Reaction-diffusion systems and delay differential equations have been extensively used over the years to model and study the dynamics of infectious diseases. In this thesis we consider two aspects of disease dynamics: spatial dynamics in a reaction-diffusion epidemic model with nonlinear incidence rate, and a delayed epidemic model with combined effects of latency and temporary immunity. The first part of the thesis is devoted to the analysis of stability and pattern formation in an SIS-type epidemic model with nonlinear incidence rate. By considering the dynamics without spatial component, conditions for local asymptotic stability are obtained for general values of the powers of nonlinearity. We prove positivity, boundedness, invariant principle and permanence of our model. The next generation matrix method is used to derive the corresponding basic reproductive number R0, and the Routh-Hurwitz criterion is used to show that for R0 ≤ 1, the disease-free equilibrium is found to be locally asymptotically stable, for R0 > 1, a unique endemic steady state exists and is found to be locally asymptotically stable. In the presence of diffusion, Turing instability conditions are established in terms of system parameters. Numerical simulations are performed to identify the spatial regions for spots, stripes and labyrinthine patterns in the parameter space. Numerical simulations show that the system has complex and rich dynamics and can exhibit complex patterns, depending on the recovery rate r and the transmission rate β. We have discovered that whenever the transmission rate exceeds the recovery rate the system exhibits stripe patterns which correspond to a disease outbreak, and in the opposite case the system settles on spot patterns which imply the absence of disease outbreaks. Also, we find that increasing the power q can lead to epidemic outbreak even at lower values of the transmission rate β. All numerical simulations use an Implicit-Explicit (IMEX) Euler's method, which computes diffusion terms in Fourier space and reaction terms in the real space. Numerical approximation of the model is benchmarked to prove stability of the numerical scheme, and the method is shown to converge with the correct order. Experimental order of convergence (EOC) and estimates for the error in both L2, H1 and maximum norms have also been computed. Also, we compare our results to those on infectious diseases and our model shows good predictions. In the second part of this thesis, we derive and analyse a delayed SIR model with bilinear incidence rate and two time delays which represent latency Τ1 and temporary immunity Τ2 periods. We prove both local and global stability of the system equilibria in the case when there are no time delays, i.e. both the latency and temporary immunity periods are set to zero. For the case when there is only latency (Τ1 > 1, Τ2 = 0) and the case when the two time delays are identical (Τ1 = Τ2 = Τ ), we show that the endemic steady state is always stable for any parameter values. For the case when there is only temporary immunity (Τ2 > 0, Τ1 = 0) and the case when there are both latency and temporary immunity in the system (Τ1 > 0, Τ2 > 0), we prove the existence of periodic solutions arising from the Hopf bifurcation. The endemic steady state undergoes Hopf bifurcation giving rise to stable periodic solutions. For the last two cases, we show interesting regions of (in)stability of the endemic steady state in the different parameter regimes. We find that by varying the transmission rate β, the natural death rate γ and the disease-induced death rate μ increase the regions of (in)stability. Also, we find that the dynamics of the system is richer when we have the two time delays in the model. Analytical results are supported by extensive numerical simulations, illustrating temporal behaviour of the system in different dynamical regimes. Finally, we relate our results to modelling infectious diseases and our results show good predictions of safety and epidemic outbreak.

This thesis is concerned with the study of the important effect of the gradient term in parabolic problems. More precisely, we study the global existence or nonexistence of solutions, and their asymptotic behaviour in finite or infinite time. Particularly when the power of the gradient term can increase to the power function of the solution. This thesis consists of five parts. (i) Steady-State Solutions, (ii) The Blow-up Behaviour of the Positive Solutions, (iii) Parabolic Liouville-Type Theorems and the Universal Estimates, (iv) The Global Existence of the Positive Solutions, (v) Viscous Hamilton-Jacobi Equations (VHJ). Under certain conditions on the exponents of both the function of the solution and the gradient term, the nonexistence of positive stationary solution of parabolic problems with gradient terms are proved in (i). In (ii), we extend some known blow-up results of parabolic problems with perturbation terms, which is not too strong, to problems with stronger perturbation terms. In (iii), the nonexistence of nonnegative, nontrivial bounded solutions for all negative and positive times on the whole space are showed for parabolic problems with a strong perturbation term. Moreover, we study the connections between parabolic Liouville-type theorems and local and global properties of nonnegative classical solutions to parabolic problems with gradient terms. Namely, we use a general method for derivation of universal, pointwise a priori estimates of solutions from Liouville type theorems, which unifies the results of a priori bounds, decay estimates and initial and final blow up rates. Global existence and stability, and unbounded global solutions are shown in (iv) when the perturbation term is stronger. In (v) we show that the speed of divergence of gradient blow up (GBU) of solutions of Dirichlet problem for VHJ, especially the upper GBU rate estimate in n space dimensions is the same as in one space dimension.

