Simulink Variants Overview

Variants in Simulink^® enable multiple design alternatives of a system to coexist in a single model, eliminating the need for separate models and facilitating reuse of model elements and artifacts. You can transition between variations by setting a variable in the workspace before simulation, choosing a variant through a user input such as a parameter or dialog box, or allowing the model to automatically select the appropriate design based on the hardware platform you are targeting.

In Simulink, you can implement design variations using variant blocks, variant parameters, or symbolic dimensions. Use variant blocks to implement structural or algorithmic variations within your model, such as different controllers or plant models. Use variant parameters to implement variations in parameter values, for example, different gain values or physical constants. Use symbolic dimensions to implement variability in signal dimensions, such as the size of an input or output port based on the selected alternative.

A meaningful combination of these variations is defined and managed as a variant configuration. Using the Variant Manager for Simulink support package, you can define multiple variant configurations and system-wide constraints. You can then apply these configurations across the model hierarchy to switch between them, reducing the complexity and potential errors that can occur when configuring each variant element individually.

Organizations can use variant configurations in their design of product lines such as automobiles, aircraft, or electronics that satisfy diverse market needs, customer preferences, and geographic requirements on a common platform. For instance, in the automotive sector, variants capture differences in parameters such as fuel consumption, emission standards, and so on, within a single model, thus reducing design redundancy, enabling consistent testing across design variations, and simplifying maintenance by integrating updates throughout the model.

Simulink variants integrate into every stage of the development process, from the design phase to code generation and up to the testing phase of the V-model in Model-Based Design as described in V-Model for System Development with Simulink Variants. During the design phase, variants allow you to incorporate multiple design alternatives into a single model using variant blocks, variant parameters, and symbolic dimensions. While designing, you can use the Variant Manager for Simulink support package to automatically generate variant configurations for all possible valid combinations of variant choices. You can use the support package to prevent invalid variant combinations and to reduce the variant model to contain only the selected variants. As development advances to the testing phase, you can use products such as Simulink Design Verifier™ and Simulink Test™ to test variant configurations against their specific requirements.

Advantages of Using Variants in Model-Based Design

This section lists the advantages of using variants, including the ability to create multiple design variations within a single model, centrally configure variants with Variant Manager, generate both common and variant-specific code for deployment, and efficiently simulate, test, and validate all design alternatives across large and complex systems.

Design Flexibility and Modularity

You can create multiple design variations within a single model to improve flexibility and modularity by using variants. Some key advantages include:

• Represent design alternatives within a single model, enabling you to switch between different configurations without maintaining multiple models.

• Rapidly prototype design possibilities as variants without creating a separate implementation for each design.

• Develop modular designs for reuse and customization, reducing complexity.

• Explore alternatives without altering the fixed components.

This diagram illustrates a system with varying modes Eco mode and Sport mode for driving, enabling quick switching between design alternatives and rapid prototyping within a single model.

Variant Management and Workflow

Using the Variant Manager for Simulink support package provides a centralized interface to manage variant configurations and constraints across your model. When you switch between design alternatives, the associated settings update consistently, reducing manual errors and streamlining testing, debugging, and code generation for complex systems. You can also generate a reduced model that uses only a subset of variant configurations from a model with design alternatives.

In this diagram, the table shows how Variant Manager for Simulink maps variant configurations such as EconomyStandard, PerformancePlus, and EconomyEnhanced, to combinations of drive modes Eco or Sport, and sensors Basic or Advanced. The blocks show varying elements for drive mode and sensor. By using Variant Manager for Simulink you can select the desired alternative for each element within a particular configuration. For example, when you select the EconomyEnhanced variant configuration, the support package selects Eco for the driving mode and Advanced for the sensor.

Variant Code Generation Capabilities

You can generate common and variant-specific code to reuse, share code with third parties, and deploy across different systems using variants. Several key advantages include:

• Reuse code for a group of models that have similar functionality but small differences.

