R7015E 2019 LP1  —  System Identification

Deadline index for scalable-learning actual date
D1 10 Sept. 2019, 14:45 (Tue.)
D2 11 Sept. 2019, 14:45 (Wed.)
D3 18 Sept. 2019, 14:45 (Wed.)
D4 20 Sept. 2019, 10:15 (Fri.)
D5 24 Sept. 2019, 14:45 (Tue.)
D6 25 Sept. 2019, 14:45 (Wed.)
D7 27 Sept. 2019, 10:15 (Fri.)
D8 1 Oct. 2019, 14:45 (Tue.)
D9 2 Oct. 2019, 14:45 (Wed.)
D10 4 Oct. 2019, 10:15 (Fri.)
D11 8 Oct. 2019, 14:45 (Tue.)
D12 9 Oct. 2019, 14:45 (Wed.)
D13 15 Oct. 2019, 14:45 (Tue.)
D14 18 Oct. 2019, 10:15

Module 0 - introduction

sub-module ID sub-module name deadline index section in the textbook raw video slides (.pdf)
M0.0 surveys D1 link link
M0.1 the system identification procedure D1 1.3 link link
M0.2 R7015E and the other courses at LTU will be done in class link link
M0.3 prerequisites will be done in class link link
M0.4 organization of the course will be done in class link link
M0.5 assessments will be done in class link link

Module 1 - models of linear time invariant systems + their predictors

sub-module ID sub-module name deadline index section in the textbook raw video slides suggested exercises*
M1.1 linear systems D1 2.1 link link 2E.4 2E.6
M1.2 prediction D1 3.2 link link 3G.1 3E.1 3E.2 3E.3
M1.3 linear models D1 4.1 link link 4E.11
M1.4 transfer function models D1 4.2 link link 4G.1, 4G.8, 4E.1
M1.5 state space models D1 4.3 link link 4G.2, 4E.3
M1.6 identifiability issues D2 4.6 link link 4E.5
lab ID lab name suggested deadline index assignment (.pdf) datasets (.zip) report template (.zip)
L1 Kalman vs. Luenberger 21 Sept. 2019, 10:00 (Fri.) link link link
auxiliary external material
analytic functions
book on Kalman filters with Python code

Module 2 - frequency-domain identification methods

sub-module ID sub-module name deadline index section in the textbook raw video slides suggested exercises*
M2.1 transient response and autocorrelation analysis D2 6.1 link link
M2.2 frequency response analysis D3 6.2 link link
M2.3 Fourier analysis D3 6.3 link link
M2.4 spectral analysis D3 6.4 link link
lab ID lab name suggested deadline index assignment (.pdf) datasets (.zip) report template (.zip)
L2 successes and pitfalls of frequency-domain identification methods 28 Sept. 2019, 10:00 (Fri.) link link link

Module 3 - least squares and maximum likelihood

sub-module ID sub-module name deadline index section in the textbook raw video slides suggested exercises*
M3.1 statistics and estimators D4 link link link
M3.2 LS estimators D4 link link link
M3.3 ML estimators D5 link link link
M3.4 MSE and bias / variance tradeoff D6 link link link
M3.5 admissibility D6 link link link
M3.6 unbiasedness and unbiased minimum variance D6 link link link
M3.7 Fisher information, Cramer-Rao and efficiency D7 link link link
M3.8 asymptotic properties D7 link link link
M3.9 theoretical properties of linear LS estimators D7 link link link
M3.10 theoretical properties of ML estimators D7 link link link
lab ID lab name suggested deadline index assignment (.pdf) datasets (.zip) report template (.zip)
L3 ML-based estimation of occupancy patterns 12 Oct. 2019, 10:00 (Fri.) link link link

Module 4 - PEM-based identification of FIR and ARX models

sub-module ID sub-module name deadline index section in the textbook raw video slides suggested exercises*
M4.1 guiding principles D8 7.1 link link
M4.2 PEM D8 7.2 link link
M4.3 PEM-based identification of FIR and ARX models D8 7.3 link link

Module 5 - Identification of ARMAX and BJ models

sub-module ID sub-module name deadline index section in the textbook raw video slides suggested exercises*
M5.1 PEM as a ML estimator D9 7.4 link link
M5.2 Instrumental Variable methods D10 7.5 - 7.6 link link
M5.3 user choices D11 15 link link
lab ID lab name suggested deadline index assignment (.pdf) datasets (.zip) report template (.zip)
L4 identification of the engine requirements for a wheel loader D17 link link link

Module 6 - model selection and validation

sub-module ID sub-module name deadline index section in the textbook raw video *slides
M6.1 general considerations D11 16.1 link link
M6.2 a priori considerations D12 16.2 link link
M6.3 selection based on preliminary data analysis D12 16.3 link link
M6.4 comparing model structures D13 16.4 link link
M6.5 model validation D13 16.5 link link
M6.6 asymptotic variances for PEM estimators D14 9.2 - 9.3 link link
additional material
some more information about nonparametric methods for sysid

Kahoots

lesson link
1 link
2 link
3 link
4 link
5 link
7 link
8 link
9 link
10 link
11 link
13 link
17 link

Auxiliary info

study guide (.tex sources)

self monitoring form (.tex sources)

potential questions during the exam

textbook (Lennart Ljung, System identification - theory for the user, Prentice-Hall, second edition) (errata)

(some) solutions of the exercises from the book from Umeå

suggestions on how to write good code