applied

sciences
Article
Reliability-Based Design Optimization of Structures Using
Complex-Step Approximation with Sensitivity Analysis
Junho Chun

School of Architecture, Syracuse University, Syracuse, NY 13203, USA; [email protected]

Abstract: Structural optimization aims to achieve a structural design that provides the best perfor-
mance while satisfying the given design constraints. When uncertainties in design and conditions
are taken into account, reliability-based design optimization (RBDO) is adopted to identify solutions
with acceptable failure probabilities. This paper outlines a method for sensitivity analysis, reliability
assessment, and RBDO for structures. Complex-step (CS) approximation and the first-order reliability
method (FORM) are unified in the sensitivity analysis of a probabilistic constraint, which streamlines
the setup of optimization problems and enhances their implementation in RBDO. Complex-step
approximation utilizes an imaginary number as a step size to compute the first derivative without
subtractive cancellations in the formula, which have been observed to significantly affect the accuracy
of calculations in finite difference methods. Thus, the proposed method can select a very small
step size for the first derivative to minimize truncation errors, while achieving accuracy within the
machine precision. This approach integrates complex-step approximation into the FORM to compute
sensitivity and assess reliability. The proposed method of RBDO is tested on structural optimization
problems across a range of statistical variations, demonstrating that performance benefits can be
achieved while satisfying precise probabilistic constraints.







Citation: Chun, J. Reliability-Based

Keywords: reliability-based design optimization; sensitivity analysis; complex-step approximation
Design Optimization of Structures
Using Complex-Step Approximation
with Sensitivity Analysis. Appl. Sci.
2021, 11, 4708. https://doi.org/ 1. Introduction
10.3390/app11104708 Structural systems operate in variable and uncertain environments that impact build-
ing performance and integrity. While maintaining a satisfactory level of safety and reliabil-
Academic Editor: Sang-Hyo Kim ity, structural systems are expected to satisfy their design criteria, such as performance and
serviceability. Uncertainties and tolerances are commonly present and even anticipated
Received: 5 April 2021
during the design process and construction. The prediction of building response to these
Accepted: 15 May 2021
events has a crucial role in the design and optimization of structures. In the optimal design
Published: 20 May 2021
of structural systems, variations in the material, loads, and parameters of a structure are
random variables with a certain probability distribution rather than fixed coefficients. With
Publisher’s Note: MDPI stays neutral
the existence of such uncertainties, the problem of reliability analysis and its reflection
with regard to jurisdictional claims in
must be addressed during the design process of structures. The inclusion of reliability
published maps and institutional affil-
analysis within the design process could drastically reduce the total cost of development
iations.
and the risk of failure.
Structural design optimization can be viewed as a process that methodically searches
for improved engineering solutions and designs while considering requirements and
constraints that exist in the field. In deterministic design optimization (DDO) [1], the
Copyright: © 2021 by the author.
uncertainties of a structural system, such as material properties, forces, models, and sizes,
Licensee MDPI, Basel, Switzerland.
are taken into account subjectively and indirectly, by means of safety factors or load com-
This article is an open access article
binations with coefficients specified in the design codes. The procedures of deterministic
distributed under the terms and
design optimization consider the effects of uncertainties in a qualitative way, without incor-
conditions of the Creative Commons
poration of quantitative characterization of the probabilistic nature of system and design.
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
Non-deterministic optimization has emerged as an alternative form to accurately model
4.0/).
the safety-under-uncertainty aspect of the optimization problem. Methods for considering

Appl. Sci. 2021, 11, 4708. https://doi.org/10.3390/app11104708 https://www.mdpi.com/journal/applsci

Appl. Sci. 2021, 11, 4708 2 of 20

uncertainty in the optimal design process can be categorized into two groups. Reliability-
based design optimization (RBDO) aims to meet a specific target failure probability for
objectives and/or constraint functions that include uncertainty parameters. Robust design
optimization (RDO), on the other hand, minimizes the sensitivity of the objectives and
constraints to the uncertainty design parameters.
The growing interest in the identification of feasible, reliable, and robust solutions in
structural design optimization has led to significant advances in theoretical and numerical
developments, such as simulation and analysis approaches and optimization algorithms.
Sandren and Cameron [2] proposed a robust design optimization method for structures
with considerations of parameter variation and sensitivity to variation by using genetic and
non-linear programming algorithms. Marano et al. [3] developed a robust optimal design
criterion for a tuned mass damper device under random vibrations. The robustness of
optimal solutions was obtained by solving a multi-objective optimization problem in which
the mean and standard deviations of the deterministic objection were minimized. A robust
design of a vibration absorber under uncertainties in mass and stiffness of the dynamic
system was utilized to demonstrate the robust design approach by Zang et al. [4]. Robust
design optimization was performed by a direct first-order perturbation method based on a
Taylor expansion. The development of nonlinear programming-based RDO [5–7] in various
fields with diverse applications and a thorough review of RDO methods are illustrated in
the literature, e.g., by Messac and Ismail-Yahaya [8], as well as Zhang et al. [4].
In RBDO, a design at each iteration step is checked to determine whether it satisfies
probabilistic constraints by reliability analysis, which requires repetitive computational
efforts [9]. To assess the probabilistic constraints in RBDO, various methods have been
developed [10,11]. The reliability index approach (RIA) [9] and its inverse, the performance
measure approach (PMA) [10,12], utilize the first-order reliability method (FORM) [13] for
reliability analysis. RBDO is performed using a nested approach, such that each iteration in
the design optimization process involves a sub-loop of iterations for reliability assessment
at the current stage of design. This is often called a “double-loop” approach because each
step of the optimization process involves another iteration for the reliability analysis [13].
As a result of the repeated iteration process, the computational cost associated with the
utilization of the double-loop approach can prohibitively be expensive. To improve the
computational efficiency of reliability analysis and optimization procedure, many RBDO
approaches, such as decoupling [14,15] and single-loop (SL) methods [16–20], have been
proposed. The decoupling method decouples the RBDO procedure by constructing a
sequence of cycles with deterministic optimization followed by the reliability analysis of
the converged solution in the feasible domain. The SL approach is based on the Karush–
Kuhn–Tucker (KKT) optimality conditions to approximate the solution of the sub-loop
optimization. As a result, the sub-loop for reliability analysis is replaced by a deterministic
constraint, which transforms a double-loop RBDO problem into an equivalent single-
loop optimization problem. RBDO methods have been applied to many fields, including
structural engineering [17,21], mechanical engineering [22,23], aerospace engineering [24],
and material science [25]. A comprehensive review on RBDO can be found in the work of
Song et al. [26], Frangopol and Maute [27], and Tsompanakis et al. [28].
In DDO, RDO, and RBDO, sensitivity analysis of probabilistic constraints is crucial to
the utilization of efficient optimization algorithms. The dependence on reliability with re-
spect to the design inputs or values can be evaluated using sensitivity analysis. There have
been multiple approaches in sensitivity analysis, such as Monte Carlo Simulation-based
approaches [29,30], and simulation-based methods [31,32]. However, these approaches
generally require high computational power that might restrict their applicability, espe-
cially when sensitivity analysis needs to be performed with respect to many parameters.
The difficulty of sensitivity analysis increases when the underlying functions and models
are complex, or the analytical derivation of gradients is not feasible due to the implicit
system. In this paper, the sensitivity analysis of failure probability, a probabilistic con-
straint of RBDO, is proposed using complex-step (CS) approximation integrated with
Appl. Sci. 2021, 11, 4708 3 of 20

