JP2018155675A - Analysis support system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an analysis support system capable of supporting a sufficient analysis by decreasing a load in the on-site, while satisfying the needs required by the on-site even when data having wide information in the onsite is measured.SOLUTION: An analysis support system according to the present invention measures on-site data which is a response at the time when a predetermined input is given to a measured object in an on-site, and analyzes a state of the measured object by a computer based on a portion of the response of the on-site data, and accumulates a plurality of on-site data in the data base. The on-site analysis device capable of sharing this accumulated on-site data with the other analysis device is compared with the other response including a portion of the response of the shared on-site data in an off-site, thereby analyzing a state of the measured object.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an analysis support system that analyzes data acquired by a sensor or the like.

  Conventionally, Patent Document 1 is known as a technique for analyzing information measured at a so-called site (hereinafter referred to as on-site) such as a manufacturing factory or a maintenance factory. In this gazette, characteristics are measured on-site, quality is determined, and characteristics and pass / fail data are stored. Then, the data is transmitted to a laboratory, an analysis company, etc. (hereinafter referred to as off-site) that specially analyzes data, and analysis such as statistical processing is performed.

JP 2003-98210 A

  As described in Patent Document 1, as the data to be measured on-site, for example, in the case of the length of an object, only one piece of information is obtained, so the length of the product is measured on the production line, and the quality is determined. As long as the process is executed, a particularly complicated analysis is not required, and only one data is processed. Therefore, in the off-site processing, only one data needs to be statistically processed. However, for example, when spectral data is measured in a non-contact manner by light irradiation, data in a wide wavelength range can be acquired unlike simple length information. When analyzing all of this data on-site, the analysis may not be able to keep up with the on-site production line tact time. In addition, depending on the on-site environment, it is difficult to install an expensive analysis device, and it cannot be said that it meets the on-site needs.

  The purpose of the present invention is to support sufficient analysis by reducing the burden on-site while meeting the needs required on-site, even when data with a wide range of information is measured on-site. Is to provide a simple analysis support system.

In the analysis support system of the present invention, an on-site measurement unit that measures on-site data that is a response when a predetermined input is given to an object to be measured on-site,
An on-site analysis unit for analyzing the state of the object to be measured by a computer based on a part of the response of the on-site data;
A database for storing a plurality of the on-site data;
A sharing unit that shares the on-site data stored in the database with other analysis devices;
An on-site analyzer with
Off-site analysis of the state of the object to be measured by comparing other responses including part of the response of the on-site data shared by the sharing unit between the plurality of on-site data. An off-site analysis device having a site analysis unit;
Equipped with.

  Therefore, on-site analysis is performed based on a part of the data obtained on-site, so that quick analysis that meets on-site needs can be realized. Also, by accumulating the obtained on-site data, including some, and sharing this accumulated data off-site, based on other responses including some on-site performed off-site. Sufficient analysis can be performed and on-site analysis can be supported.

1 is a system diagram illustrating an analysis support system according to a first embodiment. 3 is a schematic diagram illustrating an on-site analysis program design method according to Embodiment 1. FIG. 2 is a diagram illustrating a calibration curve of Example 1. FIG. It is the schematic showing a part of screen displayed with the on-site analysis program of Example 1. FIG. 1 is a schematic diagram illustrating the operation of an off-site analysis apparatus according to a first embodiment. It is the schematic showing the spectrum data analysis process performed by the off-site analysis program of Example 1. FIG.

  FIG. 1 is a system diagram illustrating an analysis support system according to the first embodiment. In Example 1, as a specific example, a production line for producing resin parts from resin pellets is shown on-site. In addition, laboratories and analysis companies that are equipped with various high-function analyzers and provide on-site analyzers used on-site are shown as off-site. Although the on-site and the off-site are illustrated as being connected via the network 6, they are not necessarily connected via the network 6, and data may be transmitted / received by sending a storage medium or the like.

