JPWO2022153645A5 - - Google Patents
Download PDFInfo
- Publication number
- JPWO2022153645A5 JPWO2022153645A5 JP2022575086A JP2022575086A JPWO2022153645A5 JP WO2022153645 A5 JPWO2022153645 A5 JP WO2022153645A5 JP 2022575086 A JP2022575086 A JP 2022575086A JP 2022575086 A JP2022575086 A JP 2022575086A JP WO2022153645 A5 JPWO2022153645 A5 JP WO2022153645A5
- Authority
- JP
- Japan
- Prior art keywords
- waveform image
- model data
- model
- waveform
- generation unit
- 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.)
- Pending
Links
- 238000004519 manufacturing process Methods 0.000 claims 16
- 238000001746 injection moulding Methods 0.000 claims 14
- 238000013500 data storage Methods 0.000 claims 13
- 238000002347 injection Methods 0.000 claims 12
- 239000007924 injection Substances 0.000 claims 12
- 238000001514 detection method Methods 0.000 claims 11
- 230000002159 abnormal effect Effects 0.000 claims 6
- 238000010801 machine learning Methods 0.000 claims 5
- 239000011347 resin Substances 0.000 claims 4
- 229920005989 resin Polymers 0.000 claims 4
- 238000009825 accumulation Methods 0.000 claims 3
- 238000000034 method Methods 0.000 claims 3
- 230000005856 abnormality Effects 0.000 claims 2
- 230000006399 behavior Effects 0.000 claims 2
- 239000000463 material Substances 0.000 claims 2
- 238000005259 measurement Methods 0.000 claims 2
- 230000035945 sensitivity Effects 0.000 claims 2
Claims (16)
前記入力部においてデジタルデータに変換された検出信号をグラフ化して波形画像を生成する波形画像生成部と、
各々の前記センサが正常状態である条件の下で波形画像生成部で生成された波形画像、ならびに当該波形画像に関する品質情報を共に入力して両者を結び付けたモデルデータを生成するモデルデータ生成部と、
前記モデルデータ生成部で生成されたモデルデータを記憶するモデルデータ記憶部と、
前記波形画像生成部により製品の製造過程において前記センサで検出して得られる波形画像を判定対象とし、判定対象となる前記波形画像に対して前記モデルデータ記憶部に記憶されたモデルデータに含まれる前記波形画像を比較して両者の類似度を算出し、前記製品の正常および異常を判定する画像比較部と、を備える計測装置。 an input unit for inputting and A/D-converting detection signals from a plurality of sensors arranged in a manufacturing apparatus;
a waveform image generator that generates a waveform image by graphing the detection signal converted into digital data in the input unit;
a model data generation unit that receives both the waveform image generated by the waveform image generation unit under the condition that each of the sensors is in a normal state and quality information about the waveform image, and generates model data that combines the two; ,
a model data storage unit that stores the model data generated by the model data generation unit;
The waveform image generated by the waveform image generation unit detected by the sensor in the manufacturing process of the product is determined, and is included in the model data stored in the model data storage unit for the waveform image to be determined. An image comparison unit that compares the waveform images to calculate a degree of similarity between them and determines whether the product is normal or abnormal.
前記波形画像生成部は、射出成形処理の１サイクルごとに得られる各々の前記センサからの検出信号に基づく波形のグラフを各センサに対して１枚ずつ作成し、このグラフを画像に変換して波形画像を生成する、請求項１に記載の計測装置。 The manufacturing device is an injection molding device,
The waveform image generator creates a waveform graph for each sensor based on the detection signal from each sensor obtained in each cycle of the injection molding process, and converts the graph into an image. 2. The metrology device of claim 1, which generates a waveform image.
前記波形画像生成部は、射出成形処理の１サイクルごとに得られる各々の前記センサからの検出信号に基づく複数の波形のグラフを１画面内に集約して作成し、集約した前記波形のグラフを画像に変換して波形画像を生成する、請求項１に記載の計測装置。 The manufacturing device is an injection molding device,
The waveform image generation unit aggregates and creates a plurality of waveform graphs based on the detection signals from the sensors obtained in each cycle of the injection molding process in one screen, and creates a graph of the aggregated waveforms. 2. The measuring device according to claim 1, which converts to an image to generate a waveform image.
前記波形画像生成部は、前記射出成形装置の金型に配置された圧力センサにより検出された検出信号をグラフ化して波形画像を生成し、
前記モデルデータ生成部は、前記波形画像生成部で生成された波形画像、ならびに当該波形画像に対する、バリ、オーバーパック、ショートショット、ランナーバランス、フローマークまたはヒケの少なくとも一つに関する品質情報を共に入力して両者を結びつけたモデルデータを生成し、
前記モデルデータ記憶部は、前記モデルデータ生成部で生成されたモデルデータを記憶し、
前記画像比較部は、前記波形画像生成部により製品の製造過程において前記圧力センサで検出して得られる波形画像を判定対象とし、判定対象となる前記波形画像に対して前記モデルデータ記憶部に記憶されたモデルデータに含まれる前記波形画像を比較して両者の類似度を算出し、前記製品の正常および異常を判定する、請求項１から請求項５のいずれか１項に記載の計測装置。 The manufacturing device is an injection molding device,
The waveform image generation unit generates a waveform image by graphing a detection signal detected by a pressure sensor arranged in a mold of the injection molding apparatus,
The model data generator inputs the waveform image generated by the waveform image generator and quality information regarding at least one of burrs, overpacks, short shots, runner balance, flow marks, and sink marks for the waveform image. and generate model data that connects both,
The model data storage unit stores the model data generated by the model data generation unit,
The image comparison unit determines a waveform image obtained by detecting the pressure sensor with the pressure sensor in the manufacturing process of the product by the waveform image generation unit, and stores the waveform image to be determined in the model data storage unit. The measuring device according to any one of claims 1 to 5, wherein the waveform images included in the obtained model data are compared to calculate a degree of similarity between the two to determine whether the product is normal or abnormal.
前記波形画像生成部は、前記射出成形装置の射出部に配置された圧力センサにより検出された検出信号をグラフ化して波形画像を生成し、
前記モデルデータ生成部は、前記波形画像生成部で生成された波形画像、ならびに当該波形画像に対する、射出工程の樹脂流動挙動または可塑化安定性の少なくとも一つに関する品質情報を共に入力して両者を結びつけたモデルデータを生成し、
前記モデルデータ記憶部は、前記モデルデータ生成部で生成されたモデルデータを記憶し、
前記画像比較部は、前記波形画像生成部により製品の製造過程において前記圧力センサで検出して得られる波形画像を判定対象とし、判定対象となる前記波形画像に対して前記モデルデータ記憶部に記憶されたモデルデータに含まれる前記波形画像を比較して両者の類似度を算出し、前記製品の正常および異常を判定する、請求項１から請求項５のいずれか１項に記載の計測装置。 The manufacturing device is an injection molding device,
The waveform image generation unit generates a waveform image by graphing a detection signal detected by a pressure sensor arranged in an injection unit of the injection molding apparatus,
The model data generation unit inputs the waveform image generated by the waveform image generation unit and quality information related to at least one of resin flow behavior in the injection process and plasticization stability for the waveform image, and combines the two. generate the linked model data,
The model data storage unit stores the model data generated by the model data generation unit,
The image comparison unit determines a waveform image obtained by detecting the pressure sensor with the pressure sensor in the manufacturing process of the product by the waveform image generation unit, and stores the waveform image to be determined in the model data storage unit. The measuring device according to any one of claims 1 to 5, wherein the waveform images included in the obtained model data are compared to calculate a degree of similarity between the two to determine whether the product is normal or abnormal.
前記入力部においてデジタルデータに変換された検出信号をグラフ化して波形画像を生成する波形画像生成部と、
各々の前記センサが正常状態である条件の下で波形画像生成部で生成された波形画像、ならびに当該波形画像に関する品質情報を共に入力して両者を結び付けたモデルデータを生成するモデルデータ生成部と、
前記モデルデータ生成部で生成されたモデルデータを記憶するモデルデータ記憶部と、
前記モデルデータ記憶部に記憶されているモデルデータを読み込み、機械学習により波形画像から品質を分類する学習済モデルを構築するモデル生成部と、
前記モデル生成部で生成された学習済モデルを記憶する学習済モデル記憶部と、
前記波形画像生成部により製品の製造過程において前記センサで検出して得られる波形画像を判定対象とし、判定対象となる前記波形画像を前記学習済モデル記憶部に記憶された学習済モデルへ入力し、前記製品の正常および異常を判定する推論部と、を備える計測装置。 an input unit for inputting and A/D-converting detection signals from a plurality of sensors arranged in a manufacturing apparatus;
a waveform image generator that generates a waveform image by graphing the detection signal converted into digital data in the input unit;
a model data generation unit that receives both the waveform image generated by the waveform image generation unit under the condition that each of the sensors is in a normal state and quality information about the waveform image, and generates model data that combines the two; ,
a model data storage unit that stores the model data generated by the model data generation unit;
a model generation unit that reads the model data stored in the model data storage unit and constructs a learned model that classifies quality from the waveform image by machine learning;
a learned model storage unit that stores the learned model generated by the model generation unit;
The waveform image generation unit detects a waveform image obtained by detecting the sensor in the manufacturing process of the product as a judgment object, and inputs the waveform image to be judgment object to the learned model stored in the learned model storage unit. , and an inference unit that determines whether the product is normal or abnormal.
前記波形画像生成部は、前記射出成形装置の金型に配置された圧力センサにより検出された検出信号をグラフ化して波形画像を生成し、
前記モデルデータ生成部は、前記波形画像生成部で生成された波形画像、ならびに当該波形画像に対する、バリ、オーバーパック、ショートショット、ランナーバランス、フローマークまたはヒケの少なくとも一つに関する品質情報を共に入力して両者を結びつけたモデルデータを生成し、
前記モデルデータ記憶部は、前記モデルデータ生成部で生成されたモデルデータを記憶し、
前記モデル生成部は、前記モデルデータ記憶部に記憶されているモデルデータを読み込み、機械学習により波形画像から品質を分類する学習済モデルを構築し、
前記学習済モデル記憶部は、前記モデル生成部で生成された学習済モデルを記憶し、
前記推論部は、前記波形画像生成部により製品の製造過程において前記センサで検出して得られる波形画像を判定対象とし、判定対象となる前記波形画像を前記学習済モデル記憶部に記憶された学習済モデルへ入力し、前記製品の正常および異常を判定する、請求項９に記載の計測装置。 The manufacturing device is an injection molding device,
The waveform image generation unit generates a waveform image by graphing a detection signal detected by a pressure sensor arranged in a mold of the injection molding apparatus,
The model data generator inputs the waveform image generated by the waveform image generator and quality information regarding at least one of burrs, overpacks, short shots, runner balance, flow marks, and sink marks for the waveform image. and generate model data that connects both,
The model data storage unit stores the model data generated by the model data generation unit,
The model generation unit reads the model data stored in the model data storage unit, builds a learned model that classifies the quality from the waveform image by machine learning,
The learned model storage unit stores the learned model generated by the model generation unit,
The inference unit determines a waveform image obtained by detecting the sensor in the manufacturing process of the product by the waveform image generation unit, and uses the waveform image to be determined as a learning target stored in the learned model storage unit. 10. The measuring device according to claim 9 , inputting to a finished model to determine normality and abnormality of said product.
前記波形画像生成部は、前記射出成形装置の射出部に配置された圧力センサにより検出された検出信号をグラフ化して波形画像を生成し、
前記モデルデータ生成部は、前記波形画像生成部で生成された波形画像、ならびに当該波形画像に対する、射出工程の樹脂流動挙動または可塑化安定性の少なくとも一つに関する品質情報を共に入力して両者を結びつけたモデルデータを生成し、
前記モデルデータ記憶部は、前記モデルデータ生成部で生成されたモデルデータを記憶し、
前記モデル生成部は、前記モデルデータ記憶部に記憶されているモデルデータを読み込み、機械学習により波形画像から品質を分類する学習済モデルを構築し、
前記学習済モデル記憶部は、前記モデル生成部で生成された学習済モデルを記憶し、
前記推論部は、前記波形画像生成部により製品の製造過程において前記センサで検出して得られる波形画像を判定対象とし、判定対象となる前記波形画像を前記学習済モデル記憶部に記憶された学習済モデルへ入力し、前記製品の正常および異常を判定する、請求項９に記載の計測装置。 The manufacturing device is an injection molding device,
The waveform image generation unit generates a waveform image by graphing a detection signal detected by a pressure sensor arranged in an injection unit of the injection molding apparatus,
The model data generation unit inputs the waveform image generated by the waveform image generation unit and quality information related to at least one of resin flow behavior in the injection process and plasticization stability for the waveform image, and combines the two. generate the linked model data,
The model data storage unit stores the model data generated by the model data generation unit,
The model generation unit reads the model data stored in the model data storage unit, builds a learned model that classifies the quality from the waveform image by machine learning,
The learned model storage unit stores the learned model generated by the model generation unit,
The reasoning unit determines a waveform image obtained by detecting the sensor in the manufacturing process of the product by the waveform image generation unit as a determination target, and uses the waveform image as a determination target for learning stored in the learned model storage unit. 10. The measuring device according to claim 9 , which inputs to a finished model and determines normality and abnormality of the product.
前記推論部は、前記モデル生成部で構築した学習済モデルを使用して、前記製品の正常および異常を判定する、請求項９から請求項１２のいずれか１項に記載の計測装置。 The model generation unit for constructing a trained model for classifying quality from waveform images by machine learning is arranged outside the measurement device,
13. The measuring device according to any one of claims 9 to 12 , wherein the inference unit uses the learned model constructed by the model generation unit to determine whether the product is normal or abnormal.
前記製品の製造過程において各センサの検出に基づいて得られる波形画像を判定対象とし、判定対象となる前記波形画像に対して、前記モデル蓄積モードで得られた前記モデルデータの波形画像を比較することで製品の品質を判定する品質判定モードと、を備えた計測方法。 A model that creates model data that links waveform images obtained under conditions in which good products are molded with each sensor in a normal state, and quality information corresponding to the waveform images, in order to compare product quality judgments. an accumulation mode;
A waveform image obtained based on the detection of each sensor in the manufacturing process of the product is determined, and the waveform image of the model data obtained in the model accumulation mode is compared with the waveform image to be determined. a quality judgment mode for judging product quality by
前記モデルデータを読み込み、機械学習により波形画像から品質を分類する学習済モデルを構築する学習モードと、
前記製品の製造過程において各センサで検出して得られる波形画像を判定対象とし、判定対象となる前記波形画像を前記学習済モデルへ入力し、前記製品の正常および異常を判定する品質判定モードと、を備えた計測方法。 A model that creates model data that links waveform images obtained under conditions in which good products are molded with each sensor in a normal state, and quality information corresponding to the waveform images, in order to compare product quality judgments. an accumulation mode;
a learning mode for reading the model data and constructing a trained model that classifies quality from the waveform image by machine learning;
A quality judgment mode for determining whether the product is normal or abnormal by inputting the waveform image to be determined to the trained model into the waveform image obtained by detecting each sensor in the manufacturing process of the product. , a method of measurement.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2021002552 | 2021-01-12 | ||
| PCT/JP2021/039667 WO2022153645A1 (en) | 2021-01-12 | 2021-10-27 | Measurement device, measurement method, and injection molding device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| JPWO2022153645A1 JPWO2022153645A1 (en) | 2022-07-21 |
| JPWO2022153645A5 true JPWO2022153645A5 (en) | 2023-05-02 |
Family
ID=82447114
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP2022575086A Pending JPWO2022153645A1 (en) | 2021-01-12 | 2021-10-27 |
Country Status (2)
| Country | Link |
|---|---|
| JP (1) | JPWO2022153645A1 (en) |
| WO (1) | WO2022153645A1 (en) |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH07205242A (en) * | 1994-01-13 | 1995-08-08 | Sumitomo Heavy Ind Ltd | Method and apparatus for monitoring quality and deciding propriety of molded form of injection molding machine |
| JP4297280B2 (en) * | 2005-03-08 | 2009-07-15 | 日精樹脂工業株式会社 | Waveform display method and apparatus for injection molding |
| JP6517728B2 (en) * | 2016-05-12 | 2019-05-22 | ファナック株式会社 | Device and method for estimating wear amount of check valve of injection molding machine |
| JP6772963B2 (en) * | 2017-06-05 | 2020-10-21 | トヨタ自動車株式会社 | Abnormality diagnosis device and abnormality diagnosis method |
| JP7151564B2 (en) * | 2019-03-14 | 2022-10-12 | Ｕｂｅマシナリー株式会社 | Injection molding machine and method for judging quality of molded products |
-
2021
- 2021-10-27 WO PCT/JP2021/039667 patent/WO2022153645A1/en active Application Filing
- 2021-10-27 JP JP2022575086A patent/JPWO2022153645A1/ja active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6557272B2 (en) | State determination device | |
| JP6616375B2 (en) | State determination device | |
| US6546311B2 (en) | Injection moulding system and method that determine optimum injection moulding conditions | |
| US11213986B2 (en) | Method for determining an actual volume of an injection moldable compound in an injection molding process | |
| US20240042665A1 (en) | Injection molding condition generation system and method | |
| CN105751463B (en) | Injection molding system | |
| US20200101649A1 (en) | Device for assisting molding condition determination and injection molding apparatus | |
| WO2020099557A1 (en) | Method and system for improving a physical production process | |
| TW202133968A (en) | Anomaly sensing device, anomaly sensing method, and computer program | |
| TW202142847A (en) | Abnormality detector, abnormality detecting method and computer program | |
| JP4364828B2 (en) | Molding machine monitoring apparatus, method and program | |
| JPWO2022153645A5 (en) | ||
| US20220339837A1 (en) | Operation Quantity Determination Device, Molding Apparatus System, Molding Machine, Non-Transitory Computer Readable Recording Medium, Operation Quantity Determination Method, and State Display Device | |
| JP7344754B2 (en) | Learning model generation method, computer program, setting value determination device, molding machine and molding device system | |
| CN108608627A (en) | Molding machine mode locking force setting method and its system | |
| WO2022075181A1 (en) | State determination device and state determination method | |
| Lopes et al. | Part quality prediction in an injection moulding process using neural networks | |
| WO2022153645A1 (en) | Measurement device, measurement method, and injection molding device | |
| WO2023032375A1 (en) | Method for adjusting molding condition parameters, computer program, device for adjusting molding condition parameters, and molding machine | |
| JP2021066057A (en) | Injection molding machine management device and injection molding machine | |
| WO2024106002A1 (en) | Molding condition correcting device, molding machine, molding condition correcting method, and computer program | |
| WO2024057461A1 (en) | Determination device and determination method | |
| US20240140010A1 (en) | Information processing device, injection molding machine, and non-transitory computer readable medium storing program | |
| US11642822B2 (en) | Largest empty corner rectangle based process development | |
| KR102500376B1 (en) | Monitoring apparatus using sensor signal in injection mold and method thereof |