the FORM. Li [33] showed sensitivity analysis in the FORM-based RBDO using the CS.
This paper provides details on the general integration of CS in the FORM and discusses
uniform applicability to structural RBDO. The sensitivity analysis is performed using a
very small complex value, such that the outcome is not significantly affected by a selec-
tion of steps, unlike the finite difference method. In addition, the proposed method can
be easily implemented in RBDO and is free from round-off and subtractive cancellation
errors when compared with the finite difference method. Furthermore, sensitivity analysis
and assessment of the failure probability can be simultaneously performed using the CS
approximation simply by taking real and imaginary parts of numerical results so that
computational efficiency in analysis and optimization increases. This paper presents the
RBDO method utilizing the FORM and presents a sensitivity analysis of probabilistic
constraints with respect to design and random variable parameters to use a gradient-based
optimization algorithm.

2. Reliability Analysis and Reliability-Based Design Optimization

2.1. Reliability Analysis Using the First-Order Reliability Method
The time-invariant reliability problem can be characterized by an n-vector of random
variables x = ( x1 , x2 , x3 , . . . , xn )T and the subset Ω of their outcome space, which defines
the failure events of the limit-state function g(x). The failure probability Pf is given by the
n-fold integral
Pf = P(x ∈ {Ω ≡ g(x) ≤ 0}) = g(x)≤0 k(x)dx
R
(1)
where k(x) is the joint probability density function (PDF) of x. By transforming the random
variables into the standard normal space (u = T(x)), the failure probability integral can be
written as R
Pf = G(u)≤0 ϕn (u)du (2)
where G (u) is the limit-state function transformed from the original space into the standard
normal space and ϕn (·) is the n-variate standard normal density function for the random
vector. The component reliability problem utilizes a reliability method to compute the
failure probability. The FORM linearizes the limit-state function G (u) at a point u*, which
is obtained by solving a constraint optimization problem as

arg min {kuk | G (u) = 0 } (3)

u∗

The point u* indicates a location in the linearized limit-state function, in which the
distance from the origin in the standard normal space is the minimum. This point is
commonly known as the “design point” or “most probable point” (MPP). The linearized
limit-state function is written as

G (u) = ∇ G (u∗ )(u − u∗ ) (4)

 
where ∇ G (u∗ ) = ∂u1∗ G, ∂u2∗ G, . . . , ∂u∗n G denotes the gradient row vector. The lineariza-
tion represents the failure domain G (u) ≤ 0, that is, the half-space defined by β − αu ≤ 0.
∇ G (u∗ )
β = αu∗ is the reliability index and α = − k∇G(u∗ )k denotes the normalized negative vector
at the design point (refer to Figure 1). This representation leads to

G (u) = k∇ G (u∗ )k( β − αu) (5)

Appl. Sci. 2021, 11, 4708 4 of 20

Figure 1. Illustration of the first-order reliability method (FORM) for a component reliability problem.

The probability content of the half-space, which is defined by the reliability index,
gives the approximation of the failure probability as

Pf ≈ Φ(− β) (6)

where Φ(·) denotes the cumulative distribution function (CDF) of the standard normal
distribution.
It should be noted that there are various algorithms for approximation of the design
point in Equation (3) and the reliability index [34,35]. The accuracy of the FORM depends
upon how well the actual limit-state surface is represented by the linear approximation. For
nonlinear limit-state functions having a larger number of random variables, the FORM may
not identify accurate design points or encounter a convergence issue while searching for
the most probable points for complex systems. The nonlinearity of the limit-state surface is
caused by various factors, such as the nonlinear relationship between random variables
and the transformation of random variables from the original space to standard normal
space. The second-order reliability method (SORM) [13,36,37] can improve the assessment
given by FORM by including information about the curvature. The SORM approximates
the limit-state surface by using a hyper-paraboloid with tangent hyperplane and main
curvatures at the design point. Furthermore, Der Kiureghian and Dakessian [38] showed
that a limit-state function might have multiple design points. An optimization algorithm
utilized to solve Equation (3) may converge to a local design point that results in a gross
approximation error. Several search algorithms [38–40] integrating multiple design points
have been proposed as an effort to improve the accuracy of reliability analysis.

2.2. Reliability-Based Design Optimization

One of the goals in structural optimization is to find the optimal solutions in a given
design domain subjected to tractions and displacement boundary conditions while satisfy-
ing given design constraints. When the reliability analysis of a structure under uncertainty
is incorporated into structural optimization, it is referred to as RBDO. RBDO aims to achieve
the optimal design under given probabilistic constraints, arising from uncertainties in ma-
terial properties or loads. The component RBDO (CRBDO) approach seeks to satisfy the
probabilistic constraint regarding each failure event. A CRBDO problem can be formulated as

min f obj (d, µx )

d,µx
s.t. Pf i = P[ gi (d, x) ≤ 0] ≤ Pf i t , i = 1, . . . , nc (7)
dl ≤ d ≤ du
Appl. Sci. 2021, 11, 4708 5 of 20

where f obj is the objective function; d is a vector of deterministic design variables; µx is

the vector of the means of random variable x; P[ gi (d, x) ≤ 0] is the failure probability of
the i-th limit-state function gi (·); Pf i t represents the target failure probability of the i-th
limit-state; nc is the number of probabilistic constraints; and dl and du are the lower bounds
of design variables and upper bounds of design variables, respectively. This problem is
solved by generating and solving a sequence of explicit subproblems that are approxima-
tions of Equation (7). Optimization algorithms, such as sequential linear programming
(SLP), sequential quadratic programming (SQP; [41]), convex linearization (COLIN; [42]),
optimality criteria (OC; [43]), and method of moving asymptotes (MMA; [44]), incorporate
information about the sensitivity of objective and probabilistic constraint functions to solve
subproblems.

2.3. Sensitivity Analysis

Sensitivity analysis of the probabilistic constraint aims to evaluate the dependence of
reliability on design parameters or random variables. To utilize an efficient gradient-based
optimization algorithm in RBDO, sensitivity analysis is integral. The sensitivity of the
failure probability with respect to the design parameter d [45] can be computed by a chain
rule as
dPf dΦ(−β)
∇d P f = dβ(d,x∗ )
·∇d β(d, x∗ ) = dβ(d,x∗ )
·∇d β(d, x∗ ) = − ϕ( β)·∇d β(d, x∗ ) (8)

where
∇d β(d, x∗ ) = 1
k∇ G (d,u∗ )k
·∇d g(d, x∗ ) (9)
Similarly, the sensitivity of the probability approximation with respect to a set of
distribution parameters θ, such as the mean and standard deviation, is obtained from
dPf dΦ(−β)
∇θ P f = dβ(d,x∗ )
·∇θ β(d, x∗ ) = dβ(d,x∗ )
·∇θ β(d, x∗ ) = − ϕ( β)·∇θ β(d, x∗ ) (10)

where
∇θ β(d, x∗ ) = αJu,θ (x∗ , θ) (11)
where Ju,θ is the Jacobian of the θ to u transformation.

