JP2021149733A - Countermeasure selection support system, and method therefor - Google Patents

Countermeasure selection support system, and method therefor Download PDF

Info

Publication number
JP2021149733A
JP2021149733A JP2020050591A JP2020050591A JP2021149733A JP 2021149733 A JP2021149733 A JP 2021149733A JP 2020050591 A JP2020050591 A JP 2020050591A JP 2020050591 A JP2020050591 A JP 2020050591A JP 2021149733 A JP2021149733 A JP 2021149733A
Authority
JP
Japan
Prior art keywords
countermeasure
countermeasure method
selection support
method selection
support system
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.)
Granted
Application number
JP2020050591A
Other languages
Japanese (ja)
Other versions
JP7484281B2 (en
Inventor
大介 八木
Daisuke Yagi
大介 八木
秀徳 高井
Hidenori Takai
秀徳 高井
浩樹 渡邉
Hiroki Watanabe
浩樹 渡邉
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.)
Resonac Corp
Original Assignee
Showa Denko Materials Co 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 Showa Denko Materials Co Ltd filed Critical Showa Denko Materials Co Ltd
Priority to JP2020050591A priority Critical patent/JP7484281B2/en
Publication of JP2021149733A publication Critical patent/JP2021149733A/en
Application granted granted Critical
Publication of JP7484281B2 publication Critical patent/JP7484281B2/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Landscapes

  • General Factory Administration (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

To provide a countermeasure selection support system which, when a defect is generated in a film manufacturing process, determines a factor of the generation to implement a countermeasure therefor.SOLUTION: The present invention is directed to a countermeasure selection support system for supporting selection of countermeasures for generation of defect of a film manufactured by a manufacturing apparatus in a manufacturing line. The system has a storage device for storing a plurality of countermeasures for defect generation, and a processor for executing a countermeasure support program 302. The processor executes the countermeasure support program 302, which leads to reception of measured values from a plurality of sensors provided in the manufacturing apparatus, output of a model prediction value based on the measured values during a learning mode, reception of the measured values and visual inspection data of a manufactured film during a monitoring mode, calculation of priority of the countermeasure based on deviation of the model prediction value from the measured values during the monitoring mode, and display of the countermeasure and the priority on a client computer 203.SELECTED DRAWING: Figure 5

Description

本発明は、フィルム製造プロセスにおける欠陥要因を判別し、作業者の対策方法選定を支援する対策方法選定支援システム、及びその方法に係る。 The present invention relates to a countermeasure method selection support system that determines a defect factor in a film manufacturing process and supports the selection of a countermeasure method by an operator, and a method thereof.

連続的に搬送される長尺状可撓性の基材上に、所望の膜厚で塗布膜を形成するフィルム製造プロセスにおいて、生産性および歩留まりを向上させるためには、欠陥発生時に迅速かつ的確な対策方法を実施することが必要である。 In the film manufacturing process of forming a coating film with a desired film thickness on a long and flexible substrate that is continuously conveyed, in order to improve productivity and yield, it is necessary to quickly and accurately occur when a defect occurs. It is necessary to implement various countermeasures.

しかし、基材の搬送や塗布、乾燥等、複数の工程が連続して存在するフィルム製造プロセスでは、搬送中の基材の不具合や塗布液の粘度等の物性値の異常、塗工装置の設定値の最適化不足、乾燥不足等様々な欠陥要因が考えられる。このため、欠陥画像等の情報からこれらの欠陥要因を判別することが困難であり、現場での対策方法は作業者の勘と経験を頼りに試行錯誤しながら実施されている。また、対策方法の実施は一度製造ラインを停止して行うため、ライン停止による生産性の悪化が避けられない現状であり、製造ラインを停止することなく欠陥要因を特定し対策方法を実施することが望ましい。 However, in a film manufacturing process in which a plurality of processes such as transfer, coating, and drying of a base material are continuously present, defects in the base material during transportation, abnormalities in physical property values such as viscosity of a coating liquid, and setting of a coating device are used. Various defect factors such as insufficient optimization of values and insufficient drying can be considered. For this reason, it is difficult to determine the causes of these defects from information such as defect images, and on-site countermeasures are implemented by trial and error, relying on the intuition and experience of the operator. In addition, since the production line is stopped once and the countermeasure method is implemented, the current situation is that productivity deterioration due to the line stop is unavoidable. Therefore, it is necessary to identify the cause of the defect and implement the countermeasure method without stopping the production line. Is desirable.

関連する先行技術としては、特許文献１があり、製造ラインの装置に設けられたセンサの測定値、アクチュエータおよびそのドライバからのフィードバックデータ、製品検査データ、製造ラインの管理データおよびメンテナンス記録を含むデータを回収し、回収以前に蓄積されたデータの時間変化と照合し、その後の各装置の状態や故障等の発生時期を予測する管理システムが開示されている。 As a related prior art, there is Patent Document 1, which includes measured values of sensors provided in equipment of a production line, feedback data from an actuator and its driver, product inspection data, management data of a production line, and maintenance records. A management system is disclosed that collects data, collates it with the time change of data accumulated before collection, and predicts the state of each device and the time when a failure occurs after that.

特開２０１７−３２３８３号公報Japanese Unexamined Patent Publication No. 2017-323383

上述のように、フィルム製造工程は、基材の搬送工程や溶剤を含む塗布物の塗工工程、溶剤を乾燥させるための乾燥工程等の複数の連続する工程からなるため、欠陥発生時の候補として、搬送中の基材の不具合や塗布液の粘度等の物性値の異常、塗工装置の設定値の最適化不足、乾燥不足等様々な要因が考えられる。このため、欠陥発生時に根本要因を絞り込むことが難しく、要因に対応した適切な対策方法を実施することが困難である。 As described above, the film manufacturing process comprises a plurality of continuous processes such as a transfer process of the base material, a coating process of the coating material containing a solvent, and a drying process for drying the solvent. Various factors can be considered, such as a defect in the base material during transportation, an abnormality in physical property values such as the viscosity of the coating liquid, insufficient optimization of the set value of the coating device, and insufficient drying. Therefore, it is difficult to narrow down the root cause when a defect occurs, and it is difficult to implement an appropriate countermeasure method corresponding to the factor.

この課題に対して、上述した特許文献１に開示の方法では、センサの情報と欠陥発生時の欠陥要因とが対応していないため、センサの異常データを検知した後、どのような対策方法を実施すればよいかは、作業者が試行錯誤しなければならない。 In response to this problem, in the method disclosed in Patent Document 1 described above, the sensor information and the defect factor at the time of defect occurrence do not correspond to each other. Therefore, what kind of countermeasure method should be taken after detecting the abnormality data of the sensor. It is up to the operator to make trial and error as to whether or not it should be carried out.

本発明の目的は、上記の課題を解決するため、欠陥の要因と現場で実施されている対策方法との対応を予め紐づけておき、対策方法とその優先度を作業者に視覚的に提示することが可能な対策方法選定支援システム、及び方法を提供することを目的とする。 An object of the present invention is to link the cause of the defect with the countermeasure method implemented in the field in advance in order to solve the above-mentioned problems, and visually present the countermeasure method and its priority to the operator. The purpose is to provide a countermeasure method selection support system and methods that can be taken.

