Towards a Real-Time In-Flight Ice Detection System via Computational Aeroacoustics and Bayesian Neural Networks

Abstract

In-flight rotor icing presents a serious problem in the operation of rotorcraft in cold climates, as complex ice shapes can significantly degrade the aerodynamic performance and handling characteristics of rotorcraft. Reliable real-time detection of ice formation is a critical enabling technology in improving rotorcraft safety. In this work, we propose a novel approach towards developing a real-time in-flight ice detection system using computational aeroacoustics and Bayesian neural networks. In particular, an icing simulation code based on a fully unsteady collection efficiency approach is coupled with the aeroacoustic solver available in the open- source software suite SU2, in order to compute far-field noise signatures corresponding to discrete iced rotor blades in various icing conditions. Additionally, Bayesian neural networks are constructed from the dataset thus generated to enable rapid predictions together with uncertainty information, of aerodynamic performance indicators from acoustic signal, as a first step in developing an in-flight ice detection and warning system.

Publication
AIAA Aviation 2019 Forum
Date
Citation
B. Y. Zhou, N. R. Gauger, M. Morelli, A. Guardone, J. Hauth, and X. Huan. Towards a Real-Time In-Flight Ice Detection System via Computational Aeroacoustics and Bayesian Neural Networks. In AIAA Aviation 2019 Forum, AIAA paper 2019–3103, Dallas, TX, 2019. https://dx.doi.org/10.2514/6.2019-3103

BibTeX

@inproceedings{Zhou2019a,
address = {Dallas, TX},
author = {Zhou, Beckett Y. and Gauger, Nicolas R. and Morelli, Myles and Guardone, Alberto and Hauth, Jeremiah and Huan, Xun},
booktitle = {AIAA Aviation 2019 Forum},
doi = {10.2514/6.2019-3103},
number = {2019-3103},
title = {{Towards a Real-Time In-Flight Ice Detection System via Computational Aeroacoustics and Bayesian Neural Networks}},
year = {2019}
}