3. Complex-Step Approximation and Finite Difference Method

The complex-step (CS) derivative method [46] was introduced by Squire and Trapp [47]
and has been proven to be more efficient for the first-order derivative calculation than
the conventional finite difference method [48]. In the CS derivative approximation, an
imaginary number multiplied by the step size h is utilized. The first derivative is approxi-
mated by a single function evaluation with it. To derive the CS derivative and identify its
associated errors, consider the differentiable function f (·) and the point y on the real axis
(y ∈ R). The Taylor series expansion of f (y + ih) gives

h2 (2) 3 4
f (y + ih) ≈ f (y) + ih f 0 (y) − 2! f (y) − i h3! f (3) (y) + h4! f (4) (y) + . . . (12)

where i is the imaginary unit. Rearranging Equation (12) and taking the imaginary part of
both sides yield
=[ f (y+ih)]
f 0 (y) ≈ + O h2


h
(13)
where =[·] denotes the imaginary part of the function f. When a small h is selected, terms
with the order h2 and higher can be truncated. Evaluating the function f at the imaginary
argument y + ih, and dividing it by h gives an approximation to the first derivative. The
CS approximation in Equation (13) does not cause subtractive cancellation and results in
calculations without the associated round-off error. Note that the real part of Equation (13)
multiplied by h is equal to f (y). Therefore, the first derivative approximation and function
value can be computed simultaneously, which will result in increased computational
Appl. Sci. 2021, 11, 4708 6 of 20

efficiency for analysis and optimization. The level of accuracy in the first derivative using
the conventional finite difference method can vary with a change in the step size, which is
generally selected arbitrarily. This level indicates that the actual effect of the step size in the
finite difference method on the result makes it difficult to predict prior to actual calculation.
It is noteworthy that sensitivity analysis using the CS approximation simply requires
the evaluation of the function with a complex variable without analytical derivations of
gradients. Therefore, the CS approximation allows for sensitive analysis in situations when
the analytic derivative of gradients cannot be explicitly expressed or when each analysis
becomes computationally expensive.
More accurate finite difference approximations for the first derivative can be de-
rived by combining different Taylor series expansions. For instance, subtracting the two
expansions at the points f (x−h) and f (x+h) gives

f ( x +h)− f ( x −h)
f 0 (x) = + O h2


2h
(14)

An error from Equation (14), which is referred to as the second-order central differ-
ence method, is proportional to h2 . Similarly, second-order forward difference (FDM2 )
and fourth-order central difference methods method (CDM4 ) for approximating the first
derivative are
(
−3 f ( x )+4 f ( x +h)− f ( x +2h)
FDM2 : + O h2


0
f (x) = 2h (15)
− f ( x +2h)+8 f ( x +h)−8 f ( x −h)+ f ( x −2h)
CDM4 : + O h4


12h

Although the high-order finite difference method leads to less truncation error, which
can also be minimized by taking a relatively small h, subtractive cancellations involved in
formulas may lead to significant round-off errors, unlike the CS approximation.

4. Proposed Method of Sensitivity Analysis

The proposed method utilizes the CS approximation to compute the sensitivity of a
probabilistic constraint with respect to a vector of design variables. This section introduces
procedures of the CS derivative approximation for gradients and sensitivity analysis.
Consider the vector of random variables x = ( x1 , x2 , x3 , . . . , xn )T and the vector of design
variable d = (d1 , d2 , d3 , . . . , dz )T . The sensitivity of a reliability index in the FORM with
respect to a vector of design variables can be derived using the CS approximation as
∗ ∗ d,x∗ ) T
h i
∇d β(d, x ) = k∇G(1d,u∗ )k ·∇d g(d, x∗ ) = k∇G(1d,u∗ )k · ∂g(∂dd,x1 ) , . . . , ∂g(∂dd,x ) , . . . , ∂g(∂d
∗
z
k (16)
1 =[ g(d+ihe1 ,x∗ ),...,g(d+ihek ,x∗ ),...,g(d+ihez ,x∗ )]T
= k∇G(d,u∗ )k · h

where ek denotes the k-th column of a z-th order identity matrix (Iz ). Because the limit-state
function is defined in terms of the random variables x, it is necessary to carry out these
calculations in the original space. Results are then transformed in the standard normal
space to perform the FORM. The gradient vector ∇ G (u∗ ) can be rewritten as
−1 (x∗ )
∇ G (d, u∗ ) = ∇x∗ g(d, x∗ )Ju,x (17)

−1 is the inverse of the Jacobian of the x to u transformation. Note that the Jacobian
where Ju,x
is dependent on the distribution types of random variables. For instance, the Jacobian of
transformation Ju,x for statistically independent random variables x is

Ju,x = diag[ Jii ], i = 1, . . . , n

f (x ) (18)
Jii =  −1 hR xii i i
ϕ Φ −∞ f i ( xi )dxi

where f i ( xi ) is the marginal PDF of xi and ϕ(·) is the univariate standard normal PDF. If
the random variables x are normally distributed with the correlation matrix R = LLT , in
Appl. Sci. 2021, 11, 4708 7 of 20

which L is a lower-triangular matrix obtained by the Cholesky decomposition of R, the

Jacobian is
Ju,x = L−1 D−1 (19)
where D is the diagonal matrix of standard deviations. Furthermore, the Jacobian matrix
of dependent nonnormal random variables can be derived [49] as

f (x )
J11 = R x1 1
ϕ(Φ−1 [ −1∞ f 1 ( x1 )dx1 ])
Jij =0 i<j Zxi
f i ( xi | x1 ,...,xi−1 ) where Fi ( xi | x1 , . . . , xi−1 ) = f i ( xi | x1 , . . . , xi−1 )dxi (20)
= i=j>1
ϕ(Φ−1 [ Fi ( xi | x1 ,...,xi−1 )]) −∞
1 ∂Fi ( xi | x1 ,...,xi−1 )
= i>j
ϕ(Φ−1 [ Fi ( xi | x1 ,...,xi−1 )]) ∂x j

where f i ( xi | x1 , . . . , xi−1 ) denotes the conditional PDF of xi for given x1 , . . . , xi−1 .

