Prognostics and Health Management

  • Research
  • >
  • Prognostics and Health Management

Prognostics and health management (PHM)

Prognostics and health management (PHM) is a multifaceted discipline that protects the integrity of components, products, and systems of systems by avoiding unanticipated problems that can lead to performance deficiencies and adverse effects on safety.

Prognostics is the process of predicting a system’s remaining useful life (RUL). By estimating the progression of a fault given the current degree of degradation, the load history, and the anticipated future operational and environmental conditions, PHM can predict when a product or system will no longer perform its intended function within the desired specifications.

Human laboratory pursues to develop the methodology that diagnoses health and predicts the remaining useful life (RUL) of engineered systems in real-time. This research can lead to maximizing facility availability and reducing maintenance costs. It can make stable facility operations by minimizing the occurrence of failures.

  • A novel feature learning research using deep learning and statistical analysis
  • Deep learning-based data-driven PHM method
  • Model-based PHM method with digital twin

PHM2.png

 

Reference: M.G.Pecht et al, "Prognostics and Health Management of Electronics," IEEE Press, Wiley

 

Acquiring fault data is a very difficult point in the PHM research field for training deep learning architecture. Collecting PHM data needs a lot of effort to emulate faulty situations. To solve this problem, Human Lab. has several testbeds that can simulate fault situations under various conditions. Also, Human Lab. is conducting many research projects that check the applicability of the proposed PHM techniques required in the real field. For more information, please check the above details.

 

홈페이지 1.png

Motor testbed for fault diagnosis research
(Fault type: motor winding faults per various load conditions)

 

홈페이지 2.png

Bearing & Rotor testbed for fault diagnosis research
(Fault type: bearing faults, rotor unbalance, shaft misalignment per various load conditions)

 

홈페이지 3.png

Bearing accelerated life testbed for prognostic research

(Fault type: bearing run-to-failure)

 

홈페이지 4.png

 

Fan testbed for heating, ventilation, & air conditioning (HVAC) fault diagnosis research

(Fault type: fan unbalance, belt looseness)

 

 

Established Dataset #1

Title: Vibration, Acoustic, Temperature, and Motor Current Dataset of Rotating Machine Under Varying Load Conditions for Fault Diagnosis

Link: https://data.mendeley.com/datasets/ztmf3m7h5x

 

Established Dataset #2

Title: Vibration and Current Dataset of Three-Phase Permanent Magnet Synchronous Motors with Stator Faults

Link: https://data.mendeley.com/datasets/rgn5brrgrn

 

Established Dataset #3 (Data files are divided into three parts because of the limitation of data volume)

Title: Vibration and Motor Current Dataset of Rolling Element Bearing Under Varying Speed Conditions for Fault Diagnosis

Link1: https://data.mendeley.com/datasets/vxkj334rzv

Link2: https://data.mendeley.com/datasets/x3vhp8t6hg

Link3: https://data.mendeley.com/datasets/j8d8pfkvj2

 

Established Dataset #4

Title: Vibration and Motor Current Dataset of Rolling Element Bearing Under Run-to-Failure

Link: TBA (maybe distributed after December 2022).

 

Collaboration & on-going project (last updated on November, 2022)

  • 전류신호 이미지화 기법을 이용한 컴프레서 저널베어링 마모 진단기술 개발, LG전자
  • 하모닉 감속기 고장 특성 분석, 한국생산기술연구원
  • 음향진동 자료 기반 인공지능 고장진단체계의 K1전차 동력장치 진단 연구, 국방과학연구소, 민군협력진흥원
  • 정상데이터 기반 구동모터 고장진단 기술개발, LS전선
  • 악조건 속에서의 진동,전류,영상데이터를 이용한 회전체 고장진단 및 수명예측, 삼성중공업
  • 인공지능 기반 음향인식 기술을 이용한 수배전반 스마트 안전진단 플랫폼 개발, 동서발전
  • 인공신경망 및 전산역학 기반 풍력타워 구조 건전성 감시 기법 개발, 한전전력연구원
  • 디지털트윈 및 인공신경망 기반 공조기 건전성 모니터링 기법 개발, KAIST
  • 진동신호 기반 제조설비 이상진단을 위한 모션증폭기술 및 디지털트윈 모델링 기법 개발, KAIST
  • 영구자석 전동기 고장진단을 위한 고장데이터 구축 사업, KAIST

 

 

홈페이지 6.png