parsim

Simulate dynamic system multiple times in parallel or serial

collapse all in page

Syntax

simOut = parsim(in)
simOut = parsim(in,Name=Value)
futureOut = parsim(designStudy)

Description

simOut = parsim(in) simulates a model using the inputs specified in the SimulationInput object, in. The parsim function uses an array of SimulationInput objects to run multiple simulations.

simOut = parsim(in,Name=Value) simulates a model in parallel using the inputs specified in the SimulationInput object and the options specified using one or more name-value arguments.

futureOut = parsim(designStudy) simulates massive simulations with the simulink.multisim.DesignStudy object.

The parsim function requires the Parallel Computing Toolbox™ license to run the simulations in parallel. The parsim function runs the simulations in serial if a parallel pool cannot be created or if Parallel Computing Toolbox is not used.

example

Examples

collapse all

Simulate the model vdp in rapid accelerator mode.

Load the model.

openExample('simulink_general/VanDerPolOscillatorExample')
mdl = "vdp";
load_system(mdl)

This step builds the Rapid Accelerator target:

Simulink.BlockDiagram.buildRapidAcceleratorTarget(mdl);

Create a SimulationInput object and use the setModelParameter function to set RapidAcceleratorUpToDateCheck to "off".

in = Simulink.SimulationInput(mdl);
in = setModelParameter(in,SimulationMode="rapid-accelerator");
in = setModelParameter(in,RapidAcceleratorUpToDateCheck="off");

Simulate the model.

out = parsim(in);
Open Script

Simulate the model, CSTR, in parallel by sweeping over a variable. An array of SimulationInput objects is used to perform the sweep.

Specify sweep values.

FeedTempSweep = 250:10:300;

Create an array of SimulationInput objects.

for i = length(FeedTempSweep):-1:1
in(i) = Simulink.SimulationInput('CSTR');
in(i) = in(i).setVariable('FeedTemp0',FeedTempSweep(i));
end

Simulate the model in parallel.

out = parsim(in, 'ShowSimulationManager', 'on')

[28-Aug-2024 10:44:10] Checking for availability of parallel pool...
Starting parallel pool (parpool) using the 'Processes' profile ...
28-Aug-2024 10:45:15: Job Queued. Waiting for parallel pool job with ID 2 to start ...
Connected to parallel pool with 8 workers.
[28-Aug-2024 10:45:41] Starting Simulink on parallel workers...
[28-Aug-2024 10:46:32] Configuring simulation cache folder on parallel workers...
[28-Aug-2024 10:46:33] Loading model on parallel workers...
[28-Aug-2024 10:46:49] Running simulations...
[28-Aug-2024 10:47:40] Completed 1 of 6 simulation runs
[28-Aug-2024 10:47:41] Received simulation output (size: 20.18 KB) for run 1 from parallel worker.
[28-Aug-2024 10:47:41] Completed 2 of 6 simulation runs
[28-Aug-2024 10:47:41] Received simulation output (size: 20.22 KB) for run 2 from parallel worker.
[28-Aug-2024 10:47:41] Completed 3 of 6 simulation runs
[28-Aug-2024 10:47:41] Received simulation output (size: 20.61 KB) for run 3 from parallel worker.
[28-Aug-2024 10:47:44] Completed 4 of 6 simulation runs
[28-Aug-2024 10:47:44] Received simulation output (size: 35.84 KB) for run 4 from parallel worker.
[28-Aug-2024 10:47:45] Completed 5 of 6 simulation runs
[28-Aug-2024 10:47:45] Received simulation output (size: 37.76 KB) for run 6 from parallel worker.
[28-Aug-2024 10:47:45] Completed 6 of 6 simulation runs
[28-Aug-2024 10:47:45] Received simulation output (size: 39.48 KB) for run 5 from parallel worker.
[28-Aug-2024 10:47:45] Cleaning up parallel workers...

out = 

1x6 Simulink.SimulationOutput array

Input Arguments

collapse all

Inputs and configurations for parallel simulations, specified as an array of Simulink.SimulationInput objects.

Inputs and configurations for massive parallel simulations, specified as a simulink.multisim.DesignStudy object.

