Socio cultural Inspired Metaheuristics

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 ...

Cohort Intelligence: A Socio-inspired Optimization

Method 1st Edition Anand Jayant Kulkarni

https://textbookfull.com/product/cohort-intelligence-a-socio-
inspired-optimization-method-1st-edition-anand-jayant-kulkarni/

Proceedings of the 2nd International Conference on Data

Engineering and Communication Technology ICDECT 2017
Anand J. Kulkarni

https://textbookfull.com/product/proceedings-of-the-2nd-
international-conference-on-data-engineering-and-communication-
technology-icdect-2017-anand-j-kulkarni/

Hokkien Theatre Across The Seas A Socio Cultural Study

Caroline Chia

https://textbookfull.com/product/hokkien-theatre-across-the-seas-
a-socio-cultural-study-caroline-chia/

Employing People with Disabilities Good Organisational

Practices and Socio cultural Conditions Ewa
Giermanowska

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/

The Paradox of Creativity in Art Education Bourdieu and

Socio cultural Practice Kerry Thomas

https://textbookfull.com/product/the-paradox-of-creativity-in-
art-education-bourdieu-and-socio-cultural-practice-kerry-thomas/

Islamic Marketing Understanding the Socio Economic

Cultural and Politico Legal Environment 1st Edition
■edomir Nestorovi■ (Auth.)

https://textbookfull.com/product/islamic-marketing-understanding-
the-socio-economic-cultural-and-politico-legal-environment-1st-
edition-cedomir-nestorovic-auth/

Female Olympians: A Mediated Socio-Cultural and

Political-Economic Timeline 1st Edition Linda K. Fuller
(Auth.)

https://textbookfull.com/product/female-olympians-a-mediated-
socio-cultural-and-political-economic-timeline-1st-edition-linda-
k-fuller-auth/

Cultural Psychology Third Edition Steven J. Heine

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.

More information about this series at http://www.springer.com/series/7092

Anand J. Kulkarni Pramod Kumar Singh
• •

Suresh Chandra Satapathy •

Ali Husseinzadeh Kashan

•

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

ISSN 1860-949X ISSN 1860-9503 (electronic)

Studies in Computational Intelligence
ISBN 978-981-13-6568-3 ISBN 978-981-13-6569-0 (eBook)
https://doi.org/10.1007/978-981-13-6569-0

Library of Congress Control Number: 2019931506

© Springer Nature Singapore Pte Ltd. 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.

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

This edited volume intends to disseminate state-of-the-art knowledge and

development in the field of socio-cultural inspired algorithms to engineers, scientists
and researchers working in the areas of optimization, evolutionary computation and
computational intelligence. Akin to the evolutionary and swarm-based optimization
algorithms, the socio-cultural algorithms too belong to the category of metaheuristics
and are inspired from natural and social tendencies observed in humans by learning
from one another through social interactions. The socio-behavioral models, inspired
from social evolution observed in humans, are fairly recent developments in the field
of nature-inspired metaheuristics and serve as a powerful algorithmic framework to
model self-adaptation in an evolutionary system that provides guiding strategies
which help to explore the search space efficiently.
The volume critically elaborated complete state-of-the-art review of several
contemporary socio-inspired optimization algorithms. It may certainly help in
locating new research avenues to interested researchers. The volume also discusses
associated original and pioneering research contributions with a description of the
methodology, complete mathematical formulation and typical problem solving
along with the comparative solutions and analysis. More specifically, the volume in
detail discussed several applications of well-known socio-inspired cohort intelli-
gence (CI) algorithm. The associated rigorously tested applications are mainly from
the domain of combinatorial optimization, discrete structural optimization,
mechanical engineering design, control system design and rescue autonomous
robotics. Furthermore, the application of teaching–learning-based optimization
(TLBO) algorithm was demonstrated in the areas of pattern recognition and clus-
tering for verification of plant diseases. The volume also introduces a novel
expectation algorithm (ExA) validated on several unconstrained problems from
standard test bed. Moreover, a modified version of league championship algorithm
(LCA) referred to as premier LCA (PLCA) and improved cultural algorithm
(ICA) were successfully proposed for solving complex structural problems. An
emerging social group optimization (SGO) algorithm was applied for solving
practically important variations of antenna array synthesis problem. A variation of
social emotional optimization algorithm (SEOA) is proposed for feature selection in

