TY - GEN
T1 - Turning dynamic sensor measurements from gas turbines into insights
T2 - ASME Turbo Expo 2019: Turbomachinery Technical Conference and Exposition, GT 2019
AU - Karlstetter, Roman
AU - Widhopf-Fenk, Robert
AU - Hermann, Jakob
AU - Rouwenhorst, Driek
AU - Raoofy, Amir
AU - Trinitis, Carsten
AU - Schulz, Martin
N1 - Publisher Copyright:
Copyright © 2019 ASME.
PY - 2019
Y1 - 2019
N2 - Gas turbine power plants generate an ever growing amount of high frequency dynamic sensor data. One of the applications of this data is the protection against problems induced by combustion dynamics, as, e.g., with the ArgusOMDS system developed by IfTA. In the light of digitalization, this data has the potential to also be used in other areas and ultimately transform maintenance, repair and overhaul approaches. However, current solutions are not designed to cope with the large time windows needed for a general analysis and this can hinder development of advanced machine analysis algorithms. In this work, we present an end-to-end approach for large scale sensor measurement analysis, employing data mining techniques and enabling machine learning algorithms. Our approach covers the complete data pipeline from sensor measurement acquisition to analysis and visualization. We demonstrate the feasibility of our approach by presenting several case studies that prove the benefits over existing solutions.
AB - Gas turbine power plants generate an ever growing amount of high frequency dynamic sensor data. One of the applications of this data is the protection against problems induced by combustion dynamics, as, e.g., with the ArgusOMDS system developed by IfTA. In the light of digitalization, this data has the potential to also be used in other areas and ultimately transform maintenance, repair and overhaul approaches. However, current solutions are not designed to cope with the large time windows needed for a general analysis and this can hinder development of advanced machine analysis algorithms. In this work, we present an end-to-end approach for large scale sensor measurement analysis, employing data mining techniques and enabling machine learning algorithms. Our approach covers the complete data pipeline from sensor measurement acquisition to analysis and visualization. We demonstrate the feasibility of our approach by presenting several case studies that prove the benefits over existing solutions.
UR - http://www.scopus.com/inward/record.url?scp=85075516119&partnerID=8YFLogxK
U2 - 10.1115/GT2019-91259
DO - 10.1115/GT2019-91259
M3 - Conference contribution
AN - SCOPUS:85075516119
T3 - Proceedings of the ASME Turbo Expo
BT - Ceramics; Controls, Diagnostics, and Instrumentation; Education; Manufacturing Materials and Metallurgy
PB - American Society of Mechanical Engineers (ASME)
Y2 - 17 June 2019 through 21 June 2019
ER -