User`s guide
Identifying Low-Order Transfer Functions (Process M odels)
• Stabilit y — Behaves the same way as the Prediction option, but also
forcesthemodeltobestable. Formoreinformation about model stability,
see “Unstable Models” on page 8-69.
•
Filter —Specifyacustomfilter to open the Estimation Focus dialog box,
where you can enter a filter, as described in “Simple Passband Filter” on
page 1-111 or “Defining a Custom Filter” on page 1-112. This prefiltering
applies only for estimating the d ynamics from input to output. The
disturbance model is determ ined from the estimation data.
At the comm a nd line. Specify the focus as an argument in the estimation
command
pem using the same options as in the GUI. For example, use this
command to optimize the fit for simulation and estimate a disturbance model:
pem(data,'P1D','dist','arma2','Focus' ,'Simulation')
Options for Initial States
Because the process models are dynamic, you need initial states that
capture past input properties. Thus, you must specify how the iterative
algorithm treats initial states. This information supports the estimation
procedures “How to Estimate Process Models Using the GUI” on page 3-23
and “Estimating Process Models at the Command Line” on page 3-29.
In the System Identification Tool GUI. Set Initial state to one o f the
following o ptions:
•
Zero — S ets all initia l states to zero.
•
Estimate — Treats the initial states as an unknown vector of param eters
and estimates these states from the da ta.
•
Backcast — Estim ates initial states u sing a ba c kward fi ltering method
(least-squares fit).
•
U-level est — Estimates both the initial states and the InputLevel
model property that represents the input offset level. F or multiple inputs,
the input level for each input is estimated individually. Use if you included
an integrator in the transfer function.
•
Auto — A uto m atically choose s one of the preceding options based on the
estimation data. If initial states hav e negligible effect on the prediction
errors, the initial states are set to zero to optimize algorithm performance.
3-39