v
vi Preface

classification problems with application to vocal segmentation of the selected

songs. Furthermore, socio-evolution and learning optimization (SELO) was applied
to practically important crop planning problem choosing the most suitable crop that
can be grown during a particular period.
Every chapter submitted to the volume was critically evaluated by at least two
expert reviewers. The critical suggestions by the reviewers certainly helped and
influenced the authors of the individual chapter to enrich the quality in terms of
experimentation, performance evaluation, representation, etc. The volume may
serve as a complete reference for the socio-cultural algorithms.

Pune, India Anand J. Kulkarni

Gwalior, India Pramod Kumar Singh
Bhubaneswar, India Suresh Chandra Satapathy
Tehran, Iran Ali Husseinzadeh Kashan
Singapore, Singapore Kang Tai
Contents

Optimum Design of Four Mechanical Elements Using Cohort

Intelligence Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Kishor Marde, Anand J. Kulkarni and Pramod Kumar Singh
A Self-organizing Multi-agent Cooperative Robotic System:
An Application of Cohort Intelligence Algorithm . . . . . . . . . . . . . . . . . . 27
Palash Roychowdhury, Siddarth Mehra, Rishi Devarakonda,
Prakhar Shrivastava, Siddharth Basu and Anand J. Kulkarni
Elitist TLBO for Identification and Verification of Plant Diseases . . . . . 41
Tamanna Jena, T. M. Rajesh and Mallanagouda Patil
Feature Selection for Vocal Segmentation Using Social Emotional
Optimization Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Poreddy Rajasekharreddy and E. S. Gopi
Simultaneous Size and Shape Optimization of Dome-Shaped
Structures Using Improved Cultural Algorithm . . . . . . . . . . . . . . . . . . . 93
Shahin Jalili, Yousef Hosseinzadeh and Timon Rabczuk
A Socio-based Cohort Intelligence Algorithm
for Engineering Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Ishaan R. Kale, Anand J. Kulkarni and Suresh Chandra Satapathy
Solution to Small Size 0–1 Knapsack Problem Using Cohort
Intelligence with Educated Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Mandar S. Sapre, Harsh Patel, Kishan Vaishnani, Raj Thaker
and Apoorva S. Shastri
Maximizing Profits in Crop Planning Using Socio Evolution
and Learning Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
G. Jaya Brindha and E. S. Gopi

vii
viii Contents

Application of Variations of Cohort Intelligence in Designing

Fractional PID Controller for Various Systems . . . . . . . . . . . . . . . . . . . 175
Pritesh Shah and Anand J. Kulkarni
Expectation Algorithm (ExA): A Socio-inspired Optimization
Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Apoorva S. Shastri, Aishwary Jagetia, Amit Sehgal, Meet Patel
and Anand J. Kulkarni
Premier League Championship Algorithm: A Multi-population-Based
Algorithm and Its Application on Structural Design Optimization . . . . 215
Ali Husseinzadeh Kashan, Shahin Jalili and Somayyeh Karimiyan
Socio-inspired Optimization Metaheuristics: A Review . . . . . . . . . . . . . 241
Meeta Kumar and Anand J. Kulkarni
Social Group Optimization Algorithm for Pattern Optimization
in Antenna Arrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
V. V. S. S. S. Chakravarthy, P. S. R. Chowdary,
Suresh Chandra Satapathy, Jaume Anguera and Aurora Andújar
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
About the Editors

Anand J. Kulkarni holds a Ph.D. in Distributed Optimization from Nanyang

Technological University, Singapore, M.S. in AI from the University of Regina,
Canada, and Bachelor of Engineering from Shivaji University, India. He worked as
a Research Fellow on a cross-border supply-chain disruption project at Odette
School of Business, University of Windsor, Canada. Currently, he is working as
Head and Associate Professor at the Symbiosis Institute of Technology, Pune,
India. His research interests include optimization algorithms, multi-objective opti-
mization, multi-agent systems, complex systems, swarm optimization, game theory
and self-organizing systems. He is the founder and chairman of the OAT Research
Lab. Anand has published over 40 research papers in peer-reviewed journals and
conferences along with two books.

