Biomedical Engineering Reference
In-Depth Information
.
5.. Calculate.
I
[ ]
distance
. for. each. of. the. remaining. nondominated. solu-
tions.
i
= 2,3,⋯,(
l
−1).by
I i
[
+
1]
−
I i
[
−
1]
m
m
I i
[ ]
=
I i
[ ]
+
.
(2.8)
distance
distance
f
(max)
f
(min)
−
m
m
.
.
where
l
.is.the.total.number.of.solutions.in.the.nondominated.set.
I
,.
I
[
i
]
m
.is.the.
m
th.objective.function.value.of.the.
i
th.solution.in
I
,.and.
f
(min)
.and.
(max)
f
m
.are.the.minimum.and.maximum.values.of.the.
m
th.
objective.function,.respectively.
.
6.. Update.
m
.by.setting.
m
=
m
+ 1..If.
m
≤
M
.where.
M
.is.the.total.number.
of.objective.functions,.go.to.step.3.
A.solution.with.a.smaller.crowding.distance.means.that.it.is.more.crowded..
In.other.words,.this.solution.and.its.other.surrounding.solutions.are.close.
in.distance.
2.2.3.2.2 Crowded-ComparisonOperator
The. crowded-comparison. operator,. implemented. in. the. selection. process,.
aims.to.obtain.more.uniformly.distributed.nondominated.solutions..If.any.two.
solutions. picked. have. different. ranks,. the. solution. with. the. lower. rank. will.
be.chosen..However,.if.their.ranks.are.the.same,.the.solution.with.the.larger.
crowding.distance.(i.e.,.located.in.a.less-crowded.region).is.preferred.
2.2.3.2 Elitism Strategy
The.procedure.for.executing.elitism.is.outlined.as.follows:
.
1.. Assuming.that.the.
t
th.generation.is.considered,.combine.the.original.
population.
P
t
.and.the.offspring.population.
Q
t
.to.form.a.population.
R
t
.
.
2.. Sort.
R
t
.by.the.fast.nondominated.sorting.approach.and.then.identify.
different.nondominated.fronts.
F
,
F
.
,
1
2
.
3.. Include.these.fronts.in.the.new.population.
P
t
1
.one.by.one.until.
P
t
1
.
+
+
with.size.
N
.is.full.
If.there.is.insufficient.room.to.include.all.members.of.a.particular.front,.say.
F
i
.into.
P
t
1
,.the.irst
j
.members.of.
F
i
.with.larger.crowding.distances.will.be.
selected.to.fill.
+
+
2.2.4 Strength Pareto evolutionary algorithm 2
The. strength. Pareto. evolutionary. algorithm. 2. (SPEA2). [69]. is. an. improved.
version.of.SPEA.[70].that.includes.three.distinct.features:.strength.value.and.
Search WWH ::
Custom Search