• Colloquium Winter Term 2022/2023

    The CeNoS Colloquium starts at 4.30 p.m. as zoom session. If you are interested in the zoom link, please send an email to okamp@uni-muenster.de


    Collocation-Based EEG/MEG Inverse Problem Solution Based on Helmholtz Reciprocity Principle: The Same MNE?

    Sergey N Makarov
    Worcester Polytechnic Institute (Worcester MA) and Martinos Faculty at Massachusetts General Hospital (Charlestown MA)

    Helmholtz reciprocity principle makes it possible to expand an unknown cortical dipole density into global bases - cortical fields of sequentially excited electrodes (EEG) or coils (MEG). Some results will be given and a connection to the minimum norm estimation (MNE) formulation will be discussed.



  • Colloquium Summer Term 2022

    The CeNoS Colloquium starts at 16.30 s.t as zoom session. If you are interested in the zoom link, please send an email to okamp@uni-muenster.de


    Online Colloquium / ZOOM Session

    Data-Driven Analysis of Complex Systems for a Sustainable Future

    Dr. Benjamin Schäfer
    Karlsruher Institut für Technologie (KIT) / Institut für Automation und angewandte Informatik (IAI)

    The transition to a sustainable energy system raises numerous complex challenges, as power generation becomes more uncertain, while simultaneously more operational data becomes available. Still, a coherent approach to fully utilize such data is missing. Within this talk, I will sketch our efforts in developing a data-driven framework, combining data analysis, machine learning and modelling approaches to study complex (energy) systems. I will emphasize the need for open data but also for transparency in physical and machine (learning) models. In particular, I will show how we may extract scaling laws, obtain forecasts, and identify drivers in different power grids from a purely data-driven perspective.


    Online Colloquium / ZOOM Session

    Modelling and controlling turbulent vortex dynamics in cardiac systems and active fluids

    Prof. Dr. Markus Bär
    Physikalisch-Technische Bundesanstalt (PTB), Berlin 

    Many relevant experimental systems and applications exhibit irregular dynamics such as spatiotemporal chaos or turbulence.  In this talk, I will present two examples of such behavior where control of such states by transforming their dynamics into homogeneous or periodic steady states or regular periodic patterns is desirable. The specific applications are (i) defibrillation of cardiac tissue by periodic pacing and (ii) turbulence in active suspensions. For the cardiac system, we used simulations as well as data analysis not only to identify suitable conditions e. g. pacing strength and periods [1, 2], but also to determine the mechanism of defibrillation in this specific modus.  For active turbulence, a simple phenomenological deterministic continuum model was used for quantitative modelling of experimental control in bacterial suspensions. Detailed investigations showed how the turbulent collective dynamics could be transformed into more regular, periodic spatiotemporal patterns of different symmetries using a periodic array of obstacles. This is in line with recent experimental findings using confinement by an arrays of pillars [3].  A recent study addresses the competition between the controlling effect of confinement and a nonlinear advection term behind the emergence of turbulence in active fluids demonstrating that the breakdown of control occurs through an Ising-like phase transition in a coarse-grained description of the model [4].



    [1] P. Buran, M. Bär, S. Alonso and T. Niedermayer. Chaos 27, 113110 (2017).

    [2] P. Buran, T Niedermayer, and M Bär. Preprint on https://www.biorxiv.org (2021).

    [3] H. Reinken, D. Nishiguchi, S. Heidenreich, A. Sokolov, M. Bär, S. H. L. Klapp, and I. Aranson. Comm. Phys. 3, 76 (2020).

    [4] H. Reinken, S. Heidenreich, M. Bär, S. Klapp. Phys. Rev. Lett. 128, 048004 (2022).



    Online Colloquium / ZOOM Session

    Machine Learning for Nanophotonics and Nanophotonics for Machine Learning

    Prof. Dr. Carsten Schuck
    Physikalisches Institut  (WWU) /  Center for NanoTechnology (CeNTech) / Center for Soft Nanoscience (SoN)

    Machine Learning techniques offer novel approaches to a wide range of complex problems that may defy human intuition. One area where this is particularly relevant is the field of nanophotonic design. Nanostructured Photonic Integrated Circuits (PICs) are replacing their electrical equivalents in high bandwidth telecommunication as well as sensing applications and play a key role in modern quantum technology. However, designing nanophotonic circuit components today relies primarily on combining intuitive concepts, such as interference and coupled mode theory, with brute force parameter optimization. In view of the vastness of possible device configurations, such conventional design approaches only consider an infinitesimal part of the solution space. We here show how to go beyond this paradigm by employing an inverse design technique that relies on reinforcement learning for finding non-intuitive configurations of nanophotonic devices that exceed conventional designs in both performance and compactness.