Pramod Kumar Singh is a Professor in Information and Communication

Technology at the ABV-Indian Institute of Information Technology and
Management, Gwalior, India. He completed his B.Tech. from the Kamla Nehru
Institute of Technology, Sultanpur, UP, India, and M.Tech. and Ph.D. from the
Indian Institute of Technology, Kharagpur, India. His research interest includes
computational intelligence, multi-objective optimization, data mining, text analyt-
ics, and machine learning. He is member of IEEE and ACM. He has been Associate
Editor of IJSI and Guest Editor of JOCS. Besides, he is on the reviewer board of
various reputed journals published by IEEE (including Transactions), Elsevier,
Springer, ACM, and Taylor and Francis. He has published more than 70 research
papers in various international journals and conferences.

Suresh Chandra Satapathy is a Professor at the School of Computer

Engineering, KIIT, Odisha, India. Previously, he was a Professor and Head of the
Department of CSE at ANITS, AP, India. He received his Ph.D. in CSE from
JNTU, Hyderabad and M.Tech. in CSE from the NIT, Odisha. He has more than 27
years of teaching and research experience. His research interests include machine
learning, data mining, swarm intelligence and applications. He has published more
than 98 papers in respected journals and conferences and has edited numerous

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

Kishor Marde, Anand J. Kulkarni and Pramod Kumar Singh

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.

Keywords Design of mechanical elements · Single-objective optimization ·

Cohort intelligence algorithm · Penalty function approach

K. Marde · A. J. Kulkarni (B)

Symbiosis Institute of Technology, Symbiosis International (Deemed University),
Pune 412115, Maharashtra, India
e-mail: anand.kulkarni@sitpune.edu.in; kulk0003@uwindsor.ca
K. Marde
e-mail: kishor.marde@sitpune.edu.in
A. J. Kulkarni
Odette School of Business, University of Windsor, 401 Sunset Avenue, Windsor, ON N9B3P4,
Canada
P. K. Singh
ABV - Indian Institute of Information Technology and Management Gwalior, Gwalior 474015,
Madhya Pradesh, India
e-mail: pksingh@iiitm.ac.in; pksingh7@gmail.com

© Springer Nature Singapore Pte Ltd. 2019 1

A. J. Kulkarni et al. (eds.), Socio-cultural Inspired Metaheuristics,
Studies in Computational Intelligence 828,
https://doi.org/10.1007/978-981-13-6569-0_1
2 K. Marde et al.

1 Introduction

Several nature-/bioinspired metaheuristic techniques have been proposed so far, such

as Swarm Intelligence (SI) methods and Evolutionary Algorithms (EAs). The major
SI methods include Ant Colony Optimization (ACO) [1], PSO [2], ABC [3], bacterial
foraging optimization algorithm [4], bat algorithm [5], cuckoo search algorithm [6],
glowworm swarm optimization [7], firefly optimization [8], predator–prey algorithm
[9], etc. The EA-based methods include Genetic Algorithm (GA) [10, 11], evolu-
tionary strategies [12], biogeography-based optimization algorithm [13], differential
evolution [14], artificial immune system [15], memetic algorithms [16], learning
classifier systems [17], etc. There are few other optimization algorithms that are also
available such as backtracking search algorithm [18], harmony search algorithm [19],
random optimization algorithm [20], random search algorithm [21], scatter search
algorithm [22], tabu search algorithm [23], Teaching–Learning-Based Optimization
(TLBO) [24], etc. Cohort Intelligence (CI) algorithm is an AI-based optimization
methodology [25]. It is a socio-inspired optimization method in which cohort candi-
dates learn from one another through interaction and competition to achieve a goal
which is common to all. In every learning attempt, every candidate chooses certain
candidate in the cohort from which it can learn certain qualities. This may make the
follower candidate improve its current behavior. The quality here refers to the vari-
ables in the system, and the behavior refers to the objective function. It is important to
mention here that the decision to follow other candidate is based on the roulette wheel
approach which helps to incorporate the uncertainty in the system. As demonstrated
in the previous studies [25, 26], due to the roulette wheel approach, there are chances
that a candidate may follow another candidate with worse behavior; however, it helps
the individual candidate and the entire cohort to jump out of the local minima. The
cohort is assumed to have converged when there is no significant change in the can-
didates’ behavior for considerable number of learning attempts. So far, CI has been
tested for solving several unconstrained test problems [25] as well as combinato-
rial problems such as 0–1 Knapsack problems [27], and real-world problems from
healthcare domain and large-sized cross-border shipper problem [26]. In addition,
a Modified CI (MCI) and its hybridized version with K-means were also developed
solving several cases of clustering problems [28]. This paper investigates the abil-
ity of the CI algorithm for solving problems from mechanical engineering design
domain. The well-studied problems such as volume minimization of the closed coil
helical spring problem, weight minimization of the hollow shaft problem, weight
minimization of the belt pulley drive problem, and manufacturing cost minimization
of a helical spring problem were solved. These problems are continuous and con-
strained in nature. A generic constraint handling technique such as penalty function
approach [29] was incorporated into the CI method.
Optimum Design of Four Mechanical Elements … 3

