Model-based Identification and Estimation (etit-214)

Model-based identification and estimation
V-UE; 2+1 SWS; ECTS-Points: 4
Lecturer: Dr. Alexander Schaum
 

Overview

This lecture gives an introduction to identification and estimation methods by addressing the following topics:
 
  • System identification using first-principle  (white and grey box identification) as well as purely data-driven approaches (black box identification)
  • State and parameter estimation using observers and Kalman filters
     

Dates

See UnivIS (lecture ID 080084) and UnivIS (lecture ID 080086). 

Examination:

In order to find appropriate dates for the oral examination please contact the lecturer under alsc@tf.uni-kiel.de
The examination will take place in groups (<=2) and constits of the discussion of the Projects (Exercise 5, Exercise 10) as well as some general questionaire. It can be carried out within the following time intervals:
17.07 - 26.07.2019
12.08 - 30.08.2019
09.09 - 20.09.2019
For the possibility of taking the exam at the beginning/middle of october please contact the lecturer.

 

Downloads

Lecture notes

  • Introduction
  • Part1: Linear continuous-time systems (solution and behavior)
  • Part2: Linear discrete-time systems
  • Part 3: Identification (general aspects, white-box identification)
  • Part 4: Grey-box identification (mean-squared error minimization)
  • Part 5: Black-box identification
  • Part 6: Observers for LTI systems
  • Part 7: Stochastic observer design (Kalman-Bucy-Filter)
  • Part 8: Sampled data stochastic observer design and joint state and parameter estimation of continuous-time systems

 

Exercises

 

References

Are provided at the end of the notes to the classes (see Downloads).

 

 

 

Research