A New Bio-inspired Algorithm:

Chicken Swarm Optimization

Xianbing Meng1,2, Yu Liu2, Xiaozhi Gao1,3, and Hengzhen Zhang1
1

College of Information Engineering, Shanghai Maritime University,

1550 Haigang Avenue, Shanghai, 201306, P.R. China
2
Chengdu Green Energy and Green Manufacturing R&D Center,
355 Tengfei Road No. 2, Chengdu, 610200, P.R. China
3
Department of Electrical Engineering and Automation, Aalto University School
of Electrical Engineering, Otaniementie 17, FI-00076 Aalto, Finland
[email protected], [email protected]

Abstract. A new bio-inspired algorithm, Chicken Swarm Optimization (CSO),

is proposed for optimization applications. Mimicking the hierarchal order in the
chicken swarm and the behaviors of the chicken swarm, including roosters,
hens and chicks, CSO can efficiently extract the chickens swarm intelligence
to optimize problems. Experiments on twelve benchmark problems and a speed
reducer design were conducted to compare the performance of CSO with that of
other algorithms. The results show that CSO can achieve good optimization results in terms of both optimization accuracy and robustness. Future researches
about CSO are finally suggested.
Keywords: Hierarchal order, Chickens behaviors, Swarm intelligence, Chicken Swarm Optimization, Optimization applications.

Introduction

Bio-inspired meta-heuristic algorithms have shown proficiency of solving a great

many optimization applications [1, 2]. They exploit the tolerance for imprecision and
uncertainty of the optimization problems and can achieve acceptable solutions using
low computing cost. Thus the mate-heuristic algorithms, like Particle Swarm Optimization (PSO) [3], Differential Evolution (DE) [2], Bat Algorithm (BA) [1], have attracted great research interest for dealing with optimization applications.
New algorithms are still emerging, including krill herb algorithm [4], and social
spider optimization [5] et al. All these algorithms extract the swarm intelligence from
the laws of biological systems in nature. However, to learn from the nature for developing a better algorithm is still in progress.
In this paper, a new bio-inspired optimization algorithm, namely Chicken Swarm
Optimization (CSO) is proposed. It mimics the hierarchal order in the chicken swarm
and the behaviors of the chicken swarm. The chicken swarm can be divided into
several groups, each of which consists of one rooster and many hens and chicks.
Different chickens follow different laws of motions. There exist competitions between
different chickens under specific hierarchal order.
Y. Tan et al. (Eds.): ICSI 2014, Part I, LNCS 8794, pp. 8694, 2014.
Springer International Publishing Switzerland 2014

A New Bio-inspired Algorithm: Chicken Swarm Optimization

87

The rest of paper is organized as follows. Section 2 introduces the general biology
of the chicken. The details about the CSO are discussed in Section 3. The simulations
and comparative studies are presented in section 4. Section 5 summaries this paper
with some conclusions and discussions.

General Biology

As one of the most widespread domestic animals, the chickens themselves and their
eggs are primarily kept as a source of food. Domestic chickens are gregarious birds
and live together in flocks. They are cognitively sophisticated and can recognize over
100 individuals even after several months of separation. There are over 30 distinct
sounds for their communication, which range from clucks, cackles, chirps and cries,
including a lot of information related to nesting, food discovery, mating and danger.
Besides learning through trial and error, the chickens would also learn from their
previous experience and others for making decisions [6].
A hierarchal order plays a significant role in the social lives of chickens. The preponderant chickens in a flock will dominate the weak. There exist the more dominant
hens that remain near to the head roosters as well as the more submissive hens and
roosters who stand at the periphery of the group. Removing or adding chickens from
an existing group would causes a temporary disruption to the social order until a specific hierarchal order is established [7].
The dominant individuals have priority for food access, while the roosters may call
their group-mates to eat first when they find food. The gracious behavior also exists in
the hens when they raise their children. However, this is not the case existing for
individuals from different groups. Roosters would emit a loud call when other chickens from a different group invade their territory [8].
In general, the chickens behaviors vary with gender. The head rooster would positively search for food, and fight with chickens who invade the territory the group
inhabits. The dominant chickens would be nearly consistent with the head roosters to
forage for food. The submissive ones, however, would reluctantly stand at the periphery of the group to search for food. There exist competitions between different
chickens. As for the chicks, they search for the food around their mother.
Each chicken is too simple to cooperate with each other. Taken as a swarm,
however, they may coordinate themselves as a team to search for food under specific
hierarchal order. This swarm intelligence can be associated with the objective problem to be optimized, and inspired us to design a new algorithm.

