Optimisation et simulation d'une chaîne logistique : application au secteur de l'agriculture. (Supply chain optimization and simulation : an application to agricultural sector)

To overcome the new challenges facing agricultural sector, imposed by globalisation, changing market demands and price instability, the crop production supply chain must particularly be very reactive, flexible, with a high yield and at low cost. Its improving and eventual re-configuration can lead to an upgrade in efficiency, responsiveness, business integration and make it able to confront the market competitiveness. The thesis is thus placed in this particular context and aims to support decision making in crop harvesting activity, which is considered the pivotal stage in the cereal production circuit owing to its high cost and impact on the returns earned. Managing the harvest activity involves gathering, transportation and storage operations, performed by a collection of agricultural holdings geographically dispersed. Besides its practical relevance, this thesis forms part of the Operational Research (OR) and more specifically, refers to the linear and stochastic programming, discrete event simulation, and their coupling. In addition, the synergy created between OR, inferential and predictive statistics, geographical information system tools makes the decision support system competitive, efficient and responsive