JP2020173581A - Data analysis method, data analysis device, and computer program - Google Patents

Abstract

To provide a data analysis method capable of correctly evaluating a degree of association of characteristics between products manufactured from the same material lot by a different manufacturing line, a data analysis device, and a computer program.SOLUTION: A data analysis method comprises an acquisition step, and an evaluation step. The acquisition step acquires a dataset containing independent data collected in a manufacturing process and not depending on a specific process to be analyzed during the manufacturing process, and dependent data collected in the manufacturing process and depending on the specific process. The evaluation step evaluates relevance between datasets using all independent data and dependent data corresponding to the specific process.SELECTED DRAWING: Figure 8

Description

The present invention relates to data analysis methods, data analyzers, and computer programs.

In the raw materials that are the basis of products, the characteristics of each raw material (for example, particle size) vary. Conventionally, a method for analyzing data including such variations has been known (see, for example, Patent Document 1). The technique described in Patent Document 1 evaluates log data recorded when an exposure apparatus performs a predetermined control process by using a graphical lasso. Non-Patent Document 1 describes an evaluation method for the degree of relevance of each characteristic in the data.

JP-A-2017-167310

Written by Kazunori Yamaguchi, Written by Motohisa Hirono, Supervised by Masahiko Muchinaka, "Introduction to SEM Causal Analysis JUSE-StatWorks Official Text (Series 6 Useful for Real Part)", Nikka Giren Publishing Co., Ltd., July 2012

The relevance of the various data collected during the manufacturing process may depend on the manufacturing line. Therefore, the analysis of various data collected in the manufacturing process is carried out by dividing the data for each manufacturing line (hereinafter, also referred to as "stratification"). However, the number of samples of raw material lot data, etc., which does not depend on the production line, is reduced by stratification. As described above, there is a problem that the degree of relevance evaluated from the data of a small number of samples may greatly deviate from the true value. In addition, when the correlation coefficient is used as the degree of relevance, a high correlation coefficient may be obtained by chance. In such a case, when the Gaussian graphical model is applied, the variance-covariance matrix and the correlation coefficient matrix are definite. There was a problem that the value was not satisfied and the degree of relevance could not be evaluated.

The present invention has been made to solve at least a part of the above-mentioned problems, and an object of the present invention is to correctly evaluate the degree of relevance of characteristics between products manufactured from the same raw material lot by different production lines.

The present invention has been made to solve at least a part of the above-mentioned problems, and can be realized as the following forms.

(1) According to one embodiment of the present invention, a data analysis method is provided. This data analysis method includes data collected in the manufacturing process, independent data that does not depend on the specific process to be analyzed in the manufacturing process, and dependent data that depends on the specific process. It includes a step of acquiring a data set and a step of evaluating the relevance of the data set by using all of the independent data and the dependent data corresponding to the specific process.

According to this configuration, all of the independent data that does not depend on a specific process and the dependent data corresponding to a specific process (that is, the data divided for each specific process) are evaluated for the relevance of the dataset. , Data corresponding to a specific process) and. For this reason, it is possible to suppress a decrease in the number of samples of independent data as compared with the case where the entire data set is divided into specific processes, and the problem caused by the decrease in the number of samples (relevance is high). It is possible to suppress the occurrence of deviating from the true value (the positiveness of the covariance matrix and the correlation coefficient matrix becomes unsatisfactory when applying a Gaussian graphical model), and evaluate the degree of relevance correctly. ..

(2) In the data analysis method of the above embodiment, the specific process may be a production line that realizes the production process.
According to this configuration, since the specific process is a production line, it is possible to evaluate the relevance of the data set including the independent data before processing on the production line and the dependent data after processing.

(3) In the data analysis method of the above embodiment, the independent data is data representing various characteristics of each raw material used in the manufacturing process, and the dependent data is acquired in each process in the manufacturing process. The data may represent at least one of various characteristics of the processed product and processing conditions for processing each raw material.
According to this configuration, the relationship between various characteristics of the raw material as independent data can be evaluated, and at least one relationship between various characteristics of the processed product as dependent data and the processing conditions in the manufacturing process can be evaluated. Can be evaluated.

