10. Identification of Linear Dynamical Systems - part 2

Learning Goals: You should be able to model random disturbances and estimate models  from data.  You should understand the definition of covariance function $R_w(\tau)$and spectral density $\Phi_w(\omega)$ of a discrete random time sequence $w(t)$ and how these functions influence the typical behavior or the signals. You should be able to estimate the spectrum of signals using Periodograms, Welch's method and BlackmanTukey method. You should be able to estimate transfer functions using spectral analysis.

Reading: Ch 9 + 10.1-5 in Ljung and Glad  (same chapters in Ljung+Glad+Hansson the 2nd edition of the book)

Lecture 10 slides are available here Download available here

Lecture 10 video is available here Download available herePlay media comment.

or on Youtube 

 

Matlab code:

realizations01.m Download realizations01.m

covplot.m Download covplot.m

morecovplots.m Download morecovplots.m

spec_leakage.m Download spec_leakage.m

periodogram01.m Download periodogram01.m

welch.m Download welch.m

blackmantukey.m Download blackmantukey.m

spa01.m Download spa01.m

etfefig.m Download etfefig.m