5 ECTS credits
135 h study time
Offer 1 with catalog number 4023819ENR for all students in the 1st semester at a (E) Master - advanced level.
Course charter
• This course describes the various steps one must go through for obtaining a linear model of a dynamical process. It starts with the choice (design) of the measurement setup, the choice (design) of the excitation signal, the choice of the parametric model (discrete time, continuous time, parametric versus non-parametric noise model...), the estimation of the parametric model (identification toolboxes in Matlab), till finally the model selection and the model validation. Hereby the influence of each error source (stochastic measurement errors, systematic measurement errors, nonlinear distortions, time-varying effects, model errors...) on the result is studied in detail.
Overview of the content
• Nonparametric Estimation of the Frequency Response Function and the Noise Variance
o Estimation in open and closed loop
o Impact of nonlinear behaviour – concept of the best linear approximation
o Impact of time-variation – concept of the best linear time-invariant approximation
o Multiple-input, multiple output systems
• Parametric models for linear time-invariant systems
o Impact measurement setup – zero-order-hold versus band-limited
o Lumped versus distributed models
o Generalized relationship between input-output DFT spectra
o Multiple-input, multiple output systems
• Improved Nonparametric Frequency Response Function Estimators
o Local polynomial method
o Local rational methods (new)
o Modeling of the time-varying effects
• Parametric Estimation Plant Model using a Known Noise Model
o Frequency domain maximum likelihood solution (generation initial estimates, minimization cost function, calculation covariance)
o Model validation
o Model selection
o Multiple-input, multiple output systems
• Estimation Parametric Plant Model with an Estimated Nonparametric Noise Model
o Sample maximum likelihood – independent repeated experiments
o Sample maximum likelihood method – one experiment
o Multiple-input, multiple output systems
• Estimation Parametric Plant and/or Noise Models
o Classical time-domain approach – prediction error method
o Frequency domain maximum likelihood solution
o Identification in open and closed loop
o Multiple-input, multiple output systems
• Guidelines
Overview of the project
o Design of a special class of periodic excitation signals for measuring dynamical systems: quantification of the level of the noise, the nonlinear distortions and the time-varying effects.
o Use of an advanced data acquisition set up for measuring dynamical systems: uploading and downloading the signals via Matlab.
o Measurement and identification of electronic circuits available in the lab: nonlinear dynamical systems and time-varying dynamical systems.
o Use of existing system identification toolboxes in Matlab (Time- or Frequency Domain System Identification toolboxes) and/or development of own Matlab routines.
Course material
• Digital course material (Recommended):
o Presentation used during lectures are available on the website of the department ELEC
o Frequency Domain System Identification Matlab toolbox available for the students
• Additional study material is available in the following books:
o P. Eykhoff (1974). System Identification, John Wiley and Sons, London (UK).
o L. Ljung (1999). System Identification: Theory for the User, 2nd edition, Prentice-Hall, Upper Saddle River, NJ (USA).
o R. Pintelon and J. Schoukens (2012). System Identification: A Frequency Domain Approach, 2nd edition, Wiley-IEEE Press, Hoboken, NJ (USA).
o J. Schoukens, R. Pintelon, and Y. Rolain (2012). Mastering System Identification in 100 Exercises, Wiley-IEEE Press, Hoboken, NJ.
o T. Söderström (2018). Errors-in-Variables Methods in System Identification, Springer, Cham (Switzerland).
o T. Söderström and P. Stoica (1998). System Identification, Prentice-Hall, Englewood Cliffs, (NJ (USA).
General expectation
• After following this course, the student should be able to solve independently a system identification problem. A critical interpretation of the measurements and the results obtained is of key importance.
Detailed outcomes
• To successfully complete this course, the student should be able to:
o Choose an appropriate measurement setup for a given identification problem (prediction/control or physical interpretation).
o Design an excitation signal for measuring simultaneously in a user-defined frequency band the dynamic behaviour of a system, the noise level, the level of the nonlinear distortions, and the level of the time-varying effects.
o Select and validate a model (discrete- or continuous-time, dynamic order) from noisy input, noisy output observations.
o Select an appropriate estimator for a particular identification problem.
o Analyse and interpret critically measurement and identification results.
o Make a motivated choice between non-parametric and parametric noise models.
o Write a clear and concise report of your identification project. Notice unexpected/peculiar phenomena in measurements and try finding an explanation. Highlight and discuss unexpected/peculiar identification results and try finding an explanation.
o Present, explain and motivate your identification project in an oral session.
• In addition, the student must have acquired insight into:
o The problems that arise when the input signal is not known exactly and the possible solutions.
o The difficulties of identifying dynamical systems operating in closed loop, and the possible solutions.
o The peculiarities of identifying multiple-input, multiple-output systems.
Contribution to the program outcomes of the Master in Electronics and Information Technology Engineering
The Master in Engineering Sciences has in-depth knowledge and understanding of
1. exact sciences with the specificity of their application to engineering
3. the advanced methods and theories to schematize and model complex problems or processes
The Master in Engineering Sciences can
4. reformulate complex engineering problems in order to solve them (simplifying assumptions, reducing complexity)
6. correctly report on research or design results in the form of a technical report or in the form of a scientific paper
8. collaborate in a (multidisciplinary) team
9. work in an industrial environment with attention to safety, quality assurance, communication and reporting
10. develop, plan, execute and manage engineering projects at the level of a starting professional
11. think critically about and evaluate projects, systems and processes, particularly when based on incomplete, contradictory and/or redundant information
The Master in Engineering Sciences has
12. a creative, problem-solving, result-driven and evidence-based attitude, aiming at innovation and applicability in industry and society
13. a critical attitude towards one’s own results and those of others
15. the flexibility and adaptability to work in an international and/or intercultural context
16. an attitude of life-long learning as needed for the future development of his/her career
The Master in Electronics and Information Technology Engineering:
17. Has an active knowledge of the theory and applications of electronics, information and communication technology, from component up to system level.
18. Has a profound knowledge of either (i) nano- and opto-electronics and embedded systems, (ii) information and communication technology systems or (iii) measuring, modelling and control.
20. Is able to analyze, specify, design, implement, test and evaluate individual electronic devices, components and algorithms, for signal-processing, communication and complex systems.
21. Is able to model, simulate, measure and control electronic components and physical phenomena.
Factors that determine the judgement
• The quality of the written report (language, scientific content, critical attitude with respect to own results and those in the literature).
• The quality of the oral discussion about the identification project.
• The amount of project work performed (number of measurements and devices identified, own software developed, comparison of different methods).
• The understanding of the basic concepts of the course (open book exam).
We expect that you already have prior knowledge
• A good knowledge of system theory, basic statistics, basic numerical analysis, and basic data-driven modelling
• Basic lab skills for measuring dynamical systems
The final grade is composed based on the following categories:
Oral Exam determines 100% of the final mark.
Within the Oral Exam category, the following assignments need to be completed:
Grading
• The oral open book exam counts for 100% of the final mark.
• Notes:
o The report must be turned in by the date that was communicated.
o It is the responsibility of the student to make an appointment for making the measurements and for defending the project. If no appointment is made, this is equivalent to an absence.
This offer is part of the following study plans:
Master of Electronics and Information Technology Engineering: Standaard traject (only offered in Dutch)
Master of Electrical Engineering: Standaard traject BRUFACE J