Air Traffic Control Approach using Distributed CSPs

The Constraints Satisfaction Problems (CSPs) have a great role in the Artificial Intelligence domain. They allow to represent in a simple form a great number of real problems. This is the framework of our contribution. It consists to propose a distributed CSP representation of air traffic conflicts and to propose a methodology of resolution to this problem. In this perspective, we first propose an amelioration of a descriptionresolution expertise within the framework of formalizing the air traffic controllers way of thinking by a system of production rules (SPR). Then, we deduce distributed constraints from the inference engine results that we have adapted to the SPR. This leads us to the definition of a distributed CSP related to air traffic conflicts problem, to which we propose a resolution method.