The rest of the paper is organized as follows: Sect. 2 proposes constrained CI

method in detail. Section 3 describes the four mechanical engineering problems, the
detailed CI solutions, and comparison with other algorithm solutions. The conclu-
sions and future directions are discussed at the end of the paper.

2 Constrained Cohort Intelligence Method

Consider a mechanical engineering constrained problem in the sense of minimization.

Minimize f (x)  f (x1 , . . . , xi , . . . , x N )

Subject to h i (x)  0, i  1, . . . , q
gi (x) ≤ 0, i  1, . . . , k
ϑiminimum ≤ xi ≤ ϑimaximum , i  1, . . . , N
 
ϑi  ϑiminimum , ϑimaximum (1)

For handling constraints, the problem is transformed into

Minimize ϕ(x, R)  f (x) + T (g(x), h(x), R)

ϑiminimum ≤ xi ≤ ϑimaximum , i  1, . . . , N (2)

where T is the penalty function controlled by the parameter R,

 p 
 
s
T (g(x), h(x), R)  R [h i (x)] +
2
[gi (x)]2
(3)
i0 i0

Consider a cohort with M candidates with every candidate  m(m  1, . . . , M), a

set of associated qualities/attributes/characteristics xm  x1m , xim , x Nm which makes
its behavior ϕ(xm , R). As a natural tendency, every candidate observes its own behav-
ior as well as behavior of every other candidate  (m) inthe cohort and tries to follow
better behavior than its current behavior, i.e., ϕ x(m) , R < ϕ(xm , R) based on certain
probability. This way, every candidate follows behavior of one another and entire
cohort behavior is improved. The cohort is considered saturated/converged if none
of the candidates improves its behavior for certain number of learning attempts. The
steps of implementation of CI are described as follows [25].
Step 1: Initialize with number of candidates M in the cohort, sampling interval ϑi ϑi
for each quality xi , i  1, . . . , N and set up the sampling interval reduction factor
r ∈ [0, 1], convergence parameter ε and number of variations k.
Step 2: For minimization problem, the probability of selecting the behavior ϕ(xm , R)
of every associated candidate m(m  1, . . . , M) is calculated as follows:
4 K. Marde et al.

1/ϕ(xm , R)
pm  M
, m  1, . . . , M (4)
m1 1/ϕ(xm , R)

Step 3: Every candidate m  1, . . . , M generates a random number rand ∈ [0, 1]

and
 using a roulette wheel approach decides to follow corresponding behavior
m{?]
ϕ xm[?] , R and associated qualities xm[?]  x1m[?] , . . . , xim[?] , . . . , x N . The super-
script [?] indicates that the behavior is selected by candidate m and not known in
advance.
Step 4: Every candidate m(m  1, . . . , M) shrinks the sampling interval ϑim[?] , i 
1, . . . , N associated with every variable xim[?] , i  1, . . . , N to local neighborhood
as follows:
m[?] m[?] m[?]
ϑi ∈ xi − ϑiminimum − ϑimaximum × r/2 , xi + ϑiminimum − ϑimaximum × r/2 (5)

