Professional Documents
Culture Documents
PDF Socio Cultural Inspired Metaheuristics Anand J Kulkarni Ebook Full Chapter
PDF Socio Cultural Inspired Metaheuristics Anand J Kulkarni Ebook Full Chapter
Anand J. Kulkarni
Visit to download the full and correct content document:
https://textbookfull.com/product/socio-cultural-inspired-metaheuristics-anand-j-kulkarn
i/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
https://textbookfull.com/product/cohort-intelligence-a-socio-
inspired-optimization-method-1st-edition-anand-jayant-kulkarni/
https://textbookfull.com/product/proceedings-of-the-2nd-
international-conference-on-data-engineering-and-communication-
technology-icdect-2017-anand-j-kulkarni/
https://textbookfull.com/product/hokkien-theatre-across-the-seas-
a-socio-cultural-study-caroline-chia/
https://textbookfull.com/product/employing-people-with-
disabilities-good-organisational-practices-and-socio-cultural-
conditions-ewa-giermanowska/
Natural History Dioramas – Traditional Exhibits for
Current Educational Themes: Socio-cultural Aspects
Annette Scheersoi
https://textbookfull.com/product/natural-history-dioramas-
traditional-exhibits-for-current-educational-themes-socio-
cultural-aspects-annette-scheersoi/
https://textbookfull.com/product/the-paradox-of-creativity-in-
art-education-bourdieu-and-socio-cultural-practice-kerry-thomas/
https://textbookfull.com/product/islamic-marketing-understanding-
the-socio-economic-cultural-and-politico-legal-environment-1st-
edition-cedomir-nestorovic-auth/
https://textbookfull.com/product/female-olympians-a-mediated-
socio-cultural-and-political-economic-timeline-1st-edition-linda-
k-fuller-auth/
https://textbookfull.com/product/cultural-psychology-third-
edition-steven-j-heine/
Studies in Computational Intelligence 828
Anand J. Kulkarni
Pramod Kumar Singh
Suresh Chandra Satapathy
Ali Husseinzadeh Kashan
Kang Tai Editors
Socio-cultural
Inspired
Metaheuristics
Studies in Computational Intelligence
Volume 828
Series Editor
Janusz Kacprzyk, Polish Academy of Sciences, Warsaw, Poland
The series “Studies in Computational Intelligence” (SCI) publishes new develop-
ments and advances in the various areas of computational intelligence—quickly and
with a high quality. The intent is to cover the theory, applications, and design
methods of computational intelligence, as embedded in the fields of engineering,
computer science, physics and life sciences, as well as the methodologies behind
them. The series contains monographs, lecture notes and edited volumes in
computational intelligence spanning the areas of neural networks, connectionist
systems, genetic algorithms, evolutionary computation, artificial intelligence,
cellular automata, self-organizing systems, soft computing, fuzzy systems, and
hybrid intelligent systems. Of particular value to both the contributors and the
readership are the short publication timeframe and the world-wide distribution,
which enable both wide and rapid dissemination of research output.
The books of this series are submitted to indexing to Web of Science,
EI-Compendex, DBLP, SCOPUS, Google Scholar and Springerlink.
Kang Tai
Editors
Socio-cultural Inspired
Metaheuristics
123
Editors
Anand J. Kulkarni Pramod Kumar Singh
Department of Mechanical Engineering ABV-Indian Institute of Information
Symbiosis Institute of Technology Technology and Management Gwalior
Symbiosis International (Deemed Gwalior, Madhya Pradesh, India
University)
Pune, Maharashtra, India Ali Husseinzadeh Kashan
Faculty of Industrial and Systems
Odette School of Business
Engineering
University of Windsor
Tarbiat Modares University
Windsor, Canada
Tehran, Iran
Suresh Chandra Satapathy
School of Computer Engineering
Kalinga Institute of Industrial Technology
(KIIT), Deemed to be University
Bhubaneswar, Odisha, India
Kang Tai
School of Mechanical and Aerospace
Engineering
Nanyang Technological University
Singapore, Singapore
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface
v
vi Preface
vii
viii Contents
ix
x About the Editors
volumes for Springer AISC and LNCS. In addition to serving on the editorial board
of several journals, he is a senior member of the IEEE and a life member of the
Computer Society of India, where he is the National Chairman of Division-V
(Education and Research).
