Model Predictive Control Toolbox 3.1.1

Introduction

Model Predictive Control Toolbox™ provides MATLAB® functions, a graphical user interface (GUI), and Simulink® blocks for designing and simulating model predictive controllers in MATLAB and Simulink. These controllers optimize the performance of multi-input/multi-output systems that are subject to input and output constraints.

The toolbox lets you define an internal plant model used by the model predictive controller in three ways. You can estimate the model from experimental data (with System Identification Toolbox™), obtain it from a linearized Simulink model, or specify it directly as a linear time invariant object, such as a transfer function, or a state space model. The plant model can include delays.

You can implement the model predictive controller by generating C code (with Real-Time Workshop®).

Key Features

• Graphical user interface and MATLAB commands for designing and simulating model predictive controllers
• Ability to define an internal linear plant model from experimental data or linearized Simulink model
• Simulink blocks for designing and simulating model predictive controllers directly in Simulink
• Control of nonlinear plants using multiple model predictive controllers with bumpless control transfer
• Ability to handle time-varying constraints and weights, off-diagonal weights, and custom unmeasured disturbance models
• Ability to generate C code for application deployment (with Real-Time Workshop)

Using one of the two blocks available in Model Predictive Control Toolbox to design and simulate a controller directly in Simulink. Click on image to see enlarged view.