Refine
Has Fulltext
- yes (8)
Document Type
- Doctoral Thesis (7)
- Habilitation Thesis (1)
Language
- English (8)
Is part of the Bibliography
- yes (8)
Keywords
- Synchronisation (4)
- Netzwerke (3)
- networks (3)
- Datenassimilation (2)
- Kuramoto model (2)
- Kuramoto-Modell (2)
- Synchronization (2)
- synchronization (2)
- Chimäre-Zustände (1)
- Complex Network (1)
Institute
Synchronization – the adjustment of rhythms among coupled self-oscillatory systems – is a fascinating dynamical phenomenon found in many biological, social, and technical systems.
The present thesis deals with synchronization in finite ensembles of weakly coupled self-sustained oscillators with distributed frequencies.
The standard model for the description of this collective phenomenon is the Kuramoto model – partly due to its analytical tractability in the thermodynamic limit of infinitely many oscillators. Similar to a phase transition in the thermodynamic limit, an order parameter indicates the transition from incoherence to a partially synchronized state. In the latter, a part of the oscillators rotates at a common frequency. In the finite case, fluctuations occur, originating from the quenched noise of the finite natural frequency sample.
We study intermediate ensembles of a few hundred oscillators in which fluctuations are comparably strong but which also allow for a comparison to frequency distributions in the infinite limit.
First, we define an alternative order parameter for the indication of a collective mode in the finite case. Then we test the dependence of the degree of synchronization and the mean rotation frequency of the collective mode on different characteristics for different coupling strengths.
We find, first numerically, that the degree of synchronization depends strongly on the form (quantified by kurtosis) of the natural frequency sample and the rotation frequency of the collective mode depends on the asymmetry (quantified by skewness) of the sample. Both findings are verified in the infinite limit.
With these findings, we better understand and generalize observations of other authors. A bit aside of the general line of thoughts, we find an analytical expression for the volume contraction in phase space.
The second part of this thesis concentrates on an ordering effect of the finite-size fluctuations. In the infinite limit, the oscillators are separated into coherent and incoherent thus ordered and disordered oscillators. In finite ensembles, finite-size fluctuations can generate additional order among the asynchronous oscillators. The basic principle – noise-induced synchronization – is known from several recent papers. Among coupled oscillators, phases are pushed together by the order parameter fluctuations, as we on the one hand show directly and on the other hand quantify with a synchronization measure from directed statistics between pairs of passive oscillators.
We determine the dependence of this synchronization measure from the ratio of pairwise natural frequency difference and variance of the order parameter fluctuations. We find a good agreement with a simple analytical model, in which we replace the deterministic fluctuations of the order parameter by white noise.
Synchronization of coupled oscillators manifests itself in many natural and man-made systems, including cyrcadian clocks, central pattern generators, laser arrays, power grids, chemical and electrochemical oscillators, only to name a few. The mathematical description of this phenomenon is often based on the paradigmatic Kuramoto model, which represents each oscillator by one scalar variable, its phase. When coupled, phase oscillators constitute a high-dimensional dynamical system, which exhibits complex behaviour, ranging from synchronized uniform oscillation to quasiperiodicity and chaos. The corresponding collective rhythms can be useful or harmful to the normal operation of various systems, therefore they have been the subject of much research.
Initially, synchronization phenomena have been studied in systems with all-to-all (global) and nearest-neighbour (local) coupling, or on random networks. However, in recent decades there has been a lot of interest in more complicated coupling structures, which take into account the spatially distributed nature of real-world oscillator systems and the distance-dependent nature of the interaction between their components. Examples of such systems are abound in biology and neuroscience. They include spatially distributed cell populations, cilia carpets and neural networks relevant to working memory. In many cases, these systems support a rich variety of patterns of synchrony and disorder with remarkable properties that have not been observed in other continuous media. Such patterns are usually referred to as the coherence-incoherence patterns, but in symmetrically coupled oscillator systems they are also known by the name chimera states.
The main goal of this work is to give an overview of different types of collective behaviour in large networks of spatially distributed phase oscillators and to develop mathematical methods for their analysis. We focus on the Kuramoto models for one-, two- and three-dimensional oscillator arrays with nonlocal coupling, where the coupling extends over a range wider than nearest neighbour coupling and depends on separation. We use the fact that, for a special (but still quite general) phase interaction function, the long-term coarse-grained dynamics of the above systems can be described by a certain integro-differential equation that follows from the mathematical approach called the Ott-Antonsen theory. We show that this equation adequately represents all relevant patterns of synchrony and disorder, including stationary, periodically breathing and moving coherence-incoherence patterns. Moreover, we show that this equation can be used to completely solve the existence and stability problem for each of these patterns and to reliably predict their main properties in many application relevant situations.
