Asymptotic mean-square stability analysis and simulations of a stochastic model for the human immune response with memory

In this thesis we extended a deterministic model for the Human Immune response, consisting of a non-linear system of differential equations with distributed time delay, which was introduced by Beretta, Kirshner and Marino in 2007, by incorporating stochastic perturbations with multiplicative noise around equilibria of the deterministic model. Our aim is to study the robustness of the equilibria of the deterministic model for the human immune response system with respect to fluctuations due to considering the human body as a noisy enviroment. We do this by analysing the asymptotic mean square stability of the equilibria of our stochastic model. Our work could be divided roughly into two parts. In the first part we analyse the stability of a general non linear system of stochastic differential equations with distributed memory terms by studying the stability properties of the linearisation in the first approximation. First of all we state, using Halanay's inequalities, comparison results useful in the investigation of exponential mean square stability of linear stochastic delay differential systems with distributed memory terms. Then we provide conditions under which asymptotic mean square stability of a nonlinear system of stochastic delay differential equations is implied by the exponential mean square stability of linearised stochastic delay system in the first approximation. In the second part we apply the theoretical results obtained in the first part to investigate the stochastic stability properties of the equilibria of our stochastic model of human immune response. The theoretical results are illustrated by numerical simulations and an uncertainty and sensivity analysis of our stochastic model, suggesting that the deterministic model is robust with respect to the stochastic perturbations.