See discussions, stats, and author profiles for this publication at: [Link]

net/publication/282574837

Optimization of Steel Structures Using Particle

Swarm Technique

Article · January 2014

CITATIONS READS

0 149

4 authors, including:

O. El-Mahdy Osama Ahmed Kamal

Benha University Benha University
31 PUBLICATIONS 3 CITATIONS 31 PUBLICATIONS 185 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Master and PhD thesis View project

All content following this page was uploaded by O. El-Mahdy on 18 April 2016.

The user has requested enhancement of the downloaded file.

Optimization of Steel Structures Using Particle Swarm Technique

O. KAMAL, O. EL-MAHDY, M. Nour** and G. EL-KOMY 


Professor,  Assistant Professor, Graduate Student
Faculty of Engineering at Shoubra, Benha University, Cairo, Egypt

ABSTRACT

The computational drawbacks of existing numerical methods have forced researchers to rely
on heuristic algorithms. Heuristic methods are powerful in obtaining the solution of
optimization problems. The Particle Swarm Optimization (PSO) method is a numerical
optimization technique that simulates the social behavior of birds, fishes and bugs. Similar to
birds seek to find food, the optimum design process seeks to find the optimum solution. In the
present study, a combination between nominal PSO and Gaussian PSO is applied to get the
optimum design for steel structures. A modified fly back technique is used to deal with
constraints. Standard design problems selected from literatures are used to evaluate the
performance of the proposed technique. The comparison showed that design obtained using
the present algorithm is more efficient and economical than that provided by other design
approaches.

KEYWORDS
Particle Swarm Optimization (PSO); Steel; Frames; Design Codes; Gauss Distribution; Fly
Back Mechanism

1. INTRODUCTION

Structural design optimization is a critical and challenging activity that has received
considerable attention in the last two decades. High number of design variables, largeness of
the search space and controlling great number of design constraints are major preventive
factors in performing optimum design in a reasonable computation time. Despite these facts,
designers and owners have always desired to have optimal structures.

The developments in the stochastic search techniques in numerical optimization have

provided efficient optimum design tools for structural designers. The basic idea behind these
stochastic search techniques is to simulate the natural phenomena. These stochastic search
techniques do not require the derivatives of the objective function and constraints and they
use probabilistic transition rules not deterministic rules [1]. The particle swarm optimization
(PSO), that was only a few years ago a curiosity, has now attracted the interest of researchers
around the globe.

In the present study, a combination between nominal PSO and Gaussian PSO is applied in
a new technique to get the optimum design for steel structures. The development of the paper
is as follows: Section 2 presents the general formulation of the particle swarm optimization
approach. Section 3 literature review of PSO and structural steel design are presented.
Section 4 presents the use of Gauss Distribution in PSO. In Section 5, the problem
formulation and constraint handling technique are discussed. In section 7, three different
case studies are analyzed to demonstrate the effectiveness of the approach in finding optimal
structural optimization solutions. Finally, the paper closes with main concluding remarks.

1
2. PARTICLE SWARM OPTIMIZATION

In the simplest version of PSO, each member of the particle swarm is moved through a
problem space by two elastic forces. One attracting it with random magnitude to the best
location so far encountered by the particle, it is called Lbest (local best). The other attracting it
with random magnitude to the best location encountered by any member of the swarm which
is typically denoted by Gbest (global best). The position and velocity of each particle are
updated at each time step until the swarm as a whole converges to an optimum [2 and 3].

The original PSO formulae as described in [4] are:

  
Vi t 1  Vi t  c1 r1t pit  xit  c 2 r2t p g  xit  (2.1)

xit 1  xit  Vi t 1 (2.2)

Where:
 i= (1, 2, 3…………, N), and N is the number of particles.
 c1 and c2 are positive constants, referred to as cognitive and social parameters,
respectively.
t t
 r1 and r2 are random vectors with components uniformly distributed in (0, 1).
 xit = ( x1t , x2t ,… xDt ), the location of the i th particle at t th iteration, and D is the
number of design variables.

Vi t = ( V1t , V2t ,…., VDt ), the velocity vector of the ith particle at tth iteration.

 pit = ( p1t , p2t ,…., pDt ), represents the best previous position vector of the ith
particle at tth iteration.
pg
 = ( p1 , p2 ,…., pD ), the symbol g represents the index of the best particle
among all the particles in the population.

