THE STUDY OF CUCKOO OPTIMIZATION ALGORITHM FOR PRODUCTION PLANNING PROBLEM

International Journal of Computer-Aided Technologies (IJCAx) Vol.2, No.3, July 2015
DOI:10.5121/ijcax.2015.2301 1
THE STUDY OF CUCKOO OPTIMIZATION
ALGORITHM FOR PRODUCTION PLANNING
PROBLEM
A. Akbarzadeh1
, E. Shadkam2
1,2
Department of Industrial Engineering, Faculty of Eng.; Khayyam University, Mashhad,
Iran
2
Department of Industrial Engineering, Isfahan University of Technology, Isfahan, Iran
ABSTRACT
Constrained Nonlinear programming problems are hard problems, and one of the most widely used and
common problems for production planning problem to optimize. In this study, one of the mathematical
models of production planning is survey and the problem solved by cuckoo algorithm. Cuckoo Algorithm is
efficient method to solve continues non linear problem. Moreover, mentioned models of production
planning solved with Genetic algorithm and Lingo software and the results will compared. The Cuckoo
Algorithm is suitable choice for optimization in convergence of solution.
KEYWORDS
Meta-heuristic algorithms, Cuckoo Optimization Algorithm, Lot Sizing, Production Planning.
1. INTRODUCTION
The optimization problems are so important in many different fields of science like physics,
chemistry and engineering and their purpose is to find the best possible answer for an obvious
problem.
The need of searching is the reason that many different searching algorithms have been provided.
The evolutionary algorithms are a group of algorithms based on random searching that is inspired
by the natural biological evolution modeling. They work on possible answers that have a superior
feature and the longer generation survival that bring a closer estimate of the optimized answer.
One of the newest evolutionary algorithms is the Cuckoo optimization algorithm. The current
version of this algorithm is used to solve the continuous optimization problems and proved its
performance in this field. Because of the ability of this algorithm in solving the continuous
problems, it also has been discredited so that can solve the discrete issues too. The production
planning is one of the problems that has a wide practical usage and is solvable in the discrete
space. Therefore many expanded solving methods provide for this issue at all times. Considering
the importance of this kind of problems, it comes to our mind to solve this problem using the
Cuckoo algorithm.

2
One of the main models of production planning is to determine the lot sizing that has been
investigated in different cases like multi period, multi item and that has capacity constraints with
probable demand, removing the shortage cost and adding the level of services constraint by Bijari
and at al. [1]
Also another model about determining the lot sizing and production scheduling with purpose of
maximizing the profit has been investigated that in this model, the flexibility of choosing demand
is imagined (the amount of chosen demands to comply is calculated by the model). In this article,
a mathematical model has been suggested for PGLSP problem and it has been investigated [2].
Merzifonlou lu and Geunes studied the production planning problem with the hypothesis of
flexibility in demand and the absence of capacity constraint. Not only they determined the
amount of production in each period, but also they introduced a decision variable that can specify
the percentage of accepted demand of each product in each period. The objective function of this
problem is the same as the maximum profit objective function [6]. On the other hand, many
algorithms and meta-heuristic algorithms have been provided in order to solve the production
timing problem like “Ant colony optimization algorithm” [7], “Simulated annealing”[8], “Genetic
algorithm” [9,10,11,12,13,14,15,16,17,18] and “Tabu search” [9]. Many articles are written about
helpful nonlinear programming optimization [3], picture segmentations [4], Wind farms capacity
determination [5], etc.
In the second section of this article, the Cuckoo algorithm will be explained, the third section
belongs to analyzing the production planning model, the Cuckoo algorithm will be used in the
mentioned problem in the fourth section and at last the conclusion will be provided.
2.1. Introducing the Cuckoo optimization algorithm
The Cuckoo optimization algorithm expanded by X. S. Yang and S. Deb in 2009 [20].
The Cuckoos laying method combined with the Levy Flight were the first clues of this algorithm
creation instead of simple random isotropic hike. Later, R.Rajabioun investigated the Cuckoo
optimization algorithm in detail in 2011 [21].
Like other evolutionary algorithms, this one begins with an initial population. This method works
as the following steps. We assume that this population owns some eggs. First they put these eggs
in other kind of birds’ nests then wait until the host bird maintain these eggs beside her eggs. In
fact, this lazy bird perfectly makes other birds to play an involuntary role in her generation
survival. Some of the eggs that have less similarity to the host bird’s eggs will be recognized and
destroyed.
In fact the cuckoos improve and learn how to lay eggs more like the target host bird’s eggs
continuously and the host birds learn how to recognize the fake eggs.
More number of survived eggs in each zone shows more suitability of that zone, and more
number of survived eggs, more attention pay to that zone and in fact this is the parameter that the
Cuckoo optimization algorithm wants to optimize.

