WO2019004253A1

WO2019004253A1 - Device and method for predicting operation results in manufacturing system

Info

Publication number: WO2019004253A1
Application number: PCT/JP2018/024301
Authority: WO
Inventors: 博史天野; 多鹿　陽介; 裕一樋口; 太一清水
Original assignee: パナソニックＩｐマネジメント株式会社
Priority date: 2017-06-27
Filing date: 2018-06-27
Publication date: 2019-01-03

Abstract

The present invention predicts operation results about a combination of components which have not actually been operated in a manufacturing system. This device for predicting the operation results comprises a processor and a computer-readable recording medium, wherein the computer-readable recording medium includes: a command for instructing the processor to receive operation results about multiple combinations of components, said combinations being combined from among multiple types of components that can be included in the manufacturing system; and a command for instructing the processor to predict, on the basis of the received operation result, operation results for combinations which, among the multiple combinations, are those for which operation results were not obtained. One of the multiple types of components includes a manufacturing device or parts constituting a manufacturing device.

Description

Apparatus and method for predicting operation results in a manufacturing system

The present disclosure relates to technology for predicting operation results in a manufacturing system.

In general, factories are required to reduce the power and manufacturing time used for production. For example, the system disclosed in Patent Document 1 stores a factor that makes the power consumption per unit number of products higher than a reference value, and calculates the degree of occurrence of such factor as a factor point. This makes it possible to easily identify the factor when the power per unit of product increases.

International Publication No. 2017/002242

However, in the system of Patent Document 1, although the factor of the power increase under the conditions that have actually been operated can be identified, such a condition can not be done for the conditions that have not been actually operated.

The present disclosure actually operates on a manufacturing apparatus that may be included in a manufacturing system, or a part that configures such a manufacturing apparatus, and at least one component of the manufacturing system that has not been operated in practice. The purpose is to predict the operation result when it is made to

An apparatus for predicting operation results in a manufacturing system according to the present disclosure is combined between one or more processors, the one or more processors, and a plurality of types of components that may be included in the manufacturing system. Receiving the operation result obtained by actually operating each of the plurality of combinations with respect to the plurality of combinations of the constituent elements, and of the plurality of combinations, the operation result is not actually obtained yet Predicting one or more operation results for the combination based on the received one or more non-transitory computer-readable storage media including instructions for causing the operation results to be predicted. One of the plurality of types of components includes a manufacturing apparatus or a part that constitutes the manufacturing apparatus.

According to this, the actual operation is performed on the manufacturing apparatus included in the manufacturing system or a part constituting such manufacturing apparatus and at least one component of the manufacturing system, which has not been actually operated. Operation results can be predicted.

A computer-implementable method of predicting operation results in a manufacturing system according to the present disclosure is for a plurality of combinations of the components combined between a plurality of types of components that may be included in the manufacturing system; Receiving, by the processor, an operation result obtained by actually operating each of the plurality of combinations, and an operation result for a combination of the plurality of combinations for which the operation result has not yet been obtained. Predicting by the processor based on the received operation results. One of the plurality of types of components includes a manufacturing apparatus or a part that constitutes the manufacturing apparatus.

According to the present disclosure, the manufacturing apparatus included in the manufacturing system or a part (part etc.) constituting such manufacturing apparatus and at least one component (operator, part, parameter etc.) of the manufacturing system For combinations that have not been operated, it is possible to predict the operation results (power, manufacturing time, quality, etc.) when actually operated. By using this prediction result, an optimal combination can be grasped before production to make the operation result better, and a more efficient production plan can be formulated.

FIG. 1 is a sequence diagram of a manufacturing system and an apparatus for predicting operation results in the manufacturing system according to an embodiment of the present invention. FIG. 2 is a block diagram showing a configuration example of a prediction device according to an embodiment of the present invention. FIG. 3 is a diagram showing an example of the result table. FIG. 4 is a flowchart showing an example of processing in the component element information acquisition block of FIG. FIG. 5 is a flowchart showing an example of processing in the operation result acquisition block of FIG. FIG. 6 is a flowchart showing an example of processing in the operation result prediction block of FIG. FIG. 7 is a flowchart showing an example of processing in the correlation coefficient calculation block of FIG. FIG. 8 is a flowchart showing an example of processing in the evaluation value prediction block of FIG. FIG. 9 is a diagram showing an example of a result table including predicted evaluation values.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. Components indicated by the same reference numerals in the drawings are identical or similar components.

