display diagram showing an example of the target shape displayed on the screen, Figure 17 is a flow sheet showing a method for correcting the asymmetrically obtained actual shape, and Figure 18 is a situation in which the asymmetrical actual shape is corrected. FIG.

In the following description, the same reference numerals will be used for the elements common to the roll rolling mill 2 shown in FIGS. 19 to 21, and the description thereof will be omitted.

Further, the following embodiments are merely examples embodying the present invention, and are not intended to limit the technical scope of the present invention.

The aluminum foil rolling target shape adjusting device 1 according to the present embodiment is applied to Broglie-Smeen line control as a process system that systemizes operator control know-how. Prior to the explanation, the above-mentioned aluminum foil rolling target shape will be explained with reference to FIG.

As shown in the figure, the aluminum foil rolling target shape adjusting device 18 includes a rolling data collection section 7, a rolling situation analysis section 8, a control target generation section 9, an Axilan candidate thruster 11, a target shape generating section 12, and an Axilan. It mainly consists of the effect evaluation section 10, and includes a rolling situation analysis knowledge base DI that stores the wisdom of operation J-, a control target setting knowledge base D, and an Axilan inference knowledge base Ds (knowledge for maintaining consistency of rules and Axilan validity) ), and working memories Mr, Mz, M! that temporarily store various data. , Ma, Ms.
It is equipped with A summary of the processing contents in each of the above sections will be explained below.

In the above-mentioned aluminum foil rolling target shape 11 adjustment device 1;, J:, first, inference processing is activated by key human power from the terminal G of the aluminum foil rolling mill 2 example, and the operating condition data and evening data from the aluminum foil rolling machine a2 are started. It is input via the shape control section 3.

■Rolling data collection unit Rolling data collection unit 715 receives operating condition data including actual shape data from the shape control unit 3 and writes it into working memory M8 ■Rolling status analysis unit Rolling status analysis unit 8jJ5, The two-dimensional shape data is analyzed to determine the rolling state of the aluminum foil 53. It is determined to what extent the actual shape data recorded in "Ipchi" 2 corresponds to actual shape patterns that are classified in advance into several types of *fl patterns and stored in the rolling situation analysis knowledge base D.

At the same time, the current target shape data is analyzed.

■Control target generation unit The control target generation unit 9 generates the results of the analysis of the actual shape data by the rolling situation analysis unit 8F: and the terminal device 6 by the operator 5.
Based on the input from the controller, a target of the direction in which the actual shape of the aluminum foil 53 is to be changed is set.

■Axis hit candidate inference unit■-(1) Rule inference The axis candidate inference unit 11 uses the above-mentioned control goals and operating conditions as a condition part, and the axirun, etc. for realizing the control goals mentioned above as a conclusion part. The axilan determined to be valid is written into the working memory Ms by the rule It theory that applies the rules stored in the IF-THEN type axilan inference knowledge base D. During this writing, the following processing is executed.

■-(2) Elimination of contradictions and redundancies When contradictory axilan mosquito supplements are deducted, the axilan of the more important control objective is applied.

■-(3) Learning invalid axillanes In the above rules, if there are vA axilan candidates for a certain control target, a priority is given to the axilan candidates, and the highest priority Apply Axilan only. Executed to achieve a certain control goal! lhAxilan that has no effect is not repeatedly applied even if the same control target is set next time.

(2) Target shape generation unit The target shape generation unit 12 generates new target shape data based on the applied axilan, and outputs it to the shape control unit 3. The shape wj controller 3 controls the aluminum foil rolling II2 based on this target shape data.

■Axilan effect evaluation unit The axilan effect evaluation unit 10 evaluates whether or not the II control of the aluminum foil rolling 112 based on the applied axilan was effective, based on the results of data analysis and the arrangements made to the operator 5. At this time, the axillane evaluated to be invalid is stored in the working memory M6, and the axillane is stored in the working memory M6 and the axillane is evaluated as invalid. Il#! Spear! It will be referred to during the next axis candidate It discussion in 111.

This example will be described in detail below.

The above aluminum foil rolling density shape adjustment device [1 is shown in Fig. 1 and 1]
As shown in Figure 12, the actual shape of aluminum 1153 is 5ui.
Target shape data serving as a guideline for the control is output to the shape control section 3 that controls the injection amount or temperature of the coolant 58 so as to i, and operating condition data is input from the shape control section 3.

Another possible method for adjusting the actual shape of the aluminum foil 53 is to control the push-up blade of the push-up roll 60 that urges the lower roll 52 upward toward the upper roll 52. In this embodiment, only the control of the coolant 58 will be described below.

■ Collection of operating condition data In the aluminum foil rolling target shape adjusting device 1, the inspection roll 4 is used to measure the elongated portion 5 of the aluminum foil 53 at the time of rolling.
Element 4.4 and a sensor that detects the actual shape data showing the stretched portion 55 (FIG. 20). It is provided as an assembly on the downstream side of the rolling roll 52 in the conveyance direction (arrow K), and outputs the actual shape data to the rolling data collecting section 7 (FIG. 2) via the shape controlling section 3. The rolling data collection section 7 writes the operating condition data (Table 1) transferred from the shape control section 3 at predetermined time intervals into the work memory M and displays it.

■Analysis elements of actual shape data 4. The rolling situation analysis knowledge pace D1 for calculating the elongation state and tension m of the aluminum foil 53 during rolling (respective actual shape classification items) and their degree from the actual shape data detected by the (sensor) As shown in Figure 3 (B), for example, the real shape classification items such as "edge tension" to "touch stretch" and the real shape classification detected from the element 4° and corresponding to the real shape data indicating the real shape. It stores a specific program for each item to specify the item.