3
The Cuckoo optimization algorithm is as the rest:
Figure 1: The Cuckoo optimization algorithm diagram
The variable values of the problem should take shape out of an array In order to solve an
optimization problem. That array is called “habitat”.
In an optimization problem, the next var
N of a habitat will be a var
1 N
× array that shows the
current living location of cuckoos. This array describes as:
[ ]
1 2 var
, ,..., N
habitat X X X
=
The suitability (profit) in the current habitat obtains by evaluating the profit function ( p
f ) in the
habitat. So:
1 2 var
( ) ( , ,..., )
N
profit f b habitat f b X X X
= =
A habitat matrix in size of pop var
N N
∗ will be prepared for starting an optimization algorithm,
then for each habitat a random number of eggs will be allocated.

4
The laying radius will be calculated by considering the number of eggs that each cuckoo lays and
also the distance between the cuckoos and the current optimized zone. After that the cuckoos
begin to lay in that zone. The laying radius calculates as:
( )
'
hi low
Number of current cuckoo s eggs
ELR a Var Var
Total number of eggs
= × × −
Then each cuckoo begins to lay her eggs in the nests within her ELR
So after each laying round, p% of eggs (usually 10%) that is less profitable (their profit function
is in lower level) destroys. Other chicks power up and grow in the host nests.
2.2.The cuckoo’s migration
The cuckoos live in their environments while they are growing up and get older but when the
laying time comes, they migrate to better habitats where the eggs have more chance to survive.
After composing the groups in general different living locations (justified region or search space
of the problem), the group with the best location will be targeted and other cuckoos will migrate
there.
When the grown cuckoos live all around the environments, it is hard to find out each cuckoo
belongs to which group. For solving this problem, the cuckoos will be grouped by the “K-means”
method which is a classic way of grouping (finding a K between 3 and 5 is usually acceptable).
When the cuckoos migrate to the target, they don’t travel the direct way. They just travel (λ %) of
the way with the deflection of (φ) as it is shown in figure 2.
Figure 2: the cuckoos’ migration to the target
These two parameters (λ,φ) helps cuckoos to explore a larger area. λ is a random number between
0 and 1 and φ is a number between
6
π
− and
6
π
.
The migration formula is:
Next Habitat current Habitat Goal Point current Habitat
X X F(X X )
= + −

5
After some iteration, all of the cuckoos gathered in an optimized point where the eggs have the
most similarity to the host ones and access to the richest food sources is available. This location
contains the most probable profit and the least possible number of killed eggs. Convergence more
than 95% of all cuckoos in a single point ends the COA.
3. INTRODUCING THE MODEL AND ITS VARIABLES AND PARAMETERS
In this section, one of the production planning models will be introduced. The assumption is that
a number of products have to be manufactured in a specific number of periods with the condition
of minimizing its cost. Determining the optimized number of each product is the purpose. The
parameters and the decision variables that is needed for the production planning problem, is given
in the table 1 and 2.
Table1: the parameters of the model
Number of all products
N
Number of all sources for producing
M
Price of the product i in the period t
ܲ௜௧
Cost of the product i in the period t
‫ܥ‬୧୲
Maintenance cost of the product i in the period t
ℎ୧୲
The source capacity m in the period t
‫݌ܽܥ‬୫୲
Maximum capacity of source i in the period t
‫ܯ‬୧୲
The occupied capacity for the product i
‫ݕ‬୧
The requirement of product i in the period t
‫ܦ‬୧୲
The intake of source m for the product i
ܽ୧୫
The preparing cost of the product i in the period t
‫ܣ‬୧୲
Table2: the decision variables of the model
The amount of sold product i in the period t
ܑ࢙‫ܜ‬
The amount of product i in the period t
‫ݔ‬୧୲
Dearth of the product i in the period t
‫ܤ‬୧୲
Holding amount of the product i in the period t
‫ܫ‬୧୲
The delayed demand of the product i transported from the period t-1 to the
period t
‫ݎ‬୧୲
The objective function purpose is maximizing the income of the sold products with reducing the
production costs and maintenance and preparation costs. The objective function will be changed