Ali Husseinzadeh Kashan completed his Ph.D. and postdoc in Industrial Engineering
at Amirkabir University of Technology, Iran. Currently, he is an Associate Professor
at the Department of ISE, Tarbiat Modares University, Iran. He received an ‘Outstanding
Young Scientist of Industrial Engineering’ award from the Academy of Sciences of the
Islamic Republic of Iran in 2016, and has been awarded honors for his research by
various academic associations in Iran. His research interests include mathematical
modelling and algorithms, computational intelligence and engineering design with
nature-inspired optimization and simulation-based optimization. Dr. Husseinzadeh
Kashan has published over 70 research papers in peer-reviewed journals and conferences
and has served as a reviewer for several outstanding journals such as: IEEE Transactions
on Evolutionary Computations, Omega, Computers & Operations Research, Journal of the
Operational Research Society, Computers & Industrial Engineering, and International
Journal of Production Research.
Kang Tai obtained his B.Eng. (1st Class Honors) in Mechanical Engineering from
the NUS in 1990. In 1992 he was awarded the NTU Singapore Overseas Scholarship
to pursue his Ph.D. at the Imperial College of Science, Technology and Medicine
London, after which he returned to NTU in 1995 and he is currently an Associate
Professor there. From 2001 to 2004 he was also concurrently appointed as a Faculty
Fellow of the Singapore-MIT Alliance. He teaches various UG and PG courses. His
research interests include design automation, optimization, evolutionary algorithms,
computational geometry, system identification and mathematical modeling of
industrial processes. He has also published more than 70 research papers.
Optimum Design of Four Mechanical
Elements Using Cohort Intelligence
Algorithm
Abstract In this study, Cohort Intelligence (CI) algorithm is implemented for solv-
ing four mechanical engineering problems such as design of closed coil helical
spring, belt pulley drive, hollow shaft, and helical spring. As these problems are
constrained in nature, a penalty function approach is incorporated. The perfor-
mance of the constrained CI is compared with other contemporary algorithms such
as Genetic Algorithm (GA), Particle Swarm Optimization, Artificial Bee Colony
(ABC), Teaching–Learning-Based Optimization (TLBO), and TLBO with Differ-
ential Operator (DTLBO). The performance of the constrained CI was better than
other algorithms in terms of objective function. The computational cost was quite
reasonable, and the algorithm exhibited robustness solving these problems.
1 Introduction
1/ϕ(xm , R)
pm M
, m 1, . . . , M (4)
m1 1/ϕ(xm , R)
START
Every candidate follows the best behavior from within its behaviors
N
Cohort
behavior saturated?
Y
N
Convergence
Y
Accept the current
behavior as final solution
STOP
The problem statement and mathematical model of closed coil helical spring are
adopted from [30, 31]. The spring is made up of a wire coiled in the form of a helix
which is primarily intended for compressive and tensile load (Fig. 2). The cross
section of wire may be circular, square, or rectangular. It can be bent in the form
of helix. Basically, two types of helical springs are used that are compressive and
tensile helical spring. The helical spring is said to be closely coiled, when there is
small gap between adjacent coils. In other words, helix angle is very small and plane
containing each coil is almost right angle to axis of helix. External loads act along
the axis of the spring. In both cases, wire is subjected to torsion. Due to torsion, shear
stress is produced in the spring.