The gradient of the limit-state function ∇x∗ g(d, x∗ ) in Equation (17) is then computed
as
∂g(d, x∗ ) ∂g(d, x∗ ) ∂g(d, x∗ ) T
 
∇x∗ g(d, x∗ ) = , . . . , , . . . , (21)
∂x1∗ ∂xk∗ ∂xn∗
Application of the CS approximation to Equation (21) yields

∇x∗ g(d, x∗ ) = 1h ·=[ g(d, x∗ + ihe1 ), . . . , g(d, x∗ + ihek ), . . . , g(d, x∗ + ihen )]T (22)

Note that taking a real part of g(d, x∗ + ihek ), k = 1, . . . , n in Equation (22) results
in a value of g(d, x∗ ). This indicates that both the value and sensitivity of the limit-state
function can be obtained from Equation (22). Using Equation (8), the sensitivity of the first-
order probability approximation with respect to a vector of design variables is obtained by
computing
∗ ∗ ∗ )]T
∇d Pf = − ϕ( β)· =[ g(d+ihe1 ,xh)k∇
,...,g(d+ihek ,x ),...,g(d+ihez ,x
g(d,x∗ )J−1 (x∗ )k
(23)
x∗ u,x

Note that the gradients in Equation (21) are also utilized in the FORM to compute a
failure probability of a limit-state function.
When the improved Hasofer and Lind algorithm [50] is utilized to solve Equation (3),
a sequence of points u by using a recursive formular is identified as

ui+1 = ui + λi si , i = 1, 2, 3, . . . . (24)

where λi is a search scale factor and si is the search direction vector obtained by
 
i G (d,ui )
T
s = + α u αi − ui
i i (25)
∇ G (d,ui )

The gradient ∇ G d, ui can be computed by using the CS approximation as

      1 h      iT  
∇ G d, ui = ∇xi g d, xi Ju−,1x xi = ·= g d, xi + ihe1 , . . . , g d, xi + ihek , . . . , g d, xi + ihen −1
Ju,x x
i
(26)
h
The proposed RBDO method using the CS approximation (RBDO−CS) performs the
reliability analysis by identifying the design points based on gradients in Equations (22) and
(26), and sensitivity analysis using Equation (23) to utilize a gradient-based optimization
algorithm.
Appl. Sci. 2021, 11, 4708 8 of 20

5. Numerical Applications
5.1. Comparative Study of the Accuracy of Complex-Step and Finite Difference Methods
The effect of step size on the first derivative by finite difference and CS methods is
investigated. Consider the following highly nonlinear function:
ex
f (x) = √ (27)
sin3 ( x )+cos3 ( x )

The four finite difference methods (first-order forward (FDM), second-order forward
(FDM2 ), second-order central (CDM2 ), and fourth-order central difference (CDM4 ) approx-
imations) and CS approximation discussed in Section 3 are utilized to compute the first
derivative at x = −0.35 with varying step sizes.
The numerical results in Figure 2a show that the first derivatives by the CS method are
not sensitive to the selection of step size. Finite difference methods, however, break down
as the step size reaches the last value. Although higher-order finite difference methods
result in fewer truncations errors, subtractive cancellation and their associate round-off
errors affect the calculations. It should be noted that the number of function evaluations
in higher-order finite difference methods is higher than one in the CS approach. Errors
between the exact derivatives and approximated derivatives with varying step sizes are
plotted in Figure 2b. The comparison results confirm that the CS approximation is more
accurate for small step sizes, whereas the finite difference approach cannot achieve full
accuracy.

Figure 2. (a) Comparison of first derivatives by finite difference and complex-step (CS) methods,
(b) absolute errors from computing the first derivatives with varying step sizes.

5.2. RBDO of the Short Column under a Probabilistic Strength Constraint

The proposed RBDO−CS approach is applied to identify the optimal member size of
a short column subjected to an axial force P and bending moments M1 and M2 , as shown
in Figure 3. Assume an elastic perfectly plastic material with yield stress σy . The limit-state
function regarding the failure of the column is defined by
 q
M1 M2 P
g(d, x) = 1 − S1 (d)σy
− S2 (d)σy
− A(d)σy
(28)
Appl. Sci. 2021, 11, 4708 9 of 20

T
where x = M1 , M2 , P, σy denotes the vector of random variables, q = 2 is the limit-state
function parameter, d is the design variable, A(d) = 1.5d2 is the cross-sectional area, and
S1 (d) = 0.375d3 and S2 (d) = 0.25d3 are the flexural moduli of the plastic column section.
Assume that M1 , M2 , P, and σy have the Nataf distribution with the marginal distributions
and coefficient of variance (c.o.v) listed in Table 1. The objective function for this problem
is taken as the minimization of the cross-sectional area, and the probabilistic constraint is
defined in terms of yield strength. The RBDO problem is formulated as

min A(d)
d
s.t. Pf = P[ g(d, x) ≤ 0] ≤ Pf t (29)
dl ≤ d ≤ du

Figure 3. Short column dimension and loading profiles.

Table 1. Description of random variables for the short column optimization problem.

Random Marginal Correlation Coefficient

Mean c.o.v
Variables Distribution M1 M2 P σy
M1 , kNm Normal 250 0.3 1.0
M2 , kNm Normal 125 0.3 0.5 1.0
P, kN Gumbel 2500 0.2 0.3 0.3 1.0
σy , MPa Weibull 40 0.1 0 0 0 1.0

Table 2 summarizes the target failure probability and optimization parameters. A sen-
sitivity analysis is performed by the proposed method to utilize a gradient-based optimiza-
tion algorithm—the method of moving asymptotes (MMA; [44]). The optimization process
is repeated until the convergence criteria on the maximum number of interactions or the
relative change in the design variable is satisfied. The initial reliability analysis of the limit-
state function g d0 , x by the FORM results in u∗ = (−2.0084, −0.9605, −0.3399, 0.4619)T ,

β = −2.2989, and Pf = 0.9892. In this study, the single loop approach using the Karush–
Kuhn–Tucker conditions (SL–KKT) [18] is utilized to compare the proposed RBDO–CS
method. The CS approximation is also implemented to compute sensitivities in the SL–KKT.
The same optimization parameters and convergence criteria are considered for the compar-
ison. Table 3 represents the optimal area and design variables obtained by the proposed
RBDO–CS and SL–KKT methods, as well as the averaged computational time carried out
to five optimization iterations (a modern (2020) eight-core 3.8 GHz Intel process (16 GB
RAM) is used for the analyzes). Because the SL–KKT approach does not perform reliability
analysis during optimization procedures, the failure probability of limit-state function is
computed by the FORM with optimal column design. The optimal results from both ap-
proaches are in agreement with each other. Figure 4 shows the convergence histories of the
RBDO–CS and the SL–KKT methods regarding the objective function, design variable, and
failure probability. The proposed RBDO–CS method enables the identification of a feasible
solution while satisfying the probabilistic constraint. Although the proposed RBDO–CS
Appl. Sci. 2021, 11, 4708 10 of 20

method performing reliability analysis requires a slightly increased computational time for
five iterations of the optimization procedure when compared to the SL–KKT method, the
number of iterations required to identify the converged solution is lower. Of note, when the
limit-state function is highly nonlinear, the second-order reliability method (SORM; [13])
leads to a more accurate assessment of reliability. The limit-state function in Equation (28) is
nonlinear and the SORM is utilized to assess failure probabilities of the optimized column
by the RBDO–CS, and SL–KKT methods. The SORM results in Table 3 indicate that the
first-order linearization of the limit-state function may lead to a less conservative design
when it has a high degree of nonlinearity.

Table 2. Parameters for the target failure probability and optimization.

Upper Bound Convergence

Target Failure Probability Pf t Initial Design d0 , m Lower Bound dl , m
du , m Criterion
0.01 0.25 0.1 2.0 10−4

Table 3. Comparison between results from the RBDO−CS and SL−KKT approaches for the column design problem.

