Quick Start — Optimization-Based Linear Control Design

In this quick start, you get an overview of the typical tasks for optimization-based linear control design using the SISO Design Tool:

1. Open a SISO Design Tool session.

2. Configure a project for optimization-based control design.

3. Specify the controller parameters to design.

4. Specify the design requirements.

5. Design the controller.

6. Evaluate the controller design.

Note   The same workflow applies to optimization-based control design for LTI models created at the command line using Control System Toolbox software. To learn how to create LTI models, see Linear (LTI) Models in the Control System Toolbox documentation.

Prerequisites for optimization-based linear control design include:

• Simulink Compensator Design Task that contains a linearized version of the Simulink model and, optionally, any response plots you configure.

  For more information on how to linearize a Simulink model for control design, see Designing Compensators in the Simulink Control Design documentation.

• Time- and frequency-domain design requirements

To design a controller using optimization methods:

1. Open a SISO Design Tool session by typing the following command at the MATLAB prompt:

   sisotool('projectname.mat')

   

   The command also opens a SISO Design for SISO Design Task window by default and any response plots you configured when you linearized the Simulink model using Simulink Control Design software.

2. Configure a project for optimization-based control design by clicking Optimize Compensators in the Automated Tuning tab of the SISO Design Task.

   

   This action creates a new Response Optimization node in the Control and Estimation Tools Manager.

   

3. Specify the controller parameters to design in the Compensators tab.

   

4. Specify the design requirements.

   1. In the Design requirements tab, click Add new design requirement. Specify the design requirements, for example Bode magnitude lower limit, in the New Design Requirement dialog box.

      

      In the SISO Design window, the yellow region with the black line segment represents the design requirement on the response plot.

      

      The Design Requirements tab also lists the design requirement.

      

   2. Repeat step a to specify additional time- and frequency-domain requirements.

5. Start the optimization to design the controller by clicking Start Optimization in the Optimization tab.

   

6. Evaluate the controller design.

   1. Examine the system's response in the response plot, for example the Bode plot, to see if it meets the requirements. The system's response must lie in the white region in order to meet the design requirement.

      

   2. Examine the controller parameter values in the Compensator tab.

      

7. Write the controller parameter values into the Simulink model. To do so, click Update Simulink Block Parameters in the SISO Design Task node.

See Also: Design an Optimization-Based PID Controller for a Linearized Simulink Model.

Types of Time- and Frequency-Domain Design Requirements for Optimization-Based Control Design

Design an Optimization-Based PID Controller for a Linearized Simulink Model

© 1984-2012- The MathWorks, Inc.    -   Site Help   -   Patents   -   Trademarks   -   Privacy Policy   -   Preventing Piracy   -   RSS