Thesis
French
ID: <
10670/1.4ksa3q>
Abstract
From the perspective of best practice assessment of daily mobility, it should refocus the methods and tools for decision aid around the actors in travel: users. In this logic MICROBILIS model was developed to evaluate the adaptation strategies of users relative to their environmental transport. Three streams have been mobilized: the micro-simulation of assignment models, graph theory and multi-agent systems. The environment is modeled from a microscopic simulator of movements and a cellular graph, defining the network capacity. The simulations allow to find the empirical relationships of the dynamics of traffic on the sections and highlight upper capacity constraints at intersections. The transition to the simulation of a large network induces the complexity of the environment and the multiplication of particular cases. It was not possible to make this transition without reducing the initial assumptions, making it unrepresentative of reality.