CN108687137B - complete equipment control device, rolling mill control device, control method and storage medium - Google Patents

complete equipment control device, rolling mill control device, control method and storage medium Download PDF

Info

Publication number
CN108687137B
CN108687137B CN201810267684.XA CN201810267684A CN108687137B CN 108687137 B CN108687137 B CN 108687137B CN 201810267684 A CN201810267684 A CN 201810267684A CN 108687137 B CN108687137 B CN 108687137B
Authority
CN
China
Prior art keywords
control
plant
actual data
output
learning
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201810267684.XA
Other languages
Chinese (zh)
Other versions
CN108687137A (en
Inventor
服部哲
高田敬规
田内佑树
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hitachi Ltd
Original Assignee
Hitachi Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hitachi Ltd filed Critical Hitachi Ltd
Publication of CN108687137A publication Critical patent/CN108687137A/en
Application granted granted Critical
Publication of CN108687137B publication Critical patent/CN108687137B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
    • B21B—ROLLING OF METAL
    • B21B37/00—Control devices or methods specially adapted for metal-rolling mills or the work produced thereby

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Feedback Control In General (AREA)
  • Control Of Metal Rolling (AREA)

Abstract

the present invention relates to a plant control device, a rolling mill control device, a plant control method, a rolling mill control method, and a storage medium storing a program for learning an optimal operation method for actual data without deteriorating a state of a plant to be controlled. The present invention is a control method for a mode in which a control target complete equipment recognizes a combination of actual data of the control target complete equipment, the control method including: a control method learning device for learning a combination of actual data and control operation of the whole set of the apparatus to be controlled; and a control execution device for controlling the entire equipment to be controlled based on a combination of the learned actual data and the control operation, wherein the control execution device includes a control rule execution unit, a control output determination unit, and a control output suppression unit, and the control method learning device includes a control result satisfaction determination unit, a learning data creation unit, and a control rule learning unit.