上述の目的を達成するために、本発明においては、製造ラインの製造装置で製造されるフィルムの欠陥発生時の対策方法選定を支援する対策方法選定支援システムであって、欠陥発生時の複数の対策方法を格納する記憶装置と、学習モードと監視モードを含む対策方法支援プログラムを実行するプロセッサを有する計算機サーバを備え、プロセッサが対策方法支援プログラムを実行することで、製造装置に設置された複数のセンサから測定値を受信し、学習モード中は測定値に基づきモデル予測値を出力し、監視モード中は測定値と製造されたフィルムの外観検査データを受信し、欠陥発生時に対策方法について、モデル予測値と監視モード中の測定値の乖離に基づき対策方法の優先度を算出する対策方法選定支援システムを提供する。 In order to achieve the above object, the present invention is a countermeasure method selection support system that supports selection of a countermeasure method when a defect occurs in a film manufactured by a manufacturing apparatus of a production line, and is a plurality of countermeasure method selection support systems when a defect occurs. A plurality of computers installed in a manufacturing device by having a storage device for storing countermeasure methods and a computer server having a processor for executing a countermeasure method support program including a learning mode and a monitoring mode, and the processor executing the countermeasure method support program. Receives the measured value from the sensor of, outputs the model predicted value based on the measured value in the learning mode, receives the measured value and the appearance inspection data of the manufactured film in the monitoring mode, and describes the countermeasure method when a defect occurs. We provide a countermeasure method selection support system that calculates the priority of countermeasure methods based on the discrepancy between the model predicted value and the measured value in the monitoring mode.

また、上記の目的を達成するため、本発明においては、計算機サーバにより、製造ラインの製造装置で製造されるフィルムの欠陥発生時の対策方法選定を支援する対策方法選定支援方法であって、計算機サーバは、欠陥発生時の複数の対策方法を格納する記憶装置と、学習モードと監視モードを含む対策方法支援プログラムを実行するプロセッサを有し、プロセッサが対策方法支援プログラムを実行することで、製造装置に設置された複数のセンサから測定値を受信し、学習モード中は測定値に基づきモデル予測値を出力し、監視モード中は測定値と製造されたフィルムの外観検査データを受信し、欠陥発生時に対策方法について、モデル予測値と監視モード中の測定値の乖離に基づき対策方法の優先度を算出する対策方法選定支援方法を提供する。 Further, in order to achieve the above object, in the present invention, the computer server is a countermeasure method selection support method for supporting the selection of a countermeasure method when a defect occurs in a film manufactured by a manufacturing apparatus on a production line, which is a computer. The server has a storage device that stores a plurality of countermeasures when a defect occurs and a processor that executes a countermeasure support program including a learning mode and a monitoring mode. The processor executes the countermeasure support program to manufacture the server. It receives measurements from multiple sensors installed in the device, outputs model predictions based on the measurements during learning mode, receives measurements and visual inspection data of the manufactured film during monitoring mode, and is defective. As for the countermeasure method when it occurs, we provide a countermeasure method selection support method that calculates the priority of the countermeasure method based on the difference between the model predicted value and the measured value in the monitoring mode.

本発明により、作業者の技術に依存することなく、欠陥発生から対策方法選定にかかる時間を短縮することが可能となり、フィルム製造プロセスにおける生産性の向上を図ることができる。 According to the present invention, it is possible to shorten the time required for selecting a countermeasure method from the occurrence of defects without depending on the skill of the operator, and it is possible to improve the productivity in the film manufacturing process.

実施例１に係る、フィルム製造の製造プロセスを説明するためのフロー図。The flow chart for demonstrating the manufacturing process of film manufacturing which concerns on Example 1. FIG. 実施例１に係る、対策方法選定支援システムを説明するための構成図。The block diagram for explaining the countermeasure method selection support system which concerns on Example 1. FIG. 実施例１に係る、対策方法選定支援システムを構成する計算機サーバを説明するための構成図。FIG. 5 is a configuration diagram for explaining a computer server constituting a countermeasure method selection support system according to the first embodiment. 実施例１に係る、対策方法選定支援システムを構成するクライアント計算機を説明するための構成図。FIG. 5 is a configuration diagram for explaining a client computer constituting a countermeasure method selection support system according to the first embodiment. 実施例１に係る、対策方法支援プログラムにおける学習モードと監視モードを説明するためのフロー図。The flow diagram for demonstrating the learning mode and the monitoring mode in the countermeasure method support program which concerns on Example 1. FIG. 実施例１に係る、クライアント計算機の表示装置の表示画面を説明するための構成図。FIG. 6 is a configuration diagram for explaining a display screen of a display device of a client computer according to a first embodiment. 実施例１に係る、対策方法選定支援システムで出力される対策方法優先度グラフを説明するための対策方法優先度グラフの図。The figure of the measure method priority graph for explaining the measure method priority graph output by the measure method selection support system which concerns on Example 1. FIG. 実施例１に係る、対策方法実施前後における対策方法優先度グラフの変化を説明するための図。The figure for demonstrating the change of the countermeasure method priority graph before and after the implementation of the countermeasure method which concerns on Example 1. FIG. 実施例１に係る、対策方法に紐づいた要因、要因に関連するターゲット変数、ターゲット変数のモデルの対応を表す図。The figure which shows the factor associated with the countermeasure method, the target variable related to the factor, and the correspondence of the model of the target variable which concerns on Example 1. FIG. 実施例１に係る、対策方法の優先度の時系列データを表す図。The figure which shows the time-series data of the priority of the countermeasure method which concerns on Example 1. FIG.

本発明を実施するための形態を図面に従い説明するのに先立ち、本発明の対策方法選定支援システムを概説する。本発明の対策方法選定支援システムは、フィルム製造装置を監視する、少なくとも計算機サーバより構成される監視システム、監視システムで出力した結果を表示する表示装置と表示プロセッサを有するクライアント計算機より構成される。 Prior to explaining the embodiment for carrying out the present invention in accordance with the drawings, the countermeasure method selection support system of the present invention will be outlined. The countermeasure method selection support system of the present invention is composed of a monitoring system consisting of at least a computer server that monitors the film manufacturing apparatus, a display device that displays the result output by the monitoring system, and a client computer having a display processor.

計算機サーバは、フィルム製造ラインの製造装置の設定値と、製造装置に設置されたセンサの測定値と、製造されたフィルムに対して外観を検査する外観検査装置の外観検査データと、欠陥発生時の複数の対策方法と、実際に実施した対策方法とを格納データとして格納した記憶装置と、対策方法支援プログラムと、この対策方法支援プログラムを実行するプロセッサを備える。 The computer server uses the set values of the manufacturing equipment of the film manufacturing line, the measured values of the sensors installed in the manufacturing equipment, the visual inspection data of the visual inspection device that inspects the appearance of the manufactured film, and when a defect occurs. It is provided with a storage device that stores a plurality of countermeasure methods and the actually implemented countermeasure methods as stored data, a countermeasure method support program, and a processor that executes this countermeasure method support program.

