Generally, real-time management of sewer systems uses two main tools. The first software one, dedicated to weather forecast is known by their imprecision and uncertainty. The main issue of the second one concerns their computational complexity, which becomes a burden, as they require temporal important steps of decision. This leads to an important problem, where the obtained results are generally different from the really waited ones.
The main idea and objective of the CARDIO project, which are the core of this thesis, is to deal with this important problem using a new paradigm based on “automatic” concepts rather than applied from “hydrology” point of view. Such approach allows the realization of a large amount of numerical simulation in a fast way during a very short time (a few seconds). This allows dispensing of weather forecast by using real-time meditative pluviometric data directly. The objective is to achieve a system where the decision-making is made based on reliable data and also, the error correction is permanent.