RBDO Optimal Optimal Pf Averaged Computational Time Carried Total Number

Method f* (m2 ) d* (m) FORM SORM out to Five Iterations, (Seconds) of Iterations
RBDO−CS 0.2293 0.3909 0.0100 0.0126 0.0130 11
SL−KKT 0.2293 0.3908 0.0101 0.0127 0.0108 12

Figure 4. Convergence history of the proposed RBDO−CS and SL−KKT approaches: (a) optimal
area; (b) design variable; and (c) failure probability.

5.3. RBDO of the Ductile Frame Structure Subjected to Probabilistic Moment Strength Constraints
The frame structure under uncertainty in the magnitudes of external forces and
moment capacities is illustrated in Figure 5a. Assume that the frame structure is constructed
with ductile members that have plastic moment capacities mi , i = 1, . . . , 5, at joints. Under
the externally applied forces h, v, this frame may fail due to any of the sway, beam, and
Appl. Sci. 2021, 11, 4708 11 of 20

combined mechanisms illustrated in Figure 5b. Using the principle of virtual work, the
limit-state functions regarding these mechanisms are formulated [13] as

g1 (d, x) = m1 + m2 + m4 + m5 − hd1
g2 (d, x) = m2 + 2m3 + m4 − vd2 (30)
g3 (d, x) = m1 + 2m3 + 2m4 + m5 − hd1 − vd2

where x = (m1 , m2 , . . . , m5 , h, v)T is the vector of random variables, and d = (d1 , d2 )T is

the vector of design variables (d1 : height, d2 : width). Assumed distributions and second
moments of the random variables are tabulated in Table 4. The maximization of the height
and width of the ductile frame structure and the three limit-state functions in terms of
the moment capacities are considered the objective and probabilistic constraint functions,
respectively. The RBDO problem of the ductile structure can be stated as

max f (d) = d1 + 2d2

d
s.t. Pf 1 = P[ g1 (d, x) ≤ 0] ≤ Pf 1 t
Pf 2 = P[ g2 (d, x) ≤ 0] ≤ Pf 2 t (31)
Pf 3 = P[ g3 (d, x) ≤ 0] ≤ Pf 3 t
dl ≤ d ≤ du

Figure 5. (a) Ductile frame structure, and (b) failure mechanisms.

Table 4. Description of random variables for the ductile frame optimization problem.

Random Variables Marginal Mean c.o.v Correlation

Distribution
m1 , kNm
m2 , kNm
ρmi m j = 0.3
m3 , kNm Joint lognormal 150 0.2 , i 6= j
m4 , kNm
m5 , kNm
h, kN Gumbel 50 0.4 Independent
v, kN Gamma 60 0.2 Independent
Appl. Sci. 2021, 11, 4708 12 of 20

A target failure probability of 0.003 is assigned to all probabilistic constraints. The

initial design and parameters for RBDO are summarized in Table 5. Using the FORM with
d0 , the following values for the three constraints are obtained:

Limit-State Function β Pf
 
g 1 d0 , x 1.452 0.073
 
g 2 d0 , x 1.435 0.076
 
g 3 d0 , x 0.701 0.242

Table 5. Target failure probability and parameters for design variables and optimization.

Target Failure Initial Design Upper Bound Convergence

Lower Bound dl , m
Probability Pf t d0 , m du , m Criterion
Pf i t = 3 × 10−3
(7.0, 7.0) (1.0, 1.0) (10.0, 10.0) 10−4
(i = 1, 2, 3)

The convergence histories of the objective function, design variables, and failure
probabilities given in Figure 6 demonstrate the quick identification of converged solutions
from the RBDO−CS method while satisfying the probabilistic constraints. To verify the
results of RBDO−CS, the SL−KKT approach is utilized for the ductile structure with
the same design and optimization parameters in Tables 4 and 5. Figure 6 shows the
comparison of results from RBDO−CS and SL−KKT methods, which are in agreement.
Values of design variables and failure probabilities of the limit-state functions obtained
by the FORM and SORM at optimal solutions are listed in Table 6. Note that the failure
probability of g1 (d∗ , x∗ ) is less than the target failure probability, which indicates that this
constraint is inactive at the local optimum.

Figure 6. Convergence histories of the proposed RBDO−CS and SL−KKT methods: (a) optimal
objective function; (b) design variables; and (c) failure probabilities.
Appl. Sci. 2021, 11, 4708 13 of 20

Table 6. Comparison between results from the RBDO−CS and SL−KKT methods for the frame structure problem.

RBDO Optimal Optimal Pfi Averaged Computational Time Carried Total Number of
Method f∗ (m) d∗ (m) FORM SORM out to Five Iterations, (Seconds) Iterations

i=1 0.0004 0.0004

RBDO−CS 13.658 (3.362,5.148) i=2 0.0030 0.0028 0.105 12
i=3 0.0030 0.0032
i=1 0.0004 0.0012
SL−KKT 13.667 (3,364,5.151) i=2 0.0030 0.0027 0.031 14
i=3 0.0030 0.0033

The RBDO−CS method results in the identification of more optimal solutions when
compared to the results from SL−KKT method. Furthermore, the SORM results indicate
that the failure probabilities of limit-state functions at d∗ obtained by RBDO−CS method
are closer to the target failure probabilities than by the SL−KKT method.

5.4. RBDO of the Truss Structure under Probabilistic Strength Constraints

The proposed RBDO−CS method is applied to identify the optimal element sizes of
an indeterminate truss structure by McDonald and Mahadevan [51] (see Figure 7). The
deterministic design variables d = ( A1 , A2 , A3 , A4 , A5 , A6 )T are the cross-sectional areas
of elements. The applied force FA and yield strengths σi , i = 1, . . . , 6 of the elements are

T
the random variables x = σ1 , σ2 , . . . , σy , FA , which are assumed to follow the normal
distribution. All random variables are statistically independent of each other, and the
means and standard deviations of them are summarized in Table 7. In this example,
RBDO aims to minimize the volume of the truss structure while satisfying the probabilistic
constraints in terms of yield strengths. In this example, the truss element forces are derived
in terms of the applied load with the assumption of the two diagonal elements carrying
equal forces. Limit-state functions are defined as:
√
gi (d, x) = Ai σi − 22 FA , i = 1, 2
A1 , A2 , A3 , A4 , A5 , A6 ≥ 0 (32)
= Ai σi − 21 FA , i = 3, 4, 5, 6

Figure 7. Six-element truss structure.

Appl. Sci. 2021, 11, 4708 14 of 20

Table 7. Description of random variables for the truss optimization problem.

Random Marginal Mean Standard Correlation

Variables Distribution Deviation
FA , kips Normal 1000 100 Independent
σ1 , ksi
σ2 , ksi
σ3 , ksi
Normal 36 3 Independent
σ4 , ksi
σ5 , ksi
σ6 , ksi

The RBDO problem can be defined as follows:

√
min f (d) = 2 ∑2i=1 Ai + ∑6j=3 A j
d
s.t. Pf i = P[ gi (d, x) ≤ 0] ≤ Pf i t , i = 1, . . . , 6 (33)
dl ≤ d ≤ du

The initial design variables d0 are set to 12 in2 , and the corresponding failure prob-
abilities of the constraints by the FORM are Pf i = (0.999, 0.999, 0.865, 0.865, 0.865, 0.865).
The proposed RBDO−CS method with the parameters listed in Table 8 is able to find an
optimal solution quickly, whose convergence history is plotted in Figure 8. Furthermore,
the optimal designs are confirmed with the SL−KKT approach.