This thesis is concerned with modelling the spread of diseases amongst host populations and the epidemics that result from this process. We are primarily interested in how networks can be used to model the various heterogeneities observable in real-world populations. Firstly, we start with the full system of Kolmogorov/master equations for a simple Susceptible-Infected-Susceptible (SIS) type epidemic on an arbitrary contact network. From this general framework, we rigorously derive sets of ODEs that describe the exact dynamics of the expected number of individuals and pairs of individuals. We proceed to use moment closure techniques to close these hierarchical systems of ODEs, by approximating higher order moments in terms of lower order moments. We prove that the simple first order mean-field approximation becomes exact in the limit of a large, fully-connected network. We then investigate how well two different pairwise approximations capture the topological features of theoretical networks generated using different algorithms. We then introduce the effective degree modelling framework and propose a model for SIS epidemics on dynamic contact networks by accounting for random link activation and deletion. We show that results from the resulting set of ODEs agrees well with results from stochastic simulations, both in describing the evolution of the network and the disease. Furthermore, we derive an analytic calculation of the stability of the disease-free steady state and explore the validity of such a measure in the context of a dynamically evolving contact network. Finally, we move on to derive a system of ODEs that describes the interacting dynamics of a disease and information relating to the disease. We allow individuals to become responsive in light of received information and, thus, reduce the rate at which they become infected. We consider the effectiveness of different routes of information transmission (such as peer-to-peer communication or mass media campaigns) in slowing or preventing the spread of a disease. Finally, we use a range of modelling techniques to investigate the spread of disease within sheep flocks. We use field data to construct weighted contact networks for flocks of sheep to account for seasonal changes of the flock structure as lambs are born and eventually become weaned. We construct a range of network and ODE models that are designed to investigate the effect of link-weight heterogeneity on the spread of disease.

This thesis discusses a two dimensional non-smooth dynamical system described by an autonomous ordinary differential equation. The right hand side of the differential equation is assumed to be discontinuous. We provide a local theory of existence, uniqueness and exponential asymptotic stability and state a formula for the basin of attraction. Our conditions are sufficient. Thetheory generalizes smooth dynamical systems theory by providing contraction conditions for two nearby trajectories at a jump. Such conditions have only previously been studied for a two dimensional nonautonomous differential equation. We provide an example of the theory developed in this thesis and show that we can determine stability of a periodic orbit without explicitly calculating it. This is the main advantage of our theory. Our conditions require to define a metric. This however, can turn out to be a difficult task, and at present, we do not have a method for finding such a metric systematically. The final part of this thesis considers an application of a nonsmooth dynamical system to biomechanics. We model an elderly person stepping over an obstacle. Our model assumes stiff legs, and suggests a gait strategy to overcome an obstacle. This work is in collaboration with Professor Wagner's research group at Institute for Sport Science at the University of Mϋnster. However, we only present work developed independently in this thesis.

This work explores mechanisms for pattern formation through coupled bulksurface partial differential equations of reaction-diffusion type. Reaction-diffusion systems posed both in the bulk and on the surface on stationary volumes are coupled through linear Robin-type boundary conditions. In this framework we study three different systems as follows (i) non-linear reactions in the bulk and surface respectively, (ii) non-linear reactions in the bulk and linear reactions on the surface and (iii) linear reactions in the bulk and non-linear reactions on the surface. In all cases, the systems are non-dimensionalised and rigorous linear stability analysis is carried out to determine the necessary and sufficient conditions for pattern formation. Appropriate parameter spaces are generated from which model parameters are selected. To exhibit pattern formation, a coupled bulk-surface finite element method is developed and implemented. We implement the numerical algorithm by using an open source software package known as deal.II and show computational results on spherical and cuboid domains. Theoretical predictions of the linear stability analysis are verified and supported by numerical simulations. The results show that non-linear reactions in the bulk and surface generate patterns everywhere, while non-linear reactions in the bulk and linear reactions on the surface generate patterns in the bulk and on the surface with a pattern-less thin boundary layer. However, linear reactions in the bulk do not generate patterns on the surface even when the surface reactions are non-linear. The generality, robustness and applicability of our theoretical computational framework for coupled system of bulk-surface reaction-diffusion equations set premises to study experimentally driven models where coupling of bulk and surface chemical species is prevalent. Examples of such applications include cell motility, pattern formation in developmental biology, material science and cancer biology.