First, a method for identifying actual shape classification items will be described. The actual shape data detected as a tension distribution from the element 4° is obtained in the form of an elongation distribution in the foil width direction, as shown in FIG. 3(a).

The actual shape data shown in the figure consists of an end portion, a quarter portion, a center portion A, and a center portion B from the outside, and the entire center portion is further composed of the center portion A and the center portion B. In this case, the tension portion 55 is located in the center portion B, and the extension portions 54 are located in the quarter portions on both sides.

The actual shape classification items stored in the rolling situation analysis knowledge base D1 are mainly classified into five types as shown below, as shown in FIG. 3(b).

(1) "Edge tension"...When the elongation rate at the edge shows a low value at the end, it is considered to be edge tension. ■Whether the value at the end is the minimum value of the whole.

■Judged by the extent of the difference in elongation rate between the end and quarter parts.

(2) ``r end section''...Contrary to the case of edge binding, this refers to a case where the value at the end section is significantly larger than that at other parts.

In many cases, it is preferable to have a real shape where the edges extend to some extent, but if the edges extend too much, it is considered a problematic shape.

(3) "Quarter elongation"...Judged based on how large the elongation rate value of the most elongated part near the part corresponding to the set zero point in the target shape is compared to that at the end. . As mentioned above, the actual shape in the quarter portion is the easiest to stretch, and most of these types of shapes appear in actual actual shapes.

(4) ``R middle tension'': The tension state (low elongation rate) at the center is compared and judged with that at the ends.

(5) "Medium elongation"...The difference between the elongation rate value of the most elongated part in the center and the elongation value of the part elongated the most overall (in most cases, the end or quarter part) is small. Is it?
Or, if it is negative, it is determined that there is medium growth.

For mid-stretch, there are two types of stretch: ■ A straight mid-stretch where the central part is elongated. .

Other unique real shape classification items include "non-target" below.
, there is "zero point inappropriateness".

"Asymmetric"...Usually 1, the target shape is symmetrical in the foil width direction, and the actual shape is the symmetrical shape, but
A shape in which this symmetry is broken is called an asymmetric shape. The criteria for this judgment are as follows: (1) The difference in elongation rate between the maximum value and the elongation rate between the minimum value at both the left and right end portions t is large.

■One end is specified as end tension and the other end as end extension.

It is considered asymmetric if any of the following holds true.8 Of course, both may hold true.

“Inappropriate zero point”: This refers to a case where the zero point location set in the target shape and the location showing the maximum elongation rate (^) in the actual shape do not match.Normally, the quarter portion a of the rolling roll 52 Heat tends to accumulate, and the quarter part of the actual shape corresponding to the quarter part a is the easiest to stretch.Therefore, when setting the target shape, place 7 The point is set 1. Then, it is necessary to adjust so that if these parts are in rows, they will be 7 people.

Based on the detected actual shape data, it is determined to which actual shape classification item the current shape state corresponds according to the method shown in section 62 of "Identification method" in FIG. 3 0)). As described above, this method is stored in the rolling condition analysis knowledge base D as a program and 1-.

”-j-, one or more actual shape classification items that are the cause of the actual shape data input from the working memory M, are stored in the rolling situation analysis unit Om,! :; and said specific block 1
Calculated by cogram.

Normally, when a certain real shape is in that state, it is not always the case that only one of the real shape data is selected. Therefore, it is often judged that there are multiple real shape classification items such as tl+i:Z. In that case, there are real shape classification items with strong causal relationships with real shape data and classification items with weak correlation. be. The strength of such a causal relationship, that is, the degree of the shape state, is called certainty.

The rolling situation analysis unit 8 narrows down the actual shape data into one or more actual shape classification items with a certain degree of certainty derived by an appropriate function, and stores the actual shape classification items and their fI body in a working memory M. .

to be memorized. For example, aluminum foil 53 is 13! Within the range of the actual shape data input from the four elements 4o from the end where the plexus is placed, the elongation rate difference or2 between the part with a high elongation rate and the end, and the maximum elongation rate r in the entire actual shape data.
When the ratio β2/α2 of the difference β2 between 15 depending on the value of
A confidence level from 0 to 1 is added. The same holds true for other real shape classification items.

In this case, statistical characteristic information of the roll rolling mill 2, such as the average, change tendency 1 variance, correlation, three-dimensional pattern recognition, etc., is calculated based on the actual shape data within a predetermined past time from the working memory M. , based on the continuous characteristic information U2, that is, by using the statistical characteristic information as a variable for the certainty factor, the certainty factor for each of the real shape classification items may be calculated.

For example, for some reason, the actual shape at a certain point in time may differ from the actual shape at points before and after that point by a pole O#I. R-physically speaking, the cause is an abnormality in the shape of the raw material sheet before rolling, and when the actual shape of the "edge tension" continues, the state of "edge elongation" is detected for a moment, and the "edge tension" state is detected for a moment to restore the "edge tension" state. This is a case where the situation continues.

Therefore, by applying the average value of the statistical characteristic information of the actual shape data at several points in the past up to the above-mentioned certain point, it is possible to judge the rolling state of the trend without being influenced by the above-mentioned noise elements. .

Next, a case where a change tendency of actual shape data within a past predetermined time is employed as the statistical characteristic information will be described in detail below. This tendency is calculated, for example, according to the processing procedure of flowchart 1 shown in FIG. 4(b). During the rolling operation of the aluminum foil 53, the actual shape data of the aluminum foil 53 obtained from the inspection roll 4 on the roll rolling mill 2 side is collected by the rolling data collection unit 7 at predetermined time intervals (
340).

Next, the actual shape data is compared with the actual shape classification item in the rolling condition analysis section 8, and is specified as a certain actual shape classification item, and the degree thereof is expressed by a level number represented by a natural number O to 5. That is, the rolling state of the aluminum foil 53 is determined (S41).

