Modelling is a powerful method for understanding complex systems, which works by simplifying them to their most essential components. The choice of the components is driven by the aspects studied. The tool chosen to perform this task will determine what can be modelled, the maximum number of components which can be represented, as well as the analyses which can be performed on the system. Performance Evaluation Process Algebra (PEPA) was initially developed to tackle computer systems issues. Nevertheless, it possesses some interesting properties which could be exploited for the study of epidemiological systems. PEPA's main advantage resides in its capacity to change scale: the assumptions and parameter values describe the behaviour of a single individual, while the resulting model provides information on the population behaviour. Additionally, stochasticity and continuous time have already proven to be useful features in epidemiology. While each of these features is already available in other tools, to find all three combined in a single tool is novel, and PEPA is proposed as a useful addition to the epidemiologist's toolbox. Moreover, an algorithm has been developed which allows converting a PEPA model into a system of Ordinary Differential Equations (ODEs). This provides access to countless additional software and theoretical analysis methods which enable the epidemiologist to gain further insight into the model. Finally, most existing tools require a deep understanding of the logic they are based on and the resulting model can be difficult to read and modify. PEPA's grammar, on the other hand, is easy to understand since it is based on few, yet powerful concepts. This makes it a very accessible formalism for any epidemiologist. The objective of this thesis is to determine precisely PEPA's ability to describe epidemiological systems, as well as extend the formalism when required. This involved modelling two systems: the bubonic plague in prairie dogs, and measles in England and Wales. These models were chosen as they exhibit a good range of typical features, allowing to thoroughly test PEPA. All features required in each of these models have been analysed in detail, and a solution has been provided for representing each of these features. While some of them could be expressed in a straightforward manner, PEPA did not provide the tools to express others. In those cases, we determined methods to approach the desired behaviour, and the limitations of said methods were carefully analysed. In the case of models with a structured population, PEPA was extended to simplify their expression and facilitate the writing process of the PEPA model. The work also required the development of an algorithm to derive ODEs adapted to the type of models encountered. Finally, the PEPAdum software was developed to assist the modeller in the generation and analysis of PEPA models, by simplifying the process of writing a PEPA model with compartments, performing the average of stochastic simulations and deriving and explicitly providing the ODEs using the Stirling Amendment.

The main goal of my studies has been to search for the optimal control strategy of controlling epidemics when taking into account both economical and social costs of the disease. Three control scenarios emerge with treating the whole population (global strategy, GS), treating a small number of individuals in a well-defined neighbourhood of a detected case (local strategy, LS) and allowing the disease to spread unchecked (null strategy, NS). The choice of the optimal strategy is governed mainly by a relative cost of palliative and preventive treatments. Although the properties of the pathogen might not be known in advance for emerging diseases, the prediction of the optimal strategy can be made based on economic analysis only. The details of the local strategy and in particular the size of the optimal treatment neighbourhood weakly depends on disease infectivity but strongly depends on other epidemiological factors (rate of occurring the symptoms, spontaneously recovery). The required extent of prevention is proportional to the size of the infection neighbourhood, but this relationship depends on time till detection and time till treatment in a non-nonlinear (power) law. The spontaneous recovery also affects the choice of the control strategy. I have extended my results to two contrasting and yet complementary models, in which individuals that have been through the disease can either be treated or not. Whether the removed individuals (i.e., those who have been through the disease but then spontaneously recover or die) are part of the treatment plan depends on the type of the disease agent. The key factor in choosing the right model is whether it is possible - and desirable - to distinguish such individuals from those who are susceptible. If the removed class is identified with dead individuals, the distinction is very clear. However, if the removal means recovery and immunity, it might not be possible to identify those who are immune. The models are similar in their epidemiological part, but differ in how the removed/recovered individuals are treated. The differences in models affect choice of the strategy only for very cheap treatment and slow spreading disease. However for the combinations of parameters that are important from the epidemiological perspective (high infectiousness and expensive treatment) the models give similar results. Moreover, even where the choice of the strategy is different, the total cost spent on controlling the epidemic is very similar for both models. Although regular and small-world networks capture some aspects of the structure of real networks of contacts between people, animals or plants, they do not include the effect of clustering noted in many real-life applications. The use of random clustered networks in epidemiological modelling takes an impor- tant step towards application of the modelling framework to realistic systems. Network topology and in particular clustering also affects the applicability of the control strategy.

