Decision Analytics Journal 3 (2022) 100043

Contents lists available at ScienceDirect

Decision Analytics Journal

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

Optimum design of truss structures by Material Generation Algorithm with

discrete variables
Mahdi Azizi a ,∗, Milad Baghalzadeh Shishehgarkhaneh b , Mahla Basiri a
a Department of Civil Engineering, University of Tabriz, Tabriz, Iran
b Department of Civil Engineering, Islamic Azad University of Tabriz, Tabriz, Iran

ARTICLE INFO ABSTRACT

Keywords: Optimization is an act of decision-making for reaching a point in which the overall behaviour of the considered
Material Generation Algorithm system is acceptable by the experts. In this paper, the optimum design of truss structures is considered utilizing
Truss structure the Material Generation Algorithm (MGA) as one of the recently developed metaheuristic algorithms in which
Metaheuristic algorithm
the basic and advanced principles of chemistry have been used as an inspirational concept. For numerical
Optimization
investigations, the benchmark problems in the truss optimization field, including the 25-bar, 72-bar and 200-
bar truss structures, are used while a penalty method is utilized accordingly for constraint handling purposes.
Multiple optimization runs are conducted for statistical purposes, and the results are compared to the results
of other metaheuristic algorithms. The obtained results proved that the MGA could provide very acceptable
and optimal design sections for the considered problems, which results in the lowest possible weight.

1. Introduction Optimal design of structures means using the latest methods and
techniques of modern structural science and earthquake to achieve the
Optimizing a system means minimizing or maximizing a function, most economical and safest possible design for the structure. This plan
measuring the system’s performance. Designers used to have to develop considers two factors of optimizing the behaviour of the structure and
many knowledge-based models to get a model with superior attributes. the project’s economy. Finally, in addition to the structure’s safety, it
Simultaneously, using intelligent approaches such as optimization al- significantly reduces construction costs and avoids wastage of natural
gorithms, these processes could well be completed quicker and more materials and resources. In recent decades, metaheuristic algorithms
accurately. Optimization solutions provided by experts are only pos- have been utilized as an optimization technique for the optimum design
sible for minimal problems, and the use of computers is essential of different structural systems. Kaveh, et al. [43] proposed an advanced
for the scientific problems we face daily. Metaheuristic algorithms
charged system search (ACSS) algorithm for optimizing the large-scale
are some searching techniques in which the higher-level methodolo-
frame structures using box-shaped sections for columns. Tahsin Öztürk,
gies are utilized to conduct searching procedures to reach optimal
et al. [44] utilized teaching–learning-based optimization to optimize
solutions. Particle Swarm Optimizer (PSO) [1], Grey Wolf Optimizer
reinforced concrete counterfort retaining walls. Azizi [45] applied the
(GWO) [2], Chaos Game Optimization (CGO) [3], Genetic Algorithm
improved charged system search in the optimization of the member-
(GA) [4], Atomic Orbital Search (AOS) [5], Ant colony Optimization
ship functions and the rule base of the fuzzy controller. Carvalho,
(ACO) [6], Crystal Structure Algorithm (CSA) [7], Material Generation
Algorithm (MGA) [8], Stochastic Paint Optimizer (SPO) [9], Dynamic et al. [46] investigated the optimum design of truss structures utilizing
Water Strider Algorithm (DWSA) [10], Remora Optimization Algorithm the differential evolution algorithm. Khodadadi, et al. [47] presented a
(ROA) [11], Artificial Gorilla Troops Optimizer (GTO) [12], African multi-objective version of Crystal Structure Algorithm (MOCryStAl) to
Vultures Optimization Algorithm (AVOA) [13], Gaze Cues Learning- solve numerous real-world engineering design problems. Li and Xu [48]
based Grey Wolf Optimizer (GGWO) [14], Whale Optimization Al- developed the improved version of the wolf pack algorithm and utilized
gorithm (WOA) [15], Ant Lion Optimizer (ALO) [16,17], Harmony these algorithms in the optimal design of truss structures. Kaveh,
Search (HS) algorithm [18], and Tribe-Charged System Search [19], et al. [49] used an advanced charged system search (ACSS) algorithm
are the most recent metaheuristic algorithms which have been devel- for the optimum design of steel structures. Mergos and Mantoglou [50]
oped for optimization purposes in different fields [20,21]. However, used a flower pollination algorithm for the optimal design of reinforced
Table 1 provides more information concerning different metaheuristic concrete retaining walls. Kaveh, et al. [51] investigated the applica-
algorithms. bility of multiple standard metaheuristic algorithms for the optimal

∗ Corresponding author.
E-mail addresses: [email protected] (M. Azizi), [email protected] (M. Baghalzadeh Shishehgarkhaneh),
[email protected] (M. Basiri).

https://doi.org/10.1016/j.dajour.2022.100043
Received 11 March 2022; Received in revised form 29 March 2022; Accepted 31 March 2022
Available online 5 April 2022
2772-6622/© 2022 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

Table 1
Historical to well-known metaheuristic optimization algorithms.
Name Authors Year Classifications Inspiration
Simulated annealing (SA) Kirkpatrick, et al. [22] 1983 Physics-based The annealing procedure of the metal
working
Genetic algorithm (GA) Holland [23] 1992 Evolutionary-based Charles Darwin’s theory of natural
evolution
Particle Swarm Kennedy and Eberhart 1995 Swarm-based The motion of bird flocks and schooling
Optimization (PSO) [24] fish
Harmony search (HS) Geem, et al. [18] 2001 Population-based The principles of the musicians’
improvisation of the harmony
Shuffled frog leaping Eusuff, et al. [25] 2006 Population-based Natural memetics
algorithm (SFLA)
River formation dynamics Rabanal, et al. [26] 2007 Swarm-based Drops of water from the river bed
(RFD)
Cuckoo search (CS) Yang and Deb [27] 2009 Swarm-based The brood parasitism of some cuckoo
species
Flower pollination Yang [28] 2012 Swarm-based The pollination behaviour of flowers
algorithm (FPA)
Golden Ball (GB) Osaba, et al. [29] 2014 Population-based Soccer concepts
Dynamic Virtual Bats Topal and Altun [30] 2016 Population-based Bat’s echolocation behaviour
Algorithm (DVBA)
Sine Cosine Algorithm Mirjalili [31] 2016 Population-based Sine and cosine functions
(SCA)
Grasshopper Optimization Saremi, et al. [32] 2017 Swarm-based The behaviour of grasshopper swarms in
Algorithm (GOA) nature
Harris hawks optimization Heidari, et al. [33] 2019 Population-based The cooperative behaviour and chasing
(HHO) style of Harris’ hawks
Group teaching Zhang and Jin [34] 2020 Population-based Group teaching mechanism
optimization algorithm
(GTOA)
Black Widow Optimization Hayyolalam and 2020 Population-based The unique mating behaviour of black
Algorithm (BWO) Pourhaji Kazem [35] widow spiders
Black Hole Mechanics Kaveh, et al. [36] 2020 Physics-based The black hole phenomenon
Optimization (BHMO)
Electron radar search Rahmanzadeh and 2020 Physics-based The electron discharge mechanism
algorithm (ERSA) Pishvaee [37]
Lévy flight distribution Houssein, et al. [38] 2020 Physics-based The Lévy flight random walk for
(LFD) exploring unknown large search spaces
Slime mould algorithm Li, et al. [39] 2020 Population-based The oscillation mode of slime mould in
(SMA) nature
Equilibrium Optimizer Faramarzi, et al. [40] 2020 Physics-based Control volume mass balance models
(EO) used to estimate both dynamic and
equilibrium states
Giza Pyramids Harifi, et al. [41] 2021 Population-based The ancient past
Construction (GPC)
Crystal Structure Algorithm Talatahari, et al. [7] 2021 Physics-based The principles underlying the formation
(CryStAl) of crystal structures
Atomic Orbital Search Azizi [5] 2021 Physics-based Principles of quantum mechanics
(AOS)
Ali Baba and the forty Braik, et al. [42] 2022 Population-based The famous tale of Ali Baba and the
thieves (AFT) forty thieves