    Nanophotonics, on the other hand, also provides means for novel hardware implementations of neural networks. Optical artificial neural networks (OANNs) allow for extremely energy-efficient signal processing at the speed of light, therewith overcoming the performance limitations of corresponding implementations on electronic hardware, such as CPUs or GPUs, that are subject to the von Neumann bottleneck. Here we show a path towards realizing the linear and nonlinear building blocks of OANNs on a nanophotonic platform that could be scaled to large system size, thus realizing a photonic neural network processing unit. In particular, we leverage a variety of molecular compounds developed in the context of the collaborative research center on “Intelligent Matter” (CRC 1459) to realize nonlinear activation functions and develop interfaces between nanophotonic OANN components and photoresponsive molecular systems. We further study the performance of such nanophotonic neural networks that are trained in the presence of noise, naturally occurring in ensembles of molecular compounds but absent in digital electronic implementations.


    Online Colloquium / ZOOM Session

    What is the Use of Nonlinear Dynamics in Practice? - Use Cases from Critical Infrastructures and Life Sciences

    Dr. Arndt Telschow
    Cuculus GmbH, Erfurt

    Nonlinear dynamics and nonlinear effects are ubiquitous in complex systems, including ecosystems, technical and social systems. In the lecture, I will use selected examples from critical infrastructure and the life sciences to show the great benefits that nonlinear approaches can have in practice. The use cases are partly based on my work at the Westfälische Wilhelms-Universität Münster and partly on my current work at the software company Cuculus (www.cuculus.com). Use cases discussed include: (1) Forecasting. I present the ZONOS 24insight software module for forecasting electricity consumption and generation (see https://24insight.zonos.de for a live demo). (2) Fraud Detection. I present the results of a white paper describing a new approach to identifying electricity theft. (3) Anticipate critical transitions. I present a new method for quantifying resilience in complex systems and show how it can be used to predict tipping points. By analyzing empirical data, the method is used to anticipate (i) power grid collapse, (ii) global climate change (end of last greenhouse earth), bacterial population collapse, physiological changes and intracellular switching.


    Online Colloquium / ZOOM Session

    Design and Theory of Self-Avoiding Active Matter

    Dr. Mazi Jalaal
    University of Amsterdam 

    Active matters present systems made of micro-scale agents that consume energy from the environment. In many types of such systems, the active agent leaves a trace behind, creating memory in space and time. In this talk, I will first give a brief overview of different classes of active matter with such a feature and then present a frugal and tuneable experimental framework in which one could synthetically design active particles with a self-avoiding memory. I will then present a theoretical framework based on mathematical billiards and show how active matter with memory can result in complex, chaotic patterns in a simple deterministic dynamical system. 


    Additional date / Two talks

    1. Talk:  On-off Intermittency due to Parametric Lévy Noise

    Dr. Adrian van Kan
    University of California, Berkeley | UCB · Department of Physics 

    Instabilities arise in many physical systems at some parameter threshold. Typically the system is embedded in an uncontrolled noisy environment. The fluctuating properties of the environment affect the control parameters of the instability, which leads to multiplicative noise. The result of multiplicative noise close to an instability threshold is on-off intermittency, which is characterised by an aperiodic switching between a large-amplitude “on” state and a small-amplitude “off” state.
    I present a new type of intermittency, Lévy on-off intermittency, which arises from multiplicative α-stable white noise close to an instability threshold [1]. We study this problem in the linear and nonlinear regimes, both theoretically and numerically, for the case of a pitchfork bifurcation with fluctuating growth rate. In a recently introduced point-vortex model of 3-D perturbations in 2-D flows [2], the perturbation amplitude obeyed  such an equation. We compute the stationary distribution analytically and numerically from the associated fractional Fokker-Planck equation in the Stratonovich interpretation.
    We characterize the system in the parameter space (α, β) of the noise, with stability parameter α ∈ (0, 2) and skewness parameter β ∈ [−1, 1]. Five regimes are identified in this parameter space, in addition to the well-studied Gaussian case α = 2. Three regimes are located at 1 < α < 2, where the noise has finite mean but infinite variance. They are differentiated by β and all display a critical transition at the deterministic instability threshold, with on-off intermittency close to onset. Critical exponents are computed from the stationary distribution. Each regime is characterised by a specific form of the density and specific critical exponents, which differ starkly from the Gaussian case. A finite or infinite number of moments may converge, depending on parameters. Two more regimes are found at 0 < α ≤ 1. There, the mean of the noise diverges, and no transition occurs. In one case the origin is always unstable, independently of the distance μ from the deterministic threshold. In the other case, the origin is conversely always stable, independently of μ. In summary, we demonstrate that an instability subject to non-equilibrium, power-law-distributed fluctuations can display substantially different properties than for Gaussian thermal fluctuations, in terms of statistics and critical behavior.

    [1]  van Kan, A., Alexakis, A. and Brachet, M.E., 2021. Lévy on-off intermittency. Physical Review E, 103(5), p.052115.
    [2] van Kan, A., Alexakis, A. and Brachet, M.E., 2021. Intermittency of three-dimensional perturbations in a point-vortex model. Physical Review E, 103(5), p.053102.