対策方法支援プログラムの学習モード中は、記憶装置から、外観検査データをもとに外観に欠陥が発生していない正常時のデータを受信し、その正常データを用いて、対策方法に対応する要因ごとに、欠陥が発生していない時の各要因の予測モデルを生成する。生成した予測モデルは、記憶装置に保存される。予測モデルは、製造装置の設定パラメータおよび設置センサのパラメータの中から、要因ごとに要因を最も直接的に表す変数を目的変数、目的変数と関係性を持つと思われる変数を説明変数として選定し、正常時のデータを用いて、説明変数から目的変数のモデル予測値とその誤差を出力する。ここで、予測モデルは、ガウス過程回帰、ニューラルネットワーク等の機械学習モデルより構築される。 Countermeasure method During the learning mode of the support program, normal data with no appearance defects is received from the storage device based on the visual inspection data, and the normal data is used to determine the factors corresponding to the countermeasure method. For each, a prediction model of each factor when no defect occurs is generated. The generated prediction model is stored in the storage device. In the prediction model, the variable that most directly represents the factor for each factor is selected as the objective variable, and the variable that seems to be related to the objective variable is selected as the explanatory variable from the setting parameters of the manufacturing equipment and the parameters of the installed sensor. , The model predicted value of the objective variable and its error are output from the explanatory variables using the normal data. Here, the prediction model is constructed from machine learning models such as Gaussian process regression and neural networks.

対策方法支援プログラムの監視モード中は、記憶装置から設備の設定値と、測定値と、要因の予測モデルとを受信し、複数の要因について、要因の予測モデルから得られるモデル予測値と、設備の設定値とセンサの測定値から計算される実測値との乖離に基づき、要因の異常度を算出し、その異常度を元に当該対策方法の優先度を算出し、対策方法の内容と、対応する優先度をクライアント計算機に出力する。 Countermeasures During the monitoring mode of the support program, the equipment setting value, measured value, and factor prediction model are received from the storage device, and for multiple factors, the model prediction value obtained from the factor prediction model and the equipment Based on the discrepancy between the set value of and the measured value calculated from the measured value of the sensor, the degree of abnormality of the factor is calculated, and the priority of the countermeasure method is calculated based on the degree of abnormality. Output the corresponding priority to the client computer.

ここで、要因qの異常度Sqは、各要因の目的変数における実測値とモデル予測値の乖離スコアTqの対数尤度で定義され、SqおよびTqは次の式１、式２及び式３で定義される。 Here, the anomaly degree Sq of the factor q is defined by the log-likelihood of the deviation score Tq between the measured value and the model predicted value in the objective variable of each factor, and Sq and Tq are given by the following equations 1, 2 and 3. Defined.

Figure 2021149733
Figure 2021149733

Figure 2021149733
Figure 2021149733

Figure 2021149733
Figure 2021149733

yqは要因qの目的変数の実測値、y’qは目的変数のモデル予測値、σqはモデル予測値の誤差、fq(x)はtqの確率密度関数である。確率密度関数は、正常データからtqの確率分布を求め、この確率分布にカイ二乗分布を当てはめることにより得る。 y q is the measured value of the objective variable of factor q, y'q is the model predicted value of the objective variable, σ q is the error of the model predicted value, and f q (x) is the probability density function of t q. The probability density function is obtained by finding the probability distribution of t q from normal data and applying the chi-square distribution to this probability distribution.

クライアント計算機の表示プロセッサは、計算機サーバから、対策方法と対策方法の優先度を受信し、クライアント計算機が有する表示装置に、対策方法の内容と対応する優先度とを対策方法優先度グラフとして作業者に視覚的に表示する。作業者はこの対策方法優先度グラフを参考にして欠陥の要因を解消する適切な対策方法を実施することができる。 The display processor of the client computer receives the countermeasure method and the priority of the countermeasure method from the computer server, and displays the content of the countermeasure method and the corresponding priority on the display device of the client computer as a countermeasure method priority graph. Visually display on. The worker can implement an appropriate countermeasure method to eliminate the cause of the defect by referring to this countermeasure method priority graph.

続いて、本発明を実施するための形態を、図面を用いて詳述する。なお、以下に説明する実施形態は特許請求の範囲にかかる発明を限定するものではなく、また実施形態の中で説明されている諸要素及びその組み合わせの全てが発明の解決手段に必須であるとは限らない。 Subsequently, a mode for carrying out the present invention will be described in detail with reference to the drawings. It should be noted that the embodiments described below do not limit the inventions claimed in the claims, and all of the elements and combinations thereof described in the embodiments are indispensable for the means for solving the invention. Is not always.

実施例１は、要因に対応する対策方法に優先度をつけて視覚的に作業者に提示する対策方法選定支援システム、及び方法の実施例である。すなわち、製造ラインの製造装置で製造されるフィルムの欠陥発生時の対策方法選定を支援する対策方法選定支援システムであって、欠陥発生時の複数の対策方法を格納する記憶装置と、学習モードと監視モードを含む対策方法支援プログラムを実行するプロセッサを有する計算機サーバを備え、プロセッサが対策方法支援プログラムを実行することで、製造装置に設置された複数のセンサから測定値を受信し、学習モード中は測定値に基づきモデル予測値を出力し、監視モード中は測定値と製造されたフィルムの外観検査データを受信し、欠陥発生時に対策方法について、モデル予測値と監視モード中の測定値の乖離に基づき対策方法の優先度を算出する対策方法選定支援システム、及びその方法の実施例である。 The first embodiment is an example of a countermeasure method selection support system and a method in which the countermeasure methods corresponding to the factors are prioritized and visually presented to the operator. That is, it is a countermeasure method selection support system that supports the selection of countermeasure methods when defects occur in the film manufactured by the production equipment of the production line, and is a storage device that stores a plurality of countermeasure methods when defects occur, and a learning mode. It is equipped with a computer server that has a processor that executes a countermeasure method support program including a monitoring mode, and when the processor executes the countermeasure method support program, it receives measured values from multiple sensors installed in the manufacturing equipment and is in learning mode. Outputs the model predicted value based on the measured value, receives the measured value and the appearance inspection data of the manufactured film during the monitoring mode, and regarding the countermeasure method when a defect occurs, the difference between the model predicted value and the measured value in the monitoring mode. This is a countermeasure method selection support system that calculates the priority of countermeasure methods based on the above, and an example of the method.

