• Medientyp: Elektronische Hochschulschrift; E-Book
  • Titel: Propagation of fluctuations and detection of hidden units in network dynamical systems
  • Beteiligte: Hähne, Hauke [Verfasser:in]
  • Erschienen: Carl von Ossietzky Universität Oldenburg: /oops/ - Oldenburger Online-Publikations-Server, 2019-08-15
  • Umfang: text
  • Sprache: Nicht zu entscheiden
  • Entstehung:
  • Anmerkungen: Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
  • Beschreibung: Networks of dynamical systems model a broad range of complex phenomena in natural and technological systems. The analysis of the interplay of dynamics and topology is divided into two major classes of problems: While forward approaches deduce properties of the dynamical evolution from the topology and the system parameters, inverse methods infer system properties from observations of the dynamics. With this thesis, we contribute to both fields: First, we analyze the spreading of fluctuations in power grids induced by volatile wind power feed-in with analytical, experimental, and numerical methods. Second, we consider the case in which a network dynamical system is only partially accessible for observations and derive a novel method to detect nodes that are hidden from measurement. We test our approach on systems with various, possibly noisy, dynamics and demonstrate successful detection and quantification of hidden nodes even in the case where only very few nodes are measured.