 |
|
A genetic algorithm in the Generational form has in the end a structure as follow:
do ng generation
do nind individuals
translate bits into variables
compute objectives => interface to analysis
end do
do some statistics on the population individuals
do Create a new population:
by cross over:
select individual
and reproduce
by mutation:
select individuals
and mutate
end do
end do
while a Steady-state option would look like:
do ng*ind generation
select a random location
create a new individual by:
cross over:
select individual
and reproduce
or by:
mutation:
select individuals
and mutate
translate bits into variables
compute objective <=> interface to analysis
end do
The key points of the GA are therefore the operators used for selection and reproduction that highly influence the robustness and the efficiency of the algorithm.