For example, from the end where the aluminum foil 53 is pressed,
Within the range of the actual shape data input from one element 4゜, the elongation rate difference α2 between the part with the highest elongation rate and the end, and the difference β between the maximum and minimum elongation rates in the entire actual shape data
2 exceeds a predetermined setting value, it is determined that the actual shape at this time includes the actual shape classification item "edge elongation", and the The degree of "edge elongation" is expressed by a level number represented by a natural number 0 to 5.

Further, in step S42, the number Ht of shape points from -5 to +5 is determined for the actual shape data at time T, and is stored in a storage section (not shown) that stores the number H2 of shape points within a predetermined time.

This embodies an empirical handling method when the operator 5 specifies real shape data into an arbitrary real shape classification item,
For example, in order to make it easier for the operator 5 to understand when displayed on the terminal 6, the certainty of the real shape classification item to which a certain real shape data fits and the real shape whose elongation/tension state contradicts the real shape classification item. The certainty factor of the classification item is converted into true coordinates consisting of positive and negative natural numbers centered on 0.

For example, if the real shape classification item is "edge elongation" and the degree is level 5, the shape score is +5 points, and if it is level 1, it is +1 point. Also, if the level is 5 for "edge tension", it will be given 15 points, and if it is level 1, it will be given -1 point. "
0 points will be given if the item does not fall into the category of ``edge elongation'' or ``edge tension.''

Then, if it is necessary to change or adjust the target shape based on the input from the terminal device 6 of the two roll rolling mills according to a request from the operator 5' or the input from the rolling situation analysis section 8 (S43), in step S44, The change tendency of the actual shape within a predetermined time up to the present is calculated.

For example, the shape change tendency at time T is calculated from the past number of shape points Ht-1+11, . . . , Hl. Here, the past 10 shape points H,,.

. . , Hl. First, from the maximum and minimum values of the shape point numbers Ht-*+ .

H4-saw(Hz-v, -, Ht)-sin(
Ht-w, -, Ht) Then, the shape point difference H4
is less than 2, aluminum F! It is determined that the actual shape of No. 53 is in a stable state and there is no tendency for the shape to change.

If H4 is 3 or more, it is first determined whether the actual shape is in a periodic changing state.For example, shape points H&-9+
..., Ht, the number of times the number of shape points changes in the increasing direction is Ho, and the number of times the number of shape points changes in the decreasing direction is H-.
Then, as shown in the following formula, lH, -H-1≦3. And 3≦H4≦4 In other words, when the absolute value of the difference between Ho and H- is 3 or less and the shape point difference H4 is 3 or more and 4 or less, it is determined that the actual shape tends to change periodically.

If Hd is 2 or more and does not satisfy the above formula, it is assumed that a shape change tendency is recognized, and it is determined what kind of trend there is.For example, the current shape score H
1 and the shape score H1-9 of the past 10 points are compared.

That is, as shown in Table 2, Hl is smaller than tit-*.
When 2 also shows a large value, it indicates an elongation tendency in which the ends are being stretched; conversely, when it shows a small value, it is determined that there is a tension tendency in which the ends are becoming tense. Further, if H2 and Hl-9 are equal, it is determined that the actual shape is stable, and the step of correcting the number of levels of the actual shape classification item (S45), which will be described later, is performed.
For example, if the shape change tendency at time T is "edge elongation tendency" and the actual shape classification item at that time is identified as "edge elongation" and the number of levels is 3. As shown in Figure 4, etc., a correction action is taken to increase the number of levels by 2 for the actual shape classification item at that time (345), which is useful for future shape control. So, if we consider the current shape judgment and shape change tendency, level 3 in the above example is inappropriate, and realistically the degree of "edge elongation" is almost level 5 (edge elongation This is because it is considered to be comparable to a rapidly progressing state.

Therefore, in step S47 in FIG. 4(a), the target shape data is adjusted to obtain the desired real shape based on the latest real shape classification item specified as described above and the number of levels after correction. Changes and adjustments are made and output to the shape control section 3 (S4B).

On the other hand, if it is determined in step 344 that the actual shape tends to change periodically, the control gain for changing and adjusting the target shape data is reduced by 30% (
346), thereby suppressing the tendency of periodic changes in the actual shape.

On the other hand, if dispersion is applied as the statistical characteristic information, the rolling state can be determined from the degree of dispersion of the actual shape data regarding the shape of the end of the aluminum foil 53 due to temporal changes, for example. That is, when the degree of dispersion is large, it indicates that the rolling condition is unstable. Therefore, in such a case, the shape control section 3 executes the control axis shown in the following (i) and (ii) to stabilize the rolling state.

At this time, (i) change and set the coolant amount control gain to a lower value.

(ii) Since it is possible that the rolling roll 52 is not sufficiently heated, the amount of coolant is decreased overall or the elongation rate of the target shape is increased overall.

Further, as an example where the characteristic information is a correlation, the correlation between the level of quarter elongation and the level of edge tension is shown in FIG. 5(a).
In this case, it can be said that there is a positive correlation between the two, which indicates that quarter elongation and edge tension are likely to occur simultaneously.

In such a case, it is empirical knowledge that a shape adjustment method such as reducing the amount of coolant at the end in order to eliminate the end tension is not effective. On the other hand,
As shown in FIG. 5(b), if the relationship between the two is uncorrelated, the shape adjustment method taken in the case of positive correlation is effective. The actual shape is adjusted using the 11!1 method.

Furthermore, as the statistical characteristic information, 3 related to the actual shape data
It can also be by recognition of dimensional patterns.

In Figure 0 showing the three-dimensional pattern P in Figure 6,
An arrow M indicated by a two-dot chain line indicates each time, ~t.