図１は実施例１の対策方法選定支援システムが対象とするフィルム製造の製造プロセスの一例を示す。同図に示すフィルム製造工程は、長尺状可撓性の基材を搬送して送り出し(s01)、溶剤を含む塗布物を塗工機により所望の膜厚で塗布して塗布膜を形成し(s02)、乾燥機で溶剤を乾燥して除去させる工程(s03)を経て、最後に外観検査機により外観検査(s04)を行う製造プロセスである。ここで、各工程には各工程設備に、例えば基材の張力、塗布液の液物性、温度、塗布物の厚み等の複数のセンサデバイスからなるセンサが設置されており、各種センサ測定値と設備の設定値と外観欠陥検査装置で得られる外観欠陥情報とが、データ収集システムによりフィルム製造設備データサーバに蓄積されている。 FIG. 1 shows an example of a film manufacturing manufacturing process targeted by the countermeasure method selection support system of the first embodiment. In the film manufacturing process shown in the figure, a long flexible base material is conveyed and sent out (s01), and a coating material containing a solvent is applied to a desired film thickness by a coating machine to form a coating film. (S02), a manufacturing process in which the solvent is dried and removed by a dryer (s03), and finally the appearance inspection (s04) is performed by an appearance inspection machine. Here, in each process, a sensor composed of a plurality of sensor devices such as the tension of the base material, the physical properties of the coating liquid, the temperature, the thickness of the coating material, etc. is installed in each process equipment, and various sensor measurement values are obtained. The set value of the equipment and the appearance defect information obtained by the appearance defect inspection device are stored in the film manufacturing equipment data server by the data collection system.

図２は本実施例に係る対策方法選定支援システムの全体構成図である。対策方法選定支援システムは、フィルム製造設備の各種のセンサ測定値と、設備の設定値と、外観欠陥検査装置からの外観欠陥情報とを収集するフィルム製造設備データサーバ201と、計算機サーバ202と、クライアント計算機203から構成される。これらフィルム製造設備データサーバ201と計算機サーバ202とクライアント計算機203はネットワーク204を介して接続されている。計算機サーバ202とクライアント計算機203の詳細は後述する。ネットワーク204は、有線ネットワークでも無線ネットワークでもよい。本実施例では、場所Aにフィルム製造設備データサーバ201が、場所Bに計算機サーバ202が、場所Cにクライアント計算機203が設置されている。フィルム製造設備データサーバ201、計算機サーバ202、クライアント計算機203は同一の場所に配置されていてもよい。 FIG. 2 is an overall configuration diagram of a countermeasure method selection support system according to this embodiment. The countermeasure selection support system includes a film manufacturing equipment data server 201 and a computer server 202 that collect various sensor measurement values of the film manufacturing equipment, equipment setting values, and appearance defect information from the appearance defect inspection device. It consists of a client computer 203. The film manufacturing equipment data server 201, the computer server 202, and the client computer 203 are connected to each other via the network 204. Details of the computer server 202 and the client computer 203 will be described later. The network 204 may be a wired network or a wireless network. In this embodiment, the film manufacturing equipment data server 201 is installed at the location A, the computer server 202 is installed at the location B, and the client computer 203 is installed at the location C. The film manufacturing equipment data server 201, the computer server 202, and the client computer 203 may be arranged in the same place.

図３は、計算機サーバ202の一構成例を示す図であり、計算機サーバ202は通常のコンピュータ構成を備え、記憶装置301と、記憶装置301などに記憶された対策方法支援プログラム302を実行する実行プロセッサ303を有する。記憶装置301は、フィルム製造設備データサーバ201から受信する複数のセンサデバイスが測定した複数のセンサ測定値、設備設定値、外観欠陥情報と、更には外観検査データ、欠陥発生時の対策方法、及びフィルム製造中に実施した対策方法等を格納する。対策方法支援プログラムの詳細は後述する。 FIG. 3 is a diagram showing a configuration example of the computer server 202. The computer server 202 has a normal computer configuration, and executes the storage device 301 and the countermeasure method support program 302 stored in the storage device 301 or the like. It has a processor 303. The storage device 301 includes a plurality of sensor measurement values, equipment setting values, appearance defect information measured by a plurality of sensor devices received from the film manufacturing equipment data server 201, appearance inspection data, countermeasures when defects occur, and countermeasures. Stores countermeasures taken during film production. The details of the countermeasure support program will be described later.

図４は、クライアント計算機203の一構成例を示す図であり、クライアント計算機203は表示装置401と計算機サーバ202と通信可能な表示プロセッサ402を備えている。表示プロセッサ402は、計算機サーバ202の出力を処理する。表示装置401の表示内容は、対策方法支援プログラム302の実行により計算機サーバ202から出力される内容であり、詳細は後述する。 FIG. 4 is a diagram showing a configuration example of the client computer 203, and the client computer 203 includes a display device 401 and a display processor 402 capable of communicating with the computer server 202. The display processor 402 processes the output of the computer server 202. The display content of the display device 401 is the content output from the computer server 202 by executing the countermeasure method support program 302, and the details will be described later.

図５は、実行プロセッサ303が実行する対策方法支援プログラム302のシステムフローの一例を示す図である。対策方法支援プログラム302は学習モードと監視モードの少なくとも２つの機能を持つ。学習モードはフィルム製造前に機能するモードであり、監視モードはフィルム製造時に機能するモードである。
以下、学習モードと監視モードの詳細を説明する。 FIG. 5 is a diagram showing an example of the system flow of the countermeasure method support program 302 executed by the execution processor 303. Countermeasure method The support program 302 has at least two functions, a learning mode and a monitoring mode. The learning mode is a mode that functions before film production, and the monitoring mode is a mode that functions during film production.
The details of the learning mode and the monitoring mode will be described below.

学習モードは、記憶装置301に蓄積された過去の製造装置の設定値と設置センサの取得値、外観検査データ、欠陥情報などから、欠陥が発生していない場合のデータを抽出し、フィルムに欠陥が発生していない時の正常時のデータを抽出し、その抽出した正常時のデータを用いて欠陥要因ごとに、要因を定量的に表すターゲット変数の予測モデルを、ターゲット変数と関連性があるセンサパラメータと設備パラメータから作成し、記憶装置301に記憶する。予測モデルは、製造装置の設定パラメータおよびセンサパラメータの中から、要因ごとに要因を最も直接的に表す変数を目的変数、目的変数と関係性を持つ変数を説明変数として選定し、正常時のデータを用いて、説明変数から目的変数のモデル予測値とその誤差を算出して出力する。ここで、予測モデルは、ガウス過程回帰、ニューラルネットワーク等の機械学習モデルより構築される。 In the learning mode, data when no defect has occurred is extracted from the past set value of the manufacturing device and the acquired value of the installation sensor, visual inspection data, defect information, etc. accumulated in the storage device 301, and the film is defective. The normal data when is not occurring is extracted, and the extracted normal data is used to predict the target variable that quantitatively expresses the factor for each defect factor, which is related to the target variable. Created from sensor parameters and equipment parameters, and stored in the storage device 301. In the prediction model, the variable that most directly represents the factor for each factor is selected as the objective variable and the variable that is related to the objective variable is selected as the explanatory variable from the setting parameters and sensor parameters of the manufacturing equipment. Is used to calculate and output the model predicted value of the objective variable and its error from the explanatory variables. Here, the prediction model is constructed from machine learning models such as Gaussian process regression and neural networks.