  On-site, the resin pellet 2 as a raw material is conveyed to the production line 3. The production line 3 has a near-infrared spectrometer 1 (hereinafter referred to as a small NIR 1) having a measurement wavelength range of 900 nm to 1700 nm. The small NIR 1 irradiates the resin pellet 2 with light, acquires spectral data from the reflection, and transmits it to the computer 4. Since the computer 4 does not require a large processing capacity, it may be a notebook personal computer or the like. The computer 4 has an on-site analysis program for simply analyzing the state of the resin pellet 2. The on-site analysis program extracts part of the spectral data from the wavelength region related to the moisture content of the resin pellet 2, and based on this extracted part of the data, qualifies as a material. judge. If the material is qualified, OK is displayed on the monitor 4a connected to the computer 4. If the material is unqualified, NG is displayed.

  The resin pellet 2 is manufactured by injection molding into a mold. The manufactured product is subjected to visual inspection and strength test in the inspection process, and the inspection result is transmitted to the computer 4. Various data such as spectral data obtained in the above process, analysis results with on-site analysis program, inspection results, inspection conditions, manufacturing environment temperature, humidity, atmosphere, date, weather, etc., process manager, etc. The database 5 is linked to the unique information (manufacturer, lot number, delivery time, storage state, etc.) of the resin pellet 2 (hereinafter referred to as a data unit) and connected to the computer 4 via the network 6. Accumulated. All data units stored in the database 5 are described as bulk data. The small NIR 1, the on-site analysis program, and the database 5 constitute an on-site analysis apparatus.

  Off-site has a large computer 10 capable of analyzing bulk data. The large computer 10 has an off-site analysis program for analyzing the bulk data stored in the database 5. The off-site analysis program reads data in all wavelength regions of the accumulated spectral data and analyzes the change amount and change tendency of the spectral data at each wavelength. Then, identification of the wavelength range in which the spectral data has changed, extraction of the change amount and change tendency, comprehensive analysis of the causal relationship between the change amount and change tendency and other data in the data unit, and the analysis result Is displayed on the monitor 10a.

  If the cause is identified from the analysis result of the bulk data and it is determined that the detailed sample analysis is not particularly necessary, a report on correction of the on-site analysis program, improvement of the on-site environment, and the like is created. On the other hand, when it is narrowed down to a plurality of causes from the analysis result of the bulk data, but it is determined that the cause needs to be specified based on the detailed sample analysis, the necessary high-function analyzer is selected. If detailed sample analysis is required, on-site requests for samples (OK, NG resin pellets 2, products, etc.) are requested.

  Off-site is a Fourier transform type near infrared spectroscopic analyzer 11 (FTNIR), Fourier transform type infrared spectroscopic analyzer 12 (FTIR), ultraviolet excitation visible light spectroscopic analyzer 13, Raman scattering spectroscopic analysis as a high performance analyzer. Device 14 (Raman), optical microscope device 15 (OM), electron microscope device 16 (SEM), Karl Fischer moisture analyzer 17, gas chromatography mass spectrometer 18 (GC-MS), nuclear magnetic resonance analyzer 19 (NMR) ). Although the above eight types of analyzers have been described from the viewpoint of resin pellets 2 and resin products, thermal analysis, viscoelastic analysis, elemental analysis, chemical analysis, X-ray diffraction analyzer, small-angle X-ray scattering analysis are performed as necessary. An analyzer or an analyzer of an X-ray photoelectron spectrometer can also be used. For example, the FTNIR 11 has a wider measurable wavelength range than the small NIR 1 that measures on-site, and can measure with high resolution and high sensitivity. When the necessary high-function analyzers are narrowed down by the analysis of the off-site analysis program, the minimum required analyzer is selected from the analyzers. The high-function analyzer and the off-site analysis program constitute an off-site analyzer.

