FAULT DETECTION AND RECONFIGURATION APPLIED TO A
HELICOPTER SWASHPLATE ACTUATOR
Thomas Rakotomamonjy
∗
, Dominique Tristrant
∗
, Adrien Cabut
∗∗
∗
ONERA — The French Aerospace Lab,
∗∗
Formerly intern at ONERA
thomas.rakotomamonjy@onera.fr;dominique.tristrant@onera.fr;adrien.cabut@gmail.com
Keywords: helicopter, fault diagnosis, reconfigurable control, swashplate
Abstract
The objective of this paper is to evaluate the fea-
sibility and the performance of some fault de-
tection and reconfiguration techniques applied to
the main rotor swashplate of a helicopter. The
swashplate displacements and rotations allow to
change the rotor blades pitch angle and is thus re-
sponsible for controlling the magnitude and ori-
entation of the rotor lift and thrust forces: a fail-
ure affecting the actuators moving the swashplate
could have very serious consequences upon flight
safety.
A simplified model of the kinematics of a
swashplate, able to take into account the actua-
tor displacements, has been developed and im-
plemented into a numerical aero-mechanics heli-
copter simulation code. Then, a method based on
Principal Analysis Component has been found to
give very good results for the detection of actua-
tor failures such as partial or total jamming. Fi-
nally a new method for the reconfiguration of the
control law in the presence of the fault has been
developed, which is based upon a LQ criterion,
and has also been successfully tested in simula-
tion.
Main notations
X
T
= transposee of X
˙
X = derivative of X w.r.t. time
λ
i
, i ∈ {1; 2;3} = length of i
th
actuator
θ
p
= longtudinal attitude of
upper plate
φ
p
= lateral attitude of upper plate
Λ
p
= distance between respective
centers of upper and lower plates
u, v, w = helicopter body translational
velocities
p, q, r = helicopter body angular velocities
φ, θ = helicopter body bank and
pitch angles
A, B = state and input matrices
of the linear helicopter model
X, Y , U = state, output and input vector
1 Introduction
Fault Detection, Identification and Reconfigura-
tion (or FDIR for short) is a major concern in
aeronautics. The objective is to assess, isolate
and counteract any minor or major failure which
might have an incidence upon the flight perfor-
mances and/or security [2]. This thematic has
been widely applied for fixed-wing aircraft: due
to the configuration symmetry and the redun-
dancy of the control surfaces, a new flight equi-
librium point is usually achievable in case of an
actuator failure on an airplane, by using the func-
tional ones [3, 11].
But helicopters did not receive as much at-
tention in the literature: most applications of
FDIR techniques for rotary-wing aircraft deal
with small, multirotor UAVs or with tandem
helicopters[8], both able to use control redundan-
cies to overcome the loss of a control element.
However, for a standard helicopter with one main
rotor, the consequences of an actuator failure can
be much more critical, since it is a naturally un-
1
