We propose methods and tools to help strategic network designing dedicated to public transportation in urban environment. This class of poblems can fit into a larger framework : mobility network design problems. First, we survey network design problems. Then, we formalize the concept of mobility network design problems with elastic demand varying with the quality of service. We present various modeling and extensions that we studied on small instances with exact resolution. We solve bigger instances with GRASP and Tabu Metaheuristics. We use a partcular route description named "geodesic". We also use an approximation of the objective function to significantly accelerate its evaluation. The last method involves Benders'sheme. In the appendices, we propose design paterns based on generic programming that aim to design both flexible and efficient software components. We also present a coupling method between optimization and simulation named "model enhancement"