The initial values of xit and Vit can be assigned randomly. The initial value for Pit will be
xit. The initial value of Pg can be determined by evaluating all particles in the group and
selecting the initial position of the best particle. Figure (1) represents the velocity updating in
PSO.

Another parameter called constriction factor (k) is introduced to insure a convergence of

PSO. A simplified method of incorporating it appears in the following equations:

   
Vi t 1  k[.Vi t  c1 r1 . p it  x it  c 2 r2 . p gt  x it ] (2.3)

2
k (2.4)
2     2  4

Where
  c1  c 2 , and   4 (2.5)

2
 Y xt+1
Vgbest
Vt+1
Vpbest

xt Vt

X
xt : current location of particle,
xt+1: new location of particle,
Vt : current velocity,
Vt+1: modified velocity,
Vpbest : velocity based on local best,
Vgbest : velocity based on global best

Fig. (1): PSO Position Updating

Mathematical back ground for Eqs. (2.3, 2.4 and 2.5) is widely discussed and proofed by
Ozcan, Mohan, Clerc and Kennedy [5 and 6].

3. LITERATURE REVIEW OF PSO AND STRUCTURAL STEEL DESIGN

In the following paragraphs the most important researches who employ the PSO in the
optimization of structural design are presented.

Fourie and Groenwold [7] applied the particle swarm optimization algorithm (PSOA) to
the optimal design of structures with sizing and shape variables. Some new operators, namely
elite particle and elite velocity, had been used.

The application of the PSO to the optimal sizing design of truss structures was studied by
Schutte, and Groenwold, [8]. A simple methodology was presented to accommodate the
stress and displacement constraints during initial iterations. The construction factor and
dynamic inertia weight were applied to PSO and a penalty function was used as a constraints
handling technique.

Bochenek and Foryś [9] considered structural optimization against instability using PSO.
The standard maximization of critical load was performed both for single and double
buckling load. The modified optimization for post-buckling behavior was also performed.

Design and optimization of steel structures for fire resistance was also considered by
Jármai, et al. [10] using modified PSO. A pressure vessel supporting frame was considered
using square hollow section columns and square or rectangular hollow section for beams.
Overall and local buckling constraints were considered and the steel frame was designed
according to Eurocodes (1) and (3).

The early work of use of PSO to optimize the design of steel structure was presented by
Heinisuo, M. et al. [11]. In their work, the design of welded steel beams for a typical structure
was considered.

3
 The tubular truss optimization using heuristic algorithm had been considered by Jalkanen
[12]. The multi-criteria topology, shape and sizing optimization problem were formulated
based on both material and manufacturing cost. Design constraints satisfied the requirements
of steel design rules of Eurocode (3).

Kaveh and Talatahari presented a Discrete Particle Swarm Ant Colony Optimization
(DPSACO) for design of steel truss [13] and steel frame [14]. The particle swarm ant colony
optimization applied the particle swarm optimizer with passive congregation for global
optimization and ant colony optimization worked as a local search.

Later, Kaveh and Talatahari [15 and 16] presented a hybrid algorithm based on the Particle
Swarm Optimization with Passive Congregation (PSOPC), the Ant Colony Algorithm
(ACO), and the Harmony Search (HS) approach to solve engineering optimization problems.
The idea of ACO worked as a local search and HS utilized to handle the boundary
constraints.

A refined version of particle swarm optimization technique for the optimum design of
steel trusses was proposed by Doğan et al. [17]. In their work, an additional velocity term was
defined and added to let the particle move randomly in certain directions in the close
neighborhood of its current position, to avoid the stuck in local optimum. The penalty
function technique was used to handle the ASD-AISC code constraints.

The composite steel box girder was investigated under size, shape and topology
optimization using Particle Swarm Optimization by Ghasemi and Dizangian [18]. The
objective function was the minimization of total weight of the structure under strength and
serviceability constraints, enforced by penalty functions. All design requirements of
AASHTO and Iranian Codes of Practice (ICP) for loading of bridges were considered.

A parallel version for PSO optimizer was studied by Yang and Zhang, [19] to solve large
truss topological optimization problems. It was the first modified PSO that can solve truss
topology optimization problem with more than 200 design variables. This modified PSO was
based on Lbest model of PSO. In their model, the information linked between different
particles were set by some topology technique, i.e. one particle could share the information
only with some particles like its neighbors, not with all the other ones. The Lbest model tried
to prevent premature convergence by maintaining diversity of potential problem solutions.

