logo-polimi
Loading...
Risorse bibliografiche
Risorsa bibliografica obbligatoria
Risorsa bibliografica facoltativa
Scheda Riassuntiva
Anno Accademico 2014/2015
Scuola Scuola di Ingegneria Industriale e dell'Informazione
Insegnamento 088981 - MODEL IDENTIFICATION
Docente Formentin Simone , Piroddi Luigi
Cfu 10.00 Tipo insegnamento Corso Integrato

Corso di Studi Codice Piano di Studio preventivamente approvato Da (compreso) A (escluso) Insegnamento
Ing Ind - Inf (Mag.)(ord. 270) - CO (482) COMPUTER SCIENCE AND ENGINEERING - INGEGNERIA INFORMATICA* AZZZZ088981 - MODEL IDENTIFICATION
088977 - MODEL IDENTIFICATION AND ADAPTIVE SYSTEMS
088978 - MODEL IDENTIFICATION AND DATA ANALYSIS

Programma dettagliato e risultati di apprendimento attesi

 

Objectives

Engineering requires mathematical models to solve analysis and design problems. Yet mathematical models are often difficult to obtain based on simple physical laws. In such a case, one can resort to automatic procedures to obtain a suitable description of the system from experimental observations of the input and output variables. In other words, these identification procedures allow for the mapping of available data into models, and are the main topic of this course. The focus of the first part of the course will be mainly on prediction error minimization (PEM) methods which are at the core of signal processing and data analysis. The second part will concern more advanced topics, such as adaptive filtering, identification and control methods, state-space methods for multivariable system identification and for state estimation (Kalman filtering), nonlinear model identification methods. The course has both a theoretical and a practical flavor, with seminars on applications of the illustrated techniques in various fields.

 

Course program
1. From data to model. Laws and models in engineering and science. Problems of prediction, time series analysis, filtering, clustering, control. Model accuracy versus complexity. Data treatment. Basic elements of an identification problem: system, model, criterion, optimization, parameter uncertainty, model validation.
2. Signals and systems. Deterministic and stochastic descriptions of signals and systems. Stochastic processes and stationary stochastic processes. Covariance function and spectrum. Stationary stochastic processes generated as output of dynamic systems. Time series and I/O models (AR, MA, ARMA, ARX, ARMAX, Box&Jenkins).
3. Linear time-invariant system identification. Taxonomy of system identification approaches. Theory of linear optimal prediction (Kolmogorov-Wiener). Prediction error minimization (PEM) methods. LS, WLS. Exact Maximum-Likelihood estimation of AR and ARMA parameters. The Output Error method. Yule-Walker equations and the Durbin-Levinson algorithm.
4. Analysis and practical aspects. Model validation and complexity. Cross-validation. Identifiability and model structure selection. Design of the experimental condition. Data preprocessing. Robustness (outliers).
5. Linear time-varying system identification. Recursive methods for parameter identification: Recursive Least Squares (RLS), Extended Least Squares (ELS), Recursive Maximum Likelihood (RML). Adaptive methods for parameter identification: forgetting factor. Adaptive and predictive control. Applications.
6. Adaptive filtering and active noise control applications. Adaptive digital filtering. Standard and modified LMS algorithm. The active noise control setting. Broadband feedforward control algorithms: FXLMS, leaky FXLMS, FULMS. Narrowband feedforward control. Periodic disturbance rejection. Disturbance frequency estimation. Multiple channel control. Implementation issues. Applications.
7. Kalman filtering. Estimation, prediction and filtering based on the Kalman Filter. Using the Kalman filter for parameter estimation (gray-box identification). Extended Kalman Filter for non-linear systems. Virtual sensors. Applications.
8. Multivariable system identification. Subspace identification methods. Applications.
9. Nonlinear system identification. Nonlinear model classes: block-oriented models, NARX/NARMAX models, neural networks, Radial Basis Functions. Model structure selection: batch and recursive methods. Applications.