(Provide on-site analysis equipment)
At off-site, grasp on-site needs and design a simple on-site analysis program. FIG. 2 is a schematic diagram illustrating an on-site analysis program design method according to the first embodiment. On-site, it is required to stabilize the amount of water contained in the resin pellet 2. For example, if the elapsed time from when the resin pellet 2 is delivered on-site to the production line 3 is long, the resin pellet 2 may adsorb moisture. If injection molding is performed in a state where the amount of water contained in the resin pellet 2 is large, voids are generated in the product due to vaporization of the water in the resin, which may result in insufficient strength and appearance defects. The upper part of FIG. 2 shows spectral data obtained by measuring a plurality of resin pellet 2 samples. The resin pellet 2 made of nylon absorbs light near 1450 nm due to the influence of OH groups due to the absorption of moisture. That is, as shown in the enlarged view in the lower part of FIG. 2, when there are few OH groups, the spectrum data near 1450 nm tend to be convex upward. However, when moisture is adsorbed, the tendency changes from an upward tendency to a gradual decrease. By utilizing this change tendency at 1450 nm, moisture content can be measured in a non-contact and non-destructive manner.

  FIG. 3 is a diagram illustrating a calibration curve of Example 1. The horizontal axis represents the actual water content measured by the Karl Fischer moisture measuring device 17, and the vertical axis represents the predicted water content measured by the small NIR 1. A line corresponding to the actually measured moisture content and the predicted moisture content is set as a calibration curve. Since it can be said that the result of measurement without greatly deviating from this calibration curve (for example, 2.0 wt% or less, measurement points A, B, and C shown in FIG. 3) can accurately predict the moisture content with the small NIR1, A predetermined area including the calibration curve is set as an OK area. On the other hand, if the resin pellet 2 contains an excessive amount of water (for example, 2.0 wt% or more, measurement points D and E shown in FIG. 3), voids are generated. Set. Further, the resin pellet 2 including an excessive amount of water (for example, 2.8 wt or more, measurement point F shown in FIG. 3) far away from the daily management area is also set as the NG area. By creating a pass / fail judgment file with these conditions and incorporating it into the on-site analysis program, only the data near 1450 nm, which is part of the spectrum data measured with the compact NIR1, is used and qualified as a material. Can be determined.

  Here, when the resin pellet 2 is measured by the small NIR 1, spectral data of 900 nm to 1700 nm can be obtained, so that it is conceivable to analyze using data in the entire range. However, if all the data is analyzed, calculation load and calculation time are required, and there are cases where the tact time of the on-site production line cannot be met. In addition, what can be routinely managed on-site is the amount of water, and it is difficult to routinely measure and manage data analysis due to other deterioration or the like. Therefore, by focusing on some of the data that is considered to be the most important on-site, we will provide a program that can realize quality control that matches the tact time without increasing the size of the device or complicating the program. It was.

(Operation of on-site analyzer)
FIG. 4 is a schematic diagram illustrating a part of a screen displayed by the on-site analysis program according to the first embodiment. As a result of analysis by the on-site analysis program, if it is determined that the region is an OK region, OK is displayed on the monitor 4a, and the process proceeds to the next measurement result without any problem. At this time, even if the result of the on-site analysis program is OK, if the result is NG in the inspection process, NG is displayed on the monitor 4a. Therefore, it is necessary to investigate the cause that cannot be determined by the on-site analysis program. If it is determined that the result of the on-site analysis program is the NG area, NG is displayed on the monitor 4a. At this time, if NG is displayed with a low frequency, the resin pellet 2 may be simply excluded from the production line. However, if the frequency of NG increases, the cause of the problem must be investigated.

  In the on-site analysis program, as shown in FIG. 4 (i), a “detailed analysis request button” is displayed when NG is displayed on the monitor 4a. When the on-site manager or the like presses the detailed analysis request button, a “database sharing permission button” is displayed as shown in FIG. 4 (ii). Here, of the bulk data stored in the database 5, a data range that permits sharing is set. This data range may be selected as appropriate from within the data unit, or all may be selected. Further, a period during which data is accumulated may be set. When the database sharing permission button is pressed in this state, a signal for permitting access to the database 5 is transmitted to the off-site together with a detailed analysis request notification. On the next screen, as shown in FIG. 4 (iii), the statuses “database sharing” and “detailed analysis requested” are displayed, and “access permission start button” is displayed.