Particle Swarm method based optimum design algorithm for unbraced steel frames was
presented by Doğan and Saka [1]. In the optimum design algorithm the design constraints
were imposed in accordance with LRFD-AISC. In the design of beam-column members the
combined strength constraints were considered by takeing into account the lateral tensional
buckling of the member. The algorithm developed selected optimum W sections for beams
and columns of unbraced frame from 272 W-sections list. This selection was carried out such
that design constraints imposed by the LRFD-AISC were satisfied and the minimum frame
weight was obtained.

Kaveh et al. [20] introduced a new hybrid advanced algorithm by using the abilities of
heuristic Particle Swarm Ant Colony Optimization (HPSACO) and a Hybrid Big Bang–Big
Crunch algorithm (HBB–BC). In their study, the advantages of the HPSACO and HBB–BC
ere combined to improve the performance of the resulted algorithm. They considered three
main steps as global searching step, local searching step and location controlling step. These

4
steps all together improved the exploration and exploitation abilities of the algorithm. Their
proposed method was tested on frame structures from the literatures.

A modified algorithm called Accelerated Particle Swarm Optimization (APSO) was

developed by Talatahari et al. [21] for finding optimum design of steel frame structures.
APSO showed some extra advantages in convergence for global search. The modifications on
standard PSO effectively accelerated the convergence rate of the algorithm and improved the
performance of the algorithm in finding better optimum solutions.

4. GAUSS-PSO COMBINATION

As a member of stochastic search algorithms, PSO has drawbacks. Although PSO

converges to an optimum much faster than other evolutionary algorithms, it usually cannot
improve the quality of the solutions as the number of iterations is increased [22]. Many
adjustments have been made to the original algorithm that introduced in [4] over the past
decade in order to improve general performance by incorporate either the capabilities of other
evolutionary computation techniques. Some researchers have studied a version that
eliminates the current position term, and simply places the particle at a new position based on
previous successes.

One of the important modifications on PSO is the using of Gauss Distribution instead of
the velocity and position update equations of PSO. It is greatly simplifies the particles swarm
by stripping away the velocity rule, but performance seems not good as canonical one in
some tested problems [23].

In Bare-Bones formulations of PSO, which described by Kennedy [24] particles proposed

to move according to a probability distribution rather than through the addition of velocity,
which called “velocity-free” PSO.

In a first study of bare-bones PSO, the particle update rule was replaced with a Gaussian
distribution of mean (Pi+Pl)/2 and standard deviation |Pi-Pl|, thus Eqs. (2.1) and (2.2) will be
replaced by Eq. (2.6) [13 to 18]:
 ( P  Pl 
x it 1  G i , Pi  Pl  (2.6)
 2 

where xit+1 is the position of the particle to be updated, G (mean, s.d.) is a Gaussian random
number generator , pi and pl are the best position reached by the ith particle and the best
position reached by any particle in the neighborhood of the ith particle, respectively.

In the present paper, the authors introduce a new strategy to deal with the Gauss
Distribution and PSO by combining them in the search. First, fast search is started using the
normal PSO, which give a fast trapped to local optimum after a number of iterations, then the
Gaussian-PSO starts the fine tuning search. This strategy gives good results and fast
convergence compared to using normal PSO only. Furthermore, Gaussian-PSO is more
simple than the other hybridization methods.

5
5. PROBLEM FORMULATION AND CONSTRAINTS HANDLING

The main concept in the design of steel frames is to find the sections for columns and
beams. The design should be carried out in such a way that the frame satisfies the
serviceability and strength requirements specified by the design code of practice while the
economy is observed in the overall or material cost of the frame. These are the design
variables, constraints and objective function, respectively.

Similar to other stochastic optimization methods, the PSO algorithm is defined for
unconstrained problems [25]. One of the most simple and effective method to deal with
constraints is the fly-back mechanism, which purposed firstly by Hu et al. [26].

In the current study, a modified fly-back mechanism is used to handle the problem
specific constraints which can be described as follows:

1- During the initialization process:

a. All particles are started with feasible solutions.
2- When updating the global best memory (Gbest):
a. If the new solution is infeasible, return back to the previous feasible solution.
b. Between two feasible solutions, the one having better objective function value
is preferred.
3- When updating the personal best memories (Pbest):
a. Any feasible solution is preferred to any infeasible solution.
b. Between two feasible solutions, the one having better objective function value
is preferred.
c. Between two infeasible solutions, the one having smaller sum of constraint
violation is preferred.

This modified technique is powerful in dealing with constraints and more simplified from
other techniques such as penalty function, as no additional parameters are needed for the
applied constraints – handling technique.