FIG. 1 is a sequence diagram of a manufacturing system 110 and an apparatus 130 for predicting operation results of the manufacturing system according to an embodiment of the present invention. FIG. 2 is a block diagram showing a configuration example of the prediction device 130 according to the embodiment of the present invention. The prediction device 130 of FIG. 2 is specifically a computer system, and includes a processor 132, an input device 134, an output device 136, a network interface 138, a memory 142, a file storage device 144, and a bus 146. including.

Processor 132 communicates with other components via bus 146. The network interface 138 transmits and receives data to and from a communication network such as the Internet. The network interface 138 is connected to a communication network by wire or wirelessly, and is connected to the manufacturing system 110 via the communication network. The prediction device 130 may be connected to the manufacturing system 110 without passing through the communication network.

The memory 142 and the file storage device 144 are one or more volatile or nonvolatile non-transitory computer readable storage media. The memory 142 includes, for example, a random access memory (RAM) and a read only memory (ROM), and stores data and instructions. The file storage device 144 may include a RAM, a ROM, an electrically erasable programmable read only memory (EEPROM), a semiconductor memory such as a flash memory, a magnetic recording medium such as a hard disk drive, an optical recording medium, a combination thereof, and the like. Where embodiments of the invention are implemented in software, for example, microcode, assembly language code, or higher level language code may be used. The memory 142 or the file storage device 144 stores a program described by these codes and including an instruction that implements the functions of the embodiment of the present invention. By operating according to such a computer program, processor 132 achieves its functions.

Input device 134 may include a touch screen, a keyboard, a remote control, a mouse, and the like. The output device 136 may include a flat panel display such as a liquid crystal display or an organic EL display.

Manufacturing system 110 may be, for example, a manufacturing apparatus and its control unit, a manufacturing line to which a plurality of such manufacturing apparatuses are connected, a manufacturing floor on which a plurality of such manufacturing lines are arranged, or a plurality of such manufacturing floors. Is a manufacturing plant with

In the present specification, components of the manufacturing system are components necessary for manufacturing a product, such as a manufacturing apparatus, manufacturing data, parts of a manufacturing apparatus, parts of a product, an operator operating a manufacturing apparatus, etc. . The components of the manufacturing system are not limited to these, and may be any components necessary for manufacturing a product in the manufacturing process. The products include not only finished products but also intermediate products which do not reach it. In addition, the operation result is the power required for manufacturing the product, the time required for manufacturing, the quality of the product, and the like.

The manufacturing system 110 starts manufacturing at block 12, performs processing to manufacture a product at block 14, and ends manufacturing at block 16. At block 20, the predictor 130 receives information about the components of the manufacturing system 110. Specifically, the processor 132 of the prediction device 130 identifies the manufacturing device identification information as a component of the manufacturing system 110, and the operator identification information as a component of the manufacturing system 110 that operates the device. Are received from the manufacturing system 110.

The processor 132 performs an operation result obtained by actually operating each of the plurality of combinations of the plurality of combinations of components combined among the plurality of types of components that may be included in the manufacturing system 110. To receive. Here, one of the plurality of types of components includes, for example, a manufacturing apparatus or a part that configures the manufacturing apparatus, and the other includes, for example, an operator who can operate the manufacturing apparatus. More specifically, at block 40, the processor 132 receives, from the manufacturing system 110, operation results for manufacturing in the manufacturing system 110, for example, information regarding the operation of the manufacturing apparatus, for the plurality of combinations of components, and the result table Write to memory 142 as part of The manufacturing system 110 and the prediction device 130 repeat the above operation while changing the combination of the manufacturing device and the operator who can operate the device.

Thereafter, the processor 132 predicts, based on the received operation result, an operation result for a combination for which an operation result has not yet been obtained among the plurality of combinations of components. More specifically, at block 60, the processor 132 operates, for example, among combinations of a plurality of manufacturing apparatuses and a plurality of operators who can operate the apparatuses, which combination has not actually been performed yet. Predict the results. The processor 132 outputs the predicted operation result as a prediction result table.

