Program

Update: 13 MARCH 2020

DUE TO THE STRICT RULES OF THE MINISTERY OF HIGHER EDUCATION, THE SPRING SCHOOL IS CANCELLED

PROGRAM OF THE 4th SPRING SCHOOL on

Data-Driven Model Learning of Dynamic Systems

GENERAL INFORMATION

A tentative time schedule for the Spring School is available here

Note that modifications can still be brought to this time schedule.

For the computer exercices,  participants  should bring their own laptop with one of the latest versions of Matlab (version R2014a at least)  installed with stand alone license. The Matlab System Identification Toolbox must be available.

PROGRAM AT A GLANCE

TUESDAY 14 APRIL (from 14:00) - WEDNESDAY 15 APRIL (all day) - THURSDAY 16 APRIL (late afternoon)

Lecturer: Xavier Bombois, CNRS Research Director, Laboratoire Ampère, Ecole Centrale de Lyon 

Topic: linear system identification

Theme 1: Introduction;concepts; identification cycle

Theme 2: Parametric (prediction error) identification methods: prediction criterion and model structures, linear and pseudo-linear regressions, conditions on data, statistical and asymptotic properties, model set selection and model validation

Theme 3: Non-parametric identification (ETFE)

Theme 4: Experiment design.

Exercises : getting hands on the different concepts using computer exercises (Matlab System Identification toolbox).

THURSDAY 16 APRIL (from morning till the mid-afternoon)

Lecturer: Paul Van den Hof, Professor, TU Eindhoven, The Netherlands

Topic: dynamic network identification

Theme 1: closed-loop identification

Theme 2: dynamic network identification

FRIDAY 17 APRIL (morning)

Lecturer: Laurent Bako, Associate Professor, Laboratoire Ampère, Ecole Centrale de Lyon

Topic: hybrid system identification

Theme 1: From sparsity-inducing optimization to robust regression

Theme 2: Application to hybrid system identification

FRIDAY 17 APRIL (afternoon)

Lecturer: Xavier Bombois, CNRS Research Director, Laboratoire Ampère, Ecole Centrale de Lyon

Topic: design of optimal identification experiments

Theme 1: Formulation as an optimization problem, accuracy and cost constraint

Theme 2: convexification of the optimization problem, parametrization of the to-be-design power spectrum

Theme 3: Alternative formulations, least costly experiment design