AI agent skill · by @matlab

Simulating Simulink Models

Configures Simulink simulations non-destructively using SimulationInput objects — parameter overrides without modifying the model, batch sweeps via parsim, custom input signals via Dataset, and simulation data retrieval via logsout. Use when running sim()/parsim() with setVariable, setBlockParameter, setExternalInput, or when performing parameter sweeps and multi-run analysis. Not needed for one-shot simulations without configuration.

Archived 2026-10-23 · Source updated 2026-10-07. Check the original repository for the latest version.

SKILL.md instructions

# Simulating Simulink Models with the sim Command

Use this skill when you need to **configure** a simulation non-destructively — parameter overrides, custom inputs, batch execution, or structured output access. For persistent, reusable pass/fail behavioral testing (especially of individual subsystems), use `testing-simulink-models` instead. For trivial one-shot simulations without configuration, a direct `sim()` call suffices without this skill.

## When to Use

- Overriding model or block parameters non-destructively (setVariable, setBlockParameter, setModelParameter) — without modifying the .slx file
- Passing custom input signals to root-level Inport blocks via setExternalInput with a Dataset
- Running parameter sweeps or batch simulations (SimulationInput arrays, parsim, Fast Restart)
- Accessing logged signal data (logsout) for analysis after simulation

## When NOT to Use

- Trivial one-shot simulations without parameter overrides or custom inputs — a direct `sim('ModelName')` call works without this skill
- Writing declarative Gherkin-based tests → use `testing-simulink-models`
- Testing an individual subsystem or component → use `testing-simulink-models` (requires Simulink Test; auto-creates a harness, compiles only the subsystem — much faster than `sim()` which always compiles the entire model)
- Adding, connecting, or deleting blocks → use `building-simulink-models`
- Checking model structure for unconnected ports → use `model_check` tool directly
- Generating requirements from model behavior → use `generate-requirement-drafts`

## Minimal working pattern

Always simulate using `Simulink.SimulationInput` and `Simulink.SimulationOutput`:

```matlab
in = Simulink.SimulationInput('MyModel');
in = in.setModelParameter('StopTime', '10');
out = sim(in);
```

## Setting parameters

Use `SimulationInput` methods to configure the simulation:

```matlab
% Model-level parameters (StopTime, SolverType, SimulationMode, etc.)
in = in.setModelParameter('StopTime', '10', 'SolverType', 'Fixed-step');

% Block parameters — resolve path from blk_X ID (never type block names manually)
blkPath = Simulink.ID.getFullName('MyModel:5');
in = in.setBlockParameter(blkPath, 'Gain', '5');

% MATLAB workspace variables used by the model
in = in.setVariable('Kp', 1.2);
```

## Input signals

Pass input signals through Inport blocks using a `Simulink.SimulationData.Dataset`. Elements are matched to Inport blocks **by index position** — the first element maps to the Inport with port number 1, the second to port number 2, and so on.

```matlab
dt = 0.01;
N = 1000;
t = dt*(0:N)';
u = sin(2*pi*t);

ts = timeseries(u, t);

ds = Simulink.SimulationData.Dataset;
ds{1} = ts;

in = in.setExternalInput(ds);
out = sim(in);
```

You can also use `timetable` as an input format:

```matlab
secs = seconds(t);
tt = timetable(secs, u);

ds = Simulink.SimulationData.Dataset;
ds{1} = tt;

in = in.setExternalInput(ds);
```

## Discovering logged data

First, discover what kinds of logged data the model produces using `who`, then inspect signal names within `logsout`:

```matlab
in = Simulink.SimulationInput('MyModel');
out = sim(in);

% See what logging properties exist (logsout, yout, tout, etc.)
who(out)

% List individual signal names within logsout
disp(out.logsout.getElementNames);
```

## Accessing logged data

Logged signals are available through `out.logsout`. Access them directly by name:

```matlab
% Plot a logged signal
plot(out.logsout.get('signalName').Values)

% Get time and data separately
sig = out.logsout.get('signalName').Values;
plot(sig.Time, sig.Data)
```

## Multiple simulations

When running many simulations, create an array of `Simulink.SimulationInput` objects:

```matlab
in = repmat(Simulink.SimulationInput('MyModel'),N,1);
for k = 1:N
    in(k) = Simulink.SimulationInput('MyModel');
    in(k) = in(k).setVariable('gain', gains(k));
end
out = sim(in);
```

To enable fast restart for iterative sweeps (compiles the model only once):

```matlab
out = sim(in, 'UseFastRestart', 'on');
```

## Parallel simulation (parsim)

To run multiple simulations in parallel, use `parsim` instead of looping over `sim`:

```matlab
for k = 1:N
    in(k) = Simulink.SimulationInput('MyModel');
    in(k) = in(k).setVariable('gain', gains(k));
end
out = parsim(in);
```

`parsim` also supports `'UseFastRestart','on'` for faster batch runs.

## Guardrails

- **Never** use `set_param`, `load_system`, or `open_system` to drive simulation — `SimulationInput` replaces all of these.
- **Never** wrap `SimulationOutput` access in `try-catch` or `isfield` — `sim` either returns a valid object or throws. `SimulationOutput` has no `isfield` method.
- **Never** create unnecessary intermediate variables for logged data — access directly via `out.logsout.get('name').Values`.
- **Always** use `in`/`out` as variable names for `SimulationInput`/`SimulationOutput`.
- **Always** use `setExternalInput` with a `Dataset` — don't pass comma-separated lists of variables.

----

Copyright 2026 The MathWorks, Inc.

----