Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.

  

Parameter Identification of Bilinear Dynamical Systems (Expectation Maximization using Factor Graphs)

Author(s):

Andreas Isler
Conference/Journal:

Master Thesis FS 11 (10009)
Abstract:

In this thesis a novel identification procedure for bi linear dynamical systems based on Expectation Maximization (EM) using factor graphs as introduced. Factor graphs are a method of representing factorisable functions and allow the implementation of various distributed and modular algorithms e.g. in system theory. EM is an iterative optimization approach for solving maximum likelihood problems. The combination of EM using factor graphs became a relevant topic during the last decade. In the thesis the two known concepts factor graphs and EM are introduced, before their combined application is presented. Finally the novel approach for bi linear systems is introduced. The thesis concludes with exemplary results, including strength and weaknesses of the new approach.

Year:

2011
Type of Publication:

(12)Diploma/Master Thesis
Supervisor:

M. Morari

No Files for download available.
% Autogenerated BibTeX entry
@PhdThesis { Xxx:2011:IFA_4143
}
Permanent link