Optimum Design of Four Mechanical Elements … 7
Terminology:
U Volume of spring (cm3 )
Nc Number of active coils
C Spring index
K Stiffness of spring or spring constant (kgf/cm)
G Modulus of rigidity (Kgf/cm2 )
S Allowable shear stress
Fmax Maximum working load (kg)
The objective for design of closed coil helical spring is minimization of the spring
volume. For given application, estimate the minimum volume of helical spring under
8 K. Marde et al.
π2
U (Nc + 2)Dd 2 (6)
4
The basic procedure of design of helical spring consists of the following con-
straints [31], and the same constraints are used in this example.
Stress constraint: The shear stress must be less than specified value, which is
represented as
D
S − 8C f Fmax ≥0 (7)
π d3
where
4C − 1 0.615 D
Cf + , C (8)
4C − 4 C d
Here, maximum working load (Fmax ) is set to be 435.6 kg, and allowable shear
stress (S) is set to be 13288.02 kgf/cm2 .
Configuration constraint: The free length of the spring must be less than the
maximum specified value. The spring constant (K) can be determined using the
following expression:
Gd 4
K (9)
8Nc D 3
Fmax
δl (10)
K
It is assumed that the spring length under maximum working load Fmax is 1.05
times the solid length. Thus, the free length is given by the following expression.
lmax − l f ≥ 0 (12)
Optimum Design of Four Mechanical Elements … 9
d − dmin ≥ 0 (13)
Dmax − (D + d) ≥ 0 (14)
C −3≥0 (15)
The deflection under preload must be less than the maximum specified. The deflec-
tion under preload is expressed as
Fp
δp (16)
K
where F p is equal to 136.08 kg.
The constraint is given by the expression
δ pm − δ p ≥ 0 (17)
Fmax − F p
l f − δp − 1.05(Nc + 2)d ≥ 0 (18)
K
The above constraint should be equality, and it is intuitively clear that at conver-
gence and the constraint function will always be zero.
The deflection from preload to maximum load must be equal to the specified
value. These two made an inequality constraint since it should always converge to
zero. It can be represented as
Fmax − F p
− δw ≥ 0 (19)
K
10 K. Marde et al.
100
candidate1
90 candidate2
candidate3
Behavior (Volume)
80 candidate4
candidate5
70
60
50
40
0 5 10 15 20 25 30 35 40 45 50
Learning Attempts
1 46.0968 0.88 585 0.7102 1.2700 15.0000 0.0000 7.0109 0.2022 5.6397 0.0000 15.0774 3.0288 12.5326
2 46.0968 1.01 575 0.7102 1.2700 15.0000 0.0000 7.0051 0.2022 5.6397 0.0000 15.0774 3.0374 12.5398
3 46.0968 0.84 555 0.7102 1.2700 15.0000 0.0000 7.0089 0.2022 5.6397 0.0000 15.0774 3.0317 12.5351
4 46.0972 0.95 630 0.7105 1.2700 15.0000 0.0000 7.0281 0.2025 5.6394 0.0000 15.0776 3.0032 12.5113
5 46.0971 0.86 570 0.7105 1.2700 15.0000 0.0000 7.0272 0.2025 5.6394 0.0000 15.0776 3.0045 12.5124
6 46.1041 0.90 595 0.7074 1.2700 15.0000 0.0000 6.8027 0.1994 5.6425 0.0000 15.0748 3.3369 12.7894
7 46.0985 0.84 560 0.7111 1.2700 15.0000 0.0000 7.0684 0.2031 5.6388 0.0000 15.0782 2.9433 12.4615
