A Dynamical Systems Analysis of Movement Coordination Models

Abstract

In this thesis, we present a dynamical systems analysis of models of movement coordination, namely the Haken-Kelso-Bunz (HKB) model and the Jirsa-Kelso excitator (JKE). The dynamical properties of the models that can describe various phenomena in discrete and rhythmic movements have been explored in the models' parameter space. The dynamics of amplitude-phase approximation of the single HKB oscillator has been investigated. Furthermore, an approximated version of the scaled JKE system has been proposed and analysed. The canard phenomena in the JKE system has been analysed. A combination of slow-fast analysis, projection onto the Poincare sphere and blow-up method has been suggested to explain the dynamical mechanisms organising the canard cycles in JKE system, which have been shown to have different properties comparing to the classical canards known for the equivalent FitzHugh-Nagumo (FHN) model. Different approaches to de fining the maximal canard periodic solution have been presented and compared. The model of two HKB oscillators coupled by a neurologically motivated function, involving the effect of time-delay and weighted self- and mutual-feedback, has been analysed. The periodic regimes of the model have been shown to capture well the frequency-induced drop of oscillation amplitude and loss of anti-phase stability that have been experimentally observed in many rhythmic movements and by which the development of the HKB model has been inspired. The model has also been demonstrated to support a dynamic regime of stationary bistability with the absence of periodic regimes that can be used to describe discrete movement behaviours.