Availability Evaluation of Finite Element Analysis Data for Data Supplementation with Motor Failure Diagnosis Algorithm Training Data

The classic method of diagnosing motor failure is to analyze the signal characteristics of the motor. If a deep learning algorithm is used, the above process is omitted. This is replaced by a lot of data. However, the number of data that can be collected is limited. This is because the degree of fai...

Full description

Saved in:
Bibliographic Details
Main Authors: Han, Ji-Hoon, Choi, Eui-Jin, Hong, Sun-Ki
Format: Conference Proceeding
Language:English
Subjects:
Deep learning
Failure analysis
FEM
Finite element analysis
Motor fault diagnosis
Neural networks
Shape
Systems engineering and theory
Training data
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The classic method of diagnosing motor failure is to analyze the signal characteristics of the motor. If a deep learning algorithm is used, the above process is omitted. This is replaced by a lot of data. However, the number of data that can be collected is limited. This is because the degree of failure and the condition of the motor are different in each situation. To solve this problem, research on data propagation using FEM (Finite Element Method) is in progress. FEM is used to check whether the propagated data is similar to classical theory and whether it can be used as training data for deep learning algorithms. Through these experiments, it was confirmed that the FEM data is data that helps in learning the deep learning algorithm that can classify the initial failure state.
ISSN:2771-7577
DOI:10.1109/SCEMS56272.2022.9990660