design of large-scale truss structures. Daqiqnia, et al. [52] applied This paper considers the optimization of truss structures as one of
an intelligent methodology in a residential dwelling toward reducing the most recent challenges in the structural optimization field. The
energy. Kao, et al. [53] developed a new methodology based on the dif- Material Generation Algorithm (MGA) is utilized as the metaheuristic
ferential evolution algorithm for the optimal design of truss structures. optimization algorithm developed by Talatahari, et al. [8] recently,
Omidinasab and Goodarzimehr [54] investigated the optimal design employing the basic and advanced chemistry and material science
of truss structures by developing a hybrid metaheuristic algorithm by principles. Since the MGA algorithm is one of the recently proposed
combining the PSO and GA. Talatahari, et al. [55] proposed a new fuzzy metaheuristic algorithms, its capability in dealing with optimum design
adaptive charged system search (CSS) for global optimization in dealing of truss structures have not been evaluated; so, another critical purpose
with unconstraint, constrained, and engineering design problems. Ser- of this study could be assessing the performance of the MGA algorithm
pik [56] investigated structural truss systems’ optimum size and shape in reducing the weight of truss structures and comparing its results with
design through job search-inspired strategy and genetic operations. alternative algorithms. For numerical investigations, the benchmark
Almeida, et al. [57] investigated the optimum truss structure design problems in the truss optimization field, including the 25-bar, 72-bar
by utilizing the multi-material topology optimization method. Lee, and 200-bar truss structures, are used while a penalty method is utilized
et al. [58] applied the harmony search (HS) heuristic algorithm for accordingly for constraint handling purposes. Multiple optimization
structural size optimization problems. Khodadadi and Mirjalili [59] runs are conducted for statistical purposes, and the results are com-
used generalized normal distribution optimization (GNDO) algorithm pared to the results of other metaheuristic algorithms. Nonetheless,
for optimum design of truss structures. Kaveh, et al. [60] proposed the main advantages of the recently proposed MGA metaheuristic algo-
an efficient hybrid optimization algorithm based on invasive weed rithm are three-fold: fast convergence behaviour, being parameter-free,
optimization algorithm and shuffled frog-leaping algorithm for design and the lowest possible objective function evaluation. Furthermore, the
optimization of truss structures. Zheng, et al. [61] applied a hybrid in- MGA algorithm can provide competitive answers in dealing with truss
vasive weed optimization (HIWO) algorithm in dealing with economic structures’ optimum design compared to other answers reported in the
dispatch (ED) problems in power systems. last pieces of literature. In contrast, the MGA algorithm, like other