Step 5: Each candidate m(m  1, . . . , M) samples k qualities from within

updated sampling interval ϑim[?] , i  1, . . . , N associated with every quality
xim[?] , i  1, . . . , N and compute a set of associated v behaviors, i.e., ϕ m,v 
ϕ(xm , R)1 , . . . , ϕ(xm , R) j , .., ϕ(xm , R)v , and selects the best behavior ϕ ∗ (xm , R)
from within. This makes the cohort available with M updated behaviors represented
as
      
ϕ M  ϕ ∗ x1 , R , . . . , ϕ ∗ xm , R , .., ϕ ∗ x M , R .

Step 6: The cohort behavior could be considered saturated, if there is no significant

improvement in the behaviors ϕ ∗ (xm , R) of every candidate m(m  1, . . . , M) in
the cohort, and the difference between the individual behaviors is not very significant
for successive considerable number of learning attempts, i.e., if
  n  n−1 
 
1. max ϕ M − max ϕ M ≤ε
     
 n n−1 
2. min ϕ M − min ϕ M  ≤ ε and
  M n  M n−1 
 
3. max ϕ − min ϕ  ≤ ε, every candidate m(m  1, . . . , M) expands
the sampling interval ϑim[?] , i  1, . . . , N associated with every quality xim[?] , i 
1, . . . , N to its original one ϑiminimum ≤ xi ≤ ϑimaximum , i  1, . . . , N .
Step 7: If either of the two criteria listed below is valid, accept any of the M behaviors
in the cohort as the final objective function value ϕ ∗ (x, R) as the final solution and
stop, else continue to Step 2.
1. If maximum number of attempts exceeded.
2. If cohort saturates to the same behavior (satisfying the condition in Step 6) for
τmax times (Fig. 1).
Optimum Design of Four Mechanical Elements … 5

START

Initialize number of candidates in the cohort, quality variations ,

and set up interval reduction factor

The probability associated with the behavior being followed by every

candidate in the cohort is calculated

Using roulette wheel approach every candidate selects behavior to

follow from within the available choices

Every candidate shrinks/expands the sampling interval of every qual-

ity i based on whether condition of saturation is satisfied

Every candidate forms k behaviors by sampling the qualities from

within the update sampling intervals

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

Fig. 1 Cohort Intelligence (CI) algorithm flowchart [25]

6 K. Marde et al.

3 Application to Mechanical Engineering Problems

As mentioned before, we considered mechanical engineering problems such as vol-

ume minimization of the closed coil helical spring problem, weight minimization
of the hollow shaft problem, weight minimization of the belt pulley drive problem,
and manufacturing cost minimization of a helical spring problem which were solved.
These problems have been solved using AI-based optimization techniques such as
GA [10, 11, 30], PSO [2], ABC [3], TLBO [24], and DTLBO [31]. As these prob-
lems are constrained in nature, these algorithms were required to be incorporated
with certain constrained handling techniques. Singh et al. (2005) used penalty func-
tion approach for handling constraints, whereas another feasibility-based rule was
used in association with other optimization techniques listed above. Similarly, in the
constrained CI algorithm discussed above, penalty function approach was incorpo-
rated to solve these problems. The constrained CI algorithm and mathematical model
of each problem were coded in MATLAB 8.3.0.535 (R2014a), and simulations were
run on a Windows platform using Intel(R) Core(TM)2 Duo T6600, 2.20 GHz proces-
sor speed with 3.00 GB RAM. Every problem was solved 20 times. The individual
problem and solution details are discussed below.

3.1 Problem 1: Closed Coil Helical Spring

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

Fig. 2 Closed coil helical spring [31]

Terminology:
U Volume of spring (cm3 )
Nc Number of active coils

d Wire diameter (cm)

D Mean coil diameter (cm)
dmin Minimum diameter of wire (cm)

Dmax Maximum outside diameter of spring (cm)

Cf Stress factor or Wahl factor

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)

δl Deflection under the maximum load (cm)

Fp Preload compressive force (kg)

δp Deflection under the preload (cm)

δw Deflection from preload to maximum load (cm)

δ pm Maximum deflection under preload (cm)

lf Free length of spring (cm)

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.

several constraints. The optimization criterion is to minimize the volume of a closed

coil helical spring under several constraints.
In this paper, the same objective function is considered, those considered by [31].
The volume of the spring (U) can be minimized subjected to the constraints discussed
as follows:

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

where shear modulus (G) is equal to 80854.6 kgf/cm2 .

