Multivar – A MATLAB Based MIMO Control System Design Application

A newly developed, MATLAB-based, MIMO control design application, named Multivar has been designed for researchers, engineers and engineering students who are investigating and designing MIMO control systems. This application has greater operation flexibility than its predecessors because of its two modes - Function mode (which allows users to access the individual functions in any stage of the control design process) and GUI mode (a wizard-based approach utilizing the functions from the Function mode). The application can support more input model types than its predecessors as it is capable of supporting LTI systems that either contain or do not contain time delays. It also provides support in the areas of model creation, conversion, approximation and analysis, input-output pairing, decoupling, controller design, and evaluation, while adding techniques that are not found in other toolboxes such as SVD analysis and decoupling. In addition, it allows the user to save or export the final control design or compare with saved designs so that the users might select the design that best satisfies their needs.

I. Introduction

With the development of diverse control theories and approaches, the possible control design strategies for any single MIMO system have become numerous. As a result, the plethora of control methodologies and theories may cause the control engineer or engineering student to be bogged down with the various analytical and design techniques and to become indecisive as to what technique the user might apply next. Thus there is a need for a fast, flexible and productive approach to control design where the user would be allowed to select the design approach with the best performance out of the various approaches. Since the analytical and design techniques used in MIMO control system design are often mathematically complex in nature and often require the use of faster computing technology to perform calculations, MATLAB is a popular choice for researchers, engineers and students carrying out investigations on MIMO control systems. Although other MATLAB-based MIMO toolboxes have been created in the past, they have often been limited to one type of model input or tend to have less in either flexibility or features. Thus, this paper seeks to develop a new MATLAB-based MIMO toolbox for control system analysis and design to address these need. The new toolbox Multivar has been developed using MATLAB 2013a. Additional information and a copy of the new toolbox Multivar, is available on the MATLAB File Exchange [1].