Course material

Lecture notes 

-         Lecture notes (theory) (pdf)– in English:

o   Part I: Prediction Theory

1.   Stochastic processes

2.   Spectral analysis

3.   Predictors (1)

4.   Predictors (2) and examples

o   Part II: Identification

1.   Introduction and PEM approach

2.   Maximum likelihood

3.   Asymptotic analysis

4.   Whiteness test

5.   LS procedure

o   Part III: Durbin Levinson algorithm and minimum variance control + examples on GMV

o   Part IV: State space methods

1.   State-space models

2.   Identification of state-space models

o   Part V: Adaptive control

1.   Recursive least squares

2.   Adaptive control

o   Part VI: Nonlinear identification with neural networks

-         Exercise lecture notes (pdf) – in English

o   Part I: Stochastic processes + matlab function for comparing the realizations

o   Part II: matlabllE>matlab function for Anderson whiteness test + function used in class for showing the Anderson Test for different processes

o   Part III: Canonical forms and optimal predictors

o   Part IV: Identification problems + Example

o   Part V: Control problems

o   Part VI: Bayesian estimation

o   Part VII: Kalman filtering and prediction

o   Lab lectures: Matlab software tools for model identification and data analysis+ solutions to exercizes (matlab *.m file) + Material for 01/12/2017 (NEW). 

Exams

-         Guidelines for the exam and possible theoretical questions.

-         Past examinations

o   30/08/2018, solutions (.pdf)

o   05/07/2018, solutions (.pdf)

o   20/06/2018, solutions (.pdf)

o 13/02/2018, solutionsargin-left:54.0pt;text-indent:-18.0pt;line-height:150%;mso-list: l0 level2 lfo2'>o   26/01/2018, solutions (.pdf)

o   Academic year 2016/2017, all exams (.pdf)

o   Academic year 2015/2016, all exams (.pdf)

o   Academic year 2014/2015, all exams (.pdf)

o   Academic years 1999/2000-2002/2003 (10 credits – in Italian) (.zip)

o   Academic year 2003/2004 (10 credits – in Italian) (.pdf: prima parte, seconda parte)

o   Academic years 2010/2011-2013/2014 (10 credits – for Automation Engineering students – in Italian) (temi2010, temi2011-2012, temi2013, temi2014)

o   At the website http://corsi.dei.polimi.it/IMAD/IMAD_CO/index.htm, in “Course Materials” it is possible to retrieve related material in English, e.g., solved exercizes.

o   Academic years 2005/2006-2010/2011 (5 credits, i.e. analysis of stochastic processes, model identification, minimum variance control  - for Aerospace Engineering students – in Italian) (.zip)

o   Academic years 2006/2007-2012/2013 (7.5 credits – for Automation Engineering students) (temi2007,temi2010,temi2011-2012, temi2013)

Other material

-         Textbooks NEW

-         Equations of the Kalman Predictior and Generalized Minimum Variance Control (.pdf)

-         Software tools:

o   Introduction to MATLAB and SIMULINK – in Italian (.ppt)

o   IDENTIFICATION LAB MATIERIAL – in Italian:

1.   slides (.pdf)

2.   solutions to the exercizes (.zip)

o   Introduction to MATLAB SYSTEM IDENTIFICATION TOOLBOX – in Italian (.ppt)

o   IMAD LAB (Simulink toolbox for prediction and identification) (.zip)

-         Examples – in Italian:

o   Esempio su "Controllo a Minima Varianza Generalizzata"  (.ppt)

o   Identificazione del sistema del comando di corrente di un MRD (.pdf)

-         Exercizes – in Italian:

o   Esercizio sull'analisi spettrale (.pdf)

o   Realizzazione di un processo e spettro (.pdf)

o   Forma canonica di un processo stazionario (.pdf)

o   Predittore a uno e due passi (.pdf)

o   Identificazione dei parametri di un modell AR(2) data una realizzazione di un processo ARMA(2,1) (.pdf)

o   Stima spettrale con metodi diretti (.pdf)

o   Controllo a minima varianza (.pdf)

-         Papers and notes

o   S. Bittanti. Kolmogorov, non solo probabilità. Da Le Scienze, Dicembre 1988, numero 244 (.pdf).

o   S. Bittanti. History and Prehistory of the Riccati Equation (.pdf).

o   Adaptive and robust identification (.pdf).

o   Kalman filter (.pdf).

o   Minimum variance control (.pdf)

o   Non linear systems (.pdf).