大学院
学内のオンライン授業の情報漏洩防止のため，URLやアカウント、教室の記載は削除しております。
最終更新日：2024年10月18日
授業計画や教室は変更となる可能性があるため、必ずUTASで最新の情報を確認して下さい。
UTASにアクセスできない方は、担当教員または部局教務へお問い合わせ下さい。
最終更新日：2024年10月18日
授業計画や教室は変更となる可能性があるため、必ずUTASで最新の情報を確認して下さい。
UTASにアクセスできない方は、担当教員または部局教務へお問い合わせ下さい。
Prognostics and System Health Management
Prognostic and System Health Management (PHM) is an engineering field aiming to provide a
thorough analysis of the health condition of a system and its components. It can help predict
impending faults and determine their remaining-useful-life. An efficient PHM system can
speed up fault analysis by providing early indication of what components of the system are
likely to fail and will need maintenance in the near future.
This course provides the essential concepts and methods of building a PHM system It gives an
overview on fault diagnosis and failure prognosis, techniques and applications in various areas
including structural health monitoring (SHM), aerospace maintenance and in-situ monitoring
technologies.
thorough analysis of the health condition of a system and its components. It can help predict
impending faults and determine their remaining-useful-life. An efficient PHM system can
speed up fault analysis by providing early indication of what components of the system are
likely to fail and will need maintenance in the near future.
This course provides the essential concepts and methods of building a PHM system It gives an
overview on fault diagnosis and failure prognosis, techniques and applications in various areas
including structural health monitoring (SHM), aerospace maintenance and in-situ monitoring
technologies.
時間割／共通科目コード
コース名
教員
学期
時限
3734-117
GEN-AA6o18L3
Prognostics and System Health Management
KHAN MOHAMMAD SAMIR
木曜4限
マイリストに追加
マイリストから削除
講義使用言語
英語
単位
2
実務経験のある教員による授業科目
NO
他学部履修
可
開講所属
工学系研究科
授業計画
Introduction to PHM basics (Samir)
The role of data in PHM (Samir)
PHM standards and architectures/frameworks: (Samir)
Machine learning methods - basics, supervised/unsupervised, Bayesian inference (Yairi)
Deep learning methods - NN, CNN, RNN (Samir)
Model based methods 1 - Inference for dynamical systems : Kalman filtering, Sequential Monte Carlo (Yairi)
Anomaly detection - Mahalanobis, SVM, GMM, isolation forest (Samir)
Model based methods 2 - Learning dynamical systems: EM algorithm, subspace method, Deep neural networks (Yairi)
Fault diagnostics - data-driven vs model based (Samir)
Fault prognostics - data drive vs model based (Samir)
Cost analysis of PHM (Samir)
RUL Regression (Yairi)
The ‘No fault found’ problem (Samir)
Future of PHM technology (Samir)
授業の方法
The first lecture on Apr. 7 will be hybrid. All the rest of the sessions will be on-line only.
成績評価方法
TBD
履修上の注意
指示しない