The boundary conditions are the limitations of sections area according to slandered tables
of steel sections, or the thicknesses of steel plates available in the market in case of built up
section. In the present work, before updating memories and check constraints, each design
variable passes through a filter which works as follows:
1- Pass the design variable if it satisfying the boundary conditions.
2- Otherwise, regenerate the design variable to satisfying the boundary conditions.

Algorithmic flow chart of the proposed technique is shown in Fig. 2

6
 Start
N= No. of Particles
D=No. of Design Variables
Set N, D, C1, C2, C1=C2=2.1
r1, r2, T, G R1 & r2 are random numbers
T= No. of Iterations
G= Duration of Normal PSO
Randomly Initialize
Particle Velocities Xi
with feasible Values

No
If T=1 If T > G

Yes
No Update new
Particle’s
Gauss Positions
T=T+1 Normal PSO
Distribution and
Velocities

Yes

Check if the
Regenerate Satisfying
Boundary Limits No
Limits
are Satisfied
Set New
r1 & r2 Yes

Structure Analysis
Evaluate Objective
Program to Get
Function (Fitness) F(x)
Straining Actions

Update The Personal Evaluate Violation of

Best Lbest Constraints
No Update The Global Best
Gbest
Constraints Handling
(Modified Fly Back
Mechanism)
Check
Stopping
Criteria
Yes

END

Fig. (2): Algorithmic Flow Chart

6. NUMERICAL EXAMPLES

To demonstrate the efficiency of the algorithm, three well-known examples selected from
the literatures are designed using the proposed method. In these designs, the duration of the
normal PSO is considered as the number of iterations in which the algorithm use PSO in
search before starting Gauss search (denoted as G). The value of C1 and C2 in Eq. (2.1) are
considered to be 2.1 and construction factor K is used.

7
Example No. (1): Two Bars Truss

The first example is a benchmark example introduced in [7 and 27]. It is a simple truss
structure depicted in Fig. 3. The member cross sections represent the design variables. Arora
[27] gives an analytical solution of f = 8.046×103 kg which is the same truss weight obtained
using the purposed algorithm. A number of trials has been done to investigate the effect of
the duration of normal PSO, each trial has 400 iterations and take less than 5 minutes. The
convergence history is showed in Table 1 and Figs. 4 and 5.

Table 1: Iterations Results for Example No. (1)

No. of No. of Duration Start Optimum Iteration of

Trial
Iterations Particles of PSO Weight Weight (Kg) Optimum
No.
(T) (N) (G) (Kg) Weight

1 400 5 50 14822.19 8046.24 150

2 400 5 100 18061.35 8046.31 234
3 400 5 150 16109.72 8046.24 291
4 400 5 200 15771.68 8046.38 259
5 400 5 100 14822.19 8046.28 225
6 400 10 100 14577.46 8046.24 145
7 400 15 100 10699.74 8046.24 152
8 400 20 100 11586.28 8046.24 128

E = 207 GPa
ρ = 7850 kg/m3
Fy = 160 MPa
Amin = 0.01 m2

0.3 m

0.4 m
P=1200 KN

Fig. (3): Two Bars Truss

8
 Fig. (4): Iteration History for Example No. 1.
(No. of particles =5 and changing the duration of PSO search)

Fig. (5): Iteration History for Example No. 1.

(No. of particles varied and the duration of PSO search is 100 iterations)

Example No. (2): Two – bay Three Storey Steel Frame

This Frame was considered by Kaveh and Malakoutirad [28]. Figure (6) shows the
configuration and applied loads of the frame. The 15 members of the structure have been
categorized into seven groups. Each group has the same cross section and represents a design
variable as indicated in the figure.

9
The material properties are:
E = 200 GPa
ρ = 7850 kg/m3
Fy= 235 MPa

a) Loading of the frame b) Member grouping

Fig. (6): Two-bay Three-Storey Frame

Fig. (7): Iteration History for Example No. 2.

(No. of particles = 5 and the duration of PSO search is 100 iterations)
The constraints include the design specifications of AISC-ASD for each member. The
allowable inter-storey drift is 12 mm and allowable sway of the top storey is 36 mm. The
population sizes considered by Kaveh was 425 particles and 5000 iterations were used.
Figure (8) shows the iteration history of the compared example.

10
 Figure (7) shows the iteration history for this example using current purposed algorithm.
This example demonstrates the effeteness of Gauss distribution on the search. In this work,
only 5 particles and 200 iterations are used to obtain the same results with a total running
time less than 26 minutes.

