SubjectsSubjects(version: 863)
Course, academic year 2019/2020
Experimental Identification - AP445012
Title: Experimental Identification
Guaranteed by: Department of Computing and Control Engineering (445)
Actual: from 2019
Semester: both
Points: 0
E-Credits: 0
Examination process:
Hours per week, examination: 3/0 other [hours/week]
Capacity: winter:unlimited / unknown (unknown)
summer:unknown / unknown (unknown)
Min. number of students: unlimited
Language: English
Teaching methods: full-time
For type: doctoral
Note: you can enroll for the course in winter and in summer semester
Guarantor: Mareš Jan doc. Ing. Ph.D.
Interchangeability : N445021, P445012
Annotation -
Last update: Pátková Vlasta (19.11.2018)
The main goal of the course is to familiarize the students with the common generally valid principles of system identification both analogue and discrete
Aim of the course -
Last update: Pátková Vlasta (19.11.2018)

Students will be able to:

• work with common identification methods

• work with their implementation

• evaluate their efficiency

• students will take part in one large complex work (beginning with data measuring up to model formulation)

• students will use MATLAB interface for model identification

Literature -
Last update: Pátková Vlasta (19.11.2018)

R:Tangirala A.K., Principles of System Identification:Theory and Practice, CRC Press, Boca Raton, 2015, ISBN 978-1-4398-9602-0

A:Ljung L.,Systém Identification. Theory for the User,Prentice Hall PTR,N.J.,1999,0136566952

Learning resources -
Last update: Pátková Vlasta (19.11.2018)

internal materials

Teaching methods -
Last update: Pátková Vlasta (19.11.2018)

lectures, projects and consultations

Syllabus -
Last update: Pátková Vlasta (19.11.2018)

1. Experimental identification - basic scheme, wide and narrow definition of the branch

2. Structure model selection and process dynamics

3. Signals and their characteristics, signal discretization. Input testing signals, their selection

4. Adequacy criteria for models and processes, least square method, its modification

5. Identification methods classification, deterministic models in time and frequency domain

6. Dynamic system identification via transfer characteristics - project I

7. Methods of Strejc, Broid, progressive integration - project II

8. Frequency domain identification - method of Kardašov-Karnjušin

9. Frequency domain identification - correlation and spectral analysis, statistic dynamics methods

10. Stochastic models of discrete type, noise models, drift description

11. Parameter estimation methods for discrete models, least square - simple and weighted

12. Generalized, extended, repeated LSM, maximum credibility method

13. Numeric solution of LSM, recursive methods, robust identification, bootstrap, jackknife - project III

14. Large project

Entry requirements -
Last update: Pátková Vlasta (19.11.2018)


Registration requirements -
Last update: Pátková Vlasta (19.11.2018)


Course completion requirements -
Last update: Pátková Vlasta (19.11.2018)
  • making all projects
  • final discusion on them with the demonstration of knowledge