In the global effort to achieve sustainable development, new adaptive approaches of urban planning call for a reliable understanding of Urban Morphodynamics: the evolution of the spatial structure of an urban system from elementary interactions between its components. In this thesis, theoretical and methodological contributions to Urban Morphodynamics are proposed from the integration of the two main strands of literature in quantitative urban geography: Location Theory, which studies the economic determinants of the location of economic activities, and Complex Systems, which describe how cities self-organize from the dynamic interactions of many individuals. After an historical review in the first part, its second part addresses two problems of Urban Morphodynamics, gathering concepts and methods from both Location Theory and Complex Systems. First, using mathematical modelling, it exposes the scaling properties of the equilibrium monocentric city model. Second, coupling analytical developments with agent-based simulations, it explores the dynamic growth of a non-monocentric city model. In a third part, it proposes a hybrid modelling framework that combines the behavioural foundations of Discrete Choice Theory with the nonlinear dynamic models of Synergetics. This framework is then applied to a core-periphery model of New Economic Geography in order to forecast regional development trajectories. In a fourth and last part, in light of the proposed models, this thesis argues that the theoretical foundations of Location Theory and Complex Systems can be reconciled in an operational urban morphodynamic theory.