    2. Talk: Pressure-driven Wrinkling of Soft Inner-lined Tubes

    Dr. Ben Foster
    University of California, Berkeley | UCB · Department of Physics 

    A simple equation modelling an inextensible elastic lining of an inner-lined tube subject to an imposed pressure difference is derived from a consideration of the idealised elastic properties of the lining and the pressure and soft-substrate forces. Two cases are considered in detail, one with prominent wrinkling and a second one in which wrinkling is absent and only buckling remains. Bifurcation diagrams are computed via numerical continuation for both cases. Wrinkling, buckling, folding, and mixed-mode solutions are found and organised according to system-response measures including tension, in-plane compression, maximum curvature and energy. Approximate wrinkle solutions are constructed using weakly nonlinear theory, in excellent agreement with numerics. Our approach explains how the wavelength of the wrinkles is selected as a function of the parameters in compressed wrinkling systems and shows how localised folds and mixed-mode states form in secondary bifurcations from wrinkled states. Our model aims to capture the wrinkling response of arterial endothelium to blood pressure changes but applies much more broadly.



  • Colloquium Winter Term 21/22

    The CeNoS Kolloquium starts at 16.30 s.t as zoom session. If you are interested in the zoom link, please send an email to okamp@uni-muenster.de


    Smart grid communication and cyber security: a computer-science perspective

    Prof. Anna Remke
    Institut für Informatik, WWU

    Within smart grids the safe and dependable distribution of electric power highly depends on the security of Supervisory Control and Data Acquisition (SCADA) systems and their underlying communication protocols. Existing network-based intrusion detection systems for Industrial Control Systems (ICS) are usually centrally applied at the SCADA server and do not take the underlying physical process into account. A recent line of work proposes an additional layer of security via a process-aware approach applied locally at the field stations. We broaden the scope of process-aware monitoring by considering the interaction between neighboring field stations, which facilitates upcoming trends of decentralized energy management (DEM). Local security monitoring is lifted to monitoring neighborhoods of field stations, therefore achieving a broader grid coverage w.r.t. security.


    The role of non-conservative interactions in nonequilibrium complex systems

    Dr. Sarah Loos
    International Center for Theoretical Physics, Trieste, Italy

    The complex world surrounding us, including all living matter and various artificial complex systems, is mostly far from thermal equilibrium. A major goal of modern statistical physics and thermodynamics is to unravel the fundamental principles and in particular the collective phenomena of such nonequilibrium systems, like the swarming of fish, flocking of birds, or pedestrian crowd dynamics. A key novel concept to describe and classify nonequilibrium systems is the stochastic entropy production, which explicitly quantifies the breaking of time-reversal symmetry. However, so far, little attention has been paid to the implications of non-conservative interactions, such as time-delayed (i.e., retarded) or nonreciprocal interactions, which cannot be represented by Hamiltonians contrasting all interactions traditionally considered in statistical physics. Time-delayed and nonreciprocal interactions indeed emerge commonly in biological, chemical and feedback systems, and are widespread in engineering and machine learning. In this talk, I will use simple time- and space-continuous models to discuss technical challenges and unexpected physical phenomena induced by nonreciprocity [1,2] and time delay [3,4].

    [1] Loos and Klapp, NJP 22, 123051 (2020)
    [2] Loos, Hermann, and Klapp, Entropy 23, 696 (2021)
    [3] Loos and Klapp, Sci. Rep. 9, 2491 (2019)
    [4] Holubec, Geiss, Loos, Kroy, and Cichos, Phys. Rev. Lett. in press (2021)


    Complexity meets energy: A slightly biased, high-variance survey on modeling power grid dynamics

    Dr. Katrin Schmietendorf
    Center for Nonlinear Science, WWU

    In the course of the energy transition the ‚traditional‘ power grid with large conventional plants injecting constant feed-in adjustable to the current demand is progressively being transformed into a decentralized grid with small and medium generating units, many of which are volatile and display feed-in variations on time scales ranging from annual to sub-seconds. This development poses novel challenges with respect to stable and reliable grid operation, frequency quality, storage facilities, control schemes and grid design. Against this backdrop, power system analysis has evolved from a primarily engineering topic to an interdisciplinary research area involving, inter alia, statistics, meteorology, computer science, turbulence research, network and complex systems theory.

    About a decade ago, Filatrella et al. uncovered that the most basic generator model standardly used in electrical engineering corresponds to a modification of the Kuramoto model, a paradigmatic model for self-organized synchronization. Building on this relationship, a new branch of research, also referred to as Kuramoto-like power grid modeling, evolved within the nonlinear dynamics community. In Kuramoto-like power grid modeling power systems are analyzed as networks of coupled oscillators. Various application-oriented issues have been addressed so far, e. g. power outages due to severe perturbations and transmission-line overloads, cascading failures, the impact of fluctuating feed-in on power quality and grid stability, storage and new control strategies.

    The purpose of my talk is to give a survey across the major branches of Kuramoto-like power grid modeling, summarize the main results obtained so far and attempt an outlook to future focuses of research, both thematically and methodologically.  Despite its summary character, the talk will put emphasis on findings me and collaborators obtained during my doctoral research which concern the impact of short-term wind power fluctuations on grid stability and power quality. The overview and outlook is also intended to serve as basis for identifying potential synergy effects and points of contact for follow-up research projects at the WWU.