Fig. (8). Iteration History for Example No. 2 as shown in [28]

Example No. (3): Two – bay Six Storey Steel Frame

The third example is two-bay, six storey steel frame shown in Fig. (9). This frame was
considered firstly by Kameshki [29] using GA, and Kaveh et al. [16] using Big – Bang
algorithm and later considered by Dogan and Saka [1] using PSO. The frame consists of
thirty members that are collected in eight groups as shown in the figure. The allowable inter-
storey drift is 12 mm while the lateral displacement of the top storey is limited to 72 mm. The
modulus of elasticity is 200 GPa.

According to Dogan and Saka, the optimum design is determined after 6500 iterations
and the minimum weight of the frame is 7533 kg. In the present work, an optimum weight of
7047 kg is obtained after only 250 iterations using 15 particles. Figure (10) shows the
iteration history of the design.

7. CONCLUSIONS
A particle swarm optimizer and Gaussian-PSO are combined to develop an optimum
design algorithm for steel frames. Constraints are handled with modified fly-back
mechanism. The design algorithm is mathematically quite simple and effective in finding
solutions of combinatorial optimization problems. The PSO is acting as fast optimizer and the

11
 Fig. (9). Two-bay Six-Storey Frame

Fig. (10): Iteration History for Example No. 3.

(No. of particles = 15 and the duration of PSO search is 100
i i )
12
 Gaussian–PSO plays the fine tuning search. The technique reduces the number of
iterations required to reach the optimum design (100 to 400 iterations) and also allow using
smaller number of particles (5 to 15 particles). The studied examples show that 100 iterations
are enough for PSO fast search before starting Gaussian–PSO. Moreover, the studied
examples show that the modified fly-back mechanism is powerful in particle swarm
optimization technique.

REFERENCES

1. Doğan, E. and Saka, M., “Optimum Design of Unbraced Steel Frames to LRFD–
AISC Using Particle Swarm Optimization”, Advances in Engineering Software, No.
46, pp 27–34, 2012.
2. Voss, M. S., “Social programming using functional swarm optimization,” in
Proceedings of the IEEE Swarm Intelligence Symposium (SIS 2003), Indianapolis,
Indiana, USA, pp. 103-109, 2003.
3. Langdon, W. B., Poli, R., Holland, O. and Krink, T.,"Understanding Particle Swarm
Optimization by Evolving Problem Landscapes", in Proceedings of the IEEE Swarm
Intelligence Symposium, Pasadena, California, USA, pp.30–37, 8-10 June 2005.
4. Kennedy, J., and Eberhart, R., "Particle swarm optimization", Proceedings of the
IEEE International Conference on Neural Networks (ICNN), Vol. 4, pp.1942-1948,
Perth, Australia, 1995.
5. Ozcan, E. and Mohan, C. K., "Analysis of a simple particle swarm optimization
system", Intelligent Engineering Systems Through Artificial Neural Networks, Vol. 8,
pp. 253–258, 1998.
6. Clerc, M. and Kennedy, J., “The Particle Swarm: Explosion, Stability, and
Convergence in a Multidimensional Complex Space”, IEEE Transactions on
Evolutionary Computation, Vol. 6, No. 1, 2002.
7. Fourie, P.C. and Groenwold, A.A., “The Particle Swarm Optimization Algorithm in
Size and Shape Optimization”, Struct. Multidisc Optim., No. 23, pp 259–267,
Springer-Verlag 2002.
8. Schutte, J.F. and Groenwold, A.A., “Sizing design of truss structures using particle
swarms”, Struct. Multidisc Optim., No. 25, pp 261–269, 2003.
9. Bochenek, B. and Foryś, P., "Structural Optimization Against Instability Using
Particle Swarms", Proceedings of the 6th World Congresses of Structural and
Multidisciplinary Optimization (WCSMO), Rio de Janeiro, Brazil, 30 May - 03 June,
2005.
10. Jármai, K., Rodrigues, J.P.C. and Farkas, J., "Optimization of Steel Structures for Fire
Resistance Using PSO", Proceedings of the 6th World Congresses of Structural and
Multidisciplinary Optimization (WCSMO), Rio de Janeiro, Brazil, 30 May - 03 June,
2005.
11. Heinisuo, M., Saikkonen, M., Haapio, J. and Miles, J.," Welded Steel Beam Design
Using Particle Swarm Analysis", In Proceedings of CIB 24th W78 Conference,
Maribor, 26-29 June, 2007.
12. Jalkanen, J., “Tubular Truss Optimization Using Heuristic Algorithm”, Thesis for the
degree of Doctor of Technology, Tampere University of Technology, 2007.
13. Kaveh, A. and Talatahari, S., "A Hybrid Particle Swarm and Ant Colony
Optimization for Design of Truss Structures", Asian Journal of Civil Engineering,
Vol. 9,No. 4, pp.329-348, 2008.