As shown in Figure 0, which shows the temporal position transition of element 4° where the elongation rate has detected the maximum value, the elongation rate of element 4. is several elements 4.0 in the foil width direction.
It meanders within its width. If this is the case, the coolant 58 is injected not only to the part that is currently at its maximum elongation, but also equally intensively to the rolling rolls 52 corresponding to the several elements within the width of 4 degrees, including that part. There must be.

This applies coolant 5 only to the area that is currently experiencing maximum growth.
This is because even if 8 is injected, the elongated area only moves to the adjacent area. Statistical characteristic information represented by such a three-dimensional pattern P, such as ``The elongated portion of the actual shape meanders depending on the time'', can of course be determined by the numerical calculation algorithm described above.
A recognition method using a neural network that performs pattern recognition of actual shape data of the aluminum foil 53 is effective.

As shown in FIG. 7, in the bilayer network 20, a plurality of nitrogen layers 15 that calculate and output input data through threshold processing are conceptually arranged as an input layer, an intermediate layer, and an output layer, and there is a gap between each layer. They are connected via a connecting part 16. And, the neural network 20 is
The pattern data of the actual shape data of the aluminum foil 53 is used as input data, the actual shape classification items and their degrees are used as output data, and the correspondence relationship between the two is learned by changing the connection weight of the connection section 16. Therefore, when new real shape data is input to the trained neural network 20, the real shape data is specified in one of the real shape classification items along with its degree. Such a neural network 20 may be applied in the rolling situation analysis section 8.

That is, the rolling situation analysis section 8 is a reliability calculation means and a characteristic information calculation means.

Here, as a calculation method for determining the actual shape state of the aluminum foil 53, an example of numerical calculation using a specific method determined for each actual shape classification item has been mainly shown. It goes without saying that similar computational effects can be obtained by making decisions based on a rule base (not shown) that stores decision knowledge.

The actual shape data input from the working memory M to the rolling situation analysis unit 8 and subjected to calculation may be one type of data value at one point in time within the predetermined time, or several types of data values at one point in time. , or the value of one type of data at several points in time, or the value of several types of data at several points in time, and the data necessary for the calculation may be used as appropriate.

In this manner, the rolling situation analysis unit 8 determines the actual shape classification items and their reliability for the current actual shape data of the aluminum foil 53, and writes them into the working memory M2.

That is, the rolling condition analysis section 8 is a plant condition classification means, and the rolling condition analysis knowledge base D is a pattern storage means.

■Generation of control target The control target data (shape change target (Fig. 8) and its importance), which is the key to appropriately setting or changing the target shape, is generated by the operator 5 in the control target generation unit 9. It is input from the terminal device 6 on the roll rolling mill 2 side, or is automatically generated based on rolling status data such as actual shape classification items and their reliability in the working memory M2.

This automatic generation will reflect the operational policy (e.g., ``For a given bus, roll the edges with a large extension'').
, or a rule such as "If there is a conflict between the information input by the operator 5 and the automatically generated information, the operator's input information is given priority" is applied with reference to the control target setting knowledge base D2.

For example, if the detected actual shape data includes "edge elongation" as in the example above, and the confidence level at that time is 0.8, five shapes are required to eliminate "edge elongation". "I want to stretch the edges" is selected from among the change targets, and a degree of importance corresponding to the confidence level (0, 8) is assigned to the selected shape change target.

Then, the shape change target and its importance are stored in the working memory M3.

The situation in which the importance of the shape change target is calculated in the control target generation unit 9 from the certainty of the actual shape classification item specified as described above will be described in detail below.

The degree of importance is a measure of whether or not it is necessary to change the target shape data, and is shown in the graph (Fig. 9) showing the relationship between the actual shape data and the confidence level of the actual shape classification item indicating the degree of edge tension, for example. shown.

In the figure, an example of "edge tension" among the actual shape classification items is shown.

An importance calculation function (f(x)) that gives the degree (importance) of the necessity of changing the target shape is defined for each of the actual shape classification items such as edge tension and edge extension. The importance calculation function fl() for `` is defined as follows.

That is, the edge tightness degree (confidence) is compared with the first threshold Ll, and as shown in FIG. 9, if it exceeds the first threshold Ll, the actual shape classification item is selected for changing the target shape. The degree of importance corresponding to the degree of certainty shown in the figure is calculated. On the other hand, if the confidence level is equal to or less than the first threshold, 0 is given to the importance level, and the actual shape classification item is not used when changing the target shape. 1 threshold value is set individually, each confidence level is compared with the first threshold value, and it is determined for each item whether or not the target shape should be changed.

Whether or not there is a need to change the target shape is determined for each actual shape classification item as described above, and may also be determined based on whether the average of the combined importance of each item exceeds a threshold. be.

Next, a method for determining whether or not it is necessary to change the target shape from the actual shape classification items selected for changing the target shape based on the synthesized importance levels or their averages will be described in detail.

If the actual shape classification items are SI+S1+..., Si, then the importance calculation function f-! for each item defined for them is SI+S1+..., Si. Composite importance calculation function g that combines ()
() is defined as follows.

g (fs + (x+), ..., f,! (I
t)) The function g() is gmi(Σr-i(Xt))/i.

Or g=Σrst(Xt) However, 0≦fst(Xm)≦1°f*=(x is expressed in the form of a summation average or a summation, such as strongly monotonically increasing xl.

Here, when r (Ll ) is (1-+), strong monotonically increasing means that when Ll<Lx, f (L, )<f (
It means L-ri.

In this way, the composite importance indicating the degree of necessity for starting the inference process is calculated from the importance of each item in the axilan candidate inference section 11, and the composite importance is set to a predetermined second threshold Ls.
(not shown), inference processing starts,
Target shape data is changed appropriately.