The deflection under maximum working load is given by

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.

l f  δl + 1.05(Nc + 2)d (11)

Thus, constraints are given by

lmax − l f ≥ 0 (12)
Optimum Design of Four Mechanical Elements … 9

where lmax is set equal to 35.56 cm.

The wire diameter must exceed the specified minimum value, and it should satisfy
the following condition:

d − dmin ≥ 0 (13)

where dmin is equal to 0.508 cm.

The outside diameter of the coil must be less than maximum specified value, and
it is

Dmax − (D + d) ≥ 0 (14)

where Dmax is equal to 7.62 cm.

The mean coil diameter must be at least three times greater than wire diameter to
ensure that spring is not tightly wound, and it is represented as

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)

where δ pm is equal to 15.24 cm.

The combined deflection must be consistent with the length and same can be
represented as

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

Fig. 3 CI convergence plot for closed coil helical spring

where δw is made equal to 3.175 cm.

During optimization, the ranges for different variables are kept as follows:

0.508 ≤ d ≤ 1.016, 1.270 ≤ D ≤ 7.620, 15 ≤ Nc ≤ 25.

Therefore, the above mentioned problem is a constraint optimization with a single-

objective function subjected to eight constraints.
The closed coil helical spring problem was solved 20 times with number of candi-
dates M  5, the number of variations k  115, and the sampling interval reduction
factor r  0.9. The solution including objective function (Spring Volume U) 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 1. The CI solutions were compared with several optimization algorithms
such as GA, PSO, ABC, TLBO, and DTLBO. The best, mean, and worst objective
function (Volume) values, average, and minimum computational time of CI solving
the problem are presented in Table 2. It is observed that CI algorithm yielded better
solutions than DTLBO, TLBO, PSO, GA, and ABC. The improvements were 0.44%,
0.49%, 0.89%, 0.91%, and 1.22% as compared to DTLBO, TLBO, ABC, PSO, and
GA, respectively. The CI convergence plot is presented in Fig. 3. 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.
Table 1 Performance of the proposed Cohort Intelligence (CI) for closed coil helical spring
Trials Obj Time (s) FE d D Nc Equation (7) Equation (12) Equation (13) Equation (14) Equation (15) Equation (17) Equation (18) Equation (19)

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)

Fig. 4 Hollow shaft [31]

3.2 Problem 2: Design of Hollow Shaft

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

d0 Outside diameter of hollow shaft, cm

d1 Inside diameter of hollow shaft, cm
T Twisting moment or torque.
G Modulus of rigidity, kgf/cm2
θ Angle of twist
J Polar modulus
K Ratio of inner diameter to outer diameter
Tcr Critical buckling load, kg/cm

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

Ws  cross sectional area × length × density

π 2 
Ws  d0 − d12 Lρ (20)
4

Substituting the values of L, ρ as 50 cm and 0.0083 kg/cm3 , respectively, one

finds the weight of the shaft (Ws ) and it is given by
 
Ws  0.326d02 1 − K 2 (21)

[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

Now, θ applied should be greater than T L/G J ; that is, θ ≥ GT LJ .

 of θ , T 4, G, J as4 2π
 /180 per m length, 1.0 × 10 kg/cm,
5
Substituting the values
0.84 × 10 kg/cm and (π/32)d0 1 − K , respectively, one gets the constraints
6 2

as
 
d04 1 − K 4 − 1736.93 ≥ 0 (23)

The critical buckling load (Tcr ) is given by the following expression:

14 K. Marde et al.

π d03 E(1 − K )2.5

Tcr ≤ √  0.75 (24)
12 2 1 − γ 2

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

d03 E(1 − K )2.5 − 0.4793 ≤ 0 (25)

The ranges of different variables are mentioned below:

7 ≤ do ≤ 25 (26)

0.7 ≤ K ≤ 0.97 (27)

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.

