Main Content

This example shows how to estimate the frequency response of a portion
of a Simulink^{®} model using the **Model Linearizer**. To estimate the frequency
response, you specify the portion of the model you want to estimate, the operating point for
estimation, and the input signal to use for estimation.

Open the Simulink model.

```
sys = 'scdDCMotor';
open_system(sys)
```

To open the **Model Linearizer**, in the Simulink model window, in the **Apps** gallery, click
**Model Linearizer**.

By default, **Model Linearizer** uses the linearization analysis points defined
in the model (the *model I/Os*) to determine where to inject the test
signal and where to measure the frequency response. The model
`scdDCMotor`

contains predefined linear analysis points: an input point
at the compensator output, and an open-loop output after the unit gain block. For this
example, use these predefined model I/Os to obtain the frequency response of the inner
loop of the model with the outer loop open.

If you want to obtain the frequency response of a different portion of the model, on
the **Estimation** tab of **Model Linearizer**, use the
**Analysis I/Os** drop-down list. Analysis points for estimation work
the same way as analysis points for linearization. For more information about linear
analysis points, see Specify Portion of Model to Linearize.

You perform frequency response estimation at a steady-state operating point of the
model. You can compute or specify an operating point in **Model Linearizer** using
the **Operating Point** drop-down list. By default, **Model
Linearizer** uses the operating point defined by the model initial conditions. For
this example, use that operating point. For more information about operating points, see
About Operating Points.

Frequency response estimation injects an input signal into the input analysis point you specify for estimation. For this example, configure a sinestream signal, which is a series of sinusoidal perturbations at frequencies you specify. (For more information about input signals, see Estimation Input Signals.)

On the

**Estimation**tab, in the**Input Signal**drop-down list, select`Sinestream`

. The Create sinestream input dialog box opens.To specify frequency points for the input signal, click . In the Add frequencies dialog box, specify the frequency range and number of points for the input signal. The frequency points you specify are the frequencies at which

**Model Linearizer**computes the estimated response.For this example, specify a range from 0.1 to 100 rad/s. Also, specify 100 logarithmically spaced frequencies.

Click

**OK**. The added points are visible in the frequency content viewer of the Create sinestream input dialog box.Select all these frequency points for the estimation input signal.

Specify the amplitude of the input signal. Enter

`1`

in the**Amplitude**field. When you specify a scalar value,**Model Linearizer**uses the same amplitude for all frequencies.Click

**OK**to create the sinestream input signal. The new input signal,`in_sine1`

, appears in the**Linear Analysis Workspace**.

You can now estimate the frequency response and generate a frequency-domain plot of
the result. To do so, click
**Bode**. The estimated frequency response, appears in the
**Linear Analysis Workspace** as the `frd`

model
`estsys1`

.

To export the estimated frequency response model to the MATLAB^{®} workspace for further analysis, drag it from the **Linear Analysis
Workspace** section of the Data Browser to the **MATLAB
Workspace** section.

The simulation results viewer in **Model Linearizer** lets you examine further
details of the frequency response estimation. For more information, see Analyze Estimated Frequency Response.