The plasmasphere is a dynamic region of cold, dense plasma surrounding the Earth. Its shape and size are highly susceptible to variations in solar and geomagnetic conditions. Having an accurate model of plasma density in the plasmasphere is important for GNSS navigation and for predicting hazardous effects of radiation in space on spacecraft. The distribution of cold plasma and its dynamic dependence on solar wind and geomagnetic conditions remain, however, poorly quantified. Existing empirical models of plasma density tend to be oversimplified as they are based on statistical averages over static parameters. Understanding the global dynamics of the plasmasphere using observations from space remains a challenge, as existing density measurements are sparse and limited to locations where satellites can provide in-situ observations. In this dissertation, we demonstrate how such sparse electron density measurements can be used to reconstruct the global electron density distribution in the plasmasphere and capture its dynamic dependence on solar wind and geomagnetic conditions.
First, we develop an automated algorithm to determine the electron density from in-situ measurements of the electric field on the Van Allen Probes spacecraft. In particular, we design a neural network to infer the upper hybrid resonance frequency from the dynamic spectrograms obtained with the Electric and Magnetic Field Instrument Suite and Integrated Science (EMFISIS) instrumentation suite, which is then used to calculate the electron number density. The developed Neural-network-based Upper hybrid Resonance Determination (NURD) algorithm is applied to more than four years of EMFISIS measurements to produce the publicly available electron density data set.
We utilize the obtained electron density data set to develop a new global model of plasma density by employing a neural network-based modeling approach. In addition to the location, the model takes the time history of geomagnetic indices and location as inputs, and produces electron density in the equatorial plane as an output. It is extensively validated using in-situ density measurements from the Van Allen Probes mission, and also by comparing the predicted global evolution of the plasmasphere with the global IMAGE EUV images of He+ distribution. The model successfully reproduces erosion of the plasmasphere on the night side as well as plume formation and evolution, and agrees well with data.
The performance of neural networks strongly depends on the availability of training data, which is limited during intervals of high geomagnetic activity. In order to provide reliable density predictions during such intervals, we can employ physics-based modeling. We develop a new approach for optimally combining the neural network- and physics-based models of the plasmasphere by means of data assimilation. The developed approach utilizes advantages of both neural network- and physics-based modeling and produces reliable global plasma density reconstructions for quiet, disturbed, and extreme geomagnetic conditions.
Finally, we extend the developed machine learning-based tools and apply them to another important problem in the field of space weather, the prediction of the geomagnetic index Kp. The Kp index is one of the most widely used indicators for space weather alerts and serves as input to various models, such as for the thermosphere, the radiation belts and the plasmasphere. It is therefore crucial to predict the Kp index accurately. Previous work in this area has mostly employed artificial neural networks to nowcast and make short-term predictions of Kp, basing their inferences on the recent history of Kp and solar wind measurements at L1. We analyze how the performance of neural networks compares to other machine learning algorithms for nowcasting and forecasting Kp for up to 12 hours ahead. Additionally, we investigate several machine learning and information theory methods for selecting the optimal inputs to a predictive model of Kp. The developed tools for feature selection can also be applied to other problems in space physics in order to reduce the input dimensionality and identify the most important drivers.
Research outlined in this dissertation clearly demonstrates that machine learning tools can be used to develop empirical models from sparse data and also can be used to understand the underlying physical processes. Combining machine learning, physics-based modeling and data assimilation allows us to develop novel methods benefiting from these different approaches.
The Earth's inner magnetosphere is a very dynamic system, mostly driven by the external solar wind forcing exerted upon the magnetic field of our planet. Disturbances in the solar wind, such as coronal mass ejections and co-rotating interaction regions, cause geomagnetic storms, which lead to prominent changes in charged particle populations of the inner magnetosphere - the plasmasphere, ring current, and radiation belts. Satellites operating in the regions of elevated energetic and relativistic electron fluxes can be damaged by deep dielectric or surface charging during severe space weather events. Predicting the dynamics of the charged particles and mitigating their effects on the infrastructure is of particular importance, due to our increasing reliance on space technologies.
The dynamics of particles in the plasmasphere, ring current, and radiation belts are strongly coupled by means of collisions and collisionless interactions with electromagnetic fields induced by the motion of charged particles. Multidimensional numerical models simplify the treatment of transport, acceleration, and loss processes of these particles, and allow us to predict how the near-Earth space environment responds to solar storms. The models inevitably rely on a number of simplifications and assumptions that affect model accuracy and complicate the interpretation of the results. In this dissertation, we quantify the processes that control electron dynamics in the inner magnetosphere, paying particular attention to the uncertainties of the employed numerical codes and tools.