2
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

proposed metaheuristic algorithms, cannot provide exact solutions; in bonding (Fig. 1.b) involves the transfer of an electron. One of the
other words, the MGA is an approximate method in finding the best reacting atoms loses one or more electrons, and the other atom gains
and optimum solutions. In other words, the indispensable hypothesis one or more electrons, so the ionic bonding is stronger than hydrogen
of this study is whether the MGA algorithm can provide competitive and van der Waals bonding.
and reasonable answers in dealing with truss structures or not. The first position updating process in the MGA is conducted utilizing
the concept of the chemical compound in which the PTEs are assumed
2. Material Generation Algorithm to stand in the ground state so they can be affected by the lights
(photons), colliding with other moving particles and magnetic fields.
The matter is everything that includes physical objects and occupies The acts of these phenomena on PTEs result in gaining or losing elec-
space. The most common definition of matter is that matter is anything trons. The overall tendency of PTEs for participating in the mentioning
that has volume and mass. However, scientists still have no consensus process is modelled through a probability distribution function in a
on the exact definition of matter. Until the twentieth century, the continuous way which leads to a new material (Fig. 2). These aspects
term ‘‘matter’’ included ordinary matter made up of atoms and did not are mathematically modelled as follows:
include other energy phenomena such as light or sound. This notion [ 1 2 𝑘 𝑑
]
𝑀𝑎𝑡𝑛𝑒𝑤1 = 𝑃 𝑇 𝐸𝑛𝑒𝑤 𝑃 𝑇 𝐸𝑛𝑒𝑤 ⋯ 𝑃 𝑇 𝐸𝑛𝑒𝑤 ⋯ 𝑃 𝑇 𝐸𝑛𝑒𝑤 ,
of matter has now been extended to anything with mass, even at rest,
but these definitions are still inaccurate because the mass of an object k = 1, 2, … , 𝑑. (3)
itself can arise from the motion and interaction of energies (possibly 𝑘
𝑃 𝑇 𝐸𝑛𝑒𝑤
𝑟
= 𝑃 𝑇 𝐸𝑟12 ± 𝑒− , 𝑘 = 1, 2, … , 𝑑. (4)
without mass); Therefore, a universal definition and a fundamental
( ) 1 −(𝑥−𝜇)2
and comprehensive concept for it is not yet available in physics today. 𝑘
𝑓 𝑃 𝑇 𝐸𝑛𝑒𝑤 |𝜇, 𝜎 2 = √ .𝑒 2𝜎 2 , 𝑘 = 1, 2, … , 𝑑. (5)
The word matter is also freely used as a general term for anything 2𝜋𝜎 2
or all visible physical objects. Everything we can touch in everyday where 𝑟1 is a random integer number in the range of [1, 𝑛] and 𝑟2
life is made up of atoms. These materials consist of atoms, which are is another random integer number in the range of [1, 𝑑]; 𝑃 𝑇 𝐸𝑟12 is
𝑟
made up of interactions of subatomic particles, usually made up of randomly selected 𝑃 𝑇 𝐸 related to the existing solution candidate or
a nucleus containing protons and neutrons and a cloud of electrons material 𝑒− is a probabilistic factor for considering the process of losing,
in orbit around the nucleus. The MGA, one of the recently proposed sharing and gaining electrons; 𝑃 𝑇 𝐸𝑛𝑒𝑤 𝑘 is the new configuration of
metaheuristic algorithms, is formulated based on material science and decision variables for the newly generated material (𝑀𝑎𝑡𝑛𝑒𝑤1 ).
the general aspects of the material generations in nature in which the The chemical reaction is a process that leads to an interaction
chemical reaction and chemical compounds are in perspective. In the between chemicals in which the substances that initially participate in
first step, the initialization process is conducted in which the initial a chemical reaction as reactants produce products during the reaction.
values for the Periodic Table Elements (PTEs) as decision variables are Because a chemical change usually characterizes chemical reactions,
determined as follows: the products produced are generally different from the reactants. Clas-
⎡𝑀𝑎𝑡1 ⎤ ⎡𝑃 𝑇 𝐸1 𝑃 𝑇 𝐸1𝑗 ⎤
1 𝑃 𝑇 𝐸12 ⋯ ⋯ 𝑃 𝑇 𝐸1𝑑 sically, chemical reactions are changes that involve the rapid movement
⎢𝑀𝑎𝑡2 ⎥ ⎢⎢𝑃 𝑇 𝐸 1 𝑃 𝑇 𝐸22 ⋯ 𝑃 𝑇 𝐸2𝑗 ⋯ 𝑃 𝑇 𝐸2𝑑
⎥
⎥
of electrons to form and break chemical bonds. However, the general
⎢ ⎥ 2
⋮ ⎥ ⎢ ⋮ ⋮ ⋮ ⋱ ⋮ ⎥ concept of a chemical reaction, especially the concept of a chemical
𝑀𝑎𝑡 = ⎢ =⎢ ⎥,
⎢ 𝑀𝑎𝑡𝑖 ⎥ ⎢𝑃 𝑇 𝐸𝑖1 𝑃 𝑇 𝐸𝑖2 ⋯ 𝑃 𝑇 𝐸𝑖𝑗 ⋯ 𝑃 𝑇 𝐸𝑖𝑑 ⎥
equation, applies to elementary particle changes and nucleus reactions.
⎢ ⋮ ⎥ ⎢ ⋮ ⋮ ⋮ ⋱ ⋮ ⎥ Chemical reactions (Fig. 3) take place inside every living thing and
⎢ ⎥
⎣𝑀𝑎𝑡𝑛 ⎦ ⎢⎣𝑃 𝑇 𝐸 1 𝑃 𝑇 𝐸𝑛2 ⋯ 𝑃 𝑇 𝐸𝑛𝑗 ⋯ 𝑃 𝑇 𝐸𝑛𝑑 ⎥
⎦ allow the living thing to survive, grow and reproduce. In addition,
𝑛
{ chemical reactions produce many materials, including petrochemicals,
𝑖 = 1, 2, … , 𝑛. agricultural chemicals, ceramics, polymers and rubbers (elastomers),
(1)
𝑗 = 1, 2, … , 𝑑. natural oils, etc., while humankind is heavily dependent on these
( )
products.
𝑃 𝑇 𝐸𝑖𝑗 (0) = 𝑃 𝑇 𝐸𝑖,𝑚𝑖𝑛
𝑗 𝑗
+ 𝑈 𝑛𝑖𝑓 (0, 1) . 𝑃 𝑇 𝐸𝑖,𝑚𝑎𝑥 𝑗
− 𝑃 𝑇 𝐸𝑖,𝑚𝑖𝑛 ,
{ In the main loop of the MGA, the second position updating process
𝑖 = 1, 2, … , 𝑛. is conducted using the chemical reaction concept in material science
(2)
𝑗 = 1, 2, … , 𝑑. in which the generation of new materials is in perspective as follows
(Fig. 4):
where 𝑑 and 𝑛 are the problem dimensions and the initial number of ∑𝑙
materials respectively; 𝑃 𝑇 𝐸𝑖𝑗 (0) denotes the initial values for the jth 𝑚=1 (𝑝𝑚 .𝑀𝑎𝑡𝑚𝑗 )
𝑀𝑎𝑡𝑛𝑒𝑤2 = ∑𝑙 , 𝑗 = 1, 2, … , 𝑙. (6)
decision variable or periodic table element regarding the ith solution
𝑗 𝑗 𝑚=1 (𝑝𝑚𝑗 )
candidate or material; 𝑃 𝑇 𝐸𝑖,𝑚𝑖𝑛 is the lower and 𝑃 𝑇 𝐸𝑖,𝑚𝑎𝑥 is the upper
where 𝑀𝑎𝑡𝑚 denotes on the 𝑚th existing material; 𝑝𝑚 is a probability
bounds of the variables; and 𝑈 𝑛𝑖𝑓 (0, 1) is a random number.
function based on Gaussian distribution; and 𝑀𝑎𝑡𝑛𝑒𝑤2 denotes on the
The main loop of the MGA and the position updating process in
newly generated material.
this loop is developed based on the chemical compounds and chemical
The flowchart and pseudo-code of the MGA are illustrated in Figs. 5
reactions. Chemical bonds are the forces that hold atoms or molecules
and 6, respectively.
together and fall into two categories:
The ‘‘Big O notation’’, which is often used in computer science and
• Interatomic bonds: Covalent bond–ionic bond–metal bond is employed in this study for future research on the MGA, is one of the
• Intermolecular bonds: van der Waals force–hydrogen bond well-known methodologies for measuring the computational complex-
ity of algorithms. By indicating 𝑛 and 𝑑 as the total number of initial
These bonds can be between two identical atoms or two different solution candidates and the dimension of the optimization problem, re-
atoms; in the first case, it is called a Homonuclear molecule. In the spectively, the computational complexity of generating position vectors
second case, it is called a Heteronuclear molecule. The strength of a and calculating the objective function is O(𝑛 × 𝑑) and O(𝑛)×O(𝑓 (𝑥)),
chemical bond is determined by its electronegativity. The number of where 𝑓 (𝑥) denotes the optimization problem’s objective function. Each
chemical bonds in different molecules varies from one bond in simple line in the MGA’s main loop has a computational complexity of MT,
two-atom molecules to many bonds in macromolecules. A bond is called where MT is the total number of iterations. Because two new solution
a covalent bond (Fig. 1.a), in which atoms share their single electrons candidates are formed in each loop based on the chemical compound
while this bond is formed between non-metal and non-metal. An ionic and chemical reaction schemes, the position updating procedure for

3
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

Fig. 1. Schematic presentation of ionic (a) and covalent (b) compounds.

as follows:
∑
𝑒
𝑊 𝑒𝑖𝑔ℎ𝑡 (𝑨) = 𝜌𝑖 𝐼𝑖 𝐴𝑖 , 𝑖 = 1, 2, … , 𝑒. (7)
𝑖=1

𝛿𝑚𝑖𝑛 ≤ 𝛿𝑖 ≤ 𝛿𝑚𝑎𝑥 , 𝑖 = 1, 2, … , 𝑛. (8)

𝜎𝑚𝑖𝑛 ≤ 𝜎𝑖 ≤ 𝜎𝑚𝑎𝑥 , 𝑖 = 1, 2, … , 𝑒. (9)
𝜎𝑖𝑏
≤ 𝜎𝑖 ≤ 0, 𝑖 = 1, 2, … , 𝑛𝑐. (10)
{ }
𝑨 ∈ 𝑺 = 𝐴1 , 𝐴2 , … , 𝐴𝑖 (11)

where 𝑨 represents a vector including the cross-sectional area of the

design sections (𝐴𝑖 ); 𝜌𝑖 is the density of the utilized material; 𝐼𝑖 is the
length of the structural elements; 𝑛 and 𝑒 are the total numbers of
Fig. 2. Schematic presentation of generating a new material by chemical compounds. nodes and elements in the structure; 𝑛𝑐 represents the total number
of structural elements subjected to compressive loading; 𝜎𝑖 and 𝛿𝑖 are
the nodal stress and displacement in the structure; 𝜎𝑖𝑏 is the allowable
each of the solution candidates has a computational complexity of buckling stress, and 𝑺 is the predefined set of discrete cross-sectional
O(MT×𝑛×𝑑×2). Furthermore, Besides, for objective function evaluation areas.
in the main loop, the complexity of O(MT×𝑛 × 𝑑×2)×O(𝑓 (𝑥)) is found Since structural design optimization is a constraint optimization
out. problem, a constraint handling approach should be utilized for conduct-
ing the optimization procedure. For this purpose, a penalty function is
3. Problem statement determined as follows, which is part of the penalty constrain handling
method utilized in this paper:
In this section, a structural design optimization problem is for-
mulated in which a weight minimization procedure is conducted by 𝑓𝑝𝑒𝑛𝑎𝑙𝑡𝑦 (𝑨) = (1 + 𝜀1 .𝑣)𝜀2 × 𝑊 𝑒𝑖𝑔ℎ𝑡(𝑨) (12)
considering the required design constraints of the related codes and ∑
𝑞
{ }
standards. The objective function is regarded as the structure’s overall 𝑣= 𝑚𝑎𝑥 0, 𝑔𝑖 (𝑨) (13)
weight, while discrete design variables are considered for assigning 𝑖=1

predefined design sections to the structural elements during the opti- where 𝑞 represents the total number of design constraints; 𝑣 is the
mization process. The mathematical presentation of these aspects are summation of the violated design constraints; 𝑔𝑖 (𝑨) represents the ith

4
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

Fig. 3. Schematic presentation of a chemical reaction.