• Share code generated from a model with multiple design variations with third parties, allowing them to select and use the variant that best fits their needs.

• Generate code for one or all variants and easily switch the active variant at compile time, system initialization, or run time, depending on your deployment workflow.

• Automatically generate code specific to each variant, reducing manual effort and ensuring accurate implementation of functionality specific to each variant.

This diagram illustrates a system with varying drive modes, Eco and Sport, and sensors, Basic and Advanced, along with the corresponding generated code. The generated code includes common code for fixed elements and variant-specific code for variant configurations EconomyStandard, PerformancePlus, and EconomyEnhanced. When you compile the code for EconomyEnhanced, the resulting executable includes the common code and only the code specific to the Eco mode and the Advanced sensor. To compile code for a different variant configuration EconomyStandard, you are not required to regenerate the code. You select the variant configuration EconomyStandard prior to compilation, and the code generator produces an executable tailored for EconomyStandard by including the common code and only the code specific to the Eco mode and the Basic sensor.

Testing and large-scale design

You can manage and validate multiple design alternatives by integrating variant testing into your workflow. Several key advantages include:

• Simulate every design possibility for a given test suite.

• Distribute the testing process for large-scale designs across a cluster of multicore computers.

• Map different test suites to design alternatives for efficient management of design-specific tests.

Tip

Limit variants during large-scale design and testing to reduce complexity and validation effort, avoid longer test times, and reduce maintenance overhead.

This diagram illustrates how a test suite connects to multiple design alternatives for efficient variant testing. The test suite includes drive mode selection and actions, mapped to designs EconomyStandard, PerformancePlus, and EconomyEnhanced. This setup enables you to test different designs, distribute tests across multicore clusters, and map test suites to design alternatives efficiently.

Variant Implementation Techniques in Simulink

You can use several variant implementation techniques in Simulink to manage design alternatives within a single model.

Switch Between Model Structures

Use hierarchical variant blocks to encapsulate multiple implementations of a component in a separate hierarchy of the model.

Variant block	Use
Variant Subsystem	The Variant Subsystem block provides a structured hierarchy where each design variation is encapsulated within its own subsystem, enabling you to add or remove variations directly within the block. You can switch between the variations using variant controls. For information on how to use a Variant Subsystem block, see Implement Variations in Separate Hierarchy Using Variant Subsystems. You can use Variant Subsystem blocks to synthesize hardware design alternatives into hardware description language (HDL) code and to generate C/C++ code for software design alternatives. For more information, see Usage of Different Subsystem Types (HDL Coder) and Generate Code for Variant Subsystem Blocks (Simulink Coder).  Create Multiple Implementations of Component in Separate Hierarchy using Variant Subsystem Block This diagram illustrates a vehicle with three possible engine configurations: Engine_2_cyl_petrol, Engine_4_cyl_petrol, and Engine_8_cyl_petrol. You can implement each engine model as a separate subsystem inside the Variant Subsystem block and then switch between the subsystems based on the selected engine configuration during simulation. For example, Engine_2_cyl_petrol is active if the value of the variable engConf is set to EC.cyl2.
Variant Assembly Subsystem	The Variant Assembly Subsystem block is an alternative configuration of the Variant Subsystem block that enables you to manage design variations in a model by using an external source such as an enumeration class or a MATLAB® function without modifying or accessing the model itself. For more information, see Add or Remove Variant Choices of Variant Assembly Subsystem Blocks Using External Files.

Note

If you have a Subsystem block within your model and need to incorporate multiple design alternatives, use the Simulink.VariantUtils.convertToVariantSubsystem function to convert the Subsystem block into a Variant Subsystem block. If you want to manage multiple design alternatives from external sources, use the Simulink.VariantUtils.convertToVariantAssemblySubsystem function to convert it into a Variant Assembly Subsystem block.

