Agent-based simulation for the study of complex dynamic systems. 

Agent-based simulation is increasingly used to study dynamic complex systems. The principle of this approach is to explicitly represent the entities making up the system under study in the form of autonomous, interacting computer entities - the agents. It offers a natural, easy-to-understand modeling framework, enabling the development of truly experimentable virtual worlds. Recent years have seen the development of increasingly detailed agent-based models, and by the integration of new ways of interacting with them. These developments have been made possible by new modeling software platforms offering tools that are easy to access, even for non-computer scientists, but increasingly powerful. This presentation will focus on my work in this field, with 3 main research axes: the integration of the spatial dimension in simulations, the modeling of human behavior and, finally, participative simulation. I will give concrete examples of projects carried out in these 3 areas.