A specific example is given below: The importance of "edge tension" is 0.4 The importance of "quarter elongation" is 0.6 The importance of other ("edge elongation", "medium elongation", "middle tension") When , the composite importance is 0, 4 + 0.6 + O = 1.0 in the case of summation. If the second threshold value RI at this time is 0.9, the combined importance is greater, so the inference process is started. As a method for determining the degree of importance serving as the trigger, the above-described total average method may be employed.

In this embodiment, as mentioned above, when the confidence level for each item exceeds the threshold value given for each item, for example,
The inference process is also started when the degree of the difference exceeds the threshold L8 shown in FIG.

Further, when the composite importance exceeds a second threshold value, an activation signal may be simultaneously output to an alarm device (not shown) to drive the alarm device.

Next, the target shape inference process will be explained.

■Application of Axilan■-(1) Rule inference Rolling process including operating condition data including current target shape data and current actual shape data, extracted actual shape classification items and their confidence levels processed as described above. situational data,
The control target data including the shape change target and its importance are transferred from the working memories M, , M, , M to the action candidate inference unit 11, respectively. The action candidate inference unit 11 collates each transferred data with the condition part of the rule stored in the axillary inference knowledge base D3, and as a result of the collation, determines a rule that satisfies that all the condition parts are true. is extracted, and the target shape change data (FIG. 8, hereinafter referred to as axilan) in the conclusion part of the rule is selected.

As already mentioned, the inference process that draws a conclusion (target shape to be adopted) corresponding to the conditions such as operating condition data, rolling status data, and MW Menoki data as described above does not rely on the knowledge (know-how) of experienced people. In the present invention, such inference processing is automated. Rules for such automatic inference are accumulated and stored in the action inference knowledge base D. Such a rule is expressed in the form "if [condition part], then [conclusion part]" and is expressed in the form of a logical product as shown below.

If [control target data] and (operating condition data, rolling status data), then [designation of target shape adjustment parameter and its change degree (action and its degree)] Here, the target shape adjustment parameter is -4, this is an element that determines the target shape data (target elongation weight distribution for the bus).

The target shape adjustment parameters that make up the conclusion section of each rule may not include all the parameters shown in Table 4; in many cases, some of the parameters necessary to satisfy the condition section Only the shape adjustment parameters are listed along with their degree of change.

For example, as shown in FIG. 10, in rule example 1,
If the actual shape of the aluminum foil 53 is determined to have quarter elongation, and there is no zero point under the part with the largest elongation near the quarter part, then A rule specifying an action such as "to have" is stored in the Axilan inference knowledge base D1.

(2) Eliminating contradictions and redundancies On the other hand, as shown in rule example 4 shown in Table 3, additional conditions may be added to the rules.

For example, in a real shape, when edge tension and quota elongation occur at the same time, rule example 2 and rule example 3 may be selected. The unit 11 is unable to resolve these contradictions and an error occurs. Therefore, for example, by setting an additional condition in the condition part of rule example 2 and making it rule example 4, this problem can be solved. That is, in Rule Example 4, the importance of wanting to stretch the edges and the importance of wanting to stretch the quota part are both greater than the first threshold, but the importance of wanting to stretch the quota part is less than 0.4. In this case, the target value for the end level (the elongation rate at the end in the width direction) is increased. As a method for resolving such contradictions or redundancies, it is also possible to give priority to the shape change target with the higher degree of importance.

On the other hand, there are cases where two or more types of shape change targets are simultaneously selected for certain actual shape data, and the contents of the axilan corresponding to each shape change target are the same.

For example, the actual shape classification items (Figure 3 (b)) of the current actual shape data are simultaneously specified as "edge tension" and "r medium elongation", and the degree is determined as an action derived from each actual shape classification item. This is a case where "Raise the edge level" (Figure 8), which has the same content but a difference, is selected at the same time. As a rule applied in such a case, a rule for setting the degree of action according to the certainty of the specified real shape classification item is stored in advance in the axilan inference knowledge base D.

Therefore, when the same actions with different degrees are selected at the same time as described above, instead of executing these axirans at the same time, only the one with the greater degree of action is executed, so that the execution of the action is changed. Redundancy is avoided. Here, conversely, only actions with a small degree may be executed, or actions that match the average value of the degrees of each action may be selected or generated.

■-(3) Learning invalid actions Furthermore, as shown in Figure 8, for the shape change goal of 1,
There are several types of action candidates with priorities added. Then, when a certain shape change target is selected, the action with the highest priority is executed.

The priority is not fixed (it is checked for each inference). For example, when inference is executed in the action candidate inference unit 11, which action is adopted for which shape change target is stored in the working memory M5. At the next time of inference, the action effect evaluation unit 10 determines whether the previous shape change goal has been achieved (effect evaluation) by comparing the importance levels of the previous time and this time.

As a result, if the current target shape data that has been changed based on the previous shape change target and its importance is determined to be valid by the action effect evaluation unit 10, that is, the importance is lower than the previous value. If so, it is assumed that the adopted action is valid, and the action inference knowledge base D
The priority stored in , is incremented. Conversely, if it is determined that the action is invalid, the previously applied action that was invalid and the rule that inferred the selection of that action are stored in the working memory M4.