FIG. 3 is a diagram showing an example of the result table. The result table has cells corresponding to multiple combinations of components combined between multiple types of components (eg, manufacturing equipment and operators) that may be included in the manufacturing system 110. The result table is, for example, a table having cells corresponding to all combinations of a plurality of operators and a plurality of manufacturing apparatuses, as shown in FIG. In each cell, the evaluation value for the corresponding combination is recorded. The result table has “-” in each cell, which means “not evaluated” as an initial value. For example, as the power consumption required for manufacturing is smaller, a larger evaluation value is recorded, and the range of the evaluation value is, for example, 0-5.

FIG. 4 is a flow chart illustrating an example of the process in block 20 of FIG. 1 for receiving component information. At block 22, processor 132 receives from manufacturing system 110 identification information of the operator who performed the manufacturing. At block 24, the processor 132 determines whether the received identification information of the operator is present in the result table stored in the memory 142. If the identification information of the operator exists, the process proceeds to block 26. If not, the process proceeds to block 28. At block 28, processor 132 additionally records the row of identification information for the operator in the result table, stores the result table in memory 142, and processing proceeds to block 26.

At block 26, processor 132 receives from manufacturing system 110 identification information of the manufacturing device that performed the manufacturing. At block 30, the processor 132 determines whether the received identification information of the manufacturing device is present in the result table stored in the memory 142. If the identification information of the manufacturing apparatus is present, the process is terminated. If not, the process proceeds to block 32. At block 32, the processor 132 adds and records the row of identification information of the manufacturing device in the result table. The processor 132 stores the result table in the memory 142 and ends the process.

FIG. 5 is a flowchart showing an example of processing in the operation result acquisition block 40 of FIG. At block 42, processor 132 receives operational results for manufacturing from manufacturing system 110. The operation result is, for example, the power required for manufacturing. At block 44, the processor 132 normalizes the received operation results to values of 0-5. At block 46, the processor 132 records the normalized operation results in the results table as an evaluation value corresponding to the combination of the manufactured operator and the manufactured manufacturing device. The processor 132 stores the result table in the memory 142 and ends the process.

For example, the evaluation value when the manufacturing apparatus a is operated by the operator A is recorded in the cell belonging to the row of the operator A and the column of the manufacturing apparatus a in the result table of FIG. In the case of such a combination of the operator A and the manufacturing apparatus a, for example, “5” is recorded when the power consumption required for manufacturing is very small. For example, when the power consumption required for manufacturing is very large when the manufacturing apparatus m is operated by the operator N, the row of the manufacturing apparatus m belongs to the row of the operator N in the result table of FIG. In the cells belonging to, “1” is recorded as an evaluation value, which means that the power consumption required for manufacturing is very large. In the example of the result table of FIG. 3, only the combination of the operator B and the manufacturing apparatus b is not evaluated, and the others are evaluated.

FIG. 6 is a flowchart showing an example of processing in the operation result prediction block 60 of FIG. In the correlation coefficient calculation block 70, the processor 132 calculates the correlation coefficient between the operators recorded in the result table from the cells having the evaluation value of the cells of the result table. In the evaluation value prediction block 90, the processor 132 calculates the operation result for the combination for which the operation result has not yet been obtained from the correlation coefficient obtained in the block 70 and the evaluation value in the result table, that is, an unevaluated combination. Predict the rating value for

FIG. 7 is a flowchart showing an example of processing in the correlation coefficient calculation block 70 of FIG. At block 72, the processor 132 sets one of the operators registered in the result table to the operator i = 0. At block 74, the processor 132 determines whether correlation coefficients with other operators have been calculated for all the operators registered in the result table. Specifically, if the value i has reached a value smaller by one than the number of operators, the process is ended, and if it has not, the process proceeds to block 76. At block 76, the processor 132 sets the operator whose correlation coefficient is to be calculated with the operator i to the operator j = i + 1. At block 78, processor 132 determines, for operator i, whether correlation coefficients with all other operators have been calculated. Specifically, the process proceeds to block 84 if the value j has reached the number of operators, or proceeds to block 80 if it has not.

