상세정보
학술지
인공지능 모델을 이용한 공작기계의 스핀들 고장 진단
- 발행일
- 202105
- 출처
- 대한기계학회논문집 A, v.45 no.5, pp.401-408
- ISSN
- 1226-4873
- 출판사
- 대한기계학회
- 협약과제
- 20PT1600, 엣지 컴퓨팅 기반 공작기계 핵심요소 상태 진단 및 보안 기술 개발 , 나중찬
- 초록
- The development of information and communication technology (ICT) is accelerating innovation in the traditional manufacturing sector. The smart factory which is a state-of-the-art factory collects data in real time through various sensors. Recently, researches on applying the artificial intelligence technology to these collected data to detect machine failures has gained a lot of attention. In this study, we built a test bench to check the possibility of applying the artificial intelligence technology for the fault diagnosis of the spindle of machine tools. We collected failure data by changing the off-center of the spindle using bolts. Further, we used artificial intelligence models (CNN, LSTM, and auto-encoder) to analyze the accuracy of seven types of fault classifications. In addition, the method of data collection, data process, and model development is proposed to effectively apply the artificial intelligence technology to machine tool domains.
- KSP 제안 키워드
- Artificial intelligence models, Artificial intelligence technology, Auto-Encoder(AE), Data Collection, Data process, Information and communication technology(ICT), Machine tool, Manufacturing sector, Off-center, Real-Time, Smart Factory