Advances in Engineering Software 119 (2018) 48–59

Contents lists available at ScienceDirect

Advances in Engineering Software

journal homepage: www.elsevier.com/locate/advengsoft

Research paper

An efficient approach for optimal sensor placement and damage T

identification in laminated composite structures
⁎
D. Dinh-Conga,c, H. Dang-Trungb,c, T. Nguyen-Thoib,c,
a
Division of Construction Computation, Institute for Computational Science, Ton Duc Thang University, Ho Chi Minh City, Vietnam
b
Division of Computational Mathematics and Engineering, Institute for Computational Science, Ton Duc Thang University, Ho Chi Minh City, Vietnam
c
Faculty of Civil Engineering, Ton Duc Thang University, Ho Chi Minh City, Vietnam

A R T I C L E I N F O A B S T R A C T

Keywords: This paper proposed an efficient approach for optimal sensor placement and damage identification in laminated
Optimal sensor placement composite structures. This approach first utilized a model reduction technique, namely iterated improved re-
Damage identification duced system (IIRS) method, to develop a reduced order model for optimal sensor placement (OSP), and then the
Laminated composite structures OSP strategy using Jaya algorithm is conducted by formulating and solving an optimization problem for finding
Jaya algorithm
the best sensor locations. The objective function of the optimization problem is defined based on the correlation
Model reduction
between the flexibility matrix obtained from an original finite element model and the corresponding one cal-
Limited sensors
culated from IIRS method. Next, the approach uses the measured incomplete modal data from optimized sensor
locations for detecting and assessing any stiffness reduction induced by damage. In order to do this, the damage
identification problem is formulated as an optimization problem where the damage extent of elements and the
modal flexibility change are taken as the continuous design variables and the objective function, respectively.
The Jaya algorithm is again adopted to solve the optimization problem for determining the actual damage sites
and extents. Numerical simulations of a three cross-ply (0°/90°/0°) beam and a four-layer (0°/90°/90°/0°) la-
minated composite plate are carried out to demonstrate the applicability and efficiency of the proposed ap-
proach.

1. Introduction structures can be formulated as an optimization problem, where the

location and degree of damage are found by minimizing the objective
With significantly increasing applications of composite materials in function which is commonly defined in terms of the difference between
mechanics, aerospace, marine, civil and many other industries, struc- the vibration data measured by modal testing and those calculated from
tural health monitoring (SHM) for composite materials have gained analytical model. For generating data from the analytical model used in
much attention from the scientific and engineering communities. A health monitoring studies, the damage can be simulated by either re-
reliable and effective damage diagnose method is extremely important ducing the stiffness of elements in damaged areas [5,11,12] or using
to ensure the conditions of integrity and safety of structures made of crack models [13–15] or delamination models [16–18]. As a result, the
composite materials. Over the last few decades, many works and studies traditional finite element (FE) analysis or isogeometric analysis [19–22]
have been focused on vibration-based global SHM techniques for sol- can be employed as a tool for damage diagnosis through the model
ving damage identification problems in composite structures. The basic updating process which typically requires an optimization algorithm. In
idea of these techniques is that the change of either modal parameters order to meet this requirement, several meta-heuristic optimization
(natural frequencies [1], mode shapes [2]) or their variations (fre- algorithms have been applied as intelligent searching techniques to deal
quency-response function [3], curvature mode shapes [4], flexibility with the problem. For example, Su et al. [16] used genetic algorithms
matrix [5], etc.) can be used as signals to detect and locate damages in (GAs) and artificial neural networks (ANNs) for quantitative assessment
the structures. For more detailed information on these techniques, the of delamination in glass fiber-reinforced epoxy (GF/EP) composite la-
reader can refer to good review articles [6,7]. Besides, several recent minates. In their study, the efficiency of the GA was compared with
studies related to flaw detection problems in smart composite structures ANNs in term of both the prediction precision and computational cost.
were presented in Refs. [8–10]. Qian et al. [17] proposed a hybrid optimization algorithm featuring
In essence, the problem of damage identification in composite cooperative particle swarm optimization (PSO) with simplex method

⁎
Corresponding author at: Division of Computational Mathematics and Engineering, Institute for Computational Science, Ton Duc Thang University, Ho Chi Minh City, Vietnam.
E-mail addresses: dinhcongdu@tdt.edu.vn (D. Dinh-Cong), dangtrunghau@tdt.edu.vn (H. Dang-Trung), nguyenthoitrung@tdt.edu.vn (T. Nguyen-Thoi).

https://doi.org/10.1016/j.advengsoft.2018.02.005
Received 12 December 2017; Received in revised form 6 February 2018; Accepted 18 February 2018
0965-9978/ © 2018 Elsevier Ltd. All rights reserved.
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