For example, the shape change target shown in FIG. 8 is based on the rolling status data from the rolling status analysis unit 8 or the input data from the operator 5, and the previous time "I want to lengthen the edges" is determined to be 0 in importance.
6, "Raise the level of the edge" which has the highest priority among the accompanying actions (Priority level) is selected, and the level of the edge (
The target shape data with increased elongation rate) is output to the shape control unit 3, and the target shape data and shape change target are applied at the same time.

its importance (0,6), its action, and its priority (
1) is stored in the action inference knowledge base D. Then, when determining the current shape change goal, if the shape change goal "I want to lengthen the edges" is ashes with the previous importance of 0.6 or higher, the edge tension state of the actual shape in question will be In many cases, it has not been improved (this means that the action applied last time was invalid.On the contrary, this time I want to “stretch the edge”)
is selected with an importance level of less than 0.6, it is determined that the previous action was valid.

Therefore, the invalidated action and the rule that inferred the action are written into the working memory M. Even if the same shape change target is selected at the next inference and the invalid action is selected, the invalid action will not be applied, and the next highest priority action that is determined to be appropriate will be applied. This data is then used to change the appropriate target shape data. As a result, if the next axilin applied to achieve the shape change target is determined to be effective, the priority of the action is raised and the priority of the invalid action is lowered. The priority changed as described above is written in the priority section of the Axilan inference knowledge base D.

In this way, the aluminum foil rolling target shape adjusting device 1
Is there a problem with aluminum because the target shape obtained by this method has no effect on the actual shape? Even if the actual shape of I53 continues, in the current inference, an action different from the previous action is selected even if the shape change target is the same as in the previous inference. As a result, invalid rules are not repeatedly applied, and the actual shape is appropriately changed.

The action candidate inference unit 11 checks in detail whether or not the action proposed as a candidate is appropriate according to the check tree shown in FIG. be registered.

In addition, in the action inference knowledge base D3, a rule set consisting of a plurality of rules having the same shape change target in the condition part is set for each shape change target as described above, and a main unit that stores the rule set is set in the action inference knowledge base D3. A save area is secured to store the area and the rules removed from the rule set. Note that the rules for which inference processing has been completed are sequentially removed from the main area to the save area.

In the check tree shown in the figure, first, check whether a rule that matches the currently selected shape change target and the shape change target exists in the rule set to which the rule belongs, that is, the rule corresponding to the above rule set is in the main area. Case CI is checked to see if any remain. If case C1 is applicable, the process proceeds to case C3 to check additional conditions related to operating conditions, etc. in the condition section of the rule.

If the case C3 does not apply, that is, if all the condition parts including the incidental conditions do not match, the rule is removed from the rule set in the main area to the save area. If the above-mentioned supplementary conditions are satisfied, in case C4, whether the action of the rule in which all the above-mentioned condition parts were satisfied has been applied in the past and its effect was not recognized;
That is, it is checked whether it is currently registered in the working memory M4. If case C4 is applicable, that rule is removed from the rule set (main area). If case C4 does not apply, the existence of mutually contradictory actions is checked in case C5. As described above, not only one shape change target is set, but a plurality of shape change targets may be selected. In this case, the actions associated with each shape change target often conflict with each other.

Therefore, as a result of the check, if there is an action that is inconsistent with the currently checked action among the actions that are considered to be valid and have already been registered in the working memory M%, the rule that has one of the actions , if the importance of the currently selected shape change target is greater, the action of that rule is registered in the work memory M5, and the already registered contradictory action is removed to the save area.

On the other hand, if the importance of the shape target with the already registered action is greater, the registered action is left in the working memory MS, and the rule with the currently checked action is added to the rule set. (main area) to the save area.

If case C5 does not apply, that is, if the above-mentioned actions do not contradict each other, it is checked in case C1 whether the shape change targets of these actions are the same. If case C6 applies, a check is made to see if there is a difference in priority between the above actions (case C1), and if there is a difference in priority, only the action with the higher priority is left in working memory M5. , the lower priority action is deleted from the working memory M.

If there is no difference in priority, both actions are left together in the working memory Ms.

As described above, if case CI does not apply, it is determined that there is no rule in the rule set that matches the shape change target selected this time, and the process proceeds to check case C8.

If case C8 applies, that is, all rules in each rule set have been checked, but if it is determined that all action candidates are nonconforming due to the checks in cases C8 to Cma, past invalid information in working memory M4 is At the same time, since it is not possible to select an appropriate action candidate under the current priority, reset the priority and select from the rule with the reset action candidate of the priority. After recovering each rule set, the action check reasoning is restarted from the beginning.

On the other hand, if case C8 does not apply, then the action candidates that are judged to be valid after verifying all the rules in each rule set should have been selected and stored in the working memory Ms, so the inference regarding the action check is end.

In this way, rules in a rule set consisting of rules that have a certain shape change goal in common are sequentially removed from the rule set (main area) after checking the validity of their actions one by one. Stored in the evacuation area.

Then, the above rule check is executed until the rule set of the main area related to the shape change target becomes empty. In other words, the case where the rule set is empty does not correspond to case C1.

If two or more shape change targets are selected,
The validity check of the above action is also executed for another shape change target set in the same manner as above (01 to C).

In this way, the above-mentioned check routine checks the inconsistency, priority, effectiveness record, etc. between the action candidates already registered in the working memory M5 and the action candidates to be newly registered, and then checks the target shape data. The validity of the actions to be applied to the changes and the consistency of the rules are maintained.

■ Generation of target shape Next, the action and its degree registered in the work memory M5, that is, the "edge level" is raised.

Target shape change data such that the degree is 0.8'' is transferred to the target shape generation unit 12.

The target shape generation unit 12 changes the values of the target shape adjustment parameters shown in Table 4 and FIGS. 12(a) to 12(b) based on the target shape change data. , the aforementioned Axilan raises the ′ end level. If the degree is 0.8'', the target shape adjustment parameter a related to the elongation rate of the end portion.

The value of is changed and set according to the degree of the action, and the target shape data changes. Further, the shape control section 3 controls the coolant 58 of the roll rolling mill 2 based on the input changed target shape data.

