System Identification

AR Models

For a single output signal the counterpart of the ARX model is the AR model:

(3-40)

The arx command also covers this special case

m = arx(y,na)

but for scalar signals more options are offered by the command

m = ar(y,na)

which has an option that allows you to choose the algorithm from a group of several popular techniques for computing the least squares AR model. Among these are Burg's method, a geometric lattice method, the Yule-Walker approach, and a modified covariance method. (See the Command Reference chapter for details.) The counterpart of the iv4 command is

m = ivar(y,na)

which uses an instrumental variable technique to compute the AR part of a time series.

ARX Models

General Polynomial Black-Box Models