Fig. 4. Schematic presentation of generating a new material by chemical reactions.

design constraints; 𝜀1 and 𝜀2 denotes the control values for determining

the penalty during the optimization process.

4. Numerical investigations

In this section, the structural details of the truss structures and

the results of the optimization procedures are presented. A total of 30
independent optimization runs are conducted in each case for statistical
purposes, while the results of the MGA are compared to the result of
other metaheuristic approaches in the literature.

4.1. 25-bar truss structure

This truss structure has 25 members and ten nodes with stress and
displacement limitations of ±40 KSI and ±0.35 in, respectively. The
modulus of elasticity is 104 KSI, and the density of the utilized steel
material is 0.1 lb/in3. The discrete design variables for this problem
is as S = {0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.1, 1.2, 1.3,
1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.2, 2.3, 2.4, 2.6, 2.8, 3.0, 3.2,
3.4} (in.2) while the complete description of loading scenario and other
characteristics of this problem are provided by Ho-Huu, et al. [62] and
the schematic presentation of this structure is illustrated in Fig. 7.
The convergence history of the MGA in dealing with the 25-bar
truss design example is illustrated in Fig. 8, in which the convergence
curves for the best and all of the 30 independent optimization runs are
presented. Fig. 5. Flowchart of the MGA.

The best result of the multiple optimization runs by the MGA are
presented in Table 2 for the 25-bar truss problem in which the discrete
design variables are also provided for comparative purposes. The lowest multiple runs by MGA and the worst run are also very competitive
possible weight for the structure is calculated for the MGA. The results because the maximum number of function evaluations is set to 2000
of other alternative metaheuristics are derived from the literature for in this design example, which is a minimal value for this kind of
having a better perspective on the capability of the MGA. It is evident complex problem. Nonetheless, it is clear that almost all metaheuristic
that the MGA can provide 484.8542 lb for the 25-bar truss problem, algorithms and the MGA calculate the same value for the weight of
which is the best and lowest possible amount that has been calculated the 25-bar truss structure, hovering around 484 lb, not considering the
so far. Besides, the statistical results demonstrate that the mean of SGA. Regarding the mean weight of the 25-bar truss structure, the MBA

5
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

Fig. 6. Pseudo code of the MGA.

Fig. 7. 25-bar truss structure [9].

Fig. 9, in which the capability of the constraint handling approach is

in perspective.

4.2. 72-bar truss structure

This truss structure has 72 members and 20 nodes with stress

limitations of ±25 ksi. The modulus of elasticity is 104 ksi and the
density of the utilized steel material is 0.1 lb/in 3 . The discrete design
variables for this problem is as S = {0.111, 0.141, 0.196, 0.25, 0.307,
0.391, 0.442, 0.563, 0.602, 0.766, 0.785, 0.994, 1, 1.228, 1.266, 1.457,
1.563 1.62, 1.8 1.99, 2.13, 2.38, 2.62, 2.63, 2.88, 2.93 3.09, 1.13 3.38,
3.47, 3.55 3.63, 3.84, 3.87, 3.88, 4.1 4.22 4.49, 4.59, 4.8, 4.97 5.12,
5.74, 7.22, 7.97 8.53 9.3 10.85, 11.5, 13.5, 13.9, 14.2, 15.5, 16 16.9
18.8 19.9 22, 22.9, 24.5, 26.5, 28, 30, 33.5} (in.2) while the complete
description of loading scenario and other characteristics of this problem
Fig. 8. Convergence history of MGA for 25-bar truss structure.
are provided by Ho-Huu, et al. [62] and the schematic presentation of
this structure is illustrated in Fig. 10.
The convergence history of the MGA in dealing with the 72-bar
calculated the lowest value, accounting for 484.885 lb, followed by DE. truss design example is illustrated in Fig. 11, in which the convergence
Furthermore, the lowest standard deviation has been given by the MBA curves for the best and all 30 independent optimization runs are
algorithm in the 25-bar truss structure example; in contrast, the MGA presented.
algorithm can provide the SD of approximately 1.74 in this example. In Table 3, the best result of multiple optimization runs by the MGA
The design constraints, including the displacement and stress lim- alongside the statistical results is presented for the 72-bar truss prob-
itations for the best optimization run by the MGA are presented in lem. The discrete design variables are also provided for comparative

6
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

Table 2
Comparative results of MGA and other approaches in dealing with 25-bar truss problem.
Design SGA [63] HS [58] HPSO [64] ECBO [65] MBA [66] DE [62] AEDE [62] MGA
variables (in.2 )
𝐴1 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000
𝐴2 0.5000 0.3000 0.3000 0.3000 0.3000 0.3000 0.3000 0.3000
𝐴3 3.4000 3.4000 3.4000 3.4000 3.4000 3.4000 3.4000 3.4000
𝐴4 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000
𝐴5 1.5000 2.1000 2.1000 2.1000 2.1000 2.1000 2.1000 2.1000
𝐴6 0.9000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
𝐴7 0.6000 0.5000 0.5000 0.5000 0.5000 0.5000 0.5000 0.5000
𝐴8 3.4000 3.4000 3.4000 3.4000 3.4000 3.4000 3.4000 3.4000
Weight (lb) 486.2900 484.8500 484.8500 484.8500 484.8500 484.8540 484.8540 484.8542
Worst weight – – – – 485.0480 485.3800 486.1000 491.6923
(lb)
Mean weight – – – 485.8900 484.8850 484.9100 485.0140 487.0783
(lb)
Standard – – – – 0.0720 0.1310 0.2730 1.7437
deviation (lb)

HPSO: Heuristic Particle Swarm Optimization

MBA: Mine Blast Algorithm
SGA: Steady-State Genetic Algorithms
DE: Differential Evolution
AEDE: Adaptive Elitist Differential Evolution
ECBO: Enhanced Colliding bodies optimization
HS: Harmony Search Algorithm

Fig. 9. Displacement and stress design constraints for 25-bar truss problem.

Fig. 10. 72-bar truss structure [67].

purposes. By comparing the results of the MGA to other methods, it

can be concluded that the MGA can provide 389.3342 lb, which is the
around 390 lb, followed by the MGA. However, considering standard
lowest possible weight for this structure based on the reported results
deviation (SD), the lowest SD value of 0.8113 has been given by the
in the literature. In addition, the MGA has more acceptable statistical MGA algorithm, which is followed by IMBA with 0.84. Meanwhile, the
results, including the mean (390.1562 lb) and worst (392.3447 lb) CBO algorithm provided the highest value of SD (24.8) in the 72-bar
values. On the other hand, SGA calculated much more weight for the truss structure.
72-bar truss structure, registered at nearly 427 lb. WCA and IMBA could The design constraints, including the displacement and stress lim-
provide the lowest mean weight for the mentioned example, hovering itations for the best optimization run by MGA regarding the two