(4) In the data analysis method of the above embodiment, the evaluation of the relevance is independent of the variance-covariance matrix or the correlation coefficient matrix obtained in a state where the data set is divided for each specific process. For the elements corresponding to the data, input a matrix using the covariance matrix or the correlation coefficient matrix obtained without dividing the data set for each specific process into the Gaussian graphical model. May be good.
According to this configuration, a variance-covariance matrix or a correlation coefficient matrix is used to evaluate the relevance of the dataset. In the variance-covariance matrix or correlation coefficient matrix of the divided dependent data dataset, the element corresponding to the independent data is replaced with the variance-covariance matrix or correlation coefficient matrix of the independent data set. As a result, it is possible to calculate the relevance without causing an error in the Gaussian graphical model due to the accidentally high correlation coefficient between the variables of the independent data, and to evaluate the relevance as a network between the variables.

The present invention can be realized in various aspects, for example, a data analyzer, a data analysis system, a data analysis method, a control method of the data analyzer, and a computer program for executing these devices and methods. , This can be realized in the form of a server device for distributing the computer program, a non-temporary storage medium for storing the computer program, or the like.

It is a block diagram of the data analyzer as one Embodiment of this invention. It is an example of a data set including raw material data and product data. This is an example of a correlation coefficient matrix calculated from raw material data. A dataset that corresponds to the production line. A dataset that corresponds to the production line. It is a part of the correlation coefficient matrix calculated from the data set corresponding to the production line. It is a correlation coefficient matrix in which a part of the correlation coefficient matrix shown in FIG. 6 is replaced. It is a flowchart of the data analysis method in this embodiment. It is a flowchart of the data analysis method in 2nd Embodiment.

<First Embodiment>
FIG. 1 is a block diagram of a data analyzer 10 as an embodiment of the present invention. As shown in FIG. 1, the data analyzer 10 exhibits various characteristics of the raw material powders A and B in the process from the processing of the raw material powders A and B on the production lines L1 and L2 to the production of the product. The raw material data DTS to be represented and the product data DTL1 and DTL2 to represent various characteristics of the product are acquired. The data analyzer 10 evaluates the relationship between various characteristics in each acquired data.

The data analyzer 10 includes a CPU (Central Processing Unit) 1, a ROM (Read Only Memory) 2, a RAM (Random Access Memory) 3, and a storage unit 9. The CPU 1 functions as the acquisition unit 4 and the evaluation unit 5 by expanding and executing the computer program stored in the ROM 2 in the RAM 3. The storage unit 9 stores various data such as raw material data DTS and product data DTL1 and DTL2. The storage unit 9 is composed of a hard disk drive (HDD: Hard Disk Drive) or the like.

In the manufacturing process of the present embodiment, as shown in FIG. 1, a raw material in which a binder is added to two types of raw material powders A and B is divided into a production line L1 and a production line L2. In each of the production lines L1 and L2, the raw materials are agitated, compression-molded in the shape of gears, sintered for a predetermined temperature and a predetermined time, and shipped as a product. The acquisition unit 4 acquires the raw material data DTS and the product data DTL1 and DTL2 collected in the manufacturing process. The raw material data DTS is independent data that does not depend on the production lines L1 and L2, which are the analysis targets of the present embodiment, before the binder is added. On the other hand, the product data DTL1 and DTL2 are dependent data depending on the production lines L1 and L2. As a method of acquiring the raw material data DTS and the product data DTL1 and DTL2, the data may be input via an input unit such as a keyboard or a mouse (not shown), or the data may be acquired by wired communication or wireless communication. .. In the following, the case where the relevance of the collected data is evaluated for each production line L1 and L2 will be illustrated. Therefore, the production lines L1 and L2 correspond to a specific process.

FIG. 2 is an example of a data set DS including raw material data DTS and product data DTL1 and DTL2. FIG. 2 shows various characteristics associated with the six product IDs as a table. For example, in product ID 1, the raw material lot ID is a, the purity of the raw material powder A is 6.2, the water content of the raw material powder A is 1.1, the production line is the production line L1, and the number of rotations of the stirring blade during stirring. Is 3.2, the weight before compression is 5.3, and the hardness of the product is 10.5. It should be noted that the numerical values representing these characteristics represent numerical values with respect to the normalized reference value, and therefore there is no unit. In other embodiments, the unnormalized measurements themselves may be the data.

