Coursecode: tn311
Coursename: System Identification B

DUT creditpoints: 5
ECTS creditpoints: 8
Subfaculty of Mechanical Engineering and Marine Technology
Lecturer(s): Hof, dr.ir. P.M.J. Van den

Tel.: 015-2784509
Catalog data:
Experimental modelling of dynamical systems; methodology. Discrete-time signals and system analysis. Identification of transfer functions. Representations of linear models; black-box models; parametrised model sets. Identification by prediction error minimization; least squares methods. Approximate modelling; algorithms. Experiment design and data analysis; identification from closed loop data; model validation. MATLAB toolbox.

Courseyear: 3, 4
Semester: 0/4/0/0/0
Hours p/w: 4
Other hours: -
Assessment: Oral
Assessm.period(s): all year around
(see academic calendar)
Prerequisites: wb2307
Follow up: --
Detailed description of topics:
  • Methodology of identification; experimental modelling of dynamical systems; objectives and prospects.
  • Discrete-time signals and systems.
  • Identification of non-parametric models; transfer function estimation; spectral estimation.
  • Prediction error identification of black box models; model structures; identification criteria; asymptotic properties; least squares methods; algorithmic aspects; approximate modelling.
  • Model structure selection and model validation.
  • Experiment design and data processing.
  • Closed loop identification.
  • MATLAB-toolbox
Course material:
Lecture Notes: P.M.J. Van den Hof, "System Identification" (1998). (in English).
References from literature:
  • L. Ljung, "System Identification - Theory for the User". Prentice Hall, Englewood Cliffs, NJ, 1987.
  • T. Söderström and P. Stoica, "System Identification". Prentice Hall, Hemel Hempstead, UK, 1989.
Remarks (specific information about assesment, entry requirements, etc.):
Goals:
This course is designed to provide the students with engineering knowledge and skills to design experiments, analyse data and to identify and validate models of dynamical systems, as appearing in industrial process control and mechanical servo systems. The methods as introduced in the course are supported by excercises on simulation and real-life examples, employing appropriate and user-friendly software tools.
Computer use:
Personal computers are used for working out the exercises and for performing the take-home exam.
Laboratory project(s):
Design content:
The course is finished with an oral examination that is based on an assignment to identify and validate models from an experimental laboratory setup.
Percentage of design: 25%