Description

Complete equipment control device, rolling mill control device, control method and storage medium
Technical Field
The present invention relates to a plant control device and a control method thereof, a rolling mill control device and a control method thereof, and a storage medium storing a program, in real-time feedback control using an artificial intelligence technique such as a neural network.
Background
Conventionally, plant control based on various control theories has been performed in various plants in order to obtain a desired control result by the control.
As an example of the overall plant, for example, in the rolling mill control, as an example of the control, the fuzzy control and the neuro-fuzzy control are applied as the control theory for the shape control for controlling the rolling state of the plate. Fuzzy control is applied to shape control using a coolant, and in addition, neuro-fuzzy control is applied to shape control of the sendzimir mill. As disclosed in patent document 1, a similarity ratio between a difference between an actual shape pattern and a target shape pattern detected by a shape detector and a preset reference shape pattern is obtained, and a control output quantity for an operation end is obtained by using a control rule expressed by an operation quantity of the control operation end for the preset reference shape pattern based on the similarity ratio, thereby performing shape control to which a neuro-fuzzy control is applied. As a conventional technique, the shape control of the sendzimir mill using the neuro-fuzzy control is used.
Fig. 5 shows shape control of the sendzimir mill described in fig. 1 of patent document 1. The neuro-fuzzy control is used in the shape control of the sendzimir mill. In this example, the pattern recognition means 51 performs pattern recognition of the shape based on the actual shape detected by the shape detector 52, and calculates which of the reference shape patterns set in advance the actual shape is closest to. The control arithmetic means 53 performs control using a control rule constituted by a control operation end operation amount for a preset shape pattern as shown in fig. 6. More specifically, referring to fig. 6, the pattern recognition means 51 calculates which shape pattern (e) the difference (Δ e) between the actual shape result detected by the shape detector 52 and the target shape (e ref) is closest to 1 to 8, and the control calculation means 53 selects and executes any one of the control methods 1 to 8.
However, in the method of patent document 1, in order to verify the control rule, the operator may manually verify the control rule during rolling, and the like, and the shape change contrary to the expectation may be caused. In other words, the control rule determined as described above may not be realistic. The reason for this is that the study of mechanical characteristics is insufficient, the operating state of the rolling mill and the change of mechanical conditions, but if one rule verifies whether or not the preset control rule is the optimum one, the conditions to be considered are many and difficult. Therefore, if the control rule is once set, it is usually kept as it is as long as there is no failure.
when the control rule is not satisfied due to a change in the operation condition or the like, it is difficult to output a control accuracy of a certain degree or more because the control rule is fixed. Further, once the shape control is operated, the operator does not perform a manual operation (interference with the control), and therefore it is difficult to find a new control rule by manual intervention of the operator. Further, even when a new standard rolled material is rolled, it is difficult to set a control rule corresponding to the material.
As described above, in the conventional shape control, since the control is performed using a preset control rule, there is a problem that it is difficult to correct the control rule.
In order to solve this problem, as shown in patent document 2, the following is achieved by learning a rule that improves the shape by randomly changing a control rule while performing shape control:
1) A new control rule is found while performing shape control during rolling.
2) Since a new control rule is not predictable in advance and a control rule that cannot be predicted at all is sometimes optimal, the control operation terminal is operated at random and searched for while observing the control result.
Documents of the prior art
Patent document 1: japanese patent No. 2804161
Patent document 2: japanese patent No. 4003733
Disclosure of Invention
in the above-described conventional technique, a representative shape is set in advance as a reference shape pattern, and control is performed according to a control rule indicating a relationship with a control operation end operation amount for the reference waveform pattern. The learning of the control rule is also related to the control operation end operation amount for the reference waveform pattern, and a predetermined representative reference shape pattern is directly used. Therefore, there is a problem of shape control that reacts only to a specific shape pattern.
The reference shape pattern is determined by a person in advance based on knowledge about the target rolling mill, accumulated actual shape results, and experience of manual intervention operations, and it is difficult to cover all shapes generated in the target rolling mill and the material to be rolled. Therefore, when a shape different from the reference shape pattern occurs, control by shape control is not performed, and a shape deviation is not suppressed and remains, or a similar reference shape pattern is erroneously recognized, and an erroneous control operation is performed, and the shape is deteriorated.
As described above, in the conventional shape control, since the control rule is learned and controlled using the preset reference shape pattern and the control rule for the reference shape pattern, there is a problem that improvement of the control accuracy is limited.
as described above, the present invention provides "a plant control apparatus that recognizes a pattern of a combination of actual data of a plant to be controlled, and performs control for the plant to be controlled, the plant control apparatus including:
A control method learning device for learning a combination of actual data and control operation of the whole set of the apparatus to be controlled; and
a control execution device for controlling the whole set of equipment to be controlled according to the combination of the actual data and the control operation,
The control execution device includes:
a control rule execution unit that provides a control output in accordance with the determined combination of the actual data and the control operation of the plant to be controlled;
A control output determination section that determines whether the control rule execution section can output the control output, and notifies the control method learning device that the actual data and the control operation are erroneous; and
a control output suppressing unit that prevents the control output from being output to the plant to be controlled when the control output determining unit determines that the actual data of the plant to be controlled is deteriorated when the control output is output to the plant to be controlled,
The control method learning device includes:
a control result adequacy determination unit that determines whether the control result regarding whether the actual data is better or worse than before the control is good or worse after a time delay until the control effect is exhibited by the actual data when the control execution device actually outputs the control output to the plant to be controlled;
A learning data creation unit that obtains teacher data by using the control result quality and the control output in the control result quality determination unit; and
a control rule learning unit for learning the actual data and the teacher data as learning data,
The control method learning device performs learning to obtain individual combinations of actual data and control operations for a plurality of control targets based on the states of the control target plant, and uses the obtained combinations of actual data and control operations as the determined combinations of actual data and control operations of the control target plant in the control rule execution unit. ".
The present invention is a rolling mill control device to which a complete equipment control device is applied, wherein the complete equipment to be controlled is a rolling mill, and the actual data is the shape of the output side of the rolling mill. ".
In addition, the present invention is "a plant control method applied to a plant control apparatus for performing control in a mode in which a plant to be controlled recognizes a combination of actual data of the plant to be controlled, the plant control apparatus including:
A control method learning device for learning a combination of actual data and control operation of the whole set of the apparatus to be controlled; and
A control execution device for controlling the whole set of equipment to be controlled according to the combination of the actual data and the control operation,
The control execution means provides a control output in accordance with the determined combination of the actual data and the control operation of the plant to be controlled, determines whether the control output can be output, notifies the control method learning means that the actual data and the control operation are erroneous, and prevents the control output from being output to the plant to be controlled in the case where it is determined that the actual data of the plant to be controlled is deteriorated in the case where the control output is output to the plant to be controlled,
A control method learning device determines whether the control result regarding whether the actual data is better or worse than before the control after a time delay until the control effect is expressed in the actual data when the control execution device actually outputs the control output to the control target plant, obtains teacher data using the control result and the control output, learns the actual data and the teacher data as learning data, obtains unique combinations of actual data and control operations for a plurality of control targets according to the state of the control target plant, and uses the obtained combinations of the actual data and the control operations as the determined combinations of the actual data and the control operations of the control target plant in the control execution device. ".
In addition, the invention is a rolling mill control method, which is applied with a complete equipment control method and is characterized in that the complete equipment to be controlled is a rolling mill, and the actual data is the shape of the output side of the rolling mill. ".
the present invention is "a storage medium storing a program for realizing, by a computer system, a complete equipment control apparatus that performs control in accordance with a pattern in which a complete equipment to be controlled recognizes a combination of actual data of the complete equipment to be controlled,
The computer system is provided with:
A control method learning device for learning a combination of actual data and control operation of the whole set of the apparatus to be controlled; and
a control execution device for controlling the whole set of equipment to be controlled according to the combination of the actual data and the control operation,
The programs are a control rule execution program, a control output determination program, and a control output suppression program for realizing the processing of the control execution means, wherein the control rule execution program provides a control output in accordance with the determined combination of the actual data and the control operation of the control target whole plant; a control output decision program decides whether or not the control rule execution program can output a control output, and notifies the control method learning device that the actual data and the control operation are erroneous; a control output suppression program that prohibits output of a control output to the plant to be controlled when the control output determination program determines that the actual data of the plant to be controlled is deteriorated when the control output is output to the plant to be controlled,
The programs are a control result adequacy determining program, a learning data making program, and a control rule learning program for realizing the processing of the control method learning device, wherein the control result adequacy determining program is used for realizing the processing of the control result adequacy determining, wherein in the case where the control execution device actually outputs the control output to the control target complete equipment, whether the control result as to whether the actual data becomes good or deteriorated as compared with that before the control is good or bad is determined after a time delay until the control effect is expressed in the actual data; the learning data creation program obtains teacher data by using the control result and the control output in the control result adequacy determination program; the control rule learning program learns the actual data and the teacher data as learning data,
the control method learning device performs learning to obtain individual combinations of actual data and control operations for a plurality of control targets based on the states of the control target plant, and uses the obtained combinations of actual data and control operations as the determined combinations of actual data and control operations of the control target plant in the control rule execution program. ".
By using the present invention, the control rules of the shape mode and the operation method used in the shape control can be automatically corrected and optimized in the control. Therefore, the control accuracy is improved, the start period of the control unit is shortened, and the control unit can cope with the aging change.