As shown in FIG. 2, the product IDs 1 to 3 are manufactured by the production line L1, and the product IDs 4 to 6 are manufactured by the production line L2. The purity of the raw material powder A and the water content of the raw material powder A shown in FIG. 2 are independent data that do not depend on the production lines L1 and L2. On the other hand, the rotation speed of the stirring blade, the weight before compression, and the hardness are dependent data depending on the production lines L1 and L2. Product data as dependent data DTL1 and DTL2 are data representing various characteristics in each process in the manufacturing process and processing conditions for processing each raw material. In addition, as other independent data in this embodiment, the particle size of the raw material powders A and B and the like can be mentioned. In addition, other dependent data in this embodiment include the torque of the stirring blade as the processing conditions, the compressive load, the sintering temperature, and the sintering time, and the weight after compression and the sintering as the characteristics of the processed product. Later weight and so on.

The evaluation unit 5 evaluates the relevance of various characteristics in the data set DS shown in FIG. 2 by using all of the raw material data DTS and the product data DTL1 and DTL2. The evaluation unit 5 evaluates the relationship between various characteristics of the raw material powders A and B by using all of the raw material data DTS that do not depend on the production lines L1 and L2. The evaluation unit 5 of the present embodiment calculates the correlation coefficient matrix using the raw material data DTS shown in FIG.

FIG. 3 is an example of a correlation coefficient matrix calculated from the raw material data DTS. The correlation coefficient between the purity of the raw material powder A shown in FIG. 3 and the water content of the raw material powder A is 0.7. This correlation coefficient matrix is derived from the six raw material lots IDa to IDf in FIG. Next, the evaluation unit 5 divides the data set DS of FIG. 2 into a production line L1 and a production line L2.

FIG. 4 is a data set DS1 corresponding to the production line L1. FIG. 5 is a data set DS2 corresponding to the production line L2. The data set DS1 shown in FIG. 4 is composed of three raw material data DTS1s of raw material lots IDa to IDc processed on the production line L1 and product data DTL1. The data set DS2 shown in FIG. 5 is composed of three raw material data DTS2 of raw material lots IDd to IDf processed on the production line L2 and product data DTL2.

The evaluation unit 5 calculates a correlation coefficient matrix for each of the divided data sets DS1 and DS2 shown in FIGS. 4 and 5. FIG. 6 is a part of the correlation coefficient matrix calculated from the data set DS1 corresponding to the production line L1. The correlation coefficient matrix shown in FIG. 6 is calculated based on various characteristics of product IDs 1 to 3. Therefore, the numerical value of the range RG1 surrounded by the thick line in FIG. 6 is calculated from various characteristics of the three raw material lots IDa to IDc. The numerical value of the range RG1 is calculated from the independent data that does not depend on the production line L1.

FIG. 7 is a correlation coefficient matrix in which a part of the correlation coefficient matrix shown in FIG. 6 is replaced. When the evaluation unit 5 calculates the correlation coefficient matrix shown in FIG. 6, it replaces the numerical value of the range RG1 corresponding to the independent data with the correlation coefficient matrix (FIG. 3) calculated from the raw material data DTS. Therefore, the numerical value of the range RG1 surrounded by the thick line in FIG. 7 is calculated from various characteristics in all six raw material lots of the raw material lots IDa to IDf. The evaluation unit 5 uses the correlation coefficient matrix after the range RG1 substitution shown in FIG. 7 as an input and evaluates the degree of relevance between various characteristics using a Gaussian graphical model.

FIG. 8 is a flowchart of the data analysis method in the present embodiment. In the data analysis method, first, the acquisition unit 4 acquires the data set DS including the raw material data DTS and the product data DTL1 and DTL2 (step S1). The evaluation unit 5 calculates the correlation coefficient matrix shown in FIG. 3 from the raw material data DTS that does not depend on the production lines L1 and L2 (step S2). The evaluation unit 5 divides the data set DS of FIG. 2 into the data set DS1 of the production line L1 and the data set DS2 of the production line L2 (step S3). The evaluation unit 5 calculates the correlation coefficient matrix from the respective data sets DS1 and DS2 of the production line L1 and the production line L2 (step S4). The evaluation unit 5 sets data for each production line L1 and L2 in the range RG1 (FIG. 6) corresponding to the independent data in the correlation coefficient matrix corresponding to each of the data sets DS1 and DS2 of the production lines L1 and L2. The correlation coefficient matrix (FIG. 3) obtained without dividing the sets DS1 and DS2 is replaced (step S5). The evaluation unit 5 evaluates the relationship between various characteristics by inputting the replacement correlation coefficient matrix (FIG. 7) into the Gaussian graphical model (step S6), and ends the data analysis method. In addition, step S1 corresponds to the acquisition process, and steps S2 to S6 correspond to the evaluation process.

