Conference paper
Application of the rolling pin method to detect anomalies and fault conditions
Advances in Technology to Support End User Mission - Proceedings of the 2016 Joint Conference/Symposium of the Society for Machinery Failure Prevention Technology and the International Society of Automation
May 2016
Abstract
• The fact that an event has not occurred does not indicate that the probability of the event occurring is zero - Copulas are a good tool to estimate such probabilities - The Rolling Pin Method is a computationally efficient alternative that avoids information loss with Bayesian Networks
• The Rolling Pin Method has been demonstrated to be applicable in the online detection and diagnosis of faults for simulated problems of industrial interest
• Broader Applications under consideration - Incorporate time dynamics (same variable, different time points) - Incorporate sensor faults - Estimating Mean Time Between Failure (MTBF) with confidence.
Metrics
1 Record Views
Details
- Title
- Application of the rolling pin method to detect anomalies and fault conditions
- Creators
- Jeffrey E. Arbogast - Air LiquideTaha Mohseni Ahooyi - Drexel UniversityMasoud Soroush - Drexel University, Chemical and Biological Engineering
- Publication Details
- Advances in Technology to Support End User Mission - Proceedings of the 2016 Joint Conference/Symposium of the Society for Machinery Failure Prevention Technology and the International Society of Automation
- Conference
- 2016 Joint Conference/Symposium of the Society for Machinery Failure Prevention Technology and the International Society of Automation (Dayton, Ohio, United States, 24 May 2016–26 May 2016)
- Resource Type
- Conference paper
- Language
- English
- Academic Unit
- Chemical and Biological Engineering
- Identifiers
- 991022005883704721