| Autor: |
Ghadekar, Premanand, Dargode, Anuja, Deshmukh, Jayesh, Jaisinghani, Amit, Jadhav, Madhuri, Nimbalkar, Aditya |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings; 2023, Vol. 2981 Issue 1, p1-9, 9p |
| Abstrakt: |
Airlines want to be able to forecast engine breakdowns ahead of time to improve operations and reduce flight delays. Technology advancements have increased the use of condition analysis for industrial machinery using sensor data. These advancements have created a brand-new challenge in the processing and interpretation of sensor data. The use of sensors and telemetry data to monitor engine health and condition is thought to make this sort of maintenance easier by forecasting the Time-To-Failure (TTF) of in-service engines. As a result, maintenance work might be scheduled based on TTF estimates rather than/in addition to expensive time-based preventative maintenance. This paper proposes the use of deep learning, LSTM for the Turbofan dataset, and achieved an accuracy of 97 percent with good precision and f-1 scores i.e., 0.96 which is better than conventional models to predict correctly whether an engine will fail in a particular cycle or not. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