Ce mémoire de thèse s'articule en deux parties. La première partie s'intéresse à la robustesse du nombre de reproduction de base R0 et du nombre de reproduction type (T), qui sont des seuils pour des systèmes épidémiques. Nous montrons que ces paramètres seuils ne sont pas des jauges fiables pour évaluer la distance qui sépare la Jacobiene (J) du système, calculé au point d'équilibre sans maladie à l'ensemble des matrices stables (S) si J est instable, (respectivement où l'ensemble des matrices instables (U) si J est stable). La deuxième partie se penche sur l'étude d'un modèle déterministe (S V E I R), où S représente les susceptibles, (V) les vaccinés,( E) les latents, (I )les infectieux et( R ) les immuns. Dans le dit modèle, les vaccinés sont considérés comme des « susceptibles dans une moindre mesure » du fait que le vaccin ne garantit pas une immunité totale. Le nombre de reproduction de base Rvac qui assure l'existence et l'unicité de l?équilibre endémique est déterminé. La globale stabilité de l'équilibre endémique est établie en utilisant les techniques de Lyapunov quand Rvac > 1. Ce résultat améliore un résultat de Gumel / This memory is divided in two parts. The first part talk about robustness of basic reproduction number R0 and basic reproduction number type S (T) both of them are thresholds for epidemic systems. We show that those thresholds are not good indicators to evaluate the distance between the jacobian matrix J(DFE) of system at the disease free equilibrium (DFE) and the set of stable ( S_t) or unstable (U ) matrix. The second part talk about an deterministic model ( S V E I R) ; where (S) represent the susceptibles ; (V) the vaccined ; (E) the latents,(I) the infectious, and (R) the removed. In this model, the vaccined are considered like susceptibles, because the vaccine don't confers an perfect immunity. The Basic reproduction number, (R_vac), who ensures the existence and unicity of endemic equilibrium is determined. The global stability of endemic equilibrium point is established using Lyapunov technics when (R_vac) is greater than one (R_vac)> 1

L'objectif de cette thèse est de modéliser la transmission du paludisme dans la région montagneuse de l'ouest du Kenya, en se servant des outils de systèmes dynamiques. Nous considérons deux modèles mathématiques. Le premier prend en compte une susceptibilité et une infectivité différentielle dans les métapopulations, et le second un taux de saturation des repas sanguins dans la population des moustiques. Dans le premier modèle, nous considérons plusieurs écosystèmes identifiés comme zones sensibles dans la région montagneuse de l'ouest du Kenya. Dans ce modèle, ces zones sensibles sont considérées comme nos différents patchs. Les populations de chaque patch sont divisées en deux : les enfants et les adultes. Le modèle nous permet d'évaluer le rôle de l'hétérogénéité de l'écosystème et la persistance de l'épidémie dans la région, due à la structuration d'âge. Nous prenons en compte la susceptibilité et l'infectivité différentielle afin d'étendre le modèle d'un patch en un modèle à plusieurs patchs. Après avoir subdivisé la région en n zones sensibles, nous faisons une analyse mathématique du modèle obtenu. Pour effectuer cette analyse, nous utilisons la théorie des systèmes triangulaires, des systèmes dynamiques monotones, des systèmes dynamiques non linéaires anti-monotones et le principe d'invariance de LaSalle. Un des éléments très utilisés dans notre analyse qui est un concept clé en épidémiologie, est le taux de reproduction de base, très souvent noté Ro. Cette quantité, sans dimension, est le nombre moyen de cas secondaires, engendré par un individu infectieux typique durant sa période d'infectiosité, quand il est introduit dans une population constituée entièrement de susceptibles. L'existence et la stabilité du point d'équilibre sans maladie (DFE) sont établies et nous prouvons que le DFE est globalement asymptotiquement stable lorsque Ro<1. Lorsque Ro>1, le modèle admet un point d'équilibre endémique qui est globalement asymptotiquement stable. L'analyse de notre modèle montre que la structuration d'âge réduit l'ampleur de l'infection. En utilisant les données relevées, nous faisons quelques simulations numériques afin de montrer l'impact de la métapopulation et de la structuration d'âge sur le taux de reproduction de base. Dans la seconde partie, nous formulons un modèle de paludisme avec saturation du taux d'alimentation des moustiques qui nous conduit à une incidence non linéaire. Nous démontrons que DFE est globalement asymptotiquement stable si Ro<1. Lorsque Ro>1, il existe un unique point d'équilibre endémique qui est globalement asymptotiquement stable. Des simulations numériques sont faites afin d'illustrer l'impact de la saturation d'alimentation sur le taux de reproduction de base / The objective of this thesis is to model highland malaria in western Kenya using dynamical systems. Two mathematical models are formulated ; one, on differentiated susceptibility and differentiated infectivity in a metapopulation setting with age structure, the other, a saturated vector feeding rate model with disease induced deaths and varying host and vector populations. In the first model, we consider the different ecosystems identified as malaria hotspots in the western Kenya highlands and consider the ecosystems as different patches. The population in each patch is classified as, either child or, adult. The model will aid in examining the role of ecosystem heterogeneity and age structure to the persistent malaria epidemics in the highlands. We formulate the differentiated susceptibility and infectivity model that extend to multiple patches the well known epidemiological