7
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

Table 3
Comparative results of MGA and other approaches in dealing with 72-bar truss problem.
Design SGA [63] DHPSACO [68] HPSO [64] MBA [66] CBO [69] ECBO [65] WCA [70] IMBA [70] DE [62] AEDE [62] MGA
variables (in.2 )
A1 0.1960 1.8000 4.9700 0.1960 1.6200 1.9900 1.9900 1.9900 1.9900 1.9900 1.9900
A2 0.6020 0.4420 1.2280 0.5630 0.5630 0.5630 0.4420 0.4420 0.5630 0.5630 0.5630
A3 0.3070 0.1410 0.1110 0.4420 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110
A4 0.7660 0.1110 0.1110 0.6020 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110
A5 0.3910 1.2280 2.8800 0.4420 1.4570 1.2280 1.2280 1.2280 1.2280 1.2280 1.2280
A6 0.3910 0.5630 1.4570 0.4420 0.4420 0.4420 0.5630 0.5630 0.4420 0.4420 0.5630
A7 0.1410 0.1110 0.1410 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110
A8 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110
A9 1.8000 0.5630 1.5630 1.2660 0.6020 0.5630 0.5630 0.5630 0.5630 0.5630 0.5630
A10 0.6020 0.5630 1.2280 0.5630 0.5630 0.5630 0.5630 0.5630 0.5630 0.5630 0.4420
A11 0.1410 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110
A12 0.3070 0.2500 0.1960 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110 0.1110
A13 1.5630 0.1960 0.3910 1.8000 0.1960 0.1960 0.1960 0.1960 0.1960 0.1960 0.1960
A14 0.7660 0.5630 1.4570 0.6020 0.6020 0.5630 0.5630 0.5630 0.5630 0.5630 0.5630
A15 0.1410 0.4420 0.7660 0.1110 0.3910 0.3910 0.3910 0.3910 0.3910 0.3910 0.3910
A16 0.1110 0.5630 1.5630 0.1110 0.5630 0.5630 0.5630 0.5630 0.5630 0.5630 0.5630
Weight (lb) 427.2030 393.3800 933.0900 390.7300 391.0700 389.3300 389.3340 389.3340 389.3340 389.3340 389.3342
Worst weight – – – 399.4900 495.9700 – 393.7780 389.4570 394.1700 393.3250 392.3447
(lb)
Mean weight (lb) – – – 395.4320 403.7100 391.5900 389.9410 389.8230 390.5310 390.9130 390.1562
Standard – – – 3.0400 24.8000 – 1.4300 0.8400 1.4000 1.1610 0.8113
deviation (lb)

DHPSAC: Harmony Search Algorithm

CBO: Colliding bodies optimization
WCA: Water cycle algorithm
IMBA: Improved Mine Blast Algorithm

Fig. 11. Convergence history of MGA for 72-bar truss structure.

prominent load cases of the 72-bar truss problem, are presented in

Fig. 12, in which the capability of the constraint handling approach
is in perspective.

4.3. 200-bar truss structure

This truss structure has 200 members and six nodes with stress
limitations of ±10 ksi, 30000 ksi as modulus of elasticity and 0.283
lb/in3 as the density of the utilized steel material. The discrete design
Fig. 12. Stress design constraints for 72-bar truss problem.
variables for this problem is as S = {0.100, 0.347, 0.440, 0.539,
0.954, 1.081, 1.174, 1.333, 1.488, 1.764, 2.142, 2.697, 2.800, 3.131,
3.565, 3.813, 4.805, 5.952, 6.572, 7.192, 8.525, 9.300, 10.850, 13.330,
Through a total of 30 independent optimization runs by MGA, the
14.290, 17.170, 19.180, 23.680, 28.080, 33.700} (in.2) while the
complete description of loading scenario and other characteristics of best optimum design scheme for the 200-bar truss problem is provided
this problem are provided by Ho-Huu, et al. [62] and the schematic in Table 4 alongside the statistical results. The discrete design variables
presentation of this structure is illustrated in Fig. 13. are also provided for comparative purposes in the former table. The
The convergence history of the MGA in dealing with the 200-bar lowest possible weight for the structure is calculated for the MGA.
truss design example is illustrated in Fig. 14, in which the convergence The results of other alternative metaheuristics are derived from the
curves for the best optimization run alongside the other 29 independent literature to have a better perspective on the capability of the MGA.
runs are presented. Based on the results, MGA can calculate 27415.2896 lb for the overall

8
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

results so far for this design example. In other words, the MGA al-
gorithm can provide quite competitive answers in dealing with the
200-bar truss structure. In stark contrast, the IGA algorithm calcu-
lated the highest weight for the 200-bar truss structure, accounting
for roughly 28544 lb, followed by the MABC. However, the MGA
algorithm can give superior results in the standard deviation to other
alternative metaheuristic algorithms in the 200-bar truss structure.
AEDE, meanwhile, provided the highest value of SD in this example.
The design constraints, including the stress limitations for the best
optimization run by the MGA, are presented in Fig. 15 for three
different load cases [62], in which the capability of the constraint
handling approach is in perspective.

5. Conclusion

The optimum design of truss structures is considered in this pa-

per. The Material Generation Algorithm (MGA), a recently developed
metaheuristic algorithm, is utilized as the optimization algorithm. For
numerical investigations, the benchmark problems in the truss opti-
mization field, including the 25-bar, 72-bar and 200-bar truss struc-
tures, are used while a penalty method is utilized accordingly for
constraint handling purposes. Multiple optimization runs are conducted
for statistical purposes, and the results are compared to the results of
other metaheuristic algorithms. The obtained results proved that the
MGA could provide very acceptable and optimal design sections for the
considered problems, which results in the lowest possible weight.
Regarding the 25-bar truss problem:

• MGA can provide 484.8542 lb, which is the best and lowest
possible amount that has been calculated so far.
• The statistical results demonstrate that the mean of multiple runs
by MGA and the worst run are also very competitive because the
maximum number of function evaluations is set to 2000 in this
design example, which is a minimal value for this kind of complex
problem.
Fig. 13. 200-bar truss structure [67].

Regarding the 72-bar truss problem:

• The MGA can provide 389.3342 lb, which is the lowest possible
weight for this structure based on the reported results in the
literature.
• The MGA has more acceptable statistical results, including the
mean (390.1562 lb) and worst (392.3447 lb) values.

Regarding the 200-bar truss problem:

• MGA can calculate 27415.2896 lb for the overall weight of this

structure, which is better than the previous results of the other
approaches, including the Adaptive Elitist Differential Evolution
(AEDE), which provides 27858.5000 lb, so the MGA demonstrates
better performance in this case.
• Regarding the statistical results, the MGA can provide the mean
Fig. 14. Convergence history of MGA for 200-bar truss structure. of 27880.5096 lb and worst of 28391.6911 lb, which are the
best-calculated results so far for this design example.

weight of this structure, which is better than the previous results of the
Declaration of competing interest
other approaches, including the Adaptive Elitist Differential Evolution
(AEDE), which provides 27858.5000, so the MGA demonstrates better
performance in this case. The authors declare that they have no known competing finan-
Regarding the statistical results, the MGA can give the mean of cial interests or personal relationships that could have appeared to
27880.5096 lb and worst of 28391.6911 lb, which are the best-calculated influence the work reported in this paper.