(SM) for identification of delamination in laminated composite beams. the optimal location for a given set of sensors. The correlation between
Vosoughi and Gerist [11] presented a hybrid method based on con- the flexibility matrix obtained from an original FE model and the cor-
tinuous genetic algorithm (CGA) and PSO for damage detection of la- responding one calculated from IIRS method is considered as the ob-
minated composite beams. In their work, nevertheless, the effect of jective function, where the sensor positions are defined to be the dis-
measurement noise on the accuracy of the hybrid method was not in- crete optimization variables. We also present a comparison between the
vestigated. More recently, a few studies have been focused on plate-like proposed objective function with an objective function based root mean
structures. Ashory et al. [23] used GA to solve the optimization problem square (RMS) of modal assurance criterion (MAC) [33,34] to validate
for damage location and intensity identification in composite plates. the superiority of the proposed objective function. The second one is to
Dinh-Cong et al. [12] introduced an efficient multi-stage optimization locate and quantify structural damage using incomplete modal data
procedure using a modified differential evolution algorithm (MS-MDE) obtained from optimized sensor locations. In this second part, the op-
for damage assessment in a laminated composite plate. timization-based damage identification problem is first formulated by
It is noted that the above-mentioned researches required full FE defining the damage extent of elements as the continuous design vari-
models to be able to determine accurately and completely the vibration ables and the modal flexibility change as the objective function, and
characteristics in composite structures, which restricts applicability to then solved by using again the Jaya algorithm. To investigate the ap-
real-world and large-scale implementations. Given that in practical si- plicability and efficiency of the proposed damage diagnosis approach,
tuations it is almost impossible to determine entire experimental modal two numerical examples including a three cross-ply (0°/90°/0°) beam
information corresponding to every node/ degree of freedom (DOF) of and a four-layer (0°/90°/90°/0°) laminated composite plate are con-
the FE structural model because the number of measurement sensors is ducted. In addition, the influence of noise in the measured incomplete
typically limited, especially for large-scale structures. Therefore, one of modal data on the accuracy of proposed approach is also examined in
the practical challenges to the problems of damage identification in the examples.
composite structures is the use of incomplete modal data instead of The remaining parts of the article are organized as follows. Initially,
complete modal data in calculation due to the limited number of the IIRS method for OSP is presented in Section 2. Then, Section 3
measurement sensors used in practice. However, there have not been generally provides the formulation of optimization-based damage de-
many reported studies in the literature for dealing with this challenge tection problem. In Section 4, the description of the Jaya algorithm for
[1,18]. discrete and continuous design variables is introduced. Section 5 shows
Generally, the more sensors are placed on a structure, the more the performance of the proposed approach through numerical ex-
information from measurement data can be obtained. However, due to amples. Finally, the concluding remarks are given in Section 6.
cost and practicality issues, there are usually only a small number of
sensors installed to a predefined set of possible locations. In fact, the 2. Formulation of optimal sensor placement problem
quality of these data, as well as the quality of damage prediction de-
pends much on the placement of sensors, and hence optimal sensor In this section, the mathematical formulation of optimal sensor
placement (OSP) is an important constituent in the SHM of composite placement (OSP) strategy is conducted by considering three following
structures. In the last two decades, many techniques have been pro- main points: (1) the OSP as an optimization problem; (2) the iterated
posed and developed to achieve OSP which can help collect the best improved reduced system (IIRS) method; and (3) the objective function
identification of structural characteristics. An overview of OSP techni- for OSP problem. The details of the three points are presented in the
ques was presented by Yi and Li [24]. Among them, combinatorial next three sub-sections.
optimization methods have been widely employed owing to its com-
putational efficiency for solving OSP problems of large-scale structures.
2.1. Optimal sensor placement as optimization problem
Recently, some meta-heuristic optimization algorithms such as im-
proved particle swarm optimization (IPSO) algorithm [25], niching
The main goal of OSP in SHM is to determine the optimal sensor
monkey algorithm (NMA) [26], firefly algorithm (FA) [27], artificial
layout that can collect as much information of structural dynamic
bee colony (ABC) algorithm [28] have been successfully applied to the
characteristics as possible. To achieve this goal, the OSP problem can be
OSP problems. Nevertheless, there are still several issues for further
formulated as a constrained optimization problem in which the sensor
improvement such as: (1) how to reduce significantly the computa-
positions are considered as the discrete design variables and the con-
tional cost of the optimization algorithms; (2) how to validate the op-
straint is typically a given limited number of sensors. The objective
timal sensor layout obtained from the OSP strategies via a damage
function, usually based on the dynamic characteristics of a structure,
detection technique; and (3) how to minimize the number of sensors
can be maximized or minimized to determine the optimal locations for
used for the problem of structural damage diagnosis.
a given limited set of sensors. Thus, the mathematical model of OSP
As an effort to fill in the above-mentioned research gaps, the current
problem can be defined by the following optimization equation.
paper hence proposes an efficient approach for optimal sensor place-
ment and damage identification in laminated composite structures. The min f (S), s ∈  +
main contributions of the paper can be addressed in three following s. t . g (S) = n,
aspects: (i) Propose a more effective OSP strategy for finding proper
Slb ≤ S ≤ Sub . (1)
sensor locations installed on laminated composite structures; (ii)
Conduct an optimization-based damage detection technique to validate where f is the objective function;S = (s1,s1, … , sn) is denoted as the
the optimal sensor layout obtained from the proposed OSP strategy for candidate sensor locations placed at nodes/ DOFs of the FE structural
structural damage detection, simultaneously showing its capacity in model; n is the given limited number of sensors; Slb and Sub represent
damage diagnosis and assessment by using the first several lower in- the vectors of lower and upper bound of S, respectively; and  + is the
complete modes; and (iii) Apply effectively the Jaya optimization al- set of positive integers.
gorithm for solving both OSP and damage diagnose problems without In the OSP problem, a numerical model is required to identify the
trapping into local optima. The present work has two main parts. The modal parameters of structural system such as natural frequency and
first one is to determine the optimal location of a given limited number mode shapes. Nevertheless, such a model will have many more nodes/
of sensors placed on a structure. For this purpose, we use a model re- DOFs, while the optimal potential sensor locations are only chosen from
duction technique, namely iterated improved reduced system (IIRS) a subset of the total nodes/ DOFs. Hence, we use a model reduction
method [29], to develop a reduced order model for OSP, and then Jaya technique, namely IIRS method [29], to eliminate those DOFs that do
algorithm as robust optimization tool [30–32] is adopted to determine not relate to the candidate sensor locations required. A brief description

49
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

of IIRS method for OSP is given in the next sub-section. 2.3. Objective function for OSP problem

2.2. Iterated IRS (IIRS) method The Modal Assurance Criterion (MAC) [37] is a good tool to check
the correlation between two mode vectors of equal order. The off-di-
Iterated IRS method employed in this present study is based on agonal elements in the MAC matrix express the consistency between
Friswell's method [29,35]. Here, we define the potential sensor loca- two modal vectors. Hence, it was recommended to use the lowest off-
tions as master DOFs denoted by m, and the eliminated locations as diagonal terms in the MAC matrix as a useful criterion to choose op-
slave DOFs denoted by s, with m + s = N (where N is total DOFs of timal sensor locations [26,28]. The MAC matrix can be constructed by
structural system discretized by FE model). For an un-damped system taking into consideration the candidate sensor positions as
with N DOFs, the generalized eigenvalue problem of a structural system (Φunre )T Φrej (S) 2
i
with the partitioned mass, stiffness matrices and mode shapes governed MACij = , (i, j = 1, 2, ...,nmod )
((Φi ) Φi )((Φrej (S))T Φrej (S))
unre T unre
(11)
by the master and slave DOFs can be written as follows
where Φunre
i is the ith columns of the mode shape matrix calculated from
⎡ Kmm Kms ⎤ ⎧ Φmm ⎫ = ⎡ Mmm Mms ⎤ ⎧ Φmm ⎫ Λ
⎢ Ksm Kss ⎥ ⎨ Φsm ⎬ ⎢ Msm Mss ⎥ ⎨ Φsm ⎬ mm full FE model; and Φire (S) is the ith columns of the mode shape matrix
⎣ ⎦⎩ ⎭ ⎣ ⎦⎩ ⎭ (2) calculated by the reduced order model, which is functions of the sensor
where K and M are, respectively, the global stiffness and mass matrices; location parameter S; and nmod is the number of identified mode
Φ is the mass-normalized eigenvectors; Λ is a diagonal matrix con- shapes. Note that each element of the MAC matrix ranges from 0 to 1.
taining corresponding eigenvalues λi (i = 1, 2, … , m); and the sub- When i = j, MAC ^ 1 if two mode shape matrices are consistent. For
scripts m and s denote the sizes of the master and slave DOFs, respec- the case of i ≠ j, a small value of the MAC off-diagonal element in-
tively. To eliminate the slave DOFs field, the second row of Eq. (2) is dicates the less correlation between corresponding mode shape vectors;
employed and rearranged to give otherwise, large off-diagonal values indicate that the two mode shape
matrices are fairly indistinguishable [33].
Φsm = K−ss1KsmΦmm + K−ss1(Msm Φmm + Mss Φsm)Λmm (3) Thus, the main goal of constructing the MAC matrix is to determine
By assuming a transformation matrix t between Φsm and Φmm, as sensor locations such that off-diagonal terms of the MAC matrix are as
small as possible. For this purpose, the root mean square (RMS) values
Φsm = tΦmm (4) of the off-diagonal elements have been employed to construct the ob-
and substituting Eq. (4) into Eq. (3) and rearranging, we obtain the jective function for OSP [25,33]. The equation of this function can be
following equation presented as follows
f (S) ≡ fRMS = 1 − RMS (12)
t = TG + K−ss1(Msm + Mss t)Φmm ΛmmΦ−mm
1
(5)

