Nowadays, crisis management logistics is facing new challenges. Indeed, geopolitical conflicts, natural disasters or emergencies can cause big damages and so require rapid and effective response. Due to their sudden occurrence, we are countering a high disturbances context: a Crisis Management Supply Chain (CMSC). Working in such an uncertain environment incites to be equipped with optimization and cooperation mechanisms assuring all the chain actors satisfaction, while acting in a collective way to reach a common objective: the crisis management.In this thesis, we focus on the definition of a modeling approach and an agent-based simulation of the Crisis Management Supply Chain. We propose a decision support system that deals with three problems: the optimal positioning of logistics zones to facilitate the flows circulation, an innovative method for solving a highly-distributed delivery scheduling problem, based on a multi-agent system, for the distribution of relief materials (food, water, clothes, etc.) to the areas affected by the disaster, and finally a need estimating agent to give an accurate forecast of resources’ consumption in the logistic zones. Blending the agent paradigm with the optimization technics helped reach our goals of implementing a large-scale decision support system. The simulation results highlight our contributions