We use a set of convenient analytical solutions for advection and diffusion equations to test the accuracy and stability of the four-dimensional Versatile Electron Radiation Belt (VERB-4D) code. We show that numerical schemes implemented in the code converge to the analytical solutions and that the VERB-4D code demonstrates stable behavior independent of the assumed time step. The order of the numerical scheme for the convection equation is demonstrated to affect results of ring current and radiation belt simulations, and it is crucially important to use high-order numerical schemes to decrease numerical errors in the model.
Using the thoroughly tested VERB-4D code, we model the dynamics of the ring current electrons during the 17 March 2013 storm. The discrepancies between the model and observations above 4.5 Earth's radii can be explained by uncertainties in the outer boundary conditions. Simulation results indicate that the electrons were transported from the geostationary orbit towards the Earth by the global-scale electric and magnetic fields.
We investigate how simulation results depend on the input models and parameters. The model is shown to be particularly sensitive to the global electric field and electron lifetimes below 4.5 Earth's radii. The effects of radial diffusion and subauroral polarization streams are also quantified.
We developed a data-assimilative code that blends together a convection model of energetic electron transport and loss and Van Allen Probes satellite data by means of the Kalman filter. We show that the Kalman filter can correct model uncertainties in the convection electric field, electron lifetimes, and boundary conditions. It is also demonstrated how the innovation vector - the difference between observations and model prediction - can be used to identify physical processes missing in the model of energetic electron dynamics.
We computed radial profiles of phase space density of ultrarelativistic electrons, using Van Allen Probes measurements. We analyze the shape of the profiles during geomagnetically quiet and disturbed times and show that the formation of new local minimums in the radial profiles coincides with the ground observations of electromagnetic ion-cyclotron (EMIC) waves. This correlation indicates that EMIC waves are responsible for the loss of ultrarelativistic electrons from the heart of the outer radiation belt into the Earth's atmosphere.
Oscillatory systems under weak coupling can be described by the Kuramoto model of phase oscillators. Kuramoto phase oscillators have diverse applications ranging from phenomena such as communication between neurons and collective influences of political opinions, to engineered systems such as Josephson Junctions and synchronized electric power grids. This thesis includes the author's contribution to the theoretical framework of coupled Kuramoto oscillators and to the understanding of non-trivial N-body dynamical systems via their reduced mean-field dynamics.
The main content of this thesis is composed of four parts. First, a partially integrable theory of globally coupled identical Kuramoto oscillators is extended to include pure higher-mode coupling. The extended theory is then applied to a non-trivial higher-mode coupled model, which has been found to exhibit asymmetric clustering. Using the developed theory, we could predict a number of features of the asymmetric clustering with only information of the initial state provided.
The second part consists of an iterated discrete-map approach to simulate phase dynamics. The proposed map --- a Moebius map --- not only provides fast computation of phase synchronization, it also precisely reflects the underlying group structure of the dynamics. We then compare the iterated-map dynamics and various analogous continuous-time dynamics. We are able to replicate known phenomena such as the synchronization transition of the Kuramoto-Sakaguchi model of oscillators with distributed natural frequencies, and chimera states for identical oscillators under non-local coupling.
The third part entails a particular model of repulsively coupled identical Kuramoto-Sakaguchi oscillators under common random forcing, which can be shown to be partially integrable. Via both numerical simulations and theoretical analysis, we determine that such a model cannot exhibit stationary multi-cluster states, contrary to the numerical findings in previous literature. Through further investigation, we find that the multi-clustering states reported previously occur due to the accumulation of discretization errors inherent in the integration algorithms, which introduce higher-mode couplings into the model. As a result, the partial integrability condition is violated.
Lastly, we derive the microscopic cross-correlation of globally coupled non-identical Kuramoto oscillators under common fluctuating forcing. The effect of correlation arises naturally in finite populations, due to the non-trivial fluctuations of the meanfield. In an idealized model, we approximate the finite-sized fluctuation by a Gaussian white noise. The analytical approximation qualitatively matches the measurements in numerical experiments, however, due to other periodic components inherent in the fluctuations of the mean-field there still exist significant inconsistencies.
Noise is ubiquitous in nature and usually results in rich dynamics in stochastic systems such as oscillatory systems, which exist in such various fields as physics, biology and complex networks. The correlation and synchronization of two or many oscillators are widely studied topics in recent years.
In this thesis, we mainly investigate two problems, i.e., the stochastic bursting phenomenon in noisy excitable systems and synchronization in a three-dimensional Kuramoto model with noise. Stochastic bursting here refers to a sequence of coherent spike train, where each spike has random number of followers due to the combined effects of both time delay and noise. Synchronization, as a universal phenomenon in nonlinear dynamical systems, is well illustrated in the Kuramoto model, a prominent model in the description of collective motion.