where TG = K−ss1Ksm is defined as the Guyan transformation matrix [36]. where the RMS is defined as
By using the transformation matrix t, the transformation between 1 nmod nmod
the master DOFs and the complete set of DOFs is represented by RMS =
nmod (nmod − 1)
∑i =1 ∑ j=1 (MACij )2 , (j ≠ i)
(13)
⎧ Φmm ⎫ = ⎡ Imm ⎤ Φ = T Φ With the same purpose, the MAC criterion can be utilized to estimate
⎨ Φ ⎢ t ⎥ mm IIRS mm
⎩ sm ⎬⎭ ⎣  ⎦ the correlation between the flexibility matrix obtained from an original
TIIRS (6) FE model and the corresponding one calculated from IIRS method.
Based on the idea, an objective function for OSP is proposed here by
where Imm is the unit matrix of size m × m. In addition, to get a specific using the diagonal terms of the flexibility matrix as follows
way for calculating the transformation matrix t or TIIRS , we conduct
nc
some following mathematics transformations: 1
Appling Eq. (6) into Eq. (2) and per multiplication by TTIIRS , we can
f (S) ≡ fMACFLEX = 1 −
nc
∑ MACFLEXk
k (14)
obtain the following reduced eigenvalue problem
In Eq. (14), we denote:
KRΦmm = MR Φmm ΛmmwithKR = TTIRS KTIRS , MR = TTIRS MTIRS (7)
(Funre T re
k ) F k (S)
2

From Eq. (7), we can get MACFLEX = , (k = 1, 2, ⋯, nc )

((Funre
k )T F unre)((F re (S))T F re (S))
k k k (15)
Φmm ΛmmΦ−mm
1
= M−R1KR (8) where
Substituting Eq. (8) into Eq. (5), the transformation matrix t can be nmod
rewritten as F* = ∑ Φi* (Λ*)−1 (Φi*)T
i=1 (16)
t = TG + K−ss1(Msm + Mss t) M−R1KR (9)
where Φi* and F* are the ith mode shape and the flexibility matrix,
As can be seen from Eq. (9), t is an implicit function and cannot be respectively; nc is the total number of columns in the flexibility matrix;
directly solved. Friswell et al. [29] proposed an IIRS technique to solve the superscripts unre and re denote the unreduced model and reduced
for Eqs. (7) and (9), in which t is updated through an iterative process model, respectively. For finding the optimal sensor placements, an
given by optimization technique namely Jaya algorithm is adopted to minimize
−1 this objective function. The brief review of the algorithm will be pre-
t k = TG + K−ss1(Msm + Mss t k − 1)(MkR− 1) KkR− 1 (10)
sented in Section 4.
where the superscript k denotes the kth iteration, with k ≥ 2. When
k = 1, t(1) = TG, which leads to Guyan's method; and k = 2, it is right 3. Optimization-based damage detection problem
the standard IRS method. Further, the convergence of the IIRS tech-
nique was proved in Ref. [35]. In optimization-based damage detection problem, the damage de-
Thereby, the modal parameters can be obtained by solving the re- tection process is generally accomplished by minimizing an objective
duced eigenvalue problem defined by MR and KR, and then they are function based on modal parameters. With this scheme, a correct choice
utilized in the OSP problem. of the objective function plays a critical role in the successful

50
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

Fig. 1. Node and element numbering of the discretized composite beam.

Table 1 In this research, the numerical results illustrated clearly its better per-
Sensor placements of the composite beam obtained by Jaya algorithm using two objective formance in comparison with other well-known optimization algo-
functions. rithms such as differential evolution (DE) and cuckoo search (CS). In
Objective Number of Sensor no. f
the present research, the Jaya algorithm is further extended to solve for
function sensors both the optimal sensor placement and damage detection problems, in
1 2 3 4 5 6 7 8 which the sensor locations and the degree of damage elements are
discrete and continuous design variables, respectively. The detail of
fRMS 6 2 4 7 12 15 16 – – 7.67 × 10 − 1
fMACFLEX 3 4 8 10 13 15 – – 3.38 × 10 − 2
Jaya algorithm for the problems with continuous or discrete variables is
fRMS 7 2 3 6 9 12 15 16 – 7.29 × 10 − 1 briefly described via four phases as follows.
fMACFLEX 3 4 7 9 11 14 15 – 5.33 × 10 − 5 Initially, an initial population consisting NP individuals is generated
fRMS 8 2 4 6 10 11 12 15 16 7.50 × 10 − 1 randomly in the search space. Each individual in NP is a vector com-
fMACFLEX 2 3 5 7 9 11 14 15 2.91 × 10 − 5
prising n design variables xj = (x1,x2, … , xn) and is initialized as

implementation of the process. Modal frequencies and mode shapes, the

if xj is a continuous variable
basic dynamic parameters, have been widely used for the problem. In
xj, i = x lj, i + rand [0, 1] × (x uj, i − x lj, i), (i = 1, 2, ...,NP; j = 1, 2, ...,n) (19)
practical applications, it is usually challenging to achieve the complete
measured modes of a structure from a given limited number of sensors else if xj is a discrete variable
and the high-order mode shapes are typically quite difficult to measure xj, i = x lj, i + round [rand [0, 1] × (x uj, i − x lj, i)], (i = 1, 2, ...,NP; j = 1, 2, ...,n)(20)

