Prof.: Luis Antonio Aguirre - UFMG

Introduction

In many scientific areas and engineering the use of mathematical models is conceived through data. System Identification (SI) studies the tools for building mathematical models through data. This short course has the goal of introduce and illustrate through case studies. The primary goal is motivate students and young researchers introducing a few challenges of the area. To do this, in the second part of the course, a few topics in non-linear system identification will be presented in accessible way, a few case studies will be also presented.

Schedule: The lessons will be divided into 2 hours modules each.

First Module
• Introduction: example using thermic-vacuum chamber
• Identification as alternative to physical modeling of the system
• Stochastic Identification as alternative to deterministic modeling
• Brief mention of the five stages of the system identification: tests, representation choice, structure choice, parameter estimation and validation of models.

Second Module
• Generation of system equations
• The classic minimum squares estimator
• ARX models parameter estimation using minimum squares
• The recursive estimator of minimum squares
• Examples

Third Module
• A few challenges of non-linear system identification
• Grey-box identification
• Examples

Fourth Module
• Case Studies

Bibliography: Aguirre, L.A., Introdução à Identificação de Sistemas, 2ª Edição, Editora UFMG, 2004.