Skip to main navigation Skip to main content
  • E-Submission

JKSPE : Journal of the Korean Society for Precision Engineering

OPEN ACCESS
ABOUT
BROWSE ARTICLES
EDITORIAL POLICIES
FOR CONTRIBUTORS
REGULAR

머신러닝을 이용한 종이용기 성형기의 캠 마모 고장 진단 알고리즘 개발

Development of Diagnosis Algorithm for Cam Wear of Paper Container Using Machine Learning

Journal of the Korean Society for Precision Engineering 2019;36(10):953-959.
Published online: October 1, 2019

1 금오공과대학교 기계시스템공학과

2 현진제업 기술연구소

1 Department of Mechanical System Engineering, Kumoh National Institute of Technology

2 Technical Research Center, Hyunjin Co., Ltd.

#E-mail: bschu@kumoh.ac.kr, TEL: +82-54-478-7398
• Received: March 8, 2019   • Revised: April 25, 2019   • Accepted: May 24, 2019

Copyright © The Korean Society for Precision Engineering

This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

  • 7 Views
  • 0 Download
  • 1 Crossref
  • 1 Scopus
prev next

Citations

Citations to this article as recorded by  Crossref logo
  • A Study on 3D Printing Conditions Prediction Model of Bone Plates Using Machine Learning
    Song Yeon Lee, Yong Jeong Huh
    Journal of the Korean Society for Precision Engineering.2022; 39(4): 291.     CrossRef

Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:

Include:

Development of Diagnosis Algorithm for Cam Wear of Paper Container Using Machine Learning
J. Korean Soc. Precis. Eng.. 2019;36(10):953-959.   Published online October 1, 2019
Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:
Include:
Development of Diagnosis Algorithm for Cam Wear of Paper Container Using Machine Learning
J. Korean Soc. Precis. Eng.. 2019;36(10):953-959.   Published online October 1, 2019
Close

Figure

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
Development of Diagnosis Algorithm for Cam Wear of Paper Container Using Machine Learning
Image Image Image Image Image Image Image
Fig. 1 Hi-1000 Paper cup forming machine
Fig. 2 3D model of bottom incurl process
Fig. 3 Test bed of bottom incurl process (a) Normal cam and (b) Abnormal cam
Fig. 4 Intentional cam failure for abnormal vibration detection experiment
Fig. 5 Decision boundary according to K
Fig. 6 Vibration data and extracted features
Fig. 7 Classified data distribution
Development of Diagnosis Algorithm for Cam Wear of Paper Container Using Machine Learning

Failure statistics of paper cup forming machine

Part Counts Total failure time (min)
Sidewall feeder 91 4540
Sidewall blanking die 147 9644
Transfer turret 53 2347
Sidewall heater station 223 8797
Mandrel turret 140 6218
Bottom blanking die 290 15759
Bottom reformer 36 2066

Experimental data acquisition system

Components Function Specification

cDAQ-9133
Data
acquisition
platform
- CPU : 1.33 GHz intel Atom
- Slot count : 8 slots
- Operating : -20 to 55oC
- System memory : 2 GB DDR

NI 9234
Acceleration
acquisition
module
- Sampling speed : 51.2 kS/s
- Resolution : 24-bit
- Dynamic range : 102 dB

8702B50
Accelerometer - Sensitivity : 9.8 mV/g
- Frequency range : 1-5000 Hz
- Output voltage : ±5 V
- Acceleration range : ±50 g

Experimental setup for data acquisition

Class Cam Speed Mass (kg) Number
1 Normal 250 3, 6 Normal cam 1
300 3, 6 Normal cam 2
2 Abnormal 250 3, 6 Abnormal cam 1
300 3, 6 Abnormal cam 2

Features for K-NN classification

Feature Mathematical expression
RMS f r m s = lim T 1 T 0 T f t 2 d t
Mean x = 1 n n i = 1 x i
Median P X m 1 2 P X m 1 2
Variance σ 2 = 1 n n i = 1 X i - X ¯ 2
Standard deviation σ X = E X 2 - E X 2
Skewness S k e w = 1 n n i = 1 X i - X ¯ σ 3
Kurtosis K u r t = 1 n n i = 1 X i - X ¯ σ 4

Classification result (K = 1)

Input Data 250 rpm 300 rpm
Training data Normal : 50 53 46
Fault : 50 47 54
Normal data Normal : 50 47 42
Fault : 0 3 8
Fault data Normal : 0 4 9
Fault : 50 46 41

Classification result (K = 5)

Input Data 250 rpm 300 rpm
Training data Normal : 50 52 47
Fault : 50 48 53
Normal data Normal : 50 48 42
Fault : 0 2 8
Fault data Normal : 0 5 10
Fault : 50 45 40

Classification result (K = 10)

Input Data 250 rpm 300 rpm
Training data Normal : 50 57 53
Fault : 50 43 47
Normal data Normal : 50 41 39
Fault : 0 9 11
Fault data Normal : 0 5 10
Fault : 50 45 40

Classification result (K = 5)

Input Data 250 rpm 300 rpm
Training data Normal : 50 52 47
Fault : 50 48 53
Normal data Normal : 50 45 42
Fault : 0 5 8
Fault data Normal : 0 2 10
Fault : 50 48 40

Classification result by Normalize (K = 5)

Input Data 250 rpm 300 rpm
Training data Normal : 50 50 50
Fault : 50 50 50
Normal data Normal : 50 50 50
Fault : 0 0 0
Fault data Normal : 0 0 0
Fault : 50 50 50

Classification rate (K= 5)

Input Data 250 rpm (%) 300 rpm (%)
Normal data Normal : 50 100 100
Fault : 0 100 100
Fault data Normal : 0 100 100
Fault : 50 100 100
Table 1 Failure statistics of paper cup forming machine
Table 2 Experimental data acquisition system
Table 3 Experimental setup for data acquisition
Table 4 Features for K-NN classification
Table 5 Classification result (K = 1)
Table 6 Classification result (K = 5)
Table 7 Classification result (K = 10)
Table 8 Classification result (K = 5)
Table 9 Classification result by Normalize (K = 5)
Table 10 Classification rate (K= 5)