accurately. To cope with the real issue, the modal flexibility is con- end if
sidered a suitable parameter because it can be approximately con-
structed by using only the measured first few frequencies and mode where x uj and x jl denote the upper and lower bounds of x j , respectively;
rand[0, 1] is a uniformly distributed between 0 and 1; round is the
shapes. In addition, it was proved to be more sensitive to damage than
function to round real values of integer design variables; NP is the
either the natural frequency or the mode shape [38]. Due to these ad-
population size. It should be noted that for OSP problem in Eq. (20), the
vantages, an objective function based on modal flexibility is defined
sensor locations are selected from a candidate set of nodes of a model
herein by using the norm of the discrepancies between the flexibility
and a single sensor could only be placed on a single node.
matrix obtained from an experimental modal and the corresponding
Next, at any Gth iteration, assume that x j, i, G is the value of the jth
one calculated from a reduced analytical model, which can be for-
variable for the ith candidate during the Gth iteration, then this value is
mulated as ′
utilized to generate a vector x i, G as
F exp − F re (x) Fro
Γ(x) = , x = (x1, ...,x ne ) ∈ [0, 1]ne
F exp Fro (17)
if xj,i is a continuous variable
where the components of vector x are the degree of damage of ne ele-
x′j,i, G = xj,i, G + r1,j, G × (xj,best, G − |xj,i, G|) − r2,j, G × (xj,worst,
ments, and they are considered as continuous design variables; F exp is
the flexibility matrix obtained from the damaged FE model, which is G − |xj,i, G|) (21)
taken as the flexibility matrix from experiments by adding noise at the else if xj,i is a discrete variable
modal parameters, while F re is the flexibility matrix obtained from re- x′j,i, G = xj,i, G + round[r1,j, G × (xj,best, G − |xj,i, G|) − r2,j,
duced analytical model; and ‖ • ‖Fro presents the Frobenius norm of a G × (xj,worst, G − |xj,i, G|)] (22)
matrix. end if
To find a set of damage variables, a robust optimization solver is
applied for minimizing the objective function Γ(x), which is expressed where xj,worst, G and xj,best, G are the value of the jth variable for the
worst candidate and the best candidate, respectively; r1,j, G and r2,j, G are
as follows
the random numbers in the range [0, 1]; the term ′′r2,j, G × (xj,worst,
G − |xj,i, G|)′′ points out the tendency of the solution avoiding the worst
finding x = {x1, x2, ⋯, x ne}
MinimizeΓ(x) solution; and the term ′′r1,j, G × (xj,best, G − |xj,i, G|)′′ points out the
S. t . 0 ≤ x i ≤ 1, (i = 1, 2, ⋯, ne ) (18) tendency of the solution toward the best solution. The absolute value of
the candidate solution |xj,i, G| helps enhance the exploration ability of
the algorithm.
4. Jaya algorithm for discrete and continuous design variables Subsequently, the values of components x′j,i, G are checked to reflect
back to the allowable region if its values exceed the corresponding
Jaya algorithm, a novel population-based search method, recently upper and lower bounds. This operation is executed as
proposed by Venkata Rao [30]. This algorithm is quite simple and easy
to use with only common controlling parameters (population size and a u u
⎧ 2x j − x ′ j, i, G if x ′ j, i, G > x j
number of generations), which escape the burden of tuning of algo- ⎪ l
x ′j, i, G = 2x j − x ′ j, i, G if x ′ j, i, G < x jl
rithmic-specific control parameters [39]. It has recently been in- ⎨
troduced for dealing with different engineering optimization problems ⎪ x ′ j , i, G otherwise (23)
⎩
[31,32,40], and its results have shown the better performance than
those of other population-based algorithms. The main idea of this al- Finally, the vector x′i,G is compared to its counterpart target in-
gorithm is that it always tries to move toward the best optimum solu- dividual xi,G. If the vector x′i,G has lower functional value, it will survive
tion and to avoid worst solution. In the previous our work [40], the to the next generation, otherwise, the target vector xi,G will be saved in
Jaya algorithm is successfully applied in structural damage assessment. the population.

51
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

Table 2
The first five frequencies of the laminated composite beam calculated by various models.

Method Modal frequency (Hz)

1 2 3 4 5

Unreduced model (FOBT) [42] 19.051 – – – –

Unreduced model (present) 19.125 38.983 61.861 85.374 109.741
IIRS method (k = 2), using 6 optimal 19.125 38.985 62.472 86.191 127.987
sensor locations from fRMS
(0.00%) (0.01%) (0.99%) (0.96%) (16.63%)
IIRS method (k = 2), using 6 optimal 19.125 38.987 61.873 85.740 113.447
sensor locations from fMACFLEX
(0.00%) (0.01%) (0.02%) (0.43%) (3.38%)

Fig. 2. The MAC values of the first eight modes for the composite beam obtained by Jaya algorithm using two objective functions.

52
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

Fig. 3. Convergence process of the Jaya algorithm for OSP (8 sensor positions) of the composite beam using two objective functions: (a) fMACFLEX; (b) fRMS.

x′i, G if f (x′i, G) ≤ f (x i, G) 5.1.1. Optimal sensor placement

x i, G + 1 = ⎧ First, we investigate the capability of the present OSP strategy for
⎨
⎩ x i, G otherwise (24)
better data acquisition. For this purpose, we assume that the strategy is
where f is the cost function which is given as either Eq. (14) for con- performed by installing 6, 7 and 8 sensors on the composite beam, re-
tinuous variable or Eq. (17) for discrete variable. spectively. In addition, it is also assumed that only the translational
DOFs of the beam are considered for possible sensor installation.
Consequently, a total of 15 nodes is available for installing the sensors,
5. Numerical examples and the nodal number varies from 2 to 16. For conducting OSP of the
composite beam, the IIRS technique is applied to obtain a reduced
In this part, two numerical examples including a three cross-ply (0°/ model consisting the translational DOFs only. Two cost functions, as
90°/0°) beam and a four-layer (0°/90°/90°/0°) laminated composite provided in Eqs. (12) and (14), are employed to determine optimal
plate are presented to verify the applicability and efficiency of the locations for the given limited number of sensors. The first eight modes
proposed approach for structural damage assessment using incomplete are selected for calculating these functions.
modal data identified from optimized sensors placement. In the fol- The OSP results of the composite beam obtained from Jaya
lowing examples, damage is simulated by reducing the stiffness of se- algorithm using two different objective functions are summarized
nele in Table 1. The results show that when the objective functions
lected elements to a certain level, i.e. K = ∑e = 1 (1 − ae ) ke , where K and
e employed for OSP are different, the corresponding optimal sensor lay-
k are the global stiffness matrix of damage structures and the stiffness
matrix of the eth element, respectively; and ae represents the damage outs will have discrepancies. That means that the OSP problem may
ratio of the eth element. For each of these examples, three different depend on the different objective functions used in the optimization
damage scenarios are considered with and without the effect of mea- process.
surement noise on identifying single and multiple structural damage The capability of capturing the vibration behavior of the
sites. The added noise level in the measured modal data is ± 0.15% beam model with two sets of sensor nodes (6 measured points) from
noise in natural frequencies and ± 3% noise in mode shapes [5]. The two objective functions are reported in Table 2. The identified
common control parameters of the Jaya algorithm in all examples are modal frequencies results from IIRS method are compared with those
given by population size (NP) = 30, maximum integration = 900, calculated by the unreduced model in the second row. It is obvious that
random number (r) = [0, 1] and stop criterion = 10−6. Each optimi- the use of 6 measurement points from the function fMACFLEX help
zation problem of OSP in all examples is repeated 10 times, and the best identify modal frequencies with more accuracy than that from the
optimal solution with the minimum objective function and a good MAC function fRMS.
index is considered as the representative OSP result. For each damage Fig. 2 shows the MAC values between the modes identified from
scenario, the average and statistical results of 10 independent runs are different sets of sensors and the corresponding modes calculated from
reported in figures and tables, respectively. All the computations are all DOFs of the FE model. From the figure, it can be seen that: (1) the
carried out in MATLAB environment. number of identified modes will increase when more sensors are
properly added; (2) with 7 measurement points from the function
fMACFLEX, the modes from 1 to 8 are well identified, while with 7
5.1. A three cross-ply (0°/90°/0°) beam measurement points from the function fRMS, only the modes from 1 to 5
are identified. This indicates that the OSP strategy based on the func-
The first example is a cross-ply (0°/90°/0°) rectangular beam with tion fMACFLEX is found to be superior to that based on the function fRMS
clamped boundary conditions at both ends, as depicted in Fig. 1. The in determining the optimal sensor positions.
geometrical parameters of the beam are given by the length L = 0.2 m, The typical convergence process of the Jaya algorithm for OSP of
the width b = 0.02 m and the thickness t = 0.02 m. The physical ma-
terial properties are given by E1 = 40 N/m2, E2 = 1 N/m2,
Table 3
G12 = G13 = 0.6E2, G23 = 0.5E2, v12 = 0.25, and the thickness of each Three different damage scenarios in the laminated composite beam.
layer is t/3. The composite beam is discretized into 16 beam elements
using first-order shear deformation theory (FSDT) as presented in Ref. Scenario 1 2 3
[41], and thus each element has two nodes with three DOFs (two
Element no. 1 5 16 1 2 9
translational displacements in the horizontal and vertical directions, Damage ratio 0.30 0.20 0.30 0.20 0.3 0.30
and one rotation) per node.

