Métodos de identificação de sistemas auxiliados por computador.

Abstract

The m a i n goal of this work was to implement some i d e n t i f i c a t i on methods, w i th emphasis on the ones which use singular value decomposition. The readers w i l l find the least square method, the instrumental variable method, a method based on the singular value decomposition of the process response and also a model order r e d u c t i on method. Some results of the application of these methods are presented at the end of t h i s work. The algorithms were grouped i n a toolbox, named ISAC (Computer A i d e d System I d e n t i f i c a t i o n ) . The ISAC has a f r i e n d l y interface, and so is not necessary to the user know the MATLAB language which is based on line commands. The toolbox has also some functions t o help the user t o find a solution t o his modeling problems, like the frequency tools. The user can also design P I D controlers w i t h this toolbox. There is a model order reduction m e t h o d based on the singular value decomposition SVD of the model impulse response.