9
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

Table 4
Comparative results of MGA and other approaches in dealing with the 200-bar truss problem.
El. No. Members in the group IGA [71] HACOHS-T [72] ARCGA [73] MABC [73] ESASS [73] DE [62] AEDE [62] MGA
1 1, 2, 3, 4 0.3470 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000
2 5, 8, 11, 14, 17 1.0810 1.0810 1.0810 1.3330 0.9540 0.9540 0.9540 0.9540
3 19, 20, 21, 22, 23, 24 0.1000 0.3470 0.1000 0.1000 0.1000 0.3470 0.3470 0.4400
4 18, 25, 56, 63, 94, 101, 132, 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000
139, 170, 177
5 26, 29, 32, 35, 38 2.1420 2.1420 2.1420 2.6970 2.1420 2.1420 2.1420 2.1420
6 6, 7, 9, 10, 12, 13, 15, 16, 27, 0.3470 0.3470 0.3470 0.3470 0.3470 0.5390 0.3470 0.3470
28, 30, 31, 33, 34, 36, 37
7 39, 40, 41, 42 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000 0.1000
8 43, 46, 49, 52, 55 3.5650 3.1310 3.1310 3.1310 3.1310 3.5650 3.1310 3.5650
9 57, 58, 59, 60, 61, 62 0.3470 0.1000 0.1000 0.1000 0.1000 0.3470 0.3470 0.1000
10 64, 67, 70, 73, 76 4.8050 4.8050 4.8050 4.8050 4.8050 4.8050 4.8050 4.8050
11 44, 45, 47, 48, 50, 51, 53, 54, 0.4400 0.4400 0.3470 0.4400 0.3470 0.5390 0.5390 0.4400
65, 66, 68, 69, 71, 72, 74, 75
12 77, 78, 79, 80 0.4400 0.1000 0.1000 0.5390 0.1000 0.1000 0.3470 0.1000
13 81, 84, 87, 90, 93 5.9520 5.9520 5.9520 5.9520 5.9520 5.9520 5.9520 5.9520
14 95, 96, 97, 98, 99, 100 0.3470 0.1000 0.1000 0.1000 0.1000 0.3470 0.1000 0.1000
15 102, 105, 108, 111, 114 6.5720 6.5720 6.5720 6.5720 6.5720 6.5720 6.5720 6.5720
16 82, 83, 85, 86, 88, 89, 91, 92, 0.9540 0.5390 0.5390 1.0810 0.4400 0.9540 0.9540 0.5390
103, 104, 106, 107, 109, 110,
112, 113
17 115, 116, 117, 118 0.3470 1.1740 1.0810 0.3470 0.5390 0.3470 0.4400 0.9540
18 119, 122, 125, 128, 131 8.5250 8.5250 7.1920 8.5250 7.1920 8.5250 8.5250 8.5250
19 133, 134, 135, 136, 137, 138 0.1000 0.1000 0.5390 0.1000 0.4400 0.1000 0.1000 0.1000
20 140, 143, 146, 149, 152 9.3000 9.3000 8.5250 9.3000 8.5250 9.3000 9.3000 9.3000
21 120, 121, 123, 124, 126, 127, 0.9540 1.3330 1.3330 0.9540 0.9540 0.9540 0.9540 1.1740
129, 130, 141, 142, 144, 145,
147, 148, 150, 151
22 153, 154, 155, 156 1.7640 0.5390 1.0810 1.7640 1.1740 1.3330 1.0810 0.4400
23 157, 160, 163, 166, 169 13.3000 13.3300 10.8500 13.3300 10.8500 13.3300 13.3300 13.3300
24 171, 172, 173, 174, 175, 176 0.3470 1.1740 0.1000 0.4400 0.4400 0.3470 0.5390 0.3470
25 178, 181, 184, 187, 190 13.3000 13.3300 13.3300 13.3300 10.8500 13.3300 14.2900 13.3300
26 158, 159, 161, 162, 164, 165, 2.1420 2.6970 1.4880 2.1420 1.7640 2.1420 2.1420 1.4880
167, 168, 179, 180, 182, 183,
185, 186, 188, 189
27 191, 192, 193, 194 4.8050 3.5650 5.9520 3.8130 8.5250 3.8130 3.8130 4.8050
28 195, 197, 198, 200 9.3000 8.5250 13.3300 8.5250 13.3300 8.5250 8.5250 9.3000
29 196, 199 17.1700 17.1700 14.2900 19.1800 13.3300 17.1700 17.1700 17.1700
Weight (lb) 28544.0140 28030.2000 28347.5940 28366.3650 28075.4880 27901.5830 27858.5000 27415.2896
Worst weight – – – – – 29652.8910 29415.0000 28391.6911
(lb)
Mean weight – – – – – 28470.1140 28425.8710 27880.5096
(lb)
Standard – – – – – 457.4670 481.5900 233.7621
deviation (lb)

IGA: Improved Genetic Algorithm

HACOHS-T: Hybridized Ant Colony–Harmony Search-Genetic Algorithm
ARCGA: Adaptive Real-Coded Genetic Algorithm
MABC: Modified Artificial Bee Colony Algorithm
ESASS: Elitist Self-Adaptive Step-Size Search

Fig. 15. Stress design constraints for 200-bar truss problem.