53
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

Fig. 4. Comparison of obtained damage identification results from Jaya algorithms for the laminated composite beam without noise and with noise from various optimal sensor
placement: (a) case 1; (b) case 2; (c) case 3.

the composite beam is shown in Fig. 3, in which the mean and best 5.1.2. Damage detection using incomplete modal data obtained from
convergence lines of the two objective functions in 8 measured points’ sensors
case are depicted. It can be seen that the average fitness has the ten- Second, we focus on damage detection in the composite beam using
dency to steadily approach to best fitness with the increase of the incomplete modal data obtained from optimized sensor locations. In
number of generations, which shows a good stability of convergence. order to examine the effect of the proposed damage identification

Table 4
The statistical results of damage identification for three scenarios of the laminated composite beam without noise and with noise from various optimal sensor locations.

Scenario Noise level Actual Jaya algorithm, using 6 optimal sensor Jaya algorithm, using 7 optimal sensor Jaya algorithm, using 8 optimal sensor
location locations locations locations

Avg. value Std. dev. Avg. NSA Avg. value Std. dev. Avg. NSA Avg. value Std. dev. Avg. NSA

1 0% α1 0.2951 0.0004 5,346 – – – – – –

3% α1 0.2322 0.0930 7,155 0.2564 0.0414 6,450 0.2790 0.0238 5,877
2 0% α5 0.1992 0.0005 5,952 – – – – – –
α16 0.2736 0.0007 – – – –
3% α5 0.1629 0.1348 7,752 0.1582 0.0571 6,387 0.1951 0.0723 6,459
α16 0.2004 0.0815 0.2483 0.0483 0.2449 0.0647
3 0% α1 0.2012 0.0007 6,249 – – – – – –
α2 0.3063 0.0005 – – – –
α9 0.2993 0.0003 – – – –
3% α1 0.1296 0.0786 7,311 0.1391 0.0667 6,546 0.1876 0.0123 6,195
α2 0.3328 0.0625 0.3216 0.0678 0.3157 0.0223
α9 0.2793 0.0773 0.2816 0.0413 0.3033 0.0286

Avg. value = average value of damage ratio with respect to Γ; Std. dev. = standard deviation with respect to Γ; Avg. NSA = an average number of structural analyses.

54
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

Fig. 5. Convergence process of the Jaya algorithm for damage detection of the composite beam: (a) without noisy data from 6 measurement points; (b) with noisy data from 8
measurement points.

Fig. 6. (a) A four-layer (0°/90°/90°/0°) square composite plate; (b) element numbering of the plate.

Table 5 of noise, the proposed method can accurately detect the actual da-
Optimal sensor placements of the composite plate obtained by Jaya algorithm using the
maged locations with only 6 sensors placed on the composite beam.
objective function fMACFLEX.
However, in noise-contaminated condition and with 6 measurement
Number of sensors Sensor no. f points, the proposed method has some false alarms elements (element 2
in scenario 1; elements 12 and 15 in scenario 2) appeared in its pre-
1 2 3 4 5 6 7 8 9 10 dictions, whereas with 8 measurement points the proposed method can
10 7 10 17 19 27 30 31 36 40 45 3.19 × 10 − 9
successfully determine the actual site of damage with negligible false
alarms. Also from the figures, it is worth mentioning that the better
damage assessment results are achieved when the more sensors are
method with the different number of sensors (6, 7 and 8 sensors), three utilized on the structure.
damage scenarios are studied in this example. The details of damaged For further investigation, the statistical results of damage detection
elements and their damage ratios of these scenarios are summarized in including the average, standard deviation, and the average number of
Table 3. The first five natural frequencies and corresponding mode structural analyses for three scenarios of the composite beam without
shapes are used for approximating the objective function as in Eq. (17). and with noisy data from a different number of sensors are provided in
To determine the damage extent and its locations, Jaya algorithm is Table 4. The results show clearly that for all three scenarios, the stan-
adopted for solving the inverse problem. dard deviation of predicted results with noise-free data from 6 mea-
The average results of damage ratio of all elements obtained by the surement points is much lower than those with corresponding noise-
Jaya algorithm for scenario 1 to 3 are shown in Fig. 4(a), (b) and (c), corrupted data. However, the standard deviation of predicted results
respectively. It can be seen that when the measured modal data is free with noise-corrupted data can be decreased by adding more sensors.

Table 6
The first six frequencies of laminated composite plate calculated by IIRS method.

Method Modal frequency (Hz)

1 2 3 4 5 6

Unreduced model (present) 0.022 0.056 0.138 0.204 0.402 0.467

IIRS method (k = 2), using 10 optimal sensor locations 0.022 0.056 0.138 0.204 0.402 0.467
(0.00%) (0.00%) (0.00%) (0.40%) (0.00%) (0.00%)

55
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

1 Particularly, the average standard deviation over all scenarios with

noise-corrupted data from 6, 7 and 8 measurement points are, respec-
tively, 9.13%, 5.09% and 3.78%. In addition, the mean error between
0.8 obtained average damage ratios and the actual damage ratios for three
1 cases with noise-corrupted data from 6, 7 and 8 measurement points are
21.23%, 16.07% and 6.72%, respectively. These results demonstrate
0.6
that the proposed method can determine both the location and severity
MAC

0.5 of structural damage with high accuracy when at least 8 sensors are
0.4 placed on the structure. Moreover, we also note that the increase of
0 measurement points will lead to the decrease of the number of struc-
0 tural analyses.
2 14 0.2 The convergence histories of the Jaya algorithm for a sample run of
4 12
6 10 each damage scenario in the case without noise and with measurement
8 8
Mode 10 12 6 Mode noise are shown in Fig. 5. It can be found that the algorithm can con-
4 0 verge rapidly to the optimal solution after 60 generations.
14 2
0
Fig. 7. The MAC values of the first fifteen modes for the composite plate. 5.2. A four-layer (0°/90°/90°/0°) laminated composite plate

Next, we consider a cantilever-laminated composite plate (0°/90°/

90°/0°), which has the dimension Lx = 0.5 m × Ly =
1 m × t = 0.04 m. The thickness of each layer is t/4. The physical
x 10-7
2.5 material parameters (Mpa) for each layer are given by E1 = 40,
Jaya algorithm: Best value E2 = 1, G12 = G13 = 0.6E2, G23 = 0.5E2, ν12 = 0.25. The FE model of
Jaya algorithm: Mean value the composite plate (as shown in Fig. 6) is discretized into a mesh 4 × 8
2
or 32 four-node Reissner–Mindlin plate elements using first-order shear
deformation theory (FSDT) as presented in Ref. [43]. Thus, the simulated
Fitness value

1.5 model has a total of 45 nodes with five DOFs (three translations and two
rotations) per node.