This thesis is concerned with the study of the Blow-up phenomena for parabolic problems, which can be defined in a basic way as the inability to continue the solutions up to or after a finite time, the so called blow-up time. Namely, we consider the blow-up location in space and its rate estimates, for special cases of the following types of problems: (i) Dirichlet problems for semilinear equations, (ii) Neumann problems for heat equations, (iii) Neumann problems for semilinear equations, (iv) Dirichlet (Cauchy) problems for semilinear equations with gradient terms. For problems of type (i), (ii), we extend some known blow-up results of parabolic problems with power and exponential type nonlinearities to problems with nonlinear terms, which grow faster than these types of functions for large values of solutions. Moreover, under certain conditions, some blow-up results of the single semilinear heat equation are extended to the coupled systems of two semilinear heat equations. For problems of type (iii), we study how the reaction terms and the nonlinear boundary terms affect the blow-up properties of the blow-up solutions of these problems. The noninuence of the gradient terms on the blow-up bounds is showed for problems of type (iv).

No description available.

Complex networks are ubiquitous in nature. Numerous neurological diseases, such as Alzheimer's, Parkinson's, epilepsy are caused by the abnormal collective behaviour of neurons in the brain. In particular, there is a strong evidence that Parkinson's disease is caused by the synchronisation of neurons, and understanding how and why such synchronisation occurs will bring scientists closer to the design and implementation of appropriate control to support desynchronisation required for the normal functioning of the brain. In order to study the emergence of (de)synchronisation, it is necessary first to understand how the dynamical behaviour of the system under consideration depends on the changes in systems parameters. This can be done using a powerful mathematical method, called bifurcation analysis, which allows one to identify and classify different dynamical regimes, such as, for example, stable/unstable steady states, Hopf and fold bifurcations, and find periodic solutions by varying parameters of the nonlinear system. In real-world systems, interactions between elements do not happen instantaneously due to a finite time of signal propagation, reaction times of individual elements, etc. Moreover, time delays are normally non-constant and may vary with time. This means that it is vital to introduce time delays in any realistic model of neural networks. In this thesis, I consider four different models. First, in order to analyse the fundamental properties of neural networks with time-delayed connections, I consider a system of four coupled nonlinear delay differential equations. This model represents a neural network, where one subsystem receives a delayed input from another subsystem. The exciting feature of this model is the combination of both discrete and distributed time delays, where distributed time delays represent the neural feedback between the two sub-systems, and the discrete delays describe neural interactions within each of the two subsystems. Stability properties are investigated for different commonly used distribution kernels, and the results are compared to the corresponding stability results for networks with no distributed delays. It is shown how approximations to the boundary of stability region of an equilibrium point can be obtained analytically for the cases of delta, uniform, and gamma delay distributions. Numerical techniques are used to investigate stability properties of the fully nonlinear system and confirm our analytical findings. In the second part of this thesis, I consider a globally coupled network composed of active (oscillatory) and inactive (non-oscillatory) oscillators with distributed time delayed coupling. Analytical conditions for the amplitude death, where the oscillations are quenched, are obtained in terms of the coupling strength, the ratio of inactive oscillators, the width of the uniformly distributed delay and the mean time delay for gamma distribution. The results show that for uniform distribution, by increasing both the width of the delay distribution and the ratio of inactive oscillators, the amplitude death region increases in the mean time delay and the coupling strength parameter space. In the case of the gamma distribution kernel, we find the amplitude death region in the space of the ratio of inactive oscillators, the mean time delay for gamma distribution, and the coupling strength for both weak and strong gamma distribution kernels. Furthermore, I analyse a model of the subthalamic nucleus (STN)-globus palidus (GP) network with three different transmission delays. A time-shift transformation reduces the model to a system with two time delays, for which the existence of a unique steady state is established. Conditions for stability of the steady state are derived in terms of system parameters and the time delays. Numerical stability analysis is performed using traceDDE and DDE-BIFTOOL in Matlab to investigate different dynamical regimes in the STN-GP model, and to obtain critical stability boundaries separating stable (healthy) and oscillatory (Parkinsonian-like) neural ring. Direct numerical simulations of the fully nonlinear system are performed to confirm analytical findings, and to illustrate different dynamical behaviours of the system. Finally, I consider a ring of n neurons coupled through the discrete and distributed time delays. I show that the amplitude death occurs in the symmetric (asymmetric) region depending on the even (odd) number of neurons in the ring neural system. Analytical conditions for linear stability of the trivial steady state are represented in a parameter space of the synaptic weight of the self-feedback and the coupling strength between the connected neurons, as well as in the space of the delayed self-feedback and the coupling strength between the neurons. It is shown that both Hopf and steady-state bifurcations may occur when the steady state loses its stability. Stability properties are also investigated for different commonly used distribution kernels, such as delta function and weak gamma distributions. Moreover, the obtained analytical results are confirmed by the numerical simulations of the fully nonlinear system.