監視モードでは、記憶装置302から指定時刻の欠陥発生時のセンサ・設備データをリアルタイムに抽出し、ターゲット変数ごとにモデル予測値と実測値との乖離を異常度として算出する。監視モード中は、記憶装置302から設備の設定値と、測定値と、要因の予測モデルとを受信し、複数の要因について、要因の予測モデルから得られるモデル予測値と設備の設定値とセンサデバイスの測定値から計算される実測値との乖離に基づき、要因の異常度を算出し、その異常度に基づき対策方法の優先度に変換し、対策方法と対応する優先度とをクライアント計算機203に送信し、作業者に提示する。 In the monitoring mode, sensor / equipment data when a defect occurs at a specified time is extracted from the storage device 302 in real time, and the deviation between the model predicted value and the measured value is calculated as the degree of abnormality for each target variable. During the monitoring mode, the equipment setting value, the measured value, and the factor prediction model are received from the storage device 302, and for a plurality of factors, the model prediction value obtained from the factor prediction model, the equipment setting value, and the sensor are received. Based on the deviation from the measured value calculated from the measured value of the device, the degree of abnormality of the factor is calculated, converted to the priority of the countermeasure method based on the degree of abnormality, and the countermeasure method and the corresponding priority are converted into the client computer 203. Send to and present to the worker.

すなわち、実行プロセッサ303は、予測モデルに基づいて第ｎ（ｎは１以上の自然数）の測定項目の測定値を予測し、センサデバイスより第ｎの測定項目の実際の測定値を受信し、予測した第ｎの測定項目の測定値と、センサデバイスより受信した第ｎの測定項目の測定値とに基づいて乖離を算出する。 That is, the execution processor 303 predicts the measurement value of the nth measurement item (n is a natural number of 1 or more) based on the prediction model, receives the actual measurement value of the nth measurement item from the sensor device, and predicts. The deviation is calculated based on the measured value of the nth measurement item and the measured value of the nth measurement item received from the sensor device.

図９に各対策方法に紐づいた要因等の対応を表すテーブル901の一例を示す。図９のテーブル901は、計算機サーバ202の実行プロセッサ303が、フィルム製造前に対策方法支援プログラム302の学習モードを実行することで作成され、記憶装置301に記憶される。同図は、対策方法Ａ、Ｂ・・・に紐づいた要因、要因に関連する各種センサデバイスからなるセンサ、ターゲット変数、ターゲット変数の予測モデルの対応表の一例を示している。学習モードでは、このように対策方法に紐づいた要因に対して、製造ラインに設置された各種センサの中から、要因に関連するセンサを選び出し、各種センサからターゲット変数とターゲット変数の説明変数を設計し、ターゲット変数の説明変数を用いて、ターゲット変数の予測モデルを機械学習モデルにより作成する。 FIG. 9 shows an example of a table 901 showing the correspondence of factors and the like associated with each countermeasure method. Table 901 in FIG. 9 is created by the execution processor 303 of the computer server 202 executing the learning mode of the countermeasure method support program 302 before film production, and is stored in the storage device 301. The figure shows an example of a correspondence table of factors associated with countermeasure methods A, B ..., a sensor composed of various sensor devices related to the factors, a target variable, and a prediction model of the target variable. In the learning mode, for the factors associated with the countermeasure method, the sensors related to the factors are selected from the various sensors installed on the production line, and the target variables and the explanatory variables of the target variables are selected from the various sensors. Design and create a prediction model of the target variable by the machine learning model using the explanatory variables of the target variable.

予測モデルにより作成されるターゲット変数のモデル予測値は、欠陥が発生していない場合のデータを用いて作成されるため、対策方法支援プログラム302の監視モードにおいて、モデル予測値と、欠陥が発生する場合のデータを用いて算出したターゲット変数の実測値とを比較して異常度を算出すると、予測値と実測値の乖離が大きくなり異常度が高く算出される。計算機サーバ202はこの異常度から、各ターゲット変数の要因に対応する対策方法の優先度を計算し、その結果を対策方法とそれらの優先度としてクライアント計算機203に送信する。 Since the model predicted value of the target variable created by the prediction model is created using the data when no defect has occurred, the model predicted value and the defect occur in the monitoring mode of the countermeasure method support program 302. When the degree of anomaly is calculated by comparing it with the actually measured value of the target variable calculated using the case data, the difference between the predicted value and the actually measured value becomes large and the degree of anomaly is calculated to be high. The computer server 202 calculates the priority of the countermeasure method corresponding to the factor of each target variable from this abnormality degree, and transmits the result to the client computer 203 as the countermeasure method and their priority.

図１０は対策方法支援プログラム302がクライアント計算機203に送信する、各対策方法の優先度の時系列データを示すテーブル1001である。テーブル1001もテーブル901同様、計算機サーバ202の記憶装置301に記憶され、データベースになる。上述の通り対策方法の優先度は、各要因の異常度に基づき算出される。なお、図１０においては、1分ごとの周期で優先度が算出されている例が示されているが、１分に限らず、この算出周期は可変であり、任意に設定可能である。 FIG. 10 is a table 1001 showing time-series data of the priority of each countermeasure method transmitted by the countermeasure method support program 302 to the client computer 203. Similar to table 901, table 1001 is also stored in the storage device 301 of the computer server 202 and becomes a database. As described above, the priority of the countermeasure method is calculated based on the degree of abnormality of each factor. Note that FIG. 10 shows an example in which the priority is calculated in a cycle of every minute, but the calculation cycle is not limited to one minute and is variable and can be set arbitrarily.

対策方法ａ、ｂ、ｃ・・・の優先度はターゲット変数ごとに計算される。ある着目ターゲット変数の優先度は、例えば、全ターゲット変数の異常度の合計値に対する着目ターゲット変数の異常度で計算される。クライアント計算機203では、計算機サーバ202が出力した対策方法と対策方法の優先度のグラフを表示装置401に表示する。フィルム製造現場の作業者はこの対策方法と優先度を確認することで、容易に実施すべき対策方法を選定することが可能である。 The priority of the countermeasure methods a, b, c ... Is calculated for each target variable. The priority of a certain target variable of interest is calculated by, for example, the degree of abnormality of the target variable of interest with respect to the total value of the degree of abnormality of all target variables. The client computer 203 displays a graph of the countermeasure method output by the computer server 202 and the priority of the countermeasure method on the display device 401. By confirming this countermeasure method and priority, the worker at the film manufacturing site can easily select the countermeasure method to be implemented.