■Evaluation of applied action Then, the operating condition data including the newly obtained actual shape data is input to the rolling data collection section 7, and the same process as the previous time is repeated. That is, the shape change target and its importance for the current actual shape are calculated by the control target generation unit 9, and compared with the previous one by the action candidate inference unit 11.

A specific example in which the above reasoning is repeated is shown in FIG. 1 if the importance is 0.8
The second inference E2 is executed, and the action candidates at that time are (A1) widening the width of the zero point and (A2) moving the zero point outward. Here, axilan A with high priority
1 was applied, and the target shape data before and after being changed thereby were as shown in the inference result E3.

However, when evaluating the effect of Axilan A1, E
The importance of l did not become smaller than 0.8, and its effect was not recognized. Therefore, the rule from which the action AI was derived is stored in the working memory M4 as being invalid.

Next, apply Axilan A2, which has the next highest priority, and perform 2
The second inference E4 is executed. In this case, the rule including the shape change target E is continuously applied to continue the rolling process, and if the operating conditions are unchanged from the first inference, the Axilan AI and action A2 are obtained as candidates. Therefore, as a result of verifying the history of the axilan effect using this second inference E4, it was found that the axilan AI was invalid last time and was stored in the working memory M4, so it was not applied this time (case C4 in Figure 11). ),
Next, Axilan A2 having a higher priority is applied to change the target shape data.

As a result, if it is determined that the quota growth has been improved, the priority of action A2 is upgraded over that of Axilan A1 and stored in the action inference knowledge base D.

In this way, when certain target shape data is given to the actual shape of the aluminum foil 53, the change in the actual shape (shape change target) and the corresponding change in the target shape data (action) are determined by The target shape data changed each time the inference is made in step 11, the actual shape data obtained as a result, the operating condition data, rolling status data, control target data, etc. used to set the target shape data. Each experience value is stored in the action inference knowledge base D as an inference rule. Then, the axigon candidate inference unit 11 calculates appropriate target shape data based on the respective experience values so as to realize the ideal actual shape (operation policy) at that time, and converts the target shape data into a target shape. It is automatically created via the shape generation section 12 and output to the shape control section 3.

As shown in FIG. 14, the aluminum foil rolling target shape adjusting device l as described above is started by the operator 5 pressing a key from the rolling mill side terminal 6 (step 21).Subsequently, the operator 5 In accordance with the input menu (FIG. 15) displayed on the screen of the terminal 6, the shape change target information is input together with the degree of importance (step 22).

Accordingly, the aluminum foil rolling target shape adjusting device 1 analyzes the rolling data transferred from the shape control unit 3 (step 23), creates an appropriate target shape by inference (step 24), and performs the target shape before and after modification. (FIG. 16) is displayed on the screen, and the corrected target shape data is output to the rolling mill 2 via the shape control section 3 (step 25).

Then, it is evaluated whether or not the corrected target shape was effective for the actual shape that had the problem (step 26), and the system waits for an input from the operator 5.

At this time, the target shape adjustment device 1 also automatically generates a shape change target, but if it is inconsistent with the shape change target input by the operator 5, the shape change target is input by the operator 5. The shape change target may be generated based on a rule that gives priority to either the shape change target or the shape determination result from the actual shape classification item.

The aluminum foil rolling target shape adjusting device 1 constantly monitors the actual shape of the aluminum foil 53, and when it is determined that it is necessary to change the target shape data, the operator 5 starts the inference, but it does not automatically For example, in the condition part of the rule, a condition for comparing the certainty factor of the actual shape classification item with a certain threshold value α is set. A specific example of the rule is given below.

If [actual shape classification item and its confidence level] and [operating conditions], then [designation of shape change target and its degree],
Specifically, if the confidence level of edge extension is <α and the bus is the second bus, then the two-way importance of wanting to extend the edge is expressed as 1.0. That is, when the actual shape changes and the degree of edge extension becomes equal to or less than the threshold value α, the above rule is applied and inference is started.

Furthermore, the aluminum foil rolling target shape adjusting device 1 has a function of improving even when the obtained actual shape data is asymmetrical in the foil width direction. As shown in FIGS. 17 and 18, if it is determined that the actual shape data is asymmetric (step 31),
The target shape, which has been symmetrical until now, is temporarily set higher by the operator 5 (step 32), which temporarily deflects the injection amount distribution of the coolant 58 injected to the rolling roll 52 in the foil width direction, This is for uniformizing the heat distribution in the rolling rolls 52, and is particularly effective at the start of operation (while the roll temperature is rising) or at the time of restart (after the rolls are rearranged). The process in step 32 is repeated until the actual shape data becomes symmetrical.

As described above, even when the operating conditions of the roll rolling mill 2 change slightly, the aluminum foil rolling target shape adjusting device 1 according to the present embodiment changes and sets the target shape in an appropriate and timely manner according to the change. The process to do so will be automatically started. Thereby, aluminum foil 53 having a predetermined actual shape can be stably produced.

The aluminum foil rolling target shape adjustment device 1 uses an element 4. in which a piezoelectric element is embedded as a sensor for detecting actual shape data of elongation and tension at the time of rolling. was adopted, but
Said element 4. It is also possible to use a plurality of air bearing type elements whose appearance is approximately - as the sensor, and to detect the actual shape data based on the change in the air pressure.

Furthermore, the material to be controlled is not limited to the aluminum foil 53, but may be copper or other metal, and its thickness is not limited.