Switch Between Signal Paths

Use Inline variant blocks to select or route signals between multiple inputs or outputs, and implement design variations at specific signal paths or blocks without adding subsystem hierarchies. Although an inline variant block does not create a new model hierarchy, its effect can extend into hierarchical elements such as Model blocks that exist in the signal path.

Variant block	Use
Variant Source and Variant Sink	The Variant Source block enables you to connect multiple source signals to its input ports, with each port representing a distinct variation. Similarly, the Variant Sink block enables you to connect multiple destination signals to its output ports, with each port representing a distinct variation. You can switch between the variations using variant controls. For more information, see Visualize Variant Implementations in a Single Layer. You can generate C/C++ code from Variant Source and Variant Sink blocks as described in Generate Code for Variant Source and Variant Sink Blocks (Simulink Coder).  Create Multiple Implementation of Component in Single Layer Using Variant Source and Variant Sink Blocks This diagram illustrates a rain-sensing wiper control system. In this system, the Variant Source block selects either the optical rain sensor or the capacitive rain sensor as input, depending on whether the value of the Sensor variable is set to Optical or Capacitive. The selected rain sensor signal, along with the vehicle speed input, is sent to the Controller subsystem. The Controller subsystem processes these inputs to determine the optimal wiper speed. The calculated wiper speed is then routed through a Variant Sink block to direct the signal to the front wiper motor if the Motor variable is set to Front, or to the rear wiper motor if the Motor variable is set to Rear. As a result, the appropriate wiper operates at the required speed, responding to both current rain conditions and vehicle speed.
Manual Variant Source and Manual Variant Sink	The Manual Variant Source and Manual Variant Sink blocks are similar to the Variant Sink and Variant Source block, but instead of using variant controls to switch between design variations, you toggle between two variations by double-clicking the block. For more information, see Provide Variation in Signal Source and Destination Using Manual Variant Source and Manual Variant Sink Blocks.
Variant Start and Variant End	The Variant Start and Variant End blocks create a bounded region to control activation and deactivation for a specific set of blocks. These blocks collectively confine activation and deactivation to the region and do not impact other parts of the model. If you have elements with varying inputs, outputs, or functionalities that would typically require a Variant Subsystem block, but you want the variant choices to appear in a single level of the model hierarchy, place a Variant Start block at the entry and a Variant End block at the exit of the variant path. For more information, see Control Variant Condition Propagation using Variant Start and Variant End Blocks. You can generate C/C++ code from Variant Start and Variant End blocks as described in Generate Code for Variant Start and Variant End Blocks (Simulink Coder).  Control and Confine Activation of Specific Set of Blocks Without Impacting Other Part of Model This diagram represents a control system that includes first-order and second-order systems. The systems are placed within the bounded region defined by the Variant Start and Variant End blocks. These variant blocks activate the blocks in the first-order system if the systemOrder variable is set to 1 and activates the blocks in the second-order system if the systemOrder variable is set to 2. The blocks located outside this region remain unaffected by the switching and are always active in the model.

Change Parameter Values

You can conditionally vary the values of block parameters in a Simulink model by using a variant parameter. To create a variant parameter object, use the Simulink.VariantVariable class. This object defines a set of values, known as choices, and variant condition expressions associated with the choices. To specify the variant condition corresponding to each choice of the variant parameter, use a variant control variable of type Simulink.VariantControl. Use the Simulink.VariantControl object to specify an activation time for the variant parameter and set a value to evaluate the variant conditions and determine the active choice of the variant parameter. During simulation, the choice associated with the variant condition that evaluates to true becomes the active value of the variant parameter.

For an overview of variant parameters, see Use Variant Parameters to Reuse Block Parameters with Different Values. Variant parameter objects support the generation of C/C++ code as described in Options to Represent Variant Parameters in Generated Code (Simulink Coder).