Chicken Swarm Optimization

Given the aforementioned descriptions, we can develop CSO mathematically. For

simplicity, we idealized the chickens behaviors by the following rules.
(1) In the chicken swarm, there exist several groups. Each group comprises a dominant rooster, a couple of hens, and chicks.

88

X. Meng et al.

(2) How to divide the chicken swarm into several groups and determine the identity of the chickens (roosters, hens and chicks) all depend on the fitness values of the
chickens themselves. The chickens with best several fitness values would be acted as
roosters, each of which would be the head rooster in a group. The chickens with worst
several fitness values would be designated as chicks. The others would be the hens.
The hens randomly choose which group to live in. The mother-child relationship
between the hens and the chicks is also randomly established.
(3) The hierarchal order, dominance relationship and mother-child relationship in a
group will remain unchanged. These statuses only update every several (G) time steps.
(4) Chickens follow their group-mate rooster to search for food, while they may
prevent the ones from eating their own food. Assume chickens would randomly steal
the good food already found by others. The chicks search for food around their mother (hen). The dominant individuals have advantage in competition for food.
Assume RN, HN, CN and MN indicate the number of the roosters, the hens, the
chicks and the mother hens, respectively. The best RN chickens would be assumed to
be roosters, while the worst CN ones would be regarded as chicks. The rest are treated
as hens. All N virtual chickens, depicted by their positions ,
1, , ,
1, ,
at time step t, search for food in a D-dimensional space. In this work, the
optimization problems are the minimal ones. Thus the best RN chickens correspond to
the ones with RN minimal fitness values.
3.1

Movement of the Chickens

The roosters with better fitness values have priority for food access than the ones with
worse fitness values. For simplicity, this case can be simulated by the situation that
the roosters with better fitness values can search for food in a wider range of places
than that of the roosters with worse fitness values. This can be formulated below.
,

1,
exp

| |

0,
,

1,

(1)

(2)

Where Randn (0,

) is a Gaussian distribution with mean 0 and standard deviation . , which is used to avoid zero-division-error, is the smallest constant in the
computer. k, a roosters index, is randomly selected from the roosters group, f is the
fitness value of the corresponding x.
As for the hens, they can follow their group-mate roosters to search for food.
Moreover, they would also randomly steal the good food found by other chickens,
though they would be repressed by the other chickens. The more dominant hens
would have advantage in competing for food than the more submissive ones. These
phenomena can be formulated mathematically as follows.
,

exp

(3)
(4)

A New Bio-inspired Algorithm: Chicken Swarm Optimization

exp

89

(5)

Where Rand is a uniform random number over [0, 1]. 1

1, ,
is an index
of the rooster, which is the ith hens group-mate, while 2
1, ,
is an index of
the chicken (rooster or hen ), which is randomly chosen from the swarm. 1
2.
Obviously,
,
, thus S2 <1< S1. Assume S1=0, then the ith hen
would forage for food just followed by other chickens. The bigger the difference of
the two chickens fitness values, the smaller S2 and the bigger the gap between the
two chickens positions is. Thus the hens would not easily steal the food found by
other chickens. The reason that the formula form of S1 differs from that of S2 is that
there exist competitions in a group. For simplicity, the fitness values of the chickens
relative to the fitness value of the rooster are simulated as the competitions between
chickens in a group. Suppose S2=0, then the ith hen would search for food in their
own territory. For the specific group, the roosters fitness value is unique. Thus the
smaller the ith hens fitness value, the nearer S1 approximates to 1 and the smaller the
gap between the positions of the ith hen and its group-mate rooster is. Hence the more
dominant hens would be more likely than the more submissive ones to eat the food.
The chicks move around their mother to forage for food. This is formulated below.
,

(6)

1,
.
Where
, stands for the position of the th chicks mother
FL
0,2 is a parameter, which means that the chick would follow its mother
to forage for food. Consider the individual differences, the FL of each chick would
randomly choose between 0 and 2.
Chicken Swarm Optimization. Framework of the CSO
Initialize a population of N chickens and define the related parameters;
Evaluate the N chickens fitness values, t=0;
While (t < Max_Generation)
If (t % G
0)
Rank the chickens fitness values and establish a hierarchal order in the
swarm;
Divide the swarm into different groups, and determine the relationship between the chicks and mother hens in a group; End if
For i = 1 : N
If i == rooster Update its solution/location using equation (1); End if
If i == hen
Update its solution/location using equation (3); End if
If i == chick Update its solution/location using equation (6); End if
Evaluate the new solution;
If the new solution is better than its previous one, update it;
End for
End while
3.2