At block 80, processor 132 calculates the correlation coefficient between operator i and operator j. Specifically, processor 132 calculates such a correlation coefficient using an evaluation value for a manufacturing apparatus for which an evaluation value is obtained for any combination of the two. At block 82, the processor 132 increments the value j by one and returns to block 78. At block 84, the processor 132 increments the value i by one and returns to block 74.

The process at block 80 will be described in more detail. The processor 132 calculates the relative evaluation value of the operator x with respect to the manufacturing apparatus y, that is,
r _xy -ra _x (equation 1)
Ask for Here, r _xy indicates the evaluation value of the operator x with respect to the manufacturing apparatus y. Also, ra _x represents an average value of evaluation values of the operator x.

The processor 132 calculates the evaluation value similarity to the manufacturing apparatus k of the operator i and the operator j, that is,
(r _ik -ra _i ) (r _jk -ra _j ) (equation 2)
Ask for When this value is negative, both operators are dissimilar, and when positive, both operators are similar.

The processor 132 calculates the correlation coefficient between the operator a and the operator x, that is,
_ax = _{y y} (r _ay -ra ' _a ) (r _xy -ra' _x ) / (√ (Σ _y (r _ay -ra ' _a ) ² )) (Σ _y (r _xy -ra' _x ) ² )) (Equation 3)
Ask for At this time, only the evaluation value for the manufacturing apparatus that has been used in combination with the operator a and has been used in combination with the operator x is the target of calculation. Here, ra ′ _x represents an average value of evaluation values of the operator x with respect to a manufacturing apparatus that has been used in combination with the operator a and has been used in combination with the operator x.

FIG. 8 is a flow chart showing an example of processing in the evaluation value prediction block 90 of FIG. At block 92, the processor 132 determines if there are unrated cells in the results table that correspond to the unrated combination. If there is an unevaluated cell, the process proceeds to block 94, and if there is no unevaluated cell, the process ends. In block 94, the processor 132 predicts, for each manufacturing device registered in the result table, the evaluation value of the unevaluated cell from the correlation coefficient between the operators and the evaluation value already obtained.

The process at block 94 will be described in more detail. From the evaluation values of all the operators x other than the operator a, the processor 132 evaluates the evaluation value when the operator a and the manufacturing apparatus y are used in combination, that is,
_{X x} ρ _ax (r _xy -ra ' _x ) (Equation 4)
Ask for

The processor 132 normalizes Expression 4 to correct the problem that the evaluation value increases as the evaluation value is determined for many operators, that is,
_{X x} ρ _ax (r _xy -ra ' _x ) / _{x x} | ρ _ax |
Ask for

Furthermore, the processor 132 can evaluate the evaluation value for the combination of the operator a and the manufacturing apparatus y, that is,
ra _ay = ra _a + _{x x} _ax (r _xy -ra ' _x ) / _{x x} | ρ _ax |
Is predicted as the operation result.

FIG. 9 is a diagram showing an example of a result table including predicted evaluation values. As a result of the above processing, an evaluation value of 1.75 is obtained as a predicted operation result for the combination of the operator B and the manufacturing apparatus b which are not evaluated in FIG. 3 and is recorded in the result table as shown in FIG. . Thereafter, a manufacturing system is configured based on the result table, and manufacturing is performed using this manufacturing system. At this time, for example, a combination with the best evaluation value or a combination with the evaluation value higher than the threshold is used for the manufacturing system.

As described above, according to the prediction device 130, the operation result (for example, power) can be predicted with high accuracy for the combination of the manufacturing device and the operator who have not been used in combination and operation. By using this prediction result, an optimal combination can be grasped before production to make the operation result better, and a more efficient production plan can be formulated.

(A) Although the combination shown in the result table has been described as being a combination of an operator and a manufacturing apparatus, it is not limited thereto. The result table may be, for example, a combination of a manufacturing grade and a manufacturing apparatus, a combination of a part A of a manufacturing apparatus and a part B of a manufacturing apparatus, a combination of a first manufacturing apparatus and a second manufacturing apparatus, or a manufacturing apparatus and its apparatus. It may indicate the operation result of the combination with the parameter to be set, and the combination may be a combination of elements constituting the manufacturing system.