〔Effect of the invention〕

As described above, the present invention provides a system drive device for driving a control system etc. for a predetermined plant.
A sensor group for detecting plant data related to one or more types of plant states, a storage means for storing the plant data, a certainty calculation means for calculating certainty for each plant state from the plant data, and the plant state. an importance calculating means for comparing the confidence level for each state with a first threshold value for each state, selecting plant data exceeding the first threshold value and calculating the importance thereof; and calculating a composite importance from the respective importance levels. , and a control means for driving the control system etc. when the combined importance exceeds a predetermined second threshold value, so that the operating state of the plant changes slightly. Even in such a case, the control system, etc. can be automatically operated in an appropriate and timely manner according to the change.

As a result, stable plant operating conditions can be maintained.

[Brief explanation of drawings]

FIG. 1 is a schematic diagram showing the system layout of an aluminum foil rolling target shape adjusting device according to an embodiment of the present invention, and FIG. 2 is a configuration diagram showing a processing flow of the aluminum foil rolling target shape adjusting device.
Figure 3 (The figure is an explanatory diagram showing the main parts of the actual shape data expressed by the elongation weight distribution in the foil width direction. Figure 3 (b) is an explanatory diagram showing the actual shape classification items of aluminum foil classified by pattern. , FIG. 4(a) is a flowchart showing the processing procedure for determining the change tendency of actual shape data, and FIG. The explanatory drawings shown in Fig. 5(a) and the same figure (
c) is a graph showing the correlation between the number of levels of two real shape classification items, Figure 6 is a three-dimensional graph showing a three-dimensional pattern due to changes in real shape data of aluminum foil over time, and Figure 7 is a conceptual neural network. Figure 8 is an explanatory diagram showing an example of the relationship between shape change targets and corresponding action candidates for actual shape classification items, and Figure 9 shows the importance of shape change targets and the actual shape classification items. A graph showing the relationship with the confidence level when it is tense. Figure 10 is an explanatory diagram showing the rules used for inference in the axilan candidate inference unit and an example of changing the target shape using the rules. Figure 11 is the application. Figure 12 is an explanatory diagram showing the processing procedure of a routine that checks the validity of the action to be performed using a check tree (Figure 12 is an explanatory diagram showing the target shape adjustment parameters used to change the target shape; (B) is a state diagram showing a change in the situation of the parameter 83, and (C) is a state diagram showing a situation where the center part of the target shape adjusted by the parameter a4 is a forward pattern. ) is a state diagram showing a situation where the central part has a reverse pattern, No. 13
The figure is a schematic explanatory diagram showing an example of inference execution for changing the target shape, Figure 14 is a flow sheet showing the processing flow of target shape adjustment, and Figure 15 is an input menu displayed on the screen of the terminal on the rolling mill side. FIG. 16 is a display diagram showing an example of the target shape displayed on the screen, FIG. 17 is a flow sheet showing a method for correcting an asymmetrically obtained actual shape, and FIG. 18 is a display diagram showing an example of the target shape displayed on the screen. 19 is a schematic perspective view showing a roll rolling mill, which is an example of the background of the present invention, and FIG. 20 is an external appearance showing the surface shape of aluminum foil after rolling. 21 is an explanatory diagram showing the correlation between the cross-sectional shape of the rolling roll, the actual shape of the aluminum foil, and the target shape for controlling the actual shape, and FIG. 22 is an explanatory diagram showing the correlation between the operational target shape of the aluminum foil and the This is a graph simultaneously showing the target shape set in the process. [Explanation of symbols] 1... Aluminum foil rolling target shape adjusting device 2... Roll rolling mill 7... Rolling data collection unit (sampling means) M,
... Working memory (sampling means) 8 ... Rolling situation analysis section (confidence calculation means) (plant state classification means) (characteristic information calculation means) 9 ... Control target generation section (importance calculation means) 11.・・・
Axilan candidate inference unit (control means) 53... Aluminum foil Dl... Rolling situation analysis knowledge base (pattern storage means) L,... First threshold value

Claims (1)

[Claims] 1. A system drive device that drives a control system or the like for a predetermined plant, comprising: a sensor group that detects plant data related to one or more types of plant states; and a storage unit that stores the plant data. and a certainty calculation means for calculating a certainty for each plant state from the plant data; and comparing the certainty for each plant state with a first threshold for each state, and determining a plant state that exceeds the first threshold. an importance calculating means for selecting and calculating the importance; and a control means for calculating a composite importance from the respective importance and driving the control system etc. when the composite importance exceeds a predetermined second threshold. A system driving device comprising: 2. A system drive device that drives a control system or the like for a predetermined plant, including a sensor group that detects plant data related to one or more types of plant conditions, a storage device that stores the plant data, and a system that stores the plant data. A plant state classification means for classifying feature patterns of a plurality of plant states in advance; a pattern storage means for storing the feature patterns of the plant states; and a confidence factor for calculating confidence for each feature pattern of each plant state from the plant data. a calculation means; and an importance calculation means for comparing the reliability of each feature pattern of the plant state with a first threshold value for each pattern, selecting a feature pattern of the plant state that exceeds the first threshold value, and calculating the importance of the feature pattern. and a control means for calculating a composite importance level from the respective importance levels and driving the control system etc. when the composite importance level exceeds a predetermined second threshold value. Device. 3. A system drive device that drives a control system, etc. for a predetermined plant, including a sensor group that detects plant data related to one or more types of plant conditions, and a sampling means that periodically collects and stores the plant data. , a characteristic information calculating means for calculating characteristic information of the plant state within a predetermined time from the plant data; a confidence calculating means for calculating a certainty for each plant state according to the characteristic information of the plant state; an importance calculating means for comparing the confidence level of each plant state with a first threshold value for each state, selecting a plant state that exceeds the first threshold value, and calculating the importance thereof; and calculating a composite importance from the respective importance levels. 1. A system driving device comprising: a control means for calculating and driving the control system etc. when the combined importance exceeds a predetermined second threshold value.