Heteroclinic cycles and networks
Stability of cycles within networks
Coupled cell systems
Local and global bifurcations
Singularities and strange attractors
Organizing Committee:
Manuela Aguiar (CMUP and FEP.UP)
Isabel Labouriau (CMUP and FCUP)
Alexandre Rodrigues (CMUP and ISEG)
Participants (updated June 2023):
Sofia Castro
Isabel Labouriau
Manuela Aguiar
Alexandre Rodrigues
Santiago Ibáñez
Alexander Lohse
Liliana Garrido-da-Silva
Telmo Peixe
Haibo Ruan
Ana Paula Dias
Soeren von der Gracht
Chris Bick
João Maurício de Carvalho
Ana Margarida Ferreira
Official Program
Mini-courses and tutorials
02/Nov (14h30m; Thursday): Sofia Castro (FEP, CMUP)
Subject: Heteroclinic dynamics
03/Nov (14h30m, Friday): Manuela Aguiar (FEP, CMUP)
Subject: Coupled Cell Networks: basic definitions and results
06/Nov (12h00m, Monday): Liliana Garrido-da-Silva (FCUP, CMUP)
Subject: Stability of cycles and networks
07/Nov (14h30m, Tuesday): Ana Paula Dias (FCUP, CMUP)
Subject: Network constraints on bifurcations of coupled cell systems
09/Nov (14h30m, Thursday): Santiago Ibañez (University of Oviedo)
Subject: Singularities: unfolding and coupling
10/Nov (14h30m, Friday): Alexandre Rodrigues (ISEG, CMUP)
Subject: Chaos in homoclinic bifurcations
Talks
13/Nov (14h30m, Monday): Alexander Lohse (Hamburg University)
14/Nov (14h30m, Tuesday): Chris Bick (Amsterdam University)
15/Nov (16h30m, Wednesday): Soeren von der Gracht (Paderborn University)
16/Nov (14h30m, Thursday): Haibo Ruan (Hamburg University of Technology)
17/Nov (14h30m, Friday): Telmo Peixe (REM, CEMAPRE, ISEG -- Lisbon University)
=========================================================================================================
Room: FC1.006
In ordinary differential equations, a saddle-saddle connection is generically not robust. However, when either symmetry or extinction hyperplanes are present these give rise to flow-invariant spaces, where this connection becomes a saddle-sink connection that is robust. A sequence of connections between consecutive equilibria is called a heteroclinic cycle. A heteroclinic network is a connected union of heteroclinic cycles.
When a cycle is part of a heteroclinic network it cannot be asymptotically stable. It can nevertheless exhibit some stability that may make the cycle visible in experiments and simulations. I shall describe several intermediate notions of stability and ways to determine them.
This mini-course will be more pedagogical than a standard scientific seminar.
=========================================================================================================
Room: FC1.031
=========================================================================================================
Room: FC1.005
=========================================================================================================
Room: FC1.031
=========================================================================================================
Room: FC1.006
Our tool will be the study of singularities (local bifurcations) that occur in coupled systems. We will see how the study of their unfoldings (the atlas of dynamics that arise around the singularity) sometimes allows us to determine properties of synchronisation/desynchronization or the existence of chaotic dynamics.
We will first approach singularities and their unfoldings from a very general perspective and then focus on three particularly interesting cases: Hopf-Zero and Hopf-Hopf singularities, and nilpotent singularities. Moreover, our dynamics will be governed by families of vector fields, i.e., we will refer to singularities of vector fields (points at which the vector fields are zero) and to couplings of systems of differential equations.
=========================================================================================================
Day: 10th of November, 14h30
Speaker: Alexandre Rodrigues (ISEG and CMUP)
Duration: 1h30m
Room: FC1.031
Title: Chaos in homoclinic bifurcations
Abstract:
Proving the existence of a homo- or heteroclinic connection in a given vector field is not an easy task. Santiago Ibanez showed how certain types of non-hyperbolic singularities (nilpotent singularities) generically unfold fields with homo- and heteroclinic connections. In this talk, we describe some global bifurcations (homo and heteroclinic) in autonomous differential equations, and their impact on the geometry of the associated dynamics. Special emphasis will be given to the emergence of chaos.
Results of this talk may be found in:
P. G. Barrientos, S. Ibáñez, A. A. Rodrigues, J. A. Rodriguez, Emergence of Chaotic Dynamics from Singularities, 32th Brazilian Mathematics Colloquium, Instituto Nacional de Matematica Pura e Aplicada (IMPA), Rio de Janeiro, 2019. xi+200 pp. ISBN: 978-85-244-0430-6, 2019
=========================================================================================================
=========================================================================================================
Room: FC1.031
In every heteroclinic network there exists at least one node (equilibrium) with more than one outgoing connection towards another node. Trajectories in a neighbourhood of the network may therefore "choose" which of these connections to follow. For a given network, it is possible to analyze systematically for which sequences of heteroclinic connections there are initial conditions near the network such that the solution follows exactly the prescribed sequence of connections. This helps to understand the long-term dynamics of a system given through an ODE. Studying such questions has led to the notion of switching dynamics and a wide range of dynamic behaviour has been unveiled subsequently, related among other things to the eigenvalues of the linearization of the vector field at the nodes.
In this talk we introduce different levels of switching and give a basic idea of conditions under which they may or may not occur. We start by looking at a simple example and finish with some questions which currently draw considerable attention.
=========================================================================================================
Room: FC1.006
Networks of identical oscillators can exhibit synchrony patterns where synchrony is localized in part of the network: A subset
of oscillators are synchronized while the rest is incoherent. While the focus has predominantly been on stable patterns, we show that for specific (nonpairwise) coupling there can also be heteroclinic cycles and networks that allow dynamical transitions between distinct localized synchrony patterns. This analysis relies on the specific choice of coupling that give rise to continuous symmetries such that the patterns are actually equilibria relative to this symmetry action. Since we cannot expect these symmetries to be present, we analyze how the synchrony patterns deform as the continuous symmetries are broken. Using numerical continuation, we identify how the synchrony patterns bifurcate in the perturbed system.
=========================================================================================================
Room: FC1.006
Many dynamical systems in fields such as neuroscience (the workings of the brain), systems biology (metabolic systems) and robotics (robot swarms) exhibit the structure of a network: they consist of nodes (neurons, proteins, robots) with connections between them. It usually does not suffice to understand the nature of the individual nodes to deduce the behavior of the network, as the specific interaction structure of a network can produce remarkable dynamics. Prominent examples include synchronization (e.g., the simultaneous firing of neurons) and highly complex branching behavior in bifurcations, phenomena that are not found in dynamical systems without the structure of a network.
Network dynamical systems are not well understood mathematically, which makes it hard to quantify and control their behavior. The reason is that most of the established machinery of dynamical systems theory fails to distinguish between networks and general dynamical systems. Thus, we need mathematical tools that are tailor-made for network problems. Several techniques have been proposed recently, and they strikingly have one thing in common: they exploit the algebraic nature of networks.
In this talk, I will give an overview over some recent results regarding the question which dynamical behavior and generic bifurcations are dictated by the network structure of a system. In particular, I will illustrate how structural and algebraic properties culminate in symmetries of the governing equations and how these can be exploited for (partial) answers. This includes classical symmetries but also more exotic concepts such as monoid and quiver representations.
=========================================================================================================
Room: FC1.006
=========================================================================================================
Room: FC1.031
=========================================================================================================
Picture by Juliane Oliveira, Sofia Castro and Isabel Labouriau