Linearization Troubleshooting Overview

If you do not get expected results when you linearize your Simulink^® model, you can diagnose and fix potential linearization issues using Simulink Control Design™ troubleshooting tools. The definition of an expected linearization result depends on your specific application.

Troubleshooting Workflow

To determine whether a linearization is successful and find potential linearization issues, first check the equations and response plots of the linearized model.

Result to Check	Signs of Successful Linearization	Signs of Unsuccessful Linearization	More Information
Linear analysis plots	Time-domain and frequency-domain response plot characteristics, such as rise time and bandwidth respectively, capture the expected dynamics of your system.	Response plot characteristics do not capture the dynamics of your system. For example: Bode plot gain is too large or too small. Pole-zero plot contains unexpected poles or zeros.	Analyze Results Using Model Linearizer Response Plots
Linear model equations	State-space matrices have expected number of states, inputs, and outputs. The linearized model can have fewer states than your Simulink model because, often, the path between linearization input and output points does not reach all the model states. Poles and zeros are in correct locations.	Zero linearization (D = 0) Infinite linearization (D = Inf)	View Linearized Model Equations Using Model Linearizer

If the response plots or model equations of the linearized system do not capture the expected dynamics of your system, check the:

• Operating point at which you linearized the model. For more information, see Check Operating Point.

• Analysis point placement in your model. For more information, see Check Analysis Point Placement.

Once you verify that the model operating point and analysis points are correct, if your model still does not linearize as expected, you can troubleshoot the linearization results using the Linearization Advisor. The Linearization Advisor is a troubleshooting tool that allows you to identify blocks in your model that are potentially problematic for linearization. For more information, see Identify and Fix Common Linearization Issues.

Once you have identified potentially problematic blocks, you can then troubleshoot the linearizations of the individual blocks using the Linearization Advisor. For more information, see Block Linearization Troubleshooting.

Troubleshoot Linearizations of Models with Special Characteristics

Some Simulink models and blocks do not linearize well or require special considerations during linearization.

Model Characteristic	Linearization Considerations	More Information
Large models	For some large complex models, you can systematically linearize specific model components. You can then check if these components linearize as expected.	Specify Portion of Model to Linearize
Models with delays	The method with which you represent time delays in your model can affect linearization results. For example, if a Bode plot shows insufficient lag in phase, the cause can be the Padé approximation of the model time delays.	Models with Time Delays Linearize Models with Delays
Multirate models	Incorrect sample time and rate conversion methods can cause poor linearization results in multirate models.	Linearize Multirate Models
Models with PWM signals	Models with pulse width modulation signals do not linearize well due to their discontinuities and high-frequency switching components. Consider specifying a custom linearization for such blocks.	Configure Models with Pulse Width Modulation Signals
Models with Model Reference blocks	Linearization is not fully compatible with model reference blocks running in accelerator simulation mode. Configure these subsystems to run in normal mode during linearization.	Linearize Models with Model References
Simscape™ networks	Simscape networks commonly linearize to zero when a set of the system equation Jacobians are zero at a given operating condition.	Linearize Simscape Networks