Open AccessRemaining Useful Life for Aircraft Engine Forecasting Based on Data-driven
Author(s) -
Wei Niu,
Wei Huang,
Hao Wang,
Jiqi Cheng,
Sen Wang
Publication year - 2022
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2183/1/012031
Subject(s) - avionics , particle swarm optimization , computer science , sequence (biology) , time series , process (computing) , series (stratigraphy) , data mining , machine learning , engineering , aerospace engineering , paleontology , biology , genetics , operating system
As for the low forecasting precision problem about time series in avionics complex system, a time series combined forecasting model of complex systems was established based on GM (grey model is referred to as GM). However, using the traditional GM(1, 1) model to predict complex system failures might result in huge forecasting errors. To solve the drawbacks of the traditional GM(1, 1) model, this work developed the traditional GM(1, 1) model (abbreviated as DGM(1, 1)), which can predict non-geometric sequences more correctly. To reduce such mistakes, a PS-DGM (1, 1) forecast model was created by combining DGM(1, 1) with the particle-swarm optimization process. The DGM(1, 1) model was utilized to generate the forecasting sequence, and then the particle-swarm optimization approach was employed to further alter them, resulting in the degradation trend prediction of avionics complex system characteristic parameters. The experimental results show that the PS-DGM(1, 1) model’s validity was confirmed by tracking and forecasting avionics equipment experiment data in a real-world setting, and that the suggested model achieves higher precision than the old one.