Drawings
Fig. 1 is a diagram showing an outline of a whole plant control apparatus according to an embodiment of the present invention.
Fig. 2 is a diagram showing a specific configuration example of the control rule execution unit 10 according to the embodiment of the present invention.
Fig. 3 is a diagram showing a specific configuration example of the control rule learning unit 11 according to the embodiment of the present invention.
Fig. 4 is a diagram showing a neural network structure in the case where the present invention is used for shape control of a sendzimir mill.
Fig. 5 is a diagram showing shape control of the sendzimir mill described in fig. 1 of patent document 1.
fig. 6 is a diagram showing a control rule in shape control of the sendzimir mill described in fig. 1 of patent document 1.
Fig. 7 is a diagram showing an outline of the input data creating unit 2.
fig. 8 is a diagram showing an outline of the control output calculation unit 3.
Fig. 9 is a diagram showing an outline of the control output determination unit 5.
Fig. 10 is a diagram illustrating a shape deviation and a control method.
Fig. 11 is a diagram showing an outline of the control quality determination unit 6.
Fig. 12 is a diagram showing the relationship between the part data and the symbol in the control output calculation unit 3.
Fig. 13 is a diagram showing the processing procedure and the processing content in the learning data creating unit 7.
Fig. 14 is a diagram showing an example of data stored in the learning data database DB 2.
Fig. 15 is a diagram showing an example of the neural network management table TB.
fig. 16 is a diagram showing an example of the learning data database DB 2.
(symbol description)
1: controlling the whole set of equipment; 2: a control input data producing section; 3: a control output calculation unit; 4: a control output suppression unit; 5: a control output determination unit; 6: a control result quality judging section; 7: a learning data creation unit; 10: a control rule execution unit; 11: a control rule learning unit; 20: a control execution device; 21: a control method learning device; DB 1: a control rule database; DB 2: outputting a decision database; DB 3: a learning data database; si: actual data; SO: controlling the output of the operation amount; s1: inputting data; s2: controlling an operation instruction of an operation end; s3: controlling the operation amount; s4: controlling the operation amount to output the possibility data; s5: data for determining whether the quality is good or not; s6: data for controlling whether the result is good or not; s7a, S7b, S7 c: teacher data; s8a, S8b, S8 c: input data (for a control rule learning unit).
Detailed Description
hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings, and the idea of the present invention and the process of realizing the present invention will be described by taking shape control of a rolling mill as an example.
First, in order to solve the above problems of the present invention, it is necessary to:
1) Instead of learning the control operation method by separately setting the reference shape mode and the control operation therefor in advance, a combination of the shape mode and the control operation is learned and used to implement the control operation.
2) Since a new control rule is not predictable in advance and a control rule that cannot be predicted at all is sometimes optimal, the control operation terminal is operated at random and searched for while observing the control result.
In order to realize this, it is necessary to change the combination of the shape mode and the control operation used in the shape control and to change the control operation so that the control result becomes favorable. For this reason, it is necessary to configure a neural network capable of learning a combination of a shape pattern and a control operation, and to change an output of the control operation of the neural network for the shape pattern generated in the rolling mill according to whether a control result is good or not.
When the shape control is performed on the rolling mill in operation, an erroneous control output may be output, and thus the shape may deteriorate and an operation abnormality such as a plate breakage may occur. When a sheet breakage occurs, it takes time to replace a roll used in a rolling mill, and loss such as waste of a rolled material during rolling is large. Therefore, it is necessary to avoid erroneous control outputs of the rolling mill as much as possible.
From the above, in the present invention, in order to realize this, for example, the simple model of the rolling mill or the like is used to verify the quality of the control operation of the neural network output, and the output that can be considered clearly as the shape deterioration is output to the control operation side of the rolling mill to prevent the shape deterioration. At this time, the neural network performs learning by setting the control operation for the shape pattern thereof to be erroneous.
Since there is a possibility that the verification method of the control operation's acceptability or ineffectiveness itself is erroneous, the control operation output regarding the neural network determined to be erroneous with a certain probability is also output to the control operation side of the rolling mill, so that it is also possible to learn an unexpected combination of the shape pattern and the control operation.
[ example 1 ]
Fig. 1 shows an outline of a whole plant control apparatus of an embodiment of the present invention. The plant control apparatus of fig. 1 comprises: a control target whole set of equipment 1; a control execution device 20 that receives actual data Si from the plant 1 to be controlled and supplies a control manipulated variable output SO determined according to a control rule (neural network) as illustrated in fig. 6 to the plant 1 to be controlled to control the plant; a control method learning device 21 that inputs actual data Si from the plant 1 to be controlled and performs learning so that the learned control rule is reflected in the control rule in the control execution device 20; a plurality of database DBs (DB1 to DB 3); and a management table TB of the database DB.
The control execution device 20 is configured with a control input data creation unit 2, a control rule execution unit 10, a control output calculation unit 3, a control output suppression unit 4, a control output determination unit 5, and a control operation disturbance generation unit 16 as main elements.
in the control execution device 20, first, the control input data generation unit 2 is used to generate the input data S1 of the control rule execution unit 10 based on the actual data Si of the rolling mill of the plant 1 to be controlled. The control rule execution unit 10 creates a control operation terminal operation command S2 from the actual data Si of the control target using a neural network (control rule) representing the relationship between the actual data Si of the control target and the control operation terminal operation command S2. The control output calculation unit 3 calculates a control operation amount S3 to the control operation side in accordance with the control operation side operation command S2. Thus, the control manipulated variable S3 is created using a neural network based on the actual data Si of the plant 1 to be controlled.
The control output determination unit 5 in the control execution device 20 determines the control operation amount output possibility data S4 to the control operation end using the actual data Si from the plant 1 to be controlled and the control operation amount S3 from the control output calculation unit 3. The control output suppression unit 4 determines whether or not the control operation amount S3 can be output to the control operation terminal based on the control operation amount output availability data S4, and outputs the control operation amount S3 determined to be available as the control operation amount output SO to be supplied to the plant 1 to be controlled. Thereby, the control manipulated variable S3 determined to be abnormal is not output to the control target plant 1. The control operation disturbance generator 16 generates disturbance for verifying the plant control device, and supplies the generated disturbance to the plant 1 to be controlled.
the control execution device 20 configured as described above refers to the control rule database DB1 and the output determination database DB3 for the execution of the processing. The control rule database DB1 is connected to both the control rule execution unit 10 in the control execution device 20 and the control rule learning unit 11 in the control method learning device 21, which will be described later, so as to be accessible. The control rule (neural network) as a result of learning in the control rule learning unit 11 is stored in the control rule database DB1, and the control rule execution unit 10 refers to the control rule stored in the control rule database DB 1. The output determination database DB3 is connected to the control output determination unit 5 in the control execution device 20 so as to be accessible.
Fig. 2 shows a specific configuration example of the control rule execution unit 10 according to the embodiment of the present invention. The control rule execution unit 10 inputs the input data S1 created by the control input data creation unit 2, and provides a control operation end operation command S2 to the control output calculation unit 3. The control rule execution unit 10 includes a neural network 101, and the control operation side operation command S2 is basically determined by the method of patent document 1 illustrated in fig. 6 in the neural network 101. In the present invention, the control rule execution unit 10 further includes a neural network selection unit 102 that selects an optimal control rule as a control rule in the neural network 101 by referring to the control rule stored in the control rule database DB1, and starts execution. As described above, the control rule execution unit 10 of fig. 2 selects and uses a necessary neural network from a plurality of neural networks divided for each operator group and control purpose. The control rule database DB1 may further include actual data (data of a work group, etc.) Si such as a neural network that can be selected and a criterion for determining whether the control target plant 1 is good or not, as data from the control target plant 1. Further, since there is a relationship that becomes a control rule if a neural network is executed, the neural network and the control rule are not distinguished in this specification and are used in the same meaning.
Returning to fig. 1, the control method learning device 21 performs learning of the neural network 101 used in the control execution device 20, and when the control execution device 20 outputs the control manipulated variable output SO to the plant 1 to be controlled, it takes time for the control effect to actually appear as a change in the actual data Si, and therefore, learning is performed using data obtained by delaying the time by an amount corresponding to the time, and in fig. 1, Z -1 indicates an appropriate time delay function for each data.
The control method learning device 21 is configured with the control result adequacy determining unit 6, the learning data creating unit 7, the control rule learning unit 11, and the adequacy determining database DB4 as main elements.
The control result adequacy determination unit 6 determines whether the actual data Si is changing in a good direction or in a bad direction using the actual data Si and the previous value of actual data Si0 from the plant 1 to be controlled and the adequacy determination data S5 stored in the adequacy determination database DB4, and outputs the control result adequacy data S6.
The learning data creation unit 7 in the control method learning device 21 creates new teacher data S7a used for learning the neural network, using data obtained by delaying input data such as the control operation terminal operation command S2, the control operation amount S3, and the control operation amount output availability data S4 created by the control execution device 20 for the same time period, and the control result adequacy data S6 from the control result adequacy determination unit 6, and supplies the new teacher data S7a to the control rule learning unit 11. The teacher data S7a is data corresponding to the control operation terminal operation command S2 output by the control rule execution unit 10, and the learning data creation unit 7 can determine, as new teacher data S7a, data obtained by estimating the control operation terminal operation command S2 output by the control rule execution unit 10 using the control result adequacy data S6 provided by the control result adequacy determination unit 6.
Fig. 3 shows a specific configuration example of the control rule learning unit 11 according to the embodiment of the present invention. The control rule learning unit 11 is configured with the input data creating unit 114, the teacher data creating unit 115, the neural network processing unit 110, and the neural network selecting unit 113 as main components. The control rule learning unit 11 obtains data S8a obtained by delaying the input data S1 from the input data creating unit 2 by a time period as an input from the outside, obtains new teacher data S7a from the learning data creating unit 7, and refers to data stored in the control rule database DB1 and the learning data database DB 2.
in the control rule learning unit 11, the input data S1 is acquired to the neural network processing unit 110 via the input data creating unit 114 after compensating for an appropriate time delay.
in the control rule learning unit 11, the teacher data creation unit 115 supplies the new teacher data S7a from the learning data creation unit 7 to the neural network processing unit 110 as the teacher data S7c including the past teacher data S7b stored in the learning data database DB 2. These teacher data S7a and S7b are stored in the learning data database DB2 as appropriate and used.