図６は、本実施例の対策方法選定支援システムにおいて、クライアント計算機203の表示装置401で表示する画面の一構成例を示す。表示画面上には少なくとも、表示ボタン601、時刻選定ボックス602、対策方法優先度グラフ603、実施対策記入ボックス604、記入対策方法保存ボタン605が存在する。作業者は時刻選定ボックス602に所望の時刻を入力し、表示ボタン601を押すことにより、該当時刻の対策方法優先度グラフ603を確認することがきる。また、作業者が実施した対策方法は、実施対策記入ボックス604に自由記述することができ、作業者が記入対策方法保存ボタン605を押すことにより、記述された対策方法が計算機サーバ202の記憶装置301に保存される。 FIG. 6 shows an example of a configuration of a screen displayed by the display device 401 of the client computer 203 in the countermeasure method selection support system of this embodiment. On the display screen, there are at least a display button 601, a time selection box 602, a countermeasure method priority graph 603, an implementation countermeasure entry box 604, and an entry countermeasure method save button 605. The operator can enter the desired time in the time selection box 602 and press the display button 601 to check the countermeasure method priority graph 603 of the corresponding time. In addition, the countermeasure method implemented by the worker can be freely described in the implementation countermeasure entry box 604, and when the worker presses the entry countermeasure method save button 605, the described countermeasure method is stored in the storage device of the computer server 202. Stored in 301.

図７は、クライアント計算機203の表示装置401で表示される対策方法優先度グラフの詳細を示す。対策方法優先度グラフ701では、対策方法支援プログラム302で算出された優先度が対策方法ごとに横軸に伸びる棒グラフで表示されている。ここで、対策方法優先度グラフ701の縦軸における対策方法の順序は、優先度を元に昇順ないし降順にソートされていても、ソートされていなくてもよい。 FIG. 7 shows the details of the countermeasure method priority graph displayed on the display device 401 of the client computer 203. In the countermeasure method priority graph 701, the priority calculated by the countermeasure method support program 302 is displayed as a bar graph extending on the horizontal axis for each countermeasure method. Here, the order of the countermeasure methods on the vertical axis of the countermeasure method priority graph 701 may or may not be sorted in ascending or descending order based on the priority.

図８は、フィルム製造中に欠陥が発生し作業者が対策を実施しようとした際に、本実施例の対策方法選定支援システムを実施した時の対策方法優先度グラフの変化を説明するための図である。対策方法優先度グラフ801は、欠陥が発生し対策方法を実施しようとした際の優先度グラフであり、対策方法優先度グラフ801では対策方法Aの優先度が最大である。作業者は表示装置401の表示画面に表示されている対策方法優先度グラフ801を確認し、対策方法Aを実施する。対策方法優先度グラフ802は、対策方法Aを実施した後の対策方法の優先度グラフであり、対策方法Aを実施する前の対策方法優先度グラフ801に比較して、対策方法Aの優先度が低くなっている。このことから、フィルム製造中に発生した欠陥の要因は、対策方法Aに対応する要因であり、本実施例の対策方法選定支援システムで選定された対策方法Aは欠陥の要因を解消する上で妥当な対策方法であったことが確認できる。 FIG. 8 is for explaining the change in the countermeasure method priority graph when the countermeasure method selection support system of this embodiment is implemented when a defect occurs during film production and the worker tries to implement the countermeasure. It is a figure. The countermeasure method priority graph 801 is a priority graph when a defect occurs and an attempt is made to implement the countermeasure method. In the countermeasure method priority graph 801, the priority of the countermeasure method A is the highest. The operator confirms the countermeasure method priority graph 801 displayed on the display screen of the display device 401, and implements the countermeasure method A. The countermeasure method priority graph 802 is a priority graph of the countermeasure method after the countermeasure method A is implemented, and is a priority graph of the countermeasure method A as compared with the countermeasure method priority graph 801 before the countermeasure method A is implemented. Is low. From this, the cause of the defect generated during film production is the factor corresponding to the countermeasure method A, and the countermeasure method A selected by the countermeasure method selection support system of this embodiment is used to eliminate the cause of the defect. It can be confirmed that it was an appropriate countermeasure.

以上説明したように、本実施例の対策方法選定支援システムによれば、フィルム製造プロセス中において欠陥が発生し、作業者が欠陥の対策方法を選定する際に、複数の欠陥要因に対して要因の状態を定量的に推定し、要因に対応する対策方法に優先度をつけて視覚的に作業者に提示して支援することができ、フィルム製造プロセスにおける生産性および歩留まりを向上させることができる。 As described above, according to the countermeasure method selection support system of this embodiment, defects occur during the film manufacturing process, and when the operator selects the countermeasure method for the defects, the factors are caused by a plurality of defect factors. It is possible to quantitatively estimate the state of the film, prioritize the countermeasures corresponding to the factors, and visually present them to the operator to support them, and improve the productivity and yield in the film manufacturing process. ..

なお、本発明は上記した実施例に限定されるものではなく、様々な変形例が含まれる。例えば、上記した実施例は本発明のより良い理解のために詳細に説明したのであり、必ずしも説明の全ての構成を備えるものに限定されるものではない。 The present invention is not limited to the above-described examples, and includes various modifications. For example, the above-mentioned examples have been described in detail for a better understanding of the present invention, and are not necessarily limited to those having all the configurations of the description.

上述した各構成、機能、計算機サーバ等は、それらの一部又は全部を実現するプログラムを作成する例を中心に説明したが、それらの一部又は全部を例えば集積回路で設計する等によりハードウェアで実現しても良いことは言うまでもない。すなわち、処理部の全部または一部の機能は、プログラムに代え、例えば、ＡＳＩＣ(Application Specific Integrated Circuit)、ＦＰＧＡ(Field Programmable Gate Array)などの集積回路などにより実現してもよい。 Each of the above-mentioned configurations, functions, computer servers, etc. has been described mainly with an example of creating a program that realizes a part or all of them, but hardware such as designing a part or all of them with an integrated circuit, for example. Needless to say, it may be realized with. That is, all or part of the functions of the processing unit may be realized by, for example, an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array) instead of the program.

201 フィルム製造設備
202 計算機サーバ
203 クライアント計算機
204 ネットワーク
301 記憶装置
302 対策方法支援プログラム
401 表示装置
402 示プロセッサ
601 表示ボタン
602 時刻選定ボックス
603 対策方法優先度グラフ
604 実施対策記入ボックス
701 対策方法優先度グラフ
801 対策方法実施前の対策方法優先度グラフ
802 対策方法実施後の対策方法優先度グラフ
901 対策方法に紐づいた要因等の対応を表すテーブル
1001 対策方法の優先度の時系列データを表すテーブル 201 Film manufacturing equipment
202 Computer server
203 client calculator
204 network
301 storage device
302 Countermeasure support program
401 Display device
402 Display processor
601 Display button
602 Time selection box
603 Countermeasure method Priority graph
604 Implementation measures entry box
701 Countermeasure method Priority graph
801 Countermeasure method Priority graph of countermeasure method before implementation
802 Countermeasure method Priority graph of countermeasure method after implementation
901 A table showing the correspondence of factors linked to the countermeasure method
1001 A table that represents the time-series data of the priority of countermeasures

Claims (13)