Parametric Analysis

There exist six parameters in CSO. Humans keep chickens primarily as a source of
food. As the food themselves, only hens can lay eggs, which can also be the source of

90

X. Meng et al.

food. Hence keeping hens is more beneficial for human than keeping roosters. Thus
HN would be bigger than RN. Given the individual differences, not all hens would
hatch their eggs simultaneously. Thus HN is also bigger than MN. Though each hen
can raise more than one chick, we assume the population of adult chickens would
surpass that of the chicks, CN. As for G, it should be set at an appropriate value,
which is problem-based. If the value of G is very big, it's not conducive for the algorithm to converge to the global optimal quickly. While if the value of G is very small,
the algorithm may trap into local optimal. After the preliminary test, G 2,20 may
achieve good results for most problems.
Furthermore, the formula of the chicks movement can be associated with the corresponding part in DE. If we set RN and MN at 0, thus CSO essentially becomes the
basic mutation scheme of DE. Hence the partial conclusions from the DE [2] can be
used. In practice, FL [0.4, 1] usually perform well.

Validation and Comparison

4.1

Benchmark Problems Optimization

Twelve popular benchmark problems [9, 10] (shown in Table 1) are used to verify the
performance of the CSO compared with that of PSO, DE and BA. The statistical results have been obtained, based on 100 independent trials, in all the case studies. The
number of iterations is 1,000 in each trial. For a fair comparison, all of the common
parameters of these methods, such as the population size, dimensions and maximum
number of generations, are set to be the same. The related parameters of these
algorithms are showed at Table 2.
Table 1. Twelve benchmark problems
Problem name
High Conditioned Elliptic
Bent Cigar
Discus
Ackley
Griewank
Sphere
Step
Powell Sum
Rastrigin
Axis parallel hyper-ellipsoid
Brown
Exponential

ID
F1
F2
F3
F4
F5
F6
F7
F8
F9
F10
F11
F12

Dimension
20
20
20
20
20
20
20
20
20
20
20
20

Bounds
[-100,100]
[-100,100]
[-100,100]
[-32,32]
[-600,600]
[-100,100]
[-100,100]
[-1,1]
[-5,10]
[-5.12,5.12]
[-1,4]
[-1,1]

Optimum
0
0
0
0
0
0
0
0
0
0
0
-1

Table 2. The related parameter values

Algorithm
PSO
DE
BA
CSO

Parameters
c1=c2=1.49445, w = 0.729
CR = 0.9, F = 0.6
0,
2,
0, 2 ,
0, 1
0.9,
RN=0.2*N, HN=0.6*N, CN=N-RN-HN, MN=0.1*N, G = 10, FL 0.5, 0.9

A New Bio-inspired Algorithm: Chicken Swarm Optimization

91

There are many variants of PSO, DE and BA. In this work, the basic BA and the
standard PSO are chosen. As for DE, the DE/rand/1/bin scheme is selected. Table 3
displays the statistical comparison of the four algorithms on twelve benchmark problems. It clearly shows that CSO is superior to PSO, DE and BA on all these problems
in terms of accuracy, efficiency and robustness.
The superiority of CSO over PSO, BA and DE should be the case. If we set RN =
CN = 0, and let S1, S2 be the parameters like c1 and c2 in PSO, thus CSO will be
similar to the standard PSO. Hence CSO can inherit many advantages of PSO and
DE. Moreover, the chickens swarm intelligence can be efficiently extracted in CSO.
Given the diverse laws of the chickens' motions and cooperation between the multigroups, the search space can be efficiently explored. Under the specific hierarchal
order, the whole chicken swarm may behave like a team to forage for food, which can
be associated with the objective problems to be optimized. All of these merits enhance the performance of CSO.
Table 3. Statistical comparison of CSO with PSO, DE and BA
Problem

F1

F2

F3

F4

F5

F6

F7

F8

Algorithm
PSO
CSO
BA
DE
PSO
CSO
BA
DE
PSO
CSO
BA
DE
PSO
CSO
BA
DE
PSO
CSO
BA
DE
PSO
CSO
BA
DE
PSO
CSO
BA
DE
PSO
CSO
BA
DE