 

Bibliography
Main textbooks (Italian):
- S. Bittanti, "Identificazione dei modelli e sistemi adattativi", Pitagora Editrice Bologna, 2003.
- S. Bittanti, "Teoria della predizione e del filtraggio", Pitagora Editrice Bologna, 2002.
- S. Bittanti, M. Campi, "Raccolta di Problemi di Identificazione, Filtraggio, Controllo Predittivo", Pitagora Editrice, Bologna, 1995. [eserciziario]

Other useful references (English):
- T. Söderström, P. Stoica, "System Identification", Prentice Hall, London (UK), 1989.
- T. Söderström, Discrete-time stochastic systems - estimation and control, Editore: Prentice Hall, 2013.
- R. Johansson, "System Modeling and Identification", Prentice Hall, Englewood Cliffs (NJ), 1993.
- B.D.O. Anderson, J.B. Moore, "Optimal Filtering", Prentice Hall, Englewood Cliffs (NJ).
- G.C. Goodwin, K.S. Sin, "Adaptive Filtering, Prediction and Control", Prentice Hall, Englewood Cliffs (NJ), 1984. 
- S.M. Kuo, D.R. Morgan, “Active Noise Control Systems”, John Wiley & Sons, 1996.

The lecture notes of the course will also be made available.

 

Requirements 
Perspective students should preferably have followed basic courses in Systems and Control Theory and Probability Theory.


Note Sulla Modalità di valutazione

The final exam is structured in three parts:

a) A written exam containing both exercises and theoretical questions concerning the first part of the course (items 1 to 4 of the course program);

b) A written exam containing theoretical questions concerning the second part of the course (items 5 to 9 of the course program);

c) A project.


Bibliografia
Risorsa bibliografica obbligatoriaS. Formentin, Model Identification and Data Analysis: Lecture Notes
Note:

The lecture notes will be made available during the course.

Risorsa bibliografica obbligatoriaL. Piroddi, Model Identification and Adaptive Systems: Lecture notes http://home.deib.polimi.it/piroddi/mias.html#lez
Note:

The lecture notes will be made available during the course.

Risorsa bibliografica obbligatoriaS. Bittanti, Identificazione dei modelli e sistemi adattativi, Editore: Pitagora Editrice Bologna, Anno edizione: 2003, ISBN: 88-371-1200-9
Risorsa bibliografica obbligatoriaS. Bittanti, Teoria della predizione e del filtraggio, Editore: Pitagora Editrice Bologna, Anno edizione: 2002, ISBN: 88-371-1092-8
Risorsa bibliografica obbligatoriaS. Bittanti, M. Campi, Raccolta di Problemi di Identificazione, Filtraggio, Controllo Predittivo, Editore: Pitagora Editrice, Bologna, Anno edizione: 1995, ISBN: 88-371-0792-7
Risorsa bibliografica obbligatoriaS.M. Kuo, D.R. Morgan, Active Noise Control Systems, Editore: John Wiley & Sons, Anno edizione: 1996, ISBN: 0471134244
Risorsa bibliografica facoltativaT. Söderström, P. Stoica, System Identification, Editore: Prentice Hall, London (UK), Anno edizione: 1989, ISBN: 0-13-881236-5
Risorsa bibliografica facoltativaT. Söderström, Discrete-time stochastic systems - estimation and control, Editore: Prentice Hall, Anno edizione: 2013, ISBN: 978-1852336493
Risorsa bibliografica facoltativaR. Johansson, System Modeling and Identification, Editore: Prentice Hall, Englewood Cliffs (NJ), Anno edizione: 1993, ISBN: 0-13-482308-7
Risorsa bibliografica facoltativaB.D.O. Anderson, J.B. Moore, Optimal Filtering, Editore: Prentice Hall, Englewood Cliffs (NJ), Anno edizione: 1979, ISBN: 0-13-638122-7
Risorsa bibliografica facoltativaG.C. Goodwin, K.S. Sin, Adaptive Filtering, Prediction and Control, Editore: Prentice Hall, Englewood Cliffs (NJ), Anno edizione: 1984, ISBN: 0486469328

Mix Forme Didattiche
Tipo Forma Didattica Ore didattiche
lezione
60.0
esercitazione
40.0
laboratorio informatico
0.0
laboratorio sperimentale
0.0
progetto
0.0
laboratorio di progetto
0.0

Informazioni in lingua inglese a supporto dell'internazionalizzazione
Insegnamento erogato in lingua Inglese
Disponibilità di materiale didattico/slides in lingua inglese
Disponibilità di libri di testo/bibliografia in lingua inglese
Possibilità di sostenere l'esame in lingua inglese
Disponibilità di supporto didattico in lingua inglese
schedaincarico v. 1.6.1 / 1.6.1
Area Servizi ICT
21/01/2020