Table 8. Target failure probability and parameters for design variables and optimization.

Target Failure Initial Design Upper Bound Convergence

Lower Bound dl , in2
Probability Pf t d0 , in2 du , in2 Criterion
Pf i t = 1.667 × 10−4
(12, 12) (5, 5) (50, 50) 10−3
(i = 1, . . . , 6)

Figure 8. Convergence histories of the proposed RBDO−CS method: (a) optimal volume; (b) design
variables; and (c) failure probabilities.

Table 9 represents the optimal design and failure probabilities of constraints. The
results of the SORM and MCS with 108 samples (c.o.v = 0.02) confirm the failure probability
Pf i of the optimal design by the proposed method. It should be noted that the FORM
Appl. Sci. 2021, 11, 4708 15 of 20

and SORM approximations of the probability of failure are sufficiently accurate for this
problem. It is because the limit-state functions in Equation (32) are linear, and random
variables following normal distributions are statistically independent of each other.

Table 9. Results of the RBDO−CS and SL−KKT approaches for the truss optimization problem.

Optimal f∗ ×L, in3 Optimal Ai∗ , in2 Pfi

RBDO−CS/SL−KKT Element RBDO−CS/SL−KKT FORM SORM MCS
1 31.37 1.667 × 10−4 1.667 × 10−4 1.655 × 10−4
2 31.37 1.667 × 10−4 1.667 × 10−4 1.672 × 10−4
3 22.18 1.667 × 10−4 1.667 × 10−4 1.684 × 10−4
177.46
4 22.18 1.667 × 10−4 1.667 × 10−4 1.667 × 10−4
5 22.18 1.667 × 10−4 1.667 × 10−4 1.675 × 10−4
6 22.18 1.667 × 10−4 1.667 × 10−4 1.654 × 10−4

5.5. RBDO of the Truss Cantilever Structure Subjected to Probabilistic Displacement Constraints
Consider a truss cantilever structure consisting of 56 bar elements and 21 nodes
(Figure 9c,d). The truss cantilever structure subjected to two external forces at nodes 5 and
21 contains three-roller supports as shown in Figure 9a. All of the truss elements have the
same modulus of elasticity E. The objective of RBDO is to minimize the weight of the truss
cantilever structure and constraints are vertical displacements (u5y , u21y ) at nodes 5 and
21. Limit-state functions are defined as follows:

g1 (d, x) = u5 − u5y (d, x)

(34)
g2 (d, x) = u21 − u21y (d, , , x)

where u5 , u21 are the maximum allowable displacements in the vertical direction, respec-
tively. Two external

forces and modulus of elasticity are assumed to be random variables
x = F5y , F21y , E , and cross-sectional areas are design variables d = ( Ai , i = 1, . . . , 56).
Table 10 provides marginal distributions, mean values, and second moments of the random
variables. Parameters for reliability analysis and optimization are listed in Table 11. The
RBDO problem can be written in a nested formulation as follows:

min f (d) = ∑56

i =1 A i L i
d l u
s.t. Pf i = P[ gi (d, x) ≤ 0] ≤ Pf i t , i = 1, 2 d ≤ d ≤ d (35)
with K(d, x)u(d, x) = f(x)

where K is the global stiffness matrix of the truss cantilever structure, u is the global
displacement vector, and f is the global external force vector. Note that u is an implicit
function defined through the equilibrium equations K(d, x)u(d, x) = f(x). The derivative of
the displacement vector with respect to a parameter θ (= d or x) is obtained analytically as

∂K−1 (d,x) ∂f(x)

∇θ u(d, x) = ∂θ f (x) + K−1 (d, x) ∂θ
(36)

Table 10. Description of random variables for the truss structure optimization problem.

Random Marginal Mean c.o.v Correlation

Variables Distribution
F5y , kips Joint lognormal 120 0.1 ρ F5y, F21y = 0.3
F21y , kips Joint lognormal 140 0.1
E, ksi Normal 29000 0.2 Independent
Appl. Sci. 2021, 11, 4708 16 of 20

Figure 9. Truss cantilever structure: (a) dimension, loading, and boundary conditions; (b) deformed shape and displace-
ments of interest; (c) truss element numbers, and (d) node numbers.

Table 11. Parameters for the target failure probability and optimization.

¯
u5 ¯ Lower Bound Upper Bound Convergence
Pf t u 21 (in) Initial Design d0 (in2 )
(in) dl (in2 ) du (in2 ) Criterion
0.005 0.1 0.3 5.0 0.5 50.0 10−3

The sensitivity analysis in Equation (36) requires gradients of inversed stiffness matrix
and force vector, which are computationally expensive and implicitly defined. Rather
than directly solving Equation (36) for sensitivity analysis, applying the CS derivative
approximation to Equation (36) yields
   
^
= K−1 d+ihei ,x f(x)
∂K−1 (d,x)
θ = Ai : ∇ Ai u(d, x) = ∂Ai f(x) = h 
 
¯
= K−1 (d,x)f x+ih e 1
(x)
θ = F5y : ∇ F5y u(d, x) = K−1 (d, x) ∂f
∂F5y = h

¯
 (37)
= K−1 (d,x)f x+ih e 2
∂f(x)
θ = F21y : ∇ F21y u(d, x) = K−1 (d, x) ∂F21y
= h
  
¯
= K−1 d,x+ih e 3 f(x)
∂K−1 (d,x)
θ=E: ∇ E u(d, x) = ∂E f(x) = h

^ ¯
where ei is the i-th column of a 56-th order identity matrix (I56 ) and e k is the k-th column of a
3-rd order identity matrix (I3 ). Sensitivities of the limit-state functions with respect to design
variables can be computed by using Equation (37). Furthermore, sensitivities of limits-state
functions with respect to random variables are used in the FORM for reliability analysis.
Notably, the displacement vector u and its sensitivity are computed without additional step
by taking real and imaginary parts in Equation (37). The initial displacements at nodes 5
and 21 of the truss cantilever structure with mean values of random variables and uniform
cross-sectional areas of 5 in2 for all elements are 0.1339 in and 0.3557 in and corresponding
failure probabilities by the FORM are 0.928 and 0.789.
Figure 10 illustrates the convergence histories of objective and probabilistic con-
straint functions, optimized element sizes, and optimal cantilever structure. The proposed
RBDO−CS method quickly identifies the solutions satisfying probabilistic constraints and
Appl. Sci. 2021, 11, 4708 17 of 20

then reduces the volume to minimize. Optimization results in Figure 10c indicate that
elements connected to the roller support in the middle and the bottom chords are primarily
strengthened to satisfy probabilistic constraints. Table 12 provides the results from the
RBDO−CS method for the truss cantilever structure. The SL−KKT method is utilized to
verify optimization results. It is observed that the results from the proposed RBDO−CS
and SL−KKT approaches closely match. Failure probabilities of limit-state functions with
optimal design d∗ and random variables are estimated by the FORM, SORM, and MCS
with 107 samples (c.o.v = 0.03). This study confirms that the proposed RBDO−CS method
enables the efficient identification of optimal solutions. It is noted that the computational
costs of RBDO−CS and SL−KKT methods are similar with a larger number of design vari-
ables, however, the proposed RBDO−CS method identifies converged solutions faster than
the SL−KKT method. Next, the force random variables F5y , F21y are assumed to follow the
normal distributions with the same means and standard deviations. This is to investigate
the effect of probabilistic distribution types on the optimal design and to demonstrate the
general applicability of the proposed method. The minimum objective function value is
larger than that of the lognormal distribution case, as illustrated in Figure 11. Further-
more, the impact of statistical correlation between random forces on the optimal design
is investigated by altering correlation coefficients ρ F5y , F21y from 0.0 to 0.6. The optimal
volumes obtained by the RBDO−CS method are shown in Figure 11. A positive correlation
among the random forces results in higher volume, i.e., more conservative design. This is
because positively correlated forces F5y , F21y increase the vertical displacements, and thus
the failure probabilities.