(Operation of off-site analysis equipment)
FIG. 5 is a schematic diagram illustrating the operation of the off-site analysis apparatus according to the first embodiment. When the detailed analysis request notification and the access permission signal to the database 5 are received off-site, the database 5 is accessed within a range where the access permission is obtained, and the permitted data is received. In off-site, a large amount of spectrum data is read from bulk data. In addition, the chemical structure is specified from the material information, and the spectrum peak is selected based on the chemical structure. In addition to selecting spectrum peaks from the spectrum analysis results and manufacturing conditions, the environment humidity, environment temperature, and atmosphere are read from the environment information, and spectrum analysis is performed from the bulk data and the selected spectrum peaks. For example, the change amount or change tendency of the spectrum peak is extracted. In addition, as for the method of extracting the change amount and change tendency of the spectrum peak, there are parametric analysis, non-parametric analysis, univariate analysis, multivariate analysis, linear analysis, nonlinear analysis, and other statistical methods known to those skilled in the art. Can be mentioned. Multivariate analysis is analysis that determines patterns in apparently disordered data, including principal component analysis ("PCA"), discriminant analysis ("DA"), PCA-DA, canonical correlation ( "CC"), cluster analysis, partial least squares ("PLS"), predictive linear discriminant analysis ("PLDA"), neural networks, and pattern recognition techniques, but are not particularly limited.

  FIG. 6 is a schematic diagram illustrating a spectrum data analysis process executed by the off-site analysis program according to the first embodiment. The resin pellet 2 is made of nylon-6, and its chemical formula is represented by a chain structure below in FIG. When nylon-6 absorbs moisture, the spectral peak value near 1450 nm decreases due to the effect of the O-H group. Further, when nylon-6 deteriorates, the spectral peak value that protrudes downwards near 1050 nm gradually changes due to the N—H group, and the spectral peak value near 1340 nm decreases due to the C—H group. Here, some spectral data was used on-site, but off-site can be used to grasp the changing tendency of the molecular structure other than the water content by expanding the wavelength range of the spectral data. . In addition, since a large amount of bulk data is processed, information that cannot be grasped on-site can be extracted. Specifically, the extracted change amount and change tendency can be analyzed in association with manufacturing conditions and environmental information. Therefore, it becomes possible to grasp the causal relationship between the resin pellet 2 and the product that has become NG and the environmental condition, and the causal relationship with the manufacturing condition, and an appropriate improvement measure can be proposed for on-site.

  Next, when it can be determined that the cause of NG (measurement point F shown in FIG. 3) generated on-site by the above-described data analysis has been investigated, analysis using a high-function analyzer is not performed. On the other hand, if it cannot be determined that the cause of NG has been sufficiently investigated, the sample is requested to be provided on-site based on the extracted information. Since the analysis using the bulk data has already been performed, the high-function analyzers that are considered necessary are narrowed down. Therefore, the required sample can be specifically requested from the off-site. When a sample is provided from on-site, the principal component analysis (qualitative analysis or quantitative analysis) of the sample is performed by the selected high-performance analyzer. Based on the identified cause, a report is generated and reported on-site. In some cases, the accuracy of analysis of the on-site analysis apparatus can be improved by resetting the calibration curve of the pass / fail determination file or resetting the NG area and updating the on-site pass / fail determination file.

