1
Intelligent Helicopter Turbine Engine Fault Diagnosis Using Multi-
Head Attention
Yong Hun Park
1,a
, Hwan In Oh
1,a
, In Tae Kim
1
, So Jung Lee
1
, Se Hee Moon
1
, Gyu Jin Park
1
, Jeong Kyu Park
1
,
and Joon Ha Jung
1,*
1
Department of Industrial Engineering, Ajou University, Suwon, Gyeonggi-do, 16499, Republic of Korea
dydtm1225@ajou.ac.kr
ohhi0928@ajou.ac.kr
kitkit8142@ajou.ac.kr
judy0807@ajou.ac.kr
anstpgml5035@ajou.ac.kr
kjko46443@ajou.ac.kr
pjkp0310@ajou.ac.kr
joonha@ajou.ac.kr
ABSTRACT
A turbine engine provides power to the helicopter, enabling
the helicopter to travel and hover in the air. Since the
rotorcraft operates at high altitudes, ensuring safety and
maintaining a healthy operational status are crucial at all
times. Therefore, a prognostics and health management
(PHM) system for the turbine engine must be implemented to
predict any anomalies or faults to prevent catastrophic
accidents. This research proposes a novel fault diagnosis
method for helicopter turbine engines based on operational
data acquired from actual aircraft. First, the proposed method
predicts engine torque using other operational data while
accounting for uncertainty. A Bayesian regression approach
is employed to predict the engine torque. The torque margin,
defined as the difference between the actual torque and the
estimated torque, is then used to diagnose engine faults.
Specifically, a multi-head attention mechanism is
incorporated to capture interactions between various engine
parameters. Additionally, domain adaptation techniques are
applied to enhance the model's generalization performance,
ensuring robustness across diverse operating conditions. The
proposed method is validated using seven different datasets,
each acquired from a helicopter engine. Four datasets were
used for training, while the remaining three were allocated
for testing and validation. The results indicated that the
proposed method accurately predicted torque. Furthermore,
the fault diagnosis showed promising results, leading to a
3rd-place finish in the 2024 PHM Society Data Challenge in
terms of validation score.
1. INTRODUCTION
A helicopter, as a rotary-wing aircraft, operates by utilizing a
turbine engine to drive the rotor blades, enabling vertical
takeoff, landing, and flight. The turbine engine is the primary
power source that makes these maneuvers possible.
However, it is constantly exposed to various environmental
factors and harsh operational conditions. Over time, this
exposure can lead to a degradation in engine performance,
which may result in severe accidents. Therefore, to prevent
such accidents, a reliable engine monitoring and fault
diagnosis system is essential.
Previous studies have primarily focused on developing
monitoring and fault diagnosis methods based on engine
performance data. For instance, Fentaye et al. (2021) applied
a modular CNN to address the issue of fault detection and
classification in helicopter turbine engines. Zhao et al. (2022)
utilized transfer learning with an Extreme Learning Machine
to improve the prediction of engine torque using operational
data. Hu et al. (2024) employed adversarial transfer learning
with a Gaussian model to enhance fault diagnosis under
varying conditions. However, these AI-based approaches
present several challenges. First, AI-based methods often fail
to provide information regarding the uncertainty of their
predictions. This lack of uncertainty information is
particularly problematic in critical systems such as aviation,
where operators need not only the prediction but also an
understanding of how reliable that prediction is. Without
knowledge of prediction uncertainty, operators may over-rely
on the AI model's output, even when it is potentially
inaccurate. For example, an AI model might predict that an