Name-Value Arguments

collapse all

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Before R2021a, use commas to separate each name and value, and enclose Name in quotes.

Example: ShowProgress="on"

Additional files to attach to the parallel pool, specified as a cell array of character vectors. The parsim function does not support using From File blocks to load data from MAT files in subdirectories.

Option to show progress of simulations, specified as "on" or "off". Set to "on", to see the progress of the simulations in the MATLAB® Command Window. The progress is hidden when set to "off".

Note

When the progress is shown, a message 'Cleaning up parallel workers..' might be displayed before the completion of the last few simulations. This message does not depend on the completion of the simulations. Simulations complete when the outputs are fetched from the future. For more information, see Simulink.Simulation.Future.

Option to run simulations in background, specified as "off" or "on". Set to "on" to run simulations asynchronously, keeping the MATLAB command prompt available for use. To terminate the simulations running in the background, use the cancel function of the Simulink.Simulation.Future objects.

When you specify the 'RunInBackground' argument as "on", the parsim function returns Simulink.Simulation.Future objects. Use the Simulink.Simulation.Future object to access the SimulationOutput object.

Function to run once per worker before starting simulations, specified as a function handle.

Example: SetupFcn=@()openProject("Modelex/Model_example.prj")

Note

When buildRapidAcceleratorTarget is used in the SetupFcn and the model has external inputs specified,​ either set LoadExternalInput to "off" or ensure that the specified external input is available on the workers to prevent compilation error.

Function to run once per worker after running simulations, specified as a function handle.

Option to manage model dependencies, specified as "on" or "off". When ManageDependencies is set to "on", model dependencies are automatically sent to the parallel workers if required. If ManageDependencies is set to "off", explicitly attach model dependencies to the parallel pool.

Option to log diary information for parallel simulations and PreSimFcn and PostSimFcn, specified as 'on' or 'off'. You can use the setPreSimFcn to print diagnostics to the diary to help debug parsim and batchsim setup issues.

Option to use fast restart, specified as "off" or "on". When UseFastRestart is set to "on", simulations run on the workers using fast restart. For more information, see Script Iterative or Batch Simulations Using Fast Restart.

When performing a parameter sweep that varies properties of a Simscape™ component, such as the mass and inertia of a Solid block in Simscape Multibody™, specify the Simscape block parameter as run-time configurable. For more information, see About Simscape Run-Time Parameters (Simscape).

Note

The parsim function does not support fast restart for models that contain To File blocks.

Option to transfer variables to parallel workers, specified as "off" or "on". When TransferBaseWorkspaceVariables is set to "on", variables used in the model and everything defined in the base workspace are transferred to the parallel workers.

The TransferBaseWorkspaceVariables option is not recommended for large scale simulations. Transferring unnecessary base workspace data leads to wasted network and memory usage. Instead, consider using project startup scripts, model callback functions, or the SetupFcn name-value argument to define the base workspace data.

Option to open Simulation Manager, specified as "off" or "on". When ShowSimulationManager is set to "on", you can use the Simulation Manager to monitor simulations.

Option to skip remaining simulations if simulation error occurs, specified as "off" or "on". Setting StopOnError to "on" stops the execution of simulations if a simulation error occurs.

A location path for saving simulation outputs (including logging and input metadata) while running parsim the simulink.multisim.DesignStudy object. The OutputLocation is specified as one of these:

  • 'cluster-storage' – By default, OutputLocation uses a backing job or parallel pool storage to writer simulation results. The lifetime of the storage is tied to the lifetime of the underlying parallel job. For interactive pools, the job is deleted on pool shutdown.

  • 'memory' – In-memory storage where output data is stored in-memory and returned directly to the MATLAB client.

  • 'locationPath' – Absolute path, specified as a string or a character vector.

Output Arguments

collapse all

Simulation results and metadata, returned as an array of Simulink.SimulationOutput objects, or a vector. The Simulink.SimulationOutput object contains all data logged from simulation as well as metadata about the simulation, including timing information and diagnostics.

Extended Capabilities

expand all

Version History

Introduced in R2017a

See Also

Objects

Tools

Functions

Topics