Giampiero CAMPA
 
Date of final exam: 25/02/2000

E-mail: campa@cemr.wvu.edu

Tutor: Prof.  M. Innocenti, Università di Pisa

  ___________________________________________________________________________________________________________


Multivariable systems: tools for modeling, simulation, analysis, and robust control synthesis
  ___________________________________________________________________________________________________________

Advisor:

Prof.  M. Innocenti, Università di Pisa

Summary of the thesis

This thesis covers 5 different topics related to multivariable systems:

  1. A comparison between different control techniques applied to a linearized underwater vehicle model, specifically Sliding Mode Control, m-synthesis and LMI-synthesis, the controls were tested and compared on the nonlinear model.
  2. A novel approach to robust (nonlinear) control, which only requires the augmentation of a preexisting linear control with a neural network, is presented for both the output feedback and state feedback cases.
  3. A new powerful graphical integrated environment that allows multivariable system analysis, multivariable control synthesis using up to 10 different robust control techniques, closed loop analysis and control system optimization.
  4. A new Simulink matrix library allows some operations with matrices directly in the Simulink environment. These blocks eliminate the need to write S-functions either as fast-but-long c files or small-but-slow m-files, replacing them with fast-and-small Simulink diagrams. As a result of using these blocks the simulation of nonlinear multivariable systems could be greatly improved in modularity, speed, and implementation ease.
  5. A nonlinear model with six degrees of freedom of an underwater vehicle. This model is completely written in Matlab, the notation follows closely the standard one, and each of the 10 Matlab functions is fully explained and commented, so it should provide an example on how to translate the nonlinear equations of a mechanical system into a running simulation.

 

  _______________________________________