Localisation des fuites d'eau par le calage de modèle fonde sur l'algorithme génétique.

Abstract

The identification of the quantity of water lost in the water networks and the leak detection has a major importance of the public water services, as well for motives of exploitation and for their planning. In this thesis, an approach based on the optimization presented to quantify and locate the leaks in the water network in the Mostaganem city water supply network, through the process of hydraulic calibration model. The model calibration is formulated as a nonlinear optimization problem solved by the use of the genetic algorithm. The accuracy of leakage detection depends on the quality of the network observed data and the type of approach used for calibration.