<s>
Finite	B-Application
element	I-Application
model	O
updating	O
is	O
the	O
process	O
of	O
ensuring	O
that	O
finite	B-Application
element	I-Application
analysis	I-Application
results	O
in	O
models	O
that	O
better	O
reflect	O
the	O
measured	O
data	O
than	O
the	O
initial	O
models	O
.	O
</s>
<s>
It	O
is	O
part	O
of	O
verification	B-Application
and	I-Application
validation	I-Application
of	O
numerical	O
models	O
.	O
</s>
<s>
The	O
process	O
is	O
conducted	O
by	O
first	O
choosing	O
the	O
domain	B-Algorithm
in	O
which	O
data	O
is	O
presented	O
.	O
</s>
<s>
The	O
domains	O
used	O
include	O
time	O
domain	B-Algorithm
,	O
frequency	O
domain	B-Algorithm
,	O
modal	O
domain	B-Algorithm
,	O
and	O
time-frequency	O
domain	B-Algorithm
.	O
</s>
<s>
The	O
third	O
task	O
is	O
to	O
formulate	O
a	O
function	O
which	O
has	O
the	O
parameters	O
that	O
are	O
expected	O
to	O
be	O
design	O
variables	O
,	O
and	O
which	O
represents	O
the	O
distance	O
between	O
the	O
measured	O
data	O
and	O
the	O
finite	B-Application
element	I-Application
model	O
predicted	O
data	O
.	O
</s>
<s>
For	O
nonlinear	O
analysis	O
,	O
more	O
specific	O
methods	O
like	O
response	O
surface	O
modeling	O
,	O
particle	B-Algorithm
swarm	I-Algorithm
optimization	I-Algorithm
,	O
Monte	B-Algorithm
Carlo	I-Algorithm
optimization	I-Algorithm
,	O
and	O
genetic	B-Algorithm
algorithms	I-Algorithm
can	O
be	O
used	O
.	O
</s>
<s>
Recently	O
,	O
finite	B-Application
element	I-Application
model	O
updating	O
has	O
been	O
conducted	O
using	O
Bayesian	O
statistics	O
which	O
gives	O
a	O
probabilistic	O
interpretation	O
of	O
model	O
updating	O
.	O
</s>