製造ラインの製造装置で製造されるフィルムの欠陥発生時の対策方法選定を支援する対策方法選定支援システムであって、
欠陥発生時の複数の対策方法を格納する記憶装置と、学習モードと監視モードを含む対策方法支援プログラムを実行するプロセッサを有する計算機サーバを備え、
前記プロセッサが前記対策方法支援プログラムを実行することで、前記製造装置に設置された複数のセンサから測定値を受信し、学習モード中は前記測定値に基づきモデル予測値を出力し、監視モード中は、前記測定値と製造された前記フィルムの外観検査データを受信し、欠陥発生時に前記対策方法について、前記モデル予測値と前記監視モード中の前記測定値の乖離に基づき前記対策方法の優先度を算出する、
ことを特徴とする対策方法選定支援システム。 It is a countermeasure method selection support system that supports the selection of countermeasure methods when defects occur in the film manufactured by the manufacturing equipment on the production line.
It is equipped with a storage device that stores multiple countermeasures when a defect occurs, and a computer server that has a processor that executes a countermeasure support program including a learning mode and a monitoring mode.
When the processor executes the countermeasure method support program, it receives measured values from a plurality of sensors installed in the manufacturing apparatus, outputs model predicted values based on the measured values during the learning mode, and is in the monitoring mode. Receives the measured value and the appearance inspection data of the manufactured film, and when a defect occurs, the priority of the countermeasure method is based on the discrepancy between the model predicted value and the measured value in the monitoring mode. To calculate,
A countermeasure method selection support system characterized by this. 請求項１記載の対策方法選定支援システムであって、
前記計算機サーバは、複数の前記対策方法と対応する前記優先度を出力する、
ことを特徴とする対策方法選定支援システム。 The countermeasure method selection support system according to claim 1.
The computer server outputs the priority corresponding to the plurality of countermeasure methods.
A countermeasure method selection support system characterized by this. 請求項２記載の対策方法選定支援システムであって、
前記計算機サーバと通信可能であって、複数の前記対策方法と対応する前記優先度を表示する表示プロセッサを有するクライアント計算機を備える、
ことを特徴とする対策方法選定支援システム。 The countermeasure method selection support system according to claim 2.
A client computer having a display processor capable of communicating with the computer server and displaying the priority corresponding to the plurality of countermeasures is provided.
A countermeasure method selection support system characterized by this. 請求項３記載の対策方法選定支援システムであって、
前記表示プロセッサは、複数の前記対策方法と対応する前記優先度のグラフと、時刻選択ボックスと、表示ボタンと、作業者が実施した対策方法を入力する入力ボックスを表示装置に表示可能である、
ことを特徴とする対策方法選定支援システム。 The countermeasure method selection support system according to claim 3.
The display processor can display on the display device a graph of the priority corresponding to the plurality of the countermeasure methods, a time selection box, a display button, and an input box for inputting the countermeasure method implemented by the operator.
A countermeasure method selection support system characterized by this. 請求項１記載の対策方法選定支援システムであって、
前記センサは、複数のセンサデバイスを含み、複数種類の測定項目について測定値を測定し、
前記プロセッサは、前記モデル予測値に基づいて第ｎ（ｎは１以上の自然数）の測定項目の測定値を予測し、前記センサデバイスより前記第ｎの測定項目の測定値を受信し、予測した前記第ｎの測定項目の測定値と、前記センサデバイスより受信した前記第ｎの測定項目の測定値とに基づいて乖離を算出する、
ことを特徴とする対策方法選定支援システム。 The countermeasure method selection support system according to claim 1.
The sensor includes a plurality of sensor devices and measures measured values for a plurality of types of measurement items.
The processor predicts the measurement value of the nth measurement item (n is a natural number of 1 or more) based on the model prediction value, receives the measurement value of the nth measurement item from the sensor device, and predicts the measurement value. The deviation is calculated based on the measured value of the nth measurement item and the measured value of the nth measurement item received from the sensor device.
A countermeasure method selection support system characterized by this. 請求項５記載の対策方法選定支援システムであって、
前記プロセッサは、前記乖離に基づき、前記第ｎの測定項目に対応する対策方法の優先度を算出する、
ことを特徴とする対策方法選定支援システム。 The countermeasure method selection support system according to claim 5.
The processor calculates the priority of the countermeasure method corresponding to the nth measurement item based on the deviation.
A countermeasure method selection support system characterized by this. 請求項1記載の対策方法選定支援システムであって、
前記プロセッサは、前記学習モード中に得られた前記モデル予測値に対して、欠陥が発生していないデータを抽出し、抽出した前記データをもとに、複数の前記対策方法の各々について、対応する要因を直接的に表す変数を目的変数とし、前記目的変数と関係性を持つ変数を説明変数とし、前記目的変数の予測モデルを、前記説明変数から機械学習モデルにより作成し、作成した前記予測モデルを前記記憶装置に格納する、
ことを特徴とした対策方法選定支援システム。 The countermeasure method selection support system described in claim 1
The processor extracts data that does not cause defects with respect to the model predicted value obtained during the learning mode, and based on the extracted data, responds to each of the plurality of countermeasure methods. The variable that directly represents the factor to be used is used as the objective variable, the variable that is related to the objective variable is used as the explanatory variable, and the prediction model of the objective variable is created from the explanatory variable by the machine learning model, and the created prediction is performed. Store the model in the storage device,
A countermeasure method selection support system characterized by this. 計算機サーバにより、製造ラインの製造装置で製造されるフィルムの欠陥発生時の対策方法選定を支援する対策方法選定支援方法であって、
前記計算機サーバは、欠陥発生時の複数の対策方法を格納する記憶装置と、学習モードと監視モードを含む対策方法支援プログラムを実行するプロセッサを有し、
前記プロセッサが前記対策方法支援プログラムを実行することで、前記製造装置に設置された複数のセンサから測定値を受信し、学習モード中は前記測定値に基づきモデル予測値を出力し、監視モード中は前記測定値と製造された前記フィルムの外観検査データを受信し、欠陥発生時に前記対策方法について、前記モデル予測値と前記監視モード中の前記測定値の乖離に基づき前記対策方法の優先度を算出する、
ことを特徴とする対策方法選定支援方法。 It is a countermeasure method selection support method that supports the selection of countermeasure methods when defects occur in the film manufactured by the manufacturing equipment of the production line by the computer server.
The computer server has a storage device that stores a plurality of countermeasure methods when a defect occurs, and a processor that executes a countermeasure method support program including a learning mode and a monitoring mode.
When the processor executes the countermeasure method support program, it receives measured values from a plurality of sensors installed in the manufacturing apparatus, outputs model predicted values based on the measured values during the learning mode, and is in the monitoring mode. Receives the measured value and the appearance inspection data of the manufactured film, and when a defect occurs, determines the priority of the countermeasure method based on the discrepancy between the model predicted value and the measured value in the monitoring mode. calculate,
Countermeasure method selection support method characterized by this. 請求項８記載の対策方法選定支援方法であって、
複数の前記対策方法と対応する前記優先度を表示装置に表示する、
ことを特徴とする対策方法選定支援方法。 The countermeasure method selection support method according to claim 8.
Displaying the priority corresponding to the plurality of countermeasures on the display device,
Countermeasure method selection support method characterized by this. 請求項９記載の対策方法選定支援方法であって、
複数の前記対策方法と対応する前記優先度のグラフと、時刻選択ボックスと、表示ボタンと、作業者が実施した対策方法を入力する入力ボックスとを前記表示装置に表示する、
ことを特徴とする対策方法選定支援方法。 The countermeasure method selection support method according to claim 9.
A graph of the priority corresponding to the plurality of the countermeasure methods, a time selection box, a display button, and an input box for inputting the countermeasure method implemented by the operator are displayed on the display device.
Countermeasure method selection support method characterized by this. 請求項８記載の対策方法選定支援方法であって、
前記センサは、複数のセンサデバイスを含み、複数種類の測定項目について測定値を測定し、
前記プロセッサは、前記モデル予測値に基づいて第ｎ（ｎは１以上の自然数）の測定項目の測定値を予測し、前記センサデバイスより前記第ｎの測定項目の測定値を受信し、予測した前記第ｎの測定項目の測定値と、前記センサデバイスより受信した前記第ｎの測定項目の測定値とに基づいて乖離を算出する、
ことを特徴とする対策方法選定支援方法。 The countermeasure method selection support method according to claim 8.
The sensor includes a plurality of sensor devices and measures measured values for a plurality of types of measurement items.
The processor predicts the measurement value of the nth measurement item (n is a natural number of 1 or more) based on the model prediction value, receives the measurement value of the nth measurement item from the sensor device, and predicts the measurement value. The deviation is calculated based on the measured value of the nth measurement item and the measured value of the nth measurement item received from the sensor device.
Countermeasure method selection support method characterized by this. 請求項１１記載の対策方法選定支援方法であって、
前記プロセッサは、前記乖離に基づき、前記第ｎの測定項目に対応する対策方法の優先度を算出する、
ことを特徴とする対策方法選定支援方法。 The countermeasure method selection support method according to claim 11.
The processor calculates the priority of the countermeasure method corresponding to the nth measurement item based on the deviation.
Countermeasure method selection support method characterized by this. 請求項８記載の対策方法選定支援方法であって、
前記プロセッサは、前記学習モード中に得られた前記モデル予測値に対して、欠陥が発生していないデータを抽出し、抽出した前記データをもとに、複数の前記対策方法の各々について、対応する要因を直接的に表す変数を目的変数とし、前記目的変数と関係性を持つ変数を説明変数とし、前記目的変数の予測モデルを、前記説明変数から機械学習モデルにより作成する、
ことを特徴とする対策方法選定支援方法。 The countermeasure method selection support method according to claim 8.
The processor extracts data that does not cause defects with respect to the model predicted value obtained during the learning mode, and based on the extracted data, responds to each of the plurality of countermeasure methods. A variable that directly represents the factor to be used is used as an objective variable, a variable having a relationship with the objective variable is used as an explanatory variable, and a prediction model of the objective variable is created from the explanatory variables by a machine learning model.
Countermeasure method selection support method characterized by this.
JP2020050591A 2020-03-23 2020-03-23 Countermeasure selection support system and method Active JP7484281B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2020050591A JP7484281B2 (en) 2020-03-23 2020-03-23 Countermeasure selection support system and method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2020050591A JP7484281B2 (en) 2020-03-23 2020-03-23 Countermeasure selection support system and method