As described above, Example 1 has the following effects.
(1) Small NIR1 (onsite) that measures spectral data (onsite data) that is the response when light is irradiated (given predetermined input) to the resin pellet 2 (object to be measured) onsite Measuring section),
An on-site analysis program (on-site analysis unit) for analyzing the state of the resin pellet 2 by the computer 4 based on a spectrum peak value at 1450 nm which is a part of the response of the spectrum data;
A database 5 for storing a plurality of spectral data;
A detailed analysis request button for sharing spectrum data stored in the database 5 with an off-site analysis device (other analysis device), a database sharing permission button (sharing unit),
An on-site analyzer with
The state of the resin pellet 2 by comparing the spectrum peak shapes at 1050 nm and 1340 nm, which are other responses including part of the response of the spectrum data shared by the database sharing permission button, off-site, between the plurality of spectrum data An off-site analysis apparatus having an off-site analysis program (off-site analysis unit) for analyzing
Equipped with.
Therefore, on-site analysis is performed based on a part of the data obtained on-site, so that quick analysis that meets on-site needs can be realized. Also, by accumulating the obtained on-site data, including some, and sharing this accumulated data off-site, based on other responses including some on-site performed off-site. Sufficient analysis can be performed and on-site analysis can be supported.

(2) On-site data is spectral data representing a response when irradiated with light. Spectral data can be acquired simultaneously in addition to the wavelength of interest when light is irradiated. Therefore, more appropriate analysis support can be implemented by using some responses on-site and effectively utilizing other responses including some responses off-site.
(3) Since it is a database sharing permission button displayed on the monitor 4a (computer screen) as a sharing unit, it is possible to select sharing of the database 5 based on on-site determination. Therefore, since information management can be appropriately performed on the site side, a database sharing destination can be freely set.
(4) The on-site analysis program creates a pass / fail judgment file (analysis unit) that analyzes whether the spectral peak value at 1450 nm (part of the response) of the spectral data satisfies a predetermined condition (for example, 2.0 wt% or less). Have
The off-site analysis program updates the pass / fail determination file of the on-site analysis program based on the analysis result.
Therefore, an appropriate improvement measure can be proposed for on-site.

(5) The database 5 is arranged on the network 6,
The sharing unit transmits a signal permitting access from the off-site analysis device to a predetermined range set by the on-site analysis device among the plurality of spectrum data stored in the database 5 to the off-site analysis device. It is.
Therefore, since shared data can be set in the database, all information including confidential information can be stored in the database on-site.
(6) The off-site analysis apparatus includes an FTNIR 11 (off-site measurement unit) having higher resolution than the small NIR 1.
Therefore, a more detailed analysis can be performed as necessary.
(7) The high-performance analyzer (off-site measurement unit) has a plurality of high-performance analyzers that measure data different from the spectrum data measured by the small NIR 1 for the resin pellet 2;
The off-site analysis program selects an optimum high-function analyzer from a plurality of high-function analyzers based on the analysis result.
Therefore, the analysis period and analysis cost can be suppressed.
(8) The high-performance analyzer is an optical microscope, image analysis using a scanning or transmission electron microscope, structural analysis related to the transmission electron microscope, infrared spectroscopy, nuclear magnetic resonance analysis, near infrared spectroscopy, ultraviolet. Excitation visible light spectroscopy, Raman spectroscopy, Auger electron spectroscopy, elemental analysis, thermal analysis, viscoelastic analysis, chemical analysis, analysis related to scanning probe microscope, nanoindentation analysis, mechanical strength test, X-ray scattering analysis And two or more means selected from the group consisting of X-ray photoelectron spectroscopy, X-ray diffraction analysis, X-ray small angle scattering analysis, neutron diffraction analysis, neutron small angle scattering analysis, and X-ray absorption fine structure analysis.
Therefore, the object to be measured can be measured from all sides, and the cause can be investigated accurately.

(Other examples)
The present invention has been described based on the first embodiment. However, the present invention can be applied to other embodiments without departing from the scope of the present invention. In Example 1, the present invention was applied with respect to the moisture absorption of the resin pellets. For example, the presence or absence of oil adhesion on the metal surface may be measured on-site, or the curing determination of the adhesive may be measured on-site. It is also possible to identify the type of raw material with the powder raw material unopened, measure the compounding ratio of the molded resin product and the amount of additive on-site, and use photocatalytic coating. The film thickness of a coating such as the above may be measured on-site, or the deterioration determination of oil such as engine oil may be measured on-site.