similarly, the input data S8a from the control input data creating unit 2 is supplied to the neural network processing unit 110 in the input data creating unit 114 as the input data S8c including the total of the past input data S8b stored in the learning data database DB 2. These input data S8a and S8b are stored in the learning data database DB2 as appropriate and used.
The neural network processing unit 110 is configured by a neural network 111 and a neural network learning control unit 112, and the neural network 111 takes in the input data S8c from the input data creating device 114, the teacher data S7c from the teacher data creating unit 115, and the control rule (neural network) selected by the neural network selecting unit 113, and stores the neural network finally specified in the control rule database DB 1.
The neural network learning control unit 112 controls the input data creation device 114, the teacher data creation unit 115, and the neural network selection unit 113 at appropriate timings to obtain inputs to the neural network 111, and also controls the processing results to be stored in the control rule database DB 1.
here, the neural network 101 in the control execution device 20 of fig. 2 and the neural network 111 in the control method learning device 21 of fig. 3 are both neural networks having the same concept, and the basic concept of the use is explained as follows. First, the neural network 101 in the control execution device 20 is a predetermined content neural network, and when the input data S1 is supplied, the control operation terminal operation command S2 that is a corresponding output is obtained, and it can be said that the neural network is used for processing in one direction. In contrast, the neural network 111 in the control method learning device 21 is used to obtain a neural network that satisfies the input/output relationship by learning when the input data S1, the input data S8c of the control operation terminal operation command S2, and the teacher data S7c are set as learning data.
The basic processing method in the control method learning device 21 configured as described above is considered as follows. First, when the content of the control operation amount output availability data S4 is "ok", the control operation amount output SO is output to the control target plant 1, and when the content of the control result ok data S6 is "ok" (the actual data Si changes in the ok direction), it is determined that the control operation terminal operation command S2 output by the control rule execution unit 10 is correct, and the learning data is created such that the output of the neural network becomes the control operation terminal operation command S2.
On the other hand, when the content of the control manipulated variable output availability data S4 is "no", or when the content of the control result adequacy data S6 is "no" (the actual data Si changes in a direction of deterioration) by outputting the control manipulated variable output SO to the plant 1 to be controlled, it is determined that an error has occurred in the control manipulated variable operation command S2 output by the control rule execution unit 10, and the learning data is created SO that the output of the neural network is not output. At this time, the neural network output is configured to output 2 kinds of outputs of + direction and-direction to the same control operation terminal as the control output, and the learning data is created so that the control operation terminal operation command S2 on the side to be output is not output.
In the control rule learning unit 11 illustrated in fig. 3, the following processing is performed as a result of data processing by the neural network learning control unit 112. Here, first, learning of the neural network 101 for the control rule execution unit 10 is performed using learning data which is a combination of S8c obtained by delaying the input data S1 to the control execution device 20 by a time and the teacher data S7c created by the teacher data creation unit 115. Actually, the control rule learning unit 11 includes the same neural network 111 as the neural network 101 of the control rule execution unit 10, and an operation test is performed under various conditions to learn a response at that time, and as a result of the learning, a control rule that produces a more favorable result is obtained. Since learning needs to be performed using a plurality of learning data, the learning process is performed by taking out a plurality of past learning data from the learning data database DB2 in which the learning data created in the past is accumulated, and the learning data of the present time is stored in the learning data database DB 2. The learned neural network is stored in the control rule database DB1 for use by the control rule executing unit 10.
In the learning of the neural network, the learning may be performed by using the past learning data together every time new learning data is created, or the learning may be performed by accumulating the learning data to some extent (for example, 100 pieces of learning data) and then using the past learning data together.
In addition, the control result quality determination unit 6 performs quality determination based on the quality determination criterion from the quality determination database DB 4. In the determination of the acceptability of the control result, since the determination result differs depending on the control purpose, a plurality of neural networks corresponding to a plurality of control purposes are created, teacher data is created and learned depending on the control purpose even if the input data is the same, and a plurality of teacher data are created for 1 time of input data and used for learning of the neural networks corresponding to the respective teacher data, thereby enabling the neural networks corresponding to a plurality of control purposes to be learned at the same time. Here, the plurality of control purposes mean, for example, in the case of shape control, which portion (plate edge portion, central portion, asymmetric portion, or the like) is to be controlled preferentially in the plate width direction, and any of a plurality of control target items (plate thickness, tension, rolling load, or the like, for example) is to be controlled preferentially.
In the case of the above configuration, when the neural network 101 for controlling the rule executing unit 10 once learns, a new control operation is not performed. Therefore, the new operation method is randomly generated at a proper time by the control operation disturbance generation unit 16, and the control operation is executed by applying the control operation amount S3, thereby learning the new control method.
The overall plant control method will be described in detail below, with respect to shape control in the sendzimir mill shown in patent document 1. The shape control will be described with reference to the following specification A, B.
The specification a is a specification regarding priority, and is information having priority in the board width direction. For example, in shape control, it is often difficult to control the mechanical characteristics to a target value in the entire width direction. Therefore, the following 2 specifications a1, a2 regarding priority are set in the board width direction. Here, the priority specification a1 is "priority the board end portions", and the priority specification a2 is "priority the center portion", and control is performed in accordance with 2 priority orders, i.e., a1 and a 2. In the case of implementing the control, any of the specifications a1 or a2 regarding the priority is considered.
The specification B is a specification for coping with a condition known in advance. For example, since the relationship between the shape mode and the control method varies under various conditions, it is considered necessary to divide the relationship by, for example, defining specification B1 as a sheet width and defining specification B2 as a steel type. Since the above-described changes occur, the degree of influence on the shape of the shape manipulation end changes.
in this case, the plant 1 to be controlled is a sendzimir mill, and the actual data is the actual shape result. The sendzimir mill is a multi-roll mill for cold rolling of hard materials such as stainless steel. In the sendzimir mill, small-diameter work rolls are used to provide strong rolling of hard materials. Therefore, it is difficult to obtain a flat steel sheet. As a countermeasure, a multi-roll structure and various shape control sections are adopted. In the sendzimir mill, the upper and lower 1st intermediate rolls generally have a single cone, and in addition to being displaceable, there are also 6 split rolls and 2 rolls called AS-U. In the example described below, the detection data of the shape detector is used as the actual data Si of the shape, and the shape deviation as the difference from the target shape is used as the input data S1. The control manipulated variable S3 is set to the AS-U of #1 to # n and the roller movement amounts of the upper and lower 1st intermediate rollers.
Fig. 4 shows a neural network structure in the case of use in shape control of a sendzimir mill. Here, the neural network is the neural network 101 when used in the control rule executing unit 10, and the neural network shown in the neural network 111 when used in the control rule learning unit 11, but the structures are the same.
In the case of the shape control of the sendzimir mill shown in fig. 4, the actual data Si from the plant 1 to be controlled is the actual data of the sendzimir mill including the data of the shape detector (here, the data is output as the shape deviation which is the difference between the actual shape and the target shape), and the control input data creation unit 2 obtains the normalized shape deviation 201 and the shape deviation class 202 as the input data S1. Thus, the input layers of the neural networks 101 and 111 are composed of the normalized shape deviation 201 and the shape deviation class 202. In fig. 4, the shape deviation level 202 is input to the neural network input layer, but the neural network may be switched according to the level.
The output layer is constituted by an AS-U operation degree 301 and a 1st intermediate operation degree 302 corresponding to the AS-U and the 1st intermediate roll which are shape control operation ends of the sendzimir mill. In each operation degree, AS-U, there are an AS-U opening direction (a direction in which a roll gap (an interval between upper and lower work rolls of a rolling mill) is opened) and an AS-U closing direction (a direction in which the roll gap is closed) for each AS-U. The 1st intermediate roll has a 1st intermediate roll opening direction (a direction in which the 1st intermediate roll moves outward from the center of the rolling mill) and a 1st intermediate roll closing direction (a direction in which the 1st intermediate roll moves toward the center of the rolling mill) with respect to the upper and lower 1st intermediate rolls. For example, when the shape detector has 20 sections and the shape deviation level 202 has 3 levels (large, medium, and small), the input layer has 23 inputs. When the number of the bed plates of the AS-U is 7 and the upper and lower 1st intermediate rollers can be displaced in the plate width direction, the output layer is 18 in total of 14 AS-U operation degrees 301 and 41 intermediate operation degrees. The number of intermediate layers and the number of neurons in each layer are set in time. As will be described later with reference to fig. 8, the shape control operation end of the sendzimir mill as the output layer is configured to output 2 kinds of outputs, i.e., + direction and-direction, to the control operation ends.
fig. 10 illustrates a shape deviation and a control method. Here, fig. 10 shows a control method in the case where the shape deviation is large in the upper part, and fig. 10 shows a control method in the case where the shape deviation is small in the lower part. The height direction is the size of the shape deviation, the horizontal axis direction is the plate width direction, both sides of the plate width indicate the plate end portions, and the center indicates the plate center portion. As shown in the upper part of fig. 10, when the shape deviation is large, the entire shape is corrected preferentially to the local shape deviation in the plate width direction. On the other hand, as shown in the lower part of fig. 10, when the shape deviation is small, it is prioritized to reduce the local shape deviation.
Since it is necessary to change the control method according to the magnitude of the shape deviation, the shape deviation level 202 is set as shown in fig. 4 and supplied to the neural networks 101 and 111, and the magnitude of the shape deviation is determined. Regarding the shape deviation, for example, the result of normalization to 0 to 1 may be used regardless of the size of the shape deviation. This is an example, and it may be considered to directly input the shape deviation to an input layer of the neural network without normalizing the shape deviation, or to change the neural network itself according to the size of the shape deviation (for example, 2 neural networks are prepared, and the neural network used when the shape deviation is large and the neural network used when the shape deviation is small are distinguished).
the neural networks 101 and 111 having the configuration shown in fig. 4 described above are made to learn the operation method for the shape mode, and the learned neural networks are used to perform shape control. Even a neural network having the same structure can output different control outputs for the same shape pattern with different characteristics according to the learned condition.