10
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

References [27] X.-S. Yang, S. Deb, Cuckoo search via Lévy flights, in: 2009 World Congress on
Nature & Biologically Inspired Computing (NaBIC), IEEE, 2009, pp. 210–214.
[1] R. Eberhart, J. Kennedy, A new optimizer using particle swarm theory, in: [28] X.-S. Yang, Flower pollination algorithm for global optimization, in: International
MHS’95. Proceedings of the Sixth International Symposium on Micro Machine Conference on Unconventional Computing and Natural Computation, Springer,
and Human Science, 1995, pp. 39–43. 2012, pp. 240–249.
[2] S. Mirjalili, S.M. Mirjalili, A. Lewis, Grey wolf optimizer, Adv. Eng. Softw. 69 [29] E. Osaba, F. Díaz, E. Onieva, Golden Ball: A Novel meta-heuristic to solve
(2014) 46–61, http://dx.doi.org/10.1016/j.advengsoft.2013.12.007. combinatorial optimization problems based on soccer concepts, Appl. Intell. 41
[3] S. Talatahari, M. Azizi, Chaos game optimization: a novel metaheuristic algo- (2013) 145–166, http://dx.doi.org/10.1007/s10489-013-0512-y.
rithm, Artif. Intell. Rev. 54 (2) (2021) 917–1004, http://dx.doi.org/10.1007/ [30] A.O. Topal, O. Altun, Dynamic Virtual Bats Algorithm (DVBA) for global numer-
s10462-020-09867-w. ical optimization, in: 2014 International Conference on Intelligent Networking
[4] J.H. Holland, Adaptation in Natural and Artificial Systems: An Introductory and Collaborative Systems, 2014, pp. 320–327.
Analysis with Applications To Biology, Control, and Artificial Intelligence, MIT [31] S. Mirjalili, SCA: A Sine Cosine Algorithm For solving optimization problems,
Press, 1992. Knowl.-Based Syst. 96 (2016) 120–133, http://dx.doi.org/10.1016/j.knosys.2015.
[5] M. Azizi, Atomic orbital search: A novel metaheuristic algorithm, Appl. Math. 12.022.
Model. 93 (2021) 657–683, http://dx.doi.org/10.1016/j.apm.2020.12.021. [32] S. Saremi, S. Mirjalili, A. Lewis, Grasshopper Optimisation Algorithm: Theory
[6] M. Dorigo, V. Maniezzo, A. Colorni, Ant system: optimization by a colony of and application, Adv. Eng. Softw. 105 (2017) 30–47, http://dx.doi.org/10.1016/
cooperating agents, IEEE Trans. Syst. Man Cybern. B 26 (1) (1996) 29–41, j.advengsoft.2017.01.004.
http://dx.doi.org/10.1109/3477.484436. [33] A.A. Heidari, S. Mirjalili, H. Faris, I. Aljarah, M. Mafarja, H. Chen, Harris hawks
[7] S. Talatahari, M. Azizi, M. Tolouei, B. Talatahari, P. Sareh, Crystal structure optimization: Algorithm and applications, Future Gener. Comput. Syst. 97 (2019)
algorithm (CryStAl): A metaheuristic optimization method, IEEE Access 9 (2021) 849–872, http://dx.doi.org/10.1016/j.future.2019.02.028.
71244–71261, http://dx.doi.org/10.1109/ACCESS.2021.3079161. [34] Y. Zhang, Z. Jin, Group teaching optimization algorithm: A novel metaheuristic
[8] S. Talatahari, M. Azizi, A.H. Gandomi, Material generation algorithm: A novel method for solving global optimization problems, Expert Syst. Appl. 148 (2020)
metaheuristic algorithm for optimization of engineering problems, Processes 9 113246, http://dx.doi.org/10.1016/j.eswa.2020.113246.
(5) (2021) 859, http://dx.doi.org/10.3390/pr9050859. [35] V. Hayyolalam, A.A. Pourhaji Kazem, Black Widow Optimization Algorithm: A
[9] A. Kaveh, S. Talatahari, N. Khodadadi, Stochastic paint optimizer: theory and novel meta-heuristic approach for solving engineering optimization problems,
application in civil engineering, Eng. Comput. (2020) 1–32, http://dx.doi.org/ Eng. Appl. Artif. Intell. 87 (2020) 103249, http://dx.doi.org/10.1016/j.engappai.
10.1007/s00366-020-01179-5. 2019.103249.
[10] A. Kaveh, A. Dadras Eslamlou, N. Khodadadi, Dynamic water strider algorithm [36] A. Kaveh, M. Seddighian, E. Ghanadpour, Black hole mechanics optimization: a
for optimal design of skeletal structures, Period. Polytech. Civ. Eng. 64 (3) (2020) novel meta-heuristic algorithm, Asian J. Civ. Eng. 21 (2020) http://dx.doi.org/
904–916, http://dx.doi.org/10.3311/PPci.16401. 10.1007/s42107-020-00282-8.
[11] H. Jia, X. Peng, C. Lang, Remora optimization algorithm, Expert Syst. Appl. 185 [37] S. Rahmanzadeh, M. Pishvaee, Electron radar search algorithm: a novel devel-
(2021) 115665. oped meta-heuristic algorithm, Soft Comput. 24 (2020) 1–23, http://dx.doi.org/
[12] B. Abdollahzadeh, F. Soleimanian Gharehchopogh, S. Mirjalili, Artificial gorilla 10.1007/s00500-019-04410-8.
troops optimizer: A new nature-inspired metaheuristic algorithm for global [38] E.H. Houssein, M.R. Saad, F.A. Hashim, H. Shaban, M. Hassaballah, Lévy flight
optimization problems, Int. J. Intell. Syst. 36 (10) (2021) 5887–5958, http: distribution: A new metaheuristic algorithm for solving engineering optimization
//dx.doi.org/10.1002/int.22535. problems, Eng. Appl. Artif. Intell. 94 (2020) 103731, http://dx.doi.org/10.1016/
[13] B. Abdollahzadeh, F.S. Gharehchopogh. S. Mirjalili, African vultures optimization j.engappai.2020.103731.
algorithm: A new nature-inspired metaheuristic algorithm for global optimization [39] S. Li, H. Chen, M. Wang, A.A. Heidari, S. Mirjalili, Slime mould algorithm: A
problems, Comput. Ind. Eng. 158 (2021) 107408, http://dx.doi.org/10.1016/j. new method for stochastic optimization, Future Gener. Comput. Syst. 111 (2020)
cie.2021.107408. 300–323, http://dx.doi.org/10.1016/j.future.2020.03.055.
[14] M.H. Nadimi-Shahraki, S. Taghian, S. Mirjalili, H. Zamani, A. Bahreininejad, [40] A. Faramarzi, M. Heidarinejad, B. Stephens, S. Mirjalili, Equilibrium optimizer:
GGWO: Gaze Cues learning-based grey wolf optimizer and its applications for A novel optimization algorithm, Knowl.-Based Syst. 191 (2020) 105190, http:
solving engineering problems, J. Comput. Sci. 61 (2022) 101636, http://dx.doi. //dx.doi.org/10.1016/j.knosys.2019.105190.
org/10.1016/j.jocs.2022.101636. [41] S. Harifi, J. Mohammadzadeh, M. Khalilian, S. Ebrahimnejad, Giza pyramids
[15] S. Mirjalili, A. Lewis, The whale optimization algorithm, Adv. Eng. Softw. 95 construction: an ancient-inspired metaheuristic algorithm for optimization, Evol.
(2016) 51–67, http://dx.doi.org/10.1016/j.advengsoft.2016.01.008. Intell. 14 (2021) http://dx.doi.org/10.1007/s12065-020-00451-3.
[16] S. Mirjalili, P. Jangir, S. Saremi, Multi-objective ant lion optimizer: a multi- [42] M. Braik, M. Ryalat. H. Al-Zoubi, A novel meta-heuristic algorithm for solving
objective optimization algorithm for solving engineering problems, Appl. Intell. numerical optimization problems: Ali Baba and the forty thieves, Neural Comput.
46 (1) (2017) 79–95. Appl. 34 (2022) http://dx.doi.org/10.1007/s00521-021-06392-x.
[17] S. Mirjalili, The ant lion optimizer, Adv. Eng. Softw. 83 (2015) 80–98, http: [43] A. Kaveh, N. Khodadadi, B.F. Azar, S. Talatahari, Optimal design of large-scale
//dx.doi.org/10.1016/j.advengsoft.2015.01.010. frames with an advanced charged system search algorithm using box-shaped
[18] Z.W. Geem, J.H. Kim, G.V. Loganathan, A new heuristic optimization algorithm: sections, Eng. Comput. 37 (4) (2021) 2521–2541.
Harmony search, Simulation 76 (2) (2001) 60–68, http://dx.doi.org/10.1177/ [44] H. Tahsin Öztürk, T. Dede, E. Türker, Optimum design of reinforced concrete
003754970107600201. counterfort retaining walls using TLBO, Jaya Algorithm Struct. 25 (2020)
[19] S. Talatahari, M. Azizi, Tribe-charged system search for global optimization, 285–296, http://dx.doi.org/10.1016/j.istruc.2020.03.020.
Appl. Math. Model. 93 (2021) 115–133, http://dx.doi.org/10.1016/j.apm.2020. [45] M. Azizi, Optimization of fuzzy controller for nonlinear buildings with improved
12.007. charged system search, Struct. Eng. Mech. an Int’l J. 76 (6) (2020) 781–797.
[20] M. Azizi, R.G. Ejlali, S.A. Mousavi Ghasemi, S. Talatahari, Upgraded Whale [46] J.P.G. Carvalho, É.C.R. Carvalho, D.E.C. Vargas, P.H. Hallak, B.S.L.P. Lima, A.C.C.
Optimization Algorithm for fuzzy logic based vibration control of nonlinear steel Lemonge, Multi-objective optimum design of truss structures using differential
structure, Eng. Struct. 192 (2019) 53–70, http://dx.doi.org/10.1016/j.engstruct. evolution algorithms, Comput. Struct. 252 (2021) 106544, http://dx.doi.org/10.
2019.05.007. 1016/j.compstruc.2021.106544.
[21] S. Talatahari, M. Azizi, An extensive review of charged system search algorithm [47] N. Khodadadi, M. Azizi. S. Talatahari, P. Sareh, Multi-objective crystal structure
for engineering optimization applications, in: S. Carbas, A. Toktas, D. Ustun algorithm (MOCryStAl): Introduction and performance evaluation, IEEE Access
(Eds.), Nature-Inspired Metaheuristic Algorithms for Engineering Optimization 9 (2021) 117795-117812, http://dx.doi.org/10.1109/ACCESS.2021.3106487.
Applications, Springer Singapore, Singapore, 2021, pp. 309–334. [48] Y.-C. Li, P.-D. Xu, Improved wolf pack algorithm for optimum design of truss
[22] C.D. Kirkpatrick, S. Gelatt Jr, M.P. Vecchi, Optimization by simulated annealing, structures, Civ. Eng. J. 6 (2020) 1411–1427, http://dx.doi.org/10.28991/cej-
Science 220 (4598) (1983) 671–680, http://dx.doi.org/10.1126/science.220. 2020-03091557.
4598.671. [49] A. Kaveh, N. Khodadadi, S. Talatahari, A comparative study for the optimal
[23] J.H. Holland, Genetic algorithms and adaptation, in: O.G. Selfridge, E.L. Rissland, design of steel structures using CSS and ACSS algorithms, Iran Univ. Sci. Technol.
M.A. Arbib (Eds.), Adaptive Control of Ill-Defined Systems, Springer US, Boston, 11 (1) (2021) 31–54.
MA, 1984, pp. 317–333. [50] P.E. Mergos, F. Mantoglou, Optimum design of reinforced concrete retaining
[24] J. Kennedy, R. Eberhart, Particle swarm optimization, in: Proceedings of walls with the flower pollination algorithm, Struct. Multidiscip. Optim. 61 (2)
ICNN’95-International Conference on Neural Networks, IEEE, 1995, pp. (2020) 575–585, http://dx.doi.org/10.1007/s00158-019-02380-x.
1942–1948. [51] A. Kaveh, K. Biabani Hamedani, F. Barzinpour, Optimal size and geometry design
[25] M. Eusuff, K. Lansey, F. Pasha, Shuffled frog-leaping algorithm: a memetic of truss structures utilizing seven meta-heuristic algorithms: a comparative study,
meta-heuristic for discrete optimization, Eng. Optim. 38 (2) (2006) 129–154, Int. J. Optim. Civ. Eng. 10 (2) (2020) 231–260.
http://dx.doi.org/10.1080/03052150500384759. [52] A.H. Daqiqnia, S. Fard Moradinia, M. Baghalzadeh Shishehgarkhaneh, Toward
[26] P. Rabanal, I. Rodríguez, F. Rubio, Using River Formation Dynamics To Design nearly zero energy building designs: A comparative study of various techniques,
Heuristic Algorithms, Springer Berlin Heidelberg, Berlin, Heidelberg, 2007, pp. AUT J. Civ. Eng. 5 (2) (2021) http://dx.doi.org/10.22060/ajce.2021.20458.5771,
163–177. 12-12.