In the first part of this thesis, an idealized point process, valid if the characteristic timescales in the problem are well separated, is used to describe statistical properties such as the power spectral density and the interspike interval distribution. We show how the main parameters of the point process, the spontaneous excitation rate, and the probability to induce a spike during the delay action can be calculated from the solutions of a stationary and a forced Fokker-Planck equation. We extend it to the delay-coupled case and derive analytically the statistics of the spikes in each neuron, the pairwise correlations between any two neurons, and the spectrum of the total output from the network.
In the second part, we investigate the three-dimensional noisy Kuramoto model, which can be used to describe the synchronization in a swarming model with helical trajectory. In the case without natural frequency, the Kuramoto model can be connected with the Vicsek model, which is widely studied in collective motion and swarming of active matter. We analyze the linear stability of the incoherent state and derive the critical coupling strength above which the incoherent state loses stability. In the limit of no natural frequency, an exact self-consistent equation of the mean field is derived and extended straightforward to any high-dimensional case.
The two hallmark features of Brownian motion are the linear growth < x2(t)> = 2Ddt of the mean squared displacement (MSD) with diffusion coefficient D in d spatial dimensions, and the Gaussian distribution of displacements. With the increasing complexity of the studied systems deviations from these two central properties have been unveiled over the years. Recently, a large variety of systems have been reported in which the MSD exhibits the linear growth in time of Brownian (Fickian) transport, however, the distribution of displacements is pronouncedly non-Gaussian (Brownian yet non-Gaussian, BNG). A similar behaviour is also observed for viscoelastic-type motion where an anomalous trend of the MSD, i.e., <x2(t)> ~ ta, is combined with a priori unexpected non-Gaussian distributions (anomalous yet non-Gaussian, ANG). This kind of behaviour observed in BNG and ANG diffusions has been related to the presence of heterogeneities in the systems and a common approach has been established to address it, that is, the random diffusivity approach.
This dissertation explores extensively the field of random diffusivity models. Starting from a chronological description of all the main approaches used as an attempt of describing BNG and ANG diffusion, different mathematical methodologies are defined for the resolution and study of these models. The processes that are reported in this work can be classified in three subcategories, i) randomly-scaled Gaussian processes, ii) superstatistical models and iii) diffusing diffusivity models, all belonging to the more general class of random diffusivity models. Eventually, the study focuses more on BNG diffusion, which is by now well-established and relatively well-understood. Nevertheless, many examples are discussed for the description of ANG diffusion, in order to highlight the possible scenarios which are known so far for the study of this class of processes.
The second part of the dissertation deals with the statistical analysis of random diffusivity processes. A general description based on the concept of moment-generating function is initially provided to obtain standard statistical properties of the models. Then, the discussion moves to the study of the power spectral analysis and the first passage statistics for some particular random diffusivity models. A comparison between the results coming from the random diffusivity approach and the ones for standard Brownian motion is discussed. In this way, a deeper physical understanding of the systems described by random diffusivity models is also outlined.
To conclude, a discussion based on the possible origins of the heterogeneity is sketched, with the main goal of inferring which kind of systems can actually be described by the random diffusivity approach.
Selfsustained oscillations are some of the most commonly observed phenomena in biological systems. They emanate from non-linear systems in a heterogeneous environment and can be described by the theory of dynamical systems. Part of this theory considers reduced models of the oscillator dynamics by means of amplitudes and a phase variable. Such variables are highly attractive for theoretical and experimental studies. Theoretically these variables correspond to an integrable linearization of the generally non-linear system. Experimentally, there exist well established approaches to extract phases from oscillator signals. Notably, one can define phase models also for networks of oscillators. One highly active field examines effects of non-local coupling among oscillators, which is thought to play a key role in networks with strong coupling. The dissertation introduces and expands the knowledge about high-order phase coupling in networks of oscillators. Mathematical calculations consider the Stuart-Landau oscillator. A novel phase estimation scheme for direct observations of an oscillator dynamics is introduced based on numerics. A numerical study of high-order phase coupling applies a Fourier fit for the Stuart-Landau and for the van-der-Pol oscillator. The numerical approach is finally tested on observation-based phase estimates of the Morris-Lecar neuron. A popular approach for the construction of phases from signals is based on phase demodulation by means of the Hilbert transform. Generally, observations of oscillations contain a small and generic variation of their amplitude. The work presents a way to quantify how much the variations of signal amplitude spoil a phase demodulation procedure. For the ideal case of phase modulated signals, amplitude modulations vanish. However, the Hilbert transform produces artificial variations of the reconstructed amplitude even in this case. The work proposes a novel procedure called Iterative Hilbert Transform Embedding to obtain an optimal demodulation of signals. The text presents numerous examples and tests of application for the method, covering multicomponent signals, observables of highly stable limit cycle oscillations and noisy phase dynamics. The numerical results are supported by a spectral theory of convergence for weak phase modulations.