 Create Varying Value Sets for Automotive Configurations Using Variant Parameters

Consider this model that represents an automobile system with several configurations. These configurations, although similar in several aspects, can differ in values such as fuel consumption rate, engine size, or number of cylinders in the engine. Instead of designing multiple blocks that together represent all possible configurations, you can use variant parameters to model the varying values. In this model, the Fuel consumption, Engine size, and Number of cylinders blocks have variant parameters fc, es, and nc. The values of the variant parameters are based on the value of the variant control variable V. For example, if V == 1 evaluates to true, then fc is set to 30, es is set to 1000, and nc is set to 4.

Switch Based on Events or Triggers

A conditionally executed subsystem is a nonvirtual subsystem whose execution is controlled by an external signal. By specifying variant conditions in conditionally executed subsystems, you can control whether or not to provide the external signal to the subsystem. With event function blocks, you can add custom routines to the default routines and use variant conditions to control whether the system executes these custom routines. Use event-based variants when creating complex models where the execution of certain components depends on the state or activity of other components.

Event function and conditionally executed subsystem	Use
Enabled Subsystem, Triggered Subsystem, Reset, and Function-Call Subsystem	Conditional subsystems such as Enabled, Triggered, Reset, and Function-Call blocks control execution tasks such as state initialization, value resetting, and cleanup in response to external signals. With variant blocks, you can selectively provide these signals to the conditional subsystem blocks. For more information, see Propagate Variant Conditions to Control Execution of Conditional Subsystems.  Merge Distinct Function-Call Signals into Unified Output This diagram illustrates how to merge the distinct function‐call signals from different ports into a single, unified output. The Pulse Generator block drives a Variant Subsystem block that contains two charts, Chart1 and Chart2. Only Chart1 is active. The Chart1 processes the incoming signals and generates a function-call signal that activates the Function-Call Subsystem block. This subsystem receives two inputs, one from the chart and another from a Sine Wave block. The subsystem combines them to produce the output showcasing how the system merges distinct function-call signals from different sources into a single, unified output over time.
Simulink Function	You can control the execution of functions defined in the Simulink Function blocks by conditionally invoking them with a Function Caller block. To determine which function executes, you can either specify a condition for each function for individual control or inherit the condition from the caller block for centralized management through the Function Caller block. For more information, see Conditionally Execute Simulink Functions.  Execute Simulink Functions Conditionally This diagram illustrates the conditional execution of the Simulink Function block. The function-call port block within the Simulink Function block has the Enable variant condition option selected and the Variant control parameter is set to (inherit). During simulation, the Simulink Function block inherits the logical OR of the variant conditions, A == 1 and B == 1, from the corresponding Function Caller blocks Function Caller1 and Function Caller2. The Simulink Function function block executes only when at least one of its caller blocks is active, making its execution conditional.
Initialize Function, Reset Function, and Terminate Function	You can control the execution of the Initialize Function, Reset Function, and Terminate Function blocks by conditionally providing the signals to execute actions such as initializing states, resetting values, or performing cleanup. For more information, see Conditionally Execute Custom Initialize, Reinitialize, Reset, and Terminate Routines.  Provide Reset Signal Conditionally This diagram illustrates conditional activation of the reset signal. The Event Listener block within the Reset Function block has the Enable variant condition option selected. The Variant control parameter of the Event Listener block is specified as V == 1. When variant condition V == 1 evaluates to true, the reset signal is triggered, causing the state of the Discrete-Time Integrator block to reset. When V == 1 evaluates to false, the reset signal is not triggered, and the Discrete-Time Integrator block continues operating with its current state unchanged.

Adapt Signal Dimensions

Symbolic dimensions allow you to use symbols instead of fixed numbers to specify signal sizes. Use symbolic dimensions to assign different dimensions a signal without changing the model structure. See Use Symbolic Signal Dimensions. You can generate C/C++ code for models that use variant dimensions as described in Implement Symbolic Dimensions for Array Sizes in Generated Code (Embedded Coder).

 Use Symbols Instead of Fixed Numbers to Define Flexible Signal Dimensions

This diagram illustrates the use of a symbolic dimension to represent the dimension of a signal in a model. By assigning the port dimension to a symbolic variable A, the model can accept input signals of varying sizes, such as vectors, matrices, array sizes, and so on.

