Linearization Basics
Define system to linearize, plot linear response,
validate linearization results
You can linearize a Simulink® model at the default operating point defined in the model. For more information, see Linearize Simulink Model at Model Operating Point. You can also specify an operating point found using an optimization-based search or at a simulation time.
To extract the linearized response of a portion of your model, you can define specific linearization input and output points. For more information, see Specify Portion of Model to Linearize. After linearization, you can analyze and validate your results in both the time domain and frequency domain.
Apps
| Model Linearizer | Linearize Simulink models |
| Linearization Manager | Manage linear analysis points and block linearizations |
Functions
linearize | Linear approximation of Simulink model or subsystem |
linearizeOptions | Set linearization options |
getlinio | Obtain linear analysis points from Simulink model, Linear Analysis Plots block, or Model Verification block |
linio | Create linear analysis point for Simulink model, Linear Analysis Plots block, or Model Verification block |
setlinio | Save linear analysis points to Simulink model, Linear Analysis Plots block, or Model Verification block |
linlft | Linearize model while removing contribution of specified blocks |
linlftfold | Combine linearization results from specified blocks and model |
fastRestartForLinearAnalysis | Fast restart for linear analysis |
Topics
Getting Started
- Linearize Nonlinear Models
Obtain a linear approximation of a nonlinear system that is valid in a small region around an operating point. - Choose Linearization Tools
Simulink Control Design™ software lets you perform linear analysis of nonlinear models using a user interface, functions, or blocks. - Exact Linearization Algorithm
Simulink Control Design software linearizes models using a block-by-block approach. The software individually linearizes each block in a Simulink model and produces the linearization of the overall system by combining the individual block linearizations. - Linearize Simulink Model at Model Operating Point
Linearize a model at its operating point consisting of initial state values and input signals. - Linearize Plant
You can linearize a block or subsystem in your Simulink model without defining separate analysis points for the block inputs and outputs. The specified block or subsystem is isolated from the rest of the model before linearization. - Compute Open-Loop Response
You can analyze and compute the combined response of the plant and controller, excluding the effects of the feedback loop. - Configure Model Linearization Options
When linearizing models using Model Linearizer, you can adjust the resulting linearized model by changing the linearization settings. - Order States in Linearized Model
You can control the order of the states in a linearized model. This state order appears in linearization results.
Operating Point Specification
- Linearize at Trimmed Operating Point
You can linearize a Simulink model at an operating point that meets specified input, output, or state constraints. - Linearize at Simulation Snapshot
Simulate a Simulink model and extract the state and input levels of the system at specified simulation times. - Linearize at Triggered Simulation Events
You can linearize a Simulink model at specific events in time. Linearization events can be trigger-based events or function-call events.
Specify Portion of Model
- Specify Portion of Model to Linearize
Specify the subsystem, loop, or block to linearize using linear analysis points.
- How the Software Treats Loop Openings
Loop openings affect only how the software recombines linearized blocks, not how the software linearizes each block. The software ignores openings when computing operating points.
Models with Special Characteristics
- Linearize Models with Delays
When linearizing a Simulink model with continuous-time delay blocks, you can either approximate the delays or represent the delays exactly. - Linearize Multirate Models
You can linearize a Simulink model that contains blocks with different sample times.
- Linearize Models with Model References
To achieve an accurate block-by-block linearization of a model reference subsystem, first set it to run its simulation in normal mode. - Linearize Simscape Networks
You can linearize models with Simscape™ components using Simulink Control Design software. - Linearize Sparse Models
Obtain sparse linear systems from Simulink models that contain Sparse Second Order or Descriptor State-Space blocks.
Analysis and Validation
- Analyze Results Using Model Linearizer Response Plots
Analyze the time-domain and frequency-domain responses of linearized models. You can compare the responses of multiple models and view system characteristics such as stability margins and settling time. - View Linearized Model Equations Using Model Linearizer
You can view the state-space equations of your linearized model in the Model Linearizer. - Validate Linearization in Frequency Domain Using Model Linearizer
You can assess the accuracy of your linearization results by estimating the frequency response of the nonlinear model and comparing the result with the response of the linearized model. - Validate Linearization in Time Domain
You can assess the accuracy of your linearization results by comparing the simulated output of the nonlinear model and the linearized model.
Code Generation
- Generate MATLAB Code for Linearization from Model Linearizer
To reproduce your interactive linearization results at the command line, you can generate MATLAB® scripts or functions using the Model Linearizer.
Featured Examples
Trim and Linearize an Airframe
Find a steady-state operating point for an aircraft traveling at a set speed and linearize the model at that operating point.
Linearize Pneumatic System at Simulation Snapshots
Linearize a Simulink model of a pneumatic system at time-based operating point snapshots.
Linearize Pulp Paper Process Model
Linearize a Simulink model of an industrial process at a steady-state operating point.