6
into a minimizing cost function by multiplying it to (-1), because many of the existing problems
are minimizing cost problems and the model is for these kind of problems.
The mathematical model is at the rest:
N T N T N T N T N T
i=1 t=1 i=1 t=1 i=1 t=2 i=1 t=1 i=1 t=1
1
1
( 1)
( 1) ( 1)
minimize Z= pit.Sit+ cit.X hit.I Zit.Ait bit.Bit
S.T
.
.
. 0
.
it
N
i it it
i
N
im it
i
it it it i t
it it i t it it it i t
it it
y I M
a X
S D B r B
X B I I D r B
X M Z
=
=
−
− −
− + + +
+ ≤
= + +
+ + − − − =
≤
∑∑ ∑∑ ∑∑ ∑∑ ∑∑
∑
∑
In general, the first restriction is for the depository, the second one is for the used source, the third
one is the selling balance, the fourth one is holding balance and the last one is the production with
the condition of preparation.
We assume that the amount of demand for each product is apparent in each period and the
delayed demand of each period will be transported to the next period. In this article, the
production of 3 products in 5 periods is studied.
4. IMPLEMENTING THE CUCKOO ALGORITHM
There are number of input parameters in the Cuckoo optimization like other heuristic algorithm.
The amounts of these parameters affect the final answer directly. The effects of these parameters
are given in the table 3.
Table3: the changes of the input parameters
Cluster
Cuckoos
max Cuckoos
Cost
Total Time
3
7
30
6.5282E+06
25.525 s
3
7
8
6.5099E+06
12.591 s
6
7
8
6.5235E+06
26.215 s
3
30
35
6.5257E+06
28.172 s
2
7
8
6.5180E+06
12.762 s

7
So the parameters regulation is one of the most important factors that affect the final answer.
According to the amounts that are given in table 3 and our minimizing cost model, we calculated
the best answer in the shortest possible time for the model. So the best input parameters for the
model are given in table 4.
Table 4: the algorithm parameters
value
Parameters
7
Number Of Initial Cuckoos
2
Minimum Number Of Eggs Laid by Each
Cuckoo
4
Maximum Number Of Eggs Laid by Each
Cuckoo
200
Maximum Number Of Iterations
3
Number Of Clusters for KNN
8
Maximum Number of Living Cuckoo
Figure 3 shows the high isotropy of the algorithm in finding the answers of this problem.
This figure shows the high accuracy of the calculated answers of this algorithm. Comparing this
algorithm with Genetic algorithm, the Cuckoo algorithm has fewer number of initial population
and gives better answers in the shorter time.
Figure3: the isotropy speed of the algorithm for 200 iterations
If checking the operation of meta-heuristic algorithms for finding the optimized answer and
determining the production set would be in order, the model should be solved with the default
parameters with different methods. The general conclusion will be obtained by comparing the
results.
Also it is so obvious that if we implement a single meta-heuristic algorithm for a single problem a
few times, we probably will come up with different answers for the optimized (or near optimized)
quantity of the objective function. It is because of many meta-heuristic algorithms such as
0 50 100 150 200 250
6.5117
6.5118
6.5119
6.512
6.5121
6.5122
6.5123
6.5124
x 10
6
Cuckoo iteration
Cost
Value
Curent Cost = 6511715.75 , at Iteration = 201

8
Genetic and the Cuckoo algorithm use the random operators in order to find the optimized
answer.
Because the model is nonlinear and doesn’t have the optimized answer, we solved this problem
by Genetic algorithm and the Lingo 11 software. After 20 iterations, the average of the answers is
given in the table 5.
Table5: comparing the average of the answers after 20 iterations
Lingo
COA
GA
average
average
best
deviation
average
2.69e+09
6.52e+06
6.50e+06
1.38e+04
2.38e+11
1 s
16.935 s
16.674 s
-------
18 s
5. CONCLUSION
In this article, two of the meta-heuristic algorithms named Genetic algorithm and the Cuckoo
algorithm have been used for solving one of the production planning problems. Because the
Cuckoo algorithm is the expanded version of the Genetic algorithm, it gives better answers.
Also the Lingo software provides an answer near the optimized answer using the Global Solver
ability but with increasing the scale of the problem, the calculation timing will be incomputable.
The Genetic algorithm cannot find the optimized answer but it finds the answer in an acceptable
duration. The Cuckoo algorithm finds better answers of this problem in the suitable duration but
we cannot name it as the best way finding answer in contrast of other algorithms because each
year, many algorithms are published that covers the defects of earlier algorithms. But in this
problem, the Cuckoo algorithm could easily find the answer with lower number of initial
population and less iteration.
The subject of this article can be expanded in many other ways such as solving the model with
other algorithms or adding the probable demand and the transportation cost restriction to the
model in case of comparing two algorithms.
REFERENCES
[1] Shirneshan, H., Bijari, M., Moslehi Gh., Determination the lot sizing with probable demand and the
level of the service, Production and Operations Management, 1391.
[2] Sereshti, N., Bijari, M., the Maximization of profit in general lot sizing and sequencing problem, the
7th International Conference of Industrial Engineering, 1389.
[3] Abdullahi, M. Abdullahi, D., Karimpur, J., Cuckoo optimization algorithm for non linear
programming , 9th International Conference on Industrial Engineering, 1391.
[4] Mahmoudi, Sh., the segmentation of color image by Cuckoo optimization algorithm, the first PC-IT
and Electronics Engineering Conference, Islamic Azad University.

