A study on the condition based maintenance evaluation system of smart plant device using convolutional neural network

Mi Kyeong Shin, Woo Jin Jo, Hye Min Cha, Soo Hong Lee

Research output: Contribution to journalArticlepeer-review

12 Citations (Scopus)

Abstract

There are two main causes of plant accidents: poor maintenance management and human error. In this study, we implemented a smart plant maintenance system that can reduce human errors based on the conditional based maintenance (CBM) concept. Unlike smart plant technology, which focuses on existing technology, we interviewed actual engineers and implemented a system reflecting their needs. First, we implemented three methods for learning defective images using convolutional neural network (CNN) and found that blob detection processing improves learning accuracy. Second, the fitness for service API (FFS API) methodology used in the actual pitting corrosion maintenance evaluation method was used to implement the CBM system. Finally, we verified the reliability of this system by conducting validation through actual case study.

Original languageEnglish
Pages (from-to)2507-2514
Number of pages8
JournalJournal of Mechanical Science and Technology
Volume34
Issue number6
DOIs
Publication statusPublished - 2020 Jun 1

Bibliographical note

Publisher Copyright:
© 2020, The Korean Society of Mechanical Engineers and Springer-Verlag GmbH Germany, part of Springer Nature.

All Science Journal Classification (ASJC) codes

  • Mechanics of Materials
  • Mechanical Engineering

Fingerprint

Dive into the research topics of 'A study on the condition based maintenance evaluation system of smart plant device using convolutional neural network'. Together they form a unique fingerprint.

Cite this