Therefore, by separately using a plurality of neural networks according to other conditions of the shape actual data, it is possible to configure optimal control for various conditions. This is a correspondence for the specification B. The configuration of fig. 2 described above shows a specific example of the case where the above-described specification is performed. In the configuration example of fig. 2, a separate neural network is prepared for the neural network 101 used in the control law execution unit 10 according to the actual rolling situation, the name of the rolling mill operator, the type of steel of the material to be rolled, the sheet width, and the like, and is registered in the control law database DB 1. The neural network selection unit 102 selects a neural network that matches the condition at the time point, and sets the selected neural network as the neural network 101 of the control rule execution unit 10. As a condition at this point in time in the neural network selection unit 102, data of the plate width may be taken from the actual data Si in the plant 1 to be controlled, and the neural network may be selected in accordance with the data. In addition, if the plurality of neural networks used here have input layers and output layers as shown in fig. 4, the number of intermediate layers and the number of cells in each layer may be different.
Fig. 7 shows an outline of the control input data creation UNIT 2 for creating data S1 (normalized shape deviation 201, shape deviation class 202) To be input To the input layers of the neural networks 101 and 111, here, as actual data Si, shape detector data of a shape detector for detecting a plate shape at the time of rolling in a sendzimir mill which is the plant 1 To be controlled is input, and first, a shape deviation PP value (Peak To Peak value ) which is a difference between a maximum value and a minimum value of a detection result of each shape detector partition is obtained by the shape deviation PP value calculation device 210S PP, and in the shape deviation class calculation device 211, the shape deviation is classified into 3 classes of large, medium, and small in accordance with the shape deviation PP value S PP, a distribution in the plate width direction whose shape is an elongation of a material To be rolled, and I-t which represents an elongation in 10-5 UNITs is used as a UNIT.
PM PP PMHere, the classification is made such that the form deviation class becomes (large-to-1, medium-to-0, and small-to-0) by the establishment of the expression (1), the form deviation class becomes (large-to-0, medium-to-1, and small-to-0) by the establishment of the expression (2), and the form deviation class becomes (large-to-0, medium-to-0, and small-to-1) by the establishment of the expression (3).
[ formula 1 ]
SPP≥50I-UNIT…(1)
[ formula 2 ]
50I-UNIT＞SPP≥10I-UNIT…(2)
[ formula 3 ]
10I-UNIT＞SPP…(3)
as described above, the normalized shape deviation 201 and the shape deviation class 202 are created as input data to the neural network 101. The normalized shape deviation 201 and the shape deviation level 202 are input data S1 of the control rule executing unit 10.
Fig. 8 shows an outline of the control output calculation unit 3. The control output arithmetic unit 3 creates a control operation amount S3 AS an operation command to each shape control operation end based on a control operation end operation command S2 (corresponding to the AS-U operation degree 301 and the 1st intermediate operation degree 302 in the case of the shape control of the sendzimir mill) which is an output from the neural network 101 in the control rule execution unit 10. Here, 1 data example is shown for each of a plurality of AS-U operation degrees 301 and 1st intermediate operation degree 302, and each data is composed of a pair of data of an open direction degree and a closed direction degree.
In the control output calculation unit 3, the inputted AS-U operation degree 301 has outputs in the opening and closing directions of each AS-U, and therefore, an operation command to each AS-U is outputted by multiplying the difference between them by a conversion gain G ASU, and since the control output to each AS-U is the AS-U position change amount (unit is length), the conversion gain G ASU is a conversion gain from the degree to the position change amount.
The 1st intermediate operation degree 302, which is also input, has the 1st intermediate outer and inner outputs, and therefore, the operation command for shifting to each 1st intermediate roll is output by multiplying the difference by the conversion gain G 1ST, and since the control output to each 1st intermediate roll is the 1st intermediate roll shift position change amount (unit is length), the conversion gain G 1ST is a conversion gain from degree to position change amount.
With the above, the control manipulated variable S3 can be calculated. The control manipulated variable S3 is constituted by #1 to # nAS-U position change amounts (n is based on the number of seat plates of the AS-U roller), an upper 1st intermediate shift position change amount, and a lower 1st intermediate shift position change amount. Fig. 8 illustrates a system in which disturbance data from the control operation disturbance generation unit 16 is applied to the control operation terminal operation command S2.
Fig. 9 shows an outline of the control output determination unit 5. The control output determination unit 5 is composed of the rolling phenomenon model 501 and the shape correction adequacy determination unit 502, and obtains information of the actual data Si from the plant 1 to be controlled, the control manipulated variable S3 from the control output calculation unit 3, and the output determination database DB3, and provides the control manipulated variable output adequacy data S4 to the control operation terminal. With the above configuration, the control output determination unit 5 predicts the change in shape when the control manipulated variable S3 calculated by the control output calculation unit 3 is output to the rolling mill as the plant 1 to be controlled by inputting the change in shape to the known model of the plant 1 to be controlled (rolling phenomenon model 501 in the case of the embodiment of fig. 9), and when the shape is expected to deteriorate, suppresses the control manipulated variable output SO and prevents the shape from greatly deteriorating.
more specifically, the control manipulated variable S3 is input to the rolling phenomenon model 501 to predict the shape change based on the control manipulated variable S3, and the shape deviation correction amount prediction data 503 is calculated. On the other hand, by adding the shape deviation correction amount prediction data 503 to the shape detector data Si (the shape deviation actual data 504 at the current time point) from the plant 1 to be controlled to obtain the shape deviation prediction data 505, and by evaluating the shape deviation prediction data 505, it is possible to predict how the shape changes when the control operation amount S3 is output to the plant 1 to be controlled. The shape correction adequacy determining unit 502 determines whether the shape is changed in a good direction or in a bad direction based on the current shape deviation actual data 504 and the shape deviation prediction data 505, and obtains control operation amount output adequacy data S4.
The shape correction adequacy determination unit 502 specifically determines adequacy of the shape correction as described below. First, as shown in specifications a1 and a2 regarding priority of shape control, a weight coefficient w (i) in the board width direction is set in the output determination database DB3 for each of the specifications a1 and a2, considering the control priority in the board width direction. Using this, for example, the evaluation function J of the following expression (4) is used to determine whether the shape change is good or not. In equation (4), w (i) is a weight coefficient, ε fb (i) is a shape deviation actual 504, ε est (i) is a shape deviation prediction 505, i is a shape detector partition, and rand is a random number term.
[ formula 4 ]
When the evaluation function J of the formula (4) is used, the evaluation function J becomes positive when the shape is good, and the evaluation function J becomes negative when the shape is deteriorated. The rand is a random number term, and changes the evaluation result of the evaluation function J randomly. Thus, even when the shape is deteriorated, the evaluation function J is still positive, and therefore, the relationship between the shape pattern and the control method can be learned even when the rolling phenomenon model 501 is not correct. Here, the rand is changed as appropriate so as to increase the maximum value when the model of the plant 1 to be controlled is not reliable as in the beginning of the trial operation and set to 0 when stable control is to be performed by a certain learning control method.
In the shape correction adequacy determining unit 502, the evaluation function J is calculated to output the control manipulated variable output availability data S4 so that the control manipulated variable output availability data S4 becomes 1 (available) when J is equal to or greater than 0 and the control manipulated variable output availability data S4 becomes 0 (no) when J is less than 0.
the control output suppression unit 4 determines whether or not to output the control operation amount output SO to the control target plant 1 based on the control operation amount output availability data S4 as a result of the determination by the control output determination unit 5. The control manipulated variable output availability data S4 is a #1 to # nAS-U position change amount output, an upper 1st intermediate shift position change amount output, and a lower 1st intermediate shift position change amount output, and is determined as follows.
IF (control manipulated variable output availability data S4 is 0) THEN
Position change amount output of #1 to # nAS-U is 0
Upper 1st intermediate shift position change amount output equals 0
The lower 1st intermediate shift position change amount output is equal to 0
ELSE
the output of the position change amounts #1 to # nAS-U is the position change amounts #1 to # nAS-U
upper 1st intermediate shift position change amount output equal to upper 1st intermediate shift position change amount
The lower 1st intermediate shift position change amount output is the lower 1st intermediate shift position change amount
ENDIF
the control execution device 20 executes the above calculation based on the actual data Si from the plant 1 (rolling mill) to be controlled, and outputs the control manipulated variable output SO to the plant 1 (rolling mill) to be controlled, thereby performing the shape control.
Next, the outline of the operation of the control method learning device 21 will be described, in the control method learning device 21, time delay data of data used in the control execution device 20 is used, the time delay Z -1 means e -TS, which indicates that a preset time T is delayed, since the plant 1 to be controlled has a time response, there is a time delay before the actual data changes according to the control operation amount output SO, therefore, learning is performed using actual data at a time point when the delay time T elapses after the control operation is executed, in the shape control, since several seconds are required until the shape meter detects a shape change after the operation command is output to the AS-U and the 1st intermediate roll, T can be set to about 2 to 3 seconds (the delay time also changes according to the type of the shape detector and the rolling speed, SO the optimum time until the change of the control operation end to the shape change is set to T).
Fig. 11 shows an outline of the operation of the control quality determining section 6, and the shape change quality determining section 602 uses the following equation of the quality determination evaluation function J C.
[ FORMULA 5 ]
In the formula (5), ∈ fb (i) is actual shape deviation data included in the actual data Si, ∈ last (i) is the previous value of the actual shape deviation data, and wc (i) is a weighting coefficient in the width direction for determining the quality. Here, the weight coefficient wc (i) for the determination of the quality is set based on the quality determination database DB4 in accordance with the specifications a1 and a2 regarding the priority of the control. Whether the control result is good or not is determined by the good or not determination evaluation function Jc. In addition, in the case where the control operation amount output availability data S4, which is the determination result of the control output determination unit 5, is 0 (control output is not available), the shape is determined to be deteriorated although the control operation amount output to the control target plant 1 actually becomes 0.
here, when the control manipulated variable output availability data S4 is 0, the control result satisfaction data S6 is-1. In addition, according to the threshold condition (LCU is more than or equal to 0 and more than or equal to LCL), the upper limit LCU of the threshold and the addition and subtraction LCL of the threshold are preset. In this case, if the result of comparison with the health assessment evaluation function Jc is Jc > LCU, the control result health assessment data S6 is set to-1 (shape degradation), if LCU ≧ Jc ≧ 0, the control result health assessment data S6 is set to 0 (shape change in the direction of degradation), if 0> Jc ≧ LCL, the control result health assessment data S6 is set to 1 (shape change in the direction of health), and if Jc < LCL, the control result health assessment data S6 is set to 0 (shape good).
Here, the control result ok or ok data S6 — 1 indicates that the outputted control output is suppressed due to the shape deterioration, the control result ok or ok data S6 indicates that the outputted control output is maintained due to no shape change or good shape, and the control result ok or ok data S6 — 1 indicates that the outputted control amount is increased due to the possibility of further improving the shape although the shape is changed in the good direction.