Let Ω ⊂ Rn be a bounded Lipschitz domain and consider the energy functional F[u, Ω] := ∫ Ω F(∇u(x)) dx, over the space Ap(Ω) := {u ∈ W 1,p(Ω, Rn): u|∂Ω = x, det ∇u> 0 a.e. in Ω}, where the integrand F : Mn×n → R is quasiconvex, sufficiently regular and satisfies a p-coercivity and p-growth for some exponent p ∈ [1, ∞[. A motivation for the study of above energy functional comes from nonlinear elasticity where F represents the elastic energy of a homogeneous hyperelastic material and Ap(Ω) represents the space of orientation preserving deformations of Ω fixing the boundary pointwise. The aim of this thesis is to discuss the question of multiplicity versus uniqueness for extremals and strong local minimizers of F and the relation it bares to the domain topology. Our work, building upon previous works of others, explicitly and quantitatively confirms the significant role of domain topology, and provides explicit and new examples as well as methods for constructing such maps. Our approach for constructing strong local minimizers is topological in nature and is based on defining suitable homotopy classes in Ap(Ω) (for p ≥ n), whereby minimizing F on each class results in, modulo technicalities, a strong local minimizer. Here we work on a prototypical example of a topologically non-trivial domain, namely, a generalised annulus, Ω= {x ∈ Rn : a< |x| <b}, with 0 <a<b< ∞. Then the associated homotopy classes of Ap(Ω) are infinitely many when n =2 and two when n ≥ 3. In contrast, for constructing explicitly and directly solutions to the system of Euler-Lagrange equations associated to F we introduce a topological class of maps referred to as generalised twists and relate the problem to extremising an associated energy on the compact Lie group SO(n). The main result is a surprising discrepancy between even and odd dimensions. In even dimensions the latter system of equations admits infinitely many smooth solutions, modulo isometries, amongst such maps whereas in odd dimensions this number reduces to one. Even more surprising is the fact that in odd dimensions the functional F admits strong local minimizers yet no solution of the Euler-Lagrange equations can be in the form of a generalised twist. Thus the strong local minimizers here do not have the symmetry one intuitively expects!.

This thesis is concerned with the derivation, numerical analysis and implementation of high-order compact finite difference schemes for parabolic partial differential equations in multiple spatial dimensions. All those partial differential equations contain mixed derivative terms. The resulting schemes have been applied to equations appearing in computational finance. First, we develop and study essentially high-order compact finite difference schemes in a general setting with option pricing in stochastic volatility models on non-uniform grids as application. The schemes are fourth-order accurate in space and second-order accurate in time for vanishing correlation. In the numerical study we obtain high-order numerical convergence also for non-zero correlation and non-smooth payoffs which are typical in option pricing. In all numerical experiment a comparative standard second-order discretisation is significantly outperform. We conduct a numerical stability study which indicates unconditional stability of the scheme. Second, we derive and analyse high-order compact schemes with n-dimensional spatial domain in a general setting. We obtain fourth-order accuracy in space and second-order accuracy in time. A thorough von Newmann stability analysis is performed for spatial domain with dimensions two and three. We prove that a necessary stability condition holds unconditionally without additional restrictions on the choice of the discretisation parameters for vanishing mixed derivative terms. We also give partial results for non-vanishing mixed derivative terms. As first example Black-Scholes Basket options are considered. In all numerical experiments, where the initial conditions were smoothened using the smoothing operators developed by Kreiss, Thomée and Widlund, a comparative standard second-order discretisation is significantly outperformed. As second example the multi-dimentional Heston basket option is considered for n independent Heston processes, where for each Heston process there is a non-vanishing correlation between the stock and its volatility.