Signal Processing, Dynamic System Modeling

Dozenten:
Praktikum: Prof. Dr.-Ing. Bernhard U. Seeber
Clara Hollomey, PhD
Ali Saeedi
Turnus: Sommersemester
Zielgruppe: Pflichtmodul, Elite Master Program in Neuroengineering, MSNE,
Modul nur für MSNE-Studierende!
Die Vorlesung wird auf Englisch gehalten.
ECTS: 5
Umfang: 2/1/1 (Vorlesung/Übung/Praktikum)
Prüfung: schriftlich, 90 Minuten
Zeit & Ort: Vorlesung (auf Englisch): Dienstag,       08:45 - 10:15 Uhr, N2128
Übung:                           Dienstag,       10:30 - 11:15 Uhr, N2128
Praktikum:                     Donnerstag,    14:45 - 17:45 Uhr, N1135
Termine: Vorlesungs- sowie Übungsbeginn am 23.04.2019
keine VL und UE am 21.05.2019 (SVV), am 11.06.2019 sowie am 09.07.2019
Praktikum ab 25.04.2019
Praktikumstermine werden demnächst bekannt gegeben

Inhalt

This course introduces fundamental signal processing techniques applicable to a wide variety of neural and biomedical signals from different domains, e.g. inter- and intra-neuronal cell recordings, electroencephalography (EEG), electrocardiography (ECG), electromyography (EMG), as well as biologically inspired sensory data typical for the domains of multimodal interaction and robotics (such as audio and video data).

The course is devided into two parts, an introductory signal processing part and a dynamic system modeling part.


The signal processing part will include:

  • Correct sampling of continuous signals for offline as well as (blockwise) processing and analysis in real time
  • Properties of time- and frequency domain signal transformations: Laplace- and Fourier transform, z-transform,Discrete Fourier Transform, time-frequency uncertainty, effects of temporal windowing, Short-Term Fourier Transform
  • Properties of FIR and IIR filters and filter design. Minimum- and linear phase filters, phase and group delay
  • Time-frequency signal analysis including spectrograms
  • Filter banks

Examples will be given from neuronal and audio signals.


In the second part, the course addresses the modelling of dynamic systems based on pre-processed sensory data. This comprises the identification and evaluation of model structure and parameters describing the dynamical properties of the biological system to be modelled. Concepts and tools from information theory are using to quantify the ability of a biological dynamical system to process sensory information.


This course will link to control theory, controller design and information theory to illustrate the connection between biology and engineering approaches in line with dynamical systems modelling.