As described above, in the data analyzer 10 of the present embodiment, the acquisition unit 4 acquires the raw material data DTS and the product data DTL1 and DTL2 collected in the manufacturing process. The evaluation unit 5 evaluates the relevance of the data set DS by using all of the raw material data DTS before being divided into the production lines L1 and L2 and the product data DTL1 and DTL2 for each production line L1 and L2. .. That is, as the independent data used for the evaluation of the relevance, all the raw material data DTSs that do not depend on the production lines L1 and L2 are used instead of the raw material data DTS1 and DTS2 divided for each production line L1 and L2. Therefore, as compared with the case where all of the data set DS is divided into the data sets DS1 and DS2 for each production line L1 and L2, it is possible to suppress a decrease in the number of samples of the raw material data DTS which is independent data. As a result, it is possible to suppress the occurrence of problems caused by a decrease in the number of samples such as the relevance deviating from the true value, and to correctly evaluate the relevance.

Further, in the data analyzer 10 of the present embodiment, the raw material data DTS which is independent data and the product data DTL1 and DTL2 which are dependent data are separated based on the production lines L1 and L2. Therefore, the relevance of the dataset DS including the independent data and the dependent data before and after processing on the production line can be evaluated.

Further, in the data analyzer 10 of the present embodiment, the raw material data DTS, which is independent data, is data representing the characteristics of the raw material. Further, the product data DTL1 and DTL2, which are dependent data, include data representing processing conditions on the production lines L1 and L2 and data representing the characteristics of the processed products on the production lines L1 and L2. Therefore, the data analyzer 10 of the present embodiment can evaluate the relationship between various characteristics of the raw material, and can determine the relationship between the various characteristics of the processed product and the relationship between the processing conditions on the production lines L1 and L2. Can be evaluated.

Further, in the data analyzer 10 of the present embodiment, the correlation coefficient matrix is used for the evaluation of the relevance of the data set DS. The evaluation unit 5 replaces the numerical value of the range RG1 corresponding to the independent data in the divided data sets DS1 and DS2 with the correlation coefficient matrix calculated from the raw material data DTS of the independent data. This makes it possible to perform relevance calculations without error in Gaussian graphical models due to accidentally high correlation coefficients between variables in independent data. In addition, when a Gaussian graphical model is applied, it is possible to suppress the occurrence of unsatisfactory definite matrix of the correlation coefficient matrix. Therefore, the data analyzer 10 of the present embodiment can evaluate the relevance of the variables as a network.

<Second Embodiment>
FIG. 9 is a flowchart of the data analysis method in the second embodiment. The data analysis device that implements the data analysis method of the second embodiment is the same as the data analysis device 10 of the first embodiment. The evaluation of the data analysis method of the first embodiment is different from the evaluation by the Gaussian graphical model using the correlation coefficient matrix. In the second embodiment, a data analysis method different from that of the first embodiment will be described, and description of the same configuration as that of the first embodiment will be omitted.

As shown in FIG. 9, in the data analysis method of the second embodiment, the acquisition unit 4 first acquires the data set DS (FIG. 2) including the raw material data DTS and the product data DTL1 and DTL2 (step S11). ). The evaluation unit 5 calculates the correlation coefficient from the raw material data DTS that does not depend on the production lines L1 and L2 (step S12). The evaluation unit 5 divides the data set DS into the data set DS1 of the production line L1 and the data set DS2 of the production line L2 (step S13). The evaluation unit 5 has a correlation coefficient between various characteristics in the product data DTL1 and DTL2, various characteristics in the product data DTL1 and DTL2, and raw material data from the respective data sets DS1 and DS2 of the production line L1 and the production line L2. The correlation coefficient between the various characteristics in the above is calculated (step S14).

In the data analysis method of the second embodiment, the evaluation unit 5 evaluates the relevance of the respective data sets DS1 and DS2 of the product data DTL1 and DTL2 by using the raw material data DTS and the product data DTL1 and DTL2. Therefore, as in the first embodiment, the data analyzer 10 of the second embodiment relates various characteristics between the raw material data DTS in each of the data sets DS1 and DS2 divided for each production line L1 and L2. Gender can be evaluated more accurately.

<Modified example of this embodiment>
The present invention is not limited to the above-described embodiment, and can be implemented in various aspects without departing from the gist thereof. For example, the following modifications are also possible.

In the first embodiment and the second embodiment, the data analysis device 10 that implements the data analysis method has been described as an example, but the data analysis device 10 can be variously modified. For example, the data analysis device 10 includes a microphone that accepts voice input, a monitor that displays the relevance evaluation result as an image, an output device that outputs the relevance evaluation result as a log, a speaker that outputs the evaluation result by voice, and the like. You may be.