In this way, the quality determination evaluation function Jc differs because the weighting coefficient wc (i) in the board width direction changes according to the specifications a1 and a2 regarding the priority of control. Therefore, the determination result of the data S6 regarding the control result as being good or not good is also different. Therefore, the control method learning device 21 determines the control result adequacy data S6 for 2 types of specifications a1 and a2 regarding the priority of control.
Next, an outline of the learning data creating unit 7 will be described. As shown in fig. 1, the learning data creation unit 7 creates teacher data S7a for the neural network 111 used in the control rule learning unit 11, based on the determination result (control result adequacy data S6) from the control result adequacy determination unit 6, using the control operation terminal operation command S2, the control operation amount S3, and the determination result (control operation amount output availability data S4) of the control output suppression unit.
The teacher data S7a in this case is the AS-U operation degree 301 and the intermediate operation degree 302, which are outputs from the output layer of the neural network 111 shown in fig. 4. The learning data creation unit 7 creates teacher data S7a for the neural network 111 used in the control rule learning unit 11, using the control operation end operation command S2(AS-U operation degree 301, 1 intermediate operation degree 301) AS the output of the neural network 101 and the #1 to # nAS-U position change amount output, the upper 1st intermediate shift position change amount output, and the lower 1st intermediate shift position change amount output AS the control operation amount output SO.
In explaining the outline of the operation of the learning data creating unit 7, the relationship between the part data and the symbols in the control output calculating unit 3 in fig. 8 is shown in fig. 12. Here, the control operation side operation command S2, which is an output of the neural network 101, representatively shows the AS-U operation degree 301, and data on the positive side of the operation degree is OPref, data on the negative side of the operation degree is OMref, the randomly generated operation degree from the control operation disturbance generation unit 16 is the operation degree random number Oref, the conversion gain is G, and the control operation amount output SO is Cref. In this way, for the sake of simplicity, the output of the output layer of the neural network 101 from the control rule execution unit 10 is set to the positive side and the negative side of the degree of operation, and the degree of operation that occurs randomly from the control operation disturbance generation unit 16 is set to the random number of the degree of operation. The control manipulated variable output SO to the control manipulation end is set as a manipulation command value.
Fig. 13 shows the processing procedure and the processing content in the learning data creation unit 7. Here, if the explanation is made in accordance with the convention of symbols in fig. 12, the operation command value Cref is obtained by expression (6) in the first processing step 71.
[ formula 6 ]
Cref＝G·(OPref-OMref+ORref)…(6)
In the next processing step 72, the operation command value Cref is corrected to be C' ref based on the control result good/bad data S6. Specifically, the correction value C' ref of the operation command value Cref is set by expression (7) when the control result ok data S6 is-1, expression (8) when the control result ok data S6 is 0, and expression (9) when the control result ok data S6 is 1.
[ formula 7 ]
[ formula 8 ]
C′ref＝Cref…(8)
[ formula 9 ]
In step 73, the operation degree correction amount Δ Oref is obtained from the corrected operation command value C' ref by equations (10) and (11).
[ formula 10 ]
C′ref＝G·((OPref+ΔOref)-(OMref-ΔOref))…(10)
[ formula 11 ]
in the processing step 74, teacher data OP 'ref and OM' ref to the neural network 111 are obtained by equation (12).
[ formula 12 ]
as described above, the learning data creation unit 7 calculates the operation command value correction value C' ref based on the control result adequacy data S6, which is the determination result in the control result adequacy determination unit 6, with respect to the operation command value Cref actually output to the plant 1 to be controlled, as shown in fig. 12. Specifically, when the control result OK data S6 is 1, the control direction is OK, but when it is determined that the control output is insufficient, the operation command value is increased by Δ Cref in the same direction. Conversely, when the control result ok data S6 is-1, if it is determined that an error has occurred in the control direction, the operation command value is decreased by Δ Cref in the opposite direction. Since the conversion gain G is set in advance and is known, if the values on the positive side and the negative side of the degree of operation are known, the correction amount Δ Oref can be obtained. Here, Δ Cref is set by obtaining an appropriate value in advance by simulation or the like. Through the above steps, the teacher data OP 'ref and OM' ref used in the control rule learning unit 11 can be obtained by the above equation (12).
Note that, although a simple example is described in fig. 13, all of the AS-U operation degree 301 for #1 to # nAS-U and the 1st intermediate operation degree 302 for the upper 1st intermediate roll shift and the lower 1st intermediate roll shift are actually implemented AS teacher data (AS-U operation degree teacher data and 1 intermediate operation degree teacher data) of the neural network 111 used in the control rule learning unit 11.
Fig. 14 shows an example of data stored in the learning data database DB 2. To learn the neural network 111, many combinations of the input data S8a and the teacher data S7a are required. Therefore, the teacher data S7a (AS-U operation degree teacher data, 1st intermediate operation degree) created by the learning data creation unit 7 and the time lag data S8a of the input data S1 (normalized shape deviation 201 and shape deviation level) input to the control rule execution unit 10 in the control execution device 20 are combined and stored AS a set of learning data in the learning data database DB 2.
Note that, various databases DB1, DB2, DB3, and DB4 are used in the plant control apparatus of fig. 1, and fig. 15 shows a configuration of a neural network management table TB for performing management operations in association with the databases DB1, DB2, DB3, and DB 4. The management table TB includes a standard management table. Specifically, in the management table TB, the specifications are classified according to (B1) sheet width, (B2) steel type, and the specifications a1, a2 regarding the priority of control. As the plate width (B1), for example, 4 divisions of 3 feet wide, a meter wide, 4 feet wide, and 5 feet wide are used, and as the steel type, 10 divisions of steel type (1) to steel type (10) are used. Further, 2 types of specifications a1 and a2 are set as the control priority specifications a. In this case, 80 divisions are provided, and 80 neural networks are used separately according to rolling conditions.
The neural network learning control unit 112 associates the learning data, which is a combination of the input data and the teacher data shown in fig. 14, with the corresponding neural network No. (number) in accordance with the neural network management table TB shown in fig. 15, and stores the learning data in the learning data database DB2 shown in fig. 16.
The control execution device 20 creates 2-group learning data when executing the shape control of the plant 1 to be controlled. The reason for this is that 2 types of teacher data are created for the same input data and control output, since the determination of the control result as to whether the control result is good or not is made using 2 evaluation criteria, that is, the specification a1 and the specification a2, regarding the priority of control. The neural network learning control unit 112 instructs the neural network 111 to learn if teacher data is accumulated to some extent (for example, 200 sets) or newly accumulated in the learning data database DB 2.
in the control rule database DB1, a plurality of neural networks are stored in accordance with the management table TB shown in fig. 15, the neural network learning control unit 112 specifies a neural network No. to be learned, and the neural network selection unit 113 extracts the neural network from the control rule database DB1 and sets it as the neural network 111. The neural network learning control unit 112 instructs the input data creation unit 114 and the teacher data creation unit 115 to extract the input data and the teacher data corresponding to the neural network from the learning data database DB2, and to use them to perform learning of the neural network 111. Various methods have been proposed for learning the neural network, and any method may be used.
When the learning of the neural network 111 is completed, the neural network learning control section 112 writes back the neural network 111 as a learning result to the position of the neural network No. of the control rule database DB1, and the learning is completed.
The learning may be performed at fixed time intervals (for example, daily) for all the neural networks defined in fig. 15, or only the neural network No. having new learning data accumulated to some extent (for example, 100 sets) may be learned at that time point.
As described above, the shape of the rolling mill of the plant 1 to be controlled can be changed without significantly disturbing:
1) Instead of setting the reference shape pattern and the control operation therefor separately in advance to learn the control operation method, a combination of the shape pattern and the control operation is learned and used to implement the control operation.
2) Since a new control rule is not predictable in advance and a control rule that cannot be predicted at all is sometimes optimal, the control operation terminal is operated at random and searched for while observing the control result.
Further, although the neural network used in the control execution device 20 is stored in the control rule database DB1, if the stored neural network is obtained by only performing the initialization process at random, it takes time until the neural network can be learned until the corresponding control is performed. Therefore, when the control unit is constructed for the plant 1 to be controlled, the control rule is learned by simulation in advance based on the control model of the plant 1 to be controlled identified at that time point, and the neural network after completion of learning in the simulator is stored in the database, so that control with a certain level of performance can be performed from the start of the plant to be controlled.
The complete equipment control device of the present invention is actually implemented as a computer system, but in this case, a plurality of program groups are formed in the computer system, and these program groups may be stored in a storage medium.
these program groups are, for example, a control rule execution program, a control output determination program, and a control output suppression program for realizing processing of the control execution means, wherein the control rule execution program provides a control output in accordance with a determined combination of actual data and a control operation of the control target plant; the control output determination program determines whether the control rule execution program can output the control output, and notifies the control method learning device that the actual data and the control operation are erroneous; the control output suppression program prevents the control output from being output to the plant to be controlled when the control output determination program determines that the actual data of the plant to be controlled is deteriorated when the control output is output to the plant to be controlled, these program groups are, for example, a control result satisfaction/non-satisfaction determination program, a learning data creation program, and a control rule learning program for realizing the processing of the control method learning device, wherein the control result adequacy-judging program is used for realizing the processing of judging whether the control result is adequacy or not, wherein in the case where the control execution means actually outputs the control output to the control target complete equipment, whether the control result regarding whether the actual data becomes good or deteriorated as compared with that before the control is good or not is determined after a time delay until the control effect is exhibited in the actual data; the learning data creation program obtains teacher data by using the control result and the control output in the control result adequacy determination program; the control rule learning program learns the actual data and the teacher data as learning data.
The control method learning device performs learning to obtain individual combinations of actual data and control operations for a plurality of control targets based on the states of the control target plant, and uses the obtained combinations of actual data and control operations as the determined combinations of actual data and control operations of the control target plant in the control rule execution program.
In addition, when the apparatus of the present invention is applied to an actual plant, it is necessary to determine an initial value of the neural network, and in this regard, a combination of actual data and a control operation can be created by simulation using a control model of the plant to be controlled before control in the plant to be controlled is performed, and a learning period of the combination of actual data and a control operation in the plant to be controlled can be shortened.
Industrial applicability
The present invention relates to a method and a part for controlling a rolling mill, which is one of rolling apparatuses, and has no particular problem in practical use.