Best
12600.53084
0
3782.10211
0
0
0
1780594.354
0
0
0
101.92698
0
0
0
1.48288
0
0
0
0.004
0
0
0
1.867408
0
0
0
1
0
0
0
0
0

Mean
49808.05813
0
30996.60571
0
1300
0
2636386.576
0.00001
0
0
2256.99578
0
0
0
2.59402
0.35702
0
0
2.82906
0
0
0
2.94197
0
0
0
3.41
0
0
0
0
0

Worst
101417.29666
0
84110.41779
0
10000
0
3804547.807
0.0001
0
0
6307.26214
0
0
0
3.07403
11.64977
0
0
15.42094
0
0
0
4.18701
0
0
0
6
0
0
0
0
0

Std.
2124.09
1.89943e-60
30996.6057
5.99605e-12
339.7
1.35815e-62
36993.4
1.88465e-6
4.66505e-33
9.37294e-66
132.731
3.22023e-12
1.31168e-16
6.12169e-17
0.02297
0.17096
5.36538e-8
0
0.296984
1.05399e-12
1.8988e-35
4.10796e-70
0.0491192
1.94304e-12
0
0
0.103105
0
4.78569e-60
0
5.02596e-8
0

92

X. Meng et al.
Table 3. (Continued)
Problem

F9

F10

F11

F12

4.2

Algorithm
PSO
CSO
BA
DE
PSO
CSO
BA
DE
PSO
CSO
BA
DE
PSO
CSO
BA
DE

Best
10.94454
0
88.44729
8.41884
0
0
24.34652
0
0
0
4.22819
0
-1
-1
-0.41494
-1

Mean
21.26284
0
121.99296
22.70527
0
0
39.73613
0
1.29
0
5.92480
0
-1
-1
-0.20415
-1

Worst
41.78822
0
167.60654
43.9751
0
0
63.15
0
8
0
7.52686
0
-1
-1
-0.12952
-1

Std.
0.60048
0
1.57913
0.706825
2.13096e-37
1.59801e-71
0.749752
4.15726e-14
0.171595
5.71381e-72
0.0726235
2.37888e-15
9.58157e-18
0
0.0052408
1.75511e-16

Speed Reducer Design

Design of the speed reducer [11] (as shown in Fig. 1) is to design a gearbox, which
can be rotated at its most efficient speed. The gearbox is described by the face width
b
, module of teeth m
, number of teeth in the pinion z
, length of the first
shaft between bearings h1
, length of the second shaft between bearings h2
,
diameter of the first shaft d1
, and diameter of the first shaft d2
. The optimization in the design of the speed reducer is to minimize its total weight, subject to
constraints on bending stress of the gear teeth, surface stress, transverse deflections of
the shafts, and stresses in the shafts. This problem can be formulated as follows.
Minimize
0.7854
3.3333
14.9334
43.0934
7.4777
0.7854
1.508
.

Subject to
16.9

10

157.5
.

10
.

Where 2.6
3.6, 0.7
0.8, 17
28, 7.3
8.3, 7.8
1
1, 2, 3, 11 .
3.9, 5
5.5,
8.3, 2.9
Table 4 summarizes a comparison of the results achieved by CSO and other algorithms. It clearly shows that CSOs results outperform all the results achieved by the
six methods in terms of both optimization accuracy and robustness. The best solution
achieved by CSO is
(3.5, 0.7, 17, 7.308, 7.802, 3.35, 5.287) with
=
2996.60481329. The constraint values are
= (-0.07, -0.2, -0.5, -0.9, -2.33e-6, 1.06e-5,-0.7,-5.06e-5,-0.58,-0.05,-0.01), which indicates that the solution is feasible.