In the first embodiment, the input of the correlation coefficient matrix to the Gaussian graphical model is used for the evaluation of the relevance, but the input of the variance-covariance matrix to the Gaussian graphical model may be used. In this case, the relationship between various characteristics in the raw material data DTS is also expressed by the variance-covariance matrix. Further, the evaluation unit 5 does not necessarily have to input the calculated correlation coefficient matrix into the Gaussian graphical model, and may calculate the correlation coefficient matrix and end the data analysis method (FIG. 8).

In the first embodiment, the relationship between the raw material data DTS that does not depend on the production lines L1 and L2 and the data set DS including the product data DTL1 and DTL2 that depend on the production lines L1 and L2 was evaluated, but other than that. The relevance of the dataset of the combination of may be evaluated. For example, when a plurality of different raw material data delivered from different material manufacturers are processed on the production line L1, each of the plurality of raw material data is independent data that does not depend on the production line L1. In this case, for example, the evaluation unit 5 may calculate a correlation coefficient matrix representing the relevance in the plurality of independent data corresponding to each of the raw material lots.

The specific process that separates the independent data and the dependent data may be a process other than the production lines L1 and L2 that realize the production process. For example, the process may be divided into each step of stirring and sintering in the production lines L1 and L2. For example, the relevance of the collected data may be evaluated for each raw material lot. In this case, when the raw material lot IDs are different, the processing is performed assuming that the specific process is different.

The dependent data and the independent data may be data other than the data representing various characteristics. For example, the raw material data DTS may be the supplier of the raw material, the purchase price, the purchase time, and the like. Further, the product data DTL1 and DTL2 may be a measured value, an evaluation value, an index showing a difference between the measured value and the evaluation value, a selling price, and the like.

Although the present embodiment has been described above based on the embodiments and modifications, the embodiments of the above-described embodiments are for facilitating the understanding of the present embodiment, and do not limit the present embodiment. This aspect may be modified or improved without departing from its spirit and claims, and this aspect includes its equivalents. In addition, if the technical feature is not described as essential in the present specification, it may be deleted as appropriate.

1 ... CPU
2 ... ROM
3 ... RAM
4 ... Acquisition unit 5 ... Evaluation unit 9 ... Storage unit 10 ... Data analyzer A, B ... Raw material powder DS, DS1, DS2 ... Data set DTL1, DTL2 ... Product data DTS, DTS1, DTS2 ... Raw material data IDa to IDf ... Raw material Lot ID1 to ID6 ... Product L1, L2 ... Production line RG1 ... Range

Claims (6)

It ’s a data analysis method.
Acquisition to acquire a data set that includes data collected in a manufacturing process that does not depend on a specific process to be analyzed during the manufacturing process and dependent data that depends on the specific process. Process and
An evaluation step of evaluating the relevance of the data set using all of the independent data and the dependent data corresponding to the specific process.
A data analysis method that includes. The data analysis method according to claim 1.
A data analysis method, wherein the specific process is a production line that realizes the production process. The data analysis method according to claim 2.
The independent data is data representing various characteristics of each raw material used in the manufacturing process.
The dependent data is data representing at least one of various characteristics of a processed product acquired in each process in the manufacturing process and processing conditions for processing each raw material, which is a data analysis method. The data analysis method according to any one of claims 1 to 3.
The evaluation of the relevance is performed on the element corresponding to the independent data in the variance-covariance matrix or the correlation coefficient matrix obtained by dividing the data set for each specific process. A data analysis method performed by inputting a matrix using a variance-covariance matrix or a correlation coefficient matrix obtained without dividing the data set into a Gaussian graphical model for each data set. It is a data analyzer
Acquisition to acquire a data set that includes data collected in a manufacturing process that does not depend on a specific process to be analyzed during the manufacturing process and dependent data that depends on the specific process. Department and
An evaluation unit that evaluates the relevance of the data set using all of the independent data and the dependent data corresponding to the specific process.
A data analysis method that includes. It ’s a computer program
Acquisition to acquire a data set that includes data collected in a manufacturing process that does not depend on a specific process to be analyzed during the manufacturing process and dependent data that depends on the specific process. Function and
An evaluation function for evaluating the relevance of the data set using all of the independent data and the dependent data corresponding to the specific process.
A computer program that makes a computer realize.

Images