8 46.0993 0.89 590 0.7113 1.2700 15.0000 0.0000 7.0849 0.2033 5.6386 0.0000 15.0784 2.9186 12.4410
9 46.0975 0.81 535 0.7090 1.2700 15.0000 0.0000 6.9188 0.2010 5.6409 0.0000 15.0762 3.1655097 12.6464
10 46.0984 0.90 600 0.7086 1.27 15.0000 0.0000 6.8941 0.2006 5.6413 0.0000 15.0759 3.2020 12.6769
Optimum Design of Four Mechanical Elements …
11 46.0979 0.85 565 0.7088 1.2700 15.0000 0.0000 6.9064 0.2008 5.6411 0.0000 15.0761 3.1837 12.6616
12 46.0966 0.84 560 0.7097 1.2700 15.0000 0.0000 6.9720 0.2017 5.6402 0.0000 15.0769 3.0866 12.5807
13 46.0967 0.82 545 0.7101 1.2700 15.0000 0.0000 7.0017 0.2021 5.6398 0.0000 15.0773 3.0425 12.5440
14 46.0981 0.85 560 0.7087 1.2700 15.0000 0.0000 6.9003 0.2007 5.6412 0.0000 15.0760 3.1927 12.6692
15 46.0994 0.89 595 0.7113 1.2700 15.0000 0.0000 7.0858 0.2033 5.6386 0.0000 15.0784 2.9172 12.4398
16 46.0967 0.85 565 0.7095 1.2700 15.0000 0.0000 6.9573 0.2015 5.6404 0.0000 15.0767 3.1083 12.5988
17 46.0979 0.96 635 0.7108 1.2700 15.0000 0.0000 7.0523 0.2028 5.6391 0.0000 15.0780 2.9672 12.4814
18 46.0973 0.88 580 0.7091 1.2700 15.0000 0.0000 6.9274 0.2011 5.6408 0.0000 15.0764 3.1526 12.6358
19 46.0972 0.87 575 0.7105 1.2700 15.0000 0.0000 7.0310 0.2025 5.6394 0.0000 15.0777 2.9988 12.5077
20 46.0989 0.88 580 0.7112 1.2700 15.0000 0.0000 7.0774 0.2032 5.6387 0.0000 15.0783 2.9297 12.4503139
SD 0.0017 0.04 0.0010 1.67E-07 4.44E-07 0.0000
11
12 K. Marde et al.
Table 2 Comparison of closed coil helical spring results obtained by CI with other methods
Method Conventional GA [10, PSO [2] ABC [3] TLBO DTLBO CI
11, 30] [24] [31]
Worst NA 46.6932 46.6752 46.6241 46.5214 46.4322 46.1042
Mean NA 46.6821 46.6254 46.6033 46.4998 46.3192 46.0980
Best 46.5392 46.6653 46.5212 46.5115 46.3221 46.3012 46.0966
Average NA 192 108 150 132 144 53.0026
time (s)
Min. NA 180 102 138 120 132 48.6015
time (s)
This problem is adopted from [30, 31]. A shaft is rotating machine element which
usually used to transmit power from one element to another. It has to two groups:
first one is transmission shaft and second is line shaft. Transmission shaft is usu-
ally in circular cross section and used to transmit power between the source and
machines; absorbing power is called transmission shaft. Transmission shaft supports
transmission elements like pulley, gears, etc. Machine shaft is an integrated part of
the machine itself and crankshaft is an example of machine shaft. The hollow shaft
is shown in Fig. 4.
Optimum Design of Four Mechanical Elements … 13
Terminology:
Ws Weight of hollow shaft, kg
L Length of hollow shaft, cm
ρ Density, kg/cm3
In this paper, the same objective function is considered, those considered by [31].
The objective of this study is to minimize the weight of hollow shaft which is given
by expression
[31] used the following constraints, and the same constraints are used in this example.
The abovementioned objective function is subjected to the following constraints.