A New Bio-inspired Algorithm: Chicken Swarm Optimization

d2
z

93

h2

1
d

Fig. 1. Speed reducer

Table 4. Optimization results of the speed reducer design

Discussions

Mimicking the chickens behaviors, a new bio-inspired algorithm, namely Chicken

Swarm Optimization was proposed for optimization problems. The performance of
CSO is compared with that of the PSO, DE and BA on twelve benchmark problems.
Experiments show that CSO outperforms the PSO, DE and BA in terms of both optimization accuracy and robustness. Moreover, CSO can efficiently solve the speed
reducer design, which endues the CSO with a promising prospect of further studying.
One of the reasons that CSO has very promising performance is that CSO inherits
major advantages of many algorithms. PSO and the mutation scheme of DE are the
special cases of the CSO under appropriate simplifications. What is more significant
for the superiority of the CSO is that the chickens swarm intelligence can be efficiently extracted to optimize problems. The chickens' diverse movements can be conducive for the algorithm to strike a good balance between the randomness and determinacy for finding the optima. The whole chicken swarm consists of several groups,
namely multi-swarm. Through integration of the hierarchal order, chickens of the
different groups may behave as a team and coordinate themselves to forage for food.
Thus CSO can behave intelligently to optimize problems efficiently.
The innovation in this paper not only lies in efficiently extracting the chickens
swarm intelligence to optimize problems, but also making CSO innate multi-swarm
method. Multi-swarm technique is usually used to enhance performance of the population-based algorithm. As an innate multi-swarm algorithm, various multi-swarm

94

X. Meng et al.

techniques can be used to develop the different variants of CSO. Thus CSO has good
extensibility. Moreover, from the parametric analysis, the population of the hens is the
biggest in the swarm. Thus the performance of CSO largely depends on how the hens
swarm intelligence can be extracted to optimize problems. The motion of the hens can
be adaptively controlled according to the fitness value of the problem itself. With the
dynamical hierarchal order, the hens swarm can be updated. Hence CSO has the selfadaptive ability to solve the optimization problems.
More comprehensive analyses on the CSO are still need to be investigated in the
future. Moreover, we can consider there exist several roosters in a group and dynamically adjust the population of the hens and chicks in each group. Its also significant
to tune the related parameters for enhancing the algorithm performance, and design
the variants of the CSO to solve many optimization applications.
Acknowledgments. This research work was funded by the New Academic Staffs
Program of Shanghai Maritime University under Grant GK2013089.

References
1. Yang, X.S.: Bat algorithm: literature review and applications. International Journal of Bioinspired Computation 5(3), 141149 (2013)
2. Das, S., Suganthan, P.N.: Differential evolution: A survey of the state-of-the-art. IEEE
Transactions on Evolutionary Computation 15(1), 431 (2011)
3. Jordehi, A.R., Jasni, J.: Parameter selection in particle swarm optimization: A survey.
Journal of Experimental & Theoretical Artificial Intelligence 25(4), 527542 (2013)
4. Gandomi, A.H., Alavi, A.H.: Krill herd: A new bio-inspired optimization algorithm.
Communications in Nonlinear Science and Numerical Simulation 17, 48314845 (2012)
5. Cuevas, E., Cienfuegos, M., Zaldivar, D., Cisneros, M.: A swarm optimization algorithm
inspired in the behavior of the social-spider. Expert Systems with Applications 40,
63746384 (2013)
6. Smith, C.L., Zielinski, S.L.: The Startling Intelligence of the Common Chicken. Scientific
American 310(2) (2014)
7. Grillo, R.: Chicken Behavior: An Overview of Recent Science,
http://freefromharm.org/chicken-behavior-an-overview-ofrecent-science
8. Chicken, http://en.wikipedia.org/wiki/Chicken
9. Tan, Y., Li, J.Z., Zheng, Z.Y.: ICSI, Competition on Single Objective Optimization
(2014), http://www.ic-si.org/competition/ICSI.pdf
10. Yang, X.S.: Nature-inspired optimization algorithm. Elsevier (2014)
11. Robert, R., Mostafa, A.: Embedding a social fabric component into cultural algorithms
toolkit for an enhanced knowledge-driven engineering optimization. International Journal
of Intelligent Computing and Cybernetic 1(4), 563597 (2008)
12. Mezura, M.E., Hernandez, O.B.: Modified bacterial foraging optimization for engineering
design. In: Proceedings of the Artificial Neural Networks in Engineering Conference, vol. 19,
pp. 357364. Intelligent Engineering Systems Through Artificial Neural Networks (2009)
13. Akay, B., Karaboga, D.: Artificial bee colony algorithm for large-scale problems and engineering design optimization. Journal of Intelligent Manufacturing 23(4), 10011014 (2012)
14. Gandomi, A.H., Yang, X.S., Alavi, A.H.: Cuckoo search algorithm: A metaheuristic approach
to solve structural optimization problems. Engineering with Computers 29, 1735 (2013)