Figure 10. Results of the proposed RBDO−CS: (a) Convergence histories of volume and failure probability; (b) cross-
sectional areas (elements having the minimum area of 0.5 in2 are in grey); and (c) illustration of normalized cross-sectional
areas of elements (elements in blue are in tension and those in red are in compression.).
Appl. Sci. 2021, 11, 4708 18 of 20

Table 12. Comparison between results from the RBDO−CS and SL−KKT for the truss cantilever structure design.

RBDO Optimal f ∗ Pf i Averaged Computational Time Carried Total Number of

Method (ft3 ) i FORM SORM MCS out to Five Iterations, (Seconds) Iterations
1 0.0050 0.00516 0.00516
RBDO−CS 5.828 0.0637 37
2 0.0050 0.00516 0.00517
1 0.0050 0.00517 0.00518
SL−KKT 5.827 0.0591 175
2 0.0050 0.00518 0.00519

Figure 11. Optimal volumes versus correlation between external forces having different marginal
distributions.

6. Concluding Remarks
An approach to the RBDO of structures is developed using CS approximation and
the FORM. The improvement in the accuracy of probabilistic constraints by the CS deriva-
tive and its straightforward implementation are presented in this paper. The proposed
approach incorporates component reliability targets in RBDO and eliminates the need to
analytically derive gradients of functions with respect to the design and random variables
for sensitivity analysis. In addition to accounting for the statistical dependence between
two-component events, the approach can compute the sensitivities of the failure probability
to various parameters, which facilitate the use of gradient-based optimization algorithms.
The first numerical example demonstrates the accuracy of the CS approximation without
significant reliance on a step size, compared to finite difference methods. Furthermore,
the uniform applicability of the proposed approach to RBDO of structures is confirmed
through four numerical examples. This paper focuses on the CRBDO with probabilistic
constraints of multiple failure modes of a design. The CRBDO problem is formulated such
that the optimal structure satisfies each failure mode with target failure probabilities. In
some instances, the design constraint needs to be defined as a “system” event of multiple
constraints, (i.e., a logical (or Boolean) function of multiple failure modes), which is com-
monly termed a system reliability-based design optimization (SRBDO) problem. In these
cases, the sensitivity analysis of the system reliability can be more expensive and compli-
cated than a CRBDO problem due to the statistical dependency among failure events and
complexity inherent in the formula. The implementation of -step approximation in SRBDO
with diverse reliability analysis methods remains to be further explored. To improve the
accuracy of reliability analysis and enhance the use of gradient-based optimization algo-
rithms, areas for further study include sensitivity analysis with CS approximation unified
with the SORM or multiple design points approaches, and its incorporation into RBDO.
Appl. Sci. 2021, 11, 4708 19 of 20

Funding: This research was funded by the Collaboration for Unprecedented Success and Excellence
(CUSE) grant program of Syracuse University.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Conflicts of Interest: The author declares no conflict of interest.

References
1. Haftka, R.T.; Gürdal, Z. Elements of Structural Optimization; Springer: Dordrecht, The Netherlands, 1992.
2. Sandgren, E.; Cameron, T. Robust design optimization of structures through consideration of variation. Comput. Struct. 2002, 80,
1605–1613. [CrossRef]
3. Marano, G.C.; Sgobba, S.; Greco, R.; Mezzina, M. Robust optimum design of tuned mass dampers devices in random vibrations
mitigation. J. Sound Vib. 2008, 313, 472–492. [CrossRef]
4. Zang, C.; Friswell, M.; Mottershead, J. A review of robust optimal design and its application in dynamics. Comput. Struct. 2005,
83, 315–326. [CrossRef]
5. Parkinson, A.; Sörensen, C.; Pourhassan, N. A General Approach for Robust Optimal Design. J. Mech. Des. 1993, 115, 74–80.
[CrossRef]
6. Chen, W.; Allen, J.K.; Tsui, K.-L.; Mistree, F. A Procedure for Robust Design: Minimizing Variations Caused by Noise Factors and
Control Factors. J. Mech. Des. 1996, 118, 478–485. [CrossRef]
7. Sundaresan, S.; Ishii, K.; Houser, D.R. A robust optimization procedure with variations on design variables and constraints. Eng.
Optim. 1995, 24, 101–117. [CrossRef]
8. Messac, A.; Ismail-Yahaya, A. Multiobjective robust design using physical programming. Struct. Multidiscip. Optim. 2002, 23,
357–371. [CrossRef]
9. Enevoldsen, I.; Sørensen, J. Reliability-based optimization in structural engineering. Struct. Saf. 1994, 15, 169–196. [CrossRef]
10. Tu, J.; Choi, K.K.; Park, Y.H. A New Study on Reliability-Based Design Optimization. J. Mech. Des. 1999, 121, 557–564. [CrossRef]
11. Youn, B.D.; Choi, K.K.; Du, L. Enriched Performance Measure Approach for Reliability-Based Design Optimization. AIAA J. 2005,
43, 874–884. [CrossRef]
12. Youn, B.D.; Choi, K.K.; Park, Y.H. Hybrid Analysis Method for Reliability-Based Design Optimization. J. Mech. Des. 2003, 125,
221–232. [CrossRef]
13. Der Kiureghian, A. First- and Second-Order Reliability Methods. In Engineering Design Reliability Handbook; CRC Press: Boca
Raton, FL, USA, 2005.
14. Du, X.; Chen, W. Sequential Optimization and Reliability Assessment Method for Efficient Probabilistic Design. J. Mech. Des.
2004, 126, 225–233. [CrossRef]
15. Zou, T.; Mahadevan, S. A direct decoupling approach for efficient reliability-based design optimization. Struct. Multidiscip. Optim.
2006, 31, 190–200. [CrossRef]
16. Shan, S.; Wang, G.G. Reliable design space and complete single-loop reliability-based design optimization. Reliab. Eng. Syst. Saf.
2008, 93, 1218–1230. [CrossRef]
17. Nguyen, T.H.; Song, J.; Paulino, G.H. Single-Loop System Reliability-Based Design Optimization Using Matrix-Based System
Reliability Method: Theory and Applications. J. Mech. Des. 2009, 132, 011005. [CrossRef]
18. Liang, J.; Mourelatos, Z.P.; Tu, J. A Single-Loop Method for Reliability-Based Design Optimization. In Proceedings of the 28th
Biennial Mechanisms and Robotics Conference, Parts A and B, ASME International. Salt Lake, UT, USA, 28 September–2 October
2004; Volume 1, pp. 419–430.
19. Liang, J.; Mourelatos, Z.P.; Nikolaidis, E. A Single-Loop Approach for System Reliability-Based Design Optimization. J. Mech.
Des. 2007, 129, 1215–1224. [CrossRef]
20. Chun, J.; Paulino, G.H.; Song, J. Reliability-based topology optimization by ground structure method employing a discrete
filtering technique. Struct. Multidiscip. Optim. 2019, 60, 1035–1058. [CrossRef]
21. Chun, J.; Song, J.; Paulino, G.H. System-reliability-based design and topology optimization of structures under constraints on
first-passage probability. Struct. Saf. 2019, 76, 81–94. [CrossRef]
22. Grujicic, M.; Arakere, G.; Bell, W.C.; Marvi, H.; Yalavarthy, H.V.; Pandurangan, B.; Haque, I.; Fadel, G.M. Reliability-Based
Design Optimization for Durability of Ground Vehicle Suspension System Components. J. Mater. Eng. Perform. 2009, 19, 301–313.
[CrossRef]
23. Adams, B.; Eldred, M.; Wittwer, J. Reliability-Based Design Optimization for Shape Design of Compliant Micro-Electro-Mechanical
Systems. In Proceedings of the 11th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference, Portsmouth, VA,
USA, 6–8 September 2006; Volume 2, pp. 1042–1056.
24. Allen, M.; Maute, K. Reliability-based design optimization of aeroelastic structures. Struct. Multidiscip. Optim. 2004, 27, 228–242.
[CrossRef]
25. Liu, Z.; Zhu, C.; Zhu, P.; Chen, W. Reliability-based design optimization of composite battery box based on modified particle
swarm optimization algorithm. Compos. Struct. 2018, 204, 239–255. [CrossRef]
Appl. Sci. 2021, 11, 4708 20 of 20

