Predicting Helicopter Faults by Analyzing the Stability of Vibration TimeSeries

Classification Trees;Classification;Helicopters;Failures;Time Series;Explosive;Stability;Vibration;Neural Networks;Neural;Networks

Cox, Alan R Jr. ; Thom Hodgson, Committee Chair,Subhashis Ghosal, Committee Member,Ralph Smith, Committee Member,LTC Michael J. Kwinn, Jr., Committee Member,Cox, Alan R Jr. ; Thom Hodgson ; Committee Chair ; Subhashis Ghosal ; Committee Member ; Ralph Smith ; Committee Member ; LTC Michael J. Kwinn ; Jr. ; Committee Member

University:North Carolina State University

关键词: Classification Trees;    Classification;    Helicopters;    Failures;    Time Series;    Explosive;    Stability;    Vibration;    Neural Networks;    Neural;    Networks;   

Others  :  https://repository.lib.ncsu.edu/bitstream/handle/1840.16/2366/etd.pdf?sequence=1&isAllowed=y

美国|英语

来源: null

【 摘 要 】

The U.S. Army's Lead-the-Fleet (LTF) program was started to help the Army develop a better maintenance program for its helicopters. This thesis explores vibration data gathered from the U.S. Army's and the South Carolina National Guard's Vibration Management Enhancement Program (VMEP). Vibration time series are classified as either "explosive" or "stationary." This classification is then used by neural networks and classification trees to predict whether a part failed directly after a flight and is need of replacement. The belief is that this will give the maintenance personnel a better understanding of when parts fail, allowing for a more accurate replacement schedule that could save money and improve safety.