Information Technology Reference
In-Depth Information
1
Start
2
t:=0;
3
Initialize parental population
P
(
t
)
4
Repeat
5
For
k=1
To
λ
Do
6
z = 0;
7
Repeat
8
Choose
ρ
parents from
P
(
t
)
9
recombination step sizes;
10
recombination objective variables;
11
mutation step sizes;
12
If
(z mod
mod
)==0
Then
13
=
·
melt
;
14
For
j=1
To
N
Do
If
σ
j
<
Then
15
σ
j
=
16
17
Next
18
mutation objective variables;
19
fitness of
a
k
:=
f
(
x
);
20
z:=z+1;
21
Until
a
k
feasible
22
Add
a
k
to offspring population
O
(
t
);
23
Next
24
Select parental population
P
(
t
+1) from
O
(
t
);
25
t
:=
t
+1;
26
Until
termination condition
27
End
Fig. 7.4.
Pseudocode of the DSES
on the number
z
of rejected infeasible solutions: Every
infeasible trials,
is
reduced by a factor 0
<ϑ<
1 according to the equation:
:=
·
ϑ
(7.23)
The DSES is denoted by [
;
ϑ
]-DSES. Figure 7.4 shows the pseudocode of the
DSES.
7.4.2
Experimental Analysis
We summarize the results of our experimental analysis on the various con-
strained problems [72], [75]. Note that we translate equality constraints
h
j
(
x
)=
0
,
j
=1
,...,n
2
into inequality constraints
g
i
(
x
)
≤
,
i
=
n
1
+1
,...,n
2
and
g
i
(
x
)
, i
=
n
2
+1
,...,
2
n
2
and
=0
.
0001. For all ES we used
n
σ
=
N
step sizes (
N
is the dimension of the problem) and the ES mutation parame-
ter recommendations for
τ
0
and
τ
1
, see the experimental settings. We chose the
≤−
Search WWH ::
Custom Search