9
[5] Hakimi Gilani, S., Vafrakhth, H., Determine the capacity of wind power plants by Cuckoo
optimization algorithm, the second Conference on Renewable Energy and Distributed Generation
Iran, Tehran University, 2012.
[6] Merzifonluo lu, Y., and Geunes, J., Uncapacitated production and Location planing models with
demand fulfillment flexibility, International Journal of production Economics, Vol.102, No. 2, pp.
199-216, 2006.
[7] Hwa Huang, R., Yang, CL., Overlapping Production Scheduling Planning with Multiple
Objectives—An ant Colony Approach. International Journal of Production Economics 115, 2008,
pp. 163-170.
[8] Loukil, T., Teghem, J., Fortemps, Ph., A Multi-Objective Production Scheduling Case Study
Solved by Simulated Annealing. European Journal of Operational Research 179, 2007, pp. 709-722.
[9] Zegordi, S, Kamal Abadi, I., Beheshti Nia, M., A Novel Genetic Algorithm for Solving
Production and Transportation Scheduling in a Two-Stage Supply Chain. Computers & Industrial
Engineering 58, 2010, pp. 373-381.
[10] Ying Wu, L., Dong Hu, Y., Mei Xu, D., Hua, B., Solving Batch Production Scheduling using
Genetic Algorithm. Computer Aided Chemical Engineering 15, 2003, pp. 648-653.
[11] Tat Chan, W., Hu, H., An Application of Genetic Algorithms to Precast Production
Scheduling. Computers & Structures 79, 2001, pp. 1605-1616.
[12] Sankar, A.S., Ponnanbalam, S.G., Rajendran, C., A Multiobjective Genetic Algorithm for
Scheduling a Flexible Manufacturing System. International Journal of Advanced Manufacturing
Technology, 22(3), 2003, pp. 229–236.
[13] Liu, J., Tang, L., A Modified Genetic Algorithm for Single Machine Scheduling. Computers and
Industrial Engineering, 37(1), 1999, pp. 43–46.
[14] Li, Y., Ip, W., Wang, D., Genetic Algorithm Approach to Earliness and Tardiness Production
Scheduling and Planning Problem. International Journal of Production Economics 54, 1998, pp. 65-
76.
[15] Knosala, R., Wal, T., A Production Scheduling Problem using Genetic Algorithm. Journal of
Materials Processing Technology 109, 2001, pp. 90-95.
[16] Jou, C., A Genetic Algorithm with Sub-Indexed Partitioning Genes and its Application to
Production Scheduling of Parallel Machines. Computers & Industrial Engineering 48, 2005, pp. 39-
54.
[17] Ho Ko, C., Fan Wang, Sh., Precast Production Scheduling using Multi-Objective Genetic
Algorithms. Expert Systems with Applications 38, 2011, pp. 8293-8302.
[18] Chung, SH., Chan, F., Chan, H., A Modified Genetic Algorithm Approach for Scheduling of
Perfect Maintenance in Distributed Production Scheduling. Engineering Applications of Artificial
Intelligence 22,2009, pp. 1005-1014.
[19] Averbakh, I., On-Line Integrated Production Distribution Scheduling Problems with Capacitated
Deliveries. European Journal of Operational Research 200, 2010, pp. 377-384.
[20] X. S. Yang and S. Deb, Cuckoo search via Lévy Flights, In: World Congress on Nature &
Biologically Inspired Computing (NaBIC2009). IEEE Publications, pp. 210–214, 2009.
[21] R. Rajabioun, Cuckoo Optimization Algorithm, In: Applied Soft Computing journal, vol. 11,
pp.5508 - 5518, 2011.
[22] Brank, J., Sceckenbach, B., Staein, M., Deb, K., Scmeck, H., Portfolio Optimization with an
Envelope-Based Multi-Objective Evolutionary Algorithm, Eouropean Journal of Operational
Reaserch, Vol. 199, 2009, pp. 684-693.

THE STUDY OF CUCKOO OPTIMIZATION ALGORITHM FOR PRODUCTION PLANNING PROBLEM

More Related Content

Similar to THE STUDY OF CUCKOO OPTIMIZATION ALGORITHM FOR PRODUCTION PLANNING PROBLEM (20)

More from ijcax (20)

Recently uploaded (20)

THE STUDY OF CUCKOO OPTIMIZATION ALGORITHM FOR PRODUCTION PLANNING PROBLEM