Publications (2)

Publication Number Publication Date
JP2021149733A true JP2021149733A (en) 2021-09-27
JP7484281B2 JP7484281B2 (en) 2024-05-16

Family

ID=77849119

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2020050591A Active JP7484281B2 (en) 2020-03-23 2020-03-23 Countermeasure selection support system and method

Country Status (1)

Country Link
JP (1) JP7484281B2 (en)

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2012137934A (en) 2010-12-27 2012-07-19 Hitachi Ltd Abnormality detection/diagnostic method, abnormality detection/diagnostic system, abnormality detection/diagnostic program and company asset management/facility asset management system
JP7017861B2 (en) 2017-03-23 2022-02-09 株式会社日立製作所 Anomaly detection system and anomaly detection method

Also Published As

Publication number Publication date
JP7484281B2 (en) 2024-05-16

Similar Documents

Publication Publication Date Title
US10776706B2 (en) Cost-driven system and method for predictive equipment failure detection
AU2016201724A1 (en) Method and system for predicting equipment failure
JP2009080612A (en) Method for evaluating distribution, method for producing article, distribution evaluation program, and distribution evaluation system
US20190129397A1 (en) Causal relation model building system and method thereof
JP6711323B2 (en) Abnormal state diagnosis method and abnormal state diagnosis device
WO2010110365A1 (en) Factor analysis apparatus and factor analysis method
JP6778277B2 (en) Quality control equipment and quality control method
US20170176985A1 (en) Method for predicting end of line quality of assembled product
JP2019045905A (en) Behavior predicting system and behavior predicting method
JP6702297B2 (en) Abnormal state diagnosis method and abnormal state diagnosis device
TWI710873B (en) Support device, learning device, and plant operating condition setting support system
Chien et al. Bayesian decision analysis for optimizing in-line metrology and defect inspection strategy for sustainable semiconductor manufacturing and an empirical study
JP2021149733A (en) Countermeasure selection support system, and method therefor
US11676055B2 (en) System for detecting data drift in machine-learning process monitoring
JP6600120B1 (en) Management system, machine learning apparatus and management method therefor
JPWO2019016892A1 (en) Quality analyzer and quality analysis method
CN111768080A (en) Method and system for evaluating finished progress of manufactured product
JP6885321B2 (en) Process status diagnosis method and status diagnosis device
JP2009099960A (en) Quality control method, manufacturing method of semiconductor device, and quality control system
US11237550B2 (en) Ultrasonic flow meter prognostics with near real-time condition based uncertainty analysis
Tsai Development of an integrated reflow soldering control system using incremental hybrid process knowledge
JP2010224988A (en) Quality control system, quality control method, quality control program, and method for manufacturing product
WO2024116560A1 (en) Abnormality determination device and abnormality determination method
Tsarouhas Reliability, availability, and maintainability analysis of an industrial plant based on Six Sigma approach: A case study in plastic industry
Delgado et al. Benefits of multivariate statistical process control based on principal component analysis in solder paste printing process where 100% automatic inspection is already installed

Legal Events

Date Code Title Description
A621 Written request for application examination

Free format text: JAPANESE INTERMEDIATE CODE: A621

Effective date: 20230314

A977 Report on retrieval

Free format text: JAPANESE INTERMEDIATE CODE: A971007

Effective date: 20231214

A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20231219

A521 Request for written amendment filed

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20240131

TRDD Decision of grant or rejection written
A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

Effective date: 20240402

A61 First payment of annual fees (during grant procedure)

Free format text: JAPANESE INTERMEDIATE CODE: A61

Effective date: 20240415

R150 Certificate of patent or registration of utility model

Ref document number: 7484281

Country of ref document: JP

Free format text: JAPANESE INTERMEDIATE CODE: R150