  In the first embodiment, the spectrum data is used. However, the present invention is not limited to the spectrum data, and may be applied to the case where other data other than the desired data can be simultaneously acquired by one measurement. Specifically, an image of the product is taken and only the length of the product is determined from the image on-site, but other information (surface state, shape, etc.) may be used off-site. In addition, even when the specs of computers that can be brought on-site are limited, it is effective to limit the contents to be analyzed on-site.

  In the first embodiment, an example is shown in which on-site and off-site are connected via a network. However, even when the network is not connected to the network, data may be transmitted and received by sending a storage medium or the like. Similarly, the database need not be on the network, and may be built in an off-site, cloud, or on-site computer.

  In the first embodiment, the on-site small NIR 1 and the off-site FTNIR 11 have different resolutions. However, the on-site and off-site resolutions may be the same, or the off-site resolution. Can be high.

DESCRIPTION OF SYMBOLS 1 Near infrared spectroscopy analyzer 2 Resin pellet 3 Production line 4 Computer 4a Monitor 5 Database 6 Network 10 Large computer 11 Fourier transform near infrared spectroscopy analyzer 12 Fourier transform infrared spectrometer 13 Ultraviolet excitation visible light spectroscopy Device 14 Raman scattering spectrometer 15 Optical microscope device 16 Electron microscope device 17 Karl Fischer moisture analyzer 18 Gas chromatography mass spectrometer 19 Nuclear magnetic resonance analyzer

Claims (8)

An on-site measurement unit that measures on-site data that is a response when a predetermined input is given to an object to be measured on-site;
An on-site analysis unit for analyzing the state of the object to be measured by a computer based on a part of the response of the on-site data;
A database for storing a plurality of the on-site data;
A sharing unit that shares the on-site data stored in the database with other analysis devices;
An on-site analyzer with
Off-site analysis of the state of the object to be measured by comparing other responses including part of the response of the on-site data shared by the sharing unit between the plurality of on-site data. An off-site analysis device having a site analysis unit;
An analysis support system characterized by comprising The analysis support system according to claim 1,
The analysis support system, wherein the on-site data is spectrum data representing a response when irradiated with light. The analysis support system according to claim 1 or 2,
The analysis support system, wherein the sharing unit is a database sharing permission button displayed on the screen of the computer. The analysis support system according to any one of claims 1 to 3,
The on-site analysis unit is an analysis unit that analyzes whether a part of the response of the on-site data satisfies a predetermined condition,
The off-site analysis unit updates the predetermined condition of the on-site analysis unit based on an analysis result. The analysis support system according to any one of claims 1 to 4,
The database is located on a network;
The sharing unit outputs a signal permitting access from the off-site analysis device to a predetermined range set by the on-site analysis device among the plurality of on-site data stored in the database. An analysis support system characterized by being a means for transmitting to an analysis apparatus. In the analysis support system according to any one of claims 1 to 5,
The off-site analysis apparatus includes an off-site measurement unit having higher resolution and higher sensitivity than the on-site measurement unit. The analysis support system according to claim 6,
The off-site measurement unit has a plurality of off-site measurement units that measure different data from the on-site data measured by the on-site measurement unit, with respect to the object to be measured.
The off-site analysis unit selects an optimal off-site measurement unit from the plurality of off-site measurement units based on the analysis result. The analysis support system according to claim 7,
The plurality of off-site measurement units include optical analysis, image analysis using a scanning or transmission electron microscope, structural analysis related to the transmission electron microscope, infrared spectroscopy, nuclear magnetic resonance analysis, near infrared spectroscopy, ultraviolet Excitation visible light spectroscopy, Raman spectroscopy, Auger electron spectroscopy, elemental analysis, thermal analysis, viscoelastic analysis, chemical analysis, analysis related to scanning probe microscope, nanoindentation analysis, mechanical strength test, X-ray scattering method And two or more means selected from the group consisting of X-ray photoelectron spectroscopy, X-ray diffraction analysis, X-ray small angle scattering analysis, neutron diffraction analysis, neutron small angle scattering analysis and X-ray absorption fine analysis Analysis support system.