User`s guide

Identifying Input-Output Polynomial Models

You can also model the ARARMAX structure—called the extended matrix

model—by setting

nf=0.

The property-value pairs specify any model properties that conﬁgure

the estimation al go r ith m and the initial conditions. You ca n enter all

property-value pairs i n

pem as a comma-separated list without worrying

about the hierarchy of these properties in the

idpoly model object. For

more information about acces sing and setting model p roperties, see “Model

Properties” on page 2-14.

For multiple inputs,

nb, nf,andnk arerowvectorsofthesamelengthsasthe

number o f input channels:

nb = [nb1 ... nbnu];

nf = [nf1 ... nfnu];

nk = [nk1 ... nknu];

For ARMAX, Box-Jenkins, and O u tput-Error models—w hich can only be

estimated using the iterative pr ediction-error method—use the

armax, bj,and

oe estimation commands, respectively. These commands are versions of pem

with simpliﬁed syntax for these speciﬁc model structures, as follows:

m = armax(Data,[na nb nc nk])

m = oe(Data,[nb nf nk])

m = bj(Data,[nb nc nd nf nk])

Tip If your data is sampled fast, it might help to apply a lowpa ss ﬁlter to the

data before estimating the model. F or example, to mo de l only d ata in the

frequency range 0-10 rad/s, use the

Focus property, as follows :

m = oe(Data,[nb nf nk],'Focus',[0 10])

For more information about validating you model, s ee “Overview of Model

Validation and Plots” on page 8-2.

You can use

pem to reﬁne parameter estimates of an existing po ly nomial

model, as described in “Reﬁning Linear Parametric Models” on p age 3 -103.

3-63