13
 14. Kaveh, A. and Talatahari, S., "A Discrete Particle Swarm Ant Colony Optimization
for Design of Frames", Asian Journal of Civil Engineering, Vol. 9,No. 6, pp.563-575,
2008.
15. Kaveh, A. and S. Talatahari, S., “Engineering Optimization with Hybrid Particle
Swarm and Ant Colony Optimization”, Asian Journal of Civil Engineering (Building
and Housing) Vol. 10, No. 6, pp 611-628, 2009.
16. Kaveh, A., Talatahari, S. and Farahmand, B., “An Improved (HPSACO) For
Engineering Optimum Design Problems”, Asian Journal of Civil Engineering
(Building and Housing), Vol. 12, No. 2, pp 133-141, 2011.
17. Doğan, E., Hasançebi, O. and Saka, M., “A Refinement of Discrete Particle Swarm
Optimization for Large-Scale Truss Structures”, Asian Journal of Civil Engineering
(Building and Housing), Vol. 10, No. 3, pp 321-334, 2009.
18. Ghasemi, M. and Dizangian, B. “Size, Shape and Topology Optimization of
Composite Steel Box Girders Using PSO Method”, Asian Journal of Civil
Engineering (Building and Housing), Vol. 11, No. 6, pp 699-715, 2010.
19. Yang, B. and Zhang, Q. “Applying a Parallel Particle Swarm Optimizer to Truss
Topological Design”, Structural Engineering World Congress (SEWC 2011), Italy, 4-
6 April, 2011.
20. Kaveh, A., Talatahari S. and Alami, M.T.,”A New Hybrid Meta-Heuristic for
Optimum Design of Frame Structures”, Asian Journal of Civil Engineering (Building
and Housing), Vol. 13, No. 6, pp 705-717, 2012.
21. Talatahari, S., Khalili, E. and Alavizadeh S.M.,” Accelerated Particle Swarm for
Optimum Design of Frame Structures”, Hindawi Publishing Corporation,
Mathematical Problems in Engineering, Vol. 2013.
22. Bergh, F., "An Analysis of Particle Swarm Optimizers", PhD thesis, Faculty of
Natural and Agricultural Science, University of Pretoria, November, 2001.
23. Hsieh, H. and Lee, T.,”A Modified Algorithm of Bare Bones Particle Swarm
Optimization”, Proceedings of the IJCSI International Journal of Computer Science
Issues, Vol. 7, Issue 6, November 2010.
24. Kennedy, J., "How It Works: Collaborative Trial and Error", International Journal of
Computational Intelligence Research. ISSN 0973-1873, Vol.4, No.2, pp. 71–78, 2008.
25. Perez, R. E. and Behdinan, K., " Particle Swarm Optimization in Structural Design",
Chapter of "Swarm Intelligence: Focus on Ant and Particle Swarm Optimization",
Book edited by: Felix T. S. Chan and Manoj Kumar Tiwari, ISBN 978-3-902613-09-
7, Itech Education and Publishing, Vienna, Austria, December 2007.
26. Hu, X., Eberhart, R. C., and Shi, Y., “Engineering Optimization with Particle
Swarm”, Proceedings of the IEEE Swarm Intelligence Symposium 2003 (SIS 2003),
Indianapolis, Indiana, USA, pp. 53-57, 2003.
27. Arora, J.S., “Introduction to optimum design”, Elsevier Inc., New York, Third
Edition, 2012.
28. Kaveh, A. and Malakoutirad, S.,” Hybrid Genetic Algorithm and Particle Swarm
Optimization for the Force Method-Based Simultaneous Analysis and Design”,
Iranian Journal of Science & Technology, Transaction B: Engineering, Vol. 34, No.
B1, pp 15-34, 2010.
29. Kameshki(2003), E. S., “Discrete Optimum Design of Steel Frames by Genetic
Algorithm”, J. King Saud Univ., Vol. 15, Eng. Sci. (2), pp. 217-234, Riyadh, 2003.

14

View publication stats