5
ICI Bucharest © Copyright 2012-2020. All rights reserved
ISSN: 1220-1766 eISSN: 1841-429X
1. Introduction
Anomalous and faulty behaviors damage the
machines like engines, gearboxes and rotor
systems, causing security reduction, economic
losses and environmental damages because
of an increased cost of the maintenance and
processing time. The fault diagnosis is important
for anticipating the incipient fault and predicting
system failure in order to ensure the normal
operation of an engine.
It is unfeasible that human operators diagnose
anomalous faults without this causing them to
take wrong decisions, no matter if this happens
in a timely manner (Li-Juan et al., 2013; Zhang
et al. 2013).
Several researches have introduced the signal
recognition and fault diagnosis algorithms.
Statistical techniques such as Linear Discriminant
Analysis (LDA) (Zhao et al., 2014; Mjahed &
Proriol, 1989) were the rst to be exploited in this
domain. Recently, articial intelligent techniques,
such as Genetic Algorithms (GA) (Mor &
Gupta, 2014; Mjahed, 2006), Particle Swarm
Optimization (PSO) (Rini et al., 2011; Jena et al.,
2015), Articial Bee Colony Algorithm (Chen &
Xiao, 2019), Fuzzy Logic (Raj & Murali, 2013;
Xiao et al., 2013) and Articial Neural Networks
(Chandra et al., 2013; Devi & Kumar, 2014) have
been applied successfully to automatic detection
and to diagnosis. GA and PSO algorithms have
been eectively used to select the attributes of
interest (Karimova et al., 2004) and for pattern
detection and classication tasks (Hewahi, 2017;
Mjahed, 2010).
Characterizing signals with appropriate features
and their classication are the most signicant
issues for machine fault diagnosis (Londhe et al.,
2014; Maaref et al., 2018). To identify the faults
and achieve a better classication performance,
it is important that the features selected contain
necessary discriminative information. A number
of vibration intensity techniques have been used
to analyze engine front noise (Chomphan &
Kingrattanaset, 2014). In (Lajmi et al., 2017) a
fault diagnosis based on fuzzy Petri net interval
is designed.
As regards the fault diagnosis in the aeronautical
eld, several works have been elaborated. The
features for fault diagnosis and prognosis of
gearbox are proposed (Zhang et al., 2013). A
vibroacoustic technique for the fault diagnosis
in a gear transmission of a military helicopter
is presented by (Zieja et al., 2017). A genetic
algorithm-based neural network for diagnosing an
aircraft air compressor bearing is introduced by
Studies in Informatics and Control, 29(1) 5-15, March 2020
https://doi.org/10.24846/v29i1y202001
Helicopter Main Rotor Fault Diagnosis by Using
GA- and PSO- based Classiers
Soukaina MJAHED*, Salah EL HADAJ, Khadija BOUZAACHANE, Said RAGHAY
Faculty of Sciences and Technology, Department of Applied Mathematics and Computer Sciences, Cadi Ayyad
University, Marrakech, Morocco
mjahedsoukaina1@gmail.com (*Corresponding author)
Abstract: This paper presents an improvement in the recognition of faulty signals, encountered in the case of the Gazelle
helicopter’s main rotor, using Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) methods. The main focus
is on the distinction between faulty and healthy signals and then between the three subclasses of faulty signals, i.e. faulty
bearings, joints problem and mechanical loosening. This research work is divided into three parts. The rst part approaches
the two above-mentioned classes of signals at the same time, and, to this purpose, the Linear Discriminant Analysis (LDA),
Non Linear Discriminant Analysis (NLDA) and Back-propagation Neural Network (BPNN) are used. In the second and
third part of the paper, GA and PSO are employed for optimizing the hyperplanes and hypersurfaces which separate the
above-mentioned classes of signals, as well as the architecture and connection weights of a neural network (NN). Real data
are used, which correspond to the vibration signals measured during periodic technical inspections, and are characterized
by amplitudes and frequencies typical of the eight highest peaks of the Welch spectrum. The results obtained conrm the
validity of the above-mentioned approaches and comparable favorably with those of other multivariate methods. The GA- or
PSO-based neural networks` diagnosis can therefore be established for helicopter computers so that faults can be detected.
Keywords: Classication, Genetic algorithm, Particle swarm optimization, Discriminant analysis, Neural networks, Fault diagnosis.