models in one patch. Classifying the hot spots as n patches, we give its mathematical analysis using the theory of triangular system, monotone non-linear dynamical systems, and Lyapunov-Lasalle invariance principle techniques. Key to our analysis is the definition of a reproductive number, Ro, the number of new infections caused by one individual in an otherwise fully susceptible population throughout the duration of the infectious period. The existence and stability of disease-free and endemic equilibrium is established. We prove that the disease free state of the systems is globally asymptotically stable when the basic reproduction number Ro<1, and when Ro>1 an endemic equilibrium is established which is locally and globally asymptotically stable. The model shows that the age structuring reduces the magnitude of infection. Using relevant data we did some simulation, to demonstrate the role played by metapopulation and age structuring on the incidence and Ro. In the second part we formulate a model for malaria with saturation on the vector feeding rates that lead to a nonlinear function in the infection term. The vector feeding rate is assumed, as in the predator prey models, to rise linearly as a function of the host-vector ratio until it reaches a threshold Qv, after which the vector feeds freely at its desired rate. The two populations are variable and drive malaria transmission, such that when the vectors are fewer than hosts, the rate of feeding is determined by the vectors feeding desire, whereas, when the hosts are more than the vectors, the feeding rate is limited by host availability and other feeding sources may have to be sought by the vector. Malaria induced deaths are introduced in the host population, while the vector is assumed to survive with the parasite till its death. We prove that the Disease Free Equilibrium is locally and globally asymptotically stable if Ro<1 and when Ro>1, an endemic equilibrium emerges, which is unique, locally and globally asymptotically stable. The role of the saturated mosquito feeding rate is explored with simulation showing the crucial role it plays especially on the basic reproduction number

L'objectif de cette thèse est d'une part l'étude de la stabilité des équilibres de certains modèles épidémiologiques et d'autre part la construction d'un observateur pour l'estimation des états non mesurés et d'un paramètre clé pour un modèle intra-hôte. Nous proposons des extensions des modèles du type SIR, SIRS et SIS et nous étudions la stabilité globales de leur équilibres. En présence de plusieurs souches de pathogène d'un modèle SIS, on montre que le principe de compétition exclusive est vérifié: la souche qui maximise un seuil remporte la compétition en éliminant les autres souches. Il se trouve aussi que la souche gagnante est celle qui donne à l'équilibre le minimum de population hôte susceptible. Ceci peut être interprété comme étant un principe de pessimisation. En considérant ce modèle avec cette fois une loi de contact de type fréquence-dépendante, on montre que la dynamique change et qu'un équilibre de coexistence existe et qui est globalement asymptotiquement stable sous certaines conditions. Le comportement asymptotique des deux équilibres frontières est aussi prouvé. L'étude de la stabilité des états d'équilibres est essentiellement faite par la construction des fonctions de Lyapunov combiné avec le principe d'invariance de LaSalle. On considère un modèle intra-hôte structuré en classe d'âge du parasite Plasmodium falciparum avec une force d'infection général. Nous développons une méthode d'estimation de la charge parasitaire totale dont on ne sait mesurée par les méthodes actuellement connues. Pour cela nous utilisons les outils de la théorie du contrôle, plus particulièrement les observateurs à entrées inconnues, pour estimer les états non mesurés à partir des états mesurés (données). De cela nous déduisons une méthode d'estimation d'un paramètre inconnu qui représente le taux d'infection des globules rouges saines par les parasites / The purpose of this thesis is on the one hand to study stability of equilibria of some epidemic models and secondly to construct an observer to estimate the non-measured states and a key parameter in a within host model. We propose extensions of classical models SIR, SIRS and SIS and we study the global stability of their equilibria. In presence of multiple pathogen strains, we proved that competitive exclusion principle holds: the strain having the largest threshold wins the competition by eliminating the others. It turns out that the winning strain is the one for which the equilibrium gives the minimum of the susceptible host population. This can be interpreted as pessimization principle. By considering the same model with two strains and a frequency-dependent type of the contact law, we prove that dynamics changes and a coexistence equilibrium exists and it is globally asymptotically stable under some conditions. The asymptotic behavior of the two other boundary equilibria is also established. The stability study of equilibrium states is mainly done by construction Lyapunov functions combined with LaSalle's invariance principle. We consider an age-structured within-host model of the Plasmodium falciuparum parasite with a general infection force. We develop a method to estimate the total parasite burden that cannot be measured by the current methods. To this end, we use some tools from control theory, more precisely observers with unknown inputs, to estimate the non measured states from the measured ones (data). From this, we deduce a method to estimate an unknown parameter that represents infection rate of healthy reed blood cells by the parasites