11
M. Azizi, M. Baghalzadeh Shishehgarkhaneh and M. Basiri Decision Analytics Journal 3 (2022) 100043

[53] C.-Y. Kao, S.-L. Hung, B. Setiawan, Two strategies to improve the differential [64] L.J. Li, Z.B. Huang, F. Liu, A heuristic particle swarm optimization method for
evolution algorithm for optimizing design of truss structures, Adv. Civ. Eng. truss structures with discrete variables, Comput. Struct. 87 (7) (2009) 435–443,
2020 (2020) 8741862, http://dx.doi.org/10.1155/2020/8741862. http://dx.doi.org/10.1016/j.compstruc.2009.01.004.
[54] F. Omidinasab, V. Goodarzimehr, A hybrid particle swarm optimization and [65] A. Kaveh, M. Ilchi Ghazaan, A comparative study of CBO and ECBO for optimal
genetic algorithm for truss structures with discrete variables, J. Appl. Com-
design of skeletal structures, Comput. Struct. 153 (2015) 137–147, http://dx.doi.
put. Mech. 6 (3) (2020) 593–604, http://dx.doi.org/10.22055/jacm.2019.28992.
org/10.1016/j.compstruc.2015.02.028.
1531.
[66] A. Sadollah, A. Bahreininejad, H. Eskandar, M. Hamdi, Mine blast algorithm for
[55] S. Talatahari, M. Azizi, M. Toloo, Fuzzy adaptive charged system search for
global optimization, Appl. Soft Comput. 109 (2021) 107518, http://dx.doi.org/ optimization of truss structures with discrete variables, Comput. Struct. 102–103
10.1016/j.asoc.2021.107518. (2012) 49–63, http://dx.doi.org/10.1016/j.compstruc.2012.03.013.
[56] I. Serpik, Discrete size and shape optimization of truss structures based on job [67] N. Khodadadi, V. Snasel. S. Mirjalili, Dynamic arithmetic optimization algorithm
search inspired strategy and genetic operations, Period. Polytech. Civ. Eng. 64 for truss optimization under natural frequency constraints, IEEE Access 10 (2022)
(3) (2020) 801–814, http://dx.doi.org/10.3311/PPci.11840. 16188–16208.
[57] C. Almeida, F. Conde, P. Coelho, T. Pratas, Stiffness and strength-based [68] S. Talatahari, M. Azizi, H. Veladi, A. Moutabi-Alavi, S. Rahnema, Optimum
lightweight design of truss structures using multi-material topology optimization, structural design of full-scale steel buildings using drift-tribe-charged system
2021.
search, Earthq. Eng. Eng. Vib. (2021).
[58] K.S. Lee, Z.W. Geem, S.-h. Lee, K.-w. Bae, The harmony search heuristic algo-
[69] A. Kaveh, V.R. Mahdavi, Colliding bodies optimization method for optimum
rithm for discrete structural optimization, Eng. Optim. 37 (7) (2005) 663–684,
http://dx.doi.org/10.1080/03052150500211895. discrete design of truss structures, Comput. Struct. 139 (2014) 43–53, http:
[59] N. Khodadadi, S. Mirjalili, Truss optimization with natural frequency constraints //dx.doi.org/10.1016/j.compstruc.2014.04.006.
using generalized normal distribution optimization, Appl. Intell. (2022) http: [70] A. Sadollah, H. Eskandar, A. Bahreininejad, J.H. Kim, Water cycle, mine
//dx.doi.org/10.1007/s10489-021-03051-5. blast and improved mine blast algorithms for discrete sizing optimization of
[60] A. Kaveh, S. Talatahari, N. Khodadadi, Hybrid invasive weed optimization- truss structures, Comput. Struct. 149 (2015) 1–16, http://dx.doi.org/10.1016/j.
shuffled frog-leaping algorithm for optimal design of truss structures, Iran. J. compstruc.2014.12.003.
Sci. Technol. Trans. Civ. Eng. 44 (2) (2020) 405–420. [71] V. Toğan, A.T. Daloğlu, An improved genetic algorithm with initial population
[61] Z.-x. Zheng, J.-q. Li, H.-y. Sang, A hybrid invasive weed optimization algorithm
strategy and self-adaptive member grouping, Comput. Struct. 86 (11) (2008)
for the economic load dispatch problem in power systems, Math. Biosci. Eng. 16
1204–1218, http://dx.doi.org/10.1016/j.compstruc.2007.11.006.
(4) (2019) 2775–2794.
[72] M.H. Talebpour, A. Kaveh, V. Kalatjari, Optimization of skeletal structures using
[62] V. Ho-Huu, T. Nguyen-Thoi, T. Vo-Duy, T. Nguyen-Trang, An adaptive elitist
differential evolution for optimization of truss structures with discrete de- a hybridized ant colony-harmony search-genetic algorithm, Iran. J. Sci. Technol.
sign variables, Comput. Struct. 165 (2016) 59–75, http://dx.doi.org/10.1016/ Trans. Civ. Eng. 38 (2014) 1–20.
j.compstruc.2015.11.014. [73] S.K. Azad, O. Hasan¸ cebi, An elitist self-adaptive step-size search for structural
[63] S.-J. Wu, P.-T. Chow, Steady-state genetic algorithms for discrete optimization design optimization, Appl. Soft Comput. 19 (2014) 226–235, http://dx.doi.org/
of trusses, Comput. Struct. 56 (6) (1995) 979–991, http://dx.doi.org/10.1016/ 10.1016/j.asoc.2014.02.017.
0045-7949(94)00551-D.

12