METHOD OF CREATING LEARNED PREDICTION MODEL FOR PREDICTING DROSS AMOUNT OCCURRED IN MELTING FURNACE, METHOD OF PREDICTING DROSS AMOUNT GENERATED IN MELTING FURNACE, AND COMPUTER PROGRAM

Title:

METHOD OF CREATING LEARNED PREDICTION MODEL FOR PREDICTING DROSS AMOUNT OCCURRED IN MELTING FURNACE, METHOD OF PREDICTING DROSS AMOUNT GENERATED IN MELTING FURNACE, AND COMPUTER PROGRAM

Document Type and Number:

Japanese Patent JP2023125753

Kind Code:

A

Abstract:

To provide a creation method of a learned prediction model for predicting a dross generation amount in a melting furnace.SOLUTION: A method for creating a learned model includes: a step of acquiring a process state parameter for each charge (S110); a step (S130) of executing a pretreatment by applying mechanical learning to a data set of one or a plurality of process state parameters acquired in charge of m times (m is an integer of 2 or more); a step of creating a learning data set (S140); and a step of creating a learned model (S150), wherein a learning data set is created based on an n-dimensional (n is an integer of 1 or more)feature amount extracted in preprocessing, and includes one or a plurality of process target parameters indicating process basic information set for each charge.SELECTED DRAWING: Figure 5

Inventors:

YOMOGIDA SHOHEI
YAMAMOTO YUKI

Application Number:

JP2022030022A

Publication Date:

September 07, 2023

Filing Date:

February 28, 2022

Export Citation:

Click for automatic bibliography generation Help

Assignee:

UACJ CORP

International Classes:

C22B9/00; G05B23/02

Attorney, Agent or Firm:

Seiji Okuda
Shuichi Kita
Shigeyasu Murase

Previous Patent: Konnyaku noodles and their manufacturing method

Next Patent: Trajectory planning device and trajectory planning method