1 5.2.1. Optimal sensor placement

In this example, we assume that the OSP problem in SHM is fol-
0.5 lowed by installing 10 sensors on the composite plate. The selection of
the number of sensors is dependent on experiences, engineering judg-
ment and/or trial and error method. As shown in Fig. 6(b), a total of 40
0 nodes in the FE model, exclude the constrained nodes, are considered as
0 20 40 60 80
Iteration the possible candidate set of sensor locations. The first ten modes of the
plate are used for calculating the objective function. With this model,
Fig. 8. Convergence process of the Jaya algorithm for OSP of the composite plate.
only the proposed function fMACFLEX is adopted to find optimal locations
for a given limited number of sensors.
Table 5 shows the optimal locations of sensors obtained using the
Table 7
Jaya algorithm with the proposed function fMACFLEX. As can be seen in
Three different damage scenarios in the laminated composite plate. Table 6, the reduced model from NSEMR-II method using 10 optimal
sensor locations provides the identified modal frequencies results with
Scenario 1 2 3 very high accuracy. Fig. 7 shows the MAC values between the first
Element no. 3 14 19 5 23 26
fifteen modes identified from 10 optimal sensor locations and the cor-
Damage ratio 0.25 0.25 0.40 0.30 0.30 0.40 responding modes calculated the original full FE model. It is shown that
the MAC off-diagonal values are close to zero and all the diagonal
elements are close to 1, which means a good correlation between

0.3 0.3
Damage ratio
Damage ratio

0.2 0.2
3 3

0.1 0.1

0 0
8 8
7 7
6 5 6 5
4 4
Y 3
2 1 2 1 Y 3
2 1 2 1
4 3X 4 3X

(a) (b)
Fig. 9. The damage identification results obtained from the Jaya algorithm for damage scenario 1 of the composite plate with 10 measurement points: (a) noise-free; (b) noise 3%.

56
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

0.4 0.4
19 19
0.3 0.3

Damage ratio
Damage ratio
0.2 14 0.2
14

0.1 0.1

0 0
8 8
7 7
6 6 5
5 4
4
Y 3
2 1 2 1 Y 3
2 2 1
4 3X 1 4 3X

(a) (b)
Fig. 10. The damage identification results obtained from the Jaya algorithm for damage scenario 2 of the composite plate with 10 measurement points: (a) noise-free; (b) noise 3%.

26 26
0.4 0.4

0.3 23 0.3
Damage ratio

Damage ratio
23
5
0.2 0.2 5

0.1 0.1

0 0
8 8
7 7
6 5 6
5
Y 4 3 2 1 Y 4 3
2 1 2 2 1
4 3X 1 4 3X

(a) (b)
Fig. 11. The obtained damage identification results obtained from the Jaya algorithm for damage scenario 3 of the composite plate with 10 measurement points: (a) noise-free; (b) noise
3%.

related vectors as expected for MAC criterion. The typical convergence

process of the Jaya algorithm for OSP of the composite plate is depicted Table 8
The statistical results of damage identification for three scenarios of the laminated
in Fig. 8.
composite beam without noise and with noise from 10 measurement points.

5.2.2. Damage detection using incomplete modal data obtained from Scenario Noise level Actual Jaya algorithm, using 10 optimal sensor
sensors location locations
After obtaining the sensor layout for the composite plate, the modal
Avg. value Std. dev. Avg. NSA
data measured from these sensors can be used for damage assessment
by the proposed damage diagnosis method. In this example, three dif- 1 0% α3 0.2498 0.0001 16,212
ferent damage scenarios consisting of one, two and three damaged 3% α3 0.2612 0.0639 17,115
elements are considered. The details of three damage scenarios are 2 0% α14 0.2492 0.0295 15,675
listed in Table 7. The first six natural frequencies and corresponding α19 0.3983 0.0258
mode shapes are employed for calculating the objective function as in 3% α14 0.2191 0.0258 17,619
Eq. (17). α19 0.3842 0.0266

Figs. 9, 10 and 11 present the average results of damage ratio of all 3 0% α5 0.2997 0.0002 16,782
α23 0.2976 0.0011
elements obtained by the Jaya algorithm for scenario 1, 2 and 3, re-
α26 0.4000 0.0002
spectively. The graphical results show that for three scenarios, the
3% α5 0.2637 0.0463 16,815
proposed method can accurately locate all locations of damage in the 0.2683 0.0471
α23
plate even under noisy and incomplete modal data from 10 measure- 0.3862 0.0415
α26
ment points. Further, the details of statistical results including the
average, standard deviation, and the average number of structural Avg. value = average value of damage ratio with respect to Γ; Std. dev. = standard de-
analyses of the Jaya algorithm for three scenarios are reported in viation with respect to Γ; Avg. NSA = an average number of structural analyses.
Table 8. The results show that the extent of damaged elements is de-
termined by the proposed method with satisfactory accuracy. Particu- quantifying the structural damage in spite of incomplete modal data
larly, all scenarios without noise, the mean error between obtained and measurement inaccuracy.
average damage ratios and the actual damage ratios are 0.30%, while Fig. 12 illustrates the convergence histories of the Jaya algorithm
those for all scenarios with measurement noise are 9.69%. Also, the for all damage scenarios of the composite plate with and without noise
standard deviation of all predicted results is less than 5%. These results in modal data. From the figure, it is easy to see that the algorithm
demonstrate the good ability of the proposed method in detecting and converges to the minimum cost after 310 generations.

57
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

Fig. 12. Convergence process of the Jaya algorithm for damage detection of the composite plate: (a) without noisy data from 10 measurement points; (b) with noisy data from 10
measurement points.