The twisting failure can be calculated from torsion formula as given below:
T Gθ
(22)
J L
as
d04 1 − K 4 − 1736.93 ≥ 0 (23)
Substituting the values of Tcr , γ , and E to 1.0 × 105 kg/cm, 0.3, and 2.0 × 105
kg/cm2 respectively, the constraints are expressed as
7 ≤ do ≤ 25 (26)
The hollow shaft problem was solved 20 times CI with number of candidates
M 5, the number of variations k 115, and the sampling interval reduction
factor r 0.9. The solution including objective function (shaft weight Ws ) value,
Function Evaluations (FE), outer diameter of shaft do , ratio of inner diameter to outer
diameter K, and various constraint values associated with these 20 runs are presented
in Table 3. The CI solutions were compared with two optimization algorithms such
as GA and DTLBO. Comparison of weight of hollow shaft Ws , outer diameter of
shaft do , and ratio of inner diameter to outer diameter K of CI solving the problem
is presented in Table 4. It is observed that CI algorithm yielded better solutions
than DTLBO and GA. But the ratio of inner diameter to outer diameter K is not
better solution than DTLBO and GA. The improvements were 7.34% and 6.12% as
compared to DTLBO and GA, respectively. The CI convergence plot is presented
in Fig. 5. This plot clearly manifests the self-supervised learning behavior of every
candidate in the cohort. Every candidate tried to adopt the behavior of other candidate
and improve its own behavior.
This example is adopted from [30], and this example was solved previously by [31]
using DTLBO algorithm. The belts are used to transmit power over considerable
distance between two shafts by means of pulleys (Fig. 6). The belt pulley drive
consists of endless belt, driving, and driven pulley. The rotary motion of driving
pulley is converted into translatory motion of belt and again it converts into rotary
motion of driven pulley. Rotary speed of pulleys may be same or different. Generally,
in factories and workshops, stepped flat belts are used to transmit moderate amount
of power. The weight of pulley acts on shaft and bearing which affected both, due
to which shaft and bearing failures occur. It becomes essential to prevent shaft and
Optimum Design of Four Mechanical Elements … 15
Table 3 Performance of the proposed Cohort Intelligence (CI) for hollow shaft
Trials Obj Time (s) FE do K Equation (26) Constraint 2:
Eq. (25)
1 2.2253 0.7543 525 7.0000 0.9699 1461.5090 10693.58758
2 2.2253 0.7452 515 7.0000 0.9699 1461.5112 10693.60921
3 2.2253 0.7440 515 7.0000 0.9699 1461.5090 10693.71374
4 2.2253 0.7507 525 7.0000 0.9699 1461.5137 10693.35672
5 2.2253 0.7488 525 7.00007 0.9699 1461.5093 10693.68976
6 2.2253 0.7333 515 7.0000 0.9699 1461.5113 10693.48511
7 2.2253 0.7075 500 7.0000 0.9699 1461.5135 10693.3988
8 2.2253 0.7302 510 7.0000 0.9699 1461.5128 10693.5472
9 2.2253 0.7144 510 7.0000 0.9699 1461.5140 10693.4674
10 2.2253 0.7493 530 7.0000 0.9699 1461.5166 10693.3269
11 2.2253 0.7250 505 7.0000 0.9699 1461.5108 10693.5606
12 2.2253 0.7542 525 7.0000 0.9699 1461.5087 10693.7528
13 2.2253 0.7304 515 7.0000 0.9699 1461.5092 10693.6668
14 2.2253 0.7090 505 7.0000 0.9699 1461.5128 10693.6196
15 2.2253 0.7513 530 7.0000 0.9699 1461.5118 10693.5766
16 2.2253 0.7415 525 7.0000 0.9699 1461.5132 10693.6817
17 2.2253 0.7416 525 7.0000 0.9699 1461.5142 10693.4056
18 2.2253 0.7578 515 7.0000 0.9699 1461.5133 10693.3654
19 2.2253 0.7383 520 7.0000 0.9699 1461.5114 10693.5351
20 2.2253 0.7357 515 7.0000 0.9699 1461.5119 10693.5541
SD 1.6E-06 0.0149 1.21E-05 1.34E-07
Table 4 Comparison of
Method GA DTLBO CI
hollow shaft results obtained
by CI with other methods Weight of 2.3704 2.4017 2.2253
hollow shaft
Ratio of inner 0.9699 0.9685 0.9699
diameter to
outer
diameter
Outer 11.0928 10.9 7.0000
diameter of
shaft
bearing failure by minimizing the weight of flat belt drive. The belt pulley drive is
shown in Fig. 6.
