Cuttlefish optimization algorithm .28Eesa.2C Mohsin.2C Brifcani .26 Orman 2013.29 List of metaphor-based metaheuristics

the 6 cases of reflection used cuttlefish

the cuttlefish optimization algorithm population-based search algorithm inspired skin color changing behaviour of cuttlefish developed in 2013 has 2 global search , 2 local search.

the algorithm considers 2 main processes: reflection , visibility. reflection process simulates light reflection mechanism, while visibility simulates visibility of matching patterns. these 2 processes used search strategy find global optimal solution. formulation of finding new solution (newp) using reflection , visibility follows:



n
e
w
p
=
r
e
f
l
e
c
t
i
o
n
+
v
i
s
i
b
i
l
i
t
y


{\displaystyle newp=reflection+visibility}

cfa divide population 4 groups (g1, g2, g3 , g4). g1 algorithm applying case 1 , 2 (the interaction between chromatophores , iridophores) produce new solutions. these 2 cases used global search. g2, algorithm uses case 3 (iridophores reflection opaerator) , case 4 (the interaction between iridophores , chromatophores) produces new solutions) local search. while g3 interaction between leucophores , chromatophores (case 5) used produce solutions around best solution (local search). g4, case 6 (reflection operator of leucophores) used global search reflecting incoming light out modification. main step of cfa described follows:

equations used calculate reflection , visibility 4 groups described below:

case 1 , 2 g1:

r
e
f
l
e
c
t
i
o
n
[
j
]
=
r
∗

g

1


[
j
]
.
p
o
i
n
t
s
[
j
]


{\displaystyle reflection[j]=r*g_{1}[j].points[j]}

v
i
s
i
b
i
l
i
t
y
[
j
]
=
v
∗
(
b
e
s
t
.
p
o
i
n
t
s
[
j
]
−

g

1


[
i
]
.
p
o
i
n
t
s
[
j
]
)


{\displaystyle visibility[j]=v*(best.points[j]-g_{1}[i].points[j])}

case 3 , 4 g2:

r
e
f
l
e
c
t
i
o
n
[
j
]
=
r
∗
b
e
s
t
.
p
o
i
n
t
s
[
j
]


{\displaystyle reflection[j]=r*best.points[j]}

v
i
s
i
b
i
l
i
t
y
[
j
]
=
v
∗
(
b
e
s
t
.
p
o
i
n
t
s
[
j
]
−

g

2


[
i
]
.
p
o
i
n
t
s
[
j
]
)


{\displaystyle visibility[j]=v*(best.points[j]-g_{2}[i].points[j])}

case 5 g3:

r
e
f
l
e
c
t
i
o
n
[
j
]
=
r
∗
b
e
s
t
.
p
o
i
n
t
s
[
j
]


{\displaystyle reflection[j]=r*best.points[j]}

v
i
s
i
b
i
l
i
t
y
[
j
]
=
v
∗
(
b
e
s
t
.
p
o
i
n
t
s
[
j
]
−
a

v

b
e
s
t




{\displaystyle visibility[j]=v*(best.points[j]-av_{best}}

case 6 g4:

p
[
i
]
.
p
o
i
n
t
s
[
j
]
=
r
a
n
d
o
m
∗
(
u
p
p
e
r
l
i
m
i
t
−
l
o
w
e
r
l
i
n
i
t
)
+
l
o
w
e
r
l
i
m
i
t
,
i
=
1
,
2
,
.
.
.
,
n
;
j
=
1
,
2
,
.
.
.
,
d


{\displaystyle p[i].points[j]=random*(upperlimit-lowerlinit)+lowerlimit,i=1,2,...,n;j=1,2,...,d}

where




g

1




{\displaystyle g_{1}}

,




g

2




{\displaystyle g_{2}}

group1 , group2, presents




i

t
h




{\displaystyle i^{th}}

element in g, j




j

t
h




{\displaystyle j^{th}}

point of




i

t
h




{\displaystyle i^{th}}

element in group g, best best solution ,



a

v

b
e
s
t




{\displaystyle av_{best}}

presents average value of best points. while r , v 2 random numbers produced around 0 such between (-1, 1), r represents degree of reflection, v represents visibility degree of final view of pattern, upperlimit , lowerlimit upper limit , lower limit of problem domain.