Claims (12)

1. a plant control apparatus for recognizing a pattern of a combination of actual data of a plant to be controlled for the plant to be controlled, and performing control, the plant control apparatus comprising:
A control method learning device for learning a combination of actual data and control operation of the whole set of the apparatus to be controlled; and
A control execution means for executing control of the whole set of control target equipment based on the learned combination of the actual data and the control operation,
The control execution device includes:
A control rule execution unit for providing a control output in accordance with the determined combination of the actual data and the control operation of the plant to be controlled;
A control output determination section that determines whether the control rule execution section can output a control output, and notifies the control method learning device that the actual data and the control operation are erroneous; and
A control output suppressing unit that prevents the control output from being output to the plant to be controlled when the control output determining unit determines that the actual data of the plant to be controlled is deteriorated when the control output is output to the plant to be controlled,
The control method learning device includes:
A control result adequacy determination unit that determines whether the control result regarding whether the actual data is better or worse than before the control after a time delay until the control effect is exhibited by the actual data when the control execution device actually outputs the control output to the plant to be controlled;
A learning data creation unit that obtains teacher data by using the control result and the control output in the control result adequacy determination unit; and
A control rule learning unit for learning the actual data and the teacher data as learning data,
The control method learning device performs learning to obtain individual combinations of actual data and control operations for a plurality of control targets based on the states of the control target plant, and uses the obtained combinations of actual data and control operations as the determined combinations of actual data and control operations of the control target plant in the control rule execution unit.
2. the plant control apparatus according to claim 1,
In order to replace the combination of the actual data and the control operation according to the size of the actual data of the plant to be controlled, the combination of the actual data and the control operation is learned and controlled using information on the size of the actual data and information for easily standardizing the actual data to perform pattern recognition.
3. The plant control apparatus according to claim 1 or 2,
The control rule executing unit holds a combination of actual data and a control operation of a control target plant as a 1st neural network, the control rule learning unit holds a combination of actual data and a control operation as a2 nd neural network, and the 2 nd neural network obtained as a result of learning in the control method learning device is used as the 1st neural network in the control rule executing unit.
4. The plant control apparatus according to claim 1 or 2,
the control execution device includes a control operation disturbance generation unit that provides disturbance to the control output, and the control method learning device performs learning including when disturbance is applied.
5. The plant control apparatus according to claim 1 or 2,
the control method learning device obtains a plurality of combinations of actual data and control operations through learning based on a plurality of specifications determined in advance, and the control execution device selects 1 of the plurality of combinations of actual data and control operations from the plurality of combinations of actual data and control operations according to the operating state of the plant to be controlled to provide the control output.
6. The plant control apparatus according to claim 3,
The neural network to be used for learning the combination of the actual data and the operation method is changed according to the size of the actual data.
7. the plant control apparatus according to claim 1 or 2,
and changing a criterion for determining whether the control result is good or not according to the state of the plant to be controlled or the experience of the operator of the plant to be controlled, and obtaining the relationship between actual data and the operation method for the plant to be controlled and storing the relationship in a database, so that the plant to be controlled is controlled by different control methods according to the state of the plant to be controlled or the experience of the operator of the plant to be controlled.
8. The plant control apparatus according to claim 1 or 2,
Before control in the plant to be controlled is performed, a combination of the actual data and the control operation is created by simulation using a control model of the plant to be controlled, and a learning period of the combination of the actual data and the control operation in the plant to be controlled is shortened.
9. A rolling mill control apparatus to which the plant control apparatus of any one of claims 1 to 8 is applied, characterized in that,
The plant to be controlled is a rolling mill, and the actual data is the shape of the output side of the rolling mill.
10. A plant control method applied to a plant control apparatus for performing control in a mode in which a plant to be controlled recognizes a combination of actual data of the plant to be controlled, the plant control apparatus comprising:
A control method learning device for learning a combination of actual data and control operation of the whole set of the apparatus to be controlled; and
a control execution means for executing control of the whole set of control target equipment based on the learned combination of the actual data and the control operation,
the control execution means provides a control output in accordance with the determined combination of the actual data and the control operation of the control target plant, determines whether the control output can be output, notifies the control method learning means that the actual data and the control operation are erroneous, and prevents the control output from being output to the control target plant in the case where it is determined that the actual data of the control target plant is deteriorated in the case where the control output is output to the control target plant,
the control method learning device determines whether or not a control result regarding whether actual data is better or worse than before control is good after a time delay until a control effect is expressed in the actual data when the control execution device actually outputs the control output to the control target plant, obtains teacher data using the control result and the control output, learns the actual data and the teacher data as learning data, obtains a unique combination of actual data and control operation for a plurality of control targets according to a state of the control target plant by learning, and uses the obtained combination of actual data and control operation as a determined combination of actual data and control operation of the control target plant in the control execution device.
11. A rolling mill control method to which the plant control method according to claim 10 is applied,
The whole set of equipment to be controlled is a rolling mill, and the actual data is the shape of the outlet side of the rolling mill.
12. A storage medium storing a program for realizing, by a computer system, a plant control apparatus that controls a plant to be controlled by identifying a pattern of a combination of actual data of the plant to be controlled,
The computer system is provided with:
A control method learning device for learning a combination of actual data and control operation of the whole set of the apparatus to be controlled; and
A control execution means for executing control of the whole set of control target equipment based on the learned combination of the actual data and the control operation,
The programs are a control rule execution program, a control output determination program, a control output suppression program for realizing the processing of the control execution means, wherein the control rule execution program provides a control output in accordance with the determined combination of the actual data and the control operation of the plant as the object of control; the control output determination program determines whether the control rule execution program can output a control output, and notifies the control method learning device that the actual data and the control operation are erroneous; the control output suppression program prevents the control output from being output to the plant to be controlled when the control output determination program determines that the actual data of the plant to be controlled is deteriorated when the control output is output to the plant to be controlled,
The program is a control result adequacy determining program, a learning data making program, a control rule learning program for realizing the processing of the control method learning device, wherein the control result adequacy determining program is used for realizing the processing of the control result adequacy determination, wherein the adequacy of the control result as to whether the actual data becomes good or deteriorated as compared with before the control is determined after a time delay until the control effect is expressed in the actual data in the case where the control executing device actually outputs the control output to the control target plant; the learning data creation program obtains teacher data by using the control result and the control output in the control result adequacy determination program; the control rule learning program learns the actual data and the teacher data as learning data,
The control method learning means performs learning to obtain individual combinations of actual data and control operations for a plurality of control targets based on the states of the control target plant, and uses the obtained combinations of actual data and control operations as the determined combinations of actual data and control operations of the control target plant in the control rule execution program.
CN201810267684.XA 2017-04-10 2018-03-29 complete equipment control device, rolling mill control device, control method and storage medium Active CN108687137B (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2017-077350 2017-04-10
JP2017077350A JP6813416B2 (en) 2017-04-10 2017-04-10 Plant control device and its control method, rolling mill control device and its control method and program

