This is machine translation

Translated by Microsoft
Mouseover text to see original. Click the button below to return to the English version of the page.

Note: This page has been translated by MathWorks. Click here to see
To view all translated materials including this page, select Country from the country navigator on the bottom of this page.

Identify Time-Series Models at the Command Line

This example shows how to simulate a time-series model, compare the spectral estimates, estimate covariance, and predict output of the model.

Generate time-series data.

ts0 = idpoly([1 -1.5 0.7],[]);
e = idinput(200,'rgs');
y = sim(ts0,e); 
y = iddata(y);

y is an iddata object with sample time 1.

Estimate and plot the periodogram and power spectrum.

per = etfe(y);
speh = spa(y);

Estimate a second-order AR model and compare the spectrum plots.

ts2 = ar(y,2);

Define the true covariance function.

ir = sim(ts0,[1;zeros(24,1)]);
Ry0 = conv(ir,ir(25:-1:1));
ir2 = sim(ts2,[1;zeros(24,1)]);
Ry2 = conv(ir2,ir2(25:-1:1));

Estimate the covariance.

z = [y.y;zeros(25,1)];
j = 1:200;
Ryh = zeros(25,1);
for k=1:25, 
a = z(j,:)'*z(j+k-1,:);  
Ryh(k) = Ryh(k)+conj(a(:));
Ryh = Ryh/200; % biased estimate
Ryh = [Ryh(end:-1:2);Ryh];

Alternatively, you can use the Signal Processing Toolbox™ command xcorr.

Ryh = xcorr(y.y,24,'biased');

Plot and compare the covariance.


Predict model output.


See Also

| |

Related Examples

More About