26. Song, J.; Kang, W.-H.; Lee, Y.-J.; Chun, J. Structural System Reliability: Overview of Theories and Applications to Optimization.
ASCE-ASME J. Risk Uncertain. Eng. Syst. Part. A Civ. Eng. 2021, 7, 03121001. [CrossRef]
27. Frangopol, D.M.; Maute, K. Reliability-based optimization of civil and aerospace structural systems. In Engineering Design
Reliability Handbook (Chapter 24); Singhal, S., Ghiocel, D., Nikolaidis, E., Eds.; CRC Press: Boca Raton, FL, USA, 2004; pp. 1–32.
28. Tsompanakis, Y.; Lagaros, N.D.; Papadrakakis, M. Structural Design Optimization Considering Uncertainties; CRC Press: Baca Raton,
FL, USA, 2008; Volume 1.
29. Ambartzumian, R.; Der Kiureghian, A.; Ohaniana, V.; Sukiasiana, H. Multinormal probability by sequential conditioned
importance sampling: Theory and application. Probabilistic Eng. Mech. 1998, 13, 299–308. [CrossRef]
30. Genz, A. Numerical Computation of Multivariate Normal Probabilities. J. Comput. Graph. Stat. 1992, 1, 141–149. [CrossRef]
31. Wang, Z.; Song, J. Cross-entropy-based adaptive importance sampling using von Mises-Fisher mixture for high dimensional
reliability analysis. Struct. Saf. 2016, 59, 42–52. [CrossRef]
32. Kurtz, N.; Song, J. Cross-entropy-based adaptive importance sampling using Gaussian mixture. Struct. Saf. 2013, 42, 35–44.
[CrossRef]
33. Li, Z. Reliability-Based Design Optimization of Nonlinear Beam-Columns. Ph.D. Thesis, Virginia Polytechnic Institute and State
University, Blacksburg, VA, USA, 30 April 2018.
34. Liu, P.L. Der Kiureghian, A. Optimization algorithms for structural reliability. Struct. Saf. 1991, 9, 161–177. [CrossRef]
35. Bissell, D.; Ang, A.H.S.; Tang, W.H. Probability Concepts in Engineering Planning and Design: Vol. I-Basic Principles. J. R. Stat.
Soc. Ser. D Stat. 1979, 28, 226. [CrossRef]
36. Der Kiureghian, A.; Lin, H.; Hwang, S. Second-Order Reliability Approximations. J. Eng. Mech. 1987, 113, 1208–1225. [CrossRef]
37. Breitung, K. Asymptotic Approximations for Multinormal Integrals. J. Eng. Mech. 1984, 110, 357–366. [CrossRef]
38. Der Kiureghian, A.; Dakessian, T. Multiple design points in first and second-order reliability. Struct. Saf. 1998, 20, 37–49.
[CrossRef]
39. Grandhi, R.V.; Wang, L. Reliability-based structural optimization using improved two-point adaptive nonlinear approximations.
Finite Elements Anal. Des. 1998, 29, 35–48. [CrossRef]
40. Au, S.; Papadimitriou, C.; Beck, J. Reliability of uncertain dynamical systems with multiple design points. Struct. Saf. 1999, 21,
113–133. [CrossRef]
41. Boggs, P.T.; Tolle, J.W. Sequential Quadratic Programming. Acta Numer. 1995, 4, 1–51. [CrossRef]
42. Fleury, C.; Braibant, V. Structural optimization: A new dual method using mixed variables. Int. J. Numer. Methods Eng. 1986, 23,
409–428. [CrossRef]
43. Fleury, C. Structural weight optimization by dual methods of convex programming. Int. J. Numer. Methods Eng. 1979, 14,
1761–1783. [CrossRef]
44. Svanberg, K. The method of moving asymptotes—A new method for structural optimization. Int. J. Numer. Methods Eng. 1987, 24,
359–373. [CrossRef]
45. Bjerager, P.; Krenk, S. Parametric Sensitivity in First Order Reliability Theory. J. Eng. Mech. 1989, 115, 1577–1582. [CrossRef]
46. Lyness, J.N.; Moler, C.B. Numerical Differentiation of Analytic Functions. SIAM J. Numer. Anal. 1967, 4, 202–210. [CrossRef]
47. Squire, W.; Trapp, G. Using Complex Variables to Estimate Derivatives of Real Functions. SIAM Rev. 1998, 40, 110–112. [CrossRef]
48. Martins, J.R.R.A.; Sturdza, P.; Alonso, J.J. The complex-step derivative approximation. ACM Trans. Math. Softw. 2003, 29, 245–262.
[CrossRef]
49. Rosenblatt, M. Remarks on a Multivariate Transformation. Ann. Math. Stat. 1952, 23, 470–472. [CrossRef]
50. Zhang, Y.; Der Kiureghian, A. Two Improved Algorithms for Reliability Analysis. In Reliability and Optimization of Structural
Systems; Springer Science and Business Media LLC: Boston, MA, USA, 1995; pp. 297–304.
51. McDonald, M.; Mahadevan, S. Design Optimization With System-Level Reliability Constraints. J. Mech. Des. 2008, 130, 021403.
[CrossRef]