Control Bus Signal States With Variants

Buses group multiple signals or messages into a single entity, simplifying block diagrams and reducing clutter. The signals within a bus can have variant conditions propagated from connected variant blocks. If the variant condition for a signal evaluates to true, the signal is active. Otherwise, the signal is inactive. For more information, see Variant Elements Within Buses. To learn about generating C/C++ code for models that use variant blocks with buses, see Generate Code for Variant Elements Within Buses (Simulink Coder).

 Selectively Activate Signals in Variant Bus

This diagram illustrates buses with elements having different variant conditions. The model consists of two signals, a and b, which are merged to form a bus. These signals have variant conditions, V == 1 and W == 1 respectively. During simulation, if the variant condition V == 1 evaluates to true, the signal a is active. Otherwise, the signal a is inactive.

Choose Between Inline and Hierarchical Variant Blocks

You can select between variant blocks based on factors such as hierarchy support, interface flexibility, default variant specification, control ports, and so on. This table compares hierarchical and inline variant blocks.

Feature	Inline variant blocks (Variant Source and Variant Sink)	Hierarchical variant blocks (Variant Subsystem and Variant Assembly Subsystem)	Implementation
Variant choice representation	Number of ports	Subsystem, Model, or Subsystem Reference blocks	Use the Variant Source and Variant Sink blocks for signal routing, switching, or granular variation when the design variations have the same interface and only signal switching is needed. Use the Variant Subsystem block for encapsulating complex, modular, or reusable logic when the design variations require different interfaces.
Implementation of variant choices in a separate hierarchy	No	Yes	Use the Variant Subsystem block to organize large models or when each design variation is a distinct subsystem or model component.
Flexible number of inputs and outputs among variant choices (choices do not have similar interface)	No	Yes, if Propagate conditions outside of variant subsystem is selected.	Use the Variant Subsystem block when design variation have different signal requirements or structures.
Control signal support	Yes, only for single-input, single-output (SISO) blocks through data ports.	Yes, through control ports	Use control signals when you need to manage enable, trigger, or reset signals as part of the design logic.
Connection Port (Simscape) support used for modeling physical connection lines	No	Yes, only when Variant activation time is set to update diagram.	Use the Variant Subsystem block for physical modeling, for example, in Simscape™, when the connection ports are required.
Commenting out a variant choice by adding the % symbol before the variant control in the Block Parameters dialog box	No	Yes	Use the Variant Subsystem block for advanced variant management, such as temporarily disabling design variations during development or testing.

Note

You can also comment out or comment through certain elements of your model to exclude from simulation without removing them. To do so, select a model element and hover over the three dots. In the action menu that appears, select Comment Out or Comment Through.

• When you comment out a block, the block is disabled, and its input signals do not propagate through the model. Use Comment Out to temporarily disable a block without affecting the rest of the model.

• When you comment through a block, the block is bypassed, but the signals continue to propagate through the model. Use Comment Through to exclude a block from simulation while still allowing the signals to reach downstream blocks.

Unlike variant blocks, the Comment Out and Comment Through options are not controlled by variant logic, require manual changes to switch, are not reflected in generated code, and are intended primarily for testing purposes. For more information, see Comment Out and Comment Through Blocks.

For variant choices in hierarchical variant blocks, the Comment Out and Comment Through options are not available. To comment out a variant choice, add a % before the variant choice name in the block parameters dialog box.

Apply Variants in Simulink-based Modeling Domains

You can extend variant techniques beyond Simulink models to applications in Stateflow^®, AUTOSAR Blockset, Simscape, and System Composer™, maximizing reusability and scalability for evolving design requirements.

• Variants in Stateflow chart — Variant transitions allow you to define and switch between multiple design variations using variant controls within the same Stateflow chart. For more information, see Control Indicator Lamp Dimmer Using Variant Conditions (Stateflow).