(B) Although the value of each cell in the result table is described as the power required for manufacturing, the value of each cell may be, for example, a value related to quality such as manufacturing time, manufacturing yield, etc. Anything may be used.

(C) Although the result table is described as being a two dimensional table, it may be a three dimensional or more dimensional table. That is, the prediction device 130 may predict the operation result for the combination of three or more components of the manufacturing system. Although the case where the operation result about the combination for which the operation result is not obtained is actually obtained is predicted using the correlation operation as an example, the prediction method is not limited to this. As a prediction method, for example, matrix decomposition, Bayesian estimation, deep learning, etc. may be used, and operation results for combinations for which operation results have not yet been obtained are predicted based on the already obtained operation results. Any method can be used.

Some or all of the components constituting the prediction device 130 may be configured from one system LSI (large scale integration: large scale integrated circuit). The system LSI is a super-multifunctional LSI manufactured by integrating a plurality of components on one chip, and more specifically, a computer system including a microprocessor, a ROM, a RAM, and the like. . A computer program is stored in the RAM. The system LSI achieves its functions as the microprocessor operates in accordance with the computer program.

Some or all of the components that make up the prediction device 130 may be configured from an IC card or a single module that is removable from each device. The IC card or the module is a computer system including a microprocessor, a ROM, a RAM, and the like. The IC card or the module may include the above-described ultra-multifunctional LSI. The IC card or the module achieves its function by the microprocessor operating according to the computer program. This IC card or this module may be tamper resistant.

The present invention may be methods shown above. Further, the present invention may be a computer program that realizes these methods by a computer, or may be a digital signal composed of the computer program.

Further, the present invention is a computer readable recording medium capable of reading the computer program or the digital signal, such as a flexible disk, a hard disk, a CD-ROM, an MO, a DVD, a DVD-ROM, a DVD-RAM, a BD (Blu-ray disk). ), And may be recorded in a semiconductor memory or the like. Further, the present invention may be the digital signal recorded on these recording media.

In the present invention, the computer program or the digital signal may be transmitted via a telecommunication line, a wireless or wired communication line, a network represented by the Internet, data broadcasting, and the like.

The many features and advantages of the present invention are apparent from the written description, and thus, it is intended by the appended claims to cover all such features and advantages of the present invention. Moreover, the present invention should not be limited to the exact construction and operation as illustrated and described, as many modifications and variations are readily possible to those skilled in the art. Accordingly, all suitable modifications and equivalents are intended to be included within the scope of the present invention.

As described above, the present invention is useful for a method and system for predicting operation results.

110 manufacturing system 130 operation result prediction device 132 processor 142 memory

Claims

An apparatus for predicting operation results in a manufacturing system, comprising:
With one or more processors,
On said one or more processors,
The plurality of combinations of components combined among the plurality of types of components that may be included in the manufacturing system receive operation results obtained by actually operating each of the plurality of combinations. And
Predicting, based on the received operation result, an operation result of a combination for which the operation result has not yet been determined among the plurality of combinations;
One or more non-transitory computer readable storage media including instructions for causing
Equipped with
One of the plurality of types of components includes a manufacturing apparatus or a part that constitutes the manufacturing apparatus,
A device that predicts operating results in a manufacturing system.
An apparatus for predicting an operation result in the manufacturing system according to claim 1,
One of the plurality of types of components includes an operator who can operate the manufacturing apparatus
A device that predicts operating results in a manufacturing system.
A computer-implemented method for predicting operation results in a manufacturing system, comprising:
A processor is configured to obtain an operation result obtained by actually operating each of the plurality of combinations of the plurality of combinations of the plurality of combinations of the plurality of types of components that may be included in the manufacturing system. Receiving by
Predicting, by the processor based on the received operation result, an operation result of a combination of the plurality of combinations for which the operation result has not yet been determined;
Equipped with
One of the plurality of types of components includes a manufacturing apparatus or a part that constitutes the manufacturing apparatus,
A method of predicting operating results in a manufacturing system.