Fig. 13. The damage identification results obtained from the Jaya algorithm for the composite plate with 10 measurement points and noise 5%: (a) scenario 1; (b) scenario 3.

deviation of the predicted results are much larger than the previous
Table 9
results presented in Table 7. From these results, it is reasonable to
The statistical results of damage identification for the laminated composite beam with
noise from 10 measurement points and noise 5%. conclude that with this case study, the increase of measurement noise
leads to the decrease of the precision of identification. Note also that
Scenario Noise level Actual Jaya algorithm, using 10 optimal sensor when the noise level increases, the damage detection process needs
location locations more structural analyses for reaching the optimal solution.
Avg. value Std. dev. Avg. NSA
6. Conclusions
1 5% α3 0.1835 0.1548 19,236
3 5% α5 0.1315 0.0911 19,719 The article proposes an efficient approach for optimal sensor place-
α23 0.2170 0.2890 ment and damage identification in laminated composite structures. The
α26 0.2550 0.1531 sensor placement strategy is first utilized to determine the optimal sensor
layout of a given limited number of sensors placed on a structure, while
Avg. value = average value of damage ratio with respect to Γ; Std. dev = standard de-
damage identification method is proposed to locate and quantify struc-
viation with respect to Γ; Avg. NSA = an average number of structural analyses.
tural damage using incomplete modal data collected from optimized
sensor locations. The problems of optimal sensor location and damage
In this example, to further investigate the effect of measurement
identification are formulated as discrete and continuous optimization
noise on the performance of the proposed algorithm, the increase of
problems, respectively, and then solved by the Jaya algorithm. The ap-
noise level in measured modal data is considered here by adding ± 1%
plicability and efficiency of the proposed approach are verified through
instead of ± 0.15% noise in natural frequencies and ± 5% instead
two numerical examples comprising a three cross-ply (0°/90°/0°) beam
of ± 3% noise in mode shapes. Damage scenarios are assumed to be
and a four-layer (0°/90°/90°/0°) laminated composite plate. In addition,
same as scenarios 1 and 3 in Table 7. For both the damage scenarios,
various challenges such as the presence of noise in measured incomplete
the first six vibration modes are utilized for damage assessment in the
modal data and the impacts of the number of measurement locations are
plate. The final identified results using measured modal data from ten
also investigated. Based on the numerical results presented in this article,
measurement points and the noise level are reported in Fig. 13 and
some conclusions are summarized as follows
Table 9. Fig. 13 shows that although the proposed damage detection
method is still able to determine the damaged sites correctly, some false
(1) The Jaya algorithm is found to be an efficient optimization tool for
damage elements appear especially in triple damage case. Besides, the
solving both the discrete and continuous optimization problems.
statistical results in Table 9 indicate that the mean error and standard