   Control State Transition of Indicator Lamp Dimmer

  This diagram illustrates the operation of a lamp controller system that transitions between three dimmer models Dimmer1, Dimmer2, and NoDimmer. During simulation, the dimmer model Dimmer1 is active when the variable the variant condition HAS_DIMMER == 1 evaluates to true. Dimmer1 offers three brightness levels, High, Medium, and Low. The dimmer model Dimmer2 is active when the variant condition HAS_DIMMER == 2 evaluates to true and supports only High and Low brightness settings. Otherwise, NoDimmer sets the lights to full brightness.

  

• Variants in AUTOSAR Blockset components — AUTOSAR variants enable you to use variant blocks, variant parameters, and symbolic dimensions to implement AUTOSAR software components with variation points. A variation point in an AUTOSAR software component presents a choice between two or more variants. The exported ARXML code from the AUTOSAR software components contains definitions for variation point proxies and variation points. See Model AUTOSAR Variants (AUTOSAR Blockset).

   Activate or Deactivate AUTOSAR Elements Using Variant Source and Variant Sink Blocks

  This diagram illustrates the use of Variant Source and Variant Sink blocks to activate or deactivate AUTOSAR elements such as ports and runnables within a model. The Variant Source block passes the signal from Runnable_Step to trigger the myslfunc_sys function when the value of the variable A equals to 1. Otherwise, the connection is disabled and myslfunc_sys remains inactive.

  

• Variants in architecture models — Variant components enable you to represent multiple architectural implementations of a system component in a single model and switch between them using variant controls. For more information, see Design Insulin Infusion Pump Using Model-Based Systems Engineering (System Composer).

  Note

  If you have a Component (System Composer) block in your model and want to include multiple architectural configurations within it, use the makeVariant (System Composer) function to convert the Component block into a Variant Component (System Composer) block. To manage the variant choices from external sources, convert the Component block to a Variant Assembly Component (System Composer) block using the makeVariantAssembly (System Composer) function.

   Switch Between Sensor Implementations Using Variant Component Block

  This diagram illustrates how the Variant Component block enables switching between different architectural implementations of sensors within the system. The Variant Component block, named BGSensor, contains two variant choices SensorA and SensorB each representing a sensor from a different manufacturer. During simulation, the variant choice whose label you select, SensorA or SensorB, becomes active.

  

• Variants in physical network — The Variant Connector (Simscape) block and the Variant Subsystem block in label mode allow you to define variations within a physical network in your model. You can activate or deactivate the variations using variant controls during simulation. For more information, see Using Variant Connectors to Implement Variations in Physical Networks (Simscape).

   Control Current Flow in Electrical Circuits Using Primary and Nonprimary Variant Connector Blocks

  This diagram demonstrates how primary and nonprimary Variant Connector blocks activate or deactivate the flow of current to specific components within an electrical circuit. The blocks in the bounded region formed by the primary Variant Connector block VC1 and nonprimary Variant Connector block VC2 are active when the value of the variable A is set to 1. The blocks in the bounded region formed by the primary Variant Connector block VC3 and nonprimary Variant Connector block VC4 are active when the value of the variable B is set to 1.

  

Case Study: Define and Manage Variants in Automotive Assembly Line

The Use Variants in Automotive Assembly Line Production example explains how variant concepts apply on the automotive assembly line to define, manage, and switch between different vehicle configurations. Additionally, it explains the role of the Variant Manager for Simulink support package in synchronizing these variants to minimize errors and discrepancies during operations.

Additional Resources for Working with Variants

The topics below provide additional resources for using variant control elements to define and activate design variations in a model and for using the Variant Manager for Simulink support package to synchronize variations within a variant configuration.

• Define, Activate, and Manage Design Alternatives in Simulink Model

• Visualize and Interpret Variant Elements in Simulink Models

• Variant Modeling Strategies for System Development with Simulink

• Manage Variants with Variant Manager for Simulink