Uncertainty Quantification for Full-Flight Data Based Engine Fault Detection with Neural Networks
Current state-of-the-art engine condition monitoring is based on a minimum of one steady-state data point per flight. Due to the scarcity of available data points, there are difficulties distinguishing between random scatter and an underlying fault introducing a detection latency of several flights....
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-09-01
|
| Series: | Machines |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-1702/10/10/846 |
| _version_ | 1797406969282166784 |
|---|---|
| author | Matthias Weiss Stephan Staudacher Jürgen Mathes Duilio Becchio Christian Keller |
| author_facet | Matthias Weiss Stephan Staudacher Jürgen Mathes Duilio Becchio Christian Keller |
| author_sort | Matthias Weiss |
| collection | DOAJ |
| description | Current state-of-the-art engine condition monitoring is based on a minimum of one steady-state data point per flight. Due to the scarcity of available data points, there are difficulties distinguishing between random scatter and an underlying fault introducing a detection latency of several flights. Today’s increased availability of data acquisition hardware in modern aircraft provides continuously sampled in-flight measurements, so-called full-flight data. These full-flight data give access to sufficient data points to detect faults within a single flight, significantly improving the availability and safety of aircraft. Artificial neural networks are considered well suited for the timely analysis of an extensive amount of incoming data. This article proposes uncertainty quantification for artificial neural networks, leading to more reliable and robust fault detection. An existing approach for approximating the aleatoric uncertainty was extended by an Out-of-Distribution Detection in order to take the epistemic uncertainty into account. The method was statistically evaluated, and a grid search was performed to evaluate optimal parameter combinations maximizing the true positive detection rates. All test cases were derived based on in-flight measurements of a commercially operated regional jet. Especially when requiring low false positive detection rates, the true positive detections could be improved 2.8 times while improving response times by approximately 6.9 compared to methods only accounting for the aleatoric uncertainty. |
| first_indexed | 2024-03-09T03:34:29Z |
| format | Article |
| id | doaj.art-4b5da315b59141b5a7a58ca3ddd84313 |
| institution | Directory Open Access Journal |
| issn | 2075-1702 |
| language | English |
| last_indexed | 2024-03-09T03:34:29Z |
| publishDate | 2022-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Machines |
| spelling | doaj.art-4b5da315b59141b5a7a58ca3ddd843132023-12-03T14:50:44ZengMDPI AGMachines2075-17022022-09-01101084610.3390/machines10100846Uncertainty Quantification for Full-Flight Data Based Engine Fault Detection with Neural NetworksMatthias Weiss0Stephan Staudacher1Jürgen Mathes2Duilio Becchio3Christian Keller4Institute of Aircraft Propulsion Systems, University of Stuttgart, 70569 Stuttgart, GermanyInstitute of Aircraft Propulsion Systems, University of Stuttgart, 70569 Stuttgart, GermanyMTU Aero Engines AG, 80995 München, GermanyMTU Aero Engines AG, 80995 München, GermanyMTU Maintenance Hannover GmbH, 30855 Langenhagen, GermanyCurrent state-of-the-art engine condition monitoring is based on a minimum of one steady-state data point per flight. Due to the scarcity of available data points, there are difficulties distinguishing between random scatter and an underlying fault introducing a detection latency of several flights. Today’s increased availability of data acquisition hardware in modern aircraft provides continuously sampled in-flight measurements, so-called full-flight data. These full-flight data give access to sufficient data points to detect faults within a single flight, significantly improving the availability and safety of aircraft. Artificial neural networks are considered well suited for the timely analysis of an extensive amount of incoming data. This article proposes uncertainty quantification for artificial neural networks, leading to more reliable and robust fault detection. An existing approach for approximating the aleatoric uncertainty was extended by an Out-of-Distribution Detection in order to take the epistemic uncertainty into account. The method was statistically evaluated, and a grid search was performed to evaluate optimal parameter combinations maximizing the true positive detection rates. All test cases were derived based on in-flight measurements of a commercially operated regional jet. Especially when requiring low false positive detection rates, the true positive detections could be improved 2.8 times while improving response times by approximately 6.9 compared to methods only accounting for the aleatoric uncertainty.https://www.mdpi.com/2075-1702/10/10/846aircraft enginegas turbinefault detectionengine health monitoringengine condition monitoringfull-flight data |
| spellingShingle | Matthias Weiss Stephan Staudacher Jürgen Mathes Duilio Becchio Christian Keller Uncertainty Quantification for Full-Flight Data Based Engine Fault Detection with Neural Networks Machines aircraft engine gas turbine fault detection engine health monitoring engine condition monitoring full-flight data |
| title | Uncertainty Quantification for Full-Flight Data Based Engine Fault Detection with Neural Networks |
| title_full | Uncertainty Quantification for Full-Flight Data Based Engine Fault Detection with Neural Networks |
| title_fullStr | Uncertainty Quantification for Full-Flight Data Based Engine Fault Detection with Neural Networks |
| title_full_unstemmed | Uncertainty Quantification for Full-Flight Data Based Engine Fault Detection with Neural Networks |
| title_short | Uncertainty Quantification for Full-Flight Data Based Engine Fault Detection with Neural Networks |
| title_sort | uncertainty quantification for full flight data based engine fault detection with neural networks |
| topic | aircraft engine gas turbine fault detection engine health monitoring engine condition monitoring full-flight data |
| url | https://www.mdpi.com/2075-1702/10/10/846 |
| work_keys_str_mv | AT matthiasweiss uncertaintyquantificationforfullflightdatabasedenginefaultdetectionwithneuralnetworks AT stephanstaudacher uncertaintyquantificationforfullflightdatabasedenginefaultdetectionwithneuralnetworks AT jurgenmathes uncertaintyquantificationforfullflightdatabasedenginefaultdetectionwithneuralnetworks AT duiliobecchio uncertaintyquantificationforfullflightdatabasedenginefaultdetectionwithneuralnetworks AT christiankeller uncertaintyquantificationforfullflightdatabasedenginefaultdetectionwithneuralnetworks |