Geology Reference
In-Depth Information
EVOLUTIONARY OPTIMIZATION OF
PASSIVE COMPENSATOR SYSTEMS
The function
z
should be defined uniquely
and accepted by all participants of the optimiza-
tion process. The list of free parameters has to be
checked for its completeness and the ranges the
parameters may assume. Constraints may be given
by relations between parameters or physical data of
the problem like maximum stress, no penetration
of neighbouring parts or other incompatibilities.
The main advantage of passive compensation
systems is not to require an active control system
to reduce the load or the damage caused by the
earthquake. This makes them a favourite in cases
where the electric energy supply may fail and de-
struction of even small parts of the edifice would
be equivalent to a total economic loss.
To introduce the way of designing passive
compensation systems, we have first sketched the
optimization strategy to be applied. A discussion
of the basics of compensators follows. Then a
simplified model of high and slender structures
helps to optimize compensators at high towers.
Finally some of the most important results are
listed.
Evolutionary Optimization,
Terms, and Definitions
Following the theories Darwin proposed 150 years
ago, biological species develop as new members
of the population are produced by crossing their
parents' properties with an additional mutation,
a small change of these crossed properties. The
new members or individuals who better fit into
the requirements of their environment will have
a better chance to reproduce. Therefore their
properties will become dominant, while those of
individuals not fitting well will die out or play a
minor role.
To come up with a better understanding of the
basic ideas of evolutionary optimization, some
definitions may be helpful. It should be noted that
not all different schools applying evolutionary
optimization strategies use identical terms; some-
times they are even contradictory (Rechenberg
1994, pp. 15-44, Michalewicz 1996, pp. 21-32,
Gen and Cheng R. 2000, pp. 17-34). Nevertheless
the proposals made here are generally accepted
and helpful in many applications.
The central terms in evolutionary optimization
used frequently in literature are the following:
Evolutionary Optimization Methods
Optimization is one of the most often used and
least understood terms in many parts of our lives,
not only in engineering. To avoid any confusion
about its use, we have introduced a commonly
accepted definition (Steinbuch 2004, pp.188-207)
which states that:
Optimization is the process of finding the
maximum (or minimum) of a given function
z
=
(
z p p p
,
,
, ......
p
n
)
(1)
1
2
3
by varying the free parameters
p
1
, p
2
..., p
n
without
violation of given boundary conditions
p
≤ ≤
p
p
1
≤ ≤
i
n
(2)
•
Objective, Goal, Fitness:
Function mea-
suring the quality of an individual of the
population. It is defined by its parameter
values. The terms are used interchangeably
by different authors.
i
i,min
i
i
,max
and constraints
f p p
(
.
, ...
p
)
r
0
c
≤
≤
(3)
c
c
1
c
2
cm
c
related to the problem.
Search WWH ::
Custom Search