Title:
学習装置およびプログラム、異常要因推定システム、ならびに鍛造プレス装置
Document Type and Number:
Japanese Patent JP7229140
Kind Code:
B2
Abstract:
To enable an abnormality factor of a product to be estimated.SOLUTION: A learning device (100) comprises: creating means (101) that creates a learning model (111) by machine-learning, for each mold pattern, correlation between respective condition feature amounts of a plurality of kinds of processing conditions including a die height value, mold contact time and a mold temperature which are detected by sensors mounted on a forging press device and a load feature amount which is a load peak value or a load increasing tendency; and analyzing means (104) that analyzes importance of the processing conditions with respect to the load feature amount, for each mold pattern, on the basis of the learning model created by the creating means, and estimates the processing condition related to quality of a product, of the plurality of kinds of processing conditions, for each mold pattern.SELECTED DRAWING: Figure 4
More Like This:
Inventors:
Kyoji Nakatani
Koji Nishioka
Yusuke Sasaki
Minami Ryosuke
Koji Nishioka
Yusuke Sasaki
Minami Ryosuke
Application Number:
JP2019181974A
Publication Date:
February 27, 2023
Filing Date:
October 02, 2019
Export Citation:
Assignee:
Kurimoto Iron Works Co., Ltd.
International Classes:
B30B15/28; B21J5/00; B30B15/00; G05B19/18; G05B19/4155
Domestic Patent References:
| JP2019013976A | ||||
| JP6304800A | ||||
| JP7164199A |
Foreign References:
| US4723429 |
Attorney, Agent or Firm:
Patent Attorney Imy International Patent Office
Previous Patent: Tool wear diagnosis device for machine tools
Next Patent: CONCRETE PILE AND CONCRETE PIPE
Next Patent: CONCRETE PILE AND CONCRETE PIPE