58
D. Dinh-Cong et al. Advances in Engineering Software 119 (2018) 48–59

(2) In comparison with the objective function fRMS, the proposed ob- fibre Bragg grating sensors. Smart Mater Struct 2005;14:1541–53. http://dx.doi.
jective function fMACFLEX can capture better vibration modes and org/10.1088/0964-1726/14/6/047.
[17] Qian X, Cao M, Su Z, Chen J. A hybrid particle swarm optimization (PSO)-simplex
ensures a good agreement between the identified and calculated algorithm for damage identification of delaminated beams. Math Probl Eng
mode shapes. In addition, since more sensors are properly added in 2012;2012:1–12. http://dx.doi.org/10.1155/2012/607418.
the structure, the number of identified modes will be increased. [18] Rao ARM, Lakshmi K, Kumar SK. Detection of delamination in laminated compo-
sites with limited measurements combining PCA and dynamic QPSO. Adv Eng Softw
(3) When the best sensor locations are installed in the structure, the 2015;86:85–106. http://dx.doi.org/10.1016/j.advengsoft.2015.04.005.
best modal information can be obtained. This will ensure the high [19] Nguyen-Thanh N, Zhou K, Zhuang X, Areias P, Nguyen-Xuan H, Bazilevs Y, et al.
precision of the collected data which is regarded as optimal input Isogeometric analysis of large-deformation thin shells using RHT-splines for mul-
tiple-patch coupling. Comput Methods Appl Mech Eng 2017;316:1157–78. http://
data for damage diagnostics in the following step. dx.doi.org/10.1016/j.cma.2016.12.002.
(4) The proposed damage identification method can successfully de- [20] Kruse R, Nguyen-Thanh N, De Lorenzis L, Hughes TJR. Isogeometric collocation for
termine both the site and severity of damages in composite struc- large deformation elasticity and frictional contact problems. Comput Methods Appl
Mech Eng 2015;296:73–112. http://dx.doi.org/10.1016/j.cma.2015.07.022.
tures with noise in measured incomplete modal data.
[21] Chen L, Nguyen-Thanh N, Nguyen-Xuan H, Rabczuk T, Bordas SPA, Limbert G.
(5) It is worth mentioning that the proposed approach, which combines Explicit finite deformation analysis of isogeometric membranes. Comput Methods
an optimal sensor placement strategy and a damage identification Appl Mech Eng 2014;277:104–30. http://dx.doi.org/10.1016/j.cma.2014.04.015.
method, can be considered as a promising tool for practical SHM [22] Banh-Thien T, Dang-Trung H, Le-Anh L, Ho-Huu V, Nguyen-Thoi T. Buckling ana-
lysis of non-uniform thickness nanoplates in an elastic medium using the isogeo-
applications. metric analysis. Compos Struct 2017;162:182–93. http://dx.doi.org/10.1016/j.
compstruct.2016.11.092.
Acknowledgements [23] Ashory M-R, Ghasemi-Ghalebahman A, Kokabi M-J. An efficient modal strain en-
ergy-based damage detection for laminated composite plates. Adv Compos Mater
2017;3046:1–16. http://dx.doi.org/10.1080/09243046.2017.1301069.
This research is funded by Vietnam National Foundation for Science [24] Yi T-H, Li H-N. Methodology developments in sensor placement for health mon-
and Technology Development (NAFOSTED) under grant number itoring of civil infrastructures. Int J Distrib Sens Networks 2012;2012:1–11. http://
dx.doi.org/10.1155/2012/612726.
107.02-2017.08. [25] Zhang X, Li J, Xing J, Wang P, Yang Q, Wang R, et al. Optimal sensor placement for
latticed shell structure based on an improved particle swarm optimization algo-
References rithm. Math Probl Eng 2014;2014:1–12. http://dx.doi.org/10.1155/2014/743904.
[26] Yi T-H, Li H-N, Gu M, Zhang X-D. Sensor placement optimization in structural
health monitoring using niching monkey algorithm. Int J Struct Stab Dyn
[1] de Medeiros R, Sartorato M, Vandepitte D, Tita V. A comparative assessment of 2014;14:1440012http://dx.doi.org/10.1142/S0219455414400124.
different frequency based damage detection in unidirectional composite plates using [27] Zhou G, Yi T, Zhang H, Li H. A comparative study of genetic and firefly algorithms
MFC sensors. J Sound Vib 2016;383:171–90. http://dx.doi.org/10.1016/j.jsv.2016. for sensor placement in structural health monitoring. Shock Vib 2015;2015:1–10.
06.047. http://dx.doi.org/10.1155/2015/518692.
[2] Araújo dos Santos JV, Lopes HMR, Vaz M, Mota Soares CM, Mota Soares CA, de [28] Sun H, Büyüköztürk O. Optimal sensor placement in structural health monitoring
Freitas MJM. Damage localization in laminated composite plates using mode shapes using discrete optimization. Smart Mater Struct 2015;24:125034http://dx.doi.org/
measured by pulsed TV holography. Compos Struct 2006;76:272–81. http://dx.doi. 10.1088/0964-1726/24/12/125034.
org/10.1016/j.compstruct.2006.06.034. [29] Friswell MI, Garvey SD, Penny JET. Model reduction using dynamic and iterated
[3] Kessler SS, Cesnik CES. Damage detection composite materials using frequency IRS techniques. J Sound Vib 1995;186:311–23. http://dx.doi.org/10.1006/jsvi.
response methods. Compos Part B Eng 2008;33:1–19http://dx.doi.org/10.1016/ 1995.0451.
S1359-8368(01)00050-6. [30] Venkata Rao R. Jaya: a simple and new optimization algorithm for solving con-
[4] Moreno-García P, Araújo dos Santos JV, Lopes H. A new technique to optimize the strained and unconstrained optimization problems. Int J Ind Eng Comput
use of mode shape derivatives to localize damage in laminated composite plates. 2016;7:19–34. http://dx.doi.org/10.5267/j.ijiec.2015.8.004.
Compos Struct 2014;108:548–54. http://dx.doi.org/10.1016/j.compstruct.2013.09. [31] Venkata Rao R, Saroj A. A self-adaptive multi-population based Jaya algorithm for
050. engineering optimization. Swarm Evol Comput 2017. http://dx.doi.org/10.1016/j.
[5] Dinh-Cong D, Vo-Duy T, Nguyen-Minh N, Ho-Huu V, Nguyen-Thoi T. A two-stage swevo.2017.04.008.
assessment method using damage locating vector method and differential evolution [32] Singh SP, Prakash T, Singh VP, Babu MG. Analytic hierarchy process based auto-
algorithm for damage identification of cross-ply laminated composite beams. Adv matic generation control of multi-area interconnected power system using Jaya
Struct Eng 2017;20:1807–27. http://dx.doi.org/10.1177/1369433217695620. algorithm. Eng Appl Artif Intell 2017;60:35–44. http://dx.doi.org/10.1016/j.
[6] Montalvao D. A review of vibration-based structural health monitoring with special engappai.2017.01.008.
emphasis on composite materials. Shock Vib Dig 2006;38:295–324. http://dx.doi. [33] He C, Xing J, Li J, Yang Q, Wang R, Zhang X. A combined optimal sensor placement
org/10.1177/0583102406065898. strategy for the structural health monitoring of bridge structures. Int J Distrib Sens
[7] Fan W, Qiao P. Vibration-based damage identification methods: a review and Networks 2013;9:820694http://dx.doi.org/10.1155/2013/820694.
comparative study. Struct Health Monit 2011;10:83–111. http://dx.doi.org/10. [34] Muthuraman U, Hashita MMS, Sakthieswaran N, Suresh P, Kumar MR,
1177/1475921710365419. Sivashanmugam P. An approach for damage identification and optimal sensor
[8] Nanthakumar SS, Lahmer T, Zhuang X, Zi G, Rabczuk T. Detection of material in- placement in structural health monitoring by genetic algorithm technique. Circuits
terfaces using a regularized level set method in piezoelectric structures. Inverse Syst 2016;7:814–23. http://dx.doi.org/10.4236/cs.2016.76070.
Probl Sci Eng 2016;24:153–76. http://dx.doi.org/10.1080/17415977.2015. [35] Friswell MI, Garvey SD, Penny JET. The convergence of the iterated IRS method. J
1017485. Sound Vib 1998;211:123–32. http://dx.doi.org/10.1006/jsvi.1997.1368.
[9] Nanthakumar SS, Lahmer T, Rabczuk T. Detection of multiple flaws in piezoelectric [36] Guyan RJ. Reduction of stiffness and mass matrices. AIAA J 1965;3. http://dx.doi.
structures using XFEM and level sets. Comput Methods Appl Mech Eng org/10.2514/3.2874. 380–380.
2014;275:98–112. http://dx.doi.org/10.1016/j.cma.2014.03.001. [37] Allemang RJ, Brown DL. A correlation coefficient for modal vector analysis. Proc.
[10] Nanthakumar SS, Lahmer T, Rabczuk T. Detection of flaws in piezoelectric struc- 1st Int. modal Anal. Conf. 1. SEM; 1982. p. 110–6.
tures using extended FEM. Int J Numer Methods Eng 2013;96:373–89. http://dx. [38] Jun Zhao D. Sensitivity study for vibrational parameters used in damage detection.
doi.org/10.1002/nme.4565. J Struct Eng 1999;125:410–6.
[11] Vosoughi AR, Gerist S. New hybrid FE-PSO-CGAs sensitivity base technique for [39] Rao RV, Saroj A. Constrained economic optimization of shell-and-tube heat ex-
damage detection of laminated composite beams. Compos Struct 2014;118:68–73. changers using elitist-Jaya algorithm. Energy 2017;128:785–800. http://dx.doi.
http://dx.doi.org/10.1016/j.compstruct.2014.07.012. org/10.1016/j.energy.2017.04.059.
[12] Dinh-Cong D, Vo-Duy T, Ho-Huu V, Dang-Trung H, Nguyen-Thoi T. An efficient [40] Du D-C, Vinh H-H, Trung V-D, Hong Quyen N-T, Trung N-T. Efficiency of Jaya
multi-stage optimization approach for damage detection in plate structures. Adv algorithm for solving the optimization-based structural damage identification pro-
Eng Softw 2017;112:76–87. http://dx.doi.org/10.1016/j.advengsoft.2017.06.015. blem based on a hybrid objective function. Eng Optim 2017;273:1–19. http://dx.
[13] Nguyen-Thanh N, Zhou K. Extended isogeometric analysis based on PHT-splines for doi.org/10.1080/0305215X.2017.1367392.
crack propagation near inclusions. Int J Numer Methods Eng 2017;112:1777–800. [41] Vo-Duy T, Nguyen-Minh N, Dang-Trung H, Tran-Viet A, Nguyen-Thoi T. Damage
http://dx.doi.org/10.1002/nme.5581. assessment of laminated composite beam structures using damage locating vector
[14] Nguyen-Thanh N, Huang J, Zhou K. An isogeometric-meshfree coupling approach (DLV) method. J Contemp Phys 2015;50:457–65. http://dx.doi.org/10.1007/
for analysis of cracks. Int J Numer Methods Eng 2017:1–31. http://dx.doi.org/10. s11709-015-0303-0.
1002/nme.5713. [42] Khdeir AA, Reddy JN. Free vibration of cross-ply laminated beams with arbitrary
[15] Tan P, Nguyen-Thanh N, Zhou K. Extended isogeometric analysis based on Bézier boundary conditions. Int J Eng Sci 1994;32:1971–80. http://dx.doi.org/10.1016/
extraction for an FGM plate by using the two-variable refined plate theory. Theor 0020-7225(94)90093-0.
Appl Fract Mech 2017;89:127–38. http://dx.doi.org/10.1016/j.tafmec.2017.02. [43] Reddy J. Mechanics of laminated composite plates and shells: theory and analysis.
002. CRC Press; 1997.
[16] Su Z, Ling H-Y, Zhou L-M, Lau K-T, Ye L. Efficiency of genetic algorithms and ar-
tificial neural networks for evaluating delamination in composite structures using

59