16 K. Marde et al.
80
candidate1
70
candidate2
60 candidate3
Behavior (Weight) candidate4
50 candidate5
40
30
20
10
0
0 5 10 15 20 25 30 35 40 45 50
Learning Attempts
Terminology:
Wp Weight of pulley, kg
ρ Density, kg/cm3
B Width of pulley, cm
V Velocity of pulley, cm/s
T1 Belt tension in tight side, N
(continued)
Optimum Design of Four Mechanical Elements … 17
(continued)
tb Thickness of belt, cm
The objective of this problem is to minimize the weight of the pulley. The same
objective functions are considered by [31]. The weight of the pulley is considered as
objective function is to be minimized as
W p πρb d1 t1 + d2 t2 + d11 t11 + d21 t21 (28)
Assuming t 1 0.1d 1 , t 2 0.1d 2 , t11 0.1d11 and t21 0.1d21 and replacing d 1 ,
d 2 , d11 , and d21 by N 1 , N 2 , N11 , and N21 , respectively, and substituting values of N 1 ,
N 2 , N11 , and N21 , ρ (to 1000, 250, 500, 500) 7.2 × 10−3 kg/cm3 , respectively, the
objective function can be written as
[31] used the following constraints and the same constraints are used in this example.
It is subjected to the following constraints.
−T2 )
The transmitted power can be represented as P (T175 V
Substituting the expression of V in above equation, one gets
πdp Np
(T1 − T2 ) (30)
75 × 60 × 100
T2 πdp Np
P T1 1 − (31)
T1 75 × 60 × 100
Assuming T2 /T1 1/2, P 10 hp, and substituting the values of T2 /T1 and P,
one gets
18 K. Marde et al.
1 πdp Np
10 T1 1 − (32)
2 75 × 60 × 100
or
286478
T1 (33)
dp Np
Assume
2864789
30b × 1.0 ≥ (34)
d2 250
381.97
b≥ (35)
d2
bd2 − 81.97 ≥ 0 (36)
Assuming that width of the pulley is either less than or equal to one-fourth of the
diameter of the first pulley, the constraint is expressed as
b ≤ 0.25d1 (37)
d1
−1≥0 (38)
4b
The ranges of variables are mentioned as follows:
15 ≤ d1 ≤ 25, 70 ≤ d2 ≤ 80, 4 ≤ b ≤ 10
The belt pulley drive problem was solved 20 times CI with number of candidates
M 5, the number of variations k 115, and the sampling interval reduction
factor r 0.9. The solution including objective function (pulley weight W p ) value,
Function Evaluations (FE), wire diameter d, mean coil diameter D, number of active
coils Nc , and various constraint values associated with these 20 runs are presented in
Table 5. The CI solutions were compared with several optimization algorithms such
as GA, PSO, ABC, TLBO, and DTLBO. The best, mean, and worst objective function
(Weight) values, average, and minimum computational time of CI solving the prob-
lem are presented in Table 6. It is observed that CI algorithm yielded better solutions
Another random document with
no related content on Scribd:
Gutenberg™’s goals and ensuring that the Project Gutenberg™
collection will remain freely available for generations to come. In
2001, the Project Gutenberg Literary Archive Foundation was
created to provide a secure and permanent future for Project
Gutenberg™ and future generations. To learn more about the
Project Gutenberg Literary Archive Foundation and how your
efforts and donations can help, see Sections 3 and 4 and the
Foundation information page at www.gutenberg.org.
Most people start at our website which has the main PG search
facility: www.gutenberg.org.