inf5122 - Learning-Based Control in Digitalised Energy Systems (Complete module description)

inf5122 - Learning-Based Control in Digitalised Energy Systems (Complete module description)

Original version English PDF Download
Module label Learning-Based Control in Digitalised Energy Systems
Modulkürzel inf5122
Credit points 6.0 KP
Workload 180 h
Institute directory Department of Computing Science
Verwendbarkeit des Moduls
  • Master's Programme Computing Science (Master) > Angewandte Informatik
  • Master's programme Digitalised Energy Systems (Master) > Digitalised Energy System Design and Assessment
  • Master's Programme Engineering of Socio-Technical Systems (Master) > Embedded Brain Computer Interaction
  • Master's Programme Engineering of Socio-Technical Systems (Master) > Human-Computer Interaction
  • Master's Programme Engineering of Socio-Technical Systems (Master) > Systems Engineering
Zuständige Personen
  • Rauh, Andreas (module responsibility)
  • Lehrenden, Die im Modul (Prüfungsberechtigt)
Prerequisites

Basic knowledge of control of linear continuous-time and/or discrete-timesystems and/or robust control

Skills to be acquired in this module

The students identify fundamentals of learning-based control for dynamic systems.
Professional competences

The students:

  • identify fundamentals of learning-based control for dynamic systems
  • characterise problem- specific learning techniques
  • are aware of software implementations for selected test rigs.

Methological competences
The students:

  • analyse problems of learning-based control
  • generalise them independently toward novel research-oriented application scenarios.

Social competences
The students:

  • develop solution ideas for real-life control problems within an accompanying project/lab course in small teams
  • explain the obtained results in short presentations.

Self competences
The students:

  • critically reflect the achieved results of their project work
  • acknowledge limitations of various approaches for learning-based control design.
Module contents
  1. Iterative learning control (ILC)
    • Grundlegende 2D-Systemstrukturen
    • Stability criteria
    • Ausgewählte Optimierungsansätze
  2. Data-driven neural network modelling vs. first-principle modelling
    • Static function approximations
    • NARX-modelling
  3. Design of neural network- based controllers
  4. Stability of neural network-based controllers
Literaturempfehlungen
  • Moore, K.L. Iterative Learning Control for Deterministic Systems. London: Springer- Verlag. 1993
  • Jian Xin Xu; Ying Tan. Linear and Nonlinear Iterative Learning Control. Springer- Verlag. 2003
  • Bristow, D. A.; Tharayil, M.; Alleyne, A. G. "A Survey of Iterative Learning Control A learning-based method for high-performance tracking control". IEEE control systems magazine. Vol. 26. pp. 96–114. 2006
  • The Mathworks Inc. Deep Learning Toolbox – Documentation, 2021
  • Rauh, A. Folien/ Skript zur Vorlesung „Learning-Based Control in DES“
Links
Language of instruction English
Duration (semesters) 1 Semester
Module frequency every summer term
Module capacity unlimited
Teaching/Learning method V+Ü
Previous knowledge Basic knowledge of control of linear continuous-time and/or discrete-timesystems and/or robust control
Form of instruction Comment SWS Frequency Workload of compulsory attendance
Lecture 2 SoSe 28
Exercises 2 SoSe 28
Präsenzzeit Modul insgesamt 56 h
Examination Prüfungszeiten Type of examination
Final exam of module

At the end of the course

Portfolio or written exam