Publications (2)

Publication Number Publication Date
CN108687137A CN108687137A (en) 2018-10-23
CN108687137B true CN108687137B (en) 2019-12-10

Family

ID=63588226

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810267684.XA Active CN108687137B (en) 2017-04-10 2018-03-29 complete equipment control device, rolling mill control device, control method and storage medium

Country Status (3)

Country Link
JP (1) JP6813416B2 (en)
CN (1) CN108687137B (en)
DE (1) DE102018204684A1 (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP7328142B2 (en) * 2019-12-27 2023-08-16 株式会社日立製作所 Plant control device and its control method, rolling mill control device and its control method, and program
JP7495874B2 (en) 2020-12-16 2024-06-05 株式会社日立製作所 PLANT CONTROL SYSTEM, PLANT CONTROL METHOD, AND PROGRAM
JP2022107463A (en) * 2021-01-08 2022-07-21 株式会社日立製作所 Plant controller, plant control method, and program
CN116125803B (en) * 2022-12-28 2024-06-11 淮阴工学院 Inverter backstepping fuzzy neural network control method based on extreme learning machine
CN116027673B (en) * 2023-03-29 2023-06-06 中国电子科技集团公司第二十九研究所 Equipment control autonomous decision-making method based on fuzzy neural network

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS633733A (en) 1986-06-23 1988-01-08 株式会社クボタ Water feed control unit for crops
JP2804161B2 (en) * 1990-06-04 1998-09-24 株式会社日立製作所 Method and apparatus for controlling shape of Sendzimir mill
JPH04252302A (en) * 1991-01-28 1992-09-08 Kawasaki Steel Corp Control model identifying method and control method for continuous rolling mill
JP3308630B2 (en) * 1993-03-12 2002-07-29 株式会社日立製作所 Plant control apparatus and method
JP3092460B2 (en) * 1994-09-14 2000-09-25 株式会社日立製作所 Width compression machine and rolling mill
JP3223856B2 (en) * 1997-04-17 2001-10-29 日本鋼管株式会社 Rolling mill control method and rolling mill control device
JP4003733B2 (en) * 2003-10-22 2007-11-07 株式会社日立製作所 PLANT SYSTEM, CONTROL DEVICE AND CONTROL METHOD
JP4665815B2 (en) * 2006-03-31 2011-04-06 株式会社日立製作所 Plant control equipment
JP5855964B2 (en) * 2012-02-07 2016-02-09 メタウォーター株式会社 Optimal control method and optimal control device for plant equipment
JP6233423B2 (en) * 2014-02-17 2017-11-22 東芝三菱電機産業システム株式会社 Learning control device for rolling process

Also Published As

Publication number Publication date
DE102018204684A1 (en) 2018-10-11
JP2018180799A (en) 2018-11-15
CN108687137A (en) 2018-10-23
JP6813416B2 (en) 2021-01-13

Similar Documents

Publication Publication Date Title
CN108687137B (en) complete equipment control device, rolling mill control device, control method and storage medium
CN109317522B (en) Equipment control device, rolling mill control device, control method thereof, and recording medium
US20200327411A1 (en) Systems and Method on Deriving Real-time Coordinated Voltage Control Strategies Using Deep Reinforcement Learning
US10121107B2 (en) Machine learning device and method for optimizing frequency of tool compensation of machine tool, and machine tool having the machine learning device
US9043254B2 (en) Method for computer-aided closed-loop and/or open-loop control of a technical system
US10466308B2 (en) Fault level estimation method for power converters
CN106093612A (en) A kind of method for diagnosing fault of power transformer
US9934470B2 (en) Production equipment including machine learning system and assembly and test unit
JP2018180799A5 (en)
JP7033639B2 (en) Plant control device and its control method, rolling mill control device and its control method and program
JP7034046B2 (en) Control device and control method
CN110976523B (en) Control device and control method
Razmi et al. Neural network based on a genetic algorithm for power system loading margin estimation
CN113050550B (en) Plant equipment control device and method thereof, rolling mill control device and method thereof
CN109768545A (en) A kind of black-start scheme preferred method based on Vague collection
JP7495874B2 (en) PLANT CONTROL SYSTEM, PLANT CONTROL METHOD, AND PROGRAM
Cuznar et al. Optimization of cold rolling process recipes based on historical data
JP2022182538A (en) Plant control system, plant control method and program
Abraham et al. Neuro-fuzzy approximation of multi-criteria decision-making QFD methodology
JP2022183827A (en) Plant control system, rolling machine control device, plant control method, and plant control program
Ruiz et al. Fault location in electrical distribution systems using PLS and NN
CN111832784A (en) Port comprehensive energy-oriented long and short memory network power load prediction method
CN117638938A (en) Control method and device of power system and electronic equipment
Igelnik Feature selection and ranking
JP2020166599A (en) Operation plan support device, operation plan support method and operation plan support program

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant