Process monitoring of deep drawing using machine learning

摘要: This study proposes a new processing method by using the count rate of acoustic emission (AE) signal and machine learning. To analyze AE count, learning, multilayered neural networks, is implemented for deep drawing process. In press processing, quality inspection often carried out each lot in later Once failure occurs process, large number defective products may be produced due to fast speed. order prevent this, it important immediately stop just after defect occurs. The data has been used monitoring condition However, easily affected noise lacks repeatability. Also difficult handle its high frequencies target signals. Therefore, improvement recognition required thus learning approach applied this study.