3.3 Problem 3: Design of Belt Pulley Drive

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

Fig. 5 CI convergence plot for hollow shaft

Fig. 6 Belt pulley drive [31]

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

T2 Belt tension in loose side, N

d1 Diameter of first small pulley, cm

d2 Diameter of first large pulley, cm

d11 Diameter of second small pulley, cm

(continued)
Optimum Design of Four Mechanical Elements … 17

(continued)

d21 Diameter of second large pulley, cm

t1 Thickness of first small pulley, cm

t2 Thickness of first large pulley, cm

t11 Thickness of second small pulley, cm

t21 Thickness of second large pulley, cm

N1 Speed of first small pulley, rpm

N2 Speed of first large pulley, rpm

N11 Speed of second small pulley, rpm

N21 Speed of second large pulley, rpm

tb Thickness of belt, cm

σb Maximum permissible tensile stress, kg/cm3

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

W p  0.113047d12 + 0.0028274d22 (29)

[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

d2 N2 < d1 N1 , T1 < σb btb

2864789
σb btb ≥
d2 N2

Substituting σb  30.0 kg/cm2 , tb  1 cm, N 2  250 rpm in the above equation,

one gets

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.

Section 3. Information about the Project

Gutenberg Literary Archive Foundation
The Project Gutenberg Literary Archive Foundation is a non-
profit 501(c)(3) educational corporation organized under the
laws of the state of Mississippi and granted tax exempt status by
the Internal Revenue Service. The Foundation’s EIN or federal
tax identification number is 64-6221541. Contributions to the
Project Gutenberg Literary Archive Foundation are tax
deductible to the full extent permitted by U.S. federal laws and
your state’s laws.

The Foundation’s business office is located at 809 North 1500

West, Salt Lake City, UT 84116, (801) 596-1887. Email contact
links and up to date contact information can be found at the
Foundation’s website and official page at
www.gutenberg.org/contact

Section 4. Information about Donations to

the Project Gutenberg Literary Archive
Foundation
Project Gutenberg™ depends upon and cannot survive without
widespread public support and donations to carry out its mission
of increasing the number of public domain and licensed works
that can be freely distributed in machine-readable form
accessible by the widest array of equipment including outdated
equipment. Many small donations ($1 to $5,000) are particularly
important to maintaining tax exempt status with the IRS.

The Foundation is committed to complying with the laws

regulating charities and charitable donations in all 50 states of
the United States. Compliance requirements are not uniform
and it takes a considerable effort, much paperwork and many
fees to meet and keep up with these requirements. We do not
solicit donations in locations where we have not received written
confirmation of compliance. To SEND DONATIONS or
determine the status of compliance for any particular state visit
www.gutenberg.org/donate.

While we cannot and do not solicit contributions from states

where we have not met the solicitation requirements, we know
of no prohibition against accepting unsolicited donations from
donors in such states who approach us with offers to donate.

International donations are gratefully accepted, but we cannot

make any statements concerning tax treatment of donations
received from outside the United States. U.S. laws alone swamp
our small staff.

Please check the Project Gutenberg web pages for current

donation methods and addresses. Donations are accepted in a
number of other ways including checks, online payments and
credit card donations. To donate, please visit:
www.gutenberg.org/donate.

Section 5. General Information About Project

Gutenberg™ electronic works
Professor Michael S. Hart was the originator of the Project
Gutenberg™ concept of a library of electronic works that could
be freely shared with anyone. For forty years, he produced and
distributed Project Gutenberg™ eBooks with only a loose
network of volunteer support.

Project Gutenberg™ eBooks are often created from several

printed editions, all of which are confirmed as not protected by
copyright in the U.S. unless a copyright notice is included. Thus,
we do not necessarily keep eBooks in compliance with any
particular paper edition.

Most people start at our website which has the main PG search
facility: www.gutenberg.org.

This website includes information about Project Gutenberg™,

including how to make donations to the Project Gutenberg
Literary Archive Foundation, how to help produce our new
eBooks, and how to subscribe to our email newsletter to hear
about new eBooks.